Environmental and Human Health Impacts of Nanotechnology

Environmental and Human Health Impacts of Nanotechnology

Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

Environmental and Human Health Impacts of Nanotechnology Edited by JAMIE R. LEAD School of Geography, Earth and Environmental Sciences, University of Birmingham, UK EMMA SMITH Department of Biological and Chemical Sciences, The University of the West Indies, Barbados

A John Wiley and Sond, Ltd., Publication

This edition first published 2009 © 2009 Blackwell Publishing Ltd Registered office John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom. For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com. The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. The publisher and the author make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for every situation. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. The fact that an organization or Website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or Website may provide or recommendations it may make. Further, readers should be aware that Internet Websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the author shall be liable for any damages arising herefrom. Cover photo courtesy of Dr. Ralf Kaegi, Swiss Federal Institude of Aquatic Science and Technology (Eawag) Library of Congress Cataloging-in-Publication Data Environmental and human health impacts of nanotechnology / edited by Jamie R. Lead and Emma Smith. p. cm. Includes bibliographical references and index. ISBN 978-1-4051-7634-7 1. Nanoparticles—Environonmental aspects. 2. Nanoparticles—Toxicology. 3. Nanostructured materials— Environmental aspects. 4. Nanostructured materials—Health aspects. 5. Nanotechnology—Environmental aspects. 6. Nanotechnology—Health aspects. I. Lead, Jamie R. II. Smith, Emma (Emma L.) TD196.N36E58 2009 620′.5—dc22 2009009688 A catalogue record for this book is available from the British Library. Set in 10 on 12 pt Times by SNP Best-set Typesetter Ltd., Hong Kong Printed and bound in Great Britain by CPI Antony Rowe Ltd, Chippenham, Wiltshire

Contents

Preface Biographies Contributors 1. Overview of Nanoscience in the Environment Mohamed Baalousha and Jamie R. Lead 1.1 1.2 1.3 1.4 1.5 1.6 1.7

1.8

1.9

1.10 1.11 1.12 1.13

1.14 1.15 1.16

Introduction History Definitions Investment and International Efforts Development: Four Anticipated Generations Applications of Nanotechnology Potential Benefits of Nanotechnology 1.7.1 Environmental 1.7.2 Human Health Potential Adverse Effects of Nanomaterials 1.8.1 Environmental 1.8.2 Human Health Classification 1.9.1 Chemistry 1.9.2 Origin 1.9.3 Size 1.9.4 State Sources of Nanomaterials in the Environment Properties of Nanomaterials Nanomaterial Structure–Toxicity Relationship Environmental Fate and Behaviour of Nanomaterials 1.13.1 Fate in Air 1.13.2 Fate in Water 1.13.3 Fate in Soil Potential for Human Exposure Detection and Characterization of Nanomaterials Issues to be Addressed 1.16.1 Nomenclature

xiii xv xvii 1 1 2 3 6 6 7 8 8 9 10 10 12 12 12 13 13 14 14 14 15 16 17 17 19 20 21 21 21

vi

Contents

1.16.2 1.16.3 1.16.4 1.16.5

Future Development and Risk Dosimetry Methods of Detection and Characterization Environmental Fate of Nanomaterials and their (Eco)Toxicology 1.17 Conclusion 1.18 References 2. Nanomaterials: Properties, Preparation and Applications Paul Christian 2.1 2.2 2.3

2.4

2.5

2.6

2.7 2.8 2.9

Overview Introduction Nanoparticle Architecture 2.3.1 Nanoparticle Surface 2.3.2 Charge Stabilisation 2.3.3 Steric Stabilisation Particle Properties 2.4.1 Surface Plasmon Resonance 2.4.2 Catalysis 2.4.3 Quantum Confinement 2.4.4 Mechanical Performance 2.4.5 Magnetic Properties 2.4.6 Interfacial Properties 2.4.7 Other Properties Nanoparticle Preparation 2.5.1 The Challenges of Nanoparticle Synthesis: Scale Up 2.5.2 Reactivity 2.5.3 Dispersability 2.5.4 Cost 2.5.5 Methods: Natural Sources 2.5.6 Top Down 2.5.7 Bottom Up 2.5.8 Metal Nanoparticles 2.5.9 Carbon 2.5.10 Graphene 2.5.11 Carbon Black 2.5.12 Inorganic Compounds 2.5.13 Polymers Applications of Nanoparticles and Nanotechnology 2.6.1 The Past 2.6.2 The Present and Near Future Implication for Environmental Issues Conclusions References

23 23 23 23 24 24 31 31 32 35 38 41 42 45 45 46 47 48 49 49 50 53 53 53 53 54 54 55 55 59 60 62 62 63 64 65 65 67 72 73 73

Contents

3. Size/Shape–Property Relationships of Non-Carbonaceous Inorganic Nanoparticles and Their Environmental Implications Deborah M. Aruguete, Juan Liu and Michael F. Hochella, Jr 3.1 3.2 3.3

3.4 3.5

3.6 3.7 3.8

4.4

4.5

4.6

79

Introduction 79 Inorganic Nanoparticle Anatomy 80 Redox Chemistry of Nanoparticles 81 3.3.1 Photoredox Chemistry in Semiconductor Nanoparticles 81 3.3.2 Redox Chemistry in Other Nanoparticle Systems 84 Size Effects in Nanoparticle Sorption Processes 87 Nanoparticle Fate: Dissolution and Solid State Cation Movement 89 3.5.1 Basic Energetic and Kinetic Considerations of Nanoparticle Dissolution 89 3.5.2 Effects of Nanoparticle Morphology 91 3.5.3 Effects of Nanoparticle Coatings and External Substances 92 3.5.4 Case Study: The Dissolution of Lead Sulfide Nanoparticles 94 3.5.5 Solid State Cation Movement in Nanoparticles 96 Effect of Nanoparticle Aggregation on Physical and Chemical Properties 98 Environmental Implications: General Discussion, Recommendations and Outlook 99 References 101

4. Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments Mohamed Baalousha, Jamie R. Lead, Frank von der Kammer and Thilo Hofmann 4.1 4.2 4.3

vii

Introduction Definition Major Types of Environmental Colloids 4.3.1 Inorganic Colloids 4.3.2 Organic Macromolecules Intrinsic Properties of Environmental Colloidal Particles 4.4.1 Size 4.4.2 Surface Charge 4.4.3 Surface Coating by Natural Organic Matter 4.4.4 Fractal Dimension Interaction Forces Between Colloidal Particles 4.5.1 DLVO Theory 4.5.2 Stability Criteria 4.5.3 Aggregation Kinetics 4.5.4 Non-DLVO Interactions Fate and Behaviour of Colloids in Aquatic Systems 4.6.1 Aggregation

109

109 112 112 114 117 121 121 121 123 124 126 127 129 129 130 136 136

viii

Contents

4.7

4.8 4.9

4.6.2 Disaggregation: Effect of Natural Organic Matter 4.6.3 Sedimentation Behaviour Fate and Behaviour of Colloids and Nanoparticles in Porous Media 4.7.1 Saturated Porous Media 4.7.2 Unsaturated Porous Media Conclusion References

5. Atmospheric Nanoparticles Aurélie Charron and Roy M. Harrison 5.1 5.2

5.3 5.4 5.5

5.6

5.7 5.8

Introduction Sources of Atmospheric Nanoparticles 5.2.1 Sources of Primary Nanoparticles 5.2.2 Secondary Sources Chemical Composition of Atmospheric Nanoparticles Fate and Behaviour of Atmospheric Nanoparticles Atmospheric Concentrations 5.5.1 Spatial Variations 5.5.2 Temporal Variations Measurement Methods for Atmospheric Nanoparticles 5.6.1 Particle Number Concentration 5.6.2 Surface Area 5.6.3 Mass Concentration 5.6.4 Chemical Composition Conclusions References

6. Analysis and Characterization of Manufactured Nanoparticles in Aquatic Environments Martin Hassellöv and Ralf Kaegi 6.1

6.2

Introduction 6.1.1 Nanoparticles in the Aquatic Environment 6.1.2 Concepts and Definitions Relating to Analysis and Characterization Nanoparticle Analysis and Characterization Methods 6.2.1 Important Nanoparticle Characteristics 6.2.2 Sampling, NP Extraction, Sample Preparations 6.2.3 Light Scattering Methods 6.2.4 Other Electromagnetic Scattering Methods 6.2.5 Fractionation and Separation Methods 6.2.6 Microscopic Methods 6.2.7 Spectroscopic Methods 6.2.8 Surface Area Measurements with Nitrogen Gas Adsorption

140 141 141 143 146 147 147 163 163 164 164 175 178 181 182 182 185 187 189 193 194 196 200 201

211 211 212 212 214 214 224 224 229 230 237 249 251

Contents

6.3

6.4 6.5 6.6

6.2.9 Method Validation Analytical Test Strategy in NP Exposure Assessment 6.3.1 Initial Material Characterization 6.3.2 Fate and Behaviour Assessment 6.3.3 Exposure Characterization in Effect Assessment Experiments 6.3.4 Monitoring Nanopollution Conclusions Acknowledgements References

7. Ecotoxicology of Manufactured Nanoparticles Simon C. Apte, Nicola J. Rogers and Graeme E. Batley 7.1 7.2

7.3

7.4 7.5

7.6 7.7

Introduction Physico-Chemical Transformation of Nanoparticles 7.2.1 Particle Dispersion and Aggregation 7.2.2 Nanoparticle Dissolution 7.2.3 Oxidation 7.2.4 Adsorption Reactions Mechanisms of Nanoparticle Toxicity in the Environment 7.3.1 Exposure Routes 7.3.2 Nanoparticle Interactions with Cells: Cellular Uptake 7.3.3 Toxicity Mechanisms 7.3.4 Bioaccumulation Development of Valid/Realistic Toxicity Testing Protocols Review of Ecotoxicity Studies 7.5.1 Overview 7.5.2 Carbon-Based Nanoparticles 7.5.3 Metal Oxides 7.5.4 Silver 7.5.5 Copper 7.5.6 Quantum Dots 7.5.7 Iron General Conclusions and Future Directions References

8. Exposure to Nanoparticles Robert J. Aitken, Karen S. Galea, C. Lang Tran and John W. Cherrie 8.1 8.2

Introduction Physical Characteristics and Properties of Nanoparticles 8.2.1 Terminology and Definitions 8.2.2 Nanoparticle Types

ix

251 252 252 252 253 254 255 256 256 267 267 269 270 272 273 275 275 275 277 280 283 284 285 285 289 293 296 298 298 299 300 301 307 307 309 309 311

x

Contents

8.3

8.4

8.5 8.6

8.2.3 Nanoparticle Production Processes 8.2.4 Nanoparticle Behaviour Nanoparticle Exposure 8.3.1 Exposure Scenarios 8.3.2 Exposure Metrics 8.3.3 Methods of Measuring and Characterising Exposure to Nanoparticles 8.3.4 Studies Investigating Nanoparticle Exposure 8.3.5 Numbers of People Potentially Exposed Control of Exposure 8.4.1 Introduction 8.4.2 Inhalation Exposure 8.4.3 Dermal Exposure 8.4.4 Ingestion Exposure Discussion References

9. Human Toxicology and Effects of Nanoparticles Vicki Stone, Martin J.D. Clift and Helinor Johnston 9.1

9.2

9.3

9.4 9.5 9.6 9.7 9.8

Introduction 9.1.1 Toxicology – What Is It? 9.1.2 Particle Toxicology 9.1.3 Risk Assessment Ultrafine Particle Toxicology 9.2.1 Air Pollution 9.2.2 Testing the Ultrafine Particle Hypothesis 9.2.3 Reactive Oxygen Species and Oxidative Stress 9.2.4 Uptake of Nanoparticles into Cells 9.2.5 Interaction of Nanoparticles with Defence Mechanisms 9.2.6 Nanoparticle Interactions with Other Pollutants and Molecules Engineered Nanoparticles 9.3.1 Fullerenes 9.3.2 Nanotubes and Other Fibre-Like Nanostructures 9.3.3 Metals 9.3.4 Metal Oxides 9.3.5 Quantum Dots Relating Physico-Chemical Properties to Toxicity: Structure-Activity Relationships Suggestions for Future Study Designs Conclusions Abbreviations References

314 316 319 319 327 330 338 344 346 346 347 349 350 350 353

357 357 357 357 358 360 360 361 363 365 366 367 368 369 370 373 374 377 379 381 381 382 382

Contents

10. Risk Assessment of Manufactured Nanomaterials Sophie A. Rocks, Simon J. Pollard, Robert A. Dorey, Paul T. C. Harrison, Len S. Levy, Richard D. Handy, John F. Garrod and Richard Owen 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 Index

Introduction Risk Assessment Process Nanomaterials – Issues for Risk Assessment Assembling Evidence for Safety and Intervention International Case Studies Data Gaps in Risk Assessment of Nanomaterials Summary References

xi

389

389 391 399 403 406 407 410 417 423

Preface Manufactured nanoparticles (NPs) are usually defined as materials purposefully produced by human activity and which have at least one dimension between 1 and 100 nm. It is important to distinguish NPs by source; the main other NPs are incidental: that is produced indirectly by human activities including fossil fuel combustion, and natural: that is produced by processes such as chemical hydrolysis, weathering and microbial action. Other size-based definitions of NPs exist and there are a wide variety of material types which fall within this definition. Nanoscience, which is the science dealing with nanoscale materials, can be seen as simply a subset of traditional colloids science. Nevertheless, a large number of novel processes occur below this size due to effects such as exponential increases in specific surface area and surface energy, quantum effects such as quantum confinement (where wave functions are constrained by the small particle size) and undercoordination of bonds at the particle surface. Processes which occur in this size range are thus different in many ways to traditional colloid chemistry and, in general, the differences become more pronounced at smaller sizes. The current interest in nanotechnology is due to these novel properties and their exploitation in industrial processes and consumer products. Huge and exponentially growing research and development funding from government and private sources has been spent to better develop and exploit these potential uses and NPs are now used widely. Silver NPs are currently used as bacteriocides in cosmetics, fabrics, medical and health-related products and elsewhere. Titanium dioxide NPs are used in sunscreens (along with zinc oxide) and self cleaning surfaces, where they have a photocatalytic effect on organic matter due to the production of reactive oxygen species (ROS) and because of this titania is also used as a bacteriocide. Cerium dioxide is widely used as an additive to diesel to improve fuel efficiency. A wide range of other materials such as carbon nanotubes, fullerenes, gold, iron, iron oxide and more exotic species are being developed and used. The extent of the applications and the possibility of unusual and unknown ‘nano’ effects has led to concern about their environmental and human health effects in the scientific community and equal concern in industry and from regulators and policy makers. A major driver for this in some quarters is undoubtedly the example of genetically modified organisms. The extensive public backlash has made the future of that technology quite uncertain and there has been a different approach in nanotechnology to openness and acknowledgement of the risks and a commitment to reducing these risks. Public response to nanoscience and

xiv

Preface

nanotechnology is currently limited by a lack of knowledge and wider impact but is generally positive with benefits expected in health, energy and the environment to name a few. Nevertheless, it is quite feasible that this attitude will change, particularly in view of developments in next generation nanomaterials, including selforganisation and self-assembly and the increasingly researched interface between ‘bio’ and ‘nano’. There are considerable benefits to be gained from the exploitation of nanoscience but current research tells us that there are indeed potential hazards in this area. It is incumbent on the relevant communities to ensure that NPs and other nanomaterials are used appropriately and designed and tested to be of minimal hazard and that exposure is not widespread; risk needs to be minimised and seen to be minimised to allow the full benefits of nanoscience and nanotechnology to be derived. Understanding the behaviour and impacts of nanotechnology in the environment and in human health is a daunting task and many questions remain to be answered: how do we measure concentrations of NPs in complex biological and environmental media?; what are the concentrations in environmental media and in organisms?; what are the correct metrics of measurement (mass or number concentrations for instance)?; what are the sources to the environment and humans?; what are the environmental transport pathways and ultimate sinks of NPs?; are NPs bioavailable and are they subject to bioaccumulation and biomagnification?; how do NPs distribute in the sub-cellular, organ and body environments?; how are transport, bioavailability and effects related to NP physico-chemical structure? Although a substantial amount of research is being performed, the research spending on the risks of nanotechnology and the health and safety and environmental implications is still tiny in comparison to its development and exploitation. This balance is unlikely to change enormously but there are good arguments to say that this should happen and change should come quickly. The questions above and related questions remain unanswered in the main and the purpose of this volume is to collate and discuss our current knowledge and point to future areas of research which are required. We would like to acknowledge and thank a number of people and institutions which made this book possible. The UK Natural Environment Research Council (NERC) provided funding via a Knowledge Transfer Network entitled Engineered nanoparticles in the natural aquatic environment (Nanonet), which enabled all authors and editors to convene for a two-day workshop to discuss the issues and finalise the chapters. We would like to thank the chapter authors for their efforts and their timely submissions, and the patience and help of the publishing team which was essential to the editors. Jamie Lead Emma Smith March 2009

Biographies

Jamie Lead is Professor of Environmental Nanoscience in the School of Geography, Earth and Environmental Sciences, University of Birmingham, UK. Professor Lead completed his PhD at Lancaster University, UK, in 1994 after investigating lanthanide and actinide speciation in natural waters and soils. At the same institution he later undertook postdoctoral research on the impact of size of natural aquatic colloids on transition metal chemistry. In 1998, he undertook further postdoctoral work at Geneva University, Switzerland, developing and using fluorescence correlation spectroscopy to quantify diffusion coefficients of natural organic macromolecules. In 2000, he became a Lecturer at the University of Birmingham and became full Professor at Birmingham in 2008. Professor Lead is Director of the Facility for Environmental Nanoparticle Analysis and Characterisation (FENAC), which is a national UK centre collaborating with the biological community investigating nanoparticle fate and effects. He has been a visiting researcher at CSIRO, Australia, and is a Fellow of the Royal Society of Chemistry, the International Union of Pure and Applied Chemistry and the Institute of Nanotechnology. Professor Lead’s main research interests, where he has published widely, relate to the relationships between chemistry, transport and bio-uptake of pollutants, especially in relation to the nanoscale in the environment. In particular, he is interested in the structure of natural ‘nanocolloids’ and the role this has in metal and manufactured nanoparticle chemistry, fate and behaviour. He is currently collaborating extensively with the ecotoxicological community by synthesising nanoparticles of silver, cerium, iron oxide and other materials and ensuring their full characterisation. These collaborations are particularly focussed on investigating mechanisms of nanoparticle biological uptake and effects

xvi

Biographies

Dr Emma Smith is currently Lecturer in Environmental Chemistry at the University of the West Indies. She received a degree in Oceanography and Chemistry from the University of Liverpool and a Masters in Marine Resource Development and Protection with distinction from Heriot Watt University. Her PhD thesis, Unresolved Complex Mixtures of Aromatic Hydrocarbons in the Marine Environment: Solubility, Toxicity and Photodegradation Studies, was carried out at Plymouth University in conjunction with Plymouth Marine Laboratory and won the SETAC Young Scientist Award in 2000 at World Congress in Brighton. Dr Smith then worked at Plymouth University on the characterisation of bioaccumulated and unidentified agent(s) causing reduced scope for growth in mussels and the potential ecological effects of chemically dispersed and biodegraded crude oils. She then worked at the University of Toronto within the Environmental NMR Centre, evaluating climatic controls on soil organic carbon composition and potential responses to global warming. Following this Dr Smith worked with Professor Lead at the University of Birmingham implementing the Nanonet project, a Knowledge Transfer (KT) Network in the area of manufactured nanomaterials (MNs) in the natural aquatic environment. In her current position she is responsible for teaching environmental chemistry, oceanography and ecotoxicology at UWI and is working with the Caribbean Ecohealth Programme and an EU Outreach project on assessing the potential environmental and human health effects of pollution.

Contributors

Robert J. Aitken

Institute of Occupational Medicine, Edinburgh, UK

Simon C. Apte Centre for Environmental Contaminants Research, CSIRO Land and Water, Bangor, Australia Deborah M. Aruguete Department of Geosciences, Virginia Polytechnic Institute and State University, Blacksburg, USA Mohamed Baalousha School of Geography, Earth & Environmental Sciences, University of Birmingham, Birmingham, UK Graeme E. Batley Centre for Environmental Contaminants Research, CSIRO Land and Water, Bangor, Australia Aurélie Charron Transport and Environment Laboratory, INRETS – French National Institute for Transport and Safety Research, Bron, France John W. Cherrie Institute of Occupational Medicine, Edinburgh, UK Paul Christian School of Chemistry, University of Manchester, Manchester, UK Martin J. D. Clift Institute for Anatomy, Division of Histology, University of Bern, Bern, Switzerland Robert A. Dorey Microsystems & Nanotechnology Centre, Cranfield University, Cranfield, UK Karen S. Galea

Institute of Occupational Medicine, Edinburgh, UK

John F. Garrod Department for Environment, Food and Rural Affairs, London, UK

xviii

Contributors

Richard D. Handy School of Biological Sciences, University of Plymouth, Plymouth, UK Paul T. C. Harrison Institute of Environment and Health, Cranfield Health, Cranfield University, Cranfield, UK Roy M. Harrison School of Geography, Earth & Environmental Sciences, University of Birmingham, Birmingham, UK Martin Hassellöv Department of Chemistry, University of Gothenburg, Gothenburg, Sweden Michael F. Hochella, Jr Department of Geosciences, Virginia Polytechnic Institute and State University, Blacksburg, USA Thilo Hofmann Austria

Department of Environmental Geosciences, Vienna University,

Helinor Johnston Applied Research Centre for Health, Environment and Society, Edinburgh Napier University, Edinburgh, UK Ralf Kaegi Swiss Federal Institute of Aquatic Science and Technology (Eawag), Dübendorf, Switzerland Jamie R. Lead School of Geography, Earth & Environmental Sciences, University of Birmingham, Birmingham, UK Leonard S. Levy Institute of Environment and Health, Cranfield Health, Cranfield University, Cranfield, UK Juan Liu Chemical and Material Sciences Division, Pacific Northwest National Laboratory, Rickland, USA Richard Owen

School of Biosciences, University of Westminster, London, UK

Simon J. Pollard Collaborative Centre of Excellence in Understanding and Managing Natural and Environmental Risks, School of Applied Sciences, Cranfield University, Cranfield, UK Sophie A. Rocks Collaborative Centre of Excellence in Understanding and Managing Natural and Environmental Risks, School of Applied Sciences, Cranfield University, Cranfield, UK Nicola J. Rogers Centre for Environmental Contaminants Research, CSIRO Land and Water, Bangor, Australia

Contributors

xix

Vicki Stone Applied Research Centre for Health, Environment and Society, Edinburgh Napier University, Edinburgh, UK C. Lang Tran

Institute of Occupational Medicine, Edinburgh, UK

Frank von der Kammer Department of Environmental Geosciences, Vienna University, Austria

UV/VIS & metals (scaled to fit) (a.u.)

1.2

Figure 4.2 FFF–ICPMS relative particle and element size distribution of aquifer colloids. The grey area represents the UV/VIS signal at 260 nm (as turbidity) and is a measure for the total colloid concentration. The coloured traces show the distribution of the major elements iron, aluminium and manganese and of the trace element lead. The signals are scaled to fit the graph. (v.d. Kammer, Doubascoux, Lespes, unpublished.)

1.0

UV/VIS Al Fe Mn Pb

0.8 0.6 0.4 0.2 0.0 0

50

100

150

hydrodynamic radius (nm)

Figure 4.6 A tapping mode image of a humic layer that has a 1 × 1 µm2 area machined away in contact mode. Lines a–c that cut across the image are where the cross-sections below the image were taken. (Reprinted with permission from C.T. Gibson, I.J. Turner, C.J. Roberts, J.R. Lead, Quantifying the dimensions of nanoscale organic surface layers in natural waters, Environmental Science & Technology, 41, 1339–44. Copyright 2007, American Chemical Society.) Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

200

0 250

Agglomeregation State

Concentration

Shape

Surface Speciation Size

Surface Charge

+ + + + +

Surface Functionality

Porosity / Surface Area

Size Distribution

Composition

Structure / Crystallinity

Figure 6.1 The important properties of manufactured nanoparticles in aqueous media are shown, indicating that the central concept of a homogeneous solid sphere with a clean surface is often an over-simplification. All or several of these properties are needed to understand the fate and behaviour of these nanoparticles in the environment or to characterize a certain ecotoxicology experiment. Therefore, a combination of analytical methods is required to obtain a complete characterization. (Figure partly adopted from Tinke et al., 2006.)

250 nm

250 nm (a)

3.0

3.0

2.0

2.0

1.0

1.0

0

1.0

0

3.0 µm

2.0

nm 200

nm 250

0

0

1.0

2.0

3.0 µm

–250

–200 0

1.0

2.0

3.0

µm

0

1.0

2.0

3.0 µm

(b)

Figure 7.8 (a) TEM images of Escherichia coli cells Left: untreated. Right: treated with halogenated MgO nanoparticles for 60 minutes. (b) Tapping mode AFM images of E. coli cells with the corresponding cross sections below. Left: Untreated (z-height 0–920 nm). Right: treated with halogenated MgO nanoparticles for 20 minutes (z-height 0–450 nm). Note the changes in smoothness and height of the cell indicating damage to the E. coli cell envelope upon nanoparticle treatment. (Reprinted with permission from P. K. Stoimenov, R. L. Klinger, G. L. Marchin and K. J. Klabunde, Metal oxide nanoparticles as bactericidal agents, Langmuir, 18, 6679–86. Copyright 2002 American Chemical Society.)

(a)

(b)

Figure 7.9 Uptake of nanoparticles by aquatic organisms. (a) Left: Silver nanoparticles in the membrane and inside of an Escherichia coli cell; right: EDS elemental mapping showing silver distribution through the sample. (b) Daphnia magna exposed to 5 mg/l of lipid coated single-walled nanotubes showing large numbers of tubes filling the gut track (1 h exposure) and clumps of precipitated tubes around the daphnid (20 h) (bar = 200 µm). ((a) Reprinted with permission from J. R. Morones, J. L. Elechiguerra, A. Camacho et al. (2005) The bactericidal effect of silver nanoparticles, Nanotechnology, 16, 2346–53. Copyright 2005 Institute of Physics. (b) Reproduced with permission from A. P. Roberts, A. S. Mount, B. Seda et al. (2007) In vivo biomodification of lipid-coated carbon nanotubes by Daphnia magna, Environmental Science & Technology, 41, 3025–9. Copyright 2007 American Chemical Society.)

1 Overview of Nanoscience in the Environment Mohamed Baalousha and Jamie R. Lead School of Geography, Earth and Environmental Sciences, University of Birmingham, United Kingdom

1.1

Introduction

Nanotechnology and nanoscience have a long tradition but in their current forms, where it is possible to image, manipulate and quantify materials on a nanoscale, they are relatively new. Understanding the environmental and human health impacts of these nanomaterials is itself very new and also highly multidisciplinary, requiring knowledge of environmental, analytical and physical chemistry, physics, materials science, (eco)toxicology and other disciplines. The breadth of the subject makes it demanding for anyone with a serious interest in studying the subject. Therefore, in this introductory chapter the current knowledge is briefly reviewed as a preamble to the more detailed analysis in subsequent chapters. Specifically, Chapter 2 discusses the architecture (structure and composition) of nanomaterials, the methods of producing stabilized nanoparticles, the properties, preparation methods and applications of nanomaterials and nanotechnology. Chapter 3 discusses the currently available knowledge about nanomaterials (colloidal inorganic material). It focuses on the chemical behaviour of nanomaterials that is likely to determine their environmental fate and toxicity, including size, redox chemistry, sorption processes, cation diffusion kinetics and dissolution. In Chapter 4 the available knowledge of natural aquatic and terrestrial colloids (including nanoparticles) is reviewed, including the major types of natural colloidal

Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

2

Environmental and Human Health Impacts of Nanotechnology

particles and their properties which are related to environmental processes, followed by a discussion of interaction forces between colloidal particles and their fate and behaviour in aquatic and terrestrial systems. Chapter 5 reviews the available knowledge about natural and adventitious nanoparticles in the atmosphere with a focus on their sources, transformations and concentrations. The analysis and characterization of manufactured nanoparticles in the environment are discussed in Chapter 6. It gives a general overview of the key properties that describe nanomaterials and the methods for sampling, extraction and sample preparation. This is followed by an extensive discussion of analytical tools for the characterization of nanoparticles, such as fractionation, filtration, microscopy and spectroscopic methods. Chapters 7 and 8 discuss the ecotoxicology and toxicology of manufactured nanoparticles, while Chapter 9 reviews the occupational health and exposure of nanoparticles. In Chapter 10 regulation, policy and risk management are discussed. This chapter starts by presenting the risk assessment framework for chemicals and then discusses the risk assessment of nanoparticles. It also discusses the critical issues for risk assessment of nanomaterials and the approach that should be adopted for this purpose.

1.2

History

The basic concept of nanotechnology was outlined by Nobel Prize winning physicist Richard Feynman in 1959 when he said ‘the principles of physics as far as I can see, do not speak against the possibility of manoeuvring things atom by atom. It is not an attempt to violate any laws; it is something, in principle, that can be done; but in practice it has not been done because we are too big’ (Feynman, 1960, 1992). The term ‘nanotechnology’ was first used by the Tokyo Science University professor Norio Taniguchi in 1974 to describe the precision of manufacture of materials at the nanometre scale. This term ‘nanotechnology’ became popular and in use in the public domain in 1980 when Eric Drexler published his book ‘Engines of Creation’. The advent of the scanning tunnelling microscope in 1981 and atomic force microscope in 1986 enabled atom clusters to be seen for the first time (Binnig et al., 1982, 1986). However, the history of nanoparticles goes back much further. ‘Soluble’ (or colloidal) gold appeared around the fifth to fourth century BC in Egypt and China and has been used for both aesthetic and curative purposes. In 1618, the philosopher and medical doctor Francisci Antonnii published a book which is considered as the first book about colloidal (nanoparticulate) gold. In 1676, the German chemist Johann Kunckels published a book in which he spoke about ‘drinkable gold that contains metallic gold in neutral, slightly pink solution that exert curative properties for several diseases’ and concluded that ‘gold must be present in such a degree of communition that it is not visible to the human eye’. In 1818, Jeremias Benjamin Richters noticed the formation of pink or purple solutions of fine gold and yellow solutions when the particles have aggregated (Daniel and Astruc, 2004) and, in 1857, Faraday reported the formation of red solutions of gold by the

Overview of Nanoscience in the Environment

3

reduction of an aqueous solution of AuCl −4 using phosphorus in CS2 (Faraday, 1857). Shortly after that, in 1861, the term ‘colloid’ (of which nanoparticles are the smallest fraction) was coined by Graham (Graham, 1861). This brief discussion shows clearly that nanoparticle usage has a long history. The novelty today is the scale of research and industry and the ability to manipulate and design materials at the nanoscale to create large structures with fundamentally new properties and functions. This will lead to unprecedented understanding and control of the properties of materials to discover novel phenomena, process and tools. Therefore, nanotechnology will enable a wide range of discoveries in all major scientific areas.

1.3

Definitions

Various definitions for nanomaterials have been given, or are actually in debate. Most of these are based on size and imply that there is a size range between that of molecules and bulk materials, where particles have unique properties different than those of molecules or bulk material (Tratnyek and Johnson, 2006). Some of these properties arise only for particles smaller than approximately 10 nm or so, where particle size approaches the length-scale of certain molecular properties (Klabunde et al., 1996). For instance, below 10 nm, particle specific surface area increases exponentially and qualitatively similar trends apply to related properties such as the ratio of surface/bulk atoms. Another example is that of quantum confinement, which arises because the band gap of semi-conducting materials increases as particle size decreases (Klabunde et al., 1996). The decrease in haematite particle size (from 37 to 7.3 nm) greatly promotes the oxidation of aqueous manganese (II) in the presence of molecular oxygen (Madden and Hochella, 2005), quite separate from the surface area related effect. Small magnetite nanoparticles (9 nm) exhibit greater reactivity toward carbon tetrachloride (CCl4) relative to larger nanoparticles (80 nm), both on mass and surface area normalized bases (Vikesland et al., 2007). The decrease in size of ceria nanoparticle alters the oxidation state of the nanoparticles with an increase in the fraction of Ce3+ at sizes less than approximately 15 nm with complete reduction of ceria particles to Ce3+ at sizes less than approximately 3 nm (Wu et al., 2004). Size dependent inhibition of nitrifying bacteria has been observed and the inhibition was correlated to the fraction less than approximately 5 nm in the suspension (Choi and Hu, 2008). These properties of nanomaterials and others are discussed in more details in Chapter 2 and 3. Nanoscience is, however, generally defined more inclusively, as the scientific study of materials on the nanoscale, approximately defined as the length scale between 1 and 100 nm (Borm et al., 2006b). Nanotechnology, as defined by the United States National Nanotechnology Initiative, is ‘the research and technology development at the atomic, molecular or macromolecular levels, in the length scale approximately 1–100 nm; the creation, and use of structures, devices and systems that have novel properties and functions because of their small size; and ability to be controlled or manipulated on the atomic scale’ (NNI, 2004). The Royal Society

4

Environmental and Human Health Impacts of Nanotechnology

and the Royal Academy of Engineering define nanotechnology as ‘the design, characterization, production and application of structures, devices and systems by controlling shape and size at the nanometre scale’ (Royal Society and Royal Academy of Engineering, 2004). Nanomaterials are a major component of nanotechnology and can be defined as materials that have one or more dimensions in the range 1–100 nm (Lead and Wilkinson, 2006). Importantly, nanomaterials have novel properties that differ from those of the same material without nanoscale features. A recent attempt to develop a more structured approach has been published by the British Standards Institution (BSI) (BSI, 2007). In its ‘Terminology for nanomaterials’ it defines nanoscale as the ‘size range from approximately 1–100 nm’, an nano-object as a ‘discrete piece of material with one or more external dimensions in the nanoscale’ and a nanoparticle as a ‘nano-object with all three external dimensions in the nanoscale’. A nanomaterial is a ‘material having one or more external dimensions in the nanoscale or which is nanostructured’, with nanostructured being defined as ‘possessing a structure comprising contiguous elements with one or more dimension in the nanoscale’. Definitions are also provided for nanorods: ‘nanoobject with two similar external dimensions in the nanoscale and the third dimension significantly larger than the other two external dimensions’; nanofibres: ‘flexible nanorods’; and nanotubes: ‘hollow nanorods’. The term high aspect ratio nanoparticles can be used to refer to fibres, rods or tubes. These definitions are based largely on particle size and do not account for the issue of particle size distribution satisfactorily or the change in properties as a function of size. A modified definition for a nanomaterial/nanoparticle can be based on the variation of material properties with size. Other definitions are still under discussion and various relevant bodies, for example, the International Organisation for Standardisation (ISO), American Society for Testing and Materials (ASTM) and the Organisation of Economic and Co-operation Development (OECD) are currently working on precise and formal definitions and nomenclature. The nanoscale dimension in comparison to the known dimensional scale of the universe is shown in Figure 1.1 (Hochella, 2002). At the smallest end of the scale (Figure 1.1a) are fundamental particles such as electrons and quarks, which are smaller than 10−18 m, and may approach 10−30 m in size or smaller, but such dimensions are not physically measurable at least at this time. At the larger end of the scale are the size of the Earth (107 m in diameter) and the sun (109 m in diameter). The nanoscale with other related objects is described in Figure 1.1b and is in the range 1–100 nm. A nanometre is a billionth of a metre (i.e. 10−9 m). The size of a single atom is of the order of several angstroms (0.1 nm). The size of a bacterium is about 1 µm (1000 nm), approximately the limit of visibility in light microscopes. In contrast, 100 nm is approximately equal to the size of a virus. Nanoparticles, like viruses, cannot be detected through standard light microscopes, because they are smaller than wavelengths of light (approximately 400–700 nm). They can be observed only with higher resolution microscopes such as scanning electron microscope (SEM, resolution of the order of 10 nm), transmission electron microscope (TEM) and atomic force microscope (AFM), which both have a resolution of the order of <1 nm).

Overview of Nanoscience in the Environment

5

a

b

Figure 1.1 (a) The known dimensional scale of the universe. On the small end, fundamental particles such as electrons and quarks are smaller than 10−18 m, and may approach 10−30 m in size or smaller, but such dimensions are not physically measurable at least at this time. Other stops depicted along this dimensional journey include: the scale of the solid earth sciences, from atoms to Earth (10−10–107 m); the Sun (109 m in diameter) as seen from the Extreme UV Imaging Telescope on the SOHO satellite; expanding gas rings (1016 m in diameter) from supernova SN1987a as observed by the Hubble Space Telescope; infrared image of the inner portion of our own galaxy (the Milky Way is nearly 1021 m in diameter); and distant galaxies (the most distant are 1026 m away). (b) The dimensional scale of the earth sciences. Stops depicted along this dimensional journey include: scanning tunneling microscope image of lead and sulfur atoms on a galena surface (atomic size 10−10 m); crystallization nucleus of calcite (10−9–10−8 m); bacterial cells (10−6 m in length); a single crystal of quartz (10−2 m); a typical open pit mine (the Carlin Mine in Nevada, USA, 102–103 m); Mount Fuji, Japan, a composite volcano (104 m); the Red Sea from space (105 m wide and 106 m long); Earth (107 m); the Earth–Moon system as seen from Apollo 11 (4 × 108 m). (Reprinted from Geochimica et Cosmochimica Acta, 66, M. F. Hochella, There’s plenty of room at the bottom: nanoscience in geochemistry, 735–43. Copyright 2002, with permission from Elsevier.)

6

1.4

Environmental and Human Health Impacts of Nanotechnology

Investment and International Efforts

Current spending is massive and set to increase further, while nanotechnology as an industry is also in its early stages. Governments worldwide spent about US$18 billion in nanotechnology between 1997 and 2005. The worldwide investment in nanotechnology has increased from $432 million in 1997 to $4.1 billion in 2005 (Thayer, 2006). By the year 2015, $1 trillion worth of products worldwide are expected to incorporate nanotechnology in key functional components and millions of jobs are expected to be affected by nanotechnology (Roco, 2005). Therefore, nanotechnology presents a major opportunity for economic growth in many countries and will influence different aspects of human life and the environment, as will be discussed later. At least 60 countries have initiated national activities in the nanotechnology field (Roco, 2005) and similar initiatives at an international level, for example via the European Union (EU) and OECD. In 2000, the United States launched a multidisciplinary strategy program through the National Nanotechnology Initiative. Japan and Western Europe have broad programs supported by government, combining academia, industry and other end-users. There are growing programs in Asia including China, South Korea, Taiwan and Singapore. In North America, the Canadian National Research Council has created the National Institute of Nanotechnology to fund nanotechnology research whilst emerging programs have been announced in Eastern Europe. The potential impact of nanotechnology is a global issue, and international partnerships and coordination of research and policy are essential to ensure that this emerging technology becomes sustainable. An important aspect of sustainability is the quantification and minimisation of risk to human and environmental health.

1.5

Development: Four Anticipated Generations

Currently, nanotechnology uses primarily passive nanomaterials in cosmetics, electronics, structural material and so on. However, more active structures are also being used and further developed, for instance in drug delivery. Self-organisation and assembly is a particularly active area of development. Future developments are predicted to increase the more active uses of these materials. For instance, Roco predicted four overlapping generations of nanotechnology products in the period 2000–2020 (Figure 1.2): passive nanostructures, active nanostructures, systems of nanosystems and molecular nanosystems (Roco, 2005). The first generation (after 2000) involved the basic discovery and production of passive nanostructures such as the simple components of nanoparticles, nanotubes, nanolayers and nanocoatings. They have steady-state structures and functions such as chemical reactivity or mechanical behaviour during their use (Renn and Roco, 2006). The second generation (∼2005) involves active nanostructures that change their properties (morphology, shape, mechanical, electronic, magnetic, biological, etc.) during operation. Examples are nanobiodevices, transistors, targeted drugs and chemicals, energy

Overview of Nanoscience in the Environment 1st: Passive nanostructures

7

(1st generation products)

~ 2000

2nd: Active nanostructures Ex: 3D transistors, amplifiers, targeted drugs, actuators, adaptive structures ~ 2005

3rd: Systems of nanosystems Ex: guided assembling; 3D networking and new hierarchical architectures, robotics, evolutionary ~ 2010

4th: Molecular nanosystems Ex: molecular devices ‘by design’, atomic design, emerging functions

New R&D Challenges

Ex: coatings, nanoparticles, nanostructured metals, polymers, ceramics

~ 2015–2020

Figure 1.2 Four generations of products: timeline for the beginning of industrial prototyping and nanotechnology commercialization. (With kind permission from Springer Science+Business Media: J. Nanopart. Res., 7, 2005, 707–12, International Perspective on Government Nanotechnology Funding in 2005, M.C. Roco, Figure 1.)

storage devices and so on. This is now the current generation and indeed some of these structures and activities are being developed successfully for commercial exploitation. Polymer-based drug delivery, for instance, is now well advanced (Park, 2007; Yih and Al-Fandi, 2006). In this system, the third generation (∼2010 onwards) includes systems of nanosystems that might self-assemble or self-organise, networking at the nanoscale to form larger architectures (Renn and Roco, 2006). Examples are artificial organs and electronic devices based on state variables (electron spin, nuclear spin or photonic state). The fourth generation (∼2015/2020) includes molecular nanosystems, where each molecule in the nanosystem has a specific structure and plays a different role. Molecular machines might be designed by atomic manipulation and may be used as devices that will approach the way biological systems work. Whatever happens in the near-future, it is certainly clear that massive and rapid changes are about to be brought about and it is incumbent upon us to be aware of these changes and as a society to use them in a beneficial manner, while minimizing any attendant risks.

1.6 Applications of Nanotechnology Even today, there are hundreds of commercially available products using nanotechnology currently on the market including cosmetics, sunscreens, paints and coatings, catalysts and lubricants, water treatments, security printing, textiles and sport items, medical and health cares, food and food packaging, plant production products,

8

Environmental and Human Health Impacts of Nanotechnology

veterinary medicines, electronics, fuel cells, batteries and additives, paper manufacturing and weapons and explosives (PEN, 2005) (Woodrow Wilson data base, http:// www.nanotechproject.org/inventories/consumer/). Commercialization of products such as self-cleaning glasses, disinfectant tiles and filters for air purification demonstrate the early successes of nanosystems for domestic and environmental applications. Again, this commercialization is at an early stage of rapid growth; for more detail see Chapter 2.

1.7

Potential Benefits of Nanotechnology

Nanotechnology has large potential benefits to a wide rage of disciplines, such as human health, medicine and the environment. These benefits are briefly discussed below. 1.7.1

Environmental

Nanomaterials have the potential to improve the environment through the development of new solutions to environmental problems, by direct application of nanomaterials to detect, prevent and remove pollutants or by using nanotechnology to design cleaner industrial process and create environmentally-friendly products. Nanoparticles can be used to convert pollutants to less harmful chemicals in the environment using the properties of large surface area, high reactivity and enhanced transport of nanoparticles. For instance, zero-valent iron nanoparticles have been used primarily in the United States to remediate ground water contaminated with chlorinated carbon compounds such as trichloroethylene (Zhang, 2003) and for the removal of arsenic from anoxic groundwater (Kanel et al., 2005). Other nanomaterials such as zero metallophyrinogens have been found to be effective for the degradation of tetrachlorethylene, trichloroethylene and carbon tetrachloride under anaerobic conditions (Dror et al., 2005). Poly(amidoamine) dendrimers can serve as chelating agents for recovering metal ions such as Cu(II), Ag(I), Fe(III) and so on from the aqueous phase (Diallo et al., 2005) and from soils (Xu and Zhao, 2006). Nanoporous ceramic combined with self assembled monolayers of functional groups have been used to remove heavy metals from wastewater (Mattigod et al., 2006). Nanomaterials have also been used for the removal of metal contaminants from air. For instance, nancomposites of silica and titania have been used for the removal of elemental mercury vapour as an alternative to conventional activated carbon injection (Pitoniak et al., 2005) and nanostructured silica was used to capture cadmium from an exhaust combustion environment (Lee et al., 2005). Nanosensors (e.g. zinc oxide (ZnO) semiconductor, tin(IV) oxide (SnO2) semiconductor, etc.,) can also be used to detect chemicals and biological contaminants in the environment (Vaseashta and Dimova-Malinovska, 2005). Semiconductor nanostructures can play an important role in developing smart materials that can simultaneously sense and destroy contaminants from the environment (Kamat and Meisel, 2003). Such a system is extremely useful when it triggers the degradation operation on demand, that is degradation becomes operational only when the

Overview of Nanoscience in the Environment

9

system senses the presence of the contaminant in the surrounding environment. As an example, ZnO semiconductor nanoclusters have been used to detect and destroy aromatic compounds present in water with UV irradiation (Kamat et al., 2002). A variety of sensors for environmental applications have been developed in recent years, including SnO2 semiconductor systems that have been used as conductometric gas sensors (Barsan et al., 1999; Sberveglieri, 1995), a TiO2 electrode for determining the chemical oxygen demand (COD) of water (Kim et al., 2001) and porous silicon nanocrystals for detecting nerve gas agents (Sohn et al., 2000) and nitrogen containing organic compounds (Germanenko et al., 2001). Nanotechnology may also help the environment by addressing the long term sustainability of resources (water, energy, materials, ecosystems, land and air) by increasing materials and energy efficiency, reducing the need for solvents and reducing waste product volumes and concentrations (Rickerby and Morrison, 2007). Nanotechnology may reduce energy demand through more efficient and effective use of materials; the use lighter materials for vehicles, materials and geometries for more effective temperature control, materials of better electrical transmission and less dissipation and materials for the next generation of fuel cells. For instance, cerium oxide nanoparticles have been used to decrease diesel emission (Jung et al., 2005). Quantum dots, nanoparticles of semiconductors, could make more efficient solar cells (Nozik, 2002). Nanoparticles in paint technology offer the possibility of thinner, and therefore lighter, coatings, which could reduce, for example, the weight of aircraft, increase fuel efficiency and so reduce carbon dioxide emissions. Advanced filtration may enable more water recycling and desalination, which enable more energy-efficient water purification (Miyaki et al., 2000). 1.7.2

Human Health

Nanotechnology has the potential to improve the lives of people in general and especially of those with severe injuries or medical conditions. The following applications are either in the early stages of development (start of list) or discussed as future possibilities (later in the list) (Bogunia-Kubik and Sugisaka, 2002; Roco and Bainbridge, 2003; Roco, 2003; Zajtchuk, 1999): • Provision of new formulations and routes for drug delivery to difficult organs such as the brain, enormously broadening the drugs’ therapeutic potential (Yih and Al-Fandi, 2006). For example poly(butyl cyanoacrylate) nanoparticles coated with polysorbate 80 can be used to enhance the delivery of apolipoproteins to the brain (Kreuter, 2001). Ultrafine particles can mediate the delivery of [3H]dalargin to the brain (Alyaudtin et al., 2001), • Development of new vaccines (Cui and Mumper, 2003). • Replacement of diseased organs and repair of nerve damage (Yang et al., 2004a). • Early diagnoses, treatment and prevention of cancer and other diseases (Cuenca et al., 2006). • Artificial nanoscale devices can be introduced into cells to play roles in diagnostics and potentially as active components (Cornell et al., 1997; Freitas, 1998). • Mobilisation of the body’s own healing abilities to repair or regenerate damaged cells and re-grow rapidly damaged neurons (Yang et al., 2004a).

10

Environmental and Human Health Impacts of Nanotechnology

• Combining nanotechnology with biotechnology will enable repair or replacement of defective cellular components such as protein signalling receptors (Roco and Bainbridge, 2003). • Establishment of direct links between neuronal tissue and machines that would allow direct control of mechanical, electronic and even virtual objects as if they were extensions of human bodies by brain-to-brain and brain-to-machine interfaces (Roco and Bainbridge, 2003). • Continuous health monitoring and semi-automated treatment using small and cheap sensors and other implantable devices (Roco and Bainbridge, 2003).

1.8

Potential Adverse Effects of Nanomaterials

Nanotechnology can be used for harmful, as well as for beneficial, purposes and examples include in weapon construction (Glenn, 2006). Apart from the direct use of nanomaterials for harmful purposes, the rapid growth of nanotechnology industry will inevitably increase the concentration of nanomaterials in the environment, with potential consequence for human and environmental exposure. There are few studies on the effect of nanoscale materials on the environment or health, in part because of our inability to reliably quantify the concentration of nanoparticles in the environment (Chapter 6). Results are still inconclusive and the gaps in our knowledge of the environmental, health and safety implications associated with nanomaterials are large (Royal Society and Royal Academy of Engineering, 2004). Hence, it is not yet possible to draw any broad conclusions about which nanomaterials may pose hazard and/or risk. Nanomaterials may affect the environment and human health in different ways, as described below. 1.8.1

Environmental

Although less studied than human health, there is also cause for concern regarding environmental impacts, and areas such as ecotoxicology, environmental chemistry and fate and behaviour are areas of intense current research. Nanomaterials may potentially impact the environment in three possible ways: (i) direct effect on micro-organisms, invertebrates, fish and other organisms; (ii) interaction with contaminants, that may change the bioavailability of toxic compounds and/or nutrients; and (iii) changes to non-living environmental structures. 1.8.1.1 Toxicity of Nanomaterials The majority of current studies of the toxicity of nanomaterials have been performed on a limited number of nanomaterials and aquatic species, usually at high concentration and over short exposure periods (Nowack and Bucheli, 2007). There are few, if any, chronic, full life cycle or multigenerational studies available. An early study of fullerenes (C60) has shown that they induce oxidative stress in the brain of juvenile largemouth bass without a clear concentration dose–response relation (Oberdorster, 2004). Nanoclusters of C60 have been shown to generate reactive oxygen species in water under UV and polychromatic light. Fullerenol (C60(OH)24)

Overview of Nanoscience in the Environment

11

produces reactive oxygen species under similar conditions (Pickering and Wiesner, 2005). Data on potential toxicity of nanomaterials on terrestrial species (plants, soil invertebrates or micro-organisms) is increasing rapidly (Handy et al., 2008), but there is a growing need for information on more realistic conditions as well as descriptive and, eventually, predictive models of nanomaterials impacts on organisms. There are several reviews on the toxicity of nanomaterials (Borm et al., 2006a) such as carbon nanomaterials (Hurt et al., 2006), carbon nanotubes (Donaldson et al., 2006; Lam et al., 2006) and quantum dots (Hardman, 2006). Nanomaterials may pose risk to human health and to the environment but only a few specific nanoparticles have been investigated in a limited number of test systems and extrapolation of this data to other materials is not possible. Manufactured nanomaterials with new chemical and physical properties are being produced constantly and the toxicity of these is unknown. Therefore, despite a small but increasing database on the behaviour and effects of nanoparticles in the environment, no overarching and predictive models exist and their development is an urgent and active area of research. The (eco)toxicity of nanoparticles is further discussed in Chapter 7. 1.8.1.2

Nanomaterials as Carriers of Coexisting Contaminants

In the environment, contaminants are often bound to natural solid phases including nanoparticles (Lead et al., 1999; Lyven et al., 2003). Nanomaterials have the capacity to bind substantial fractions of contaminants such as trace metals and organics. Carbon nanotubes have been shown to sorb a variety of organic compounds (Long and Yang, 2001) and metals such as copper (Liang et al., 2006) and rare earth metals (Liang et al., 2005). Fullerenes will sorb organic compounds such as naphthalene (Cheng et al., 2004) and can be used for the removal of organometallic compounds (Ballesteros et al., 2000). Zero-valent iron oxide nanoparticles have been applied for the remediation of organic contaminants (Obare and Meyer, 2005; Zhang, 2003). Nanoporous ceramic sorbents have been used for the immobilization of cationic metals (Mattigod et al., 2006). It is possible that in this form contaminants may be more bioavailable and may be taken up through cell membranes more readily, although few studies are available (Navarro et al., 2008). The use of nanoparticles as drug delivery vehicles points to this being a potential problem. 1.8.1.3

Effect on Micro-Ecosystems

There are few studies demonstrating the indirect (nonbiological) impact of nanoparticles on environmental systems, although this remains a concern (Buffle, 2006). For instance, zero-valent iron nanoparticles have been used for the remediation of soil and aquifers contaminated with halogenated hydrocarbons and heavy metals (Zhang, 2003). However, as oxygen and other oxidizing materials are consumed, moderate to strong reducing/anaerobic conditions were created. Although this study did not examine consequences, it is likely that changes in, for instance, microbial ecology will have occurred. Aggregation and sedimentation of nanomaterials can be responsible for the scavenging of colloids from the water column to the sediment (Baalousha et al.,

12

Environmental and Human Health Impacts of Nanotechnology

2008a) and will lead to the formation of porous microstructures of both natural colloids and nanomaterials that can change their structure. The structure of such natural microstructures/aggregates controls the transport of sorbed trace contaminants and, thus, the potential receptor organisms. Nano- to micro-sized or larger structures that form natural colloidal and nanoparticulate material in the environment may be altered by the aggregation or other effects of natural waters. As these structures are vital for the maintenance of life through being direct energy and carbon sources, as pH and metal ion buffers and other properties, the effects of nanoparticles on these may have considerable adverse effects (Buffle, 2006). This possible impact is likely to occur on only a very small scale and at high concentrations of nanoparticles. 1.8.2

Human Health

There is now a substantial literature showing that some nanoparticles may have a toxic effect and there is a cause for concern as to their effect on human health, which has largely developed out the voluminous literature on the health impacts of adventitious nanoparticles (usually termed ultrafine, derived from combustion and other industrial processes). For instance, sunscreen titanium dioxide and zinc oxide can cause oxidative damage to DNA in vitro and in cultured human fibroblasts (Dunford et al., 1997). Nano-cerium-element doped titanium dioxide induces apoptosis of Bel 7402 human hepatoma (liver) cells in the presence of visible light (Wang et al., 2007). Inhaled ultrafine particles (UFPs) can gain access to the blood stream and can then be distributed to other organs in the body; this has been shown for synthetically produced UFPs (nanoparticles) such as C60 fullerenes which accumulate in the liver. Even large particles outside the ‘nano’ range can penetrate the stratum corneum of human skin and reach the epidermis and, occasionally, the dermis and may be taken up into the lymphatic system, while larger particles cannot (Tinkle et al., 2003). This means that there is a strong possibility that nanoparticles can be assimilated into the body through the skin, especially damaged skin. Chapter 9 details the potential effect of nanoparticles on human health.

1.9

Classification

Currently, there are hundreds of nanomaterials in use or under development that can be classified in different ways according to their chemistry, origin, size, or their state. Other nanomaterials are expected to appear in the future. 1.9.1

Chemistry

The most common method of classifying nanoparticles is by the chemistry of the core material. Carbon-based materials are mainly composed of carbon in the form of a hollow spheres, ellipsoids or tubes such as carbon black (generally not classed as a manufactured nanoparticle), carbon nanotubes (single wall (SW), multiple wall (MW)) and fullerenes (generally denoted Cn, where n may be 60, 70 or higher numbers). Metal-based nanomaterials, including quantum dots, metals (nanogold,

Overview of Nanoscience in the Environment

13

nanosilver) and metal oxides (TiO2, ZnO, CeO2, SiO2 and Al2O3) represent an important group of nanomaterials. Metal oxides represent the largest number of commercial nanomaterials. Dendrimers are highly branched, three dimensional, nano-sized polymers with controlled structure, several chain ends on their surface that can be tailored to perform specific chemical functions and a tendency to adopt a globular shape (for more discussion about molecules, dendrimers and nanoparticles, see Chapter 2). These properties have made them useful in medical chemistry application, for instance they can be used for drug delivery by placing drug molecules within the cavities of their three dimensional structure (Cloninger, 2002; Liu and Fréchet, 1999). Composites (including nanoclays) are complex structures combining several nanoparticles or nanoparticles with larger particles. 1.9.2

Origin

Nanomaterials can be classified according to their origin into manufactured, natural or adventitious. Manufactured nanomaterials are intentionally produced due to their particular properties such as carbon nanotubes, metals and metals oxides such as silver and titanium dioxide (Chapter 2). In general, when nanoparticles or nanomaterials are discussed, it is these manufactured materials which are being discussed. Natural nanomaterials are naturally occurring such as viruses, aquatic and terrestrial colloidal matter, mineral composites and ultrafine particulate matter (Chapters 4 and 5). Adventitious nanomaterials are unintentionally produced and occur as a result of industrial processes, such as diesel exhaust particles, or other friction or airborne combustion by-products, such as from road vehicles, fossil fuel and cooking (Chapter 5). 1.9.3

Size

How nanoparticles are compared to other size dependent environmentally relevant particles that have been known for decades is shown in Figure 1.3. In aquatic

1

10

100

dissolved conceptual

450

1000 Nanometers

dissolved particulate operational Water

colloids Nanoparticles PM0.1

Air

0.1

Ultrafine particles

Figure 1.3 Definition of different sizes relevant to nanoparticles. (Reprinted from B. Nowack, and T.D. Bucheli, Occurrence, behavior and effects of nanoparticles in the environment, Environment. Pol., 150, 5–22. Copyright 2007, with permission from Elsevier.)

14

Environmental and Human Health Impacts of Nanotechnology

systems, colloids are defined (Chapter 4) as materials naturally produced by processes such as erosion and degradation, within the size range 1–1000 nm (Lead and Wilkinson, 2006). In the atmosphere, particles (Chapter 5) are mainly classified according to their size (PM10, PM2.5 and PM0.1). Thus, nanoparticles correspond to PM0.1 (these are usually called ultrafines in this field) and overlap with the small fraction (<100 nm) of aquatic colloids. 1.9.4

State

Nanomaterials can be present in different states, either free, fixed or aggregates. Free nanomaterials are single individual nanomaterials. Fixed nanomaterials are those incorporated in products such as computer chips, rackets, tennis balls and so on. Aggregates are associations of nanomaterials in a network-like structure that forms due to the effect of physicochemical properties of the media in which they are contained.

1.10

Sources of Nanomaterials in the Environment

An overview of potential sources and pathways of nanomaterials in the environment is provided in Figure 1.4. As contaminants, engineered nanomaterials may be released from a point source, such as a production facility, landfill and wastewater treatment plant, or a non-point source, such as wear from material containing nanomaterials. They may be accidentally released during production and transport via a leakage from improper sealing or intentionally released, such as the use of zero-valent iron nanoparticles for soil remediation. Further, nanomaterials can be released into the environment either as single nanoparticles or as embedded in a matrix, from where release of nanomaterials will occur through the degradation of the matrix material (Kohler et al., 2007). Currently, the main source of adventitious nanomaterials is the incomplete oxidation of anthropogenic compounds by human activities, such as wood burning, fuel burning in diesel engines or cars with defective catalytic converters. Sources and pathways of natural aquatic nanoparticles are discussed in more detail in Chapter 4 and those of atmospheric nanoparticles are discussed in Chapter 5.

1.11

Properties of Nanomaterials

At the nanoscale, material properties such as mechanical, electronic, magnetic, optical, chemical, biological and others may differ significantly from the properties of micrometre-sized or bulk counterparts. For instance, they can be more chemically reactive and catalytic, have greater strength or conduct electricity more effectively. This difference in materials behaviour at nanoscale level can be explained by high specific surface area per unit mass, the increase of proportion surface atoms with the decrease in size (Oberdorster et al., 2005a), the presence of undercoordinated bonds, greater disorder at the surface or other mechanisms. Properties that

Overview of Nanoscience in the Environment

15

Figure 1.4 Some possible exposure routes for nanoparticles and nanotubes based on current and potential future applications. (Redrawn from Nanoscience and Nanotechnologies: Opportunities and Uncertainties, 2004, with permission from The Royal Society.)

may be important or different in nanoparticles and are size dependent include chemical reactivity, adsorbent properties, catalytic surface, functional group content and nature, surface charge, crystal structure, oxidation state, tendency to aggregate and dissolve and degree of toxicity. These properties are discussed in more detail in Chapters 2 and 3, their effects on the fate and behaviour of nanoparticles in the environment is discussed in Chapters 4 and 5 and their effect on toxicity is discussed in Chapters 7 and 9.

1.12

Nanomaterial Structure–Toxicity Relationship

The causes of the toxicity of nanoparticles are still unclear. So far, research results are more suggestive than definitive and no quantitative structure–activity relationship models are available. Physicochemical properties such as size, specific surface area, structure, morphology, chemical composition, solubility, reactivity, aspect ratio, photochemistry, production of reactive species and surface properties (i.e. charge and coating) can be of prime importance. For instance, size is the main characteristic

16

Environmental and Human Health Impacts of Nanotechnology

determining the extent of uptake and toxicity of many nanomaterials, which have been shown to be size dependent (Chithrani et al., 2006; Limbach et al., 2005). The size of nanoparticles may have crucial impact on where nanoparticles end up in the body in humans and other organisms. In humans, large inhalable nanomaterials tend to deposit primarily in the nose and throat while smaller particles can find their way to the upper airways. The smallest particles can penetrate deeper into the alveolar region and might penetrate to different parts of the respiratory tract. Because of their small size, nanomaterials can potentially pass through the lungs into the bloodstream and to be taken up by cells, reaching potentially sensitive sites such as bone marrow, liver, kidneys, spleen and heart. When ingested, nanomaterials can end up in the liver, the spleen, the kidneys and elsewhere. Skin penetration is poorly studied, but nanomaterials smaller than 50 nm may penetrate the skin more easily, although a number of initial studies suggest that dermal uptake is low in healthy skin (Borm et al., 2006a; Oberdorster et al., 2005b. Further, the size of nanomaterials is an important factor in determining their uptake across the gill membrane or the gastrointestinal tract (GI) of aquatic and terrestrial organisms, due to passive diffusion to the cell and the ability of nanomaterials to penetrate the cell membrane. It has been suggested, for instance, that the absolute limit for passive diffusion through fish gills is about 1 nm (Nitta et al., 2003). For more details, the reader is referred to the reviews by (Borm et al., 2006a; Oberdorster et al., 2005b) and to Chapter 9. The aspect ratio (length/thickness) of nanoparticles along with biological persistence is likely to be an important factor in their toxicology. Concerns about the aspect ratio of nanomaterials stem from the previous knowledge of fibre toxicology, especially studies on synthetic vitreous fibres and asbestos, which identify the important role of length in pulmonary bioresistance (Oberdorster et al., 2007). Fibres longer than (∼ 20 µm) cannot be phagocytosed by alveolar macrophages, causing reduced clearance and accumulation. Recent work has shown, dependent on structure, that carbon nanotubes (CNTs) act in a similar manner to asbestos and may have an even greater biological activity and therefore hazard (Poland et al., 2008). Particle composition and surface charge are also important factors determining the uptake and toxicity of nanomaterials. Mineral particle induced apoptosis was dependent on particle size, whereas composition and surface reactivity were found to be most important for the proinflammatory potential of the particles (Schwarze et al., 2007). Surface charge may alter the blood–brain barrier integrity and permeability in mammals (Lockman et al., 2004). The effect of the properties of nanomaterials on their uptake and toxicity is discussed in detail in Chapters 7 and 9.

1.13

Environmental Fate and Behaviour of Nanomaterials

The potential environmental fate and behaviour of nanomaterials are not yet well understood and available studies are scarce. Determining the fate and behaviour of nanomaterials in the environment requires understanding of potential sources of nanomaterials (Section 1.9), their fate in air, soil and water (Chapters 4

Overview of Nanoscience in the Environment

17

and 5), and their transformation, degradation and persistency. In addition, the fate of nanomaterials in the environment is likely to vary with the physical and chemical characteristics of the nanomaterials and the containing medium and the interaction of nanomaterials and other environmental contaminants. Understanding the fate and behaviour of natural colloids (nanoparticles) in water and soil (Chapter 4) and ultrafine particles in air (Chapter 5) will improve our understanding and ability to predict the fate and behaviour of manufactured nanomaterials in these systems. 1.13.1

Fate in Air

Natural and particularly adventitious (e.g. from fossil fuel combustion) nanomaterials are very abundant in the atmosphere and are present in the vicinity of any combustion process (see details in Chapter 5). Their concentration is highly elevated above the unpolluted background concentration in areas with heavy human occupancy. Atmospheric nanomaterials have three main sources: (i) primary emission, refers to those that are directly emitted from road traffic exhaust and industrial combustion; (ii) secondary emission, refers to those that are formed within the atmosphere from the condensation of low volatility vapours formed from the oxidation of atmospheric gases; and (iii) formation during diesel exhaust dilution. However, there are few, if any, data sets on manufactured nanomaterials in the atmosphere in terms of sources and concentrations. This lack of data about engineered nanomaterials in the atmosphere is due to the absence of methods capable of discriminating engineered nanomaterials from the background concentration from other sources (Harrison, 2007), which is similar to the situation in aquatic and terrestrial environments (Chapter 4). From what is known from fine and ultrafine particles, nanomaterials can undergo several processes in the atmosphere. Some nanomaterials can grow by condensation of low volatility compounds, or shrink by evaporation of adsorbed water or other volatiles, resulting in the variation in particle size distribution but not the overall number concentration (Zhang et al., 2005). Further, atmospheric nanomaterials can aggregate, resulting in an increase in particle size with a decrease in the number concentration (Gidhagen et al., 2004). They can be lost from the atmosphere by dry and wet deposition, both of which are efficient for very small particles of natural origin and so presumably also for engineered nanomaterials. This results in a reduction in particle number concentration and a shift in particle size distribution to larger sizes (Clarke et al., 2004). They can be diluted in mixing with cleaner air. For example, particles from traffic mix upward with less polluted air, leading to a reduction in the number concentration and generally an increase in particle size distribution simply because the diluted air contains a distribution of larger particle sizes (Zhang et al., 2005). These processes are discussed in detail in Chapter 5 for adventitious nanoparticles. 1.13.2

Fate in Water

The potential fate and behaviour of engineered nanomaterials, once they are released into the aquatic environment, can be understood in the light of the existing

18

Environmental and Human Health Impacts of Nanotechnology

knowledge on the fate and behaviour of natural colloidal particles (Chapter 4). This knowledge suggests that the fate of nanomaterials in the aquatic environment can be influenced by a variety of processes, such as: dispersion/diffusion, aggregation and disaggregation, interaction between nanoparticles and natural water components, sedimentation, biotic and abiotic degradation, transformation and photoreaction/light. These reactions may alter the physical and chemical properties of nanomaterials and so alter their behaviour in the aquatic environment. There have been few studies on the aqueous stability and aggregation of nanomaterials under environmentally relevant conditions. Brant et al. (2005) studied the aggregation and deposition of fullerene nanoparticles in aqueous media at variable ionic strength. They found that, while in the absence of electrolytes nC60 stayed stable over time, 0.001 M solution ionic strength (NaCl) was enough to destabilize the nC60 by screening their electrostatic charge and produce large aggregates that settle rapidly (Brant et al., 2005). The addition of humic acid has been shown to enhance the stability of fullerene suspension in the presence of sodium chloride and magnesium chloride and low concentrations of calcium choride (Chen and Elimelech, 2006). However, at high concentrations (above 10 mM) of calcium choride, enhanced aggregation of fullerene nanoparticles was observed due to bridging mechanism by humic acid aggregates (Chen and Elimelech, 2007). Other research has found similar complex interactions between natural and manufactured nanomaterials (Baalousha et al., 2008a; Giasuddin et al., 2007; Hyung et al., 2007). Extracted Suwannee River humic acid (SRHA) and natural surface water (actual Suwannee river water with unaltered natural organic macromolecules (NOM) background) has been shown to stabilize multi-wall carbon nanotubes (MWCNTs) (Hyung et al., 2007). However, extensive flocculation of multi-wall carbon nanotubes (i.e. formation of floating aggregates and partial sedimentation of other aggregates) was observed when mixed with natural waters from a lake, presumably due to the high ionic strength and the presence of divalent cations such as calcium (Baalousha et al., 2008a). Apparently, sorption of humic substances enhances the stability and inhibits the aggregation of carbon nanotubes to a certain extent (Hyung et al., 2007), However, cations, particularly divalent cations such as calcium and magnesium reduce the stability of carbon nanotubes in the absence or presence of natural organic matter surface coating (Baalousha et al., 2008a). Disaggregation is as important as aggregation in determining the fate and behaviour of nanomaterials, though few studies are available (Baalousha, 2009; Ouali and Pefferkorn, 1994). Natural organic matter has been shown to induce the disaggregation of iron oxide nanoparticle aggregates (5–10 µm) formed at pH 7, likely due to formation of a surface coating of natural organic matter on the surface and pore surface of the aggregates and thus the enhancement of surface charge (Baalousha, 2009). In addition, it has been shown that certain polymers are able to disaggregate latex particle (885 nm in diameter) aggregates (Ouali and Pefferkorn, 1994). However, polysaccharide or humic acid did not result in the disaggregation of polystyrene latex particle aggregates; this was explained by the existence of strong interparticle forces within flocs which prohibited aggregate break-up upon adsorption of natural organic matter (Walker and Bob, 2001).

Overview of Nanoscience in the Environment

1.13.3

19

Fate in Soil

The potential fate and behaviour of engineered nanomaterials in soil can be understood in the light of the existing knowledge on the fate and behaviour of natural colloidal particles (Chapter 4). Nanomaterials are small enough to travel through soil pores. However, they can be sorbed to soil particles due to their high surface area and, therefore, become immobilized. In addition, the formation of large aggregates of nanomaterials can immobilize them by filtration, sedimentation or straining in smaller pores. At the moment, little information is available on the transport and fate of nanomaterials in the natural porous environment. However, some data are available from laboratory column studies using porous media (Lecoanet et al., 2004; Lecoanet and Wiesner, 2004; Li et al., 2006; Schrick et al., 2004; Yang et al., 2007b), which suggest that transport is often relatively rapid and dependent on the type of nanomaterials. Laboratory soil column experiments on iron oxide and zero-valent iron nanoparticles show that their mobility is more limited due to the efficient filtration mechanisms of aquifer material. Field studies on iron oxide nanoparticles indicate that they may migrate only few centimetres to few metres from the point of injection and that their mobility is dependent on many factors, such as particle size, solution pH, ionic strength, soil composition and ground water flow velocity (Li et al., 2006; Schrick et al., 2004). The zero-valent nanoparticles are somewhat more mobile as they have been synthesized on supports acting as a delivery vehicle (Schrick et al., 2004; Yang et al., 2007) These delivery vehicles, including anionic hydrophilic carbon and poly(acrylic acid) (PAA), bind strongly to the iron, create highly negative surfaces, thus effectively reducing the aggregation among zero-valent iron particles, and reduce the filtration removal by aquifer materials. Laboratory soil column experiments with iron/hydrophilic carbon, iron/PAA and unsupported iron nanoparticles suggest that the anionic surface charges can enhance the transport of iron nanoparticles through soil and sand packed columns in comparison with unsupported iron nanoparticles (Schrick et al., 2004; Yang et al., 2007). In addition, the transport of iron nanoparticles (2–10 nm) through a porous media column (glass beads, unbaked sand and baked sand) can be enhanced by surface modification via surfactant sorption. Un-modified iron nanoparticles were immobile and aggregated on porous media surfaces in the column inlet area (Kanel et al., 2007). Although surfactants and polymers enhance the transport of nanoparticles, the role of natural organic matter in nanoparticle facilitated transport has not yet been investigated, but is likely to be important. Further, the characteristics of the soil matrix may influence the diffusion and transport of nanoparticles. PAA-modified nanoiron slurry has been found to travel easily through silica sand columns, but not loamy sand soil columns (Yang et al., 2007). The transport of water stable ‘nC60 particles’ underivatized C60 crystalline nanoparticles, stable in water for months through a soil column was investigated at different flow rates, while other column operating parameters remained fixed through all the experiments. The nC60 particles were observed to be more mobile at higher flow velocity due to less interaction time between the nC60 particles and the porous media (Cheng et al., 2005). Lecoanet and Wiesner (2004) studied the

20

Environmental and Human Health Impacts of Nanotechnology

transport and removal of silica, anatase and fullerene based nanoparticles in porous media. They found that the removal of anatase is less significant at higher flow rates. However, no dependence on velocity of particle passage through the porous medium was observed for silica particles and the fullerene based nanoparticles. This was explained by the very small value of the collision efficiency factor in the case of silica and the deposition of fullerene based nanoparticles on the porous media at higher flow rates after one void volume, which limits the interaction of these nanoparticles with the porous media and reduces particles removal afterwards (Lecoanet and Wiesner, 2004). The discussion above, in addition to previous knowledge on colloidal transport in porous media, suggests that the mobility of nanoparticles in soils depends on: (i) nanoparticle physical–chemical characteristics, that is size, shape, surface coatings and stability; (ii) the properties of the soil and environment, that is clay, sand, colloids, natural organic matter, water chemistry and flow rates; and (iii) the interaction of nanoparticles with natural colloidal material, that is surface coating, aggregation/disaggregation and sorption to larger particles. The more theoretical aspects of particle movement in porous media are discussed in Chapter 4.

1.14

Potential for Human Exposure

For nanomaterials to cause concern to human heath it is necessary to be exposed to them. There are multiple exposure scenarios depending on the details of manufacture, use and disposal. Throughout these scenarios the population exposed, the levels of exposure, the duration of exposure and the nature of the material to which people are exposed are all different. In an occupational context, exposure to nanomaterials can occur for workers at all phases of material life cycle. During the development of a new material, it is probable that material will be produced under tightly controlled conditions, typically in very small quantities. Once the material moves into commercial production, exposures can occur potentially during synthesis of the material or in downstream activities such as recovery, packaging, transport and storage. In these circumstances, the quantities of materials being handled will typically be much larger. Nanomaterials may also be incorporated, for example, into a composite material, which may be subsequently re-engineered or reprocessed by cutting, sawing, or finishing. Again in these circumstances the potential for exposure exists. Finally, end of life scenarios can be considered, where the material is disposed off, perhaps by incineration or some other process such as shredding or grinding. Again in these circumstances the potential for exposure to those carrying out these procedures does exist. Discharge of materials into the environment is feasible as waste or industrial pollution, directly into the air or water systems or due to deliberate release in applications such as remediation of contaminated lands. Therefore, humans may become exposed as a result of nanomaterial contamination in air, water or the food chain, or through the use of consumer products containing nanomaterials. In considering human exposure for all of these scenarios, it is necessary to consider the route of entry into the human body. In occupational exposure most

Overview of Nanoscience in the Environment

21

emphasis has (rightly) been placed on exposure by inhalation. More recently, however, emphasis has been put on dermal exposure and exposure by ingestion. For nanomaterials, given their mobility and potential for translocation, it is highly appropriate to consider these other routes. The various exposure scenarios are discussed in detail in Chapter 8.

1.15

Detection and Characterization of Nanomaterials

To assess the risk posed by nanomaterials, it is essential to be able to detect, measure and characterize them in different media (air, soil and water and toxicity test media). Properties that are important for the characterization of nanomaterials include, but are not limited to, concentration, size and size distribution, molar mass, surface area, state of dispersion/agglomeration, composition, structure, surface charge, oxidation state, solubility, reactivity and stability (Powers et al., 2006, 2007). Figure 1.5 shows an example of such a characterization of iron oxide nanoparticles (Baalousha et al., 2008b), which shows that the characterization of nanoparticles is an extensive laborious process, demanding the use of several techniques in parallel in order to achieve a high degree of accuracy and reliability. To date few quantitative analytical tools for measuring nanomaterials in natural systems are available, which results in a serious lack of information about their occurrence in the environment (Nowack and Bucheli, 2007). There are many challenges for the detection and characterization of nanomaterials, including their small size and the need to differentiate the material of interest from those similarly sized natural materials, the need for sensitive and specific techniques to measure the required metrics (size, number and surface area), the need to measure nanomaterials properties in several media and the need to measure several properties in parallel. All current measuring methods and techniques fall some way short of addressing these requirements. In addition, the diversity of nanomaterials and their properties make their identification and characterization a difficult task. Furthermore, the interaction of nanomaterials with the natural environmental or biological components provides an additional complexity to the system and so a significant analytical challenge. The challenges, methods and tools to characterize nanomaterials is discussed in detail in Chapter 6.

1.16

Issues to be Addressed

The risks and impacts of nanomaterials in the environment and for human and ecological health are still poorly understood and extensive further work is required. Some priorities to be addressed are given below. 1.16.1

Nomenclature

Formalized, standard terminology for nanotechnology is only starting to become available but is necessary to make progress. Nanotechnology has been defined in

22

Environmental and Human Health Impacts of Nanotechnology

200 nm (b) 14

12

12

10

Relative amount (%)

Relative amount (%)

(a)

10 8 6 4 2 0 5

10

4 2 0

15

Equivelent circular diameter (nm)

(c)

Concentration (ppm)

2400 1600 800 0 0

5

10

15

Hydrodynamic diameter (nm)

20

5

10

15

20

AFM height (nm)

(d)

3200 UV response (a.u.)

6

0

0

(e)

8

80 70 60 50 40 30 20 10 0 pH 2

25

(f)

pH 3

pH 4

pH 5

pH

Figure 1.5 Characterization of iron oxide nanoparticles (200 mg L−1 Fe) reference solution: (a) transmission electronic microscopy (TEM) micrograph; (b) atomic force microscopy (AFM) micrograph; (c) equivalent circular diameter histogram calculated from TEM micrograph; (d) particles height histogram measured by AFM; (e) hydrodynamic diameter for iron oxide nanoparticles at pH 2 in the absence of humic acid measured by flow-field flow fractionation; (f ) Concentration of dissolved iron in the iron oxide suspension at variable pH values. (M. Baalousha, A. Manciulea, S. Cumberland et al., Aggregation and surface properties of iron oxide nanoparticles; influence of pH and natural organic matter, Environmental Toxicology & Chemistry, 2008, 27, 1875–82. Reproduced with permission from Allen Press Publishing Services.)

Overview of Nanoscience in the Environment

23

different ways, but generally nanomaterials are defined as particles smaller than 100 nm in any one dimension. As has been seen, and as is more thoroughly discussed in the next two chapters, sizes below about 10–20 nm are of particular importance, as this is the size range in which properties change markedly. To resolve such confusion national and international standard bodies, such as the International Organization for Standardization (ISO), International Electrotechnical Commission (IEC), American National Standards Institute (ANSI), ASTM International, British Standards Institution (BSI) and others are now discussing the standardization of terminology, metrology, characterization and approaches to safety and health. 1.16.2

Future Development and Risk

Current risk assessment programs are mainly concerned with the first generation of nanomaterials or passive nanomaterials, and even here there is fundamental uncertainty arising from lack of knowledge. Further development of nanomaterials will involve larger and more complex phenomena and problems and much work remains to be done. 1.16.3

Dosimetry

Exposure and doses of nanomaterials are generally measured in terms of mass per unit volume, commonly milligrams per litre. However, there is a consensus that mass dose alone is insufficient to characterize the exposure to nanomaterials. Some studies have indicated a correlation between toxicity and particle surface area, suggesting that surface area is a better metric for measuring exposure (Oberdorster et al., 2005a, 2005b), while particle number is also an important metric. Although mass can be correlated to the surface area, large variations can occur in this correlation within different batches or due to differences in shape. Until particle structure and toxicity are quantitatively related, size, surface area and mass and number concentrations and other parameters need to be specified and measured (Oberdorster et al., 2005b). 1.16.4

Methods of Detection and Characterization

Determining the concentration and physical and chemical properties of nanomaterials in environmental and biological systems is essential. The development and testing of such methodologies is at an early stage and no published data are available. Nevertheless, this development is absolutely central to our ability to understand and predict the environmental consequences on nanotechnology. 1.16.5

Environmental Fate of Nanomaterials and their (Eco)Toxicology

Little is known in this area and there is a large range of issues to be addressed. These issues strongly overlap with the fate and behaviour of natural and incidental nanoparticles and are (non-exhaustively) summarised below: • identifying the potential sources of release and concentration of nanomaterials to the environment,

24

Environmental and Human Health Impacts of Nanotechnology

• identifying the physical and chemical factors that will influence transport and deposition of nanomaterials in the environment (air, soil and water), • determining the influence of redox and other chemical processes on nanomaterials properties and behaviour, • investigating the interaction of nanomaterials with natural environmental components, • investigating the effect of aggregation/disaggregation, sorption/desorption, the distribution of nanomaterials in the environment, • investigating the potential transformation and degradation of nanomaterials in the environment and the resulting by-products, • investigating the potential bioaccumulation/biomagnification and persistence of nanomaterials, • assessing the applicability of knowledge, data and methods on natural ultrafine particles and colloids to nanomaterials, • investigating decontamination strategies.

1.17

Conclusion

Nanomaterials have always been present in our environment, though the nanoscience and nanotechnology of manipulating matter at the nano or atomic scale is a relatively new science. Currently, the uses and range of applications of nanomaterials are expanding rapidly. Their benefits are expected to be huge, but uncertainty about the perceived risk is also a major problem both to human and environmental health and the sustainability of the nanotechnology industry. Understanding the environmental and human health impacts of nanomaterials is a crucial stage for the responsible development nanotechnology and to gain full benefit of its applications. This is a multi- and interdisciplinary task requiring knowledge of physics, environmental, analytical and physical chemistry, materials science, (eco)toxicology and others. This book is one attempt amongst many to improve our understanding of the environmental and human implications of nanotechnology.

1.18

References

Alyaudtin, R. N., A. Reichel, R. Löbenberg, et al. (2001) Interaction of poly(butylcyanoacrylate) nanoparticles with the blood–brain barrier in vivo and in vitro: J. Drug Target, 9, 209–21. Baalousha, M. (2009) Aggregation and disaggregation of iron oxide nanoparticles: influence of particles concentration, pH and natural organic matter: Sci. Tot. Environ., 407 (6), 2093–2101. Baalousha, M., N.-A. Alexa, E. Cieslak and J. R. Lead (2008a) Transport mechanisms of carbon nanotubes in the natural aquatic environment: Environmental Science & Technology (Submitted). Baalousha, M., A. Manciulea, S. Cumberland, et al. (2008b) Aggregation and surface properties of iron oxide nanoparticles; influence of pH and natural organic matter: Environ. Toxicol. Chem., 27, 1875–82. Ballesteros, E., M. Gallego and M. Valcarcel (2000) Analytical potential of fullerene as adsorbent for organic and organometallic compounds from aqueous solutions: J. Chromatography A, 869 (1–2), 101–110.

Overview of Nanoscience in the Environment

25

Barsan, N., M. Schweizer-Berberich and W. Göpel (1999) Fundamental and practical aspects in the design of nanoscaled SnO2 gas sensors: a status report: Fres. J. Anal. Chem., 365, 287–304. Binnig, G., H. Rohrer, Ch. Gerber and E. Weibel (1982) Surface Studies by Scanning Tunneling Microscopy: Phys. Rev. Lett., 49, 57–61. Binnig, G., C. F. Quate and C. Gerber (1986) Atomic Force Microscope: Phys. Rev. Lett., 56, 930–3. Bogunia-Kubik, K. and M. Sugisaka (2002) From molecular biology to nanotechnology and nanomedicine: Biosys., 65 (2–3), 123–38. Borm, P. J. A., et al. (2006a) The potential risks of nanomaterials: a review carried out for ECETOC: Part. Fibre Toxicol., 3, 1–35. Borm, P., F. C. Klaessig, T. D. Landry, et al. (2006b) Research Strategies for Safety Evaluation of Nanomaterials, Part V: Role of Dissolution in Biological Fate and Effects of Nanoscale Particles: Toxicological Sciences, 90 (1), 23–32. Brant, J., H. Lecaotnet and M. R. Wiessner (2005) Aggregation and deposition characteristics of fullerene nanoparticles in aqueous systems: J. Nanoparticle Res., 7, 533–45. BSI (2007) Terminology for nanomaterials. PAS 136:2007, British Standards Institution. Buffle, J. (2006) The key role of environmental colloids/nanoparticles for the sustainablility of life: Environ. Chem., 3, 155–8. Chen, K. L. and M. Elimelech (2006) Aggregation and Deposition Kinetics of Fullerene (C60) Nanoparticles: Langmuir, 22 (26), 10994–1001. Chen, K. L. and M. Elimelech (2007) Influence of humic acid on the aggregation kinetics of fullerene (C60) nanoparticles in monovalent and divalent electrolyte solutions: J. Colloid Interf. Sci., 309 (1), 126–34. Cheng, X., A. T. Kan and M. B. Tomson (2004) Naphthalene Adsorption and Desorption from Aqueous C60 Fullerene: J. Chemical & Engineering Data, 49 (3), 675–83. Cheng, X., A. T. Kan and M. B. Tomson (2005) Study of C60 transport in porous media and the effect of sorbed C60 on naphthalene transport: Mat. Res. Soc., 20, 3244–54. Chithrani, B. D., A. A. Ghazani and W. C. W. Chan (2006) Determining the Size and Shape Dependence of Gold Nanoparticle Uptake into Mammalian Cells: Nano Letters, 6 (4), 662–8. Choi, O. and Z. Hu (2008) Size Dependent and Reactive Oxygen Species Related Nanosilver Toxicity to Nitrifying Bacteria: Environmental Science & Technology, 42 (12), 4583–8. Clarke, A. G., L. A. Robertson, R. S. Hamilton and B. Gorbunov (2004) A Lagrangian model of the evolution of the particulate size distribution of vehicular emissions: Sci. Tot. Environ., 334–335, 197–206. Cloninger, M. J. (2002) Biological applications of dendrimers: Cur. Opin. Chem. Biol., 6 (6), 742–8. Cornell, B., V. BraachMaksvytis, L. King, et al. (1997) A biosensor that uses ion-channel switches: Nature., 387, 580–3. Cuenca, A. G., H. Jiang, S. N. Hochwald, et al. (2006) Emerging implications of nanotechnology on cancer diagnostics and therapeutics: Cancer, 107, 459–66. Cui, Z. and R. J. Mumper (2003) Microparticles and nanoparticles as delivery systems for DNA vaccines: Crit. Rev. Ther. Drug Carrier Syst., 20, 103–37. Daniel, M. C. and D. Astruc (2004) Gold Nanoparticles: Assembly, Supramolecular Chemistry, Quantum-Size-Related Properties, and Applications toward Biology, Catalysis, and Nanotechnology: Chemical Reviews, 104 (1), 293–46. Diallo, M. S., S. Christie, P. Swaminathan, et al. (2005) Dendrimer Enhanced Ultrafiltration. 1. Recovery of Cu(II) from Aqueous Solutions Using PAMAM Dendrimers with Ethylene Diamine Core and Terminal NH2 Groups: Environmental Science and Technology, 39 (5), 1366–77. Donaldson, K., R. Aitken, L. Tran, et al. (2006) Carbon Nanotubes: A Review of Their Properties in Relation to Pulmonary Toxicology and Workplace Safety: Toxicological Sciences, 92 (1), 5–22.

26

Environmental and Human Health Impacts of Nanotechnology

Dror, I., D. Baram and B. Berkowitz (2005) Use of nanosized catalysts for transformation of chloro–organic pollutants: Environmental Science & Technology, 39, 1283–90. Dunford, R., A. Salinaro, L. Ai, et al. (1997) Chemical oxidation and DNA damage catalysed by inorganic sunscreen ingredients: FEBS Lett., 418 (1–2), 87–90. Faraday, M. (1857) The Bakerian Lecture: Experimental Relations of Gold (and Other Metals) to Light: Philos. Trans. R. Soc., 147, 145–81. Feynman, R. P. (1960) There’s plenty of room at the bottom: Eng. Sci. Magazine XXIII. California. Feynman, R. P. (1992) There’s plenty of room at the bottom: J. Microelectrochem. Sys, 1, 60–6. Freitas, R. A. (1998) Exploratory design in medical nanotechnology: a mechanical artificial red cell: Artif Cells Blood Substit Immobil Biotechnol., 26, 411–30. Germanenko, I. N., S. Li and M. S. El Shall (2001) Decay Dynamics and Quenching of Photoluminescence from Silicon Nanocrystals by Aromatic Nitro Compounds: J. Physical Chemistry B, 105 (1), 59–66. Giasuddin, A. B. M., S. R. Kanel, et al. (2007) Adsorption of Humic Acid onto Nanoscale Zerovalent Iron and Its Effect on Arsenic Removal: Environmental Science & Technology, 41, 2022–7. Gidhagen, L., C. Johansson, G. Omstedt, et al. (2004) Model Simulations of NOx and Ultrafine Particles Close to a Swedish Highway, Environmental Science & Technology, 38 (24), 6730–40. Glenn, J. C. (2006) Nanotechnology: Future military environmental health considerations: Technol. Forecast. Soc. Change, 73 (2), 128–17. Graham, T. (1861) Liquid Diffusion Applied to Analysis: Philos. Trans. R. Soc., 151, 183–224. Handy, R. D., F. von der Kammer, J. R. Lead, et al. (2008) The ecotoxicology and chemistry of manufactured nanoparticles: Ecotoxicol., 4. Hardman, R. (2006) A toxicologic review of quantum dots: toxicity depends on physicochemical and environmental factors: Environ. Health Perspec., 114, 165–72. Harrison, R. (2007) Nanoparticles in the atmosphere, in Nanotechnology: consequences for human health and the environment (eds R.E. Hester and R. Harrison): RSC Publishing, Cambridge, 35–49. Hochella, M. F. (2002) There’s plenty of room at the bottom: nanoscience in geochemistry: Geochim. Cosmochim. Acta, 66 (5), 735–43. Hurt, R. H., M. Monthioux and A. Kane (2006) Toxicology of carbon nanomaterials: Status, trends, and perspectives on the special issue: Carbon, 44 (6), 1028–33. Hyung, H., J. D. Fortner, J. B. Hughes and J. H. Kim (2007a) Natural organic matter stabilizes carbon nanotubes in the aqueous phase: Environmental Science & Technology, 41 (1), 179–84. Jung, H., D. B. Kittelson and M. R. Zachariah (2005) The influence of a cerium additive on ultrafine diesel particle emissions and kinetics of oxidation: Combustion and Flame, 142 (3), 276–88. Kamat, P. V., R. Huehn and R. Nicolaescu (2002) A ‘sense and shoot’ approach for photocatalytic degradation of organic contaminants in water: J. Physical Chemistry B, 106 (4), 788–94. Kamat, P. V. and D. Meisel (2003) Nanoscience opportunities in environmental remediation: Compt. Rend. Chim., 6 (8–10), 999–1007. Kanel, S. R., D. Dhriti, M. Bruce and C. Heechul (2007) Transport of surface-modified iron nanoparticle in porous media and application to arsenic(III) remediation: J. Nanopart. Res., 9, 725–35. Kanel, S. R., B. Manning, L. Charlet and H. Choi (2005) Removal of Arsenic(III) from Groundwater by Nanoscale Zero-Valent Iron: Environmental Science & Technology, 39 (5), 1291–8. Kim, Y. C., S. Sasaki, K. Yano, et al. (2001) Photocatalytic sensor for the determination of chemical oxygen demand using flow injection analysis: Anal. Chim. Acta., 432 (1), 59–66.

Overview of Nanoscience in the Environment

27

Klabunde, K. J., J. Stark, O. Koper, et al. (1996) Nanocrystals as Stoichiometric Reagents with Unique Surface Chemistry: J. Physical Chemistry, 100 (30), 12142–53. Kohler, A. R., C. Som, A. Helland and F. Gottschalk (2007) Studying the potential release of carbon nanotubes throughout the application life cycle: J. Clean. Prod., In Press, Corrected Proof. Kreuter, J. (2001) Nanoparticulate systems for brain delivery of drugs: Adv. Drug. Deliv. Rev., 47 (1), 65–81. Lam, C. W., J. T. James, R. McCluskey, et al. (2006) A Review of Carbon Nanotube Toxicity and Assessment of Potential Occupational and Environmental Health Risks: Crit. Rev. Toxicol., 36 (3), 189–217. Lead, J. R., J. Hamilton-Taylor, W. Davison and M. Harper (1999) Trace metal sorption by natural particles and coarse colloids: Geochim. Cosmochim. Acta., 63 (11–12), 1661–70. Lead, J. R. and K. J. Wilkinson (2006) Aquatic colloids and nanoparticles: current knowledge and future trends: Environ. Chem., 3, 159–71. Lecoanet, H. F., J. Y. Bottero and M. R. Wiesner (2004) Laboratory Assessment of the Mobility of Nanomaterials in Porous Media: Environmental Science & Technology, 38 (19), 5164–9. Lecoanet, H. F. and M. R. Wiesner (2004) Velocity effects on fullerene and oxide nanoparticle deposition in porous media: Environmental Science & Technology, 38, 4377–82. Lee, M. H., K. Cho, A. P. Shah and P. Biswas (2005) Nanostructured Sorbents for Capture of Cadmium Species in Combustion Environments: Environmental Science & Technology, 39 (21), 8481–9. Li, X.-Q., D. W. Elliott and W.-X. Zhang (2006) Zero-Valent Iron Nanoparticles for Abatement of Environmental Pollutants: Materials and Engineering Aspects: Crit. Rev. Solid State Mater. Sci., 31 (4), 111–22. Liang, P., Y. Liu and L. Guo (2005) Determination of trace rare earth elements by inductively coupled plasma atomic emission spectrometry after preconcentration with multiwalled carbon nanotubes: Spectrochim. Acta Part B: Atom. Spectroscop., 60 (1), 125–29. Liang, P., Q. Ding and F. Song (2006) Application of multiwalled carbon nanotubes as solid phase extraction sorbent for preconcentration of trace copper in water samples: J. Sep. Sci., 28, 2339–2343. Limbach, L. K., Y. Li, R. N. Grass, et al. (2005) Oxide Nanoparticle Uptake in Human Lung Fibroblasts: Effects of Particle Size, Agglomeration, and Diffusion at Low Concentrations: Environmental Science & Technology, 39 (23), 9370–6. Liu, M. and J. M. J. Fréchet (1999) Designing dendrimers for drug delivery: Pharmac. Sci. Technol. Today, 2 (10), 393–401. Lockman, P. R., J. M. Koziara, R. J. Mumper and D. D. Allen (2004) Nanoparticle surface charges alter blood-brain barrier integrity and permeability: J. Drug. Target., 12, 635–41. Long, R. Q. and R. T. Yang (2001) Carbon Nanotubes as Superior Sorbent for Dioxin Removal: Journal of the American Chemical Society, 123 (9), 2058–9. Lyven, B., M. Hassellov, D. R. Turner, et al. (2003) Competition between iron- and carbonbased colloidal carriers for trace metals in a freshwater assessed using flow field-flow fractionation coupled to ICPMS: Geochim. Cosmochim. Acta., 67 (20), 3791–802. Madden, A. S. and J. Hochella (2005) A test of geochemical reactivity as a function of mineral size: Manganese oxidation promoted by hematite nanoparticles: Geochim. Cosmochim. Acta., 69 (2), 389–98. Mattigod, S. V., G. E. Fryxell, R. Skaggs and K. E. Parker (2006) Functionalized nanoporous ceramic sorbents for removal of mercury and other contaminants, Technical Proceedings of the 2006 NSTI Nanotechnology Conference and Trade Show: Nano Science and Technology Institute, 1, 355–7 Miyaki, H., S. Adachi, K. Suda and Y. Kojima (2000) Water recycling by floating media filtration and nanofiltration at a soft drink factory: Desalination, 131 (1–3), 47–53. Navarro, E., A. Baun, R. Behra, et al. (2008) Environmental behavior and ecotoxicity of engineered nanoparticles to algae, plants and fungi: Ecotoxicol., 17, 372–86.

28

Environmental and Human Health Impacts of Nanotechnology

Nitta, T., M. Hata, S. Gotoh, et al. (2003) Size-selective loosening of the blood-brain barrier in claudin-5-deficient mice: J. Cell Biology, 161 (3), 653–60. NNI (2004) National nanotechnology initiative strategic plan, www.nano.gov. Nowack, B. and T. D. Bucheli (2007) Occurrence, behavior and effects of nanoparticles in the environment: Environment. Pollut., 150, 5–22. Nozik, A. J. (2002) Quantum dot solar cells: Phys. E: Low-dimensio. Sys. Nanostruct., 14 (1–2), 115–20. Obare, S. O. and G. J. Meyer (2005) Nanostructured Materials for Environmental Remediation of Organic Contaminants in Water: J. Environ. Sci. Health, Part A, 39 (10), 2549–82. Oberdorster, E. (2004) Manufactured nanomaterials (Fullerenes, C60) nnduce oxidative stress in the brain of Juvenile Largemouth Bass: Environ. Health. Perspec., 112, 1058–62. Oberdorster, G., E. Oberdörster and J. Oberdörster (2005a) Nanotoxicology: an emerging discipline evolving from studies of ultrafine particles supplemental web sections: Environ. Health. Perspec., 113, 823–39. Oberdorster, G., et al. (2005b) Principles for characterizing the potential human health effects from exposure to nanomaterials: elements of a screening strategy: Part. Fibre Toxicol., 2 (1), 8. Oberdrster, G., V. Vicki Stone and K. Donaldson (2007) Toxicology of nanoparticles: A historical perspective: Nanotoxicol., 1, 2–25. Ouali, L., and E. Pefferkorn (1994b) Fragmentation of Colloidal Aggregates Induced by Polymer Adsorption: J. Colloid and Interface Science, 168 (2), 315–22. Park, K. (2007) Nanotechnology: What it can do for drug delivery: J. Controlled Release, 120 (1–2), 1–3. PEN (2005). The project on emerging nanotechnologies, www.nanotechproject.org. Pickering, K. D. and M. R. Wiesner (2005) Fullerol-Sensitized Production of Reactive Oxygen Species in Aqueous Solution: Environmental Science & Technology, 39 (5), 1359–65. Pitoniak, E., C. Y. Wu, D. W. Mazyck, et al. (2005) Adsorption enhancement mechanisms of silica-titania nanocomposites for elemental mercury vapor removal: Environmental Science & Technology, 39 (5), 1269–74. Poland, C. A., et al. (2008) Carbon nanotubes introduced into the abdominal cavity of mice show asbestos-like pathogenicity in a pilot study: Nat Nano, 3 (7), 423–8. Powers, K. W., S. C. Brown, B. Krishna, et al. (2006) Research strategies for safety evaluation of nanomaterials. Part VI. characterization of nanoscale particles for toxicological evaluation: Toxicological Sciences, 90 (2), 296–303. Powers, K. W., M. Palazuelos, B. M. Moudgil and S. M. Roberts (2007) Characterization of the size, shape, and state of dispersion of nanoparticles for toxicological studies: Nanotoxicology, 1 (1), 42–51. Renn, O. and M. C. Roco (2006) Nanotechnology and the need for risk governance: J. Nanopart. Res., 8, 153–91. Rickerby, D. G. and M. Morrison (2007) Nanotechnology and the environment: A European perspective: Sci. Technol. Adv. Mat., 8, 1–2, 19–24. Roco, M. C. (2003) Nanotechnology: convergence with modern biology and medicine: Cur. Opin. Biotechnol., 14 (3), 337–46. Roco, M. C. (2005) International Perspective on Government Nanotechnology Funding in 2005: J. Nanopart. Res., 7, 707–12. Roco, M. C. and W. S. Bainbridge (2003) Converging technologies for improving human performance: nanotechnology, biotechnology, information technology and cognitive science: Kluwer Academic Publishers, New York. Royal Society and Royal Academy of Engineering (2004) Nanoscience and nanotechnologies: opportunities and uncertainties, London. Sberveglieri, G. (1995) Recent developments in semiconducting thin-film gas sensors: Sens. Act. B: Chem. l, 23 (2–3), 103–9.

Overview of Nanoscience in the Environment

29

Schrick, B., B. W. Hydutsky, J. L. Blough and T. E. Mallouk (2004) Delivery Vehicles for Zerovalent Metal Nanoparticles in Soil and Groundwater: Chemistry of Materials, 16 (11), 2187–93. Schwarze, P. E., J. Ovrevik, R. B. Hetland, et al. (2007) Importance of size and composition of particles for effects on cells in vitro: Inhal. Toxicol., 19, 17–22. Sohn, H., S. Letant, M. J. Sailor and W. C. Trogler (2000) Detection of fluorophosphonate chemical warfare agents by catalytic hydrolysis with a porous silicon interferometer: Journal of the American Chemical Society, 122, 5399–400. Thayer, A. M. (2006) Investing In Nanotechnology: Chem. Eng. News, 24 February, http:// acsinfo.acs.org/cen/nanofocus/top/84/8409investing.html?print. Tinkle, S. S., J. M. Antonini, B. A. Rich, et al. (2003) Skin as a route of exposure and sensitization in chronic beryllium disease: Environ. Health Perspec., 111, 1202–8. Tratnyek, P. G. and R. L. Johnson (2006) Nanotechnologies for environmental cleanup: Nano Today, 1 (2), 44–8. Vaseashta, A. and D. Dimova-Malinovska (2005) Nanostructured and nanoscale devices, sensors and detectors: Sci. Technol. Adv. Mat., 6 (3–4), 312–8. Vikesland, P. J., A. M. Heathcock, R. L. Rebodos and K. E. Makus (2007) Particle Size and Aggregation Effects on Magnetite Reactivity toward Carbon Tetrachloride: Environmental Science & Technology, 41 (15), 5277–83. Walker, H. W. and M. M. Bob (2001) Stability of particle flocs upon addition of natural organic matter under quiescent conditions: Wat. Res., 35 (4), 875–82. Wang, L., J. Mao, G. H. Zhang and M. J. Tu (2007) Nano-cerium-element-doped titanium dioxide induces apoptosis of Bel 7402 human hepatoma cells in the presence of visible light: World J. Gastroenterol., 13, 4011–4. Wu, L., H. J. Wiesmann, A. R. Moodenbaugh, et al. (2004) Oxidation state and lattice expansion of CeO2-x nanoparticles as a function of particle size: Phys. Rev. B., 69, 125415-1–9. Xu, Y. and D. Zhao (2006) Removal of Lead from Contaminated Soils Using Poly (amidoamine) Dendrimers: Ind. Eng. Chem. Res., 45 (5), 1758–65. Yang, F., R. Murugan, S. Ramakrishna, et al. (2004) Fabrication of nano-structured porous PLLA scaffold intended for nerve tissue engineering: Biomat., 25 (10), 1891–900. Yang, G. C. C., H. C. Tu and C. H. Hung (2007) Stability of nanoiron slurries and their transport in the subsurface environment: Sep. Purif. Technol., 58, 166–72. Yih, T. C. and M. Al-Fandi (2006) Engineered nanoparticles as precise drug delivery systems: J. Cell Biochem., 97, 1184–90. Zajtchuk, R. (1999) New technologies in medicine: biotechnology and nanotechnology: Disease-a-Month, 45 (11), 453–95. Zhang, K. M., A. S. Wexler, D. A. Niemeier, et al. (2005) Evolution of particle number distribution near roadways. Part III: Traffic analysis and on-road size resolved particulate emission factors: Atm. Environ., 39 (22), 4155–66. Zhang, W.-S. (2003) Nanoscale iron particles for environmental remediation: an overview: J. Nanopart. Res., 5, 323–32.

2 Nanomaterials: Properties, Preparation and Applications Paul Christian School of Chemistry, University of Manchester, United Kingdom

2.1

Overview

The area of nanoparticle properties and preparation is vast and there are entire journals dedicated to publishing work in this area. In addition, the wide number of potential applications and proposed extensive use and implementation of nanotechnology means that an in-depth analysis of the whole area is beyond the scope of this chapter. Given the broad audience this book is intended for, this chapter aims to provide a comprehensive overview of the area with the purpose of highlighting important issues faced when studying these materials. Review articles have been cited for many of the areas discussed so that the reader may expand their knowledge further by more in-depth reading on areas of specific interest. A holistic approach to the content has been adopted to provide the reader with an overview of the key areas pertinent to the subject regardless of background. To achieve this, what a nanoparticle is and what it is not is discussed first. What properties of nanoparticles are key to their function are then examined and some of the unusual properties these particles may possess are highlighted. The challenges facing nanoparticle production are then examined, taking a closer look at several common methods. The preparation of some common types of nanoparticles is discussed and, finally, the application of these nanoparticles in devices and formulations is reviewed. This approach is designed to aid the full understanding of the function and properties of the nanomaterials in their final applications. Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

32

Environmental and Human Health Impacts of Nanotechnology

2.2

Introduction

There have been several debates over what constitutes a nanomaterial and what defines nanotechnology, as discussed in Chapter 1. However, aside from this, there is no doubt that the area has seen massive growth over the past few decades. The relative percentages of publications containing the key phrases ‘synthesis of nanomaterials’, ‘nanotechnology’ and ‘nanoscience’ are shown in Figures 2.1 and 2.2. Overall these figures show that nanotechnology itself predates the use of the term in the scientific literature, as the ability to prepare commercial nanostructured materials, such as zeolites, dates from 1956 and the interest in preparing sols of nanoparticles dates as far back as Faraday. Figure 2.2 gives some indication regarding the rates of growth and sizes of some high profile areas in nanotechnology. It clearly shows that some of the words being used to describe current technology have only been used in the scientific literature for a few years. The recent increase in interest in nanotechnology and nanomaterials and its tentative application in consumer products has led to the realisation that clear definitions are needed so that communication across the broad range of disciplines involved may be transparent and easily understood. Furthermore, for regulation to be conceived definitions are required so that regulation may be enforced. Currently the use of size as a definition of a nanoparticle is common and follows the similar application of size in the definition of the ultrafine particle in atmospheric science. The latter have been of interest for a number of years, mainly in relation to inhalation exposure in humans and air pollution. However, many of these nanomaterials

30

Percentage of all references

25

Synthesis of nanomaterials (50670) Nanotechnology (5248) Nanoscience (596)

20

15

10

5

0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Year

Figure 2.1 The number of publications with key terms in their title or abstract. (Source; MIMAS Web of Science.)

Nanomaterials: Properties, Preparation and Applications

33

50 45 Nanomachine (80)

40

Nanocomposites (10715) Nanomaterials (4065)

Percentage of all references

35

Nanomedicine (268)

30 25 20 15 10 5 0 1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

Year

Figure 2.2 The relative number of publications describing various high profile areas of nanotechnology. Total number of publications shown in brackets. Clearly the area of nanomedicine is currently seeing rapid growth; however, the area of nanocomposites is the longest lived and largest of those presented here. (Source; MIMAS Web of Science.)

were not purposely produced but formed as a by-product of another process. In addition, those which had been purposely prepared were not prepared in a form that had been optimised for dispersion in liquid media (SCENIHR, 2005). The concern with current developments in nanotechnology is that new particles will be more active, more diverse and may be released into the environment by a wider range of mechanisms than ultrafine particles. A wide range of nanoparticles is formed in nature without the influence of man; iron is stored in nanoparticles (ferritin), humic substances in soils and viruses are all on the nanoscale. The application of nanotechnology by man is also not particularly new. Man has been using nanotechnology for millennia; the ancient Egyptians added gold to glass to give a red colour and recent studies of jewellery show that this colour is due to the formation of nanoparticles of gold trapped in the glass. Similarly, in 1857, Michael Faraday reduced gold salts using white phosphorous to produce gold sols with deep red colours. Although Michael Faraday speculated on the nature of the new gold material he had prepared it was not until the advent of high resolution microscopy that these systems could be studied in detail. Towards the end of the twentieth century the enabling technologies evolved and the common tools of scanning electron microscopy (SEM), transmission electron microscopy (TEM) and scanning probe microscopy (SPM) made imaging at atomic resolutions possible. These are discussed in Chapter 6. In addition, new synthetic approaches such as chemical vapour deposition and fast ion bombardment began to make

34

Environmental and Human Health Impacts of Nanotechnology

preparing nanomaterials with designed structures possible. As a result there is a huge wealth of literature on nanomaterials and nanotechnology, but only a fraction of this has made it as far as commercialisation. There are two general areas of nanotechnology, those relating to nanomachines and those relating to more mechanically passive nanoparticles. Nanomachines are nanoscale devices which incorporate some motion or action as a result of their structure. Often these machines are single molecules and are therefore sometimes referred to as molecular machines. There are several excellent reviews (Liedl et al., 2007; Ozin et al., 2005) on the subject of molecular machines and they may essentially be divided into two types, the biologically derived and the non-biologically derived. The biologically derived examples are often based on protein or DNA molecules or assemblies which switch between two states depending on the environment in which they are placed. Typically, switching may be achieved by the addition of certain molecules to the system. Similarly, the rotary action of ATP has been harnessed to rotate a nickel bar (150 × 750 nm). This elegantly captures the ability of nature to function at this scale, the diameter of the motor itself is about 50 nm (Soong et al., 2000). However, clearly if machines, rather than the components of them, are to be prepared in this manner their overall dimensions will exceed the 100 nm limit of nanotechnology and probably result in a device which is more sensibly measured in microns. In order to push the size of these motors to even smaller scales more simple molecules have been devised. Some are based on shifting a ring along a chain in a class of molecules called rotaxanes. These molecules are, in fact, composed of a ring-shaped molecule with a rod-shaped molecule passing through it (Figure 2.3). In one example (Balzani et al., 2006) the molecule is comprised of a section which will harvest light energy. This energy is dissipated within the molecule chain by the transfer of an electron to a series of acceptors on the chain. As the electron moves through the chain it affects the interaction of the rod section with the ring section, resulting in a shifting of the ring along the rod. The concept of a motor is generally tied to a device for generating rotary motion. This is not an issue in many molecules as the thermal energy at room temperature is sufficient to make various groups on molecules spin on their axis. However, generating motion which is controlled and unidirectional is more challenging. Others have demonstrated how this may be achieved and also how it may be put to work (Pijeret et al., 2005; Vicario et al., 2006). A series of photochemical excitations and relaxation processes results in the rotation of two large groups about a formally double bond. The molecules have been specially designed to resist counter rotation and have a shape which facilitates rotation in one direction. The authors have been able to show that if similar molecules are suspended in a liquid crystalline film, excitation of the motor results in ordered motion of the regular ridges in the film. The magnitude of the process is sufficiently large as to allow a glass bar (5 × 30 µm) to be rotated on the surface. It seems that the mass use of nanotechnology in this form is still some time away. Interestingly the discussion also highlights the issue of where does nanotechnology start and chemistry/biology end? Many of the devices discussed here would be better classed as molecules rather than nanoparticles.

Nanomaterials: Properties, Preparation and Applications Light Harvester

Ring molecule

35

Acceptor 2

Ground state Spacer

Acceptor 1 Stopper Excitation Excited state Relaxation

Ground state

Figure 2.3 A representation of a rotaxane based molecular switch. As the molecule is excited with light an electron is transferred from the light harvester to acceptor 2. It is then transferred to acceptor 1, thereby switching the ring molecule from one position to the other. The switch then returns to the ground sate and the ring returns to its original position around acceptor 2.

Current state-of-the-art from a commercial perspective is focused clearly on the use of particulates rather than devices. This is generally due to the huge constraints in using nano devices under non-laboratory conditions. Whilst this places some framework for the consideration of current materials it does no necessarily exclude the nanomachines in the future and clearly in such a rapidly changing field current proposed applications are being realised increasingly rapidly. However, in order to focus the content of this work, further discussion are restricted to the production, properties and use of nanoparticles. Nanoparticles have been prepared from a wide range of materials (Table 2.1) with a diverse range of morphologies and properties. As has already been discussed they have already found commercial applications and some have been in use for thousands of years. The recent interest in the application of nanotechnology has the potential to apply a much wider range of materials in higher volume to a range of diverse applications and therefore careful application of the technology is needed.

2.3

Nanoparticle Architecture

The generally accepted definition of a nanoparticle is a particle with at least one dimension less than 100 nm (Royal Society & The Royal Academy of Engineering, 2004; NIOSH, 2004). This definition, whilst workable, leaves out some important

36

Environmental and Human Health Impacts of Nanotechnology

Table 2.1 A list of material types used in the search [(material type) and (nano* or quantum) and (synthesis or preparation)]. Some examples are given for common nanomaterials under these classes and the total number of publications referring to these materials in this manner. (Source – MIMAS Web of Science.) Material type

Common examples

Oxide

TiO2 SiO2 CeO2 Total inc. ZnO, iron oxides, etc. Cadmium, zinc, lead, mercury, bismuth, antimony, iron, indium Cadmium, zinc, lead, mercury, bismuth, antimony, iron, indium Cadmium, bismuth, antimony, iron, indium Cadmium, iron, cobalt, indium Silicon, carbon, aluminium, gallium Gallium Gold metal Silver metal, silver sulfide Ag2S Single, double and multiwalled

Sulfide Selenide Telluride Phosphide Nitride Arsenide Gold Silver Carbon nanotube

Number of hits 3161 1926 492 8801 1248 281 104 147 1356 56 4134 2970 4513

details and distinctions. It has its roots in the definition for ultrafine particles (SCENIHR, 2005), which is a term usually reserved for aerosol particles and is discussed in detail in Chapter 5. However, in the case of a nanoparticle it can be difficult to draw distinction between a nanoparticle and a molecule, particularly when some molecules may be considered to have dimensions larger than some materials which would be routinely referred to as particles. A good example of this would be a comparison of the size of a polymer containing 1000 segments with an end-to-end segment length of 0.3 nm in solution, which has a root mean squared (RMS) radius of 30 nm, and a nanoparticle of silver with a diameter of only 5 nm. Attempts to draw distinctions between a nanoparticle and a molecule have recently relied on connectivity and variability in composition. Nanoparticles routinely contain a range of sizes and their exact connectivity is not well defined, whereas most polymer molecules and large clusters, such as polyoxometallates (Long et al., 2007), have a well defined connectivity and in the case of clusters exist in a single size. There has, as yet, been no consensus about how the two may be delineated. With recent advances in nanoparticle preparation and characterisation there has even been a report of a single crystal structure of what would conventionally be considered a nanoparticle. This provides an example of a complete description of a ‘nanoparticle’s’ connectivity for a material which would be conventionally considered a nanoparticle. In some cases the use of the terms associated with colloidal systems are routinely used and are a useful way of describing nanoparticles which have been dispersed in a liquid, but this neglects those which are, for example, embedded in a polymer or form part of an extended array of nanoparticles.

Nanomaterials: Properties, Preparation and Applications

37

Everything

Colloids

1

2

3

Molecules

4 8

5

6

7

Nanoparticles

Elements

Figure 2.4 The interrelation of various types of material in the set of everything. Regions 1–8 relate to: 1 a stable colloid of an inorganic compound; 2 a stable colloid of an element; 3 a stable colloid of an element which has some molecular architecture; 4 a stable colloid of a molecule; 5 an aggregate of an inorganic compound; 6 an aggregate of an element; 7 an aggregate of an element which has some molecular architecture; 8 an aggregate of a molecule.

It is possible to describe nanoparticles as some subset of all material (Figure 2.4) in which there are subsets of elements, molecules, colloids and nanoparticles. This results in eight possible forms of nanoparticles, which can be divided into two sets; those which form stable colloids (1–4) and those which do not (5–8). Examples of these might be: 1 an inorganic compound (cadmium sulfide); 2 a element (gold); 3 an element which has some molecular architecture (sulfur); 4 a molecule (polystyrene). Similar examples could be used to describe sets 5–8 but in this case these materials do not form colloidal dispersions. These materials would therefore be precipitated forms of 1–4. Examples of these might be nanoparticles deposited onto a surface and thereby immobilised (titanium dioxide coated glass), or nanoparticles which have formed massive aggregates. This clearly demonstrates the massive diversity in what might be considered to be nanoparticle. In addition to this, the shape of nanoparticles can be massively diverse. There is a tendency to consider a particulate material to be isotropic, whereas a large range of particle morphologies has been prepared. To date the most important nanoparticles have proven to be the isotropic, the disc shaped and the wire or rod morphology. However, a much wider range of morphologies are known, including tetrapod, tear drop, dumbbell and dendrite structures (Figure 2.5) (Mana et al., 2000; Ung et al., 2007). Whilst there is massive diversity in composition and shape of nanoparticles their architecture has common components which span a range of materials. These are discussed below.

38

Environmental and Human Health Impacts of Nanotechnology Spherical: typical in non-crystalline materials and very small crystalline particles

Geometric solid: typical of crystalline materials. May be a range of shapes inc. cubic, tetrahedral, icosahedral, etc.

Dendritic: Composed of nanoscale wires the dendrite may be much larger than 100 nm.

Rod or wire: May have a range of cross sections including circular, cubic and pentagonal

Discotic: thin flat plate often wider than 100 nm. May be a range of shapes including hexagonal and irregular

Figure 2.5

Tear Drop: An extension of the spherical morphology.

Dumbbell: Formed by the growth of one material only at the ends of a rod of another material

Tetrapod: Formed by the growth hexagonal phase rods from a cubic seed crystal

Some reported morphologies for nanomaterials (see text for references).

3.5

3 SA:Volume Ratio = r

SA/Volume Ratio

3 2.5 2 1.5 1 0.5 0

0

5

10

15 20 Particle Diameter (nm)

25

30

Figure 2.6 The effect of radius on the surface area to volume ratio for a constant mass of material.

2.3.1

Nanoparticle Surface

All nanoparticles have an exceptionally high surface area to volume ratio. This is one of the key features of a nanoparticle and often has a dramatic effect on the particle physical properties as well as being key in understanding its fate and behaviour in real environments. If a simple spherical particle is considered, the surface area per unit mass scales to the inverse of the radius (Figure 2.6). The atoms at the surface of any nanoparticle contribute to a significant proportion of the whole material. The fraction of atoms at the surface of a nanoparticle can be very high. If a spherical gold particle (atomic radius = 0.144 nm) is considered, then a 2.5 nm particle has 53% of the atoms at the surface, at 10 nm this falls to 16%, at 50 nm this falls to 3% and at 100 nm this falls to 2%. Comparing this to a one micron particle at 0.2% and a seven micron particle with only 0.02% clearly shows the

Nanomaterials: Properties, Preparation and Applications

39

importance of surface chemistry when describing and considering nanoparticulate materials. However, the chemistry of the surface of a nanoparticle can also vary significantly from that of the core of the nanoparticle. In some cases this variation in properties allows some nanoparticles to form stable dispersions having a significant effect on the types of interactions the nanoparticle will have with other materials. As such it is erroneous to consider the surface chemistry of any nanoparticle to approximate well to that of the core material. An excellent example of this can be found by examining silica nanoparticles. The particle itself has the approximate molecular formula (SiO2)n where n is large (>20). This implies an infinite lattice, which is not in fact possible at the surface due to the bonding constraints surrounding the geometry of the silicon atoms. These need to form a tetrahedral structure and therefore need to form bonds in three dimensions. This results in a number of terminal Si=O or Si–OH groups on the surface of the particle. The Si–OH groups can ionise to from SiO− and H+ pairs if the particles are suspended in a suitable solvent, and this in turn gives rise to the stability of the particle dispersion. Clearly the formula would be better represented as a core of SiO2 with a shell of SiO(OH)2. This variation between the chemistry of the core and the surface is particularly relevant when considering inorganic (including carbon particles such as Carbon Black (CB), C60 or Carbon nanotubes (CNT)) or metallic nanoparticles. Although the surface chemistry of polymer nanoparticles will also be variable between the core and the surface the reasons for this are different. This means that the surface of nanoparticles prepared from what might initially appear to be an inert material such as silver can rapidly oxidise to give a very thin layer of silver oxide at the particles surface (Schnippering et al., 2007). The atoms at the surface of any inorganic material are not inert because they may lower their energy by binding to other molecules. This energy associated with the surface is called the surface energy. In fact, even bulk inorganic materials have this property and it is therefore routine, in experiments where interactions with the surface are to be studied, to clean the surface with ion bombardment or heating whilst the material is held in a high vacuum. This results in removing any adsorbed molecules for the surface. In the case of nanomaterials their surface is so large they have a massive surface energy. If the surface is not capped with a strongly bound molecule that prevents interactions between particles, the surface energy will be minimised by interactions between the particles themselves. In fact, were it not for the high surface energy of nanoparticles many of them would remain suspended indefinitely in aqueous media at room temperature and pressure. Three forces are acting on any particle suspended in a medium (Figure 2.7). Gravity pulls the particles down causing sedimentation; friction (with a component of buoyancy) works against gravity; molecules of the medium collide with the particles resulting in Brownian motion. The laws and dynamics of these processes have been well understood for over a century and have proven to be well behaved theories which have been applied in many instruments since. The competing effects of gravity and friction are described well by the Stokes law equation, which relates the rate of sedimentation to the square of the particle radius, meaning a larger particle will sediment much more rapidly than a small one. However, the particles

40

Environmental and Human Health Impacts of Nanotechnology

Friction F= 6πηr Rate of sedimentation

dx 2r2(ρ2- ρ)g = dy 9η

x ½

Gravity

Brownian motion

F = m(1-vρ)g

RTt x = 3πηrNA

Figure 2.7 The forces of gravity and friction on a suspended particle (left) and the effect of Brownian motion on a particle’s random walk in suspension as a result of collisions with solvent molecules (right). [r = particle radius, ρ = density of particle, ρ2 = density of fluid, g = acceleration due to gravity, η = viscosity of fluid, m = mass of particle, v = volume of particle, R = gas constant, T = temperature, t = time, NA = Avogadro’s number].

Table 2.2 A comparison of calculated diffusion rates (mm/h) for different particle sizes against the sedimentation rates (mm/h) for some particles of varying density (values given in brackets gcm−3) suspended in water at room temperature. Note: polyethylene floats due to its low density. Osmium given as an example due to its extreme density. Particle diameter (nm) Brownian diffusion Polyethylene (0.96) Polystyrene(1.05) Graphite (2.25) Titania anatase (3.84) Zinc oxide (5.61) Cerium (IV) oxide (7.13) Silver (10.5) Gold (18.9) Osmium (22.5)

1

10

100

500

1.8635 −8.82 × 10−8 1.10 × 10−7 2.76 × 10−6 6.26 × 10−6 1.02 × 10−5 1.35 × 10−5 2.09 × 10−5 3.94 × 10−5 4.74 × 10−5

0.5893 −8.82 × 10−6 1.10 × 10−5 0.0003 0.0006 0.0010 0.0014 0.0021 0.0039 0.0047

0.1863 −0.0009 0.0011 0.0276 0.0626 0.1016 0.1351 0.2094 0.3942 0.4735

0.0833 −0.0220 0.0276 0.6889 1.5652 2.5407 3.3784 5.2357 9.8541 11.8381

are constantly bombarded by molecules of the solvent which have an energy related to the temperature of the system. This means that the particles follow a so-called random walk. Einstein’s law of diffusion shows that the average displacement of a particle along any given axis in time (t) is proportional to the inverse square root of the particle radius. This means that the larger the particle the slower it moves due to Brownian motion. Table 2.2 shows the calculated values for the distances travelled by diffusion and those travelled due to sedimentation for a range of materials in water at room temperature. Thus, materials with densities less than that of the medium in which they are suspended do not sediment under gravity.

Nanomaterials: Properties, Preparation and Applications

41

Furthermore, it can be seen that the distances travelled due to sedimentation are much less than the average random walk due to diffusion. This means that diffusion due to Brownian motion will overcome the sedimentation process. It is also clear that for particles with high densities such as titania or gold, sedimentation will overcome Brownian motion and therefore sedimentation will occur. The implication of this is that at very low concentrations of small particles, sedimentation is not very probable without some other interaction such as attachment to another colloidal particle in suspension. So whilst in principle many very small nanoparticles should form stable dispersions without surface modification, their high surface energy generally results in aggregation, and rapid precipitation. Another reason to consider the surface of a nanoparticle is because the surfaces of nanoparticles are often modified in order to improve their compatibility with the matrix into which they are placed or to improve their dispersion stability. Often the surface of a nanoparticle will be covered with a layer of other molecules that improves the processing properties of the particles. For example, silver nanoparticles are regularly prepared with a layer of citrate ions on the surface. These ions are bound to the surface of the particle but also carry a charge, which therefore imparts a charge to the surface of the particle and results in a stable dispersion of particles once they are dispersed in water. There are essentially two methods for generating stabilised dispersions of nanoparticles. Both methods are the same as those used to generate stable dispersions of much larger colloids and rely on the principle of creating an energy barrier for the close contact of two suspended particles. Here a simplistic description of the dispersed system we will be considered briefly. Chapters 3 and 4 deal in more depth with the chemistry at the surface of the nanoparticle in terms of DVLO theory, zeta potentials and so on in relation to their environmental behaviour. 2.3.2

Charge Stabilisation

It is well known that like charges repel. Therefore, if a charge is placed on the surface of the particle an energy barrier will exist which, if large enough, will prevent aggregation. A particle is prepared so that the surface of the particle has a charge associated with it. This charge may have been deliberately attached to the surface, but in some cases it is possible that a particle may serendipitously attain charge as a result of the adsorption of molecules or ions to its surface. The charge on the surface must be neutralised by a suitable counter ion which will form a strongly associated layer at the surface of the particle. This layer usually consists of both ions and solvent. This strongly associated layer is called the Stern Layer. There is a second so-called diffuse layer beyond the Stern Layer which is less tightly bound. If a single charged point P+ at a distance r from the particle surface and distance r-δr from the Stern Layer is considered (Figure 2.8), then the net force on the point charge P will be the difference between the attraction of the negative surface and the repulsion of the positively charged Stern Layer; this may be calculated using Coulomb’s Law showing that the repulsion will be inversely proportional to both the dielectric constant of the medium and the square of the thickness of the Stern Layer. This is, in fact, a large simplification as it fails to take into account

42

Environmental and Human Health Impacts of Nanotechnology

-

-

- - -

-

-

-

- - - Electrostatic - - repulsion - - -

Particle Stern Layer Diffuse Layer

+ -

+ -

+ -

+ -

+ +

+ + + - - -

-

+ -

+ -

+ P+

δr r

-

+

Figure 2.8 A representation of the stabilisation of a nanoparticle with a charged surface. The Stern layer and diffuse layer are shown in the lower part of the diagram. The electrostatic force between the particles and point P+ will be a function of r–δr.

the diffuse layer or nature of the surface which is clearly not a point charge; the mathematics describing these phenomena are beyond the scope of this chapter. However, even this simplification has two important implications. Colloids based on charge stabilisation will not remain stable if the dielectric constant is raised too high, as might happen at high/low pH or with increasing ionic strength. Furthermore, if a divalent, or greater, ion is added, it will tend to displace the ions at the surface of the particle due to the entropic gain of this process. However, increasing the charge density at the surface of the particle will only serve to make the Stern Layer thinner and therefore reduce the repulsion between the particles. It is for these reasons that the stability of charge stabilised colloids is very dependent on the exact composition of the medium in which they are suspended. Furthermore, increasing the concentration of the particles themselves is likely to result in significant changes to the dielectric constant of the medium due to the increased association of ions with diffuse layer. It is, therefore, not easy to prepare concentrated suspension of charge stabilised colloids. The charge on the surface of these materials generally leads them to form stable dispersions in water. However, they will readily interact with other oppositely charged molecules and surfaces within the medium. The interaction of nanoparticles with natural materials has already been briefly discussed in Chapter 1 and is discussed again in Chapter 4. 2.3.3

Steric Stabilisation

The use of charge stabilisation is viable in an aqueous environment. If particles are to be suspended in a hydrophobic environment, where the dielectric constant is

Nanomaterials: Properties, Preparation and Applications

43

invariably low, then it is necessary to employ a different method of stabilisation. In fact, the use of steric stabilisation may be applied to both hydrophilic and hydrophobic media (Figure 2.9). The principle of steric stabilisation relies on the Gibbs free energy of the solvation of ‘long’ chain molecules bound to the surface of the particle. The energy of dissolving a molecule in a solvent is related to the enthalpy of solvation (∆H) and the entropy of the process, that is how much disorder is created. If a long chain molecule is bound to the surface of a particle with the chain extending into the media, then the chains are surrounded by the solvent and can be considered as being dissolved. If two particles are forced together the processes of aggregation necessitate that solvent be excluded from the region between the particles. If the Gibbs free energy of solvation of the chains is negative then the process will be thermodynamically unfavourable and there will be a barrier to close approach, thereby preventing aggregation. In many conventional colloids this is achieved by using a copolymer which has sections with a high affinity for the particle surface and sections with a high affinity for the solvent. Nanoparticles have been prepared using similar polymers. However, it is more common to use a molecule with a single anchor point and at least one long chain with an affinity for the solvent, such as thiols, phosphines, phosphine oxides, primary amines, pyridines, as well as block copolymers containing these moieties. The key feature with this type of stabilisation is strong binding to the particle surface coupled with a high affinity for the solvent in which the material is to be dispersed. The strong anchoring group ensures that the surfactant remains bound to the particle and the high affinity of

Solvent Excluded

Figure 2.9 A representation of the stabilisation of a nanoparticle with a steric stabiliser attached to the surface. The thermodynamically unfavourable exclusion of solvent from between the particles is shown at the bottom.

44

Environmental and Human Health Impacts of Nanotechnology

the long chain for the solvent provides the stabilisation of the dispersion. Generally, aqueous dispersions of this type are insensitive to mild changes in pH and quite large changes in ionic strength. Furthermore, the addition of polyvalent ions tends to have little effect on the dispersion stability. It is worth noting here the dynamics and properties of micelles and surfactants such as sodium dodecylsulfate (SDS). These molecules form what are termed association colloids. These are ordered aggregates of molecules which possess, for example, a hydrophobic core and a hydrophilic outer layer (Figure 2.10). The range and properties of these systems and their phase behaviour are extremely complex. However, their dynamics in stabilising a dispersion are important and the process of forming micelles is key to the preparation of some polymer based nanoparticles. The formation of a micelle, or the partitioning of a lyophillic surfactant at the surface of a nanoparticle, is driven by the thermodynamics of the system. However, the ability to form a stable dispersion depends on having sufficient material in the media. As a surfactant such as SDS is added to a solvent such as water it will seek to minimise the interaction of the water with the hydrocarbon chain. At low concentrations this will initially result in a layer of the surfactant forming at the air– water interface, thereby decreasing the surface tension of the liquid. As more surfactant is added it will begin to form small (nm) aggregates called micelles which present the charged head group to the aqueous media. Stabilisation of the micelle is therefore based on charge repulsion. The point at which the micelle may form is called the critical micelle concentration (cmc) and represents the minimum amount

OSO3 Na

+

Sodium dodecylsulphate

Hydrophobic hydrocarbon chain Hydrophilic head

Surface Tension

cmc

Micelle

Surfactant concentration

Figure 2.10 The parts of a surfactant (top), the structure of a micelle (bottom left) and the rapid change in surface tension of a solution of the surfactant as the critical micelle concentration (cmc) is exceeded.

Nanomaterials: Properties, Preparation and Applications

45

of surfactant needed to form a stable dispersion. The exact value of the cmc depends strongly on the exact composition of the media and the interactions there in. For example, if the surfactant is being used to disperse two different nanoparticles in the same media, the cmc is likely to differ between the two as the interaction of the aliphatic chain with the particle will differ. Furthermore, unlike most nanoparticles and colloids, association colloids are labile and form part of an equilibrium. Therefore, if a solution of a surfactant is prepared such that its concentration exceeds the cmc and it is then diluted below its cmc the micelles will disperse and no stabilisation of any particles will remain. The whole topic of colloid stability is a large area and for more detailed discussion of the theory of stability in colloidal dispersions the reader is directed to a more specialist text (Shaw, 1992).

2.4

Particle Properties

Nanoparticles may exhibit a range of properties which are either different to those of the bulk material or are a greatly enhanced form of the same properties. It is often these properties which are sought in final products and therefore some consideration regarding these should provide some greater insight into nanoparticle applications. 2.4.1

Surface Plasmon Resonance

Small particles of certain materials, in particular metals, exhibit a particular resonance effect due to the interaction of the incident light with the free electrons in the material. Light is an oscillating electromagnetic wave and, therefore, will interact with the free electrons in a metal. These interactions are present in bulk materials but, because they are essentially a surface effect, become much more pronounced in nanoparticles. In addition to surface chemistry, the shape, size and aggregation state of the particle can also have an effect on the exact profile and position of the absorption process. The wavelength at which the absorption process occurs and the intensity of the absorption depends on the dielectric constant of the particle, the dielectric constant of the surface layer on the particle and the particle size and shape. Generally, the particle size has the greatest contribution to the intensity of the resonance, but particle shape has the greatest effect on the position of the resonance. The resonance can change dramatically upon particle aggregation because the resonance effects may act through space between aggregated particles, thereby resulting in a wide range of energy absorptions. The surface plasmon resonance for silver and gold particles are probably the best known examples. Silver nanoparticles appear a yellow colour in water when capped with citrate and have an absorption maxima at 414 nm, whereas gold nanoparticles generally have a red colour under the same conditions and present an absorption maxima at 550 nm (Figure 2.11). The ability of light to interact with particles in this manner has led to the development of polarisers based on nanoparticles, nanoscale optical wave guides and solution based assays.

46

Environmental and Human Health Impacts of Nanotechnology 1

Silver

0.9

Absorbtion (au)

0.8 0.7 0.6 0.5

Gold

0.4 0.3 0.2 0.1 0 300

350

Figure 2.11

2.4.2

400

450

500 550 Wavelength (nm)

600

650

700

Surface plasmon bands for gold and silver.

Catalysis

As already discussed, nanoparticles have large specific surface areas, which makes them excellent candidates as catalysts. There are two main forms of catalyst, heterogeneous and homogeneous. Many processes are catalysed by heterogeneous catalysts which are insoluble in the reaction medium. Many of these are based on precious metals such as rhodium, platinum or palladium. The determining factor relating to the rate of a reaction in such systems is the diffusion of material to and from the catalyst surface. This issue is not simply related to the surface area of the catalyst but also to the structure of the material. For example, certain processes, such as butane hydrogenolysis on rhodium, show significantly increased activity when the reaction occurs on one face of a crystal compared to the others. Therefore, if the material can be prepared such that it contains a greater proportion of that phase the reaction rate can be increased. Furthermore, if it is a porous structure there will be limitations to the rate of the reaction related to the rates at which material can diffuse into and out of the structure. It is therefore of interest to be able to prepare free floating particles which may have high levels of specific crystal faces and present a very large surface area to the reaction medium. In practice generally the particles are present as a highly open agglomerated mass, which also has the benefit of being easier to retain in the reaction vessel as well as catalysing the reaction. Interestingly, some metals have been shown to have unusual catalytic effects compared to the bulk material. A good example of this is the ability of gold to catalyse reduction of organic materials. It is well documented that the reduction of the dye eosin is catalysed by gold nanoparticles in the presence of sodium borohydride. It has been shown that the ability of gold to participate in the reactions is closely related to particle size and that the reaction rate changes significantly when the particles are less than 10 nm (Sau et al., 2001). Clearly these properties could have a significant impact on the production of fine chemicals and commodity materials.

Nanomaterials: Properties, Preparation and Applications

2.4.3

47

Quantum Confinement

Quantum confinement is an effect which is observed when the electron hole pair, generated through photonic excitation of a semiconductor, is spatially limited due to the small size of the particle. This is an exceptionally complex area and has been reviewed in depth in the academic literature (Bukowski and Simmons, 2002). For the purposes of this book, a simplistic explanation of the effect is considered in order to gain an understanding of how it is related to nanotechnology. The effect of semiconduction can be thought of as resulting from the bonding in a material (Figure 2.12). If two atoms are bonded together there exists a set of bonding and antibonding orbitals. The electrons remain in the bonding orbital and the antibonding orbital remains unoccupied. If there are only two atoms present then the bonding and antibonding orbitals would be termed ‘molecular orbitals’ and have discrete energies. However, if a large lattice is considered then the result of having a whole series of bonds which are all similar but slightly different is a whole range of bonding and antibonding orbitals. In a large lattice there are so many of these bonding and antibonding orbitals that they begin to overlap and form a band, a pseudo-continuum of orbitals. The overall result is a valence band containing all the electrons and a conduction band which is empty. In a semiconductor the process of conduction occurs when sufficient energy is applied to the system in order to promote an electron from the valence band to the conduction band. The energy difference between the valence band and conduction band is called the band gap. In confined system it can be considered that the number of atoms and bonds in the particle is insufficient to provide all of the bonding and antibonding orbitals

Molecular Orbitals

Band Structure

Energy (au)

Conduction Band/ Antibonding Orbitals Band Gap

Valence Band/ Bonding Orbitals Cd1S1

Cd2S2

Cd3S3

Cd∞S∞

Cd500S500

Figure 2.12 The bonding and antibonding orbitals in a theoretical ‘molecule’ of cadmium sulfide (CdS) and the gradual formation of conduction and valence bands as the orbitals overlap with growth of the lattice. The intermediate state of a nanoparticle with a larger than bulk band gap is shown on the right.

48

Environmental and Human Health Impacts of Nanotechnology

for the complete band structure associated with the infinite lattice. The result of this loss of orbitals is that the particle now exhibits a band gap which is larger than that for the bulk material. In the case where the infinite approximation breaks down, the band gap of the semiconductor becomes dependent on the particle size. The overall effect of this is that the nanoparticle may exhibit electronic properties which are different from those of the bulk material from which it was prepared. Often materials which exhibit this sort of confinement are termed quantum dots. Much of the interest in quantum dots lies in their ability to efficiently harness light. It is possible to excite an electron from the valence band into the conduction band by shining light with an energy greater than that of the bandgap of the particle. This results in the promotion of an electron from the valence band to the conduction band and the formation of a positively charged hole in the valence band. These so called excitons rapidly migrate to the surface of the particle and can be used to either generate electricity by separating the pair or generate light by allowing them to recombine; so-called photoluminescence. In a nanoparticle, because the electron hole pairs are confined within the nanoparticle, there is a greatly increased probability that the recombination of the hole and electron will occur as opposed to some other relaxation process. This means that the quantum efficiency, or the fraction of light emitted per excitation photon, will be high. Quantum dots have been shown to have excellent luminescence properties and have narrower emission spectra compared to organic fluorophors. For this reason they are becoming important in fluorescence labelling of cells. Probably the most common examples are those based on cadmium selenide. These materials generally need to be coated with a layer of a second, large band gap semiconductor, in order to ensure that the electron and hole are tightly confined and therefore only relax via emissive recombination. The synthesis and some of the properties of these materials later is discussed later. It should be noted that whilst ideally the electron and hole will behave in a manner expected, there are several other possibilities that may occur depending on the exact structure of the nanoparticle, the nature of the surface species and the medium in which it has been placed. These issues, where relevant, are discussed in greater detail in later chapters. 2.4.4

Mechanical Performance

Carbon nanotubes particularly have received interest as materials with exceptionally high mechanical performance. The measurement of the mechanical properties of nanowires is extremely difficult to conduct and this demonstrates why there is a broad range of values for the mechanical properties of carbon nanotubes reported in the literature (Table 2.3). Some of the first measurements were conducted on multi-walled carbon nanotubes (see below). It was found that the breaking strain varied from 1.4 to 2.9 GPa and the Youngs modulus varied from 18 to 68 GPa. However, these values do not take into account the hollow nature of the fibres. If this is included in the calculation, the values rise from 11 to 63 GPa for strength and 270–950 GPa for modulus. Compared to conventional materials and elements this is an impressive set of values even if the lower figures are considered. (Table 2.3). It therefore seems advantageous to prepare composites from such materials,

Nanomaterials: Properties, Preparation and Applications

49

Table 2.3 Various values for the modulus and strength for a range of materials in comparison to reported values for carbon nanotubes. Material

Density (gcm−3)

Strength (MPa)

Modulus (GPa)

Reference

1.2 2.5 7.7 19.4 16.6 1.6 ∼1.9 ∼1.9

55–70 1 000 340 120 1 240 1 500 68 000 1 400

2.5–3.5 77 200 410 186 130 950 68

a b b b b c d e

Acrylic E-glass (glass fibre) Mild Steel Tungsten Tantalum Carbon Fibre MWCNT (high values) MWCNT (low values)

References: a: Harrison, 1984; b: Thostensona et al., 2001; c: Performance Composites Ltd, 2008; d: Yu et al., 2000; e: Leuenberger and Loss, 2000.

as the properties of a composite are directly related to the properties of the reinforcing fibres and their volume fraction. This also makes carbon nanotubes interesting because they have such low densities compared to conventional materials with similar mechanical properties. However, there are still several technical challenges in translating the superior properties of the carbon nanotubes into a high performance composite (Thostensona et al., 2001). 2.4.5

Magnetic Properties

There is currently a strong drive to develop nanoparticles with magnetic properties for a range of applications including data storage. This need for smaller particles is simply driven by the need to improve the density of data which may be stored on the same area of a disc. Reducing the particle dimensions by half theoretically increases the storage density by four times. Materials which exhibit permanent magnetic properties arise due to the cooperative interactions of electron spins on individual atoms. In order for a material to exhibit permanent magnetic properties, that is it can be magnetised and retains its magnetisation once the field has been removed, the strength of coupling between the spins in the material must be strong enough such that they are not randomised by thermal processes at room temperature. Quantisation of the relaxation process has been observed in molecular species (Leuenberger and Loss, 2000) and it is well known that the barrier to relaxation is related to the dimensions of the nanoparticle as well as its composition (Guardia et al., 2007). The ability to align spins which later relax back into a disordered and therefore non-magnetic state is called superparamagnetism. The field of magnetic materials is very large and too complex to deal with in detail here; however, there are some excellent reviews on the area (Duguet et al., 2006; Xu et al., 2007). 2.4.6

Interfacial Properties

It is well known in colloid science that charge stabilised particles can be used to stabilise emulsions giving the so-called Pickering emulsions (Aveyard et al., 2003;

50

Environmental and Human Health Impacts of Nanotechnology

Binks et al., 2006). The dynamics of the trapping of particles at the interface of two liquids is complex and as yet not fully understood; however, it is seen in many related systems including nanoparticle systems. The overall forces playing a role in the trapping of nanoparticles at the interface appear to be related to the ability of the particles to reduce the surface tension at the liquid–liquid interface and their relatively low diffusion coefficient, which limits the rate at which the particle may diffuse away from the surface. Currently this whole area is being studied to try and understand the factors that affect these processes. It is hoped that a greater understanding in the area will allow the development of better techniques for assembling nanoparticle arrays. One important implication for this effect is the inability to interpret simple partition coefficients from oil–water shake tests, which are often used to determine Kow and are considered important markers for undertaking further tests relating to toxicity (Chapter 10). 2.4.7

Other Properties

There are a whole host of other properties associated with nanomaterials that are not necessarily a result of their nanoscale but are important in the final application of the technology in a situation where the nanoparticles make processing of the final materials possible. A brief look is taken here at some of these, although this section is far from exclusive. 2.4.7.1

Diffusion Barriers

It has been shown that the production of composite materials consisting of exfoliated clay materials results in a massive decrease in the rate of diffusion of gasses (oxygen, nitrogen or carbon dioxide) through a polymer. This is particularly the case when the exfoliated clay composite has been stretched in one or two directions to force the layers of clay to lie in the plane of the composite sheet. The result of this is that in order for a gas molecule to pass directly through the composite it must pass through a layer of the exfoliated clay which is impossible. Therefore, the gas molecules must take a more contorted path through the composite and this has the effect of increasing the effective path length for the diffusing molecule (Figure 2.13). The overall result of this is a material that is much less permeable to gasses – a key property in many packaging materials. 2.4.7.2

Conduction

The ability to prepare concentrated solutions of nanoparticles of metals such as silver has allowed the development of conductive inks. Although conductive paints and inks based on silver have been known for some time, there is a need to optimise the properties of these inks, so that they may be printed using inkjet technology. In particular, it is important that the correct fluid properties are attained by the ink and that the ink remains stable for a reasonable lifetime. This will allow the mass production based on cheaply printed circuits. The particular focus for this work is the mass and cheap production of devices such as mobile phone aerials and RFID (radio frequency identification) tags for rapid and mass identification of goods. It

Nanomaterials: Properties, Preparation and Applications

51

Film thickness

Film thickness

Figure 2.13 The diffusion pathway though a polymer film (top) and polymer nanocomposite film (bottom). The clay nanoplates in the composite film prevent direct diffusion, so forcing the gas molecules to take a more indirect route through the film.

is possible to prepare a conducting film of silver by simple evaporation of an ink or paint to leave a film of nanoparticles in contact with each other. However, to prepare well defined, defect free tracks the coating on the nanoparticles and the solvent in which they are suspended must be removed. This is often done using a low temperature post treatment. The nanoparticles processed in this manner are generally considered to be immobilised on the surface of the substrate. 2.4.7.3

Processing

Perhaps one of the major reasons that nanoparticulate materials have received so much commercial attention is their ability to be processed in novel ways compared to the bulk material. This is particularly the case for nanoparticles of inorganic materials. The particle form lends itself well to dispersion in a range of solvents or to incorporation into a polymer. This means that low temperature processing methods may be used to prepare devices. Furthermore, the ability of an agglomerated film of nanoparticles to form a continuous film with properties similar to that of the bulk material means that nanoparticles may be used to prepare components which would ordinarily rely on vapour deposition technology and lithography. Some examples of how nanoparticle technology could and is replacing other methods of materials manufacture is discussed. Silver ink Conductive tracks for use in antenna and printed circuit board manufacture are typically manufactured using subtractive or additive processing techniques. There are two common ‘subtractive’ methods (methods that remove copper) used for the production of printed circuit boards, silk screen printing and photolithography. The vast majority of printed circuit boards are made using this method by bonding a layer of copper over the entire substrate, sometimes on both sides (creating a ‘blank printed circuit board’), then removing unwanted copper after applying a temporary mask (e.g. by etching), leaving only the desired copper traces. In the ‘additive’ processes a catalytic seed layer is deposited upon a surface and the metal, typically

52

Environmental and Human Health Impacts of Nanotechnology

copper deposited using a ‘wet’ electroless deposition process. Both approaches have the advantage of yielding high conductivity tracks but are not suited to short manufacturing runs or where interconnects between two elements on a surface need to be applied in-line. Inkjet printing of novel materials is now becoming a viable digital manufacturing technique for the additive and/or subtractive printing of conductive elements. One approach attracting much attention is the direct printing of low temperature sintering metal nanoparticles, so removing the need for subtractive or wet electroless deposition processing steps. Copper nanoparticles have to date not been successfully developed since the associated copper oxide is non-conducting. Therefore, most attention has focused upon silver. The use of nanoparticle suspensions allows many of these factors to be overcome. The small particle size alleviates the issues of blockages and therefore the ink may be readily applied using an inkjet printer. The use of nanoparticles has a further advantage in that the particles themselves may be readily sintered to form conductive tracks by the application of heat or microwave irradiation. By careful selection of particle size and coating material the temperatures required to attain a suitable track are low enough for printing to be conducted on cheap plastic films such as polyethylene terephthalate. The ultimate challenge is to attain the appropriate balance of resolution, reproducibility and costs required for mass implementation of disposable radio frequency identification tagging (RFID tags). This technology could revolutionise tracking of goods and mass scanning of multiple items without the need for barcodes. Solar cells via spin coating The major technology used in solar cells is currently based on thin film semiconductor methods. Whilst the so-called dye sensitised solar cell (DSSC) exists its technology is still under development and relies on the use of nanoparticles to function. The development of the dye sensitised solar cell is discussed in more detail later. Conventional solar cells rely on a junction between to materials with different band gaps to firstly absorb photons and then separate the associated charge carriers so that they may be harnessed to do useful work. One issue with the production of solar cells of this type is that they rely on vapour deposition technologies to prepare the thin films. This requires batch processing, high temperatures (>400 °C) and closed and carefully controlled atmospheres. These factors combine to make solar cells expensive, rigid devices with limited production volumes. These restrictions can be traced to the simple requirement of having to prepare thin films of the various components. Other methods exist which allow the preparation of thin films, for example spin coating is routinely used to prepare thin films of polymers. However, the insolubility of common semiconducting materials used in the preparation of solar cells makes this approach unsuitable. The advent of nanotechnology allows the preparation of stable suspensions of nanoparticles that may be processed in a similar manner to a solution of the bulk material. This means that spin coating technology can be used to prepare solar cells directly from nanoparticles in an on-line fashion using flexible materials as the support. The application of this technology has been repeatedly demonstrated for solar cells based both solely on nanoparticles, such as cadmium telluride and cadmium selenide, as well as cells based on semiconducting

Nanomaterials: Properties, Preparation and Applications

53

polymers, such as polythiophenes in conjunction with a nanoparticle such as cadmium selenide. Some of the current challenges lie in being able to prepare such solar cells with high efficiency and using more benign materials.

2.5

Nanoparticle Preparation

There is an ever increasing number of methods and techniques being developed to prepare and manipulate nanoparticles. Whilst some of these methods have been commercialised on a large scale many have yet to be scaled up to mass production. As already seen, nanomaterials are actually complex mixtures. Therefore, having an understanding of how they are prepared can be key to understanding their behaviour. There are several challenges currently facing the community which need to be overcome before mass production of many materials can become a reality and, as will be seen later, the exact quality, properties and amounts of materials required can vary widely from application to application. Therefore, some time is spent at the end of this chapter considering the applications in which nanoparticle have been, are being and may be used. 2.5.1 The Challenges of Nanoparticle Synthesis: Scale Up The so-called bottom-up approach to nanoparticle preparation currently offers the best route to mass produced nanoparticles. This type of preparation method essentially builds nanoparticles up from molecules or atoms. It generally requires low concentrations of nanoparticle to be prepared in order to maintain a narrow size distribution of nanoparticle diameters. Simply increasing the concentration of these reactions generally results in the formation of larger nanoparticles with a wider distribution of sizes. For some applications this is not a significant issue; for others, however, it is critical. 2.5.2

Reactivity

As already discussed, nanoparticles have very large surface areas. This can make them prone to reactions at the surface which modify and degrade the performance of the particles. Many nanomaterials will undergo oxidation or hydrolysis at the surface, resulting in the alteration of the surface chemistry of the nanoparticles. It is well known that the surface chemistry can be key to the properties of the nanoparticle. In fact, in some cases it has been speculated that this oxidation process is the reason why some nanoparticles exhibit the properties which are desired. Perhaps the only exception to this problem is oxide-based nanoparticles, those of noble metals and some polymer nanoparticles. Attempt to circumvent this problem currently focus on developing new capping agents and surfactants and in producing layered nanoparticles with chemically inert coatings. 2.5.3

Dispersability

The ability to form a stable colloidal dispersion of many nanoparticles is key to their processing and incorporation into the final product. In some cases, enhance-

54

Environmental and Human Health Impacts of Nanotechnology

ment of performance can be achieved despite some agglomeration of the nanoparticles. However, this might be improved further if a macroscopically homogeneous dispersion could be prepared. The issue of dispersion is closely related to that of reactivity in that the search for a better capping agent to prevent further reaction at the surface once the nanoparticle has been prepared must also meet the need to produce stable dispersion of nanoparticles. In addition to this, it is advantageous if the dispersions remain stable at a range of concentrations, so that very concentrated dispersion of nanoparticles may be prepared and then processed in order to prepare the final material or device. 2.5.4

Cost

The preparation of some nanoparticles requires the use of very expensive reagents and solvents. This, in many cases, increases the cost and complexity of scale up significantly. The need to find low cost, simple routes to materials is key in scaling up the preparation of certain products. 2.5.5

Methods: Natural Sources

Perhaps the simplest way to obtain a nanomaterial is to use a raw mineral with a minimal amount of post processing. A few commercial nanomaterials are prepared by simple processing of minerals. For example, asbestos has six main forms (Table 2.4), of which chrysotile, amosite and amphibole were the main forms. They are employed in applications along with small amounts of termolite. Asbestos fibres are silicate-based minerals and are either magnesium or iron silicates. They have a highly anisotropic wire-like structure with diameters in the nanometre range and lengths of several microns. It is well known now that asbestos and other fibrous materials with similar length scales are very harmful to human heath (Castleman, 2006) and are no longer in widespread use in the many countries. Another form of nanomaterial which has found commercial application is the so called nano-clays. These are a post processed form of layered silicates such as montmorillonite. These minerals have a layered structure with intercalated metal ions which serve both to balance the charge on the silicate plates and to bind the layers together. These layers may be separated by exchanging the metal ions for

Table 2.4 Forms of asbestos, their common names and formulas. (Other names may be used and can be found in Nolan et al., 1999) Name

Mineral group

Asbestos type

Formula

Chrysotile Actinolite Tremolite Grunerite Anthophyllite Riebeckite

serpentine amphibole amphibole amphibole amphibole amphibole

white

Mg3Si2O5(OH)4 Ca2(Mg,Fe)5Si8O22(OH)2 Ca2Mg5Si8O22(OH)2 (Fe(5–7), Mg(2–0)) Si8O22(OH)2 (Fe2+, Mg)7Si8O22(OH)2 Na2Fe32+(Fe 2+ , Mg )3 Si8O22(OH)2

brown grey blue

Nanomaterials: Properties, Preparation and Applications

Interchelation

55

Polymer mixing

C12H25NH3+Br-

Figure 2.14 Steps in the production of a nanocomposite based on exfoliated clays.

quaternary ammonium ions with long aliphatic chains or by melt processing with a suitable polymer such as nylon (Figure 2.14). This then results in a powder that can be further compounded with a polymer to produce a clay-reinforced nanocomposite that contains the platelets, which are well distributed through the composite. 2.5.6 Top Down Whilst some nanomaterials may be simply mined there are a whole range of materials which have to be prepared in a more laborious manner. There are two main methods for preparing nanomaterials, top down and bottom up. The top down approach works on the basis of breaking down a large piece of material into a smaller piece, in this case with dimensions in the nanometer range. This method can be used on most nanomaterials and is usually related to the patterning of a surface by either lithography and etching, or by electron or fast atom bombardment (Mendes et al., 2004). More recent interest has grown in the ability to use scanning near field pattering methods to chemically alter surfaces at resolutions down to 9 nm using conventional light (Leggett, 2006) or to use atomic force microscopy tips to plough soft polymer films on the nanoscale (Kunze, 2002). All of these methods result in essentially flat patterned surfaces; as these surfaces have some depth they are often termed 2½D techniques. If these techniques are used to produce a nanoscale pattern in a hard material they can then be used to prepare moulds and be further used for nano-imprinting and other related methods for more large scale production of devices. 2.5.7

Bottom Up

The bottom up approach relies on using small molecules to prepare the nanoparticles. For example, it is well known that the addition of a solution containing sulfate ions to a solution containing calcium ions will result in the rapid production of calcium sulfate. Calcium sulfate, being very insoluble in water, will form a precipitate, the solution will turn cloudy and the final product will settle out at the bottom

56

Environmental and Human Health Impacts of Nanotechnology

of the flask. The bottom up approach would use similar chemistry but seek to halt the growth of the particles of calcium sulfate before they became too large, thereby producing a stable dispersion of calcium sulfate where the particle sizes are less than 100 nm. The exact method used can vary enormously as will be seen here but generally the dimensions of the particles are controlled by careful control of the precipitation process. The preparation of nanoparticles by the bottom up approach relies on the principle of supersaturation (Zaizer and Lamer, 1948). In simple terms, at the start of the reaction there is a homogeneous solution of the reagents (Figure 2.15). As the reaction starts the product will begin to form; at very low concentrations it will have some solubility in the medium. As the concentration of the product increases the point of saturation will be reached. At this point the solution can no longer support the ever-increasing amounts of product being formed. However, as the product is being formed very rapidly and precipitation is limited by diffusion of the molecules in the media, which is comparatively slow, the point of saturation may be exceeded, producing a super saturated solution. At some point the supersaturation of the solution is relieved by the formation of the precipitated particles. These particles will be very small and represent the first nuclei of the final particles. It is now more favourable for the formation of more product to occur at the surface of the particle and therefore these first nuclei essentially act as seed for the growth of the particles. This is an idealised example of the processes involved and the occurrence the formation of second batches of seed nuclei is not unusual.

Supersaturation occurs

Concentration of Product

Saturation Concentration

Rapid relief of super saturation

Particle Growth

Rapid increase in product in solution

Time Figure 2.15

The various stages in the formation of an idealised nanoparticle.

Nanomaterials: Properties, Preparation and Applications

57

There is still great debate in the literature over various other mechanisms for the formation of nanoparticles and there are several other processes that may contribute to the growth stage which are not discussed in detail here. The actual kinetic of the growth process may override any thermodynamic sink for the structure of the final material and this can result in particles which contain more than one crystal phase (Christian and O’Brien, 2005, 2008) This simple method for preparing nanoparticles can take many forms and the medium in which supersaturation occurs can be as diverse as a plasma or a conventional solvent. 2.5.7.1

High Temperature Methods

There are several varieties of high temperature methods for the production of nanoparticles. Generally they are well suited to the formation of either metal oxide or metallic nanoparticles and have been used to prepare a range of commercial materials. One excellent method for the preparation of nanoparticles on a large scale is aerosol flame synthesis or spray pyrolysis. The process is relatively simple. A solution is prepared containing the ions required for the synthesis. The solution is then injected into a high temperature flame of up to 3000 K and the particles essentially form in the flame and are collected as they settle out of the atmosphere. Materials prepared by this method tend to be crystalline and well formed. The method has been used for some time for the preparation of nanoparticles of materials such as carbon, titania, silica and alumina at rates of tens of tonnes per hour (Ulrich, 1984). The result of this type of preparation methods is a nanomaterial with an uncoated surface. It is, therefore, not unusual for the particles to be aggregated and difficult to redisperse. There are some excellent reviews on the application of this method to the preparation of metal and metal oxide nanoparticleS with application in catalysis (Wooldridge, 1998; Pratsinis, 1998). 2.5.7.2 Wet Methods There are several variations on wet methods for the preparation of nanoparticles. The term refers to the use of a solvent in which the reaction is performed. There are two subsets of this method: micelle encapsulation and arrested precipitation. Many of these methods can be used with out supplying energy to the system. However, in cases where energy is required to initiate the reaction there is a lot of work investigating the use of light, cavitation, ionising radiation and microwaves instead of thermal heating. Micelle encapsulation Micelle encapsulation relies on the use of a micelle to control the particle size. As already discussed, it is well known that certain surfactants will form micelles where an oil may be dispersed in water. In fact, it is also possible to form inverse micelles, where water is dispersed in a hydrophobic medium. These water-in-oil microemulsions are used as reactors which control the particle growth, size and, in some cases, shape. Two microemulsions are prepared containing the two reagents (Figure 2.16); for example, if cadmium sulfide was to be prepared then one emulsion might contain cadmium chloride and the other sodium sulfide. Upon mixing the micelles

58

Environmental and Human Health Impacts of Nanotechnology Microemulsion of Na2S

Microemulsion of CdCl2

Mixing of Solutions

Nanoparticles of CdS are formed in micelles

Figure 2.16

Preparation of nanoparticles of cadmium sulfide in a microemulsion.

collide, combine and reform very rapidly, resulting in the mixing of the core solutions whist retaining the micelle structure. The formation of the nanoparticle is then constrained by the micelle itself. This type of approach has been applied to both oil-in-water and water-in-oil emulsions as is discussed later. Arrested precipitation There are some similarities between arrested precipitation and micelle encapsulation, notably that both employ a surfactant to control the particle size. However, whilst in the micelle method the surfactant forms structures which contain the nanoparticles, in arrested precipitation the surfactant simply partitions to the surface of the particle as the particle forms. Micelles are very unstable at elevated temperatures (>50 °C) whereas arrested precipitation methods have been applied at temperatures exceeding 300 °C. During an arrested precipitation reaction (Figure 2.17) generally a solution of one reagent is heated to the required temperature and then treated with a solution of the second reagent. It is not uncommon to conduct the whole reaction in the surfactant or capping agent alone with out any further solvent. The final particles are washed to removes some of the excess capping agent. The overall outcome of this is that flame pyrolysis type preparation methods will result in very pure materials, but with no surface coating to ensure the formation

Nanomaterials: Properties, Preparation and Applications

59

Addition of sulfur source

Solution of cadmium source in capping agent

Nanoparticles coated in capping agent

Figure 2.17 The preparation of a nanoparticle dispersion using the arrested precipitation method.

of stable suspensions, where as the micelle and arrested precipitation methods will results in stable dispersions under various conditions but with the added complexity of at least one further component to consider; the capping agent or surfactant. Furthermore, unless care is exercised in the purification of the materials there is likely to be contamination of the particle by either by products from synthesis or unreacted starting materials. Some of the starting materials for certain nanoparticles are exceptionally toxic. It is important to consider that the formation of nanoparticles by the bottom up approach is a dynamic process and that the surface of the nanoparticle cannot be considered to be unreactive. In fact, various processes are known to affect particle form long after the nanoparticles themselves have been prepared. There are two important factors. One is Oswalds ripening, which results in the sacrificial dissolution of small particles in favour of growth of the larger particles with lower surface energies. The rate of such a process is related to the distribution of particle sizes as well as the specific chemistry of the particles themselves. The other important factor is aggregation of the particles to form larger particles. These may become sintered into a polycrystalline larger particle. These processes have much to do with the fate of nanoparticles and are discussed further in Chapter 3. A brief overview of the types of methods used to prepare a range of nanomaterials is given below. These examples are by no means exhaustive. However, they will give an overview of the general methods, reagent and so on that may be used in preparing some nanomaterials. 2.5.8

Metal Nanoparticles

Metal nanoparticles are perhaps the earliest forms of nanoparticles prepared by man. The general method of preparation has changed little and generally relies on the reduction of a dissolved metal salt in the presence of a suitable capping agent or surfactant. The exact method employed depends on the metal. For example, gold

60

Environmental and Human Health Impacts of Nanotechnology

and silver nanoparticles may be prepared by simply heating an aqueous solution of gold chloride or silver nitrate with sodium citrate. This rapidly results in the formation of a red (gold) or yellow (silver) suspension of nanoparticles. This method relies on the binding of the citrate to the surface of the nanoparticle to impart stability and the colloid itself is charge stabilised; the citrate also acts as the reducing agent for the formation of the nanoparticle. The charge stabilisation of the nanoparticles means that these colloids are generally prepared at very low concentrations (<50 ppm of starting salts) and are very unstable in the presence of polyvalent metal ions. A similar result may also be achieved by reducing the metal salts in a more dilute solution of sodium citrate or thiopropanoic acid using a common reducing agent such as sodium borohydride or hydrazine. In the case of thiopropanoic acid, the terminal SH group provides the anchor to the particle surface and the acid group provides a potential point for charge formation. It is well known that alloy nanoparticles of gold and silver may also be prepared in a similar fashion due to the similarity of packing in the two lattices. More recently, silver nanoparticles have also been prepared by the so called polyol route, where a solution of metal ions in an alcohol is reduced simply by heating the solution (Sun et al., 2003) This results in reduction of the metal ions and oxidation of the alcohol. There are a number of other methods that have also been used including using surfactants such as cetyltrimethylammonium bromide (CTAB). Copper nanoparticles have been similarly prepared by the reduction of copper salts in microemulsions (Qiu et al., 1999) and a range of other metal nanoparticles can be prepared in the same manner. 2.5.9

Carbon

There are three forms of nanocarbon that have gained specific interest for applications in nanotechnology: C60, carbon nanotubes (CNT) and more recently graphene sheets. C60 is essentially a soccer ball shaped molecule of carbon. It is known to dissolve in various organic solvents. There have been several reports of dissolving in water (Lyon et al., 2006). However, it seems more likely that the result of these attempts is actually a suspension of a number of C60 units some of which, mainly at the surface, have been chemically altered. C60 or Buckminster fullerene is commonly prepared by the arc discharge method, which is one of the original methods used to prepare C60. During the preparation of the material a high voltage discharge is passed between two graphite rods in a low pressure argon atmosphere. The result is a soot which is collected on a cold finger and comprises up to 15% C60. The soot also contains so-called higher fullerenes such as C70. Carbon nanotubes are essentially hollow tubes, the wall of which is formed from a single layer of graphite which wraps round onto itself. The tubes may form Russian doll like structures with several tubes each inside another. This has resulted in the terms single-walled (SWCNT) and multi-walled (MWCNT) carbon nanotubes. The smallest single walled tubes can, in fact, be less than a nanometre in diameter and several microns long. It is now possible to selectively prepare either single-wall or multi-wall nanotubes. Depending on the exact method of preparation and purification the tubes may be closed at both ends by carbon caps, open at both

Nanomaterials: Properties, Preparation and Applications

Axis of tube

Chiral tube C(5,2)

a1

61

Zig Zag tube C(7,0)

Axis Tube

a2

na1 ma2

Basic Vector system

Armchair Tube C(4,4) Axis of tube

Figure 2.18

2D Graphene sheet

The labelling of different forms of carbon nanotubes using the vector system.

or either end or open at one end and blocked with a metal nanoparticle at the other. The carbon nanotubes themselves may take a range of forms and are usually termed as C(n,m) tubes. The numbers in the name refer to the helicity of the tube (Figure 2.18), two special cases being armchair [4,4] and zigzag [7,0] tubes. This results in a range of properties ranging from conductive to semiconducting properties. There is still a significant challenge in preparing any single form of nanotube by design and therefore most samples contain many different forms. Carbon nanotubes can be prepared by similar methods to those used to prepare C60 by simply varying the pressure of argon in the reaction atmosphere. However, control of carbon nanotube growth had been a particular challenge, along with the drive to find a cheap and scaleable method for their preparation. Recently the use of chemical vapour deposition (CVD) based methods has allowed the preparation of carbon nanotubes of both multi and single walled types by careful deposition of metal particles on a surface. The vapour–liquid–solid (VLS) method allows the preparation of long carbon nanotubes attached to a surface. Typically, a surface is coated with nanoparticles of a suitable metal, such as gold, and a vapour of argon containing a suitable carbon source, such as carbon monoxide or ethylene, is then passed over the substrate whilst it is heated to high temperatures. The metal nanoparticles melt and act as nucleation sites for the formation of the carbon nanotubes. The initial carbon is dissolved in the metal droplets and begins to form a graphite-like film. The film becomes insoluble in the metal droplet and partitions

62

Environmental and Human Health Impacts of Nanotechnology

to the surface. Further growth results in the extrusion of the carbon nanotubes from the metal droplet. Tubes prepared by this method typically have metal droplets attached to the end of the tubes and may also contain metal particles trapped within the tube or intercalated between the layers in multi walled tubes. An unsupported method based on similar principle has recently been reported (Li et al., 2004) where a yarn of nanotubes may be spun from the smoke-like output of a high temperature furnace. In all the cases there is usually some residual formation of carbon particles in the final nanotubes. The trace carbon contamination and the metals may be removed to some extent by treating with strong acids such as nitric acid. This will dissolve most metal particles (except gold), digest amorphous carbon contaminants and may also remove the carbon end caps from the nanotubes. Extended refluxing in nitric acid will eventually also begin to digest the tubes, resulting initially in the formation of carboxylate and hydroxide groups on the surface. Despite rigorous purification steps it is common to find contamination of nanotubes with various metals. 2.5.10

Graphene

Graphene, perhaps the material of most recent interest in nanotechnology, is essentially a single layer of graphite. Currently it is still being studied and has so far shown excellent properties with regards to electron transport, semicondutive properties and quantum effects (Avouris et al., 2007). It may prove to be the next material for microchip manufacture. Currently it is manufactured by molecular beam epitaxy, where a beam of precursor molecules is decomposed on the surface of a suitable substrate. This results in small areas of graphene film. The film itself is one atom thick (0.325 nm) and usually a few microns in diameter. 2.5.11

Carbon Black

Carbon black has been prepared on an on-demand basis for thousands of years in the form of lamp black. However, the first commercial scale production started in 1840 in Pennsylvania, USA. There are four main methods for producing carbon black: the channel black process, the lamp black process, the thermal black method and the furnace method. Each method produces material with slightly different compositions. The channel black process collects carbon nanoparticles as they are produced during the burning of natural gas. The carbon is collected on cooled mild steel plates which have cannels cut into them. This method results in carbon particles with a size of 9–29 nm. The lamp black process uses the burning of shallow ponds of coal tar or aromatic oils to produce the carbon particles, which have sizes of 30–200 nm. The furnace process and thermal processes are similar in that a feed of oil is fed into a hot furnace (>800 °C). The thermal process results in particles with diameters of 150–500 nm and the furnace method 13–100 nm. The exact composition of these materials can vary, especially regarding the polyaromatic hydrocarbon (PAH) content. These are often present due to incomplete reduction of the oil starting materials.

Nanomaterials: Properties, Preparation and Applications

2.5.12 2.5.12.1

63

Inorganic Compounds Oxides

Perhaps the most facile and widely used method for the preparation of metal oxides relies on the flame pyrolysis technique. This has been used to prepare a wide range of materials and is used to prepare the Degussa P45 titania commonly used in many ecotox studies. Such materials may also be prepared by the base catalysed hydrolysis of suitable salts. Silica and titania may be prepared by simple hydrolysis of an alkoxide (tetraethylorthosilicate or titanium isopropoxide) in an aqueous medium. Careful tuning of the pH will result in the formation of a stable dispersion of nanoparticles and this is a common method for the production of some silica nanoparticles. The use of various capping agents has also been employed in these types of reaction with some success. One method which has proven useful for a range of nanoparticles is hydrothermal synthesis. In this method the reaction is conducted in a sealed bomb so that the temperature may be raised above the boiling point of the liquid under standard conditions. This increase in temperature and pressure can result in the formation of different nanoparticle phases and sizes. Nanoparticles of ceria (CeO2) and zinc oxide (ZnO) have been prepared by similar methods. In these cases it is important to add a stabiliser or use a micelle to control the particle size. For example, zinc oxide is readily prepared by the reaction of zinc nitrate with a suitable base (Hartlieb et al., 2007). Ceria nanoparticles have been prepared in a similar manner (Liu et al., 2007). 2.5.12.2

Narrow Band Gap Semiconductors

Narrow band gap semiconductors represent a large range of nanoparticles. These are usually compound of metals with p-block elements such as sulfur, selenium, tellurium or phosphorous. Often these materials contain heavy metals such as cadmium, mercury, lead, indium, antimony or bismuth. Clearly the composition of these materials contains elements which, when in their free ionic form, are known to be exceptionally toxic. Whilst their current use is limited their potential is large and therefore it is worth considering their preparation. The sulfides are perhaps the easiest to prepare and are readily prepared by the rapid reaction of a solution of the metal ions with sodium sulfide in the presence of a capping agent or within a micelle. This method has been used to prepare both particle and rod shaped forms of cadmium sulfide (Simmons et al., 2002). However, the most common method for the preparation of the other chalcogenides and phosphides was initially described in 1991 (Murry et al., 1993). More recently there have been developments of similar methods to prepare metal nitrides such as gallium and aluminium nitride (Wells and Janik, 1996). In this general approach a metal precursor is dissolved in a suitable capping agent such as an alkylphoshine oxide and heated. The second precursor, such as selenium, is dissolved in a phosphine, such as trioctylphosphine, and rapidly injected into the reaction. The resulting nanoparticles are precipitated by the addition of methanol. This can result in

64

Environmental and Human Health Impacts of Nanotechnology

nanoparticles with excellent optical properties. A range of precursor molecules have been used including dialkyl adducts, such as dimethyl cadmium, as well as simple carboxylates. The capping agents used range from phosphines and phosphine oxides to amines, thiols, pyridines and carboxylic acids. Of particular note is the use of phosphine gas in the preparation of metal phosphides such as indium phosphide. Another approach to this preparation is the use of a single molecule precursor. In this case a molecule is prepared which contains all of the atoms needed to prepare the final particle. This molecule is then dissolved in a phosphine and added to a flask of hot phosphine oxide. Clearly, in many of these cases contamination by by-products is likely and high purity hard to attain or measure. Furthermore, whilst most of these nanoparticles are prepared in a form which will not suspend well in aqueous media, the addition of a polymeric surfactant is often enough to facilitate phase transfer from a hydrophobic environment into the aqueous phase. 2.5.13

Polymers

There are essentially two methods for the production of polymer based nanoparticles. In most cases, if a nanoparticle of a pure polymer is required it may most easily be prepared by using a micro emulsion method. For example, a micro emulsion of styrene in water might be prepared by using sodium dodecyl sulfate as the surfactant. A free radical initiator is added to the aqueous phase, for example hydrogen peroxide or ammonium persulfate, and the reaction heated for several hours. It is well known that this type of polymerisation tends to give excellent conversion of the monomer to the polymer. One of the reasons for the high conversion is the reaction kinetics within the micelles. Whilst transport of a radical from the water phase into the emulsion droplets is relatively slow, the reaction within the droplets is very rapid. This often results in an exponential increase in the viscosity of the monomer/polymer phase and tends to trap unreacted polymer ends within the particles. These unreacted polymer ends will still contain reactive radicals which have been used in the past to reinitiate polymerisations. The final nanoparticles will have a proportion of the surfactant bound tightly to their surface because the surfactants chain may either become entangled in the polymer chains or become grafted onto the polymer via side reactions. A second method for preparing polymer nanoparticles is to prepare a block copolymer (Figure 2.19). If a polymer is prepared with two sections that are soluble in different solvents it is possible to force them to self assemble on a molecular scale. As already discussed, polymers are molecules with lengths on the nanometre scale and, therefore, this molecular self assembly invariable results in a nanoparticle being formed. One way to achieve this is to couple to short polymers together, such as polycaprolactone and polyethylene oxide. Polycaprolactone is a water immiscible biodegradable polyester whereas polyethylene oxide is a biocompatible water miscible polyether. By preparing such a copolymer and dispersing it in water a biodegradable polymer nanoparticle may be prepared. In general, the dimensions of polymer nanoparticles are rarely less than 20 nm due to the large volumes which the molecules themselves occupy.

Nanomaterials: Properties, Preparation and Applications

65

Block copolymer

Hydrophilic section

Hydrophobic section

Nanoparticle

Polymersome

Hydrophobic area

Aqueous core

Hydrophilic area

Figure 2.19 A block copolymer for the preparation of a nano-drug delivery system. Particles (bottom left) and polymersomes (bottom right) may be formed depending on composition and processing methods.

Clearly the preparation of nanoparticles is only part of the challenge. Often the next challenge is analysing them. As complex mixtures there are many different factors to consider. Conventionally, particle size and behaviour in various media are routine. Crystal phase, capping material and purity, however, are key factors which cannot be ignored. Various methods for determining these factors are outlined in detail in Chapter 6. As a minimum it is recommended that the following information is critical, although in some cases more information may be required: • • • • •

Particle size and distribution (including surface area). Hydrodynamic size. Crystal phase of the particle. Nature of the capping agent/surface functionalisation (including surface charge). Purity.

2.6 Applications of Nanoparticles and Nanotechnology 2.6.1 The Past The use of nanomaterials is not new. They were employed by nature long before man even thought about technology. Even man has been employing them for thousands of years. The Egyptians often added gold or cobalt to molten glass to colour it and make costume jewellery. The glass took on a red (gold) colour due to the surface plasmon resonances of the particles it contained. Lamp back has been used as a pigment for inks for thousands of years before carbon black particles saw more

66

Environmental and Human Health Impacts of Nanotechnology

widespread use. In the Graeco-Roman period mixtures of calcium hydroxide and lead oxide were used to dye hair, and it has been shown recently that this results in the formation of small (about 5 nm) nanoparticles of lead sulfide in the hair itself (Walter et al., 2006). Similarly, nanoparticles were applied in the glaze layers of pottery in the ancient world and there are examples from the eleventh century of shards of pottery painted with typical Islamic patterns which contain nanoparticles of silver and copper (Darque-Ceretti et al., 2005). Michael Faraday is one of the first people to take a scientific interest in gold sols or nanoparticles. He produced gold nanoparticles by the in situ reduction of gold chloride using white phosphorous and citrate as a capping agent. Carbon black has been produced on a commercial scale for more than 150 years and, similarly, asbestos has seen a rise and fall in use over the past 70 years. More recently, widespread commercial use of carbon black and asbestos probably represents the largest market for nanomaterials. 2.6.1.1 Asbestos Asbestos is a generic term for a range of naturally occurring minerals based on fibrous silicates. Asbestos is probably one of the best known nanomaterials in the general public and its detailed history provides an excellent background to the issues relating to very small particles and especially fibres. There are several reviews and books which deal in detail with the subject (Vallarino, 2001; Williams et al., 2007) and many more research papers and reports detailing its hazards. It is worth noting that as a material it is fairly ill defined, having a range of possible chemistries and an exceptionally broad range of sizes, which often fall at the upper scale of nanotechnology. The mineral crystallises in a fibre-like form which combined with its chemical composition results in its excellent insulation properties. The material itself is ill defined by current standards of engineered nanoparticles, consisting of a wide range of diameters (25–300 nm) and lengths (0.5–>20 µm) (Langer, 1974). The hazards of asbestos are well documented and are related to a combination of the chemistry of the material and its nanoscale, fibrous form. The initial applications of asbestos were in the ship building industry during the second world war where it was employed as a insulation material. By the 1960s the main user of asbestos became the building industry, accounting for more than 74% of the total mined material in the United States (NIOSH, 1972). By the 1980s there were about 3000 industrial or commercial products which had asbestos as one of the components (Anderson et al., 1982). Regulation in the 1970s (OSHA, 1986, 1994; US Environmental Protection Agency, 1988) resulted in a rapid decline in the market and use. The use of asbestos was on a scale of thousands of tonnes per annum and figures for 1979 show that, even after regulation, 130 000 tonnes per annum were being used in floor tiling alone (OSHA, 1986, 1994; US Environmental Protection Agency, 1988). It was even considered as a component of feminine hygiene products (OSHA, 1986, 1994; US Environmental Protection Agency, 1988), although there is no evidence that the concept was ever put into production. By 1997 the total asbestos market in the United States was still as high as 21 000 tonnes per annum distributed mainly in the building industry (Figure 2.20). The huge

Nanomaterials: Properties, Preparation and Applications

67

Other (5%) Friction Products (29%)

Roofing materials (48%)

Gaskets (17%)

Paper (1%)

Figure 2.20 The relative use of asbestos in commercial and industrial applications in the United States in 1997. (Adapted from Spengler et al., 2001.)

market for asbestos and its dark history currently serves as a warning for future manufactures of nanomaterials and probably is currently serving as a break on the rate of commercialisation of nanotechnology in the west. This history of asbestos clearly shows not only the need to understand the toxic effects of nanomaterials, but also how a material with exceptional properties can rapidly find uses in a very wide range of products and applications. 2.6.1.2

Carbon Black

Carbon black containing rubbers have excellent wear properties and the neat material also unrivalled optical properties. In 1995 its world wide production was 7200 tonnes per annum (Gardiner, 1995). Its use as an ink, as already discussed, predates industrial production by thousands of years. However, as a source of black pigment it still has excellent properties. This is because the carbon particles absorb light over a wide range of the optical spectrum with little bleaching of the colour with time. Generally, an ink will actually comprise of a suspension of carbon black in a suitable carrier liquid; in ancient times this might have been an oil or water. To date, the use of carbon black in tyres accounts for the largest use of the material. The incorporation of carbon black into the rubber has a significant impact in increasing the wear properties of the rubber and thereby increasing tyre life. Several studies have been conducted on the effects of carbon black on human and animal health and there are some excellent reviews on the subject. Gardiner (1995) generally found that the carbon black itself has little adverse effect on organisms. However, recently particular concern has been placed on aerial exposures and carbon black particles containing significant levels of PAHs. It is sufficient to say that the effects of carbon black on human health are not considered to be as significant as those relating to asbestos. However, the history of carbon black give a good demonstration of how the complexity of a nanomaterial, in this case a chemically simple one, can cause difficulties in assessing their effects on human health. 2.6.2 The Present and Near Future The use of asbestos and carbon black date back more than five decades. In this section present applications of nanomaterials (<30 years) are discussed. It will particularly

68

Environmental and Human Health Impacts of Nanotechnology

focus on the use of nanomaterials in a range of applications which have been realised with some consideration regarding where these might lead in the future. 2.6.2.1

Nanocomposites

A composite is a material consisting of two phases, one which serves to bind the material together and one which serves to enhance the properties of the material. Generally, they are thought of as structural materials, that is carbon fibre or glass fibre composites, although composites are prepared for a range of other applications and from a more diverse range of materials. Once again nature has beaten man in the use of nanocomposites and they form the basis of a huge range of biologically derived structural materials. Man-made materials of this type have general focus on polymer based nanocomposites. Exfoliated clay nanomaterials as reinforcing agents in nylon-6 based composites were initially explored as possible cam-belt casing materials in automobiles in the 1990s (Usuki et al., 1993). It was found that the incorporation of the nanomaterial resulted in significant improvements in mechanical performance and thermal stability. This has been explained by the large interfacial areas and the shape of the nanomaterial itself. Clay nanocomposites have now found applications in a range of cars from different manufacturers, including timing belt and engine covers based on nylon-6 (Gao, 2004), door components and steps based on polyolefin clay nanocomposites (Cox et al., 2004; Patterson, 2004) and seat backs from polypropylene clay nanocomposite (Hussain et al., 2006). Similar materials have also found applications in packaging where the limited diffusion of oxygen through the material results in improved food lifetimes. 2.6.2.2

Immunolabelling

A specialist market has arisen in photoluminescent nanoparticles for a range of applications. Perhaps one of the more well exploited is that of fluorescent tags. Fluorescent nanoparticles based on cadmium selenide or indium phosphide have been prepared with quantum yields of <70%. These nanoparticles may be surface modified to make them hydrophilic and then further modified to add a receptor group to the surface. Using this technology it is possible to prepare a nanoparticle which will selectively bind to a site on a cell. Unlike common organic fluorescent dyes, nanoparticles may absorb light at a range of wavelengths. This means that a range of fluorescence processes may be observed during one excitation process. The use of nanoparticles in these types of application allows the use of a single excitation wavelength to initiate fluorescence. The excitation wavelength may also be selected in order to minimise the fluorescence from the tissue or container. To date the exact composition of these materials is kept as a closely guarded secret. However, it is well known that to achieve the levels of quantum yields reported in the literature a core-shell structure is needed (Section 2.4.3). 2.6.2.3

Photovoltaics

The challenge in photovoltaics is to produce a cell which will harvest all the visible light falling onto it and convert it efficiently into electrical energy. Whilst the exact architecture of the cells may differ from cell to cell the principles are the same

Nanomaterials: Properties, Preparation and Applications Semiconductor/dye

69

Junction

Semiconductor

ht Lig nt ide Inc

e-

Electrodes

h+

h+ eW

W Electrical device

Electron

Figure 2.21 A simplification of a solar cell with electrical device W connected so work may be done.

(Figure 2.21). The impact of nanotechnology on the development of the solar cell dates back almost as far as the solar cell itself (1954) (Saunders et al., 2007) However, it has not been until more recently that the use of nanoparticles in this type of application has received so much attention. There are several issues that, where they to be resolved, would make solar power a viable method for energy production in most countries. Firstly, it is important that all of the light falling onto the cell is captured. However, there are very few materials which will readily capture light at such a broad spectrum of wavelengths and also produce the effects needed to convert the incident light into energy. The second issue is the efficiency of the conversion of light into electricity itself and the third is the method of manufacture. Ideally, solar cells would be manufactured by wet chemical processing routes on a range of materials such as plastics. There are three main approaches to these problems which use nanoparticles. The first is to prepare solar cells with different materials in a nanoparticulate form. The correct layering of these materials, of sufficiently high quality, is likely to result in improvements in performance. The second approach is to use a nanoparticle as an electrode, as in the DSSC discussed earlier, and the third is to use a nanoparticle as a composite with a semiconducting polymer. This latter application is still in its infancy. There is a wide range of materials currently used and under investigation (Table 2.5). Some of these materials have yet to see commercial use; however Table 2.5 shows the broad range of materials currently under consideration. 2.6.2.4 Waste Remediation The use of iron nanoparticles for the decontamination of land is well known. It has been shown that nanoparticles of iron and those of palladium and iron are excellent

70

Environmental and Human Health Impacts of Nanotechnology

Table 2.5 A list of used and proposed nanomaterials for dye sensitised solar cell (DSSC) and thin film solar cell devices. Material

Formula

Cell type

Reference

Cadmium telluride Cadmium sulfide Copper indium gallium selenide Fullerenes Cadmium selenide Copper indium selenide Indium sulfide Zinc oxide Titanium dioxide

CdTe CdS CuInxGa(1−x)Se2 C60 CdSe CuInSe2 InSe ZnO TiO2

Thin film Thin film Thin film Thin film Thin film Thin film Thin film Thin film & DSSC DSSC

a a a b c c c c&d e

References: a: Elliott and Zhang, 2001; b: Afzaal and O’Brien, 2006; c: Hua and Dong, 2007; d: Li et al., 2006; e: Nohynek et al., 2007.

candidates for the decomposition of organic wastes, especially chloro-hydrocarbons (Wang and Zhang, 1997; Phenrat et al., 2007). Field tests of these materials have shown that they can be efficient in decomposing these contaminants. However, there is some concern regarding the depletion of dissolved oxygen during treatment (Elliott and Zhang, 2001). Similar results are now being reported for nickel/iron nanoparticles (Tee et al., 2005). Titanium dioxide has also been used in self-cleaning glass (Parkin and Palgrave, 2005) and waste water treatment (Andersson et al., 2002). Both of these applications depend on the ability of titanium dioxide to photolytically generate radicals which can then catalyse the decomposition of organic materials. 2.6.2.5

Fuel Additives

The use of cerium oxide has been shown to improve the efficiency of diesel fuels vehicles by up to 11%. Ceria nanoparticles are currently manufactured commercially have been tested on some coaches used in normal service. The reported results show excellent properties in increasing fuel efficiency and reduction in the production of ultrafine particles (UFPs). 2.6.2.6

Bactericides

Nanoparticles of silver have been generally suggested as suitable biocidal materials. These would have the advantage over silver salts because they may be immobilised in a material and therefore not be consumed with time. Whilst there are several products on the market at the moment, the exact mechanism of the biocidal effects has yet to be determined and current speculation suggests either a radical mechanism causing cell death or possible immobilisation of sulfur containing proteins (Kim et al., 2007; Jeon et al., 2003). Either mechanism seems possible and should also include the possibility of dissolution of silver ions from the nanoparticles

Nanomaterials: Properties, Preparation and Applications

71

themselves, although this may well depend on the exact conditions of the medium (Chapter 3). 2.6.2.7

Drug Delivery

There is a drive to develop new methods for delivering drugs in the human body. One of the key issues in developing a drug is that many are poorly soluble in aqueous media. Whilst solubility may be enhanced by modification of the drug molecule, such modifications have the risk of altering the action of the drug itself. There is, therefore, a need to develop carriers which will deliver the active drug without modification and allow active doses to be delivered. One method for doing this is to encapsulate the drug in a polymer particle. The drug can then exit the particle by simple diffusion or by the gradual degradation of the particle itself. A whole range of polymers have been exploited as carriers for drugs such as polyesters (poly(glycolic acid), poly (lactic acid) and poly-ε-caprolactone), polyamides and polyphosphazenes. These polymers have been prepared as co-polymers with polyethylene oxide, polyethylene glycol and polypropylene oxide in order to prepare drug carrier particles (Uhrich et al., 1999; Pillai and Panchagnula, 2001). It is well known that the response of the body to particles can vary from organ to organ. The liver tends to filter out particles on the 1–5 µm scale and the spleen tends to remove particles greater than 200 nm. However, particles smaller than 100 nm tend to penetrate blood vessel walls. Generally then, the aim is to produce particles of around 100 nm. Generally the nanoparticles are prepared by either a living polymerisation, macromonomer initiation or polymer coupling, this results in a copolymer with two sections with different solubilities in water (Section 2.5.13). A dilute solution of the polymer is then prepared and carefully added to stirred water. As the polymer precipitates the segments which are insoluble in the water aggregate to form a hydrophobic core and the water soluble segments remain on the outside giving steric stabilisation. Particles of polycaprolactone-block-polyethylene oxide prepared by this method have diameters ranging from 20–50 nm (Hua and Dong, 2007). An alternative conformation for the nanoparticle is a polymersome where the centre of the particle also contains a hydrophilic area. It has been shown that degradable systems such as these can greatly affect the delivery of drugs. A cocktail of taxol (hydrophobic) and doxorubicin (hydrophilic) were placed in a polymersome and used to treat tumors on mice. The results showed that the polymersome was much more effective than simple administration of the cocktail without the carrier. It should be noted that the exact structure of the copolymer affects both the shape and form of the nanoparticle, its degradability and its activity. 2.6.2.8

Cosmetics/Sunscreens

The use of titanium dioxide (TiO2) or zinc oxide (ZnO) nanoparticles in sunscreens and anti-ageing creams is already prevalent. These nanomaterials offer the added benefit of being transparent to visible light because the crystal sizes are too small to cause significant scattering of visible light. Generally, these particles are of a core shell nature and have an outer layer of silica or alumina which will hinder the

72

Environmental and Human Health Impacts of Nanotechnology

migration of radicals to the surface of the particles. Other cosmetics may also contain the polymer-like delivery systems described above (Section 2.6.2.7) with the aim of depositing active ingredients in the upper layers of the skin. This area has been extensively studied and a recent review of the work so far could find no toxicological hazards associated with the use of some of these types of product (Nohynek et al., 2007) although more recent reviews have suggested that the area need further investigation (Starzyk et al., 2008). 2.6.2.9

Photo-Resistant Packaging:

Attempts to reduce the degree of spoil on packaged foods has also benefited from nanoparticles. It has been found that nanoparticles of titania may be incorporated into polymers to increase their attenuation of UV light and thereby reduce spoilage of the foods. This process can necessarily produce radicals which may result in damage to the polymer film. To circumvent this problem titania with small amounts of transition metal doping, such as manganese, have been produced. The dopants trap out the radicals and reduce the degree of damage to the polymer. 2.6.2.10 Abrasives One simple but important application of nanoparticles has been in the development of abrasives for polishing surfaces in the semiconductor industry (Nennemann et al., 2006). If a material is to be patterned with an electronic circuit on the sub 100 nm scale then the surface cannot contain any defects in this size range. It has been found that nanoparticles of silica, alumina or ceria are excellent for these applications.

2.7

Implication for Environmental Issues

Clearly the number of products containing nanoparticles is likely to rise and, therefore, even though there has been release of engineered nanoparticles for millennia on a small scale, the impact of such a diverse range of particles with wide ranging variations in composition and surface functionalisation is uncertain. There are clearly issues with regulating such materials due to the fact that they are not simple molecules with a clear well defined structure which is independent of the environment in which they are placed. In fact, as is discussed in Chapter 6, the application of several physiochemical measurements to nanoparticles may be impractical or misleading. A surface functionalised nanoparticle may well have at least two melting points, one for the surface functionalisation and one for the core material. Solubility of the particle may be low, that is the ions in the particle do not easily dissociate resulting in the loss of the particle, but the particle itself may disperse well giving a stable dispersion of particles which is not in itself a solution. In addition to this, the ability of molecules bound to the surface to associate and dissociate is not considered, and is exceptionally difficult to measure. Finally, the behaviour of nanoparticles in biphasic systems will give rise to significant issues in determining log Kow and partition coefficients.

Nanomaterials: Properties, Preparation and Applications

2.8

73

Conclusions

The field of nanotechnology is extensive, permeating a range of disciplines beyond materials science and chemistry and this chapter has reviewed at a very shallow level much of the field. It has been seen that nanomaterials cannot be considered as simple entities and certainly cannot be generally studied in the same manner as small molecules. Their properties are diverse and depend on composition, chemistry and size as well as interactions with other materials. No single property can suitably be attributed to all nanoparticles, other than perhaps their intrinsically small size and high surface area. It has also been seen that the preparation of nanomaterials is diverse and their preparation by different methods may therefore result in a range of materials that may be commonly labelled under one nanoparticle type but may have a wide range of compositions and properties. These facts, therefore, make the characterisation of nanoparticles a primary concern in determining their toxicological and ecotoxicological properties; an issue which will further be dealt with in further chapters. Given the rapid rise in the number of nanoparticle types and the breadth of their applications it is of immediate importance that the factors affecting nanoparticle toxicology and environmental interactions be determined. Such a set of parameters would allow, as a minimum, a qualitative prediction of the likely effects of a new nanoparticle and, therefore, allow the design of nanoparticles with low toxicity.

2.9

References

Afzaal, M. and P. O’Brien (2006) Recent developments in II-VI and III-VI semiconductors and their applications in solar cells; Journal of Materials Chemistry, 16, 1597–602. Anderson, P., M. Grant, R. McInnes and W. Farino (1982) Analysis of fiber release from certain asbestos products, Prepared for Chemical Control Division, Office of Pesticides and Toxic Substances, US Environmental Protection Agency, Washington, DC, EPA no. 68-01-5960. Andersson, M., L. Osterlund, S. Ljungstrom and A. Palmqvist (2002) Preparation of nanosize anatase and rutile TiO2 by hydrothermal treatment of microemulsions and their activity for photocatylytic wet oxidation of phenol; Journal of Physical Chemistry B, 106, 10674–9. Aveyard, R., B. P. Binks and J. H. Clint (2003) Emulsions stabilised solely by colloidal particles; Advances in Colloid and Interface Science, 100–102, 503–46. Avouris, P., Z. Chen and V. Perebenios (2007) Carbon based electronics; Nature nanotechnology, 2, 605–15. Balzani, V., M. Clemente-Leon, A. Credi, et al. (2006) Autonomous artificial nanomotor powered by sunlight; Proceedings of the National Academy of Sciences, 103, 1178–1183. Binks, B. P., J. H. Clint, P. D. I. Fletcher, et al. (2006) Growth of gold nanoparticles films driven by the coalescence of particle-stabilised emulsion drops; Langmuir, 22, 4100–03. Bukowski, T. J. and J. H. Simmons (2002) Quantum Dot Research: Current State and Future Prospects; Critical Reviews in Solid State and Materials Sciences, 27, 119–142. Castleman, B. (2006) Asbestos products, hazards and regulation; International Journal of Heath Services, 36, 295–307. Christian, P. and P. O’Brien (2005) A new route to nanorods of cadmium sulfide; Chemical Communications, 2817–9.

74

Environmental and Human Health Impacts of Nanotechnology

Christian, P. and P. O’Brien (2008) Thermodynamic and kinetic control of crystal growth in CdS nanomaterials; Journal of Materials Chemistry, 18, 1689–93. Cox, H., et al. (2004) Nanocomposite Systems for Automotive Applications; 4th World Congress in Nanocomposites, EMC, San Francisco, September 2005, 1–3. Darque-Ceretti, E., D. Helary, A. Bouquillon and M. Aucouturier (2005) Gold like lustre: nanometric surface treatment for decoration of glazed ceramics in ancient Islam, Moresque Spain and renaissance Italy; Surface Engineering, 21, 352–8. Duguet, E., S. Vasseur, S. Mornet and J. M. Devoisselle (2006) Magnetic nanoparticles and their applications in medicine; Nanomedicine, 1, 157–68. Elliott, D. W. and W. X. Zhang (2001) Field assessment of nanoscale bimetallic particles for groundwater treatment; Environmental Science & Technology, 35, 4922–6. Gao, F. (2004) Clay/Polymer Composites: the Story; Materials Today, 7, 50–5. Gardiner, K. (1995) Effects on respiratory morbidity of occupational exposure to carbon black: a review; Archives of Environmental Health., 50, 44–52. Guardia, P., B. Batlle-Brugal, A. G. Roca, et al. (2007) Surfactant effects in magnetite nanoparticles of controlled size; Journal of Magnetism and Magnetic Materials, 316, c756–9. Harrison, R. D. (1984) Nuffield book of data (2nd edn), Longman Group Ltd, London. Hartlieb, K. J., C. L. Raston and M. Saunders (2007) Controlled scaleable synthesis of ZnO nanoparticles; Chemistry Materials, 19, 5453–9. Hua, C. and C.-M. Dong (2007) Synthesis, characterisation, effect of architecture on crystallization of biodegradable poly(E-caprolactone)-b-poly(ethylene oxide), copolymers with different arms and nanoparticles thereof; Journal of Biomedical Materials Research Part A, 689. Hussain, F., M. Hojjati, M. Okamoto and R. E. Gorga (2006) Polymer-matrix nanocomposites, processing, manufacturing and application: An overview; Journal of Composite Materials, 40, 1511–75. Jeon, H. J., S. C. Yi and S. G. Oh (2003) Preparation and antibacterial effects of Ag SiO2 thin films by sol-gel method; Biomaterials, 24, 4921–8. Kim, J. S., E. Kuk and K. N. Yu (2007) Antimicrobial effects of silver nanoparticles; Nanomedicine: Nanotechnology, Biology, and Medicine, 3, 95–101. Kunze, U. (2002) Nanoscale devices fabricated by dynamic ploughing with an atomic force microscope; Superlattices and Microstructures, 31, 3–17. Langer, A. (1974) The subject of continuous vigilance; Environmental Health Perspectives, 9, 53–6. Leggett, G. J. (2006) Scanning near-field photolithography-surface photochemistry with nanoscale spatial resolution; Chemical Society Reviews, 35, 1150–61. Leuenberger, M. N. and D. Loss (2000) Spin tunnelling and phonon-assisted relaxation in Mn12-acetate; Physical Review B, 61, 1286–302. Li, B., L. Wang, B. Kang, et al. (2006) Review of recent progress in solid-state dye-sensitized solar cells; Solar Energy Materials & Solar Cells, 90, 549–73. Li, Y., I. A. Kinloch and A. H. Windle (2004) Direct spinning for carbon nanotube fibres from chemical vapour deposition synthesis; Science, 304, 276–8. Liedl, T., T. L. Sobey and F. C. Simmel (2007) DNA-based nanodevices; Nanotoday, 2, 36–41. Liu, Y., S. He, M. Uehara and H. Maeda (2007) Wet chemical preparation of well dispersed colloidal cerium oxide nanocrystals; Chemistry Letters, 36, 764–5. Long, D-L., E. Burkholder and L. A. Cronin (2007) Polyoxometalate clusters, nanostructures and materials: From self assembly to designer materials and devices; Chem. Soc. Rev., 36, 105–21. Lyon, D., L. Adams, J. C. Falkner and P. J. Alvarez (2006) Antibacterial Activity of Fullerene Water Suspensions: Effects of Preparation Method and Particle Size; Environ. Sci. Technol., 40, 4360–6. Mana, L., E. C. Scher and A. P. Alivisatos (2000) Synthesis of soluble and processable rod-, arrow-, teardrop-, and tetrapod-shaped CdSe nanocrystals; Journal of the American Chemical Society, 112, 12700–6.

Nanomaterials: Properties, Preparation and Applications

75

Mendes, P. M., S. Jacke, K. Critchley, et al. (2004) Gold Nanoparticle Patterning of Silicon Wafers Using Chemical e-Beam Lithography; Langmuir, 20, 3766–8. Murry, C. B., D. J. Norris and M. Bawendi (1993) Synthesis and characterization of nearly monodisperse CdE (E = S, Se Te), semiconductor nanocrystallites; The Journal of the American Chemical Society, 115, 8706–15. Nennemann, A., M. Voetz, G. Hey, et al. (2006) Colloid chemical interactions of silica particles in the Cu-CMP-process; Progress in Colloid and Polymer Science, 133, 159–68. NIOSH (1972) Criteria for a recommended standard – Occupational exposure to asbestos; US Department of Health, Education, and Welfare, Public Health Service, Centers for Disease Control, National Institute for Occupational Safety and Health Washington, DC, HSM-72–10267. NIOSH (2004) What is nanotechnology, National Institute for Occupational Safety and Health, Washington, DC, 175. Nohynek, G. J., J. Lademann, J. Ribaud and M. S. Roberts (2007) Grey goo on the skin? Nanotechnology, cosmetic and sunscreen safety; Critical Reviews in Toxicology, 37, 251–77. Nolan, R. P., A. M. Langer and R. Wilson (1999) A risk assessment for exposure to grunerite asbestos (amosite) in an iron ore mine; Proceedings of the National Academy of Science, 96, 3412–9. OSHA (1986) Occupational exposure to asbestos, tremolite, anthophyllite, and Actinolite. Occupational Safety and Health Administration. Washington, DC, US Department of Labor. 29 CFR Parts 1910 and 1926, 51, 119, 22612–790. OSHA (1994) Occupational exposure to asbestos; Final rule. Occupational Safety and Health Administration, Washington, DC, US Department of Labor. 29 CFR Parts 1910, 1915, and 1926, 59, 153, 40964–1162. Ozin, G. A., I. Manners, S. Fournier-Bidoz and A. Arsenault (2005) Dream Nanomachines; Advanced Materials, 17, 3011–8. Parkin, I. G. and R. G. Palgrave (2005) Self cleaning coatings; Journal of Materials Chemistry, 15, 1689–95. Patterson, F. (2004). Nanocomposites – Our Revolutionary Breakthrough; 4th World Congress in Nanocomposites, EMC, San Francisco, September 2004, 1–3. Performance Composites Ltd (2008) online data sheet available at: http://www.performancecomposites.co.uk/carbonfibre/structuralprofile.asp Phenrat, T., N. Saleh, K. Sirk, et al. (2007) Stabilization of aqueous nanoscale zerovalent iron dispersions by anionic polyelectrolytes: adsorbed anionic polyelectrolyte layer properties and their effect on aggregation and sedimentation; Journal of Nanoparticle Research, 10, 795–814. Pijeret, D., R. A. v Delden, A. Meetsma and B. L. Feringa (2005) Acceleration of a nanomotor: Electronic control of the rotary speed of a light-driven molecular rotor; Journal of the American Chemical Society, 127, 17612–3. Pillai, O. and R. Panchagnula (2001) Polymers in drug delivery; Current Opinion in Chemical Biology, 5, 447–51. Pratsinis, S. E. (1998) Flame aerosol synthesis of ceramic powders; Prog. Energ. Combust., 24, 197–219. Qiu, S., J. Dong and G. Chen (1999) Preparation of Cu nanoparticles from water-in-oil microemulsions; Journal of Colloid and Interface Science, 216, 230–4. Royal Society & The Royal Academy of Engineering (2004) Nanoscience and nanotechnologies: opportunities and uncertainties, July 2004. London. (Also The Royal Society’s website www.royalsoc.ac.uk/policy and The Royal Academy of Engineering’s website www.raeng. org.uk.). Sau, T. K., A. Pal and T. Pal (2001) Size regime dependent catalysis by gold nanoparticles fro the reduction of eosin; Journal of Physical Chemistry B, 105, 9266–72. Saunders, J. R., D. Benfield, W. Moussa and A. Amirfazli (2007) Nanotechnology’s implications for select systems of renewable energy; International Journal of Green Energy, 4, 483–503.

76

Environmental and Human Health Impacts of Nanotechnology

SCENIHR (2005) Request for a Scientific Opinion on the Appropriateness of Existing Methodologies to Assess the Potential Risks Associated with Engineered and Adventitious Nanotechnologies. SCENIHR/002/05 (Scientific Committee on Emerging and Newly Identified Health Risks, http://ec.europa.eu/health/ph_risk/committees/04_scenihr/docs/ scenihr_o_003.pdf (March 09). Schnippering, M., M. Carrara, A. Foelske, et al. (2007) Electronic properties of Ag nanoparticle arrays. A Kelvin probe and high resolution XPS study; Physical Chemistry and Chemical Physics, 9, 725–30. Shaw, D. J. (1992) Introduction to colloid and surface chemistry (4th edn); ButterworthHeinemann Ltd, Oxford, UK. Simmons, B. A., S. Li, V. T. John, et al. (2002) Morphology of CdS nanocrystals synthesised in a mixed surfactant system; Nanoletters, 2, 263–8. Soong, R. K., G. D. Bachand, H. P. Neves, et al. (2000) Powering an Inorganic Nanodevices with a Biomolecular Motor; Science, 290, 1555–8. Spengler, J. D., J. M. Samet and J. F. McCarthy (2001) The in-door air quality handbook, McGraw-Hill, New York. Starzyk, E., A. Frydrych and A. Solyga (2008) Nanotechnology: Does it have a Future in Cosmetics? SOFW-Journal, 134, 42–52. Sun, Y. G., B. Mayers, T. Herricks and Y. N. Xia (2003) Polyol synthesis of uniform silver nanowires: a plausible growth mechanism and the supporting evidence; Nanoletters, 3, 955–60. Tee, Y. H., E. Grulke and D. Bhattacharyya (2005) Role of Ni/Fe nanoparticle composition on the degradation of trichloroethylene from water; Industrial Engineering Chemistry Research, 44, 7062–70. Thostensona, E. T., Z. Renb and T. W. Chou (2001) Advances in the science and technology of carbon nanotubes and their composites: a review; Composites Science and Technology, 61, 1899–912. Uhrich, K. E., S. M. Cannizzaro, R. S. Langer and K. M. Shakesheff (1999) Polymeric systems for controlled drug release; Chemical Reviews, 99, 3181–98. Ulrich, G. D. (1984) Flame synthesis of fine particles; Chemical Engineering News, 62, 22–9. Ung, D., Y. Soumare, N. Chakroune, et al. (2007) Growth of magnetic nanowires and nanodumbbells in liquid polyol; Chemistry of Materials, 19, 2084–94. US Environmental Protection Agency (1988) Asbestos exposure assessment. EPA Chemical Engineering Branch, Office of Pesticides and Toxic Substances, Washington, DC. Usuki, A., M. Kawasumi, Y. Kojima, et al. (1993) Swelling Behavior of Montmorillonite Cation Exchanged for omega-amino Acids by E-caprolactam; Materials Research, 8, 1174–8. Vallarino, M. S. (2001) Fibres, in The indoor air quality handbook, (eds Spengler, J. D., McCarthy J. F. and Samet J. M.), McGraw-Hill, New York. Vicario, J., N. Katsonis, B. S. Ramon, et al. (2006) Nanomotor rotates microscale objects; Nature, 440, 163. Walter, P., E. Welcomme, P. Hallegot, et al. (2006) Early use of PdS nanotechnology for an anchient hair dying formula; Nano Letters, 6, 2215–9. Wang, C. B. and N. D. Zhang (1997) Synthesizing nanoscale iron particles for rapid and complete dechlorination of TCE and PCBs.; Environmental Science and Technology, 31, 2154–6. Wells, R. L. and J. F. Janik (1996) Heterogeneous solution reactions between MBr3 (M = Ga,In), and Li3N. Formation and characterisation of nanocrystalline GaN powders; European Journal of Solid State and Inorganic Chemistry, 33, 1079–90. Williams, P. R. D., A. D. Phelka and D. J. Paustenbach (2007) A review of historical exposures to asbestos among skilled craftsmen (1940–2006); Journal of Toxicology and Environmental Health, Part B, 10, 319–77. Wooldridge, M. S. (1998) Gas-phase combustion synthesis of particles; Progress in Energy Combustion, 24, 63–87.

Nanomaterials: Properties, Preparation and Applications

77

Xu, Y. F., M. L. Yan and D. J. Sellmyer (2007) FePt nanocluster films for high-density magnetic recording; Journal of Nanoscience and Nanotechnology, 7, 206–24. Yu, M. F., O. Lourie, M. J. Dyer, et al. (2000) Strength and breaking mechanism of multiwalled carbon nanotubes under tensile load; Science, 287, 637–40. Zaizer, E. M. and V. K. Lamer (1948) The kinetics of the form and growth of monodisperse hydrosols; Journal of Colloid Science, 3, 571–98.

3 Size/Shape–Property Relationships of Non-Carbonaceous Inorganic Nanoparticles and their Environmental Implications Deborah M. Aruguete*, Juan Liu† and Michael F. Hochella, Jr.* *Department of Geosciences, Virginia Polytechnic Institute, Blacksburg, USA †Chemical and Material Sciences Division, Pacific Northwest National Laboratory, Rickland, USA

3.1

Introduction

Non-carbonaceous inorganic nanoparticles form one of the major classes of emerging synthetic nanomaterials. Such nanoparticles are already commercially used in products such as sunscreens, pigments and antibacterial coatings and continue to be heavily researched for other applications. The environment is already being exposed to these nanoparticles and this exposure will increase as more applications are developed. Most nanoscience studies have been conducted under non-environmentallyrelevant conditions and, as a result, there is a dearth of data on the environmental fate and behaviour of inorganic nanoparticles; such data is precisely what scientists and engineers need to predict the environmental impact of the nanoparticles. In many cases, environmentally relevant data may be available for their bulk counterparts. For example, data may exist pertaining to 100 µm particles of titanium dioxide as opposed to 10 nm particles of titanium dioxide. However, many of these nanoparticles have striking size (and sometimes shape) dependent properties which can be radically different from those of their bulk corresponding materials, thereEnvironmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

80

Environmental and Human Health Impacts of Nanotechnology

fore extrapolation of this data would be unwise. Understanding the size dependent differences between bulk and nanoscale properties is important for shaping future experiments in environmental nanoscience, as well as for the design of wellinformed nanotechnology legislation. In this review, an attempt is made to link what is currently known about the size dependent behaviour of inorganic nanoparticles and how this knowledge might apply in environmental or natural systems. Size and shape effects upon redox/ catalysis chemistry, sorption processes and dissolution are reviewed, and suggestions made on what sorts of future studies are necessary to answer the following question: How will nanoparticles behave differently from their bulk counterparts, and what might their environmental fate and impact be?

3.2

Inorganic Nanoparticle Anatomy

In environmental nanoscience, it is critical to understand that nanoparticles can display an astoundingly broad range of physical and chemical properties. This variety of properties generally results from differences in structure and composition between different nanoparticles. Therefore, to understand the behaviour of any given type of nanoparticle, its structural and chemical characteristics must be known. While these characteristics are addressed in more detail in Chapter 2, they are briefly reviewed here for inorganic nanoparticles in terms of environmental behaviour (organic nanoparticles are covered in more detail in Chapter 4). Commercial nanoparticles are composed of a core material and often a coating (although this is optional), as shown in Figure 3.1. This core material can in theory be any solid inorganic substance, although those most under development for applications (see Chapters 1 and 2 for more information) include metals (gold, silver, palladium and platinum) and semiconductors/insulators (metal sulfides, selenides and oxides). The core material can vary in both size and shape and confers many of the nanoparti-

Metals 1. Composition oxides of core particle sulfides material etc.

5. Aggregation Aggregated Dispersed?

2.Size of core particle

3. Core particle shape

4. Surface coating / functionalization on material (nothing, molecules, polymer…)

Figure 3.1 Flow diagram displaying variables in the ‘anatomy’ of a colloidal inorganic nanoparticle.

Size/Shape–Property Relationships

81

cle’s basic physical characteristics (e.g. optical and magnetic properties). Possible coatings include small molecular moieties, for example alkylthiols (Kloepfer et al., 2005), polymers (Harris et al., 2003), biomolecules (Gao et al., 2005; Nie et al., 2007; Rhyner et al., 2006) and inorganic coatings (e.g. a zinc sulfide film on a cadmium selenide (CdSe) nanoparticle) (Dabbousi et al., 1997; Hines and Guyot-Sionnest, 1996). Such coatings commonly affect important nanoparticle properties, such as solubility in water (Pellegrino et al., 2004), resistance against chemical degradation (e.g. passivation of a particle to prevent oxidation) (Yu et al., 2007) and affinity for different biological tissues (Bruchez et al., 1998). The specific chemical behaviour of any given nanoparticle will depend strongly upon the core material composition, the size, the shape, the coating and the aggregation state. Realistically, as a part of nanotechnology products, nanoparticles may be embedded in a particular matrix or on a substrate which can also influence nanoparticle behaviour (this matter is outside the scope of this review).

3.3

Redox Chemistry of Nanoparticles

In nature, redox reactions are an important part of phenomena such as mineral weathering, bacterial respiration and degradation of pollutants. Many nanoparticles are of interest for applications due to their ability to catalyze or directly participate in redox processes. Thus, if released into the environment, such nanoparticles may influence natural redox phenomena, including those within living organisms. Here, the varied origin and nature of redox properties in nanoparticles is discussed. (If the reader is not already versed in the basics of electronic structure and light absorption in inorganic nanoparticles, it is recommended that Chapter 2 of this book be reviewed.) 3.3.1

Photoredox Chemistry in Semiconductor Nanoparticles

When semiconductor materials absorb light of the proper energy, mobile charge carriers (electrons and holes) can be generated. If these charge carriers reach the surface of the semiconductor material, they may reduce or oxidize compounds on or near the surface, depending upon the redox potentials of the compounds. The size of the semiconductor material can affect many aspects of such redox processes. Generally, if a semiconductor nanoparticle is below a certain critical size (which depends upon the parent semiconductor material), it can exhibit quantum size effects (Alivisatos, 1996). In such cases, the wave functions of the charge carriers extend over the entire particle. Therefore, the charge carriers will not have to diffuse to participate in reactions at the particle surface (Hagfeldt and Gratzel, 1995). Quantum size effects result in the shifting of band edge energies (changes in electronic energy levels) (Alivisatos, 1996). This alters the redox potentials of charge carriers in a given nanoparticle with respect to the bulk. Therefore, a nanoparticle may be energetically able to participate in a particular redox reaction that is not possible for the parent bulk material. An excellent example of this phe-

82

Environmental and Human Health Impacts of Nanotechnology

nomenon is provided by nano-sized molybdenum disulfide (MoS2) (Abrams and Wilcoxon, 2005; Thurston and Wilcoxon, 1999). The redox potential positions of the valence and conduction bands of bulk and nanoparticulate molybdenum disulfide are shown in Figure 3.2. For comparison, the redox potentials of some environmentally or biologically relevant half reactions are included. These reactions include the generation of hydroxyl radicals (•OH) from water. Hydroxyl radicals can play a major role in oxidative damage in biological systems (Imlay, 2003; Sayre et al., 2008), as well as the degradation of organic compounds (Wilcoxon, 2000; Kamat and Meisel, 2002). Also included are the reduction of AQDS (9, 10anthraquinone-2, 6-disulfonic acid) (Sund et al., 2007), a synthetic analog of electron-shuttling molecules used in bacterial respiration, and the reduction of acetate to pyruvate (Becker and Deamer, 1986). Notably, size effects in the photocatalytic activity of molybdenum disulfide have been demonstrated. It was shown that smaller molybdenum disulfide nanoparticles (4.5 nm and below) could photocatalyze redox reactions that would degrade organic molecules, while larger nanoparticles (8–10 nm) could not (Abrams and Wilcoxon, 2005; Thurston and Wilcoxon, 1999; Wilcoxon, 2000). Results suggested that the size dependence of photocatalytic activity was due to the higher redox potential of the holes (Figure 3.2) in the smaller nanoparticles, which in turn could oxidize water and create reactive hydroxyl radicals. This is a demonstration of how nanoparticle size could influence the environmental or toxicological effects of a material. Altering

Potential vs. NHE

-1.0

Acetate/ pyruvate

0.0 MoS2 (d=4.5 nm) MoS2(bulk) MoS2 (d=8–10 nm)

AQDS (electron shuttle)

1.0 H2O/•OH 2.0

Figure 3.2 Position of the conduction and valence band edges versus the normal hydrogen electrode (NHE) for bulk and nanoparticulate MoS2, plus redox potentials for environmentally or biologically relevant half reactions. Note that by varying size, the redox properties of MoS2 are altered. For example, photoexcited 4.5 nm MoS2 nanoparticles have holes with a redox potential more positive than 1.2–1.5 V, which means these holes can oxidize water and create hydroxyl radicals. The hydroxyl radicals can then degrade organic chemicals or, potentially, cause oxidative damage in biological systems. Also displayed are redox potentials for the conversion of acetate to pyruvate and for the reduction of AQDS, a synthetic analog of electron-shuttling molecules important for bacterial respiration. (Adapted with permission from T.R. Thurston and J.P. Wilcoxon (1999), Photooxidation of Organic Chemicals Catalyzed by Nanoscale MoS2, Journal of Physical Chemistry B, 103, 11–7. Copyright (1999) American Chemical Society.)

Size/Shape–Property Relationships

83

redox potentials in nanoparticles can also change other aspects of redox processes; in cadmium sulfide (CdS) nanoparticles, nitrate reduction rates were found to increase with decreased particle size (Korgel and Monbouquette, 1997). As kinetic phenomena can be important for many geochemical processes (Lasaga, 1998), such effects may alter the geochemical impact of nanoparticles. It should be emphasized that not every type of semiconductor nanoparticle displays quantum size effects. Their manifestation depends both upon the composition of the nanoparticle and its size. Interestingly enough, even in the absence of quantum size effects, nanoparticle size may affect the kinetics of redox processes. For example, size can alter the average transit time for a charge carrier to diffuse from a nanoparticle’s interior to its surface (Hagfeldt and Gratzel, 1995). The redox reactivity of an inorganic semiconductor nanoparticle is not only determined by its core material but also by its coatings. Some coatings may enhance charge carrier transport out of the nanoparticle. For example, an electroactive ligand coating was developed for CdSe nanocrystals to improve their performance in electronic devices, such as photovoltaic cells (Milliron et al., 2003). Essentially, the energy level alignment of the ligand coating molecules favoured the transfer of charge carriers (holes) from the photoexcited CdSe nanoparticle to the ligand. Adding a layer of zinc sufide onto CdSe, on the other hand, can help to confine charge carriers, because of the energetic positions of zinc sufide band edges with respect to those of CdSe (Dabbousi et al., 1997; Hines and Guyot-Sionnest, 1996). Energy level schematics for such systems are displayed in Figure 3.3. It is evident

(a)

LUMO

(b)

ZnS

– Conduction e band edge

Valence band edge h+

CdSe nanocrystal

HOMO

Energy

Energy

CdSe

CdSe

ligand ZnS

ZnS

Figure 3.3 (a) Schematic of valence and conduction band edges in a CdSe nanoparticle and their energetic alignment with a molecular ligand coating the particle. Mobile charge carriers (the hole and electron) are generated when a photon is absorbed by the nanoparticle, as shown. Holes can be transferred to the highest occupied molecular orbital (HOMO) of the ligand. (b) Schematic of a CdSe–ZnS core–shell nanoparticle and the corresponding band edges of the core and the shell. ((a) Adapted from D.J. Milliron, A.P. Alivisatos, C. Pitois, C. Edden and J.M.J. Frechet (2003) Electroactive Surfactant Designed to Mediate Electron Transfer Between CdSe Nanocrystals and Organic Semiconductors, Advanced Materials, 15, 58–61; copyright Wiley-VCH Verlag GmbH & Co. KGaA; reproduced with permission. (b) Adapted with permission from B.O. Dabbousi, J. RodriguezViejo, F.V. Mikulec et al. (1997), Journal of Physical Chemistry B, 101, 9463–75; copyright (1997) American Chemical Society.)

84

Environmental and Human Health Impacts of Nanotechnology

that coatings must be considered carefully when predicting the environmental chemistry of a nanoparticle. 3.3.2 3.3.2.1

Redox Chemistry in Other Nanoparticle Systems Precious Metal Nanoparticles

The properties and applications of precious metal nanoparticles (gold, silver, platinum and palladium) are a very popular subject of study in nanoscience; materials such as platinum have long been used for commercial catalysis. Here the focus is upon gold as the primary example, as its behaviour arguably has many commonalities with other precious metal nanoparticles. Nanoparticulate gold is also particularly relevant as it is popular for application in nanomedicine, biolabelling and sensing (Hayat, 1989, El-Sayed et al., 2005, 2007; Huang et al., 2006, 2007a, 2007b; Oyelere et al., 2007; Sonnichsen and Alivisatos, 2005). Unlike materials such as platinum, gold in bulk form is chemically fairly inert. Nevertheless, in nanoparticulate form, it can display significant catalytic activity, both in solution and on solid supports. Gold nanoparticles can participate in many redox reactions. These include reactions of potential environmental interest, such as the oxidation of carbon monoxide (Ketchie et al., 2007; Valden et al., 1998) and the degradation of organic pollutants (Deng et al., 2005, 2007; Panigrahi et al., 2007). Precious metal nanoparticles can behave as electron transfer mediators between molecules or other species (e.g. between semiconductor nanocrystals and molecules) (Kiwi and Gratzel, 1979; Miller et al., 1981; Cozzoli et al., 2004). Reported reaction rates are significant, often comparable to the rates for reactions promoted by commonly studied catalysts (Somorjai, 1994). Size dependence of catalytic redox behaviour has been observed in colloidal gold reactions in aqueous solutions. Reports of such behaviour in the literature vary widely. For example, Sau and co-workers examined the gold nanoparticle catalysis of eosin dye reduction with sodium borohydride (NaBH4) and found two distinct size dependence regimes (Sau et al., 2001). Surface area normalized rates decreased for nanoparticle diameters from 10 to 15 nm, then increased from 15 to 46 nm. In a different study monitoring the hydrogenation of anthracene, a polycyclic aromatic hydrocarbon pollutant, nanoparticles ranging from 4.1 to 24.7 nm displayed an increasing turnover frequency (number of anthracenes reduced per surface atom per second) with decreasing size (Deng et al., 2005, 2007). Similar variation has been observed for other precious metal nanoparticle systems (Sharma et al., 2003; Duan et al., 2007). The reported variation in size dependent properties complicates attempts to predict the behaviour of precious metal nanoparticles in the environment. Nevertheless, additional studies can potentially rectify this problem. Recent studies on platinum nanoparticle catalysis indicate that nanoparticle morphology could be important in determining reaction rates (Narayanan and El-Sayed, 2005). In the aforementioned studies on colloidal precious metal nanoparticle catalysis, mostly size characterization is reported (particles are generally assumed to be approximately spherical). It is possible that preparations leading to batches with different

Size/Shape–Property Relationships

85

particle sizes also yield slightly different morphologies; thus some of the variation in catalytic activity attributed to size effects might actually be due to differences in particle shape. Hence, future studies of catalytic behaviour should include more characterization of morphology. Also, from study to study, and even within studies, the surface coatings vary, which may strongly affect the catalytic properties. More systematic comparisons of the catalytic properties of nanoparticles with different coatings would be very helpful. It should be noted that the effect of size upon a redox reaction will be dependent upon the particular energetics of that system. 3.3.2.2

Zero-Valent Iron and Iron Oxides

Iron For years, zero-valent iron (ZVI), or metallic iron, has been studied for the remediation of contaminated areas (Lo et al., 2007). As the standard reduction potential for the (Fe/Fe2+) redox reaction is −0.44 V (Atkins, 1998), metallic iron is able to reduce and transform many substances, including Cr6+, Pb2+ and halogenated organic pollutants. More recently, nanoparticulate ZVI or nZVI (including mixtures of nZVI plus an activating metal, e.g. palladium) have been investigated for use in environmental clean-up efforts (Zhang, 2003; Zhang et al., 1998; Schrick et al., 2002; Tratnyek and Johnson, 2006). nZVI is attractive for this application due to its higher specific surface area, as ZVI reductive transformation rates have been shown to be proportional to surface area (Johnson et al., 1996). Also, nZVI can be directly injected into contaminated sites, allowing for flexibility in its application (Li et al., 2006a, 2006c). Those interested in optimizing the performance of nZVI have begun to examine how physical characteristics such as nanoparticle size affect reactivity. Results have varied. For example, nZVI has been shown to reduce polychlorinated biphenyls (Lowry and Johnson, 2004), while microscale ZVI cannot. On the other hand, in another study on the reduction of nitrate, surface area normalized rate constants for 9.5 nm and 45 nm nanoparticles did not even vary by an order of magnitude (Liou et al., 2006). Interestingly enough, various nZVI preparations have been shown to alter which products result from the reductive transformations of halogenated organics (Liu et al., 2005a, 2005b; Nurmi et al., 2005). In the reduction of carbon tetrachloride, two different preparations of nZVI and microscale ZVI produced different amounts of chloroform, an undesirable product (Nurmi et al., 2005). Such findings are significant, as they mean that the chemical behaviour of nZVI might be controlled by merely altering the preparation used. The origins of nZVI chemical behaviour have not yet been exactly determined, although various studies indicate differences in crystallinity (Liu et al., 2005a) and the amount of oxides or other elements present (Liu et al., 2005b; Nurmi et al., 2005) could be critical. The iron nanoparticle sizes used are far too large to have electronic structures significantly different from the bulk, so this is an unlikely cause of their behaviour (Wang et al., 2000). One possible factor influencing the reactivity of nZVI would be nanoparticle shape and surface bonding coordination. Future studies of nZVI could utilize samples of higher uniformity in shape and size. It should be noted that nZVI has proved not only to be successful at decomposing pollutants in the laboratory. nZVI can effectively decompose pollutants in situ,

86

Environmental and Human Health Impacts of Nanotechnology

as demonstrated with multiple field tests in which nZVI was directly introduced to the environment (Li et al., 2006c; Zhang, 2003; Elliott and Zhang, 2001; Quinn et al., 2005). Given this success, use of nZVI for remediation may become more widespread. Thus, it is particularly important to consider its potential environmental effects. A recent study showed that humic acids can sorb strongly onto nZVI and even react with it, removing humic acid from solution (Giasuddin et al., 2007). While nZVI may be used with the best of intentions for remediation, it may have unintended consequences. Iron oxides It is already well established that iron oxides, including hydroxides and oxyhydroxides, play an important role in the environment. Both macro and nanoscaled iron oxides are naturally present in the environment and are involved in multiple chemical and transport processes (Davison and De Vitre, 1992; Cornell and Schwertmann, 2003; Brown et al., 1999; Dzombak and Morel, 1990). It is therefore likely that synthetic iron oxides could influence the environment as well, if released in sufficient quantities. Iron oxides are of great interest for various nanotechnologies, not only because of their intrinsic properties but also because of their low cost and low toxicity. Magnetic iron oxides have been studied for applications such as magnetic resonance imaging contrast enhancement (Lee et al., 2006) and high density data storage (White et al., 1997). Other potential nanotechnology applications include hydrogen generation (Vayssieres et al., 2005) and catalysis (Tsodikov et al., 2005; Halim et al., 2007; Liu et al., 2007). In addition to being synthesized for applications in their own right, nanoscaled iron oxides are also present in preparations of nZVI (Martin et al., 2008; Li et al., 2006a, Nurmi et al., 2005; Liu et al., 2005a). As with their bulk counterparts, nanoscale iron oxides are known to be redox active. In particular, there are multiple examples of nanoscale iron (III) oxide reduction by organic molecules (Roden, 2003; Torrent et al., 1987; Houben, 2003; Larsen and Postma, 2001). This includes molecules such as hydroquinone, a synthetic analog of biological electron transfer molecules (Anschutz and Penn, 2005). Iron (III) oxides can also participate in redox biochemistry as electron receptors for bacteria respiring under anaerobic conditions (Roden, 2003; Roden and Zachara, 1996). As well as participating directly in redox reactions, nanoscale iron oxides can act as catalysts in low-temperature systems. One reaction that multiple types of iron oxide nanoparticles (Fe3O4, ferrihydrite, α-Fe2O3) can catalyze is the decomposition of hydrogen peroxide, which can be useful for oxidizing and degrading organic pollutants in water (Filip et al., 2007; Hermanek et al., 2007; Zelmanov and Semiat, 2008; Gao et al., 2007). This catalytic activity may have significance in biological systems, as it mimics the enzyme activity of peroxidases (Gao et al., 2007). Another reaction of environmental importance that can be catalyzed (photocatalyzed) is the oxidation of sulfite (Faust et al., 1989). The effect of size upon the redox reactivity of iron oxides is still not well established. While a number of studies have examined the effect of varying particle surface area (or size) on reaction rates, results from studies vary (Roden, 2003; Liu et al., 2006; Gao et al., 2007; Schwertmann et al., 1985; Larsen and Postma, 2001;

Size/Shape–Property Relationships

87

Houben, 2003; Torrent et al., 1987). The interpretation of results from these studies is often complicated by variation in iron oxide phase amongst samples of different sizes. Further studies are needed to specifically examine whether there are any intrinsic size dependent effects present in iron oxide nanoparticles of the same phase, independent of surface area. One example of such a study concerns the oxidation rate of Mn2+ to Mn3+ by different sizes of α-Fe2O3 (haematite) (Madden and Hochella, 2005). Surface normalized oxidation rates showed that 7 nm platelets catalyzed this oxidation over an order of magnitude more quickly than 37 nm platelets. It was suggested that quantum effects probably did not play a strong role, due to strong localized bonding in haematite. However, previous calculations and measurements indicated that the surface oxygen atoms on nanoparticulate hematite would have increased Lewis basicity, enabling them to donate electron density to Mn2+ to catalyze the reaction (Noguera et al., 2002).

3.4

Size Effects in Nanoparticle Sorption Processes

In nature, the sorption of metals and organics to inorganic surfaces (mineral surfaces) can greatly influence their mobility (Hochella et al., 2008; Kretzschmar and Schafer, 2005). As inorganic nanoparticles offer a large amount of surface area relative to their volume or weight, it is expected that they would participate in sorption phenomena. Indeed, heavy metals and radionuclides have been found associated with nanoscale colloids in natural water (Hochella et al., 2005a, 2005b; Hochella and Madden, 2005; Kersting et al., 1999) and drinking water (Wigginton et al., 2007). The sorptive properties of nanoparticles have caught the attention of chemists and engineers, who are interested in using them for environmental remediation (Yavuz et al., 2006; Jeong et al., 2007; Yuan, 2004). Predicting the sorption behaviour of nanoparticles is of interest when considering both natural nanoparticles and the accidental or purposeful release of synthetic nanoparticles into natural systems. Studies have shown that on a per-mass basis, nanoparticles sorb more than their bulk counterparts (Zhang et al., 1999; Yean et al., 2005; Waychunas et al., 2005; Madden et al., 2006; Gao et al., 2004; Giammar et al., 2007). Also, size effects (independent of surface area) are expected for nanoparticle sorption. Experiments have shown that for nanoparticulate titanium dioxide and α-Fe2O3, the point of zero charge, or the pH at which the particles have zero charge, is shifted with respect to size (Guzman et al., 2006; He et al., 2008). The surface energy of nanoparticles is likely to vary with size (Zhang et al., 1999). Also, nanoparticle structures are often different from those of the bulk, displaying lowered atomic coordination and higher disorder (Rockenberger et al., 1997, 1998; Hamad et al., 1999; Marcus et al., 1991; Aruguete et al., 2007); different crystalline phases may be favoured in the nanoscale as opposed to the bulk (Dinega and Bawendi, 1999). All of these phenomena could presumably alter sorption capacity and affinity. Studies of sorption size dependence show varying results. Examples of these are summarized in Table 3.1. Many of these studies have fit their data to the Langmuir adsorption equation (McBride, 1994, Drever, 1997):

11.72 nm, 20 nm, 300 nm

As (V) and As (III)

Hg2+

5 nm, 25 nm, 75 nm Γmax for 11.72 nm particles greater

SA-normalized sorption onto 5 nm < 25 nm = 75 nm

Cu2+

7 nm, 25 nm, 88 nm

Surface geometry of atoms less favourable for bonding; charge distribution More surface sites due to less aggregation

Distorted bonding geometry at intersections of crystal faces

Kads and Γmax much larger for bulk than NP’s SA normalized sorption onto 7 nm > 25 nm = 88 nm

Pb2+

20–33 nm vs 520 nm

* One or more samples contained rutile-phase TiO2. Note: SA = surface area, NP = nanoparticle, N/A = hysteresis data not available, NOM = natural organic matter.

Fe3O4 (Yean et al., 2005)

TiO2, anatase* (Giammar et al., 2007) α-Fe2O3, hematite (Madden et al., 2006) α-FeOOH (Waychunas et al., 2005)

Cd2+

8–145 nm

Molar surface free energy higher for smaller particles Intraparticle electrostatic repulsion in aggregated NPs; Ti site disorder in small NPs Surface geometry and bonding

Kads (6 nm) > Kads (16 nm) (up to 70-fold increase) Kads much larger for bulk than for NP’s

Organic acids

6 nm, 16 nm

TiO2, anatase (Zhang et al., 1999) TiO2, anatase* (Gao et al., 2004)

Explanation suggested

Size dependence

Adsorbate

Particle sizes

Summary of particle size dependence adsorption studies.

Material

Table 3.1

Yes, greater for 20 nm than 300 nm

N/A

N/A

N/A

No, completely reversible

N/A

Hysteresis?

20 nm, 300 nm variable morphology, aggregated. NOM decreased As adsorption.

Aggregation state and morphology of particles not characterized 88 nm sample had variable morphology and more aggregation

Degree of aggregation not quantitatively assessed

Other comments

Size/Shape–Property Relationships

Γ = Γ max

89

KadsC 1 + KadsC

where C is the activity (effective concentration) of the adsorbate in solution, Kads is the adsorption reaction constant (related to the free energy change for the adsorption reaction), Γ is the number of molecules sorbed per unit area and Γmax is the adsorption capacity (the maximum number of molecules/ions per unit area that can be adsorbed). These parameters are referred to in the studies when applicable. (It should be noted that the Langmuir adsorption equation describes adsorption on a homogeneous surface. Because nanoparticle surfaces are relatively heterogeneous by their very nature, parameters derived using this model should be interpreted with caution.) As with many studies in the emerging field of nano-environmental science, interpretation of the results for size dependent trends is complicated by variation within and/or among the samples. The samples often not only vary in size but also in morphology, aggregation state and even crystal phase. All of these variables can affect particle behaviour. While these added variables can be accounted for via careful characterization, it is not always simple to do, and in some cases may be impossible. As synthetic methods advance further, it should become easier to synthesize or purchase more homogeneous and well characterised nanoparticle samples, and the results from such studies should become easier to interpret and compare.

3.5

Nanoparticle Fate: Dissolution and Solid State Cation Movement

Currently, the fate and degradation pathways of nanoparticles are unknown. One possible fate for nanoparticles is for them to dissolve. As many nanoparticles may contain toxic metals, this is a matter of concern. Here, what is known about nanoparticle dissolution is discussed, especially with respect to size and shape. Solid state cation movement and exchange processes are discussed as well, as these may also alter nanoparticle fate in the environment. 3.5.1

Basic Energetic and Kinetic Considerations of Nanoparticle Dissolution

Classically, the dependence of solubility upon particle size, assuming a spherical particle, can be expressed with a modified form of the Kelvin equation: S  2γ V  = exp   S0  RTr  where S is the solubility of particles with inscribed radius r in m, S0 is the solubility of the bulk material, γ is the surface free energy in mJ/m2, R is the gas constant in – nJ/mol⋅K, T is the temperature in K and V is the molecular volume in m3/mol (Adamson, 1982). According to this relation, as the particle dimensions decrease, the solubility increases exponentially relative to the bulk solubility. An example of

90

Environmental and Human Health Impacts of Nanotechnology

Figure 3.4 The deviation of the solubility of small grains of quartz relative to its bulk solubility (S/S0) as a function of the size of the quartz grains being dissolved according to the modified form of the Kelvin equation. The following values were used to produce this curve: T = 298 K, v¯ = 22.68 × 10−6 m3/mol, γ = 350 vmJ/m2. At a particle radius of 100 nm, the solubility is indistinguishable from the bulk value. By the time the particle radius is reduced to 1 nm, the predicted solubility is nearly three orders of magnitude higher. (Reprinted from M.F. Hochella Jr., Nanoscience and technology: the next revolution in the Earth Sciences, Earth and Planetary Science Letters, 203, 593–605, Copyright 2002, with permission from Elsevier.)

this relation is shown in Figure 3.4, which is a plot of

S versus particle radius S0

¯ values for quartz (Hochella, 2002). assuming γ and V Dissolution is generally assumed to be a spontaneous process. As long as particles are in a solution of constant undersaturation, the rate of dissolution should be constant. The relation of the normalized dissolution rate (in mol⋅m−2⋅min−1), R, can be related to the undersaturation, σ, via the relationship: R = kσ n

where k is the rate constant and n is the effective reaction order (Christoffersen et al., 1994; Budz and Nancollas, 1988). From these classical models of dissolution, smaller nanoparticles would be expected to dissolve more quickly than larger particles, and to dissolve to completion. For a number of systems, including nanoparticles of titanium dioxide (Schmidt and Vogelsberger, 2006), silica (Roelofs and Vogelsberger, 2004) and zinc oxide (Yang and Xie, 2006), smaller nanoparticles dissolve more quickly than larger nanoparticles. Despite this, it is not always clear whether classical models apply to

Size/Shape–Property Relationships

91

all nanoparticulate systems. In some systems, chemical processes not included in the classical models, such as photocatalyzed oxidation, may affect dissolution (Stouwdam et al., 2007; Aldana et al., 2001). Other experimental results indicate that small size does not always result in higher rates of dissolution. In one study of zinc oxides in aqueous systems, the same mass of nanoparticles and bulk solids dissolved at the same rate, even though the increased surface area and smaller size of the nanoparticles would warrant otherwise (Franklin et al., 2007). On the other hand, the nanoparticles in the experiment were highly aggregated, which may have lessened surface or size related effects. In some cases, dissolution at the nanoscale may be slower. Indeed, studies on various calcium phosphate minerals (Tang et al., 2001, 2003, 2004a, 2004b, 2004c, 2005; Tang and Nancollas, 2002) display a phenomenon of self-inhibited dissolution occurring primarily at the nanoscale, in which dissolution rates dwindle over time. To understand one of the means by which inhibited dissolution is possible, it is useful to consider the opposite process of nanoparticle growth from a solution. The energy of particle formation, ∆Gform, can be expressed as: ∆Gform = ∆Gv + ∆Gs where ∆Gv is the negative energy term describing the spontaneous tendency of solute to precipitate as part of a solid particle, and ∆Gs describes the excess free energy to form a new solid-liquid interface. ∆Gv depends upon the degree of saturation of the solvent. Assuming spherical particle morphology, both energy terms are functions of r, the radius of the particle. For a given level of saturation, there is a critical radius, r*, above which the magnitude of ∆Gv will be greater than that of ∆Gs and a particle can form (Tang et al., 2001). An analogous critical radius is believed to exist for dissolution processes. In dissolution, which occurs in an undersaturated solution, there is a favourable energetic driving force for units of the solid particle to become solute. However, the formation of etch pits can increase the area of the solid–liquid interface, which is energetically disfavoured. For such a system, there is a critical radius (of etch pit) at which dissolution is energetically allowed. Below this radius, dissolution is inhibited. Such inhibition phenomena have been observed in a number of systems and particularly well studied for various biologically relevant calcium phosphates (Tang et al., 2003, 2004a, 2004b, 2004c, 2005; Tang and Nancollas, 2002). It is therefore reasonable to expect the possibility of such inhibited dissolution occurring for nanoparticles of sparingly soluble compounds, because the nanoparticle dimensions may be below this critical radius. Currently, there is limited data to confirm or disprove these expectations. 3.5.2

Effects of Nanoparticle Morphology

Nanoparticles released into the environment will not only vary in size but also in their morphology, which may strongly affect dissolution. This particularly applies to cases in which the nanoparticles are crystalline. When nanoparticles of the same

92

Environmental and Human Health Impacts of Nanotechnology

crystalline substance assume different shapes, this generally means that different crystal faces comprise their surfaces. For example, consider nanoparticles of a rock salt structured mineral. A cubic nanoparticle displays {100} faces, a truncated cubo-octahedron displays {100}, {111} and {110} faces, and an octahedral nanoparticle will display {111} faces. Different crystal faces will be more or less stable (have different surface energies), depending upon their surface bonding. It is expected that less stable faces would be etched more readily than more stable faces. In our example, assuming that all other conditions are equal (same crystal structure, composition, solution undersaturation, etc.), this would mean that the three differently shaped nanoparticles might dissolve at different rates. While the energetic stability of crystal surfaces affects dissolving crystals of all sizes, it is particularly important for nanoparticles because even minimal dissolution may result in their annihilation. In principle, these concepts are simple, but applying them to quantitatively predict morphology dependent dissolution trends in nanoparticles is difficult. This is because little is known regarding the relative stabilities of nanoscale surfaces. The presence of surface defects, steps or kinks, which may be more evident on nanoparticle surfaces, will also influence the energetics of dissolution. Another complicating factor is the presence of coatings or other external substances, which are discussed in the following section. 3.5.3

Effects of Nanoparticle Coatings and External Substances

As with the surfaces of bulk materials (Zhang and Nancollas, 1990; Casey and Ludwig, 1995; Becker et al., 2005), it has been shown that external substances, particularly those that can coordinate to nanoparticle surfaces, can strongly influence nanoparticle growth and dissolution (Jun et al., 2006; Li et al., 2005, 2006b; Yin and Alivisatos, 2005). Anthropogenic nanoparticles released into the environment are likely to encounter many substances that could interact strongly with or sorb onto their surfaces, and many will already have coatings on their surfaces. One way such coatings or sorbed species may affect dissolution is by stabilizing particular crystal surfaces. Consider the partial dissolution of a truncated cubooctahedral nanoparticle composed of the rock salt structured material introduced in Section 3.5.2. The reaction coordinate for this process is displayed in Figure 3.5. Thermodynamically, the most favoured end product of dissolution for this system is a sphere. (This is not to imply that a sphere is always the most favoured shape for every system.) Imagine now adding a substance to the solution of nanoparticles which binds to and stabilizes the {100} and {110} crystal faces. This will increase the activation energy needed to obtain a spherical nanoparticle. Unless there is enough thermal energy in the system to surmount this kinetic barrier, it is likely that the process with the lower kinetic barrier (lower activation energy) will dominate. In this scenario, the {111} faces are energetically unstable relative to the other faces, so they will etch more readily. This etching results in an octahedrally shaped particle rather than a spherical particle. Coatings or external compounds can affect dissolution in other ways. For example, a coating that forms a micellar structure around a nanoparticle might reduce the

Size/Shape–Property Relationships

∆G†stab

Energy

{111}

93

{100} {110}

∆G†111

∆G†nostab Truncated cubo-octahedron

All {111} faces

Octahedron

sphere

Reaction coordinate Figure 3.5 Reaction coordinate for the partial dissolution (etching) of a truncated cubooctahedral nanoparticle. For this system, the thermodynamically favoured product is a spherical nanoparticle. In the absence of stabilizing coatings or ligands, formation of a spherical nanoparticle is kinetically favoured as well (the activation energy ∆G†nostab < ∆G†111 ). In the presence of compounds that stabilize the {100} and {110} faces, the activation energy to obtain a spherical nanoparticle increases. In this case, since this activation energy is greater than that necessary to form an octahedron (energy ∆G†stab > ∆G†111 ), formation of a sphere is no longer kinetically favoured. Etching the non-stabilized {111} faces is kinetically favoured. In the absence of enough thermal energy in the system, etching will result in an octahedral end product rather than a spherical one. (Figure adapted from Y.W. Jun, J.S. Choi and J.W. Cheon (2006) Shape control of semiconductor and metal oxide nanocrystals through nonhydrolytical colloidal routes. Angewandte Chemie, 45, 3414–39. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

activity of water at the surface. As for external compounds in the solution surrounding the nanoparticle, there are experimental examples of dissolution rates being increased by compounds such as acetic acid (Meulenkamp, 1998) or human serum albumin protein (Yang and Xie, 2006). External compounds might form stable ionic complexes with the constituent metal ions in the nanoparticle, hence energetically favouring dissolution. They also may alter pH, which again can affect nanoparticle stability. It is evident from such considerations that in any study concerning nanoparticle dissolution, as much as possible should be known regarding the composition of the coatings or the compounds in the solution surrounding the nanoparticle. These characteristics can be as significant as the composition of the inorganic part of the nanoparticle.

94

3.5.4

Environmental and Human Health Impacts of Nanotechnology

Case Study: The Dissolution of Lead Sulfide Nanoparticles

From the previous discussion of nanoparticle dissolution, it is evident that not only is size important, but morphology, coatings and molecules present in the surrounding solution. Currently, in our laboratory, the first two factors are being studied. The non-oxidative dissolution of ∼15 nm diameter lead(II) sulfide (PbS) nanoparticles in hydrochloric acid (pH 3) is being examined (Liu et al., 2007a). Bright-field transmission electron microscopy (TEM) is used to track changes in particle size and high-resolution TEM is used to measure changes in morphology and structure. The dissolution of nano-sized lead sulfide (galena) may have implications for the behaviour of both synthetic and natural nanomaterials in the environment. Lead sulfide is a low band gap semiconductor used in applications such as infrared detectors. Nanoparticles of lead sulfide are popular in nanoscience research and are commercially available. As for natural systems, it is known that nanoparticulate metal sulfides are present in some environments, and that mineral nanoparticles may be involved in the transport of heavy metals (Hochella et al., 2005b, 2008; Labrenz et al., 2000). Lead sulfide nanoparticles are synthesized under inert atmosphere in organic solution with surfactant via a previously published procedure at high temperature (Joo et al., 2003). This synthetic procedure produces monodispersed, highly crystalline nanoparticles, as confirmed with TEM and X-ray diffraction (XRD). After an initial washing procedure to remove excess free surfactant, nanoparticles are deposited onto a carbon/gold TEM grid substrate. Having the particles on a substrate helps to prevent aggregation, as this would complicate analysis. X-ray photoelectron spectroscopy (XPS) confirms that subsequent washing steps remove the majority of the surfactant (although undetectable trace amounts may remain) and that washing does not significantly affect the presence of any oxidation species on the nanoparticle surfaces. Washed, dried grids are exposed to nitrogen-purged hydrochloric acid solutions (pH 3) under constant stirring for varying periods. Images from samples exposed to the acid for different times are compared with each other using TEM measurements. Two interesting trends are summarized here. Firstly, the morphology of the lead sulfide nanoparticles changes after dissolution. From high resolution TEM measurements (Figure 3.6), the {110} and {111} faces are being etched more quickly than the {100} faces ({111} (data not shown). Such results match what might be expected from our knowledge of bulk crystals. Generally, on a crystal face, the rate at which an atom is removed from that face is inversely proportional to the number of bonds it has (Lasaga and Luttge, 2004). Atoms in the ideal bulk {110} and {111} faces have surface atomic coordination numbers of four and three, respectively, while the {100} faces have an atomic coordination number of five. Therefore, it would be expected that the {100} faces would etch more slowly than the {111} or {110} faces. At least for this system, these results indicate that some of our current knowledge about bulk crystal surfaces can be used to predict how nanoparticles might behave in the environment. Secondly, lead sulfide nanoparticles have been found to dissolve at surface area normalized rates higher than those for bulk lead sulfide by approximately one to two orders of magnitude. This difference in dissolution rate may be attributed to

Size/Shape–Property Relationships {110 }

0} {10

} 10 {1

(a)

95

{100}

(b)

{100}

d100

{1 10 }

d110

{100} { 11 0} d100

d110

(c)

Figure 3.6 (a) High resolution TEM images of a nanoparticle before dissolution (left) and a nanoparticle after 2 hours of dissolution (right). Note how the sizes of the {110} faces have increased. (b) Schematic diagrams of the distances measured on particles to determine whether the change in {110} size is statistically significant. (c) Distance ratios of d100/d110 before and after dissolution, along with mean values and 95% confidence intervals. As the {110} face size increases after 2 hours of dissolution, the value of the ratio increases.

the small size of the nanocrystals. As mentioned earlier, the modified Kelvin equation indicates that dissolution is more thermodynamically favoured for smaller particles. Also, due to their size, nanocrystals have a larger fraction of their atoms at corners and edges than bulk crystals. Such undercoordinated atoms are more active in dissolution than ones from flat surfaces. Nanoparticle morphology may

96

Environmental and Human Health Impacts of Nanotechnology

also play a role in faster dissolution. While bulk natural lead sulfide mostly displays {100} faces, the lead sulfide nanocrystals exhibit {111} and {110} faces. As described above, these faces dissolved more quickly than the {100} faces. These initial results have important implications for the dissolution behaviour of nanoparticles in the environment. Larger micro-sized particles for further size comparative rate studies are currently in the process of being synthesized. 3.5.5

Solid State Cation Movement in Nanoparticles

Another phenomenon that may affect nanoparticle degradation and fate is solid state cation movement into or out of nanoparticles. One type of cation movement is cation exchange, in which cations in solution replace cations in a lattice. Even if cation exchange does not occur significantly in the bulk form of a particular material (excluding perhaps on its surfaces), this does not preclude this process from happening fully in the nanoparticulate form or in thin films (nanoscaled films <100 nm in thickness). In the bulk forms of such materials, cation exchange is kinetically controlled by the advancement of a reaction zone, along which cations and vacancies travel. If nanoparticles are small enough, they may be as large as or smaller than this minimum reaction zone, resulting in faster cation exchange (Son et al., 2004). One fascinating example of this phenomenon is the room temperature complete exchange of silver ions for cadmium in CdSe nanoparticles. Son and co-workers synthesized spherical CdSe nanoparticles (4.2 nm in diameter) and rod shaped CdSe nanoparticles (varying dimensions) and mixed them with solutions of silver nitrate (AgNO3). (While this was done in a toluene–methanol mixture, it should be noted that cation exchange in nanostructures has also been observed for aqueous systems (Mews et al., 1994; Lokhande et al., 1992; Dloczik and Koenenkamp, 2004).) Within about 100 ms (Chan et al., 2007), CdSe nanospheres are transformed into Ag2Se nanospheres as shown in Figure 3.7. The exchange could be subsequently reversed by adding an excess of cadmium ions and a compound that forms a stable complex with silver (tributylphosphine). Notably, when the same reaction was attempted with micrometre sized powders of CdSe, cation exchange was prohibited, even over weeks. Cation exchange in this system has also been tested with Pb2+ and Cu2+, and has been demonstrated to occur with a variety of ions in metal sulfides in both nanoparticles and thin films (Robinson et al., 2007; Dloczik and Koenenkamp, 2004; Lokhande et al., 1992; Mews et al., 1994). Another interesting aspect of cation movement in nanoparticles is that the morphology of particles can be altered. As morphology can affect the chemical and physical behaviour of nanoparticles, this has important implications. Two examples are displayed in Figure 3.8. In the CdSe to Ag2Se conversion as described above, smaller nanorods are converted to nanospheres. Another example of shape change due to cation diffusion is in the synthesis of cobalt sulphide from metallic cobalt nanospheres (Yin et al., 2004, 2006). When reacting with elemental sulfur, the outwards diffusion of cobalt ions produces a hollow nanosphere. It should be noted that such cation movement processes are contingent upon favourable thermodynamic driving forces, as well as factors such as the rate of dif-

Size/Shape–Property Relationships

97

Figure 3.7 TEM images of (A) initial CdSe (diameter 4.2 nm), (B) Ag2Se transformed from the forward cation exchange reaction and (C) recovered CdSe nanocrystals from the reverse cation exchange reaction. (D to F) XRD patterns, fluorescence emission and optical absorption spectra of initial CdSe (top diffractogram or spectrum), Ag2Se (middle) and recovered CdSe (bottom) nanocrystals, respectively. In the recovered CdSe, the peak positions of the emission and absorption show a slight red shift from those of the initial CdSe, which becomes negligible for nanospheres larger than 6 nm in diameter. An additional fluorescence emission feature near 700 nm seen in the recovered CdSe (E) is due to the increased surface trap emission. Vertical lines in (E) and (F) are a guide for the comparison of peak positions. (Figure and caption from Son et al. (2004), Science, 306, 1009–12. Reprinted with permission from AAAS.)

98

Environmental and Human Health Impacts of Nanotechnology

Figure 3.8 Examples of shape change after cation exchange reactions. A and B: Conversion of nanorods to nanospheres when exchanging silver for cadmium. C and D: Conversion of solid cobalt nanospheres to hollow cobalt sulfide nanospheres, due to rapid diffusion of cobalt ions. (A and B from Son et al. (2004) Science, 306, 1009–12. C and D from Yin et al. (2004) Science, 304, 711–4. Reprinted with permission from AAAS.)

fusion for a particular atom in a given solid lattice. Cation movement will not necessarily occur for every sort of nanoparticle. Nevertheless, as nanoparticles will doubtless encounter metal ions when they are released into natural systems, it is important to keep these processes in mind.

3.6 Effect of Nanoparticle Aggregation on Physical and Chemical Properties Another structural characteristic that may well impact the behaviour and fate of nanoparticles is their degree of aggregation. It is well established that under the proper conditions nanoparticles can spontaneously self-assemble or aggregate (Shipway et al., 2000; He et al., 2008; Guzman et al., 2006; Gilbert et al., 2007; Moreau et al., 2007). Once such nanoparticles are released into chemically complex natural systems, it is reasonable to expect that some of them will aggregate and that aggregated nanoparticle systems may be as common as dispersed systems. Some examples of how aggregation may affect nanoparticle behaviour are briefly discussed here. The processes controlling aggregation are given in Chapter 4.

Size/Shape–Property Relationships

99

There are many indications that nanoparticle aggregates behave differently from their well dispersed counterparts. For example, 3 nm zinc sulfide nanoparticles have more highly ordered crystal structures when aggregated (Huang et al., 2004). The activation energies for phase transitions in titanium dioxide nanoparticles were found to be lower for less aggregated nanoparticles. This was attributed to a lower surface energy of the more aggregated nanoparticles due to interparticle interactions (Zhang and Banfield, 2007). Aggregation has also been shown to affect the thermal conductivity of nanoparticle solutions (Hong et al., 2006). Recently, aggregates of 9 nm ferrihydrite (Fe3O4) nanoparticles were found to reduce carbon tetrachloride, a common organic contaminant, more slowly than non-aggregated ferrihydrite (Vikesland et al., 2007). As per Table 3.1, in studies of metal sorption onto nanoparticles, in some cases it was suggested that the size dependent differences observed were due to aggregation (Gao et al., 2004). One of the well known effects of aggregation is its contribution to nanoparticle growth (Banfield et al., 2000; Guyodo et al., 2003; Penn and Banfield, 1998, 1999). Simply influencing growth is very important, as changing nanoparticle size may mean changing physical and chemical properties. Aggregative growth processes also may have an impact upon metal sequestration. Waychunas and co-workers (2005) found that when Zn2+ was added to mixtures of α-FeOOH (goethite) nanoparticles during aggregation, zinc was either incorporated directly into the crystal or that a secondary zinc-containing precipitate was formed. X-ray absorption spectroscopy (XAS) results from Zn2+ sorbed to the nanoparticles after aggregation were very different (Waychunas et al., 2005). The connection between degree of aggregation and chemical/physical properties is quite intriguing and invites further exploration.

3.7 Environmental Implications: General Discussion, Recommendations and Outlook Nanoparticles have size dependent properties that may affect many environmental processes. Nevertheless, it still remains difficult to predict what these actual environmental effects might be. It is clear that more research will be conducted regarding the fate and impact of nanoparticles in the environment with size dependent properties in mind. Here, commentary and recommendations on future directions in this research are discussed. First, it is critical to establish how and where nanoparticles are being released into the environment. This topic is addressed for atmospheric nanoparticles in Chapter 5, but this area continues to merit much more study. Knowing the sources and means of entry for nanoparticles will reveal which environmental exposure scenarios are the most likely. The knowledge will in turn guide researchers attempting to study the most environmentally relevant nanoparticle systems. As many nanoparticle-based technologies are still emerging as they will for some time, various release scenarios may have to be postulated. Some examples of nanoparticles in current and emerging technologies likely to be released (or they are already being released) into the environment are as follows:

100

Environmental and Human Health Impacts of Nanotechnology

(i)

Titanium dioxide and other photoactive semiconductor nanoparticles (e.g. CdSe, PbS, MoS2). Titanium dioxide and zinc oxide nanoparticles are already available in commercial sunscreen creams and multiple applications are being developed using semiconductor nanoparticles, including sensors, cells for solar power, water-splitting for hydrogen gas generation, air cleaners, and so on. A recent report on exposure modelling of manufactured nanoparticles in the environment (in Switzerland) indicated that nano-TiO2 is a substance of concern (Mueller and Nowack, 2008). (ii) Silver nanoparticles. Silver nanoparticle coatings are already on many commercially-available products (e.g. bandages, clothing) as antibacterial barriers. While Mueller and Nowack’s (2008) environmental modelling, mentioned above, concluded that nano-sized silver pose little risk in Switzerland based upon present data, the widespread use of nano-sized silver calls for additional study. (iii) Nano zero-valent iron particles. As these nanoparticles have proven to be effective at decomposing halogenated organics in the field, their continued and potentially increasing use is likely. Aside from knowing entry routes of nanoparticles into the environment, researchers will need to build a comprehensive knowledge base linking aspects of nanoparticle structure or chemistry (e.g. size, shape, coating, composition) to their behaviour in a simulated or actual environmental system. This is important because an enormous number of nanoparticle types are possible and it will be impossible to test every single one of them. Understanding trends in nanoparticle behaviour provides the ability to predict how different nanoparticles behave based upon their characteristics, even in the absence of hard data for a particular nanoparticle. Science policy makers and regulatory agencies will find this ability to predict behaviour especially useful. Towards the goal of linking nanoparticle structure and chemistry to behaviour, it is necessary for studies to use exceptionally well characterized nanoparticles. In addition to particle size and composition, features such as particle shape, coatings and aggregation state should be carefully measured or controlled if possible. Researchers should also account for any known impurities. Preferably, in any study connecting nanoparticle characteristics to behaviour, nanoparticle features should be varied carefully. For example, in a study of size dependence in nanoparticles, researchers should strive to maintain the same morphology, aggregation state and coatings between different sizes. If this is not possible due to synthetic limitations, the other varying characteristics must be taken into account when interpreting data. Given the need for well characterized, pure nanoparticles, it may behove environmental scientists to synthesize and characterize their own materials or to collaborate with nanochemistry specialists. In addition to beginning studies with well defined, well characterized nanoparticles, if possible it is important to be able to characterize the nanoparticle structure and chemical state throughout the study. Nanoparticles may change during the study, altering their morphology, degrading (e.g. oxidizing), aggregating or losing coatings. Such changes will doubtless alter the physical and chemical behaviour of

Size/Shape–Property Relationships

101

the nanoparticles. Characterizing the structures and chemical state of nanoparticles well will enable researchers to better explain changes in their behaviour. In actual natural systems, nanoparticles will have many opportunities to interact with substances such as organic matter or other natural particles, and may form complex assemblages that behave very differently from the original, pure nanoparticles. Characterization of these assemblages will doubtless be important for understanding their behaviour.

3.8

References

Abrams, B. L. and J. P. Wilcoxon (2005) Nanosize semiconductors for photooxidation. Critical Reviews in Solid State and Materials Sciences, 30 (3), 153–182. Adamson, A. W. (1982) Physical Chemistry of Surfaces, (4th edn); Wiley, New York. Aldana, J., Y. A. Wang and X. G. Peng (2001) Photochemical instability of CdSe nanocrystals coated by hydrophilic thiols. Journal of the American Chemical Society, 123 (36), 8844–50. Alivisatos, A. P. (1996) Perspectives on the physical chemistry of semiconductor nanocrystals. Journal of Physical Chemistry, 100 (31), 13226–39. Anschutz, A. J. and R. L. Penn (2005) Reduction of crystalline iron(III) oxyhydroxides using hydroquinone: Influence of phase and particle size. Geochemical Transactions, 6 (3), 60–6. Aruguete, D. M. M. A. Marcus , L. S. Li, et al. (2007) Surface structure of CdSe nanorods revealed by combined X-ray absorption fine structure measurements and ab initio calculations. Journal of Physical Chemistry C, (1), 75–9. Atkins, P. W. P. W. (1940-, 1998). Physical chemistry, xvi, 999 p.. Freeman, New York. Banfield, J. F., S. A. Welch, H. Z. Zhang, et al. (2000) Aggregation-based crystal growth and microstructure development in natural iron oxyhydroxide biomineralization products. Science, 289 (5480), 751–4. Becker, U., S. Biswas, T. Kendall, et al. (2005) Interactions between mineral surfaces and dissolved species: From monovalent ions to complex organic molecules. American Journal of Science, 305 (6–8), 791–825. Becker, W. M. and D. W. Deamer (1986) The World of the Cell. Benjamin/Cummings Publishing Company, Inc., Redwood City, California. Brown, G. E., V. E. Henrich, W. H. Casey, et al. (1999) Metal oxide surfaces and their interactions with aqueous solutions and microbial organisms. Chemical Reviews, 99, 77–174. Bruchez, M., M. Moronne, P. Gin, S. Weiss and A. P. Alivisatos (1998) Semiconductor nanocrystals as fluorescent biological labels. Science, 281 (5385), 2013–16. Budz, J. A. and G. H. Nancollas (1988) The mechanism of dissolution of hydroxyapatite and carbonated apatite in acidic solutions. Journal of Crystal Growth, 91 (4), 490–6. Casey, W. H. and C. Ludwig (1995) Silicate mineral dissolution as a ligand-exchange reaction. In: A. F. White and S. L. Brantley (Editors), Chemical Weathering Rates of Silicate Minerals. Reviews in Mineralogy. Mineralogical Society of America, Washington, D.C., pp. 583. Chan, E. M., M. A. Marcus, S. Fakra, et al. (2007) Millisecond kinetics of nanocrystal cation exchange using microfluidic X-ray absorption spectroscopy. Journal of Physical Chemistry A, 111 (49), 12210–5. Cornell, R. M. and U. Schwertmann (2003) The Iron Oxides: Structure, Properties, Reactions, Occurences and Uses. Wiley-VCH, Weinheim, 664 pp. Cozzoli, P. D., E. Fanizza, R. Comparelli, et al. (2004) Role of Metal Nanoparticles in TiO2/ Ag Nanocomposite-Based Microheterogeneous Photocatalysis. Journal of Physical Chemistry B, 108, 9623–30. Dabbousi, B. O., J. Rodriguez-Viejo, F. V. Mikulec, et al. (1997) (CdSe)ZnS core-shell quantum dots: Synthesis and characterization of a size series of highly luminescent nanocrystallites. Journal of Physical Chemistry B, 101 (46), 9463–75.

102

Environmental and Human Health Impacts of Nanotechnology

Davison, W. and R. De Vitre (1992) Iron Particles in Freshwater. In: J. Buffle and H. P. Van Leeuwen (Editors), Environmental Particles. Lewis Publishers, Ann Arbor, MI, pp. 315–55. Deng, J. P., W. C. Shih and C. Y. Mou (2005) Hydrogenation of anthracene catalyzed by surfactant-protected gold nanoparticles in aqueous solution: Size dependence. Chemphyschem, 6 (10), 2021–25. Deng, J. P., W. C. Shih, and C. Y. Mou (2007) Electron transfer-induced hydrogenation of anthracene catalyzed by gold and silver nanoparticles. Journal of Physical Chemistry C, 111 (27), 9723–8. Dinega, D. P. and M. G. Bawendi (1999) A solution-phase chemical approach to a new crystal structure of cobalt. Angewandte Chemie-International Edition, 38 (12), 1788–91. Dloczik, L. and R. Koenenkamp (2004) Nanostructured metal sulfide surfaces by ion exchange processes. Journal of Solid State Electrochemistry, 8 (3), 142–6. Drever, J. I. (1997) The Geochemistry of Natural Waters. Prentice Hall, Upper Saddle River, NJ. Duan, C. F., H. Chi, Z. F. Zhang, B. Liu, et al. (2007) Size-dependent inhibition and enhancement by gold nanoparticles of luminol-ferricyanide chemiluminescence. Journal of Physical Chemistry C, 111 (12), 4561–6. Dzombak, D. A. and F. M. M. Morel (1990) Surface complexation modeling: hydrous ferric oxide. Wiley, New York. Elliott, D. W. and W. X. Zhang (2001) Field assessment of nanoscale bimetallic particles for groundwater treatment. Environmental Science & Technology, 35, 4922–6. El-Sayed, I., X. Huang, and M. El-Sayed (2007) Multicolorimetric plasmonic gold nanoparticles for 8 optical detection of oral squamous carcinoma. Oral Oncology: 121–121. El-Sayed, I. H., X. H. Huang and M. A. El-Sayed (2005) Surface plasmon resonance scattering and absorption of anti-EGFR antibody conjugated gold nanoparticles in cancer diagnostics: Applications in oral cancer. Nano Letters, 5 (5), 829–34. Faust, B. C., M. R. Hoffmann and D. W. Bahnemann (1989) Photocatalytic oxidation of sulfur dioxide in aqueous suspensions of alpha-Fe2O3. Journal of Physical Chemistry, 93, 6371–81. Filip, J. R., Zboril, O. Schneeweiss, J. Zeman, et al. (2007) Environmental applications of chemically pure natural ferrihydrite. Environmental Science & Technology, 41 (12), 4367–74. Franklin, N. M., N. J. Rogers, S. C. Apte, et al. (2007) Comparative toxicity of nanoparticulate ZnO, bulk ZnO, and ZnCl2 to a freshwater microalga (Pseudokirchneriella subcapitata): The importance of particle solubility. Environmental Science & Technology, 41, 8484–90. Gao, L. Z., J. Zhuang, L. Nie, et al. (2007) Intrinsic peroxidase-like activity of ferromagnetic nanoparticles. Nature Nanotechnology, 2, 577–83. Gao, X. H., L. L. Yang, J. A. Petros, et al. (2005) In vivo molecular and cellular imaging with quantum dots. Current Opinion in Biotechnology, 16 (1), 63–72. Gao, Y., R. Wahi, A. T. Kan, et al. (2004) Adsorption of cadmium on anatase nanoparticleseffect of crystal size and pH. Langmuir, 20 (22), 9585–93. Giammar, D. E., C. J. Maus and L. Y. Xie (2007) Effects of particle size and crystalline phase on lead adsorption to titanium dioxide nanoparticles. Environmental Engineering Science, 24 (1), 85–95. Giasuddin, A. B. M., S. R. Kanel and H. Choi (2007) Adsorption of humic acid onto nanoscale zerovalent iron and its effect on arsenic removal. Environmental Science & Technology, 41 (6), 2022–7. Gilbert, B., G. P. Lu and C. S. Kim (2007) Stable cluster formation in aqueous suspensions of iron oxyhydroxide nanoparticles. Journal of Colloid and Interface Science, 313 (1), 152–9. Guyodo, Y., A. Mostrom, R. L. Penn and S. K. Banerjee (2003) From Nanodots to Nanorods: Oriented aggregation and magnetic evolution of nanocrystalline goethite. Geophysical Research Letters, 30 (10), 19-1–19-4.

Size/Shape–Property Relationships

103

Guzman, K. A. D., M. P. Finnegan and J. F. Banfield (2006) Influence of surface potential on aggregation and transport of titania nanoparticles. Environmental Science & Technology, 40 (24), 7688–93. Hagfeldt, A. and M. Gratzel (1995) Light-Induced Redox Reactions in Nanocrystalline Systems. Chemical Reviews, 95 (1), 49–68. Halim, K. S. A., M. H. Khedr, M. I. Nasr and A. M. El-Mansy (2007) Factors affecting CO oxidation over nanosized Fe2O3. Materials Research Bulletin, 42 (4), 731–41. Hamad, K. S., R. Roth, J. Rockenberger, T. van Buuren and A. P. Alivisatos (1999) Structural disorder in colloidal InAs and CdSe nanocrystals observed by X-ray absorption near-edge spectroscopy. Physical Review Letters, 83 (17), 3474–7. Harris, L. A. et al. (2003) Magnetite nanoparticle dispersions stabilized with triblock copolymers. Chemistry of Materials, 15 (6), 1367–77. Hayat, M. A. (Editor) (1989) Colloidal gold: principles, methods, and applications, 2. Academic Press, San Diego. He, Y. T., J. Wan and T. Tokunaga (2008) Kinetic stability of hematite nanoparticles: the effect of particle sizes. Journal of Nanoparticle Research, 10, 321–32. Hermanek, M., R. Zboril, N. Medrik, J. Pechousek and C. Gregor (2007) Catalytic efficiency of iron(III) oxides in decomposition of hydrogen peroxide: Competition between the surface area and crystallinity of nanoparticles. Journal of the American Chemical Society, 129 (35), 10929–36. Hines, M. A. and P. Guyot-Sionnest (1996) Synthesis and characterization of strongly luminescing ZnS-Capped CdSe nanocrystals. Journal of Physical Chemistry, 100 (2), 468–71. Hochella, M. F. (2002) Nanoscience and technology the next revolution in the Earth sciences. Earth and Planetary Science Letters, 203 (2), 593–605. Hochella, M. F., T. Kasama, A. Putnis, et al. (2005a) Environmentally important, poorly crystalline Fe/Mn hydrous oxides: Ferrihydrite and a possibly new vernadite-like mineral from the Clark Fork River Superfund Complex. American Mineralogist, 90 (4), 718–24. Hochella, M. F., J. N. Moore, C. V. Putnis, et al. (2008) Nanominerals, Mineral Nanoparticles, and Earth Systems. Science, 319 (5870), 1631–5. Hochella, M. F. and A. S. Madden (2005) Earth’s nano-compartment for toxic metals. Elements, 1 (4), 199–203. Hochella, M. F., J. N. Moore, C. V. Putnis, et al. (2005b) Direct observation of heavy metal-mineral association from the Clark Fork River Superfund Complex: Implications for metal transport and bioavailability. Geochimica Et Cosmochimica Acta, 69 (7), 1651–63. Hong, K. S., T. K. Hong and H. S. Yang (2006) Thermal conductivity of Fe nanofluids depending on the cluster size of nanoparticles. Applied Physics Letters, 88 (3), 031901-1–031901-3. Houben, G. J. (2003) Iron oxide incrustations in wells. Part 2: chemical dissolution and modeling. Applied Geochemistry, 18 (6), 941–54. Huang, F., B. Gilbert, H. H. Zhang and J. F. Banfield (2004) Reversible, surface-controlled structure transformation in nanoparticles induced by an aggregation state. Physical Review Letters, 92 (15), 155501-1–155501-4. Huang, X., W. Qian, I. H. El-Sayed and M. A. El-Sayed (2007a) The potential use of the enhanced nonlinear properties of gold nanospheres in photothermal cancer therapy. Lasers in Surgery and Medicine, 39, 747–53. Huang, X. H., I. H. El-Sayed, W. Qian and M. A. El-Sayed (2006) Cancer cell imaging and photothermal therapy in the near-infrared region by using gold nanorods. Journal of the American Chemical Society, 128 (6), 2115–20. Huang, X. H., P. K. Jain, I. H. El-Sayed and M. A. El-Sayed (2007b) Gold nanoparticles: interesting optical properties and recent applications in cancer diagnostic and therapy. Nanomedicine, 2, 681–93. Imlay, J. A. (2003) Pathways of oxidative damage. Annual Review of Microbiology, 57, 395–418.

104

Environmental and Human Health Impacts of Nanotechnology

Jeong, H. Y., B. Klaue, J. D. Blum and K. F. Hayes (2007) Sorption of mercuric ion by synthetic nanocrystalline mackinawite (FeS). Environmental Science & Technology, 41 (22), 7699–705. Johnson, T. L., M. M. Scherer and P. G. Tratnyek (1996) Kinetics of halogenated organic compound degradation by iron metal. Environmental Science & Technology, 30 (8), 2634–40. Joo, J., H. B. Na, T. Yu, et al. (2003) Generalized and facile synthesis of semiconducting metal sulfide nanocrystals. Journal of the American Chemical Society, 125 (36), 11100–5. Jun, Y. W., J. S. Choi and J. W. Cheon (2006) Shape control of semiconductor and metal oxide nanocrystals through nonhydrolytical colloidal routes. Angewandte ChemieInternational Edition, 45, 3414–39. Kamat, P. V. and D. Meisel (2002) Nanoparticles in advanced oxidation processes. Current Opinion in Colloid & Interface Science, 7, 282–7. Kersting, A. B., D. W. Efurd, D. L. Finnegan, et al. (1999) Migration of plutonium in ground water at the Nevada Test Site. Nature, 397 (6714), 56–9. Ketchie, W. C., Y. L. Fang, M. S. Wong, M. Murayama and R. J. Davis (2007) Influence of gold particle size on the aqueous-phase oxidation of carbon monoxide and glycerol. Journal of Catalysis, 250 (1), 94–101. Kiwi, J. and M. Gratzel (1979) Protection, size factors, and reaction dynamics of colloidal redox catalysts mediating light induced hydrogen evolution from water. Journal of the American Chemical Society, 101 (24), 7214–6. Kloepfer, J. A., S. E. Bradforth and J. L. Nadeau (2005) Photophysical properties of biologically compatible CdSe quantum dot structures. Journal of Physical Chemistry B, 109 (20), 9996–10003. Korgel, B. A. and H. G. Monbouquette (1997) Quantum confinement effects enable photocatalyzed nitrate reduction at neutral pH using CdS nanocrystals. Journal of Physical Chemistry B, 101 (25), 5010–7. Kretzschmar, R. and T. Schafer (2005) Metal retention and transport on colloidal particles in the environment. Elements, 1 (4), 205–10. Labrenz, M., G. K. Druschel, T. Thomsen-Ebert, et al. (2000) Formation of sphalerite (ZnS) deposits in natural biofilms of sulfate-reducing bacteria. Science, 290 (5497), 1744–7. Larsen, O. and D. Postma (2001) Kinetics of reductive bulk dissolution of lepidocrocite, ferrihydrite, and goethite. Geochimica Et Cosmochimica Acta, 65 (9), 1367–79. Lasaga, A. C. (1998) Kinetic theory in the earth sciences, x, 811. Princeton University Press, Princeton, N.J. Lasaga, A. C. and A. Luttge (2004) Mineralogical approaches to fundamental crystal dissolution kinetics. American Mineralogist, 89 (4), 527–40. Lee, J. H., Y. M. Huh, Y. W. Jun, et al. (2006) Artificially engineered magnetic nanoparticles for ultra-sensitive molecular imaging. Nature Medicine, 13, 95–9. Li, L., M. H. Fan, R. C. Brown, et al. (2006a) Synthesis, properties, and environmental applications of nanoscale iron-based materials: A review. Critical Reviews in Environmental Science and Technology, 36 (5), 405–31. Li, R. F., J. H. Lee, B. C. Yang, et al. (2005) Amine-assisted facetted etching of CdSe nanocrystals. Journal of the American Chemical Society, 127, 2524–32. Li, R. F., J. H. Lee, B. C. Yang, et al. (2006b) Band-edge photoluminescence recovery from zinc-blende CdSe nanocrystals synthesized at room temperature. Advanced Functional Materials, 16 (3), 345–50. Li, X. Q., D. W. Elliott and W. X. Zhang (2006c) Zero-valent iron nanoparticles for abatement of environmental pollutants: Materials and engineering aspects. Critical Reviews in Solid State and Materials Sciences, 31 (4), 111–122. Liou, Y. H., S. L. Lo, W. H. Kuan, C. J. Lin and S. C. Weng (2006) Effect of precursor concentration on the characteristics of nanoscale zerovalent iron and its reactivity of nitrate. Water Research, 40 (13), 2485–92.

Size/Shape–Property Relationships

105

Liu, G., S. Debnath, K. W. Paul, et al. (2006) Characterization and surface reactivity of ferrihydrite nanoparticles assembled in ferritin. Langmuir, 22 (22), 9313–21. Liu, J., D. Aruguete, J. Jinschek and M. F. Hochella (2007a) TEM investigation of the nonoxidative dissolution of galena (PbS) nanoparticles in a hydrochloric acid solution. Geochimica Et Cosmochimica Acta, 71 (15), A589. Liu, Q., Z. M. Cui, Z. Ma, et al. (2007b) Morphology control of Fe2O3 nanocrystals and their application in catalysis. Nanotechnology, 18, 385605-1–385605-5. Liu, Y. Q., H. Choi, D. Dionysiou and G. V. Lowry (2005a) Trichloroethene hydrodechlorination in water by highly disordered monometallic nanoiron. Chemistry of Materials, 17 (21), 5315–22. Liu, Y. Q., S. A. Majetich, R. D. Tilton, D. S. Sholl and G. V. Lowry (2005b) TCE dechlorination rates, pathways, and efficiency of nanoscale iron particles with different properties. Environmental Science & Technology, 39 (5), 1338–45. Lo, I. M. C., R. Surampalli and K. C. K. Lai (Editors) (2007) Zerovalent iron reactive materials for hazardous waste and inorganics removal. American Society of Civil Engineers, Reston, VA. Lokhande, C. D., V. V. Bhad and S. S. Dhumure (1992) Conversion of Tin Disulfide into Silver Sulfide by a Simple Chemical Method. Journal of Physics D-Applied Physics, 25 (2), 315–8. Lowry, G. V. and K. M. Johnson (2004) Congener-specific dechlorination of dissolved PCBs by microscale and nanoscale zerovalent iron in a water/methanol solution. Environmental Science & Technology, 38 (19), 5208–16. Madden, A. S. and M. F. Hochella (2005) A test of geochemical reactivity as a function of mineral size: Manganese oxidation promoted by hematite nanoparticles. Geochimica Et Cosmochimica Acta, 69 (2), 389–98. Madden, A. S., M. F. Hochella and T. P. Luxton (2006) Insights for size-dependent reactivity of hematite nanomineral surfaces through Cu2+ sorption. Geochimica Et Cosmochimica Acta, 70 (16), 4095–104. Marcus, M. A., W. Flood, M. Stiegerwald, L. Brus and M. Bawendi (1991) Structure of Capped Cdse Clusters by Exafs. Journal of Physical Chemistry, 95 (4), 1572–6. Martin, J. E., A. A. Herzing, W. L. Yan, et al. (2008) Determination of the oxide layer thickness in core-shell zerovalent iron nanoparticles. Langmuir, 24, 4329–34. McBride, M. B. (1994) Environmental Chemistry of Soils. Oxford University Press, New York. Meulenkamp, E. A. (1998) Size dependence of the dissolution of ZnO nanoparticles. Journal of Physical Chemistry B, 102 (40), 7764–9. Mews, A., A. Eychmuller, M. Giersig, D. Schooss and H. Weller (1994) Preparation, Characterization, and Photophysics of the Quantum-Dot Quantum-Well System Cds/Hgs/ Cds. Journal of Physical Chemistry, 98 (3), 934–41. Miller, D. S., A. J. Bard, G. McLendon and J. Ferguson (1981) Catalytic water reduction at colloidal metal “microelectrodes”. 2. Theory and experiment. Journal of the American Chemical Society, 103, 5336–41. Milliron, D. J., A. P. Alivisatos, C. Pitois, C. Edder and J. M. J. Frechet (2003) Electroactive surfactant designed to mediate electron transfer between CdSe nanocrystals and organic semiconductors. Advanced Materials, 15 (1), 58–61. Moreau, J. W., P. K. Weber, M. C. Martin, et al. (2007) Extracellular proteins limit the dispersal of biogenic nanoparticles. Science, 316 (5831), 1600–3. Mueller, N. C. and B. Nowack (2008) Exposure modeling of engineered nanoparticles in the environment. Environmental Science & Technology, 42, 4447–53. Narayanan, R. and M. A. El-Sayed (2005) Catalysis with transition metal nanoparticles in colloidal solution: Nanoparticle shape dependence and stability. Journal of Physical Chemistry B, 109 (26), 12663–76. Nie, S. M., Y. Xing, G. J. Kim and J. W. Simons (2007) Nanotechnology applications in cancer. Annual Review of Biomedical Engineering, 9, 257–88. Noguera, C., A. Pojani, P. Casek and F. Finocchi (2002) Electron redistribution in lowdimensional oxide structures. Surface Science, 507, 245–55.

106

Environmental and Human Health Impacts of Nanotechnology

Nurmi, J. T., P. G. Tratnyek, V. Sarathy, et al. (2005) Characterization and properties of metallic iron nanoparticles: Spectroscopy, electrochemistry, and kinetics. Environmental Science & Technology, 39 (5), 1221–30. Oyelere, A. K., P. C. Chen, X. H. Huang, I. H. El-Sayed and M. A. El-Sayed (2007) Peptideconjugated gold nanorods for nuclear targeting. Bioconjugate Chemistry, 18, 1490–7. Panigrahi, S., S. Basu, S. Praharaj, et al. (2007) Synthesis and size-selective catalysis by supported gold nanoparticles: Study on heterogeneous and homogeneous catalytic process. Journal of Physical Chemistry C, 111 (12), 4596–605. Pellegrino, T., L. Manna, S. Kudera, et al. (2004) Hydrophobic nanocrystals coated with an amphiphilic polymer shell: A general route to water soluble nanocrystals. Nano Letters, 4 (4), 703–7. Penn, R. L. and J. F. Banfield (1998) Oriented attachment and growth, twinning, polytypism, and formation of metastable phases: Insights from nanocrystalline TiO2. American Mineralogist, 83 (9–10), 1077–82. Penn, R. L. and J. F. Banfield (1999) Morphology development and crystal growth in nanocrystalline aggregates under hydrothermal conditions: Insights from titania. Geochimica Et Cosmochimica Acta, 63 (10), 1549–57. Quinn, J., C. Geiger, C. Clausen, et al. (2005) Field demonstration of DNAPL dehalogenation using emulsified zero-valent iron. Environmental Science & Technology, 39, 1309–18. Rhyner, M. N., A. M. Smith, X. H. Gao, et al. (2006) Quantum dots and multifunctional nanoparticles: new contrast agents for tumor imaging. Nanomedicine, 1 (2), 209–17. Robinson, R. D., B. Sadtler, D. O. Demchenko, et al. (2007) Spontaneous superlattice formation in nanorods through partial cation exchange. Science, 317 (5836), 355–8. Rockenberger, J., L. Troger, A. Kornowski, et al. (1997) EXAFS studies on the size dependence of structural and dynamic properties of CdS nanoparticles. Journal of Physical Chemistry B, 101 (14), 2691–701. Rockenberger, J., L. Troger, A. L. Rogach, et al. (1998) The contribution of particle core and surface to strain, disorder and vibrations in thiolcapped CdTe nanocrystals. Journal of Chemical Physics, 108 (18), 7807–15. Roden, E. E. (2003) Fe(III) oxide reactivity toward biological versus chemical reduction. Environmental Science & Technology, 37, 1319–24. Roden, E. E. and J. M. Zachara (1996) Microbial reduction of crystalline iron(III) oxides: Influence of oxide surface area and potential for cell growth. Environmental Science & Technology, 30 (5), 1618–28. Roelofs, F. and W. Vogelsberger (2004) Dissolution kinetics of synthetic amorphous silica in biological-like media and its theoretical description. Journal of Physical Chemistry B, 108 (31), 11308–16. Sau, T. K., A. Pal and T. Pal (2001) Size regime dependent catalysis by gold nanoparticles for the reduction of eosin. Journal of Physical Chemistry B, 105 (38), 9266–72. Sayre, L. M., G. Perry and M. A. Smith (2008) Oxidative stress and neurotoxicity. Chemical Research in Toxicology, 21, 172–88. Schmidt, J. and W. Vogelsberger (2006) Dissolution kinetics of titanium dioxide nanoparticles: The observation of an unusual kinetic size effect. Journal of Physical Chemistry B, 110 (9), 3955–63. Schrick, B., J. L. Blough, A. D. Jones and T. E. Mallouk (2002) Hydrodechlorination of trichloroethylene to hydrocarbons using bimetallic nickel-iron nanoparticles. Chemistry of Materials, 14 (12), 5140–7. Schwertmann, U., P. Cambier and E. Murad (1985) Properties of Goethites of Varying Crystallinity. Clays and Clay Minerals, 33 (5), 369–78. Sharma, R. K., P. Sharma and A. Maitra (2003) Size-dependent catalytic behavior of platinum nanoparticles on the hexacyanoferrate(III)/thiosulfate redox reaction. Journal of Colloid and Interface Science, 265 (1), 134–40. Shipway, A. N., M. Lahav, R. Gabai and I. Willner (2000) Investigations into the electrostatically induced aggregation of Au nanoparticles. Langmuir, 16 (23), 8789–95.

Size/Shape–Property Relationships

107

Somorjai, G. A. (1994) Introduction to surface chemistry and catalysis. Wiley, New York, 667 pp. Son, D. H., S. M. Hughes, Y. D. Yin and A. P. Alivisatos (2004) Cation exchange reactionsin ionic nanocrystals. Science, 306 (5698), 1009–12. Sonnichsen, C. and A. P. Alivisatos (2005) Gold nanorods as novel nonbleaching plasmonbased orientation sensors for polarized single-particle microscopy. Nano Letters, 5 (2), 301–4. Stouwdam, J. W., J. Shan, F. C. J. M. Van Veggel, et al. (2007) Photostability of colloidal PbSe and PbSe/PbS core/shell nanocrystals in solution and in the solid state. Journal of Physical Chemistry C, 111, 1086–92. Sund, C. J., S. McMasters, S. R. Crittenden, L. E. Harrell and J. J. Sumner (2007) Effect of electron mediators on current generation and fermentation in a microbial fuel cell. Applied Microbiology and Biotechnology, 76 (3), 561–8. Tang, R. K., M. Darragh, C. A. Orme, et al. (2005) Control of biomineralization dynamics by interfacial energies. Angewandte Chemie-International Edition, 44 (24), 3698–702. Tang, R. K. and G. H. Nancollas (2002) New mechanism for the dissolution of sparingly soluble minerals. Pure and Applied Chemistry, 74 (10), 1851–7. Tang, R. K., G. H. Nancollas and C. A. Orme (2001) Mechanism of dissolution of sparingly soluble electrolytes. Journal of the American Chemical Society, 123, 5437–43. Tang, R. K., C. A. Orme and G. H. Nancollas (2003) A new understanding of demineralization: The dynamics of brushite dissolution. Journal of Physical Chemistry B, 107 (38), 10653–57. Tang, R. K., C. A. Orme and G. H. Nancollas (2004a) Dissolution of crystallites: Surface energetic control and size effects. Chemphyschem, 5 (5), 688–96. Tang, R. K., L. J. Wang and G. H. Nancollas (2004b) Size-effects in the dissolution of hydroxyapatite: an understanding of biological demineralization. Journal of Materials Chemistry, 14 (14), 2341–6. Tang, R. K., L. J. Wang, C. A. Orme, et al. (2004c) Dissolution at the nanoscale: Selfpreservation of biominerals. Angewandte Chemie-International Edition, 43 (20), 2697–701. Thurston, T. R. and J. P. Wilcoxon (1999) Photooxidation of organic chemicals catalyzed by nanoscale MoS2. Journal of Physical Chemistry B, 103 (1), 11–7. Torrent, J., U. Schwertmann and V. Barron (1987) The Reductive Dissolution of Synthetic Goethite and Hematite in Dithionite. Clay Minerals, 22 (3), 329–37. Tratnyek, P. G. and R. L. Johnson (2006) Nanotechnologies for environmental cleanup. Nano Today, 1 (2), 44–8. Tsodikov, M. V., T. N. Rostovshchikova, V. V. Smirnov, et al. (2005) Structure and size effects in catalysis by immobilized nanoclusters of iron oxides. Catalysis Today, 105 (3–4), 634–40. Valden, M., X. Lai and D. W. Goodman (1998) Onset of catalytic activity of gold clusters on titania with the appearance of nonmetallic properties. Science, 281 (5383), 1647–50. Vayssieres, L., C. Sathe, S. M. Butorin, et al. (2005) One-dimensional quantum-confinement effect in alpha-Fe2O3 ultrafine nanorod arrays. Advanced Materials, 17 (19), 2320–3. Vikesland, P. J., A. M. Heathcock, R. L. Rebodos and K. E. Makus (2007) Particle size and aggregation effects on magnetite reactivity toward carbon tetrachloride. Environmental Science & Technology, 41 (15), 5277–83. Wang, L. S., X. Li and H. F. Zhang (2000) Probing the electronic structure of iron clusters using photoelectron spectroscopy. Chemical Physics, 262, 53–63. Waychunas, G. A., C. S. Kim and J. F. Banfield (2005) Nanoparticulate iron oxide minerals in soils and sediments: unique properties and contaminant scavenging mechanisms. Journal of Nanoparticle Research, 7 (4–5), 409–33. Wigginton, N. S., K. L. Haus and M. F. Hochella (2007) Aquatic environmental nanoparticles. Journal of Environmental Monitoring, 9, 1306–16.

108

Environmental and Human Health Impacts of Nanotechnology

Wilcoxon, J. P. (2000) Catalytic Photooxidation of Pentachlorophenol Using Semiconductor Nanoclusters. Journal of Physical Chemistry B, 104, 7334–43. Yang, Z. H. and C. S. Xie (2006) Zn2+ release from zinc and zinc oxide particles in simulated uterine solution. Colloids and Surfaces B-Biointerfaces, 47 (2), 140–5. Yavuz, C. T., J. T. Mayo, W. W. Yu, et al. (2006) Low-field magnetic separation of monodisperse Fe3O4 nanocrystals. Science, 314 (5801), 964–7. Yean, S., L. Cong, C. T. Yavuz, et al. (2005) Effect of magnetite particle size on adsorption and desorption of arsenite and arsenate. Journal of Materials Research, 20 (12), 3255–64. Yin, Y. and A. P. Alivisatos (2005) Colloidal nanocrystal synthesis and the organic-inorganic interface. Nature, 437 (7059), 664–70. Yin, Y. D., C. K. Erdonmez, A. Cabot, S. Hughes and A. P. Alivisatos (2006) Colloidal synthesis of hollow cobalt sulfide nanocrystals. Advanced Functional Materials, 16 (11), 1389–99. Yin, Y. D., R. M. Rioux, C. K. Erdonmez, et al. (2004) Formation of hollow nanocrystals through the nanoscale Kirkendall Effect. Science, 304 (5671), 711–4. Yu, W. W., E. Chang, J. C. Falkner, et al. (2007) Forming biocompatible and nonaggregated nanocrystals in water using amphiphilic polymers. Journal of the American Chemical Society, 129 (10), 2871–9. Yuan, G. D. (2004) Natural and modified nanomaterials as sorbents of environmental contaminants. Journal of Environmental Science and Health Part A–Toxic/Hazardous Substances and Environmental Engineering, A39 (10), 2661–70. Zelmanov, G. and R. Semiat (2008) Iron(3) oxide-based nanoparticles as catalysts in advanced organic aqueous oxidation. Water Research, 42, 492–8. Zhang, H. Z. and J. F. Banfield (2007) Polymorphic transformations and particle coarsening in nanocrystalline titania ceramic powders and membranes. Journal of Physical Chemistry C, 111 (18), 6621–9. Zhang, H. Z., R. L. Penn, R. J. Hamers and J. F. Banfield (1999) Enhanced adsorption of molecules on surfaces of nanocrystalline particles. Journal of Physical Chemistry B, 103 (22), 4656–62. Zhang, J. W. and G. H. Nancollas (1990) Mechanisms of growth and dissolution of sparingly soluble salts. In: M. F. Hochella and A. F. White (Editors), Mineral-water Interface Geochemistry. Reviews in Mineralogy. Mineralogical Society of America, Washington, D.C., pp. 603. Zhang, W. X. (2003) Nanoscale iron particles for environmental remediation: An overview. Journal of Nanoparticle Research, 5 (3–4), 323–32. Zhang, W. X., C. B. Wang and H. L. Lien (1998) Treatment of chlorinated organic contaminants with nanoscale bimetallic particles. Catalysis Today, 40 (4), 387–95.

4 Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments Mohamed Baalousha1, Jamie R. Lead1, Frank von der Kammer2 and Thilo Hofmann2 1

School of Geography, Earth and Environmental Sciences, University of Birmingham, United Kingdom 2 Department of Environmental Geosciences, Vienna University, Austria

4.1

Introduction

Environmental colloids have important environmental functions in aquatic and terrestrial systems. For instance, they dominate the physico-chemical speciation of trace elements and organic pollutants. A large proportion of these trace compounds (typically 40–90% or more) are adsorbed to colloids (Stumm, 1992; Stumm and Morgan, 1996). The binding of trace pollutants by colloids can be interpreted as a function of colloidal size (Lead et al., 1999), chemistry of the colloidal phases (Lienemann et al., 1997) or both (Baalousha et al., 2006a; Lyven et al., 2003). The importance of colloids in metal binding stems from the inverse relationship between size and specific surface area, although other phenomena such as quantum related effects may be important in the smallest fraction (<100 nm) of colloids (Wigginton et al., 2007). It has been found that small colloids of about 50 nm (Lead et al., 1999) or <25 nm (Lyven et al., 2003) are capable of binding the largest fraction of total trace metals. Further, Lyven et al. (2003), assumed that iron oxides and organic carbon were the main binding phases and that elements were distributed between dissolved (e.g. molybdenum), organic (e.g. copper and zinc) and iron oxides (e.g. Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

110

Environmental and Human Health Impacts of Nanotechnology

lead). However, in many cases metals were distributed to some extent among all phases due to iron–carbon interactions. Other phases such as manganese oxides and sulfides may also be important in trace element binding (Baalousha et al., 2006a). As with inorganic contaminants, colloids may significantly influence the distribution and fate and behaviour of organic contaminants. For instance, the majority of polycyclic hydrocarbons (PAHs) were found to be present in large (>20 µm) flocs (Leppard et al., 1998), which were essentially aggregates of small colloids. Marvin et al., (2004) showed that PAHs were primarily associated with particles less than 2 µm in diameter. The majority of these particles were found to be fractal aggregates of humic substance. In marine systems, the majority of polychlorinated biphenyls (PCBs) were found to be associated with particulate matter (>1.2 µm), although in the fraction <1.2 µm, colloidal binding (40–80%) was dominant (Burgess et al., 1996). Up to 93% of polychlorinated biphenyls were found to be associated with colloids in a coastal sea area (Totten et al., 2001). The interaction of selected pharmaceuticals (Maskaoui et al., 2007) and endocrine disrupting chemicals (EDCs) (Liu et al., 2005; Zhou et al., 2007) with natural colloids has also been more recently investigated. While the more hydrophobic pharmaceuticals showed a linear dependency of the Kcoc (colloidal organic carbon sorption coefficient) and the Kow (octanol–water partition coefficient), the Kcoc of the more hydrophilic EDCs were independent of the Kow, highlighting the importance of different binding mechanisms. Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) were found to be relocated from soil to groundwater associated with colloids (Hofmann and Wendelborn, 2007). The behaviour of colloidal particles is dominated by aggregation/disaggregation and sedimentation in aquatic systems (Buffle et al., 1998) and attachment to surfaces in porous media (McDowell-Boyer et al., 1986). These processes are highly influenced by solution physico-chemistry (e.g. pH and cation types and concentrations) and behaviour of natural organic molecules (Wilkinson et al., 1997a, 1997b). Aggregation of colloids results in the formation of large structures, which sediment in the water body or attach to surfaces in soils which results in their loss, thus eliminating the chemicals from a water body in a processes known as colloidal pumping (Honeyman and Santschi, 1992). Colloids are often porous and form fractal-like aggregate structures; this depends greatly on solution conditions such as pH and ionic strength (Baalousha et al., 2006b; Chen and Eisma, 1995; Senesi et al., 1996). Such porosity and conformation of colloids and their aggregates may result in sorption–desorption of chemical compounds (e.g. pollutants and nutrients) and possibly a permanent retention within the structure of colloids/aggregates (Kan et al., 1994). Further, the fractal nature of colloidal aggregates influences their sedimentation behaviour. It is now well known that contaminant and nutrient transport processes in marine and freshwater systems are dominated by the transport of particles and substances associated with them (Benedetti et al., 2002; Dai et al., 1995; Santschi et al., 1997). For decades, processes of contaminant relocation in soil and groundwater were believed to occur predominantly in a two-phase system: the mobile liquid phase and the immobile solid phase, a potentially mobile solid phase was neglected

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

111

(McCarthy and Zachara, 1989). Colloid-facilitated transport is now a well recognized process in porous media such as soils and aquifers. Small colloids compete with the solid, immobile phase for trace contaminant sorption (e.g. metals (Chen et al., 2005), organic pollutants (White et al., 2005) and nutrients (Heathwaite et al., 2005)) and increase the distances travelled by pollutants with respect to those predicted from non-colloidally bound components (Kaplan et al., 1995; Laegdsmand et al., 1999; McCarthy, 1998). The bioavailability of both metals and hydrophobic organic contaminants (HOCs) is affected by colloids. In the metals area, there are well defined equilibrium chemistry based models (biotic ligand model, BLM, and free ion activity model, FIAM), which are standard in the literature, implicating colloids in (usually) a reduction of toxicity (Campbell, 1995; Paquin et al., 2002). Dynamic models accounting for solution chemical kinetics and mass transport phenomena are being used more extensively (van Leeuwen et al., 2005). The BLM is now a regulatory tool in the Unites States and in parts of the EuropeanUnion. Although less well developed theoretically, HOCs toxicity is often related to octanol–water partitioning coefficient. Toxicity also depends on the presence of humic acids (Galle et al., 2005) and particulate or dissolved organic matter (Hodge et al., 1993). Humic substances have been shown to affect cell permeability, cell charge and nutrient availability (Kola and Wilkinson, 2005) and this may be a mechanism for their impacts on toxicity, in addition to their effect on solution speciation. From the above discussion, it is clear that colloids are important components in the environment. They can control pollutant chemistry, that is pollutant speciation, and, consequently, influence their transport and bioavailability (Doucet et al., 2006; Lead et al., 1999). Although advances have been made in understanding the behaviour and role of colloids in environmental systems in the last few decades, much is still unknown due to the intrinsic complexity of natural colloids, the lack of appropriate experimental techniques and the significant gap between coagulation theory, which was developed to describe simple systems of identical, spherical, non-living particles and the reality of natural systems that contain heterogeneous mixture of particles (Buffle, 1988). The ongoing introduction (accidentally or deliberately) of manufactured nanoparticles (NPs) will likely be controlled to a large extent by these natural colloids, especially given the likely difference in concentrations; natural colloids present at mg l−1 and NPs at µg l−1, typically in freshwaters. Thus, natural colloids are important, intrinsically, in considering the literature to better understand the likely fate and behaviour of NPs, and because they will interact directly with NPs, altering their fate and behaviour. Recent studies suggest that the introduction of manufactured NPs may have a potential harmful effect on the environment due to their potential toxicity (Lovern and Klaper, 2006; Oberdorster et al., 2006) and other indirect environmental effects (Zhang, 2003). The specific fate and behaviour of engineered nanoparticles in environmental systems, as far as it is known, was discussed in Chapter 1. In this chapter colloidal behaviour will be reviewed, as well as the parameters which determine their fate and behaviour in the aquatic and terrestrial environments. This knowledge will help to improve our somewhat sketchy understanding of the likely fate and behaviour of manufactured nanoparticles in these environments.

112

Environmental and Human Health Impacts of Nanotechnology

This chapter begins by looking at the major types of colloidal particles and their properties, which are related to environmental processes. This is followed by a discussion of interaction forces (both DLVO and non-DLVO) between colloidal particles, and the fate and behaviour of colloidal particles in aquatic and terrestrial systems. The discussion on natural aquatic and terrestrial colloids is limited to the extent to which it will aid in appreciation of the likely behaviour of manufactured nanoparticles. For a fuller discussion of natural colloids, the reader is directed elsewhere (Lead and Wilkinson, 2006a).

4.2

Definition

Traditionally, a colloidal system has been defined as a dispersion of one phase in another, where the dispersed phase is between 1 nm and 1 µm in one dimension (IUPAC, 2002). Accordingly, natural aquatic colloids are formally defined as material with one dimension between 1 nm and 1 µm (Figure 4.1), while particles are larger than 1 µm (Hofmann et al., 2003; Lead and Wilkinson, 2006b). Alternatively, colloids can be defined as organic or inorganic entities small enough to be dominated by aggregation and to remain in the water column over reasonable timescales, but large enough to have supramolecular structure and properties, for example electrical surface charge (Lead and Wilkinson, 2006a, 2006c). Particles are large enough (>1 µm) to be dominated by sedimentation, rather than aggregation (Buffle and Leppard, 1995). This definition was developed and somewhat extended by Gustafsson and Gschwend (1997), where an aquatic colloid can be defined as any constituent that provides a molecular milieu into and onto which chemicals can escape from the bulk aqueous solution, while its vertical movement (in water) is not significantly affected by gravitational settling over reasonable timescales. In practice, in much of the literature, colloids are defined as materials which permeate a filter (pore size between 0.1 and 1.0 µm, often with little standardization in the literature data) while being retained by an ultrafilter (1–100 kDa, nominal pore size). It is clear that the formal, mechanistic and practical definitions do not entirely mesh and some work on standardization is required. Within this colloidal fraction it is useful to define a nanoscale fraction (Lead and Wilkinson, 2006c; Wigginton et al., 2007), which may be thought of as between 1 and 100 nm, as with manufactured nanoparticles (Chapter 2). However, the size range below about 10–25 nm may be the range in which environmental properties such as metal binding, zeta potential and redox properties change radically compared with the bulk or larger-sized phases (Lyven et al., 2003; Madden et al., 2006; Madden and Hochella, 2005).

4.3

Major Types of Environmental Colloids

Aquatic and terrestrial colloids are highly heterogeneous in size, shape, structure, chemical composition and other properties. These colloids include organic and inorganic compounds and even biota with proportions dependent on the nature of the inputs, within media processes and outputs (Bertsch and Seaman, 1999;

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

COLLOIDS OR MACROMOLECULES

SOLUTES –10

ANALYTICAL TECHNIOUES

INORGANIC COMPOUNDS ORGANIC COMPOUNDS

1A

–9

–8

–7

1 nm

–6

113

PARTICLES log(size/m) –5

1 µm

0.45 µm

Cellular debris

Amino acids

Polysaccharides Peptides

Peptidoglycans Proteins

Fulvics

Humic aggregates

Humics

Viruses

Bacteria

Algae

Organic compounds adsorbed ‘Simple’ hydrated ions (e.g. OH–, Cl–, SO–4, HS–, Na+, Ca2+, Mg2+, Cu2+, etc.)

(b)

on inorganic particles Clays (aluminosilicates) Fe oxyhydroxides Mn oxides Metal sultides Carbonates, phosphates Amorphous SiO2 Ultrafiltration

Filtration

Electron microscopy Atomic force microscopy Flow-FFF

Confocal microscopy Optical microscopy

Sedimentation-FFF

FCS Light scattering X-Ray, neutron scattering

X-Ray absorption

LIBD

(a)

(c)

Figure 4.1 (a) Schematic representation, by size distributions, of the major environmental colloidal and particulate components; (b) typical example of natural colloids and aggregates (Rhine River), scale bar corresponds to 1 µm and (c) natural heteroaggregate of colloids and particles from Lake Bret, Switzerland, as shown by transmission electron microscopy, scale bar corresponds to 250 nm. ((a) Lead, J. R. and K. J. Wilkinson (2006) Aquatic colloids and nanoparticles: current knowledge and future trends, Environmental Chemistry, 3, 159–71. Reproduced with permission from CSIRO publishing, http://www.publish.csiro.au/journals/ ec. (c) Reprinted with permission from J. Buffle, K.J. Wilkinson, S. Stoll, M. Filella, J. Zhang, A generalized description of aquatic colloidal interactions: the three-colloidal component approach, Environmental Science & Technology, 32, 2887–99. Copyright 1998, American Chemical Society.)

Zimmermann-Timm, 2002). Hence, one of the ways to understand colloids is to study them by classes of compounds of similar composition and properties. Figure 4.1a summarizes the different types of environmental colloids together with the size range they cover and the analytical tools (see Chapter 6 for detailed discussion) that can be used to characterize them. Figures 4.1b and 4.1c show typical examples

114

Environmental and Human Health Impacts of Nanotechnology

of surface freshwater colloidal material of various sizes, as observed by transmission electron microscopy. Here, three types of colloids that are recognized as the major colloidal components in environmental systems are presented, namely inorganic colloids, humic substances and biopolymers (Buffle et al., 1998). The later two types will be presented under the same category of organic macromolecules. For more details, the reader is referred to a recent review of this area (Filella, 2006). 4.3.1

Inorganic Colloids

There are two main types of inorganic colloidal particles in oxygenated terrestrial and aquatic environments, which are aluminium phyllosilicates (e.g. clay, mica, chlorite) and oxides and hydrous oxides of iron (e.g. haematite and magnetite), manganese (e.g. pyrolusite) and silicon (e.g. SiO2). Other inorganic colloids can also be found, but they are usually minor components (e.g. other groups of silicates) or are primarily present in anoxic waters (e.g. FeS, FeS2, MnS). Sulfides have also been found to be a potentially important minor species in oxic waters. Calcium carbonate can be found in significant amount in freshwaters but is mostly in particulate form (Sigg, 1994) with weak metal binding. 4.3.1.1 Aluminium Phyllosilicates Phyllosilicates are a subgroup of silicates, an extensive group of minerals which are derived from silica (SiO2). All clay minerals, that is aluminium phyllosilicates, belong to this group. They are phyllosilicates which form parallel tetrahedral silicon sheets and octahedral aluminium sheets. The most common clay minerals within the phyllosilicates are kaolinite, illite, vermiculite and the smectite/montmorillonite group. Clay minerals from these four groups are the most abundant inorganic colloids in aquatic and terrestrial systems and are generally weathering products from soils and rocks (Berner and Holdren, 1977; Helgeson et al., 1984; Murphy and Helgeson, 1987). They usually have an irregular shape and a range of crystalline structures and sizes, covering both the colloidal and particulate range. Clay minerals control the composition of natural waters and contribute to the formation of secondary solids as many clay minerals. They have a high cation exchange capacity and different types of surface charge (permanent and variable, i.e. pH dependent). Clay particle aggregation is complicated as they have different modes of aggregation (plane-to-plane, plane-to-edge and edge-toedge) (Lagaly and Ziesmer, 2003). The stability of clays in natural water is mainly determined by charge and charge heterogeneity (Aurell and Wistrom, 2000; Chang and Sposito, 1996). As with all standard or reference materials, the application of results obtained on isolated clays in laboratory studies to their behaviour in natural environments should be performed cautiously for several reasons, given the heterogeneity, complexity and spatial and temporal variability of natural samples. Isolated and pure clays may have different surface properties compared to naturally occurring clays (Schulthess and Huang, 1990; Schulthess and Sparks, 1989), larger particle sizes are often used compared to natural colloids (Aldahan et al., 1999; Arnold et al., 2001) and given the extensive use of particle pretreatment such as drying, grinding and saturation with sodium ions, leading to the removal of

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

115

carbonates, natural organic matter (NOM) and iron coatings (Mukhopadhyay and Walther, 2001). 4.3.1.2

Oxides and Hydrous Oxides

Iron exists in the environment in two redox states, Fe(II) and Fe(III). Fe(II) is stable at low pH or in the absence of oxygen or other oxidants, is soluble and relatively free from complexation, whereas Fe(III) is stable in the presence of oxygen and insoluble at neutral pH. It is well documented that solid phase Fe(III) oxides and oxyhydroxides are formed by oxidation and hydrolysis of Fe(II) by subsurface aeration (Wolthoorn et al., 2004) at oxic/anoxic boundaries in groundwater (Christensen et al., 2001), freshwater lakes (Balistrieri et al., 1992) and coastal marine water (Gunnars et al., 2002). In freshwaters, the oxidation process typically results in colloidal particles having a mean diameter in the range of 0.05–0.5 µm (Lienemann et al., 1999; Tipping et al., 1981) and often lower in the nanoscale range. The presence of dissolved species, such as silicate, phosphate and organic matter, can affect the composition, structure, morphology and reactivity of these hydrolysis products (He et al., 1996; Kandori et al., 1992; Mayer and Jarrell, 2000). The transformation between dissolved Fe(II) species and solid Fe(III) oxyhydroxide phases at oxic–anoxic interfaces is central in the cycling of iron in aquatic environments (Davison, 1993), that is in the production and removal of particles in the different environmental compartments. These processes are highly dependent on pH and can be controlled by microorganisms (Fredrickson et al., 1998; Lovley, 1997; Pronk and Johnson, 1992). For more information, the reader is referred to the literature on iron cycling and particles in freshwater (Davison, 1993; Davison and De Vitre, 1992; Stumm and Sulzberger, 1992), the biogeochemistry of iron in seawater (Turner and Hunter, 2001) and the formation and occurrence of biogenic iron-rich minerals (Fortin and Langley, 2005). Naturally occurring iron oxide particles are very complex in structure. They exist under different crystalline and/or amorphous forms, such as haematite (α-Fe2O3), goethite (α-FeOOH), lepidocrocite (γ-FeOOH), maghaemite (γ-Fe2O3), magnetite (Fe3O4) and ferrihydrite (amorphous Fe(III) phase) (Davison and De Vitre, 1992). Iron oxides formed in natural aquatic systems are not pure oxides but contain a significant amount of other elements (Mavrocordatos et al., 2000; Mavrocordatos and Fortin, 2002; Perret et al., 2000) and are usually complexed or coated by NOM (Allard et al., 2004). In aerated sediment porewaters, freshly formed iron oxides co-precipitate heavy metals as lead, copper and zinc (v.d. Kammer et al., 2003). Iron oxides control different environmentally relevant processes in soil, freshwater, groundwater and oceans such as the attenuation of bacteria and viruses (Scholl et al., 1990), the fate and transport of trace contaminants (Tessier et al., 1996) and limited ocean primary productivity (Behrenfeld et al., 1996; Martin and Fitzwater, 1988). Iron particles are one of the main vectors of trace metals transport in aquatic systems (Lyven et al., 2003). In aquatic and terrestrial environments, mangenese (hydr)oxides are strongly affected by redox reactions. The oxidation of Mn(II) to Mn (III, IV) is thermodynamically possible at neutral pH and atmospheric oxygen concentrations, though

116

Environmental and Human Health Impacts of Nanotechnology

the process is very slow as the activation energy is high. However, oxidation rates much faster than can be accounted for by pure abiotic mechanisms have been measured and were explained by microbial catalysis of the oxidation process (Chapnick et al., 1982; Moffett and Ho, 1996; Sunda and Huntsman, 1987). It has been stated that the only process of manganese oxidation in freshwater, marine and terrestrial environments is bacteria-mediated oxidation (Filella, 2007). The microbial-mediated formation of manganese oxide particles was reported in various freshwater systems (Lienemann et al., 1997; Tani et al., 2003). Manganese (hydr)oxides have proven difficult to identify in aquatic and terrestrial environments due to their low concentrations. Manganese (hydr)oxides often occur as small crystals and are often intermixed with (hydr)oxides of iron or with organic matter (Figure 4.2), or as coatings on mineral surfaces and biofilms (Chen et al., 2000; Dong et al., 2000, 2001). Consequently, they are difficult to separate from the colloidal matrix and even more difficult to concentrate or purify. Silica (SiO2) can be encountered in the environment as the mineral quartz or its polymorphs. Silica colloids can be released during the diagenesis of amorphous silica. In addition, some plankton (diatoms, estimated to accounts for 40% of primary activity in the ocean (Nelson et al., 1995)) construct their exoskeletons from silica, which become part of colloidal or particulate pool (biogenic silica) in the aquatic environment after diatom death. Diatoms in fresh and salt water extract silica from the water to use it as a component of cell wall. The global production of biogenic silica is dominated by diatoms. The silicon cycle, the formation of biogenic silica and the factors determining the rate of silica production and removal in surface waters have been reviewed in detail elsewhere (Nelson et al., 1995; Ragueneau et al., 2000).

UV/VIS & metals (scaled to fit) (a.u.)

1.2 1.0

UV/VIS Al Fe Mn Pb

0.8 0.6 0.4 0.2 0.0 0

50

100

150

200

0 250

hydrodynamic radius (nm)

Figure 4.2 FFF–ICPMS relative particle and element size distribution of aquifer colloids. The grey area represents the UV/VIS signal at 260 nm (as turbidity) and is a measure for the total colloid concentration. The coloured traces show the distribution of the major elements iron, aluminium and manganese and of the trace element lead. The signals are scaled to fit the graph. (v.d. Kammer, Doubascoux, Lespes, unpublished.) (See colour plate section for a colour representation)

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

117

Silica colloids are present in all natural waters, and thus are expected to play an important role in environmental systems. They can play an important role in metals speciation (Osthols, 1995; Schindler et al., 1976), limit the primary production in estuaries and oceans (Conley and Malone, 1992; Dugdale et al., 1995), alter the species composition of phytoplankton in oceans (Officer and Ryther, 1980; Turner et al., 1998) and, ultimately, may have important consequences on the level of carbon dioxide in the atmosphere and, therefore, the global climate (Harrison, 2000; Treguer and Pondaven, 2000). 4.3.1.3

Sources of Inorganic Colloids

Inorganic colloidal particles can be produced by a range of different processes. In situ mobilization of particles in soil and subsurface water can be a significant source, as can chemical precipitation from supersaturated solutions, biotic processes and sources such as glacial erosion in glacial lakes and rivers, or waste disposal. The main sources of mineral colloids in surface waters are detachment, from soil surfaces (pedogenic) or by sediment resuspension, formation of solid phases by chemical precipitation, or biogenic, from primarily microbial origin. The formation of solid phases by chemical precipitation is a common process. Calcite precipitates in lakes and oceans (Stabel, 1986), metal sulfides precipitate in anoxic waters (BenYaakov, 1999; Morse and Luther, 1999), iron and manganese oxyhydroxides form in redox gradients in sediment porewaters (v.d.Kammer et al., 2003) and in transition layers of eutrophic lakes and sediments (Davison, 1993). In soils and groundwater, the main source of colloidal material is their detachment from the soil surface due to changes in solution chemistry (Kaplan et al., 1996), rainfall or soil irrigation (Bertrand and Sor, 1962). The main processes of colloid translocation, mobilization and generation in a groundwater environment are shown schematically in Figure 4.3. Colloid release is more important at high pH and low ionic strength and high flow velocity (Kaplan et al., 1993; Laegdsmand et al., 1999; Roy and Dzombak, 1996). Human activities such as waste disposal, groundwater pumping and artificial recharge can enhance the formation and mobilization of colloids by disturbing groundwater and soil solution chemistry (Gschwend and Reynolds, 1987; Hofmann and Schöttler, 1998; Liang et al., 1993). Redox gradients in groundwaters may produce iron and manganese oxide colloids due to reoxidation of anoxic groundwaters (Christensen et al., 2001; Hofmann, 2002; Wolthoorn et al., 2004). The hydrolysis of radionuclides, such as actinides may also result in the formation of colloidal particles (Bates et al., 1992). 4.3.2

Organic Macromolecules

Most dissolved, colloidal and particulate organic matter (DOM, COM and POM) in soil, sediments and natural water are still poorly characterized at the molecular level, due to their inherent chemical and structural complexity (Leenheer and Croué, 2003), although their origins, reactions and fates have been extensively studied (Dignac et al., 2000; Hedges et al., 2000; Wakeham et al., 1997; Williams and Druffel, 1988). Natural organic matter (NOM) in the aquatic or terrestrial environment can be divided into two classes of compounds: non-humic material

118

Environmental and Human Health Impacts of Nanotechnology

dissolution

translocation leachates from vadoze zone seapage

deposition

mobilisation ionic strength pH ionic composition hydraulic effects

Is↓ pH ? Ca++ → Na+

stabilisation transport

generation secondary mineral formation precipitation from over-saturation dissolution of cements

filtration

bacterial growth

Eh↓

pH ?

pCO2↑

Figure 4.3 Schematic plot of important processes influencing colloid behaviour in the subsurface environment. Mobilization usually takes place when double layers expand or by changes of the surface charge (polarity +/− to +/+ or more often −/−); hydrodynamic forces usually play a less important role for colloids. Generation occurs when new colloids are produced by precipitation from supersaturation or by dissolution of cements which contain colloidal particles (as carbonates or oxides) through changes in surrounding conditions as decrease in pH or redox potential or increase in the dissolved carbon dioxide. Removal of colloids is associated with dissolution of particles, their deposition onto the immobile matrix straining filtration in the pores (from V. D. Kammer 2005).

(e.g. protein, polysaccharides, nucleic acids and small molecules such as sugars and amino acids) and humic substances. In this section only two categories of NOM are reviewed, namely humic substances and extracellular polymeric substances (e.g. polysaccharides and proteins, usually present as fibrillar material), as they represent the major constituents of NOM and play a significant role in determining the fate and behaviour of colloids and particles (Buffle et al., 1998). Other compounds are limited by their rapid turnover or low production rates and hence are present at low concentrations in different environmental systems (Fabiano and Pusceddu, 1998; Mannino and Harvey, 2000).

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

4.3.2.1

119

Humic Substances

Humic substances are a heterogeneous mixture of low to high molar mass organic acids that are defined operationally by the standard technique used for their extraction and isolation. Humic substances are generally divided operationally into three classes of materials: fulvic acids (FA) are materials soluble in water under all pH conditions, while humic acids (HA) are soluble at pH > 2 and humins are insoluble at all pHs (Thurman and Malcolm, 1981). They cannot be fully defined as yet by structure or function, which are complex, temporally and spatially variable and highly influenced by solution conditions (Hayes et al., 1989). New technical developments have given more insight to the structures of humic substances, which have been explained in different ways including macromolecules (Swift, 1989), supramolecular association of small molecules (Piccolo, 2001), micelles (Thurman et al., 1982; Wershaw, 1999) and soft, semi-permeable spheres (Duval et al., 2005), although these structures are usually technique dependent (Lead and Wilkinson, 2006c). However, it is generally agreed that they are structurally complex macromolecules, strong adsorbers in the UV–Visible range (rich in chromophores) and are weak polyfunctional acids. Humic acids are generally terrestrial, while aquatic humic substances are dominated by FA, showing that the procedural definition has some use in interpreting geochemical behaviour. As with inorganic colloids, humic substances have different sources in different environments. The principle source of humic substances in terrestrial and freshwater environments is the degradation of higher plants, whereas the main source in marine systems is the degradation of plankton (Aiken et al., 1996; Kristensen, 1990). Terrestrial, freshwater and marine humic substances have significantly different chemical characteristics, with a significant contribution from autochthonous sources in freshwater. Freshwater humic substances, have a high C/N ratio (40 to 50), are rich in aliphatic carbon, have strong absorption in the near UV and are often depicted with a highly condensed, cyclic molecular structure. In contrast, marine humic substances have a low C/N ratio (15 to 20) (C/N ratio of fresh microbial material is between 5 and 10 to 1, for comparison), are rich in aromatic carbon, have weak absorption in the near UV and are often depicted with a more open, linear molecular structure. These differences arise from differences in the organic matter sources and formation processes of these two environments (Hedges et al., 1997). Humic substances have important environmental functions. They have an important role in regulating the chemical reactivity, speciation (Tipping, 2002), bioavailability and toxicity of metal ions in the natural environment (Koukal et al., 2003; McGeer et al., 2002; Van Ginneken et al., 2001). They also regulate the speciation (De Paolis and Kukkonen, 1997; Khan and Schnitzer, 1972), bioavailability and toxicity of organic pollutants (Haitzer et al., 1998). They play an important role in stabilizing inorganic and other colloids (Tipping and Higgins, 1982; Tipping and Ohnstad, 1984) through surface sorption and charge and steric stabilization, and have also been shown to be toxic themselves (Bernacchi et al., 1996; Qi et al., 2008).

120

Environmental and Human Health Impacts of Nanotechnology

4.3.2.2

Fibrillar Polysaccharides

In freshwaters, a number of organic compounds, in most cases polysaccharides and proteins, but also nucleic acids, peptidoglycan, lipids, lignins, and so on are produced in the water column by exudation (extracellular polymeric substances, EPS) or degradation of phytoplankton, aquatic bacteria and macrophytes. Multiple environmental factors, operating on a large scale, can impose stress (e.g. nutrient deprivation, high toxicant levels) upon some algal and bacterial species in the environment. The stress can lead to biological response such as secretion of organic macromolecules (biopolymers) to alleviate the stress (Leppard, 1995). EPS represent the most abundant organic compounds in the biosphere and constitute the largest fraction of cells. They are important in processes such as mineral dissolution (Welch et al., 1999), biomineralization (Chan et al., 2004), sediment stabilization (Dade et al., 1990), bacterial adhesion (Marshall et al., 1989), biofilm formation (Vandevivere and Kirchman, 1993) and pollutant distribution (Wolfaardt et al., 1994). This section considers mainly fibrillar polysaccharides. Other biopolymers (e.g. protein and peptidoglycan) are not considered as they represent a minor fraction of EPS or have a short turnover time (protein degrades within hours to days and peptidoglycan degrades with days to weeks of their release into the water column) (Nagatal et al., 2003; Smith et al., 1992). Fibrillar polysaccharides can be released from phytoplankton cells during all stages of growth (Strycek et al., 1992). Large amounts of polysaccharides are released during phytoplankton blooms and may comprise 80–90% of the total extracellular release (Myklestad, 1995). They may represent a significant proportion of NOM in freshwater, varying seasonally from about 5 to 30% in surface waters of lakes (Wilkinson et al., 1997a) and likely account for higher proportions (up to 80%) of NOM in marine systems (Aluwihare et al., 1997; McCarthy et al., 1998; Santschi et al., 1998; Verdugo et al., 2004). Polysaccharides are refractory enough to be found in the deep ocean and have a turnover time of hundreds of years (Guo and Santschi, 1997). Polysaccharides are generally rigid due to the large quantity of strongly bound hydration water (up to 80%), their association into double or triple helices that may be stabilized by hydrogen or calcium bridges or helices aggregation (Morris et al., 1980; Norton et al., 1984; Rees, 1981). Transmission electron microscopy (TEM) and atomic force microscopy (AFM) analysis of freshwater and marine polysaccharides suggest that they are a few nanometres in thickness with a length greater than 1 µm and variable conformation as a function of pH and ionic strength (Leppard et al., 1990; Perret et al., 1991; Santschi et al., 1998). Fibrils are important in a variety of environmental functions, such as floc formation via bridging mechanisms which enhance particle sedimentation (Buffle et al., 1998), formation of the matrix component of biofilms and facilitation of microbial adhesion to surfaces (Leppard, 1997), and binding of metal contaminants (Lamelas et al., 2005, 2006; Plette et al., 1996). Due to the complexity, which is both speciesspecific and a function of environmental conditions (Myklestad, 1995), and difficulties in isolating environmental polysaccharides, the study of their colloidal properties is usually performed on model polysaccharides purchased from chemical companies, or ideally produced in bacterial and algal culture within research laboratories

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

121

(Alasonati et al., 2007; Wilkinson et al., 1999). Far less work on these polysaccharides has been performed when compared to HS, primarily for this reason.

4.4

Intrinsic Properties of Environmental Colloidal Particles

This section focuses on the colloidal properties that are relevant to their environmental behaviour and characterization principles, especially as they might enable us to understand the fate and behaviour of manufactured nanoparticles. 4.4.1

Size

Size is the primary means of defining colloids in natural systems (Section 4.2) and is a useful parameter, as other physical and chemical parameters relevant to colloidal behaviour, for example diffusion coefficient, are influenced by, or correlate with, size. Therefore, potentially their behaviour and role in the biogeochemical cycling of trace pollutants and other processes may be understood in terms of size. Colloids, that is particles smaller than 1 µm, tend to stay in suspension, while larger particles tend to sediment (Buffle and Leppard, 1995) and so can be thought of as fundamentally different from particles, even if they are composed identically in chemical terms. As noted, at sizes <10–25 nm, environmental properties such as metal binding, zeta potential and redox properties may change radically in comparison to the bulk or larger phases of the same composition (Madden et al., 2006; Madden and Hochella, 2005). The size distribution of natural particles (Figure 4.1a) depends on the source and nature of particles, physical, chemical and biological processes, such as erosion, degradation, aggregation, disaggregation and ageing, and the physicochemical parameters of the system, such as pH, ionic strength and redox potential. Typical TEM images of natural colloidal particles from a River and a lake (Buffle et al., 1998) are shown in Figures 4.1b and 4.1c, and it is clear that it is hard to define an exact particle size due to the inherent differences in colloidal shape and to (possibly reversible) aggregation. 4.4.2

Surface Charge

There are two types of surface charge in colloids (Figure 4.4). The first is a permanent charge which arises from the isomorphous substitution of cations within the colloid, for example substitution of silicon(IV) by aluminium(III) in kaolinite. The second is a variable charge originates from chemical reactions at the colloidal surface: (i) ionization or dissociation of the surface functional groups (e.g. the dissociation of protons from carboxylic groups); (ii) dissolution of ionic solids (e.g. AgI); or (iii) specific sorption of charged species, for example simple ions such as Ca2+, surfactant ions and polyelectrolyte chains such as humic substances or synthetic surfactants. The total charge is the sum of permanent and variable charges. At pH values of natural conditions, most colloidal particles are negatively charged. The exception to this is iron oxides, which have a point of zero charge (pzc, the pH at which the overall charge equal zero) of about 7–9 depending on their

122

Environmental and Human Health Impacts of Nanotechnology (a) Ionisation of surface groups pH < 7

Al—OH+2

pH > 7

Al—O–

(c) Dissolution of ionic solids Agl I– Ag+ Ag+ + – I– I + Ag I– (d) Isomorphous substitution

(b) Ion adsorption

Clay

– e.g. SDS, – CH3(CH2)10CH2OSO3 Na+

Al3+

Si4+

Figure 4.4 The methods of charging a solid surface immersed in electrolyte. (T. Cosgrove, Charge in colloidal systems, in Colloid science: principles, methods and applications, 2005. Reproduced with permission from Blackwell Publishing.)

size and crystal structure, and so can be positively charged in many environmental compartments. In the presence of natural organic matter (i.e. humic substances), colloids generally become negatively charged and the point of zero charge shifts to lower values (Amal et al., 1992; Baalousha et al., 2008; Ramos-Tejada et al., 2003). In natural systems, for example freshwater, estuarine, marine and groundwaters, colloids have been observed to have a narrow range of electrophoretic mobilities consistent with the formation of NOM surface coating on all other types of colloids (Beckett and Le, 1990; Hunter and Liss, 1982). Thus, adsorbed NOM molecules dominate colloids surface charge and will have important consequences on their environmental functions and their fate and behaviour. In a few cases, colloids rich in iron oxides (Kaplan et al., 1995; Loder and Liss, 1985; Newton and Liss, 1987) were reported to have a positive surface charge. In aqueous media, the colloidal system as a whole is electrically neutral; oppositely charged ions surround charged particles which balance their surface charge. The distribution of ions in the vicinity of charged particle surfaces may be described by the electric double layer theory (e.g. Stern–Grahame–Gouy–Chapman), which describes the development of the potential with increasing distance from the surface (Figure 4.5). In this model ions are distributed across two layers, a compact inner layer (Stern layer), where the counterions are immobile and a diffuse outer layer, which extends over a certain distance from the particle surface and decays exponentially with increasing distance into the bulk liquid phase. The distribution of ions in the diffuse layer depends on the concentration of the electrolyte, the charge of the ions and the potential at the boundary between the compact inner layer and the diffuse outer layer. The potential at this interface is called the Stern potential. The potential at the shear plane, that is the transition plane from fixed ions and water molecules to those which can be sheared of by fluid motion, is called the zeta potential (ζ), which can be measured by electrokinetic methods (e.g. elec-

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments Stern layer

diffuse layer

123

bulk solution

_ _ _ _ _ _

+

+

+

_

+

+ +

+

_

+

_

+ Ψ Ψs ζ Ψs/e xs

1/κ

x

Figure 4.5 Schematic diagram of the diffuse double layer (DDL) forming from the surface of a colloidal particle into the bulk solution. Abbreviations: zeta potential (ζ), electrostatic potential (Ψ), electrostatic potential at the stern layer (ΨS), Euler’s number (e), Boltzmann constant (k). X is a distance from the surface, Xs is the shear plane, the distance from where ions and molecules are mobile and can be sheared off. (With kind permission from Springer Science+Business Media: Ecotoxicology, 17, 2008, 287–314, The ecotoxicology and chemistry of manufactured nanoparticles, R. D. Handy, F. von der Kammer, J. R. Lead, M. Hassellöv, R. Owen and M. Crane, Figure 2.)

trophoresis). Under conditions of very low ionic strength, the decay of the potential between the Stern layer and the shear plane is negligible and the zeta potential can be seen as an approximate of the Stern potential. For more details about the different models describing the double layer, the reader is referred to the literature (Elimelech et al., 1995a). 4.4.3

Surface Coating by Natural Organic Matter

Natural organic matter (NOM) molecules form surface films of several nanometres on macroscopic surfaces (Lead et al., 2005), manufactured nanoparticles (Baalousha et al., 2008) and natural particles (Baalousha and Lead, 2007; Hunter and Liss, 1982; Loder and Liss, 1985; Wilkinson et al., 1997b), hence the similarity in surface charge of colloidal particles in aquatic environment. Almost all environmental particles, regardless of chemical composition, are negatively charged due to the dissociation

124

Environmental and Human Health Impacts of Nanotechnology

of surface functional groups on sorbed NOM (Hunter and Liss, 1982; Loder and Liss, 1985). This adsorbed surface layer is likely to dominate the surface properties of colloids such as charge. Therefore, a useful approximation in terms of surface charge and aggregation may be to treat colloids as a single class of colloidal materials, irrespective of their nature, (Filella and Buffle, 1993; O’Melia, 1980). However, these surface coatings may be patchy (Gibson et al., 2007), depending on the nature of the underlying substrate, the NOM type and the solution conditions, meaning that this assumption must be tested in most circumstances. The presence of NOM surface coating on environmental colloids was first shown using surface charge measurements by electrophoresis. The use of TEM, AFM and field flow fractionation have given further insight into the thickness and nature of such a surface coating. The formation of surface coating on a mica surface from IHSS Suwannee River FA is shown in Figure 4.6 (Gibson et al., 2007). The thickness of film found was of about 0.4–5 nm. It has been shown that humic substances sorbs to iron oxide colloids (Baalousha et al., 2008), resulting in the formation of nanoscale surface coating. The thickness of this surface coating was found to be of the order of 0.8 nm on iron oxide particles in the presence of 25 mg l−1 humic acid, although aggregation was also increased at these concentrations due to bridging and charge neutralisation. Surface coating of colloids by NOM is likely to affect aggregation behaviour resulting in reduced aggregation through charge stabilization (Jekel, 1986) and steric stabilization mechanisms (Tipping and Higgins, 1982) or enhanced aggregation through charge neutralization and bridging mechanisms caused by fibrillar attachment (Buffle et al., 1998).

4.4.4

Fractal Dimension

Aggregation of natural colloids results in the formation of fractal aggregate structures. A fractal object has a self-similar structure at all levels of magnification, that is it can be sub-divided into parts, each of which is a reduced-size copy to the whole structure. There are three types of fractal structures: exact self-similar, quasi selfsimilar and statistical self-similar. The latter is the weakest type of self similarity, in which the fractal has a statistical numerical measure which is preserved across different scales. Natural colloidal aggregates generally fall under this type. The fractal dimension, D, can be defined as a statistical quantity that gives an indication of how a fractal structure appears to fill space. The fractal dimension can be described by a geometric power law scaling each dimensional geometry (volume (v) or mass (m) for three dimensions D3, projected area (A) for two dimensions D2, or perimeter (P) for one dimension D1) and characteristic length scales (L) of the aggregate (Lee and Kramer, 2004). D1 provides information about the morphology of the aggregate related to the irregularity of the aggregate boundary or perimeter. D2 provides information about the projected area of an aggregate and D3 provides information about the mass distribution within the aggregate. m or v ∝ LD3

A ∝ LD2

P ∝ LD1

(4.1)

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

125

Figure 4.6 A tapping mode image of a humic layer that has a 1 × 1 µm2 area machined away in contact mode. Lines a–c that cut across the image are where the cross-sections below the image were taken. (Reprinted with permission from C.T. Gibson, I.J. Turner, C.J. Roberts, J.R. Lead, Quantifying the dimensions of nanoscale organic surface layers in natural waters, Environmental Science & Technology, 41, 1339–44. Copyright 2007, American Chemical Society.) (See colour plate section for a colour representation)

126

Environmental and Human Health Impacts of Nanotechnology

A summary of studies that applied the concept of fractal dimension to environmental colloidal particles and the technique used is given in elsewhere (Filella, 2007). Clearly, only a few studies have applied the concept of fractal dimension to non-fractionated or colloidal environmental samples such as fluvial particulate matter (Lartiges et al., 2001), marine snow, diatom blooms, estuarine and marine suspended particles, or biological aggregates in wastewater treatment plants. The majority of studies have used synthetic particles such as iron oxide, goethite (Hackley and Anderson, 1989), hematite (Amal et al., 1992; Zhang and Buffle, 1996), montmorilonite or fractionated organic compounds (Chakraborti et al., 2003; Österberg and Mortensen, 1992; Rice et al., 1999; Rice and Lin, 1993; Senesi et al., 1996, 1997). Although scattered, the fractal dimension values reflect the aggregation mechanisms (Section 4.5.3), values of D3 of 1.6–1.9 indicate a diffusion limited aggregation while values about 2.1–2.3 indicate a reaction limited aggregation. Aggregate fractal dimension is an important factor controlling their fate and behaviour and their interaction with other environmental components. Fractal aggregates have higher permeability than that of a hard sphere and their permeability increases with the decrease in fractal dimension value. The settling velocity of fractal aggregates is higher than that calculated by Stokes’ law for impermeable spheres of identical size and mass and settling rate is lower for aggregates with lower fractal dimension (Johnson et al., 1996a; Li and Logan, 2001). Particle capture efficiency during sedimentation increases with the decrease in fractal dimension (Li and Logan, 1997). Adsorption/desorption hysteresis of contaminants to fractal aggregates can be explained by the blockage of the pores within the aggregates after sorption takes place, that is variation in their fractal dimension (Cheng et al., 2004). Aggregate structure also influences disaggregation rate (more details are given in Section 4.6.2).

4.5

Interaction Forces Between Colloidal Particles

The interaction forces acting between colloidal particles play an important role in determining their fate and behaviour such as stability, aggregation and sedimentation (Liang et al., 2007). As surfaces or colloidal particles approach each other to a distance smaller than a few hundred nanometres, surface forces take place and control colloidal stability and aggregation phenomena. An enormous effort has been devoted to study these forces in the last 60 years, that is since DLVO theory was elaborated. DLVO theory describes surface or particle interactions in term of two independent surface forces, the van der Waals and the electrostatic double layer forces. Since that time, research on surface forces has progressed continuously, especially in the last few decades with the invention of surface force measurement techniques such as surface force apparatus (Israelachvili and Adams, 1976) and atomic force microscopy (Binnig et al., 1986). These techniques contributed not only to confirmation of the DLVO theory, but revealed the presence of other forces called non-DLVO forces, such as the hydration, hydrophobic, steric and bridging forces. Reviewed briefly below are the current understanding of the interaction

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

127

forces between colloidal particles and the direct experimental measurements of the force as a function of surface separation carried out for particles immersed in a liquid phase. 4.5.1

DLVO Theory

The abbreviation DLVO refers to the names of Derjaguin, Landau, Verwey and Overbeek. They conducted the first successful attempts to describe colloidal stability interactions in Russia (Derjaguin and Landau, 1941) and Netherlands (Verwey and Overbeek, 1948). The DLVO theory is based on the assumption that forces between surfaces or colloidal particles can be regarded as the sum of two forces. These are the short range, attractive van der Waals and the long range, repulsive electrical double layer forces. The interplay between these two forces has many important consequences on colloid stability and aggregation: VT = VA + VR

(4.2)

where VT is the total interaction energy, VA is the attractive van der Waals energy and VR is the repulsive double layer energy. 4.5.1.1 Van der Waals Forces Van der Waals forces are always short range attractive forces and arise from spontaneous electrical and magnetic polarizations, giving a fluctuating electromagnetic field within the media and in the gap between surfaces or particles (Elimelech et al., 1995a). There are two approaches to calculate the van der Waals forces: microscopic and macroscpic. In the microscopic approach, the interaction force is the pairwise summation of all relevant interatomic interactions (Hamaker, 1937) and can be described in terms of geometrical parameters and a constant A, the “Hamaker constant”. For two spheres of equal radius, R, at a surface to surface separation distance, h, apart along the centre to centre axis, the total interaction energy can be given as (Liang et al., 2007): VA( h) = −

A  2R2 2R2 4R2    + + ln  1 −  2 2   6  h + 4 Rh ( h + 2 R) ( h + 2 R)2  

(4.3)

In the case of the interaction between a sphere and a plane at a distance, h, the interaction energy can given as: VA( h) = −

( )

A R R h  + + ln 6  h h + 2 R h + R 

(4.4)

The Hamaker constant depends on material properties such as density and polarizability. The effective Hamaker constant depends also on the dispersion medium. It is generally of the order of magnitude 10−20–10−21 J (Elimelech et al., 1995a). Aeff ≈ ( Aparticle − Amedium )

2

(4.5)

128

Environmental and Human Health Impacts of Nanotechnology

The macroscopic approach overcomes the summation assumption by considering the macroscopic electromagnetic properties of the medium (Lifshitz, 1956) where atomic structure is neglected and large bodies are treated as continuous media and forces are derived in terms of the bulk properties such as dielectric constants and refractive indices. However, the use of the macroscopic approach is limited by the computation required and the lack of appropriate dielectric data. Additional details about these two approaches and calculations of Hamaker constant can be found elsewhere (Bergstrom, 1997; Elimelech et al., 1995a; Israelachvili, 1972). 4.5.1.2

Double Layer Interaction

Colloidal particles often carry an electrical charge and therefore attract or repel each other. When two like-charged particles approach each other, their electrical double layer starts to overlap, resulting in a repulsive force which opposes further approach. For identical particles, sphere–sphere double layer interaction energy can be given by Equation 4.6. There are many expressions available based on various assumptions for sphere–sphere double layer interaction energy and readers are referred to the literature for more details (Bell et al., 1970; Carnie et al., 1994; Genxiang et al., 2001; McCormack et al., 1995; Sader et al., 1995; Stankovich and Carnie, 1996). VR( h) = 32πε R

( )γ kT ze

2

2

exp ( −κ h)

(4.6)

For small values of surface or zeta potential (ζ), this simplifies to: VR( h) = 2π eRζ 2 exp ( −κ h)

(4.7)

where ε is the permittivity of the medium, R is the particle radius, γ is dimensionless functions of the surface potentials, k is the Boltzman constant, T is the absolute temperature (Kelvin), h is the surface-surface separation between particles (m), e is the electron charge and κ is the inverse of Debye–Huckel screening length (m−1). Equation 4.7 is applicable only if κR > 5 and h << R. For the general case of electrolyte solutions containing a number of dissolved salts, κ is defined by:  e 2 ∑ ni zi2  κ=   ε kT 

(4.8)

where n is the number concentration of ion i in the solution. Inserting numerical values appropriate to aqueous solutions at 25 °C and converting the ion concentration into molar terms gives:

κ = 2.32 × 109

(∑ ci zi2 )

(4.9)

where c the concentration of ions expressed in mol l−1 and z the valency of the ions. The length 1/κ is known as the thickness of the diffuse layer. Equation 4.9 shows that the increase in ionic strength results in a decrease in the thickness of the diffuse layer and a consequent decrease in the repulsive interactions among particles. Typical values of the diffuse layer thickness, 1/κ, are in the range 1–100 nm.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

4.5.2

129

Stability Criteria

According to DLVO theory, various parameters can affect colloidal stability, such as ion type and concentration, the value of ζ potential and the particle size. According to Equation 4.9, an increase in the ionic strength results in a decrease in the diffuse layer thickness and, therefore, a decrease in the double layer repulsive force. Polyvalent electrolytes induce larger decreases in the diffuse layer thickness than monovalent electrolytes and consequently induce greater aggregation. According to Equation 4.7, the electrostatic repulsive force is proportional to the square of ζ potential, that is a doubling of the zeta potential results in quadrupling the repulsive force, and so it is a key parameter in determining the stability of colloids. However, for some environmental colloids such as humic sbstances, the physical meaning of ζ has been questioned (Duval et al., 2005). According to Equation 4.4 and Equation 4.6, both attractive and repulsive forces are proportional to the particle size. At small sizes, the value of VT (total interaction energy) is directly proportional to the size. However, at large sizes the value of VT has a more complicated variation. In all cases, electrostatic stability increases with increasing particle size. 4.5.3 Aggregation Kinetics For dilute colloidal system where only binary collisions are assumed to take place, the kinetics of particle coagulation due to Brownian motion can be described by the Smoluchowski rate equation (Holthoff et al., 1996): ∞ dN z 1 = ∑ kij N i Nj − N z ∑ kiz N i dt 2 i + j =z i =1

(4.10)

where t is the time, Nz is the concentration of z-fold aggregates and kij is the rate at which i-mers particle bind to j-mers, following the diffusion of particles toward each other. According to Smoluchowski, the aggregation rate constant for the formation of dimers from monodisperse suspension of monomers can be given by: k11 = 2ks =

8kBT 3η

(4.11)

where kB is Boltzmann constant, T is the absolute temperature and η is the viscosity of the fluid. This constant k11 is thus independent of the size of particles. Considering the van der Waals forces and the hydrodynamic interactions, the coagulation rate can be can be expressed as: ∞

  V ( h) β ( h) k11 = 2ks  ∫ exp  A  dh  2   k T ( ) + h 2 R B 0 

−1

(4.12)

where β(u) is the correction factor for the diffusion coefficient (Overbeek, 1982), h is the distance between particle surfaces and R is the particle radius. Equation 4.12 is characteristic of the fast or diffusion limited aggregation regime. In the regime of slow or reaction limited aggregation, additional repulsive forces due to electrostatic interactions may prevent the particles from aggregating. In such a

130

Environmental and Human Health Impacts of Nanotechnology

situation, only a fraction of collisions, α (collision efficiency factor), are successful. The collision efficiency factor expresses the ratio of the number of collisions leading to aggregation to the number of collisions due to Brownian motion according to Equation 4.13:

α=

∞

  1 k11 V ( h) β ( h) = = exp  A  dh   kBT   W ( k11 ) fast , VR =0  ∫0 ( h + 2 R)2

∞

 β ( h) V ( h)   dh exp  T ∫ ( 2  kBT   ) + 2 h R 0 (4.13)

W is the stability ratio, which is defined as the ratio of the fast, diffusion limited aggregation rate to the slow, reaction limited aggregation rate (Elimelech et al., 1995b). 4.5.4

Non-DLVO Interactions

Recently, DLVO theory has been found unable to fully describe colloidal behaviour in aquatic and terrestrial environments (Grasso et al., 2002; Sander et al., 2004). The structure of water, adsorbed or dissolved entities (e.g. organic molecules) close to the colloid surface in the water body may result in non-DLVO effects. Much research has been conducted to understand these forces such as hydration, hydrophobic, steric and bridging interactions, and to extend DLVO theory to account for them. Although understanding has improved significantly, much environmental research has ignored them (Grasso et al., 2002). However, in many aggregation studies, reversible aggregation cannot be fully interpreted by DLVO theory (van der Waals and electrostatic forces) alone. Therefore, in such situations, some other short range forces should be considered. 4.5.4.1

Hydration Effect

In aquatic systems most colloidal particle surfaces carry a surface charge and surface functional groups, which are expected to be hydrated, analogously to ions in solution. Therefore, aquatic colloids such as humic substances, clay particles and metal oxides are generally expected to be surrounded by a layer of water. The nature of this hydration layer can be different from that of the bulk water (Figure 4.7). This hydration layer plays an important role in the interactions of these colloids, and usually gives an extra repulsion to that induced by the double layer. For true contact to occur between particles, surfaces need to become dehydrated, hence a repulsion due to hydration occurs (Grasso et al., 2002). Direct evidence for the hydration repulsion comes from force measurements. Atomic force microscopy (AFM) enables direct measurement of forces between a planar surface and an individual colloid particle (Butt et al., 1995; Cappella and Dietler, 1999). Measurements using AFM tips or, more quantitatively, silica particles generally show good agreement with DLVO forces, although at short separation distances (2–3 nm) a deviation was observed (Figure 4.8) (Ducker et al., 1991, 1992). At these distances the DLVO theory predicts that attractive van der Waals force will exceed the repulsive double layer force, although the measured repulsive force was greater than the predicted, which has been attributed to hydration forces. This force is dependent on solution conditions. Hydration forces are more impor-

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

131

Figure 4.7 Hydration layers around a particle. The density of water molecules around a surface is highest in the proximity of the surface and decays with distance to the density observed for water molecules in the bulk solution. This results in several layers of hydration (1, 2, …. , n). (With kind permission from Springer Science & Business Media: Reviews in Environmental Science and Biotechnology, 1, 2002, 17–38, A review of non-DLVO interactions in environmental colloidal systems, D. Grasso, K. Subramaniam, M. Butkus, K. Strevett and J. Bergendahl, Figure 6.)

10 F/R (mN m–1)

F/R (mN m–1)

10

1 10–4 M

0.1

10–1 M 10–2 M 0

10

20

Distance (nm) (a)

10–3 M 30

40

pH 10.0 pH 7.0 pH 4.0 pH 3.0 pH 2.58 pH 2.0 DLVO fit

1

0.1

0.01 0

10

20

30

40

50

60

Distance (nm) (b)

Figure 4.8 The force F as a function of distance, D: (a) between a silica probe radius R = 3.5 µm and and a flat silica surface; (b) between a silica glass sphere and a flat silica plate as a function of pH. The data points represent the measurements of the surface force and the solid lines are the best fit to DLVO theory. ((a) Reprinted by permission from Macmillan Publishers Ltd, Nature, 353, 239–41, Direct measurement of colloidal forces using an atomic force microscope, W.A. Ducker, T.J. Senden and R.M. Pashley. Copyright 1991. (b) Reprinted with permission from W.A. Ducker, T.J. Senden and R.M. Pashley, Measurement of forces in liquids using a force microscope, Langmuir, 8, 1831–6. Copyright 1992, American Chemical Society.)

tant at higher ionic strength, owing to adsorbed hydrated cations as shown in Figure 4.8a (Ducker et al., 1991) and at higher pH conditions due to the higher surface charge as shown in Figure 4.8b (Ducker et al., 1992). Meagher (1992) measured hydration forces between a silica colloidal sphere and a silicon sample. It was shown that in 0.01 M CaCl2 solutions at pH 4.1, the

132

Environmental and Human Health Impacts of Nanotechnology

force–distance curve fits well the DLVO theory for distances >2 nm. However, below 2 nm, a repulsive force is probed and was attributed to hydration force (Meagher, 1992). Further, hydration forces were observed between an alumina (Al2O3) tip and a mica surface at different pH values and between a silicon nitride tip and a mica surface at high concentrations of divalent cations (>3 M) (Butt, 1991), between silica surfaces in 1,2-ethanediol and water (Atkins and Ninham, 1997), between gold surfaces in sodium chloride (Biggs et al., 1994), between silica surfaces and silicon on titanium dioxide at high pH (Larson et al., 1993) and between an alumina surface and an aluminum or silicon nitride tip (Karaman et al., 1997). The hydration force was also measured between two mica surfaces in electrolyte solution (Israelachvili and Pashley, 1983; Pashley, 1981; Pashley and Israelachvili, 1984a). They measured a short range repulsive force in addition to van der Waals forces at high salt concentrations, which varied with the type of cation in solution. The more hydrated cations, such as Mg2+ and Ca2+, gave stronger repulsive forces than the less hydrated monovalent ions, such as K+ and Cs+. Other studies have suggested that hydration forces are oscillatory and can be either attractive or repulsive (Israelachvili and Wennerstrom, 1996; Pashley and Israelachvili, 1984b). These hydration forces should be present in aquatic colloidal particles and can be dominant in those with high negative charge densities. Clearly further theoretical and experimental work is needed to explore hydration forces in environmental colloids. 4.5.4.2

Hydrophobic Interactions

A hydrophobic surface is one that has low affinity for water and has no polar or ionic groups or hydrogen bonding sites. The nature of water in contact with such a surface is different from the bulk water. Bulk water is significantly structured via the formation of hydrogen bonds between the water molecules, resulting in the formation of large clusters of hydrogen bonded water molecules. The presence of a hydrophobic surface will most likely restrict such phenomena and water confined between two hydrophobic surfaces will not be able to form clusters larger than a certain size, causing water molecules to tend to migrate to the bulk water where there is unrestricted hydrogen bonding opportunities and a lower free energy (Elimelech et al., 1995a). Hydrophobic forces can be important, giving an extra attraction between surfaces or particles. Attraction between hydrophobized mica sheets, via surface adsorption of hydrocarbon and fluorocarbon surfactants, has been directly measured (Israelachvili and Pashley, 1984) and was found to operate over a long range of about 80 nm and to be much stronger than the van der Waals force (Claesson and Christenson, 1988). The magnitude of hydrophobic forces was found to decrease with the increase in electrolyte concentration (0.01–0.1 M magnesium sulfate) (Christenson et al., 1990). Figure 4.9 shows the long range (about 80 nm) attractive hydrophobic forces measured between a silicon nitride (Si3N4) tip and a hydrophobic mica surface prepared by depositing a monolayer of cetyltrimethylammonium bromide (CTAB) on the surface of freshly cleaved mica surfaces (Teschke and de Souza, 2003). More details can be found elesewhere (Christenson and Claesson,

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

133

0.0 Dieletric Constant (E)

Force (nN)

0.5

–0.5

–1.0 0

20

80 60 40 20 0

0

10

20 30 40 Separation (nm)

40 60 Separation (nm)

80

50

100

Figure 4.9 Force vs separation curve (•) between a standard silicon nitride (Si3N4) tip and a bare mica surface (䊊) between a neutral tip and hydrophobic mica surface prepared depositing a monolayer of CTAB on the surface of freshly cleaved mica surfaces. (Reprinted from Chemical Physics Letters, 375, 540–6, O. Teschke and E.F. de Souza, Measurements of longrange attractive forces between hydrophobic surfaces and atomic force microscopy tips. Copyright 2003, with permission from Elsevier.)

2001; Eriksson et al., 1989; Ruckenstein and Churaev, 1991). It is possible for the surface of particles dispersed in environmental waters to have some degree of hydrophobicity and the hydrophobic force has been postulated to contribute to the aqueous aggregation of clay particles (Zbik and Horn, 2003). 4.5.4.3

Steric Interactions

Adsorption of natural organic matter to colloidal particles is a well known process in aquatic systems as discussed previously in Section 4.4.3 (Gibson et al., 2007; Tipping and Higgins, 1982; Wilkinson et al., 1997a). Adsorbed NOM molecules can play a very important role in the aggregation and deposition phenomena. In some cases, adsorbed NOM molecules (e.g. polysaccharides) can induce aggregation by bridging mechanisms (Section 4.5.4.4), while in other cases NOM molecules such as fulvic and humic acids can enhance colloidal stability (Wilkinson et al., 1997a) by a mechanism known as steric stabilization. The adsorbed molecule chains extend some distance into the water, giving increased stability to colloidal particles. As particles approach each other, the adsorbed layer of NOM comes into contact, resulting in the interaction between these molecules. As these molecules are hydrated, any interaction will induce hydration repulsive forces as described in the previous section. The steric stabilization effect increases with the load with NOM or the thickness of the adsorbed layer.

134

Environmental and Human Health Impacts of Nanotechnology

The stability of colloidal particles in an aquatic environment is often higher than expected on the basis of zeta potential and ionic strength, which is likely related to a steric stabilization effect induced by NOM surface coating (Jekel, 1986) and possibly to the hydration effect explained in Section 4.5.4.1. Direct measurements of forces between colloidal particles are scant, though some have been performed by AFM (Assemi et al., 2004; Mosley et al., 2003; Sander et al., 2004). Assemi et al. (2004) investigated the interaction forces between a goethite coated mica surface (positively charged) and silica colloidal probe (negatively charged). The adsorption of humic substances onto a goethite coated mica surface (imparting negative charge) induces repulsion between the goethite surface and silica colloidal probe. In addition, a high repulsion force was observed at short separation distances (typically < 2–3 nm) and was attributed to steric forces induced by the sorption of humic substances. Mosley et al. (2003) investigated the effect of adsorbed NOM, solution pH and ionic composition on the force–distance curve between natural colloids represented by surface film of iron oxides precipitated onto spherical SiO2 particles. At low ionic strength, the interparticle forces were dominated by electrostatic repulsion from the dissociation of functional groups on the NOM. At small separation distances (<10 nm) another repulsive force originating from steric interferences of NOM molecules, and at shorter distances from the hydration effect, were also present. At high ionic strength or low pH conditions, steric repulsion forces dominate due to the absence of electrostatic forces (Mosley et al., 2003). Sander et al. (2004) investigated the effect of adsorbed layer of humic acid on interparticle forces in natural colloids (iron oxide and alumina) (Sander et al., 2004). Figure 4.10a shows the interaction forces, measured by atomic force microscopy at different pH conditions, between a silica particle (4–5 µm in diameter) and a flat quartz plate, both

pH 4.0 pH 6.1 pH 9.0

1

0

6 F/R (mN/m)

F/R (mN/m)

Iron oxide

5 4

Iron oxide Humic Acid layer encountered

pH 4.0 pH 6.1 pH 9.2

3 2 1 0

5

10 15 20 Separation (nm) (a)

25

10

20 30 40 Separation (nm) (b)

50

Figure 4.10 Interaction forces as a function of separation distance at 0.001 M NaCl and at different pH conditions for (a) iron oxide coated silica and (b) iron oxide coated silica in the presence of adsorbed layer of SRHA molecules. (Reprinted with permission from S. Sander, L.M. Mosley, K.A. Hunter, Investigation of interparticle forces in natural waters: effects of adsorbed humic acids on iron oxide and alumina surface properties, Environmental Science & Technology, 38, 4791–6. Copyright 2004, American Chemical Society.)

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

135

coated with hydrous iron oxide to represent natural iron oxide particles. Figure 4.10b shows the interaction forces with the same system, but coated with NOM (Suwannee river humic acid, SRHA) (Sander et al., 2004). The adsorption of humic acid to the surfaces results in significant changes in the force–distance relationship. 4.5.4.4

Polymer Bridging

Long chain organic molecules such as polysaccharides can attach to two or more particles, thus bridging them together. Therefore, particles can form aggregates or attach to surfaces even though they may be charged and repel each other, forming open flocs. An example of the formation of such aggregates in a mixture of haematite and alginate in the presence of 6.1 mM CaCl2 is shown in Figure 4.11. Direct measurements of bridging forces have been performed by AFM (Biggs et al., 2000; Sander et al., 2004). A distance–force relation ship measured by AFM in the approach and separation force curves is shown in Figure 4.12. In the absence of SRHA (Figure 4.12a), the approach and separation curves are identical. In the presence of SRHA (Figure 4.12b), the separation curve shows a decrease in the repulsive force for NaCl in comparison to the approach curve. However, in the presence of CaCl2, the separation curve exhibits a significant hysteresis with greater attractive force close to the surface. This is explained by bridging attraction, that is humic acid molecules function as an adhesive between the surfaces. More details can be found elsewhere (Biggs, 1995; Dickinson and Eriksson, 1991; Grasso et al., 2002).

Figure 4.11 Combined hematite–alginate gel aggregate in the presence of 6.1 mM CaCl2 at pH 5.2. (a) and (b) are TEM images of the same aggregate, but at different magnifications. (Reprinted from K.L. Chen, S.E. Mylon, M. Elimelech, Aggregation Kinetics of Alginate-Coated Hematite Nanoparticles in Monovalent and Divalent Electrolytes, Environmental Science & Technology, 40, 1516–23. Copyright 2006, American Chemical Society.)

136

Environmental and Human Health Impacts of Nanotechnology 1.0 NaCl only

CaCl2-Approach CaCl2-Separation NaCl-Approach NaCl-Separation

0.5

0.4 Approach Separation

0.2 0

F/R (mN/m)

F/R (mN/m)

0.6

0.0

–0.5

10

20 30 Separation (nm) (a)

40

Bridging attraction

–1.0 5

10 15 20 25 Separation (nm) (b)

30

Figure 4.12 Normalized force as a function of surface separation distance for both approach and separation for alumina surface; (a) in 0.001 M NaCl solution at pH 6.2 and (b) with adsorbed SRHA layer at pH 6 in 0.01 M CaCl2 and 0.01 M NaCl solution. (Reprinted with permission from S. Sander, L.M. Mosley, K.A. Hunter, Investigation of interparticle forces in natural waters: effects of adsorbed humic acids on iron oxide and alumina surface properties, Environmental Science & Technology, 38, 4791–6. Copyright 2004, American Chemical Society.)

4.6

Fate and Behaviour of Colloids in Aquatic Systems

4.6.1 Aggregation 4.6.1.1

General

Electrostatic stabilization and steric stabilization are the two main mechanisms for colloid stabilization in aquatic systems, as described in Section 4.5. Unstable colloidal systems form flocs or aggregates due to interparticle attractions. Aggregation can be accomplished by a number of following methods: (i) Removal of the electrostatic barrier (Section 4.5.3) that prevents aggregation of particles. This can occur due to the increase in salt concentration (Baalousha et al., 2006b) or alteration of the pH (Baalousha et al., 2008) of a suspension to effectively neutralize or screen the surface charge of the particles. This reduces the repulsive forces that keep colloidal particles separate and allows for coagulation due to van der Waals and possibly other forces. Estuarine systems, which exhibit a sharp increase in salinity, often exhibit rapid and nearly complete aggregation, a process to be borne in mind when considering manufactured nanoparticle behaviour. Nevertheless, see the discussion of non-DLVO forces which may maintain dispersion even under these conditions. (ii) In water treatment facilities, addition of a charged polymer flocculant. Polymer flocculants can bridge individual colloidal particles by attractive electrostatic

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

137

interactions. For example, biopolymers can bridge natural colloidal particles and induce their aggregation (Chen et al., 2006) (Section 4.5.4.4). In environmental systems, the size distribution of organic and inorganic colloids is thought to follow a Pareto-type law in the nanometre to micrometre range (Filella and Buffle, 1993). This size distribution represents that of individual colloids and their aggregates. Quantitative physico-chemical aggregation theory (DLVO theory, Section 4.5.1) exists only for identical, compact, spherical particles (homoaggregation). However, there is no such general theory for aggregation of a mixture of different particles (heterocoaggregation), in particular for aggregation involving polymers. The section below summarizes the present understanding of aggregation properties of the major aquatic colloids. Surface film formation may reduce the problem to one of homoaggregation in some situations. 4.6.1.2

Role of Natural Organic Matter

Some work has investigated the effect of the different fractions of NOM (humic substances and biopolymers) on colloid stability. The available literature suggests that NOM can have different effects on natural colloids, as described in Figure 4.13. Humic substances form a surface coating on inorganic colloids, in general enhancing colloidal surface charge and, therefore, their stability. However, in the presence of a high concentration of cations, Humic substances can enhance aggregation via bridging mechanisms. The net effect will depend on surface coverage and the degree of charge alteration. For model compounds, it has been shown that the adsorption of negatively charged humic substances to positively charged iron oxide will result in destabilization only for low surface coverage (Baalousha et al., 2008; Ferretti et al., 1997; Stumm, 1992). Steric repulsion has also been suggested as a mechanism for enhanced stability caused by Humic substances. Biopolymers influence colloid stability either by being adsorbed to colloidal surfaces (Dickinson and Eriksson, 1991) or by being expelled from the area between the particles in case of non-adsorbing polymers (depletion layer) (Jenkins and Snowden, 1996; Tuinier et al., 2003). Aggregation by a depletion mechanism is not a common process in environmental colloids. The adsorption of small quantities of polymers leads to colloidal aggregation by charge neutralization or colloid bridging, whereas the adsorption of larger quantities may stabilize the colloidal suspension via steric stabilization mechanism. Analysis by microscopy of freshwater colloids, often shows small inorganic colloids embedded in networks of fibrillar materials (Figure 4.1b). Interaction of inorganic colloids with biopolymers is likely due to the minimal electrostatic repulsion because of low surface charge density of biopolymers (Section 4.3.2.2). In such a situation a highly stable colloidal suspension might produce large aggregates in the presence of biopolymers. Because biopolymers are very long in comparison with the colloid diameter, they can serve as long bridges between colloids. The attached colloid may interact with another polymer, leading to the formation of loose aggregate networks extending to a large dimension. Further, humic substances may aggregate as small spheroids along the fibril of biopolymers (Buffle and Leppard, 1995) suggesting that humic substances might interact with fibrils

138

Environmental and Human Health Impacts of Nanotechnology

Small Aggregates (Stable Suspension)

Gels

Large Aggregates (Untable Suspension)

Figure 4.13 Major types of aggregates formed in the three-colloidal component system: fulvic acid = small points; inorganic colloids = circles; rigid biopolymers = lines. Both fulvic acids and polysaccharides can also form gels, which are represented here as gray areas into which inorganic colloids can be embedded. (Reprinted with permission from J. Buffle, K.J. Wilkinson, S. Stoll, M. Filella, J. Zhang, A generalized description of aquatic colloidal interactions: the three-colloidal component approach, Environmental Science & Technology, 32, 2887–99. Copyright 1998, American Chemical Society.)

similarly to inorganic colloids. Such an aggregation process can be modelled numerically (Stoll and Buffle, 1996). 4.6.1.3

Role of Ionic Strength and Cations

According to DLVO theory, an increase in ionic strength results in the screening of surface charge, shrinkage of double layer and, subsequently, colloid aggregation. In low ionic strength freshwaters, compact reaction limited aggregation (RLA) type aggregate (Figure 4.14, Table 4.1) structures should be observed more frequently due to the low collision efficiency. However, at high ionic strengths such as marine systems, loose aggregates with a diffusion limited aggregation (DLA) type structure are more likely to be observed due to the higher collision efficiency (Leppard et al., 1986, 1997; Wilkinson et al., 1999). In the presence of high ionic strength, humic substances aggregate to small spheroids of about 10 nm or large, fractal aggregates of several micrometers (Baalousha et al., 2005, 2006b).

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

ted imi

139

e

rat

nl

LA) (= D rea (= R ctio LA) n li mit ed rate

sio

u diff

+

Figure 4.14 Schematic representation of the diffusion limited aggregation (DLA) and a reaction limited aggregation (RLA) mechanisms, leading to the formation of either loose aggregates of low fractal dimension or denser aggregates of higher fractal dimension. (Reprinted with permission from J. Buffle and G.G. Leppard, Characterization of aquatic colloids and macromolecules. 1. Structure and behavior of colloidal material, Environmental Science & Technology, 29, 2169–75. Copyright 1995, American Chemical Society.)

Table 4.1

Comparison between diffusion and reaction limited aggregation regimes.

Energy barrier Type Collision efficiency Aggregation rate Fractal dimension

DLA

RLA

Absence or primary minimum Reversible 1 Rapid (Kd) ∼1.8

Secondary minimum Irreversible <1 Slow (KR) ∼2.3

In addition, salts (sodium chloride, magnesium chloride and calcium chloride) induce the aggregation of alginate coated haematite particles through electrostatic destabilization (Figure 4.15a). However, in the presence of calcium chloride, aggregation rate was much higher than that which conventional diffusive aggregation predicts (Figure 4.15b). The observation of higher aggregation rate than that which would be observed in the presence of a simple electrolyte (Chen et al., 2006) suggests that charge neutralization mechanism is not sufficient to explain the aggregation mechanism. ‘Excess’ aggregation was explained by the formation of an alginate coated hematite gel network and the cross-linking between unadsorbed alginate, via Ca2+ bridging, that might form bridges between hematite–alginate gel structures (Chen et al., 2006).

140

Environmental and Human Health Impacts of Nanotechnology

Figure 4.15 Enhanced aggregation of alginate-coated hematite nanoparticles in the presence of CaCl2. (a) Comparison of aggregation profiles of bare and alginate-coated hematite nanoparticles in the presence of monovalent and divalent electrolytes. The bare hematite and alginate-coated hematite nanoparticles are aggregated at pH 12.2 and 5.2, respectively. For the four aggregation profiles, the particle concentration employed is 1.5 × 108 particles per ml, and the unadsorbed alginate content for the alginate-coated hematite samples is 74.8 g/l. (b) Apparent attachment efficiencies of alginate-coated hematite particles as a function of CaCl2 concentration at pH 5.2. For all the aggregation experiments, the particle concentration employed is 7.5 × 107 particles per ml, and the unadsorbed alginate content is 37.4 g/l. (Reprinted from K.L. Chen, S.E. Mylon, M. Elimelech, Aggregation Kinetics of Alginate-Coated Hematite Nanoparticles in Monovalent and Divalent Electrolytes, Environmental Science & Technology, 40, 1516–23. Copyright 2006, American Chemical Society.)

4.6.2

Disaggregation: Effect of Natural Organic Matter

Disaggregation is a very important process in determining colloidal fate and behaviour and interaction with trace contaminants, but few studies are available on the disaggregation of natural colloidal particles. Recent work (Baalousha, 2008) investigated the role played by Suwannee River humic acid on the disaggregation of synthesized iron oxide NPs. NOM has been shown to induce the disaggregation of iron oxide aggregates (5–10 µm), likely due to formation of surface coating of NOM on the surface and pore surfaces of aggregates and thus the enhancement of surface charge as confirmed by electrophoretic mobility measurements. This induces an increase of the degree of repulsion within the aggregate matrix and results in aggregate rupture. A previous study has shown that synthetic polymers are able to separate two aggregated colloids, even when the separation distance was of the order of few nanometres (primary minimum) (Ouali and Pefferkorn, 1994; Pefferkorn, 1995). There are two possible mechanisms of aggregate break-up based on aggregate structure: surface erosion, which tends to be slow, and large scale fragmentation,

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

141

which is more rapid. In surface erosion, small particles are separated from the surface of the aggregate while in large scale fragmentation the aggregates split into pieces of comparable sizes (Jarvis et al., 2005). Highly branched aggregates with a small fractal dimension break by this fragmentation mechanism, while compact aggregates with a large fractal dimension favour a surface erosion mechanism (Yeung and Pelton, 1996). Cleary more research is needed to investigate the possible disaggregation of natural environmental aggregates and the role played by natural organic molecules. The findings have clear implications for the environmental fate and behaviour of manufactured NPs. 4.6.3

Sedimentation Behaviour

The settling behaviour of a hard, non-permeable sphere can be described in a relatively straightforward manner by Stokes’ law. However, aggregation of colloidal particles in natural waters results in the formation of large, fractal and permeable aggregates (Section 4.4.4) (Johnson et al., 1996a; Lartiges et al., 2001). Thus, Stokes’ law is not suitable to describe the settling behaviour of natural colloidal aggregate. The settling behaviour of such aggregates is dependent on the drag force and permeability of the solvent through the porous aggregates. Fractal aggregates in natural waters have a heterogeneous mass distribution and porosity, resulting from the coagulation of small and more densely packed clusters into larger and overall less dense aggregates (Johnson et al., 1996a; Lartiges et al., 2001). Pores formed within the fractal aggregate will permit greater interior flow through the aggregate, resulting in a faster settling velocity. It has been demonstrated that fractal aggregates (with heterogeneous pore sizes) settle faster than predicted by Stokes’ law (Johnson et al., 1996a; Logan and Hunt, 1987), indicating that intra-aggregate flow reduces the drag for aggregates compared to that for the equivalent impermeable particles. As the fractal dimension increases, the permeability decreases and the fluid mechanics resembles more closely that of an isolated impermeable sphere (Chellam and Wiesner, 1993). The settling behaviour of fractal aggregates is not well understood since these aggregates do not behave as spheres with constant density. The settling behaviour of fractal aggregates depends on many properties, including porosity, size, permeability and buoyant density, which need be determined to predict fractal aggregate sedimentation. Several models have been developed to predict the sedimentation behaviour of fractal aggregates (Tang et al., 2002; Tang and Raper, 2002). It has been found difficult to describe mathematically the non-homogeneous distribution of pores within fractal aggregates, although this non-homogeneity has been modelled by assuming that porosity varies radially from the centre of gyration (Veerapaneni and Wiesner, 1996).

4.7

Fate and Behaviour of Colloids and Nanoparticles in Porous Media

The transport of colloids in porous media is dominated by inter alia size, shape, surface properties, the physical-chemical properties of the porous medium (grain size, surface properties) and the fluid (velocity, ionic composition, density and

142

Environmental and Human Health Impacts of Nanotechnology

viscosity). Colloid transport or deposition in porous media can be thought of as occurring in two steps (McDowell-Boyer et al., 1986): (i) transport to the vicinity of the soil or sediment grains themselves (collector) by surface filtration, straining filtration, diffusion or physical-chemical mechanisms leading to collision; and (ii) the attachment to the collector via electrostatic interactions between the colloid and the soil. Attachment and aggregation were discussed in Section 4.5. Large particles can be removed by surface filtration (a filter cake or a surface mat formation) above the media. In saturated media, colloids can be removed mechanically by sieving or straining in smaller pore spaces if colloids or aggregate dimensions exceed the pore space. Straining filtration may occur even for particles much smaller than the average grain size of narrowly distributed grains. Under strongly repulsive conditions with no physico-chemical filtration, Xu et al. (2006) determined that for straining filtration to become negligible, the ratio of grain size to colloid size must be larger than 125. For smaller colloids, physical-chemical mechanisms can be further divided into three deposition processes: diffusion, interception, and sedimentation (Figure 4.16). Diffusion limited deposition is driven by Brownian motion of the colloids, hence, it strongly depends on their size. Small colloids (<100 nm) in particular are deposited by collision to the porous media due to diffusion. Interception describes the collision between colloid and the collector if the flow path of a colloid with a defined diameter intersects the stationary collector. This process is especially important for colloids/aggregates larger 1 µm. Sedimentation of colloids is due to a difference in colloid to fluid density (mainly for larger (>200 nm) or dense colloids) (McDowell-Boyer et al., 1986). These three processes relate to the three main aggregation processes, perikinetic (diffusion), orthokinetic (shear) and differential settling. In addition, for unsaturated systems colloid retention at the stationary,

collector (grains) C

hydrodynamic flow vectors particle trajectories

A

A particle filtration by diffusion B B particle filtration by interception C particle filtration by settling

Figure 4.16 Particle filtration by (A) diffusion, (B) interception and (C) sedimentation. (Adapted from K.M. Yao (1968), Influence of suspended particle size on the transport aspect of water filtration, University of North Carolina, Chapel Hill, Dissertation.)

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

143

mobile or transitional gas–water interface has to be taken into account (Crist et al., 2004; Lenhart and Saiers, 2002; Ouyang et al., 1996; Wan and Tokunaga, 1997). While the maximum size of mobile colloids will be limited by straining filtration and pore velocity, concentration and size distribution of smaller colloids are controlled by physico-chemical filtration. Altogether, subsurface transport of small (<100 nm), individual colloids is limited mainly by the diffusion driven collision rate, whether the transport of larger colloids or aggregates (>1 µm) is limited by straining (if colloid density equals fluid density) or sedimentation (for colloid density > fluid density). 4.7.1

Saturated Porous Media

There are many approaches available to describe the transport of colloids themselves or contaminants associated with colloids in saturated porous media. As already pointed out, the reactions taking place in the porous medium can be separated into the hydraulic part (the movement of the colloids with the water stream), the collision part and the attachment/detachment part. The so-called classical filtration theory (CFT) provides adequate analytical solutions for clean bed filtration under favourable chemical conditions. Other approaches developed the advection– dispersion–deposition equation to account for charge heterogeneity and dynamic processes during the filtration of colloids. For an extensive overview, the reader is pointed towards Elimelech et al. (1995b), Kretzschmar et al. (1999) and Ryan and Elimelech (1996). Colloid transport in saturated porous media has been successfully described by the CFT. (Iwasaki, 1937) described filtration as a first order kinetic law: C P = C0 P e(− λ⋅∆x)

(4.14)

where CP is the colloid concentration at the travel distance x, C0P is the original colloid concentration, λ is the filtration factor in m−1. The colloid travel distance until certain retention is reached can be expressed as the reciprocal value of the filtration factor:

(

RT = − 1 λ ln C P C 0P

)

(4.15)

where RT is the colloid travel distance in metres at a given retention rate (e.g. ln (CP/C0P) = 6.9 for 99.9% colloid retention). The relation between the filtration factor and travel distance is given in Figure 4.17. Developed by Yao (1968) for clean bed filtration and extended by Rajagopalan and Tien (1976), the CFT enables the calculation of the filtration factor (and thus colloid transport prediction in saturated media). The equations by Yao (1968) to quantify filtration (λ) by the four different mechanisms of colloid filtration, that is sieving/straining, sedimentation, diffusion and interception, neglected the influence of adjacent collector grains and the hydrodynamic retardation (Tien and Payatakes, 1979). To account for these processes, (Rajagopalan and Tien, 1976) introduced the Happel parameter (AS) to account for hydrodynamic retardation. It can be described by the Happel sphere in cell model, in which the grains with the porous

144

Environmental and Human Health Impacts of Nanotechnology 100

Travel Distance [m]

99,99 % deposition 99,9 % deposition 10

RT = − 1 ln  C  λ  C0  1

90 % deposition 37 % deposition 0,1 0,1

1

10

100

Filtration Factor [1/m]

Figure 4.17

Relation between filtration factor and travel distance for a given deposition rate.

media are single spherical collectors surrounded by an imaginary fluid sphere. The fluid envelope contains the same amount of water as the relative volume of fluid to the collector in the entire porous media. Colloids can be transported to the imaginary single collector by different mechanisms, each with a single filtration/ deposition efficiency (γ) (Tufenkji and Elimelech, 2004):

λ=

γ 3 ( 1 − n) γ α = α ( filtration factor in a porous media ) 2dk dk

(4.16)

with

( single collector filtration efficiency )

γ = γD +γ I +γG +γS

(4.17)

and  kBT  γ D = 0, 9 AS   µd pdk v f   4 AH  γ I = AS   9πµd p2 v f 

1

 ( ρP − ρW ) gd p2  γ G = 0, 00338 AS    18 µv f

8

 dp    dk

1, 2

2

15

 dk   d  p

dp γ S = 2, 7    dk 

3

2

3

8

( filtration by diffusion)

(4.18)

( filtration by intercep ption)

(4.19)

( filtratio on by sedimentation)

(4.20)

0,4

( filtration by sieving)

(4.21)

with  (1 − y5 )  AS = 1 − 3 2 y + 3 2 y 5 + y6 

(

)

and

1

y = (1 − n) 3. ( Happel parameter )

(4.22)

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

145

dk = effective grain size in metres, n = porosity, γ = filtration efficiency of a single collector, α = collision efficiency, γD = diffusion, γI = interception, γG = gravitation/ sedimentation, γS = sieving, kB = Boltzmann constant in J/K, T = absolute temperature in K, vf = apparent flow velocity in m/s, dp = colloid size in m, g = gravitational acceleration in m2/s, µ = dynamic viscosity in kg/m/s, ρp = colloid density in kg/m3, ρW = fluid density in kg/m3, AS = Happel parameter and AH = Hamaker constant in J. Figure 4.18 shows the sum of the four different CFT mechanisms to the total filtration/deposition curve. A typical minimum in colloid filtration in porous media can be observed at 1 µm for colloids with a density of 1 g/cm3. In Figure 4.19 the effects of sediment grain size, flow velocity and colloid density on travel distance until 99.9% of all colloids are retained are shown. The grain size of the collector itself has a very strong effect on colloid retention in porous media. A ten-fold larger grain size leads to an increase of maximum travel distance from 40 cm to 15 m (gravel = 10 mm vs sand = 1 mm grain size diameter). An increase in colloid density from 1 to 3 g/cm3 reduces the maximum travel distance from 30 to 12 m. Note that density is only important for colloids >100 nm, below 100 nm there is no effect of a change in density from 1 to 3 g/cm3. A reduction of the fluid flow velocity from 10 to 1 m/d (representing typical values for aquifers) reduces the maximum travel distance from 200 to 30 m. Maximum colloid mobility can be expected at a size range close to 1 µm for colloids with a density of 1 g/cm3 and around 0.2 µm for colloids with a density of 3 g/cm3. It is important to notice that the CFT holds for a clean bed filtration approach, where the collision efficiency (i.e. the rate of successful attachment) is close to one (Ryan and Elimelech, 1996). Ripening or blocking, as well as surface heterogeneities of the collector (e.g. roughness or charge heterogeneity), cannot be described with these equations (Elimelech et al., 2000; Johnson et al., 1996b). Neither can filtration under non-favourable, repulsive conditions, marked by an energy barrier (Section 4.5).

Filtration Factor l [1/m]

100

10

1 diffusion interception

sedimentation

0,1 sieving

0,01 0,1

1

10

Colloid Size [µm]

Figure 4.18 Colloid filtration by diffusion, interception, sedimentation and sieving.

146

Environmental and Human Health Impacts of Nanotechnology 1000

1: dcol. = 10 mm, vf = 10 m/d, density = 1 g/cm?

travel distance 99,9 % [m]

2: dcol. = 10 mm, vf = 1 m/d, density = 1 g/cm? 100

3: dcol. = 10 mm, vf = 1 m/d, density = 3 g/cm? 4: dcol. = 1 mm, vf = 1 m/d, density = 3 g/cm?

10

1 2

1

3 4 0,1 0,001

0,01

0,1

1

10

colloid size [mm]

Figure 4.19 Effect of colloid density, flow velocity and collector grain diameter on filtration/ deposition efficiencies calculated by Equations 4.14 to 4.21. dcoll = grain size diameter of the collector, vf = apparent flow velocity of the fluid within the porous media.

Once the colloid is attached to the surface of the collector it can be remobilized by hydrodynamic drag or lift forces. These two forces are balanced by an adhesive force, which can be quantified in terms of free energy of adhesion using an extended DLVO type approach (Section 4.5). The fluid drag and lift forces on a retained colloid can be calculated as summarized by Ryan and Elimelech (1996). For small colloids the lift force, that is the force due to different pressure acting on the top and bottom of the particle, can be neglected. The adhesive force can be calculated using different scaling models, such as the Johnson–Kendall– Roberts model or the Derjaguin–Mullen–Toporov model (Derjaguin et al., 1975; Johnson et al., 1971; Kendall, 2001). Drag forces are typically significant only for particles larger than a few hundred nanometers and when deposition occurs in the primary minimum (according to DLVO theory, Section 4.5). However, for deposition of nanoparticles <100 nm in the secondary minimum, hydrodynamic drag forces play a larger role due to weaker thermodynamic interactions (Hahn and O’Melia, 2004). 4.7.2

Unsaturated Porous Media

In unsaturated porous media, in addition to the mechanism mentioned in Section 4.7.1, liquid–gas and liquid–solid–gas interfaces have to be taken into account. Colloids can also attach to these interfaces (Chen and Flury, 2005). Colloids in unsaturated media also are subject to straining in thin fluid films when fluid saturation becomes small (i.e. drying of the soil). This effect is called film straining and was described first by (Wan and Tokunaga, 1997; Wan and Wilson, 1994). There, the

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

147

liquid–gas interface can act as barrier for colloid movement if the water film in the unsaturated media is thinner than the diameter of the colloids. Colloids can also be retained at the menisci of pendular rings of the fluid–gas collector interface instead of the liquid–gas interface (Crist et al., 2004). Consequently, calculations using equations from the CFT for an unsaturated soil system should be regarded with care and can serve, if at all, only as a first rough estimation. Also, preferential flow in unsaturated soils, that is all phenomena where the fluid move along certain pathways while bypassing a fraction of the unsaturated porous media, has to be included in transport predictions. Even though a lot of progress on preferential flow has been made recently (Clothier et al., 2008), prediction of colloid transport is extremely difficult. Colloid mobilization in natural unsaturated porous media is highly dependent on transient flow conditions (Laegdsmand et al., 1999). Summarized by Lenhart and Saiers (2002), the transport of colloids in the unsaturated zone can be described as advection and dispersion, together with a sink-source term. Advection–dispersion is relatively well understood, but size exclusion (i.e. limiting the accessibility of colloids to parts of the unsaturated pore space) and especially the sink-source term (i.e. deposition and mobilisation) is less understood and an area of active research (Crist et al., 2004).

4.8

Conclusion

This chapter has reviewed colloids in natural aquatic and terrestrial systems and primarily their aggregation, sedimentation and transport behaviour. Colloidal structures and effects on pollution and biogeochemistry are an area of immense importance but perhaps less relevant to this chapter. With the development and increasing use of manufactured nanoparticles (NPs) (reviewed in Chapters 1 and 2), it is now necessary to consider how NPs and colloids interact and also treat the aquatic colloid literature as a rich source of information on potential NP fate and behaviour. Both NP properties (Chapters 2 and 3), their interaction with natural colloids and the subsequent changes in NP properties will need to be understood to understand NP transport and also (eco)toxicology (Chapters 7 and 9). In addition to the study of the physical interactions and fractionation of NPs with natural colloids, those interested in NP fate and behaviour would be well served by reviewing the now extensive literature on natural colloids which will facilitate development in this novel field.

4.9

References

Aiken, G., D. McKnight, R. Harnish and R. Wershaw (1996) Geochemistry of aquatic humic substances in the Lake Fryxell Basin, Antarctica: Biogeochem., 34, 157–88. Alasonati, E., B. Stolpe, M.-A. Banincasa et al. (2007) Asymmetrical flow field flow fractionation – Multidetection system as a tool for studying metal–alginate interactions: Environ. Chem., 3, 192–8.

148

Environmental and Human Health Impacts of Nanotechnology

Aldahan, A., Y. Haiping and G. Possnert (1999) Distribution of beryllium between solution and minerals (biotite and albite) under atmospheric conditions and variable pH: Chem. Geol., 156 (1–4), 209–29. Allard, T., N. Menguy, J. Salomon et al. (2004) Revealing forms of iron in river-borne material from major tropical rivers of the Amazon Basin (Brazil): Geochim. Cosmochem. Acta., 68 (14), 3079–94. Aluwihare, L. I., D. J. Repeta and R. F. Chen (1997) A major biopolymeric component to dissolved organic carbon in surface sea water: Nature, 387 (6629), 166–9. Amal, R., J. A. Raper and T. D. Waite (1992) Effect of fulvic acid adsorption on the aggregation kinetics and structure of hematite particles: J. Colloid Interf. Sci., 151 (1), 244–57. Arnold, T., T. Zorn, H. Zanker et al. (2001) Sorption behavior of U(VI) on phyllite: experiments and modeling: J. Cont. Hydrol., 47 (2–4), 219–31. Assemi, S., P. G. Hartley, P. J. Scales and R. Beckett (2004) Investigation of adsorbed humic substances using atomic force microscopy: Colloids Surf. A: Physicochem. Eng. Aspects, 248 (1–3), 17–23. Atkins, D. T. and B. W. Ninham (1997) Surface and structural forces measured between silica surfaces in 1,2-ethanediol: Colloid Surf. A: Physicochem. Eng. Aspects., 129–130, 23–32. Aurell, C. A. and A. O. Wistrom (2000) Coagulation of kaolinite colloids in high carbonate strength water: Colloids Surf. A-Physicochem. Eng. Aspects., 168 (3), 277–85. Baalousha, M. (2008) Aggregation and disaggregation of iron oxide nanoparticles; influence of particles concentration, pH and natural organic matter: Sci. Tot. Environ. (Submitted). Baalousha, M. and J. R. Lead (2007) Characterization of Natural Aquatic Colloids (<5 nm) by Flow-Field Flow Fractionation and Atomic Force Microscopy: Environmental Science & Technology, 41, 1111–17. Baalousha, M., M. Motelica-Heino, S. Galaup and P. Coustumer (2005) Supramolecular structure of humic acids by TEM with improved sample preparation and staining: Res. Microsc. Tech., 66, 299–306. Baalousha, M., F. Kammer, M. Motelica-Heino et al. (2006a) Size-based speciation of natural colloidal particles by flow-field flow fractionation, inductively coupled plasmamass spectroscopy, and transmission electron microscopy/X-ray energy dispersive spectroscopy: colloids-trace element interaction: Environmental Science & Technology, 40 (7), 2156–62. Baalousha, M., M. Motelica-Heino and P. Coustumer (2006b) Conformation and size of humic substances: Effects of major cation concentration and type, pH, salinity and residence time: Colloids Surf. A: Physicochem. Eng. Aspects, 272, 48–55. Baalousha, M., A. Manciulea, S. Cumberland et al. (2008) Aggregation and surface properties of iron oxide nanoparticles; influence of pH and natural organic matter: Environ. Toxicol. Chem., 27, 1875–82. Balistrieri, L. S., J. W. Murray and B. Paul (1992) The Cycling of Iron and Manganese in the Water Column of Lake Sammamish, Washington: Limnol. and Oceanogr., 37, 510–28. Bates, J. K., J. P. Bradley, A. Teetsov et al. (1992) Colloid Formation During Waste Form Reaction: Implications for Nuclear Waste Disposal: Science, 256 (5057), 649–51. Beckett, R. and N. P. Le (1990) The role or organic matter and ionic composition in determining the surface charge of suspended particles in natural waters: Colloids and Surfaces, 44, 35–49. Behrenfeld, M. J., A. J. Bale, Z. S. Kolber et al. (1996) Confirmation of iron limitation of phytoplankton photosynthesis in the equatorial Pacific Ocean: Nature, 383 (6600), 508–11. Bell, G. M., S. Levine and L. N. McCartney (1970) Approximate methods of determining the double-layer free energy of interaction between two charged colloidal spheres: Journal of Colloid and Interface Science, 33 (3), 335–59. Ben-Yaakov, S. (1999) pH Buffering of Pore Water of Recent Anoxic Marine Sediments: Limnol. Oceanogr., 18, 86–94.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

149

Benedetti, M., J. F. Ranville, M. Ponthieu and J. P. Pinheiro (2002) Field-flow fractionation characterization and binding properties of particulate and colloidal organic matter from the Rio Amazon and Rio Negro: Org. Geochem., 33 (3), 269–79. Bergstrom, L. (1997) Hamaker constants of inorganic materials: Adv. Colloid Interf. Sci., 70, 125–69. Bernacchi, F., I. Ponzanelli, M. Minunni et al. (1996) In vivo cytogenetic effects of natural humic acid: Mutagenesis, 11 (5), 467–69. Berner, R. A. and G. R. Holdren (1977) Mechanism of feldspar weathering; some observational evidence: Geology, 5 (6), 369–72. Bertrand, A. R. and K. Sor (1962) The Effects of Rainfall Intensity on Soil Structure and Migration of Colloidal Materials in Soils: Soil Sci. Soc. Amer. J., 26 (3), 297–300. Bertsch, P. M. and J. C. Seaman (1999) Characterization of complex mineral assemblages: Implications for contaminant transport and environmental remediation: Proc. Natl. Acad. Sci., 7, 3350–7. Biggs, S. (1995) Steric and Bridging Forces between Surfaces Bearing Adsorbed Polymer: An Atomic Force Microscopy Study: Langmuir, 11 (1), 156–62. Biggs, S., P. Mulvaney, C. F. Zukoski and F. Grieser (1994) Study of Anion Adsorption at the Gold-Aqueous Solution Interface by Atomic Force Microscopy: Journal of the American Chemical Society, 116 (20), 9150–7. Biggs, S., M. Habgood, G. J. Jameson and Y. D. Yan (2000) Aggregate structures formed via a bridging flocculation mechanism: Chemical Engineering Journal, 80 (1–3), 13–22. Binnig, G., C. F. Quate and C. Gerber (1986) Atomic Force Microscope: Phys. Rev. Lett., 56, 930–3. Buffle, J. (1988) Complexation reactoins in aquatic systems. An analystical approach, EllisHorwood, Chichester. Buffle, J. and G. G. Leppard (1995) Characterization of aquatic colloids and macromolecules. 1. Structure and behavior of colloidal material: Environmental Science & Technology, 29 (9), 2169–75. Buffle, J., K. J. Wilkinson, S. Stoll et al. (1998) A generalized description of aquatic colloidal interactions: the three-colloidal component approach: Environmental Science & Technology, 32 (19), 2887–99. Burgess, R. M., R. A. McKinney and W. A. Brown (1996) Enrichment of Marine Sediment Colloids with Polychlorinated Biphenyls: Trends Resulting from PCB Solubility and Chlorination: Environmental Science & Technology, 30 (8), 2556–66. Butt, H. J. (1991) Measuring electrostatic, van der Waals, and hydration forces in electrolyte solutions with an atomic force microscope: Biophysical Journal, 60 (6), 1438–44. Butt, H. J., M. Jaschke and W. Ducker (1995) Measuring surface forces in aqueous electrolyte solution with the atomic force microscope: Bioelectrochem. Bioenerget., 38 (1), 191–201. Campbell, P. G. (1995) Interactions between trace metals and aquatic organisms: A critique of the free-ion activity model, in Metal Speciation and Bioavailability in Aquatic Systems (A Tessier and DR Turner eds), John Wiley & Sons Ltd, Chichester, 45–102. Cappella, B. and G. Dietler (1999) Force-distance curves by atomic force microscopy: Surf. Sci. Rep., 34, 5–100. Carnie, S. L., D. Y. C. Chan and J. S. Gunning (1994) Electrical Double Layer Interaction between Dissimilar Spherical Colloidal Particles and between a Sphere and a Plate: The Linearized Poisson-Boltzmann Theory: Langmuir, 10 (9), 2993–3009. Chakraborti, R. K., K. H. Gardner, J. F. Atkinson and J. E. Van Benschoten (2003) Changes in fractal dimension during aggregation: Wat. Res., 37 (4), 873–83. Chan, C. S., G. De Stasio, S. A. Welch et al. (2004) Microbial Polysaccharides Template Assembly of Nanocrystal Fibers: Science, 303 (5664), 1656–8. Chang, F. R. C. and G. Sposito (1996) The Electrical Double Layer of a Disk-Shaped Clay Mineral Particle: Effects of Electrolyte Properties and Surface Charge Density: J. Colloid Interf. Sci., 178 (2), 555–64. Chapnick, S. D., W. S. Moore, and K. H. Nealson (1982) Microbially mediated manganese oxidation in a freshwater lake: Limnol. Oceangr., 27 (6), 1004–14.

150

Environmental and Human Health Impacts of Nanotechnology

Chellam, S. and M. R. Wiesner (1993) Fluid mechanics and fractal aggregates: Wat. Res., 27 (9), 1493–6. Chen, B., J. Hulston and R. Beckett (2000) The effect of surface coatings on the association of orthophosphate with natural colloids: Sci. Tot. Environ., 263 (1–3), 23–35. Chen, G. and M. Flury (2005) Retention of mineral colloids in unsaturated porous media as related to their surface properties: Colloid Surf. A: Physicochem. Eng. Aspects., 256 (2–3), 207–16. Chen, G., M. Flury, J. B. Harsh and P. C. Lichtner (2005) Colloid-Facilitated Transport of Cesium in Variably Saturated Hanford Sediments: Environmental Science & Technology, 39 (10), 3435–42. Chen, K. L., S. E. Mylon and M. Elimelech (2006) Aggregation Kinetics of Alginate-Coated Hematite Nanoparticles in Monovalent and Divalent Electrolytes: Environmental Science & Technology, 40 (5), 1516–23. Chen, S. and D. Eisma (1995) Fractal geometry of in situ flocs in the estuarine and coastal environments: Netherlands J. Sea Res., 33 (2), 173–82. Cheng, X., A. T. Kan and M. B. Tomson (2004) Naphthalene Adsorption and Desorption from Aqueous C60 Fullerene: Journal of Chemical & Engineering Data, 49 (3), 675–83. Christenson, H. K. and P. M. Claesson (2001) Direct measurements of the force between hydrophobic surfaces in water: Adv. Colloid Interf. Sci., 91 (3), 391–436. Christenson, H. K., J. Fang, B. W. Ninham and J. L. Parker (1990) Effect of Divalent Electrolyte on the Hydrophobic Attraction: Journal of Physical Chemistry, 94, 8004–6. Christensen, T. H., P. Kjeldsen, P. L. Bjerg et al. (2001) Biogeochemistry of landfill leachate plumes: Appl. Geochem., 16 (7–8), 659–718. Claesson, P. M. and H. K. Christenson (1988) Very long range attractive forces between uncharged hydrocarbon and fluorocarbon surfaces in water: Journal of Physical Chemistry, 92 (6), 1650–5. Clothier, B. E., S. R. Green and M. Deurer (2008) Preferential flow and transport in soil: progress and prognosis: Europ. J. Soil Sci., 59 (1), 2–13. Conley, D. J. and T. C. Malone (1992) Annual cycle of dissolved silicate in Chesapeake Bay: implications for the production and fate of phytoplankton biomass: Microb. Ecol. Progr. Ser., 81, 121–8. Crist, J. T., J. F. McCarthy, Y. Zevi et al. (2004) Pore-Scale Visualization of Colloid Transport and Retention in Partly Saturated Porous Media: Vad. Zone J., 3 (2), 444–50. Dade, W. B., J. D. Davis, P. D. Nichols et al. (1990) Effects of bacterial exopolymer adhesion on the entrainment of sand: Geomicrobiol. J., 8, 1–16. Dai, M., J. M. Martin and G. Cauwet (1995) The significant role of colloids in the transport and transformation of organic carbon and associated trace metals (Cd, Cu and Ni) in the Rhone delta (France): Marine Chem., 51, 159–75. Davison, W. (1993) Iron and Manganese in Lakes: Earth Sci. Rev., 34, 119–63. Davison, W. and R. De Vitre (1992) Iron particles in freshwater, in Environmental particles (J Buffle and HP Leeuwen eds), IUPAC series on analytical and physical chemistry of envionmental systems, Lewis Publishers, Boca Raton, FL, 315–55. De Paolis, F. and J. Kukkonen (1997) Binding of organic pollutants to humic and fulvic acids: Influence of pH and the structure of humic material: Chemosphere, 34 (8), 1693–704. Derjaguin, B. V. and L. D. Landau (1941) Theory of the stability of strongly charged lyophobic sols and of the adhesion of strongly charged particles in solutions of electrolytes: Acta physicochim. U. R. S. S., 14, 633–62. Derjaguin, B. V., J. J. Muller and Y. P. Toporov (1975) Effect on contact deformations on the adhesion of particles: J. Colloid Interf. Sci., 53, 314–216. Dickinson, E. and L. Eriksson (1991) Particle flocculation by adsorbing polymers: Advances in Colloid and Interface Science, 34, 1–29. Dignac, M.-F., P. Ginestet, D. Rybacki et al. (2000) Fate of wastewater organic pollution during activated sludge treatment: nature of residual organic matter: Wat. Res., 34 (17), 4185–94.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

151

Dong, D., Y. M. Nelson, L. W. Lion et al. (2000) Adsorption of Pb and Cd onto metal oxides and organic material in natural surface coatings as determined by selective extractions: new evidence for the importance of Mn and Fe oxides: Wat. Res., 34 (2), 427–36. Dong, D., Y. Li, B. Zhang et al. (2001) Selective chemical extraction and separation of Mn, Fe oxides and organic material in natural surface coatings: application to the study of trace metal adsorption mechanism in aquatic environments: Microchem. J., 69 (1), 89–94. Doucet, F., J. R. Lead and P. H. Santschi (2006) Role of colloids in mediating trace element fate and behaviour in aquatic systems, in Environmental Colloids: behaviour, structure and characterization (K. J. Wilkinson and J. R. Lead eds), John Wiley and Sons Inc., New York. Ducker, W. A., T. J. Senden and R. M. Pashley (1991) Direct measurement of colloidal forces using an atomic force microscope: Nature, 353 (6341), 239–41. Ducker, W. A., T. J. Senden and R. M. Pashley (1992) Measurement of forces in liquids using a force microscope: Langmuir, 8 (7), 1831–6. Dugdale, R. C., F. P. Wilkerson and H. J. Minas (1995) The role of a silicate pump in driving new production: Deep Sea Res. I: Oceanogr. Res. Pap., 42 (5), 697–719. Duval, J. F. L., K. J. Wilkinson, H. P. van Leeuwen and J. Buffle (2005) Humic substances are soft and permeable: evidence from their electrophoretic mobilities: Environmental Science & Technology, 39 (17), 6435–45. Elimelech, M., J. Gregory, G. Jia and R. A. Williams (1995a) Surface interaction potentials, in Particle deposition and aggregation: measurement, moddelling and simulation (eds M. Elimelech, J. Gregory, G. Jia and R. A. Williams)., 67: Butterworth-Heinmann, Woburn, MA, 3–33. Elimelech, M., J. Gregory, X. Jia and R. A. Williams (1995b), Particle deposition and aggregation: Measurement, modeling and simulation, Butterworth-Heinemann, Oxford. Elimelech, M., M. Nagai, C. H. Ko and J. N. Ryan (2000) Relative Insignificance of Mineral Grain Zeta Potential to Colloid Transport in Geochemically Heterogeneous Porous Media: Environmental Science & Technology, 34 (11), 2143–8. Eriksson, J. C., S. Ljunggren and P. M. Claesson (1989) A phenomenological theory of longrange hydrophobic attraction forces based on a square-gradient variational approach: J. Chem. Soc., Faraday Trans. 2, 85, 163–76. Fabiano, M. and A. Pusceddu (1998) Total and hydrolizable particulate organic matter (carbohydrates, proteins and lipids) at a coastal station in Terra Nova Bay (Ross Sea, Antarctica): Polar Biol., 19, 125–32. Ferretti, R., J. Zhang and J. Buffle (1997) Kinetics of hematite aggregation by polyacrylic acid: effect of polymer molecular weights: Colloids Surf. A: Physicochem. Eng. Asp., 121, 203–15. Filella, M. (2006) Colloidal properties of submicron particles in natural waters, in Environmental colloids and particles: behaviour, structure and characterization (eds K. Wilkinson and J. R. Lead), John Wiley & Sons Inc., New York, 18–93. Filella, M. (2007) Colloidal properties of submicron particles in natural waters, in Environmental colloids and particles (K. Wilkinson and J. R. Lead eds) IUPAC series of analystical and physical chemistry of environmental systems, John Wiley and Sons Inc., New York, 18–93. Filella, M. and J. Buffle (1993) Factors controlling the stability of submicron colloids in natural waters: Colloid Surf. A: Physicochem. Eng. Aspects., 73, 255–73. Fortin, D. and S. Langley (2005) Formation and occurrence of biogenic iron-rich minerals: Earth Sci. Rev., 72 (1–2), 1–19. Fredrickson, J. K., J. M. Zachara, D. W. Kennedy et al. (1998) Biogenic iron mineralization accompanying the dissimilatory reduction of hydrous ferric oxide by a groundwater bacterium: Geochim. Cosmochim. Acta., 62 (19–20), 3239–57. Galle, T., C. Gregoire, M. Wagner and R. Bierl (2005) Bioavailability of HOC depending on the colloidal state of humic substances: A case study with PCB-77 and Daphnia-magna: Chemosphere, 61 (2), 282–92. Genxiang, L., F. Ruijiang, J. Jun and W. H. Ping (2001) Strong and Weak Interactions of Colloidal Particles with High Surface Potentials: J. Colloid Interf. Sci., 241 (1), 81–8.

152

Environmental and Human Health Impacts of Nanotechnology

Gibson, C. T., I. J. Turner, C. J. Roberts and J. R. Lead (2007) Quantifying the dimensions of nanoscale organic surface layers in natural waters: Environmental Science & Technology, 41 (4), 1339–44. Grasso, D., K. Subramaniam, M. Butkus et al. (2002) A review of non-DLVO interactions in environmental colloidal systems: Reviews in Environmental Science and Biotechnology, 1, 17–38. Gschwend, P. M. and M. D. Reynolds (1987) Monodisperse ferrous phosphate colloids in an anoxic groundwater plume: J. Cont. Hydrol., 1 (3), 309–27. Gunnars, A., S. Blomqvist, P. Johansson and C. Andersson (2002) Formation of Fe(III) oxyhydroxide colloids in freshwater and brackish seawater, with incorporation of phosphate and calcium: Geochim. Cosmochim. Acta., 66 (5), 745–58. Guo, L. and P. Santschi (1997) Composition and cycling of colloids in marine environments: Rev. Geophys., 35, 17–40. Gustafsson, O. and P. M. Gschwend (1997) Aquatic Colloids: Concepts, Definitions, and Current Challenges: Limnol. Oceanogr., 42 (3), 519–28. Hackley, V. A. and M. A. Anderson (1989) Effects of short-range forces on the long-range structure of hydrous iron oxide aggregates: Langmuir, 5 (1), 191–8. Hahn, M. W. and C. R. O’Melia (2004) Deposition and Reentrainment of Brownian Particles in Porous Media under Unfavorable Chemical Conditions: Some Concepts and Applications: Environmental Science & Technology, 38 (1), 210–20. Haitzer, M., S. Höss, W. Traunspurger and C. Steinberg (1998) Effects of dissolved organic matter (DOM) on the bioconcentration of organic chemicals in aquatic organisms – a review: Chemosphere, 37 (7), 1335–62. Hamaker, H. C. (1937) The London–van der Waals attraction between spherical particles: Physica, 4 (10), 1058–72. Harrison, K. G. (2000) Role of increased marine silica input on paleo-pCO2 levels: Paleoceanogr., 15, 292–8. Hayes, M. H. B., P. McCarthy, R. L. Malcolm and R. S. Swift (1989) Humic substances II: In search of structure: John Wiley & Sons Ltd, Chichester. He, Q. H., G. G. Leppard, C. R. Paige and W. J. Snodgrass (1996) Transmission electron microscopy of a phosphate effect on the colloid structure of iron hydroxide: Wat. Res., 30 (6), 1345–52. Heathwaite, L., P. Haygarth, R. Matthews et al. (2005) Evaluating Colloidal Phosphorus Delivery to Surface Waters from Diffuse Agricultural Sources: J. Environ. Qual., 34 (1), 287–98. Hedges, J. I. et al. (2000) The molecularly-uncharacterized component of nonliving organic matter in natural environments: Org. Geochem., 31 (10), 945–58. Hedges, J. I., R. G. Keil and R. Benner (1997) What happens to terrestrial organic matter in the ocean?: Org. Geochem., 27 (5–6), 195–212. Helgeson, H. C., W. M. Murphy and P. Aagaard (1984) Thermodynamic and kinetic constraints on reaction rates among minerals and aqueous solutions. II. Rate constants, effective surface area, and the hydrolysis of feldspar: Geochim. Cosmochim. Acta., 48 (12), 2405–32. Hodge, V. A., G. T. Fan, K. R. Solomon et al. (1993) Effects of the Presence and Absence of Various Fractions of Dissolved Organic-Matter on the Toxicity of Fenvalerate to Daphniamagna: Environ. Toxicol. Chem., 12 (1), 167–76. Hofmann, Th. (2002) An overview of factors influencing the colloid/particle favoured relocation of contaminants.: IAHS-AISH Publication no.275, 165–8. Hofmann, Th. and U. Schöttler (1998) Behavior of suspended and colloidal particles during artificial groundwater recharge: Progr. Colloid. Polym. Sci., 111, 184–8. Hofmann, Th. and A. Wendelborn (2007) Colloid facilitated transport of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) to the groundwater at Ma Da Area, Vietnam: Environ. Sci. Pollut. Res., 14, 223–4. Hofmann, Th. et al. (2003) Aquatic Colloids: Definition and Relevance – a Review: Grundwasser, 8, 203–10.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

153

Holthoff, H., S. U. Egelhaaf, M. Borkovec et al. (1996) Coagulation Rate Measurements of Colloidal Particles by Simultaneous Static and Dynamic Light Scattering: Langmuir, 12 (23), 5541–9. Honeyman, B. D. and P. H. Santschi (1992) in Environmental Particles (J. Buffle and H. P. van Leeuwen eds), Lewis Publishers, Boca Raton, FL, 554. Hunter, K. A. and P. S. Liss (1982) Organic matter and surface charge of suspended particles in estuarine waters: Limnol. Oceanogr., 27, 322–35. Israelachvili, J. N. (1972) The calculation of van der Waals dispersion forces between macroscopic dodies: Proc. Royal Soc. London. A, Math. Phys. Sci., 331, 39–55. Israelachvili, J. N. and G. E. Adams (1976) Direct measurement of long range forces between two mica surfaces in aqueous KNO3 solutions: Nature, 262 (5571), 774–6. Israelachvili, J. N. and R. M. Pashley (1983) Molecular layering of water at surfaces and origin of repulsive hydration forces: Nature, 306, 249–50. Israelachvili, J. N. and R. M. Pashley (1984) Measurement of the hydrophobic interaction between two hydrophobic surfaces in aqueous electrolyte solutions: Journal of Colloid and Interface Science, 98 (2), 500–14. Israelachvili, J. and H. Wennerstrom (1996) Role of hydration and water structure in biological and colloidal interactions: Nature, 379 (6562), 219–25. IUPAC (2002) http://old.iupac.org/reports/2001/colloid_2001/manual_of_s_and_t/node8. html Iwasaki, T. (1937) Some notes on sand filtration.: J. Am. Wat. Works Assoc., 29, 1591– 602. Jarvis, P., B. Jefferson, J. Gregory and S. A. Parsons (2005) A review of floc strength and breakage: Wat. Res., 39 (14), 3121–37. Jekel, M. R. (1986) The stabilization of dispersed mineral particles by adsorption of humic substances: Wat. Res., 20 (12), 1543–54. Jenkins, P. and M. Snowden (1996) Depletion flocculation in colloidal dispersions: Adv. Colloid Interf. Sci., 68, 57–96. Johnson, C. P., X. Li and B. E. Logan (1996a) Settling velocities of fractal aggregates: Environmental Science & Technology, 30, 1911–18. Johnson, K. L., K. Kendall and A. D. Roberts (1971) Surface energy and the contact of elastic solids. Proceedings Royal Soc. London A Math. Phys. Sci., 324, 301–13. Johnson, P. R., N. Sun and M. Elimelech (1996b), Colloid Transport in Geochemically Heterogeneous Porous Media: Modeling and Measurements: Environ. Sci. Technol., 30 (11), 3284–93. Kan, A. T., G. Fu and M. B. Tomson (1994) Adsorption/Desorption Hysteresis in Organic Pollutant and Soil/Sediment Interaction: Environ. Sci. Technol., 28 (5), 859–67. Kandori, K., Uchida, S. S. Kataoka and T. Ishikawa (1992) Effects of silicate and phosphate ions on the formation of ferric oxide hydroxide particles: J. Mat. Sci., 27, 719–28. Kaplan, D. I., P. M. Bertsch, D. C. Adriano and W. P. Miller (1993) Soil-borne mobile colloids as influenced by water flow and organic carbon: Environ. Sci. Technol., (27), 1193–200. Kaplan, D. I., P. M. Bertsch and D. C. Adriano (1995) Facilitated Transport of Contaminant Metals Through an Acidified Aquifer: Ground water, 33, 708–17. Kaplan, D. I., M. E. Sumner, P. M. Bertsch and D. C. Adriano (1996) Chemical conditions conducive to the release of mobile colloids from ultisol profiles: Soil Sci. Soc. Amer. J., 60, 269–74. Karaman, M. E., R. M. Pashley, T. D. Waite et al. (1997) A comparison of the interaction forces between model alumina surfaces and their colloidal properties: Colloids and Surfaces A: Physicochemical and Engineering Aspects, 129–130, 239–55. Kendall, K. (2001) Molecular adhesion and its application: the sticky universe, Kluwer Academic/Plenum Publishers, New York. Khan, S. U. and M. Schnitzer (1972) The retention of hydrophobic organic compounds by humic acid: Geochim. Cosmochim. Acta., 36 (7), 745–54.

154

Environmental and Human Health Impacts of Nanotechnology

Kola, H. and K. J. Wilkinson (2005) Cadmium Uptake by a Green Alga Can Be Predicted by Equilibrium Modelling: Environmental Science & Technology, 39 (9), 3040–7. Koukal, B., C. Guθguen, M. Pardos and J. Dominik (2003) Influence of humic substances on the toxic effects of cadmium and zinc to the green alga Pseudokirchneriella subcapitata: Chemosphere, 53 (8), 953–61. Kretzschmar, R., M. Borkovec, D. Grolimund and M. Elimelech (1999) Mobile subsurface colloids and their role in contaminant transport: Adv. Agron., 66, 121–93. Kristensen, E. (1990) Characterization of biogenic organic matter by stepwise thermogravimetry (STG): Biogeochem., 9, 135–59. Laegdsmand, M., K. G. Villholth, M. Ullum and K. H. Jensen (1999) Processes of colloid mobilization and transport in macroporous soil monoliths: Geoderma, 93 (1–2), 33–59. Lagaly, G. and S. Ziesmer (2003) Colloid chemistry of clay minerals: the coagulation of montmorillonite dispersions: Adv. Colloid Interf. Sci., 100–102, 105–28. Lamelas, C., F. Avaltroni, M. Benedetti et al. (2005) Quantifying Pb and Cd Complexation by Alginates and the Role of Metal Binding on Macromolecular Aggregation: Biomacromolecules, 6 (5), 2756–64. Lamelas, C., M. Benedetti, K. J. Wilkinson and V. I. Slaveykova (2006) Characterization of H+ and Cd2+ binding properties of the bacterial exopolysaccharides: Chemosphere, 65 (8), 1362–70. Larson, I., C. J. Drummond, D. Y. C. Chan and F. Grieser (1993) Direct force measurements between titanium dioxide surfaces: Journal of the American Chemical Society, 115 (25), 11885–90. Lartiges, B. S., S. Deneux-Mustin, G. Villemin et al. (2001) Composition, structure and size distribution of suspended particulates from the Rhine River: Wat. Res., 35 (3), 808–16. Lead, J. R. and K. J. Wilkinson (2006a) Environmental Colloids: behaviour, structure and characterisation, IUPAC series of analytical and physical chemistry of environmental systems, John Wiley & Sons Inc, New York. Lead, J. R. and K. J. Wilkinson (2006b) Environmental colloids: current knowledge and future developments, in Environmental Colloids: behaviour, structure and characterisation (eds J. R. Lead and K. Wilkinson), John Wiley & Sons Inc., New York. Lead, J. R. and K. J. Wilkinson (2006c) Aquatic colloids and nanoparticles: current knowledge and future trends: Environ. Chem., 3, 159–71. Lead, J. R., J. Hamilton-Taylor, W. Davison and M. Harper (1999) Trace metal sorption by natural particles and coarse colloids: Geochim. Cosmochim. Acta., 63 (11–12), 1661–70. Lead, J. R., D. Muirhead and C. T. Gibson (2005) Characterisation of freshwater natural aquatic colloids by atomic force microscopy (AFM).: Environmental Science & Technology, 39, 6930–6. Lee, C. and T. A. Kramer (2004) Prediction of three-dimensional fractal dimensions using the two-dimensional properties of fractal aggregates: Adv. Colloid Interf. Sci., 112 (1–3), 49–57. Leenheer, J. A. and J. P. Croué (2003) Characterizing aquatic dissolved organic matter.: Environmental Science & Technology, 37, 18A–26A. Lenhart, J. J. and J. E. Saiers (2002) Transport of Silica Colloids through Unsaturated Porous Media: Experimental Results and Model Comparisons: Environmental Science & Technology, 36 (4), 769–77. Leppard, G. G. (1997) Colloidal organic fibrils of acid polysaccharides in surface waters: electron-optical characteristics, activities and chemical estimates of abundance: Colloids Surf. A: Physicochem. Eng. Aspects, 120 (1–3), 1–15. Leppard, G. G. (1995) The Characterization of Algal and Microbial Mucilages and Their Aggregates in Aquatic Ecosystems: Sci. Tot. Environ., 165 (1–3), 103–31. Leppard, G. G., J. Buffle and R. Baudat (1986) A Description of the Aggregation Properties of Aquatic Pedogenic Fulvic-Acids – Combining Physicochemical Data and Microscopic Observations: Wat. Res., 20 (2), 185–96.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

155

Leppard, G. G., B. K. Burnison and J. Buffle (1990) Transmission electron microscopy of the natural organic matter of surface waters: Anal. Chim. Acta., 232, 107–21. Leppard, G. G., M. M. West, D. T. Flannigan et al. (1997) A classification scheme for marine organic colloids in the Adriatic Sea: colloid speciation by transmission electron microscopy: Canad. J. Fisher. Aquat. Sci., 54 (10), 2334–49. Leppard, G. G., D. T. Flannigan, D. Mavrocordatos et al. (1998) Binding of Polycyclic Aromatic Hydrocarbons by Size Classes of Particulate in Hamilton Harbor Water: Environmental Science & Technology, 32 (22), 3633–9. Li, X. Y. and B. E. Logan (1997) Collision Frequencies of Fractal Aggregates with Small Particles by Differential Sedimentation: Environmental Science & Technology, 31 (4), 1229–36. Li, X. Y. and B. E. Logan (2001) Permeability of Fractal Aggregates: Wat. Res., 35 (14), 3373–80. Liang, L., J. F. McCarthy, L. W. Jolley et al. (1993) Iron dynamics: Transformation of Fe(II)/ Fe(III) during injection of natural organic matter in a sandy aquifer: Geochimica. Cosmochimica. Acta, 57 (9), 1987–99. Liang, Y., N. Hilal, P. Langston and V. Starov (2007) Interaction forces between colloidal particles in liquid: Theory and experiment: Adv. Colloid Interf. Sci., 134–135, 151–66. Lienemann, C.-P., M. Taillefert, D. Perret and J. F. Gaillard (1997) Association of cobalt and manganese in aquatic systems: Chemical and microscopic evidence: Geochim. Cosmochim. Acta., 61 (7), 1437–46. Lienemann, C.-P., M. Monnerat, J. Dominik and D. Perret (1999) Identification of stoichiometric iron-phosphorus colloids produced in a eutrophic lake: Aquat. Sci., 61, 133–49. Lifshitz, E. M. (1956) The theory of molecular attractive forces between solids, Sov Phys JETP USSR, 2, 73–83. Liu, R., A. Wilding, A. Hibberd and J. L. Zhou (2005) Partition of Endocrine-Disrupting Chemicals between Colloids and Dissolved Phase As Determined by Cross-Flow Ultrafiltration: Environmental Science & Technology, 39 (8), 2753–61. Loder, T. C. and P. S. Liss (1985) Control by Organic Coatings of the Surface Charge of Estuarine Suspended Particles: Limnol. Oceanogr., 30 (2), 418–21. Logan, B. E. and J. R. Hunt (1987) Advantages to Microbes of Growth in Permeable Aggregates in Marine Systems: Limnol. Oceanogr., 32, 1034–48. Lovern, S. B. and R. Klaper (2006) Daphnia-magna mortality when exposed to titanium dioxide and fellurene (C60) nanoparticles. Environ. Toxicol. Chem. 25, 1132–7. Lovley, D. R. (1997) Microbial Fe(III) reduction in subsurface environments: Fems Microbiol. Rev., 20, 305–13. Lyven, B., M. Hassellov, D. R. Turner et al. (2003) Competition between iron- and carbonbased colloidal carriers for trace metals in a freshwater assessed using flow field-flow fractionation coupled to ICPMS: Geochim. Cosmochim. Acta., 67 (20), 3791–802. Madden, A. S. and J. Hochella (2005) A test of geochemical reactivity as a function of mineral size: Manganese oxidation promoted by hematite nanoparticles: Geochim. Cosmochim. Acta., 69 (2), 389–98. Madden, A. S., J. Hochella and T. P. Luxton (2006) Insights for size-dependent reactivity of hematite nanomineral surfaces through Cu2+ sorption: Geochim. Cosmochim. Acta., 70 (16), 4095–104. Mannino, A. and H. R. Harvey (2000) Biochemical Composition of Particles and Dissolved Organic Matter along an Estuarine Gradient: Sources and Implications for DOM Reactivity: Limnol. Oceanogr., 45, 775–88. Marshall, P. A., G. I. Loeb, M. M. Cowan and M. Fletcher (1989) Response of microbial adhesives and biofilm matrix polymers to chemical treatments as determined by interference reflection microscopy and light section microscopy: Appl Environ Microbiol., 55, 2827–31. Martin, J. H. and S. E. Fitzwater (1988) Iron deficiency limits phytoplankton growth in the north-east Pacific subarctic: Nature, 331 (6154), 341–3.

156

Environmental and Human Health Impacts of Nanotechnology

Marvin, C. H., G. G. Leppard, M. M. West et al. (2004) Refined tunable methodology for characterization of contaminant-particle relationships in surface water: J. Environ. Qual., 33 (6), 2132–40. Maskaoui, K., A. Hibberd and J. L. Zhou (2007) Assessment of the Interaction between Aquatic Colloids and Pharmaceuticals Facilitated by Cross-Flow Ultrafiltration: Environmental Science & Technology, 41 (23), 8038–43. Mavrocordatos, D. and D. Fortin (2002) Quantitative characterization of biotic iron oxides by analytical electron microscopy: Am. Mineral., 87 (7), 940–6. Mavrocordatos, D., C. Mondi-Couture, O. Atteia et al. (2000) Formation of a distinct class of Fe-Ca(-Corg)-rich particles in a complex peat-karst system: J. Hydrol., 237 (3–4), 234–47. Mayer, T. D. and W. M. Jarrell (2000) Phosphorus sorption during iron(II) oxidation in the presence of dissolved silica: Wat. Res., 34 (16), 3949–56. McCarthy, J. F. (1998) Colloid-facilitated transport of contaminants in groundwater: mobilization of transuranic radionuclides from disposal trenches by natural organic matter: Phys. Chem. Earth, 23 (2), 171–8. McCarthy, J. F. and J. M. Zachara (1989) Subsurface transport of contaminants: Environmental Science & Technology, 23 (5), 496–502. McCarthy, M. D., J. I. Hedges and R. Benner (1998) Major Bacterial Contribution to Marine Dissolved Organic Nitrogen: Science, 281 (5374), 231–4. McCormack, D., S. L. Carnie and D. Y. C. Chan (1995) Calculations of Electric Double-Layer Force and Interaction Free Energy between Dissimilar Surfaces: J. Colloid Interf. Sci., 169 (1), 177–96. McDowell-Boyer, L. M., J. R. Hunt and N. Sitar (1986) Particle Transport Through Porous Media: Wat. Res. Res., 22, 1901–21. McGeer, J. C., C. Szebedinszky, D. G. McDonald and C. M. Wood (2002) The role of dissolved organic carbon in moderating the bioavailability and toxicity of Cu to rainbow trout during chronic waterborne exposure: Compar. Biochem. Physiol. C: Toxicol. Pharmacol., 133 (1–2), 147–60. Meagher, L. (1992) Direct measurement of forces between silica surfaces in aqueous CaCl2 solutions using an atomic force microscope: Journal of Colloid and Interface Science, 152 (1), 293–5. Moffett, J. W. and J. Ho (1996) Oxidation of cobalt and manganese in seawater via a common microbially catalyzed pathway: Geochim. Cosmochim. Acta., 60 (18), 3415–24. Morris, E. R., D. A. Rees, G. Robinson and G. A. Young (1980) Competitive inhibition of interchain interactions in polysaccharide systems: J. Molec. Biol., 138 (2), 363–74. Morse, J. W. and G. W. Luther (1999) Chemical influences on trace metal-sulfide interactions in anoxic sediments: Geochim. Cosmochim. Acta., 63 (19–20), 3373–8. Mosley, L. M., K. A. Hunter and W. A. Ducker (2003) Forces between colloid particles in natural waters: Environmental Science & Technology, 37, 3303–8. Mukhopadhyay, B. and J. V. Walther (2001) Acid-base chemistry of albite surfaces in aqueous solutions at standard temperature and pressure: Chem. Geol., 174 (4), 415–43. Murphy, W. M. and H. C. Helgeson (1987) Thermodynamic and kinetic constraints on reaction rates among minerals and aqueous solutions. III. Activated complexes and the pHdependence of the rates of feldspar, pyroxene, wollastonite, and olivine hydrolysis: Geochim. Cosmochim. Acta., 51 (12), 3137–53. Myklestad, S. M. (1995) Release of extracellular products by phytoplankton with special emphasis on polysaccharides: Sci. Tot. Environ., 165 (1–3), 155–64. Nagatal, T., B. Meon and D. L. Kirchman (2003) Microbial degradation of peptidoglycan in seawater: Limnol. Oceanogr., 48 (2), 745–54. Nelson, D. M., P. Tréguer, M. A. Brzezinski et al. (1995) Production and dissolution of biogenic silica in the ocean: Revised global estimates, comparison with regional data and relationship to biogenic sedimentation: Glob. Biogeochem. Cycles, 9, 359–72. Newton, P. P. and P. S. Liss (1987) Positively Charged Suspended Particles: Studies in an Iron-Rich River and its Estuary: Limnol. Oceanogr., 32 (6), 1267–76.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

157

Norton, I. T., D. M. Goodall, S. A. Frangou et al. (1984) Mechanism and dynamics of conformational ordering in xanthan polysaccharide: J. Molec. Biol., 175 (3), 371–94. O’Melia, C. (1980) ES&T Features: Aquasols: the behavior of small particles in aquatic systems: Environmental Science & Technology, 14 (9), 1052–60. Oberdorster, E., S. Zhu, T. M. Blickley et al. (2006) Ecotoxicology of carbon-based engineered nanoparticles: Effects of fullerene (C60) on aquatic organisms: Carbon, 44 (6), 1112–20. Officer, C. B. and J. H. Ryther (1980) The possible importance of silicon in marine eutrophication: Mar. Ecol. Prog. Ser., 3, 83–91. Österberg, R. and K. Mortensen (1992) Fractal dimension of humic acids: Europ. Biophys. J., 21, 163–7. Osthols, E. (1995) Thorium sorption on amorphous silica: Geochim. Cosmochim. Acta., 59 (7), 1235–49. Ouali, L. and E. Pefferkorn (1994) Fragmentation of Colloidal Aggregates Induced by Polymer Adsorption: Journal of Colloid and Interface Science, 168 (2), 315–22. Ouyang, Y., D. Shinde, R. S. Mansell and W. Harris (1996) Colloid-enhanced transport of chemicals in subsurface environments: A review: Crit. Rev. Environmental Science & Technology, 26, 189–204. Overbeek, J. T. (1982) Monodisperse colloidal systems, fascinating and useful: Advances in Colloid and Interface Science, 15 (3–4), 251–77. Paquin, P. R. et al. (2002) The biotic ligand model: a historical overview: Compar. Biochem. Physiol. C: Toxicol. Pharmacol., 133 (1–2), 3–35. Pashley, R. M. (1981) Hydration forces between mica surfaces in aqueous electrolyte solutions: J. Colloid Interf. Sci., 80 (1), 153–62. Pashley, R. M. and J. N. Israelachvili (1984a) Dlvo and hydration forces between mica surfaces in Mg2+, Ca2+, Sr2+, and Ba2+ chloride solutions: J. Colloid Interf. Sci., 97 (2), 446–55. Pashley, R. M. and J. N. Israelachvili (1984b) Molecular layering of water in thin films between mica surfaces and its relation to hydration forces: Journal of Colloid and Interface Science, 101 (2), 511–23. Pefferkorn, E. (1995) The role of polyelectrolytes in the stabilisation and destabilisation of colloids: Adv. Colloid Interface Sci., 56, 33–104. Perret, D., G. G. Leppard, M. Muller et al. (1991) Electron microscopy of aquatic colloids: Non-perturbing preparation of specimens in the field: Wat. Res., 25, 1333–43. Perret, D., J. F. Gaillard, J. Dominik and O. Atteia (2000) The Diversity of Natural Hydrous Iron Oxides: Environmental Science & Technology, 34 (17), 3540–6. Piccolo, A. (2001) The supramolecular structure of humic substances: Soil Sci., 166, 810–32. Plette, A. C. C., M. F. Benedetti and W. H. Van Riemsdijk (1996) Competitive Binding of Protons, Calcium, Cadmium, and Zinc to Isolated Cell Walls of a Gram-Positive Soil Bacterium: Environmental Science & Technology, 30 (6), 1902–10. Pronk, J. T. and D. B. Johnson (1992) Oxidation and reduction of iron by acidophilic bacteria: Geomicrobiol. J., 10, 153–71. Qi, S., G. J. M. Hartog and A. Bast (2008) Damage to lung epithelial cells and lining fluid antioxidant defense by humic acid: Environ. Toxicol. Pharmacol., 26 (1), 96–101. Ragueneau, O. et al. (2000) A review of the Si cycle in the modern ocean: recent progress and missing gaps in the application of biogenic opal as a paleoproductivity proxy: Glob. Planet. Change, 26 (4), 317–65. Rajagopalan, R. and C. Tien (1976) Trajectory analysis of deep-bed filtration with the spherein-cell porous media model: Am. Inst. Chem. Eng. J., 22, 523–33. Ramos-Tejada, M. M., A. Ontiveros, J. L. Viota and J. D. G. Duran (2003) Interfacial and rheological properties of humic acid/hematite suspensions: J. Colloid Interf. Sci., 268 (1), 85–95. Rees, D. A. (1981) Polysaccharide shapes and their interactions-some recent advances: Pure Appl. Chem.-1981-1, 43, 1–14.

158

Environmental and Human Health Impacts of Nanotechnology

Rice, J. A. and J. S. Lin (1993) Fractal nature of humic materials: Environmental Science & Technology, 27 (2), 413–14. Rice, J. A., E. Tombacz and K. Malekani (1999) Applications of light and X-ray scattering to characterize the fractal properties of soil organic matter: Geoderma, 88 (3–4), 251–64. Roy, S. B. and D. A. Dzombak (1996) Colloid release and transport processes in natural and model porous media: Colloid. Surf. A: Physicochem. Eng. Aspects., 107, 245–62. Ruckenstein, E. and N. Churaev (1991) A possible hydrodynamic origin of the forces of hydrophobic attraction: Journal of Colloid and Interface Science, 147 (2), 535–8. Ryan, J. N. and M. Elimelech (1996) Colloid mobilization and transport in groundwater: Colloid Surf. A: Physicochem. Eng. Aspects., 107, 1–56. Sader, J. E., S. L. Carnie and D. Y. C. Chan (1995) Accurate Analytic Formulas for the DoubleLayer Interaction between Spheres: J. Colloid Interf. Sci., 171 (1), 46–54. Sander, S., L. M. Mosley and K. A. Hunter (2004) Investigation of interparticle forces in natural waters: effects of adsorbed humic acids on iron oxide and alumina surface properties: Environmental Science & Technology, 38 (18), 4791–6. Santschi, P. H., J. J. Lenhart and B. D. Honeyman (1997) Heterogeneous processes affecting trace contaminant distribution in estuaries: The role of natural organic matter: Marine Chem., 58 (1–2), 99–125. Santschi, P. H., E. Balnois, K. J. Wilkinson et al. (1998) Fibrillar polysaccharides in marine macromolecular organic matter as imaged by atomic force microscopy and transmission electron microscopy: Limnol. Oceanogr., 43, 896–908. Schindler, P. W., B. Fnrst, R. Dick and P. U. Wolf (1976) Ligand properties of surface silanol groups. I. Surface complex formation with Fe3+, Cu2+, Cd2+, and Pb2+: Journal of Colloid and Interface Science, 55 (2), 469–75. Scholl, M. A., A. L. Mills, J. S. Herman and G. M. Hornberger (1990) The influence of mineralogy and solution chemistry on the attachment of bacteria to representative aquifer materials: J. Cont. Hydrol., 6 (4), 321–36. Schulthess, C. P. and C. P. Huang (1990) Adsorption of Heavy Metals by Silicon and Aluminum Oxide Surfaces on Clay Minerals: Soil Sci. Soc. Am. J., 54 (3), 679–88. Schulthess, C. P. and D. L. Sparks (1989) Competitive Ion Exchange Behavior on Oxides: Soil Sci. Soc. Am. J., 53 (2), 366–73. Senesi, N., F. R. Rizzi, P. Dellino and P. Acquafredda (1996) Fractal dimension of humic acids in aqueous suspension as a function of pH and time: Soil Sci. Soc. Amer. J., 60, 1613–78. Senesi, N., F. R. Rizzi, P. Dellino and P. Acquafredda (1997) Fractal humic acids in aqueous suspensions at various concentrations, ionic strengths, and pH values: Colloid. Surf. A: Physicochem. Eng. Aspects., 127 (1–3), 57–68. Sigg, L. (1994) Regulation of trace elements in lakes: role of sedimentation, in Chemical and biological regulation of aquatic systems (J. Buffle and R. R. DeVitre eds), Lewis Publishers, Boca Raton, FL, 175–96. Smith, D. C., M. Simon, A. L. Alldredge and F. Azam (1992) Intense hydrolytic enzyme activity on marine aggregates and implications for rapid particle dissolution: Nature, 359 (6391), 139–42. Stabel, H. H. (1986) Calcite Precipitation in Lake Constance: Chemical Equilibrium, Sedimentation, and Nucleation by Algae: Limonol. Oceanogr., 31, 1081–93. Stankovich, J. and S. L. Carnie (1996) Electrical Double Layer Interaction between Dissimilar Spherical Colloidal Particles and between a Sphere and a Plate: Nonlinear PoissonBoltzmann Theory: Langmuir, 12 (6), 1453–61. Stoll, S. and J. Buffle (1996) Computer Simulation of Bridging Flocculation Processes: The Role of Colloid to Polymer Concentration Ratio on Aggregation Kinetics: J. Colloid Interf. Sci., 180 (2), 548–63. Strycek, T., J. Acreman, A. Kerry et al. (1992) Extracellular fibril production by freshwater algae and cyanobacteria: Microb. Ecol., 23, 53–74.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

159

Stumm, W. (1992) Chemistry of the solid-water interface: processes at the mineral–water and particle–water interface in natural systems: John Wiley & Sons Inc., New York. Stumm, W. and J. J. Morgan (1996) Aquatic chemistry: chemical equilibria and rates in natural waters: John Wiley & Sons Inc., New York. Stumm, W. and B. Sulzberger (1992) The cycling of iron in natural environments: Considerations based on laboratory studies of heterogeneous redox processes: Geochim. Cosmochim. Acta., 56 (8), 3233–57. Sunda, W. G. and S. A. Huntsman (1987) Microbial oxidation of manganese in a North Carolina estuary: Limnol. Oceanogr., 32, 552–64. Swift, R. S. (1989) Molecular weight, size, shape, and charge characteristics of humic substances: some basic considerations, in Humic substances. II. In search of structure (eds M. H. B. Hayes, P. McCarthy, R. L. Malcolm and R. S. Swift), John Wiley & Sons Ltd, Chichester, 449–66. Tang, P. and J. A. Raper (2002) Modelling the settling behaviour of fractal aggregates – a review: Powder Technol., 123 (2–3), 114–25. Tang, P., J. Greenwood and J. A. Raper (2002) A Model to Describe the Settling Behavior of Fractal Aggregates: J. Colloid Interf. Sci., 247 (1), 210–19. Tani, Y., N. Miyata, K. Iwahori et al. (2003) Biogeochemistry of manganese oxide coatings on pebble surfaces in the Kikukawa River System, Shizuoka, Japan: Appl. Geochem., 18 (10), 1541–54. Teschke, O. and E. F. de Souza (2003) Measurements of long-range attractive forces between hydrophobic surfaces and atomic force microscopy tips: Chem. Phys. Lett., 375 (5–6), 540–6. Tessier, A., D. Fortin, N. Belzile et al. (1996) Metal sorption to diagenetic iron and manganese oxyhydroxides and associated organic matter: Narrowing the gap between field and laboratory measurements: Geochim. Cosmochim. Acta., 60 (3), 387–404. Thurman, E. M. and R. L. Malcolm (1981) Preparative isolation of aquatic humic substances: Environmental Science & Technology, 15 (4), 463–6. Thurman, E. M., R. L. Wershaw, R. L. Malcolm and D. J. Pinckney (1982) Molecular size of aquatic humic substances: Org. Geochem., 27–35. Tien, C. and A. C. Payatakes (1979) Advances in deep bed filtration: Am. Inst. Chem. Eng. J., 25, 737–59. Tipping, E. (2002) Cation binding by humic substances: Cambridge University Press, Cambridge. Tipping, E. and D. C. Higgins (1982) The effect of adsorbed humic substances on the colloid stability of haematite particles: Colloids Surf., 5 (2), 85–92. Tipping, E. and M. Ohnstad (1984) Colloid stability of iron oxide particles from a freshwater lake: Nature, 308 (5956), 266–8. Tipping, E., C. Woof and D. Cooke (1981) Iron oxide from a seasonally anoxic lake: Geochim. Cosmochim. Acta., 45 (9), 1411–19. Totten, L. A., P. A. Brunciak, C. L. Gigliotti et al. (2001) Dynamic Air-Water Exchange of Polychlorinated Biphenyls in the New York–New Jersey Harbor Estuary: Environmental Science & Technology, 35 (19), 3834–40. Treguer, P. and P. Pondaven (2000) Global change: Silica control of carbon dioxide: Nature, 406 (6794), 358–359. Tufenkji, N. and M. Elimelech (2004) Deviation from colloid filtration theory in the presence of repulsive electrostatic interactions: Implications to microbial transport., 228, U605– U606. Abs. Pap. Am. Chem. Soc. Tuinier, R., J. Rieger and C. G. de Kruif (2003) Depletion-induced phase separation in colloid-polymer mixtures: Adv. Colloid Interf Sci., 103 (1), 1–31. Turner, D. R. and K. A. Hunter (2001) The biogeochemistry of iron in seawater: John Wiley & Sons Ltd, Chichester. Turner, R. E., N. Qureshi, N. N. Rabalais et al. (1998) Fluctuating silicate : nitrate ratios and coastal plankton food webs: Proc. Natl. Acad. Sci. USA., 95 (22), 13048–51.

160

Environmental and Human Health Impacts of Nanotechnology

v. d. Kammer, F. (2005) Characterization of Environmental Colloids applying Field-Flow Fractionation – Multi Detection Analysis with Emphasis on Light Scattering Techniques, Hamburg University of Technology, Hamburg, Germany. v. d. Kammer, F., M. Baborowski, S. Tadjiki and W. V. Tümpling (2003) Colloidal particles in sediment pore waters: particle size distributions and associated element size distribution in anoxic and re-oxidized samples, obtained by FFF–ICP–MS coupling. Acta hydrochim. hydrobiol. 31 (4–5), 400–10. Van Ginneken, L., L. Bervoets and R. Blust (2001) Bioavailability of Cd to the common carp, Cyprinus carpio, in the presence of humic acid: Aquat. Toxicol., 52 (1), 13–27. van Leeuwen, H. P., R. M. Town, J. Buffle et al. (2005) Dynamic speciation analysis and bioavailability of metals in aquatic systems. Environmental Science & Technology, 39, 8545–56. Vandevivere, P. and D. L. Kirchman (1993) Attachment Stimulates Exopolysaccharide Synthesis by a Bacterium: Appl Environ Microbiol., 59, 3280–6. Veerapaneni, S. and M. R. Wiesner (1996) Hydrodynamics of Fractal Aggregates with Radially Varying Permeability: J. Colloid Interf. Sci., 177 (1), 45–57. Verdugo, P., A. L. Alldredge, F. Azam et al. (2004) The oceanic gel phase: a bridge in the DOM–POM continuum: Marine Chem., 92 (1–4), 67–85. Verwey, E. J. W. and J. Th. G. Overbeek (1948) Theory of the Stability of Lyophobic Colloids: Elsevier, Amsterdam, The Netherlands. Wakeham, S. G., C. Lee, J. I. Hedges et al. (1997) Molecular indicators of diagenetic status in marine organic matter: Geochim. Cosmochim. Acta., 61 (24), 5363–9. Wan, J. and T. K. Tokunaga (1997) Film Straining of Colloids in Unsaturated Porous Media: Conceptual Model and Experimental Testing: Environ. Sci. Technol., 31 (8), 2413–20. Wan, J. and J. L. Wilson (1994) Colloid transport in unsaturated porous media: Wat. Res. Res., 30, 857–64. Welch, S. A., W. W. Barker and J. F. Banfield (1999) Microbial extracellular polysaccharides and plagioclase dissolution: Geochim. Cosmochim. Acta., 63 (9), 1405–19. Wershaw, R. L. (1999) Molecular aggregation of humic substances: Soil Sci., 164, 803–13. White, H. K., L. Xu, A. L. C. Lima et al. (2005) Abundance, Composition, and Vertical Transport of PAHs in Marsh Sediments: Environ. Sci. Technol., 39 (21), 8273–80. Wigginton, N. S., K. L. Haus and M. F. Hochella (2007) Aquatic environmental nanoparticles: J. Environ. Monit., 9, 1306–16. Wilkinson, K. J., A. Joz-Rolland and J. Buffle (1997a) Different role of pedogenic fulvic acids and aquagenic biopolymers on colloid aggregation and stability in freshwaters: Limnol. Oceanogr., 42, 1714–24. Wilkinson, K. J., J.-C. Nègre and J. Buffle (1997b) Coagulation of colloidal material in surface waters: the role of natural organic matter: J. Cont. Hydrol., 26 (1–4), 229–43. Wilkinson, K. J., E. Balnois, G. G. Leppard and J. Buffle (1999) Characteristic features of the major components of freshwater colloidal organic matter revealed by transmission electron and atomic force microscopy: Colloids Surf. A: Physicochem. Eng. Aspects, 155 (2–3), 287–310. Williams, P. M. and E. R. M. Druffel (1988) Dissolved organic matter in the ocean: comments on a controversy: Oceanogr., 1, 14–17. Wolfaardt, G. M., J. R. Lawrence, J. V. Headley et al. (1994) Microbial exopolymers provide a mechanism for bioaccumulation of contaminants: Microb. Ecol., 27, 279–91. Wolthoorn, A., E. J. M. Temminghoff and W. H. van Riemsdijk (2004) Colloid formation in groundwater by subsurface aeration: characterisation of the geo-colloids and their counterparts: Appl. Geochem., 19 (9), 1391–402. Xu, S., B. Gao and J. E. Saiers (2006) Straining of colloidal particles in saturated porous media: Wat. Res. Res., 42 W12S16.1–W12S16.10. Yao, K. M. (1968) Influence of suspended particle size on the transport aspect of water filtration., University of North Carolina, Chapel Hill, Dissertation.

Natural Colloids and Nanoparticles in Aquatic and Terrestrial Environments

161

Yeung, A. K. C. and R. Pelton (1996) Micromechanics: A New Approach to Studying the Strength and Breakup of Flocs: J. Colloid Interf. Sci., 184 (2), 579–85. Zbik, M. and R. G. Horn (2003) Hydrophobic attraction may contribute to aqueous flocculation of clays: Colloid. Surf. A: Physicochem. Eng. Aspects., 222 (1–3), 323–8. Zhang, J. and J. Buffle (1996) Multi-method determination of the fractal dimension of hematite aggregates: Colloids Surf. A: Physicochem. Eng. Aspects, 107, 175–87. Zhang, W.-S. (2003) Nanoscale iron particles for environmental remediation: an overview: J. Nanopart. Res., 5, 323–32. Zhou, J. L., R. Liu, A. Wilding and A. Hibberd (2007) Sorption of Selected Endocrine Disrupting Chemicals to Different Aquatic Colloids: Environmental Science & Technology, 41 (1), 206–13. Zimmermann-Timm, H. (2002) Characteristics, dynamics and importance of aggregates in rivers – an invited review: Int. Rev. Hydrobiol., 87, 197–240.

5 Atmospheric Nanoparticles Aurélie Charron1 and Roy M. Harrison2 2

1 INRETS Transport and Environment Laboratory, Bron, France Division of Environmental Health and Risk Management, School of Geography, Earth and Environmental Sciences, University of Birmingham, United Kingdom

5.1

Introduction

Nanoparticles (NPs) are very numerous in the atmosphere, particularly in urban areas and close to combustion sources. They dominate the number concentration of particles in the atmosphere but represent a very small proportion of the atmospheric particle mass. The examination of particle number size distributions at various European sites showed that 70–80% of atmospheric particles have a size diameter below 100 nm (Van Dingenen et al., 2004). The knowledge of physical and chemical properties of atmospheric NPs is both health and climate relevant. There is increasing evidence that NPs may be deleterious to human health (Oberdörster, 2000; Harrison and Yin, 2000; Donaldson and Stone, 2003; Stone (Chapter 9), 2009). NPs may grow to several hundreds of nanometers in diameter and then act as cloud condensation nuclei or influence the global radiation balance by scattering or absorbing solar radiation (Yu et al., 2006; Penner et al., 2006). On the other hand, increasing particle number concentrations influences cloud cover and cloud reflectivity and reduces precipitation efficiency (AQEG, 2006). Sources, transformations and concentrations of atmospheric NPs are discussed in this chapter. NPs are generally defined as particles with one dimension ranging from 1 to 100 nm; even though, there is no widely used definition of atmospheric NPs. Particles smaller than 100 nm diameter are often called ultrafine particles and they are currently widely measured as particle number concentrations. Here, NPs are also called ultrafine particles or are discussed as particle number counts. Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

164

Environmental and Human Health Impacts of Nanotechnology

Atmospheric NPs may be either solid or liquid. They may consist partially or totally of semi-volatile material and may be involved in gas–to–particle equilibria during their evolution within the atmosphere. Atmospheric NPs may be perfect spheres or of irregular shape and may also be fractal-like agglomerates. Additionally, particles smaller than 100 nm may occupy different modes (nucleation mode for the smaller ones, Aitken mode and accumulation mode for the larger ones) and, as a consequence, they may have different sources or physical and chemical properties if their size is below 10 nm or close to 100 nm, as discussed throughout the text.

5.2

Sources of Atmospheric Nanoparticles

As with other air pollutants, NPs can be directly emitted in the atmosphere or can be formed within the atmosphere itself. The former ones are called primary NPs while the latter are called secondary NPs. Literature review suggests that a major source of outdoor NPs is emissions from vehicles and a second major source for ultrafine particles in outdoor air appears to be photochemically produced particles. Major primary sources of ultrafine particles in the indoor atmosphere appear to be smoking and cooking. 5.2.1

Sources of Primary Nanoparticles

Combustion processes generate large numbers of particles (Morawska and Zhang, 2002). According to AQEG (2005), 94% of emissions of particles less than 100 nm in diameter (by mass) for 1998 in the United Kingdom corresponded to combustion processes; including two-thirds from road transport as the major contributor (62%). The contribution of road traffic is much larger for ultrafine particles than for PM10. Recent work has also shown that some indoor sources generate very high levels of NPs. 5.2.1.1

Mobile Sources

Traffic is the major source of primary outdoor NPs. The contribution of traffic to particle number concentrations in urban areas is clearly underlined by large concentrations and diurnal variations of particle numbers measured near roads or in streets canyon (Harrison et al., 1999; Wehner et al., 2002) as illustrated in Figure 5.1 and discussed below. In roadside measurement studies, particle number size distributions of traffic influenced aerosols are generally bimodal with the strongest peak generally between 10 and 40 nm (Morawska et al., 1999a; Molnàr et al., 2002; Wehner et al., 2002), with a second mode often merged into the first one and around 50– 90 nm (Figure 5.2). Diesel engines are major sources of NPs; while petrol engines have also been reported to emit NPs (Kittelson, 1998). Variable particle number emission factors* are reported in the literature since they depend on a large number of parameters, such as the dilution ratio, the air *An emission factor defines the relationship between the amount of pollution produced by a source and a parameter indicating the source activity, for example for mobile sources either the vehicle distances travelled or the amount of fuel burned.

Atmospheric Nanoparticles 6000

N(11-30 nm) Light duty traffic

Integrated particle number (cm-3)

35000

5500

Heavy-duty traffic 5000

Particle surface area 30000

4500 4000

25000 3500 20000

3000 2500

15000 2000 10000

1500 1000

5000 500 Sun

Mon

Tue

Wed

Thu

0

Fri

Traffic counts: light duty, heavy-duty X5 ; particle surface area X2 (µm2/cm3)

40000

165

Sat 0

Figure 5.1 Median integrated particle numbers for particle sizes ranging from 11 to 30 nm from Scanning Mobility Particle sizer and median weekly cycles for light duty traffic (cars and motorcycles) and heavy duty traffic (lorries, cars, coaches) at Marylebone Road, London. (Reprinted from A. Charron and R.M. Harrison, Primary particle formation from vehicle emissions during exhaust dilution in the roadside atmosphere, Atmospheric Environment, 37, 4109–19. Copyright 2003, with permission from Elsevier.) 90000 80000

Night Morning rush hour Midday Afternoon rush hour

-3

dN/dlogDp (cm )

70000 60000 50000 40000 30000 20000 10000 0 10

100

1000

Dp (nm)

Figure 5.2 Median particle size distributions measured at Marylebone Road, London on weekdays from 1999–2001. (Reprinted from A. Charron and R.M. Harrison, Primary particle formation from vehicle emissions during exhaust dilution in the roadside atmosphere, Atmospheric Environment, 37, 4109–19. Copyright 2003, with permission from Elsevier.)

166

Environmental and Human Health Impacts of Nanotechnology

temperature and humidity, the fuel sulfur content, the vehicle types or proportion, and so on, as discussed below. Nevertheless, Beddows and Harrison (2008) found a good agreement between particle number emission factors derived from rolling chassis dynamometers equipped with suitable dilution apparatus and laboratory instruments and from various field studies. Values ranged from 2 to 70 × 1013 particles vehicle−1 km−1 for light duty vehicles and ranged from 20 to 730 × 1013 particles vehicle−1 km−1 for heavy duty vehicles (Beddows and Harrison (2008) and references therein). NPs from diesel vehicle exhaust emissions contain two predominant modes, a nucleation mode and a soot mode. The largest mass (but not number) of particles is comprised of particles with a core of elemental (graphitic) carbon with adsorbed material (organic compounds, metals) formed within the combustion chamber of the engine itself. The modal diameter of such particles is typically in the 30–100 nm size range with maximum number concentrations generally around 50–60 nm (Vogt et al., 2003; Kittelson et al., 2006a). On the contrary, the nucleation mode contains most of the particle number and comprises particles within a 6–30 nm size range (Kittelson et al., 2006a). These nucleation mode particles are made of hydrocarbons and/or sulfate and are formed through the condensation of semi-volatile vapours during the atmospheric dilution of the engine exhaust (Shi and Harrison, 1999; Shi et al., 2000). Nucleation mode particles are in the vapour phase in the tailpipe, become supersaturated and undergo gas–to–particle conversion during dilution and cooling. The origin of these very small particles is discussed below. Particle size and number concentrations emitted by diesel vehicle engines strongly depend on the conditions of dilution of the exhaust pipe output and on the sampling conditions of chassis dynamometer studies. The dilution ratio, the dilution rate, the dilution temperature, the relative humidity and the transfer line residence time are all noted as influential parameters because all affect the formation and growth of small particles from supersaturated vapours (Shi and Harrison, 1999; Shi et al., 2000). The formation of a nucleation mode in the exhaust of diesel engines also depends on the equipment of the vehicle (presence of oxidation catalyst, particulate filter, etc.), on the sulfur content of the fuel used and on the vehicle loading (Schneider et al., 2005; Vaaraslahti et al., 2004). Consequently, vehicle emissions are extremely difficult to measure repeatably and many research projects have led to different results because the protocol and operating procedures for the measurements were different. This sensitivity to conditions of dilution is also responsible for differences observed between laboratory chassis dynamometer studies and real-world measurements. Strong variability has also been observed within a single study. For example, Mathis et al. (2005) observed that peak emissions of volatile NPs only took place under specific conditions (e.g. during strong acceleration at high speed and during the regeneration of the diesel particle filter) and was poorly repeatable. The dilution ratio is one of the most important parameters that influences particle number concentrations from exhaust emissions (Shi and Harrison, 1999; Schneider et al., 2005). The dilution ratio describes the volume of diluted exhaust divided by the volume of raw exhaust. The mixing of vehicle exhaust in the atmosphere rapidly creates dilution ratios of the order of 1000 or more (Harrison, 2007).

Atmospheric Nanoparticles

167

1600 rpm 50% load

1.60E+09 1.40E+09

run 1 total dr=1090

1.20E+09

run2 total dr=54

1.00E+09

run3 total dr=28

8.00E+08 6.00E+08 4.00E+08 2.00E+08 0.00E+00 0.001

0.01

0.1

1

Dp (µm) Run

Dilution

1 2 3

ratio 1080 54 28

Number concentration (#/cm3) 7.3E+08 3.0E+08 2.9E+08

Volume concentration (µm3/cm3) 2.0E+04 1.9E+04 2.3E+04

Figure 5.3 Comparison of particle size distribution measured at different dilution ratios at engine speed 1600 rpm and 50% load. (Harrison, R. M., (2007), Nanoparticles in the atmosphere, Nanothechnology: consequences for human health and the Environment, Issues in Environmental Science and Technology, 24, 35–49. Reproduced by permission of the Royal Society of Chemistry. Adapted from Shi and Harrison (1999).)

Such dilution ratios in laboratory dilution tunnel experiments caused a decrease of the particle size and an increase of the particle numbers while the overall volume of particulate matter did not change significantly. This is illustrated in Figure 5.3 (from Harrison (2007) adapted from the work of Shi and Harrison (1999)). Rönkkö et al. (2007) confirmed the observations of Shi and Harrison; while the comparison of Giechaskiel et al. (2005) between laboratory (partial flow sampling system with constant sampling conditions) and on-road measurements for diesel Euro III passenger showed similar results for both sampling conditions despite the different dilution ratio, sampling temperature and residence time of the aerosol in dilute conditions. While in situ and modelling studies have shown that particle number concentrations measured at roadside sites depends on ambient temperature (Charron and Harrison, 2003; Gidhagen et al. 2004b; Janhäll et al., 2004; Olivares et al., 2007) as illustrated in Figure 5.4a, many laboratory chassis dynamometer studies have not found any influence of the sampling temperature (Mathis et al., 2005; Giechaskiel et al., 2005; Casati et al., 2007). In the real-world atmosphere, the process involved is likely to be nucleation after mixing with the ambient air, as demonstrated by Charron and Harrison (2003) and Olivares et al. (2007). Accordingly, Rönkkö et al. (2006) showed that the nucleation mode is formed within five metres of the moving vehicles and Uhrner et al. (2007) found the highest particle number counts at a

168

Environmental and Human Health Impacts of Nanotechnology 7

80

6

70 5

60

4

50

3

40 R2 = 0.73

2

30 20

N(11-30nm)/(100-450nm)

1

10

RH (%)

0

0 0

5

10

15

20

25

Ratio N(11-30nm)/N(100-450nm)

90

Relative humidity (%)

Ratio N(11-30nm)/N(100-450nm)

7

6

y = 0.0801x + 3.1296 R2 = 0.71

5 4 3 2 1 0

2

4

6

8

10

12

14

Temperature (Celsius degree)

Wind speed (knots)

(a)

(b)

16

18

20

Figure 5.4 (a) Relationship between the temperature and the ratio of integrated particle counts from 11 to 30 nm to integrated particle counts from 100 to 450 nm; also presented is the relationship between the relative humidity and the temperature. (b) Relationship between the wind speed and the ratio of integrated particle counts from 11 to 30 nm to integrated particle counts from 100 to 450 nm. (Reprinted from A. Charron and R.M. Harrison, Primary particle formation from vehicle emissions during exhaust dilution in the roadside atmosphere, Atmospheric Environment, 37, 4109–19. Copyright 2003, with permission from Elsevier.)

distance of 0.45 metres behind the tailpipe. Rönkkö et al. (2007) showed that the nucleation mode from a Euro IV heavy duty diesel is formed after 0.7 seconds residence time in the atmosphere. Ambient temperature seems to mainly affect particles smaller than 30–40 nm (Charron and Harrison, 2003; Olivares et al., 2007), which are almost entirely volatile (Kuhn et al., 2005; Biswas et al., 2007). Olivares et al. (2007) also observed a significant relationship between particles smaller than 20 nm and the relative humidity that was not observed by Charron and Harrison (2003), while Charron and Harrison (2003) observed that increases in wind speed increased the abundance of particles smaller than 30 nm in diameter relative to particles greater than 100 nm in diameter, presumably because of the greater dilution and lower condensation sink (Figure 5.4b). Du and Yu (2006) investigated the role of H2SO4–H2O binary homogeneous nucleation in the formation of NPs in the vehicular exhaust. They demonstrated that for vehicles running with fuel with high sulfur content (330 ppm), this significantly influences new particle formation, especially at low temperature and high relative humidity. They also showed that nanoparticle formation is significant even with fuel with low sulfur content (50 ppm). Another approach to modelling of particle exhaust emissions (Vouitsis et al., 2005) showed that a possible mechanism for new particle formation in diesel exhaust is the nucleation of sulfuric acid followed by the condensation of hydrocarbons on sulfuric acid–water nuclei. Many laboratory experiments have demonstrated that the nucleation mode (10– 40 nm) originates from the nucleation of sulfates over/within the oxidation catalyst

Atmospheric Nanoparticles

169

of the diesel vehicles (or over the diesel particle filter involving an oxidation catalyst) when sulfur rich fuels are used (Maricq et al., 2002; Vogt et al., 2003; Vaaraslahti et al., 2004; Giechaskiel et al., 2005; Ntziachristos et al., 2005; Uhrner et al., 2007) and preferentially during high speed operations (Maricq et al., 2002; Vogt et al., 2003; Giechaskiel et al., 2005; Ntziachristos et al., 2005; Uhrner et al., 2007). Strong acceleration (Uhrner et al., 2007), high engine rotational speed (Uhrner et al., 2007) and high load (Vaaraslahti et al., 2004) would also be parameters that influence the nucleation of sulfates. High particle number concentrations are measured when diesel vehicles are run with high sulfur fuel (Casati et al., 2007; Maricq et al., 2002; Vogt et al., 2003), but a large nucleation mode has also been observed with low sulfur fuel (sulfur content <40 ppm according to Ntziachristos et al., 2005; <7 ppm according to Uhrner et al., 2007). In the absence of an oxidation catalyst, Maricq et al. (2002) measured particle sizes in the exhaust of a light duty diesel lorry following a single log normal distribution with a number mean diameter of 70–83 nm. They concluded that the oxidation catalyst, whose role is to remove hydrocarbons and carbon monoxide, oxidizes sulfur dioxide into sulfur trioxide, which is in turn transformed into sulfuric acid in the presence of water. The nucleation of sulfuric acid as the exhaust cools in the atmosphere may be favoured by the low specific particle surface area available for condensation in the case of soot particle removal with a diesel particle filter (DPF). The formation of NPs is also found to correlate positively with the lubricant sulfur content (Vaaraslahti et al., 2005). Kittelson et al. (2006c) compared two diesel particulate matter abatement devices, the continuously regenerating diesel particle filter (which consists of a diesel oxidation catalyst followed by an uncatalysed filter) and the catalysed continuously regenerating trap (which consists of a diesel oxidation catalyst followed by a catalysed filter). The first device produced high concentrations of nuclei mode particles depending on the exhaust temperature (in agreement with findings of Vaaraslahti et al., 2004), while the second reduced number counts to levels which were not detectable. Grose et al. (2006) confirmed, based upon measurements of particle volatility and hydroscopicity, that NPs emitted by a diesel engine equipped with a catalytic trap are primarily comprised of sulfates. There is also evidence that the nucleation of hydrocarbons can be responsible for high nanoparticle concentrations in the exhaust of diesel vehicles. In the absence of a continuously regenerating diesel particulate filter, the emissions of a heavy duty diesel engine showed size distributions with a nucleation mode only at low load, independent of the fuel sulfur content (sulfur levels of 2–40 ppm) and independent of the lubricant sulfur level, and NPs were thought to be mainly comprised of hydrocarbons (Vaaraslahti et al., 2004; 2005). Rönkkö et al. (2006) measured the emissions of a heavy duty diesel vehicle equipped with an oxidation catalyst that met the Euro III emission standards. They showed that at low engine torque hydrocarbons have an important role in the nucleation processes, whilst at high torque the processes would be sulfur-driven. The most likely sources of material for such particles are engine lubricating oil partially burned during the combustion process and unburned fuel (Tobias et al., 2001). Sakurai et al. (2003) confirmed that diesel exhaust nanoparticle composition had a volatile component that comprised more

170

Environmental and Human Health Impacts of Nanotechnology

than 95% by unburned lubricating oil; Kittelson et al. (2006a) showed that nuclei mode particles mainly consisted of heavy hydrocarbons with volatility close to that of C24–C32 n-alkanes (as the fingerprint of lubricating oil). Very recently, Rönkkö et al. (2007) demonstrated that the nucleation mode particles had a non-volatile core formed before the dilution process and semi-volatile hydrocarbons condensed on it. The relative contribution of sulfates and hydrocarbons to nucleation of nanoparticle emissions and favourable conditions for particle formation are areas that are still poorly understood. Maricq (2007) notes that the nucleation of sulfates preferentially occurs in the exhaust of light duty diesel vehicles, while the nucleation of lubricant oil hydrocarbons preferentially occurs in the exhaust of heavy duty diesel vehicles. He interprets this difference from the different oil consumption rates and the presence or absence of an oxidation catalyst that controls the exhaust sulfate to hydrocarbon ratio. Another possible mechanism of formation of nucleation mode particles is related to the addition of a fuel additive, such as cerium or iron. Since the exhaust temperature of a diesel engine is low, the diesel particle filter requires catalytic assistance for its regeneration and this could be done either by the incorporation of a catalyst into the particle filter substrate or into the fuel. Recent research (Miller et al., 2007) has shown that the addition of iron (ferrocene) in diesel is responsible for the formation of 5–10 nm NPs via homogeneous nucleation of iron and subsequent agglomeration (Miller et al., 2007). The number and size of these NPs increase with the level of doping. The addition of cerium oxide NPs to the diesel fuel reduces significantly the number concentration of particles in the accumulation mode (mostly above 80 nm) and then the particle mass. However, it leads to a dramatic increase of nucleation mode particles that are mostly comprised of cerium (Skillas et al., 2000; Jung et al., 2005). Uncertainties still remain in number emission and chemical composition of NPs from diesel vehicles and in understanding the nucleation process under different ambient conditions. Also, most of the laboratory chassis dynamometer studies have been based upon small numbers of vehicles and studies have shown that volatile nanoparticle emissions were highly dependent on the vehicle (Mathis et al., 2005). A modelling approach of particle exhaust emissions (Vouitsis et al., 2005) demonstrated that the nucleation mode is significantly suppressed by the presence of the soot mode due to the high surface of the latter. As vehicle exhaust dilutes and cools in the atmosphere, semi-volatile species may either nucleate to form new particles or condense onto existing particles. This suggests that the reduction of the particle mass using devices such as diesel particle filters may lead to an increase of particle number concentrations if volatile species are not appreciably removed. Far fewer studies are available on petrol engines. Petrol engines have been reported to emit a lower total number concentration of particles than diesel engines (Kittelson, 1998) but higher concentrations of very small particles than diesel engines (Maricq et al., 1999a, 1999b). Ultrafine particles emitted by petrol engines comprise carbonaceous agglomerates ranging from 10 to 80 nm, consisting of a carbon core with various condensed compounds (Morawska and Zhang, 2002). The peak in the number size distributions were reported to be in the 30–60 nm size range (Ristovski et al., 1998; Maricq et al., 1999b) with particles also containing a

Atmospheric Nanoparticles

171

high fraction of volatile material (Kittelson, 1998). Kittelson et al. (2006b) investigated particle number emissions from seven petrol vehicles operated on-road and on a chassis dynamometer. Particle number emissions were highly dependent on driving mode. Higher emissions were observed during high speed highway cruise conditions, acceleration and at cold start temperatures. Contrary to previous findings, no typical particle signature was found. However, they showed that many of the particles were smaller than 10 nm and the relative contribution of light duty petrol vehicles to particle number emissions increased as particle size decreased. Such very small particles are rarely measured in the roadside atmosphere because of instrument limitations. Kittelson et al. (2006b) also showed that high mileage petrol vehicles emitted a greater number of nuclei mode particles compared to the lower mileage petrol vehicles. Interestingly, they demonstrated that when a per vehicle basis is used heavy duty vehicles produced much greater particle number concentrations than light duty vehicles (as indicated above), while particle number production is only slightly higher when a fuel-specific basis is used. Despite the smaller contribution of petrol engines than diesel engines, apportionment results show that weekend production of particle number in the United States was attributable to light duty petrol vehicles (Kittelson et al., 2006b). A fraction of NPs emitted by on-road vehicles corresponds to non-exhaust emissions. However, very few studies are currently available and the contribution of non-exhaust emissions to current (and future) NPs emissions requires further investigation. Brake wear particulate matter emissions were determined by Garg et al. (2000). On average, 35% of the brake pad mass loss was emitted as airborne particulate matter with on average 33% (in mass) of the airborne particulate matter smaller than 100 nm. Sanders et al. (2003) showed a larger proportion of brake wear debris being airborne (50–70%) but a much smaller contributions to ultrafine particle mass. However, their work demonstrated that semi-metallic and low metallic brake linings generated high numbers of particles smaller than 100 nm, particularly under harsh braking conditions. Significant amounts of ultrafine particles are produced at the road–tyre interface, as observed in a road simulator study (Dahl et al., 2006). The particles most likely consisted of mineral oils from the softening filler and fragments of the carbon-reinforcing filler material (soot agglomerates). The mean particle number diameters were between 15 and 50 nm and emission factors were found to be 3.7 × 1011 and 3.2 × 1012 particles vehicle−1 km−1 at speeds of 50 and 70 km h−1 respectively. This means that even though emission factors for particles originating from the road–tyre interface were 102–103 lower than those of current light-duty vehicles, the road–tyre interface might possibly be a significant contributor to nanoparticle emissions when cleaner vehicles are used. The characteristics of both exhaust and non-exhaust emissions of on-road vehicles are summarized in Table 5.1. 5.2.1.2

Indoor Sources

Indoor nanoparticle concentration levels are also influenced by vehicular traffic, as the main outdoor source in urban areas, and by some major indoor sources, particularly cooking and smoking.

172

Environmental and Human Health Impacts of Nanotechnology

Table 5.1

Summary of characteristics of on-road vehicle emissions. Modal diameters

Emission factors

References

Exhaust emissions

Diesel vehicles (1) Nucleation mode: 6–30 nm Soot mode: 50–60 nm Petrol vehicles (2) No typical signature From <10 nm to 80 nm

(1) Vogt et al., 2003; Kittelson et al., 2006a (2) Ritovski et al., 1998; Maricq et al., 1999b; Morawska and Zhang, 2002; Kittelson et al., 2006b (3) Morawska et al., 2005 and references therein (4) Beddows and Harrison, 2008 and references therein

Brake-wear emissions

Not yet known

Road-tyre interface emissions

15–50 nm

2–7 × 1014 particles veh−1 km−1(3) 2–5 × 1013 particles veh−1 km−1(3) Light-duty vehicles (mix fleet) 2–70 × 1013 particles veh−1 km−1(4) Heavy-duty vehicles (mix fleet) 20–730 × 1013 particles veh−1 km−1(4) Not yet known, depend on braking conditions 3.7 × 1011–3.2×1012 particles veh−1 km−1 at speeds of 50 and 70 km h−1

Garg et al., 2000; Sanders et al., 2003 Dahl et al., 2006

Peak concentrations of ultrafine particles in occupied buildings are generally associated with indoor activities. Cooking activities strongly enhance indoor particle number concentrations irrespective the use of gas or electrical stoves. The increase of ultrafine particle concentrations depends on many parameters, such as the activity type (frying, grilling, stove use, toasting, etc.), the way used to cook (gas or electrical stove, oven, etc.) and the cooking temperature. As a consequence, various impacts of cooking are noted in the published literature and results are highly variable. Indoor particle number concentrations could be increased by factors ranging from four times to 85 times (He et al., 2004; Hussein et al., 2005a; See and Balasubramanian, 2006). Also, particle modal diameters vary considerably from one study to another. He et al. (2004) found that particle number size distributions resulting from cooking activities were always unimodal, with the number median diameter ranging from 22 to 63 nm, while See and Balasubramanian (2006) observed smaller modal diameters ranging from 8 to 29 nm. Various indoor sources have been reported in the current literature. Candle burning, aroma oil evaporation and aroma lamp activities are other sources of indoor NPs that generate much less NPs than cooking or smoking; while burning incense sticks was found to be an intermediate source between candle burning and smoking (Hussein et al., 2006). Cleaning activities also greatly affect particle number concentration levels (Diapouli et al., 2007), in particular vacuuming (Hussein et al., 2005a). A gas clothes dryer generates more than 6 × 1012 ultrafine particles per drying episode; showing a bimodal distribution with a major peak at the smallest size measured (9.8 nm) and a secondary peak at 30 nm (Wallace, 2005). Finally,

Atmospheric Nanoparticles

173

opening the outside door in the absence of a strong outdoor source increased the indoor particle number concentrations by about 180%, while it decreased PM2.5 concentrations (He et al., 2004). In the absence of significant indoor sources or under unoccupied conditions, indoor particle number concentrations are smaller than those outdoors and the geometric mean diameters of the particle number size distributions are larger indoors than outdoors (Hussein et al., 2005a). In this case, an appreciable fraction of the indoor NPs is of outdoor origin and indoor and outdoor concentrations are well correlated showing that outdoor emissions influenced greatly the indoor concentration levels (Harrison et al., 2004; Diapouli et al., 2007); in particular emissions from motorized vehicles are significant (Zhu et al., 2005). Filtration and other penetration processes reduce particle number concentrations and reduce the number concentrations of small particles more than those of coarser particles. Also, deposition of aerosol particles onto indoor surfaces and volatilization (see below) reduce nanoparticle concentrations in indoor air (Riley et al., 2002). Various indoor-tooutdoor ratios (I/O) ranging from 0.1 to 0.8 have been found for ultrafine particles (Koponen et al., 2001; Harrison et al. 2004; Hussein et al., 2005a, 2006; Diapouli et al. 2007). Differences in building construction and in ventilation conditions or air exchange ratios contribute to the different indoor-to-outdoor relationships observed (Zhu et al., 2005; Guo et al., 2007). Different ranges of particle sizes measured may also contribute to the variability, since indoor-to-outdoor ratios strongly depend on the particle size (Koponen et al., 2001; Hussein et al., 2005a, 2006; Zhu et al., 2005). The ratio is larger for particles of 70–100 nm (from 0.6 to 0.9) than for particles of 10–20 nm (from 0.1 to 0.4). An I/O ratio as low as 0.1 was found for particles ranging from 8 to 25 nm in diameter (Koponen et al., 2001). In the presence of indoor activities or under occupied conditions, indoor particle number concentration levels cannot be directly estimated from outdoor number concentration measurements (Hoek et al., 2008). Infiltration of particles in indoor atmospheres may result in volatilization of the most volatile particles. It has been demonstrated that ammonium nitrate is transformed into ammonia and nitric acid when it penetrates into homes (Lunden et al., 2003). The equilibrium between ammonium nitrate and the gaseous components ammonia and nitric acid is shifted towards the gaseous components because of change of temperature, relative humidity and/or concentrations in indoor environments. Since available studies have shown that outdoor NPs are mainly comprised of volatile species (particularly those from vehicular emissions), this suggests that NPs of outdoor origin may decrease in size or evaporate entirely as they penetrate into indoor environments during periods when the indoor environment is warmer than the outdoors or because of different concentrations of volatile species. The resulting changes of outdoor particles will likely affect their toxicity and has implications for human exposure. People are also exposed to NPs in occupational indoor settings or during a daily commute. Ultrafine particle concentrations (<100 nm) may reach 108 particles cm−3 in photocopy centres with a peak concentration at a size of about 50 nm (Lee and Hsu, 2007). Some printers also generate NPs with various emission rates (from zero to the equivalent of smoking a cigarette) and various mean sizes (35–94 nm) (He

174

Environmental and Human Health Impacts of Nanotechnology

et al., 2007). However, concentration levels reached in offices with printers and without photocopiers were much lower than those measured in photocopy centres (on average 6.5 × 103 cm−3) (He et al., 2007). Exposure to NPs during a daily commute seems to be dominated by the exposure to motor vehicle emissions. Similar to particle number size distributions measured in street canyons, car in-cabin particle number size distributions were mostly bimodal with peaks at 10–30 nm and 50–70 nm in diameter (Zhu et al., 2007). Zhu et al. (2007) demonstrated that the vehicle’s ventilation particle filter offered an in-cabin protection of about 50% for particles in the 7–40 nm size range and of 20–30% for particles above 40 nm. Unlike larger particles, no source of ultrafine particles has been found in the London underground (Seaton et al., 2005). Thus, particle number concentrations were lower in the London underground than outside and similar concentration levels were found on platforms and in train cabs (Seaton et al., 2005). 5.2.1.3

Industrial Sources and Other Combustion Sources

Biswas and Wu (2005) reviewed research concerning stationary industrial sources responsible for nanoparticle emissions. Various stationary sources were found to emit NPs in large amounts, such as coal fired combustion systems, incinerators, coal/ oil/gas boilers, smelters and residential combustors. Similar to vehicular emissions, particle number size distributions depend on the composition of the fuel, on the dilution process (residence time and dilution ratio) and on the concentrations of larger particles that provide a large surface area for vapour condensation and coagulation with smaller particles. This behaviour is consistent with that of semivolatile species, while metals were found to be in important concentrations in particles smaller than 100 nm from stationary combustion sources (Tolocka et al., 2004a; Dillner et al., 2005). Bond et al. (2006) compared ultrafine particle emissions from an old coal-fired heating plant and from an industrial boiler that burns natural gas and residual oil. The number concentration maximum (3 × 108 cm−3) in the exhaust of the industrial boiler burning oil was about a factor of 30 greater than the highest value for natural gas. The emissions of the brown-coal heating plant had number maxima at particle diameters ranging between 50 and 100 nm, while the largest number concentrations of natural gas and oil particle emissions occur in particle diameters below 10 nm. Accordingly, very large concentrations of potassium and calcium were found in the aerodynamic size range of 56–100 nm in aerosols measured in Houston, USA, related to the emissions of a coal-fired power plant. Two combustion plants in Birmingham, UK, one burning gas and oil and the other associated with an incinerator, were found to be significant and stable sources of NPs between 3 and 7 nm (Shi et al., 2001a). Particle number size distributions between 11 and 450 nm have been measured since 1998 at the receptor site of Harwell in the rural Oxfordshire, UK. The activity of a coal-fired and gas burning power plant (coal-fired power station of 2000 MW and combined cycle gas turbine plant burning natural gas of 1400 MW) located 7 km away did not influence particle number size distributions measured at the measure-

Atmospheric Nanoparticles

175

ment station (Charron et al., 2004) except in the case of secondary particle formation (Charron et al., 2008b), while sulphur dioxide and PM2.5–10 concentrations were substantially influenced. This suggests that number emissions from the coal-fired power plant (possibly peaking around particle diameters of 50–100 nm as suggested by available studies) are much smaller than those from vehicles on nearby roads. The emissions of the gas burning plant may lead mostly to particles smaller than 11 nm that are not measured by the instrument deployed. Biomass burning is another significant source of NPs. Biomass burning includes various types of natural and anthropogenic combustion, such as forest fires, agricultural burning, logging and land clearing slash and waste burning (Wardoyo et al., 2007). Morawska et al. (1999a) made approximately 200 measurements during episodes of vegetation burning. They observed bimodal distributions with an Aitken mode in the number size distribution between 40–60 nm that was considered to be representative of this source. Wardoyo et al. (2007) investigated particle number size distributions emitted from the combustion of several grass species. Differences were observed between controlled laboratory conditions and airborne measurements. The latter showed smaller particles ranging from 30–60 nm during fast burning conditions and 60–210 nm during slow burning conditions, while the former showed particles with diameters up to 80 nm depending on the age of the smoke. 5.2.1.4

Natural Sources

The few studies available suggest that natural sources contribute little to primary NPs except in remote environments. A recent study (Clarke et al., 2006) has shown that in the remote marine atmosphere 60% of particles smaller than 100 nm in diameter are sea salt produced by bubbles from coastal oceanic breaking waves. Pollen fragments and virus exist as airborne agglomerates of sizes below 100 nm in diameter that contribute to the spread of disease or may be the cause of allergies (Biswas and Wu, 2005). 5.2.2

Secondary Sources

NPs are also formed in the atmosphere from the nucleation of vapours of low volatility. Semi-volatile gases involved in such processes may be either of biogenic or anthropogenic origin and may be either inorganic (mainly sulfuric acid) or organic compounds (mostly not yet identified). Semi-volatile gases are formed by oxidation of other gases in the atmosphere (sulfur dioxide, volatile organic compounds) and reactants may include photo-oxidants such as hydroxyl radicals. An example of particle number size distributions measured at Harwell, UK, on days with a large scale nucleation event is presented in Figure 5.5. The occurrence of nucleation events during the midday period and warmer seasons confirms that photochemistry plays a central role in these processes. Note that bursts of small particles of 50– 90 nm in diameter have also been observed during nighttime periods in urban areas and characterization by real-time single particle mass spectrometry showed that these ultrafine particles were entirely comprised of ammonium nitrate that has

176

Environmental and Human Health Impacts of Nanotechnology

40000

50000 29/5/01 11:00 29/5/01 12:00 29/5/01 13:00 29/5/01 14:00 29/5/01 15:00 29/5/01 16:00 29/5/01 17:00

25000 20000

10/9/01 9:00 10/9/01 10:00 10/9/01 11:00 10/9/01 12:00 10/9/01 13:00 10/9/01 14:00 10/9/01 15:00 10/9/01 16:00

45000 40000 -3

dN/dlogDp (cm )

30000

-3

dN/dlogDp (cm )

35000

15000

35000 30000 25000 20000 15000

10000 10000 5000

5000

0

0 10

100 Dp (nm)

(a)

1000

10

100

1000

Dp (nm)

(b)

Figure 5.5 Particle number size distributions measured at Harwell, UK; days of nucleation: (a) 29 May 2001; (b) 10 September 2001.

undergone a gas-to-particle conversion at nighttime low temperature and high relative humidity (Tolocka et al., 2004b). Bursts of photochemically-formed particles have been observed in various environments with various degrees of pollution such as the marine boundary layer and coastal sites (Allen et al., 1999; O’Dowd et al., 2002a, 2002c), in polar areas and boreal forests (Mäkelä et al., 1997); in Mediterranean forests (Kavouras and Stephanou, 2002), in continental areas (Birmili et al., 2003), in industrial plumes (Brock et al., 2002), in urban areas (Woo et al., 2001; Alam et al., 2003) and in the tropopause region (Young et al., 2007). Recently Kulmala et al. (2004) and Holmes (2007) provided comprehensive reviews of experimental observations of new particle formation and growth in the atmosphere spanning the various environments and possible mechanisms. They underline that processes responsible for new particle formation by nucleation and subsequent growth depend on the environment and are still not entirely understood. Various nucleation mechanisms have been proposed and demonstrated in the literature. The most widely studied are the binary water/sulfuric acid nucleation (Vehkamäki et al., 2002) and the ternary water/sulfuric acid/ammonia nucleation (Napari et al., 2002; Merikanto et al., 2007). Observations made in the free troposphere were consistent with the binary water/sulfuric acid nucleation theory (Kulmala et al., 2004), while it is likely that only ternary nucleation would be able to explain particle formation in the low troposphere (Kulmala et al., 2000; Korhonen et al., 2003). Ion-induced nucleation (Yu and Turco, 2000) and the condensation of semi-volatile organic compounds O’Dowd et al., 2002b; Kavouras and Stephanou, 2002) are other demonstrated mechanisms. New particle formation from semivolatile organic compounds may include biogenic emissions, such as emissions of organoiodine compounds from seaweeds at low tides (Allen et al., 1999; O’Dowd

Atmospheric Nanoparticles

177

et al., 2002a, 2002c), or emissions of terpenes in the boreal forests (Kavouras and Stephanou, 2002) and may also include reaction products of ozone and anthropogenic terpenes released from fresheners, cleaners and fragrances in the indoor environment (Weschler and Shields, 1999; Rohr et al., 2003). Very recently, Kulmala et al. (2007) showed that at a remote site in Finland, neutral nucleation dominates over ion-induced mechanisms. Also recently, studies of nucleation in the urban atmosphere of Atlanta, USA, came to the conclusion that newly-formed particles (<10 nm) consist of ammonium and sulfate alone (McMurry and Eisele, 2005; Smith et al., 2005; Sakurai et al., 2005; Jung et al., 2006). All these recent findings demonstrated the frequency of nucleation mechanisms involving sulfuric acid in various atmospheres. The formation of new particles needs the existence of thermodynamically stable clusters (Kulmala et al., 2000) that are mostly neutral clusters of about 2 nm in diameter (Kulmala et al., 2007). The work of Kulmala et al. (2007) showed that these clusters are continuously scavenged by coagulation and their ubiquitous presence in the atmosphere implies continuous nucleation. Their detection with traditional aerosol sizing instruments requires their growth to above 3 nm. Large pre-existing particle surface area decreases nucleation rates because it favours the coagulation scavenging of small nuclei and it depletes the non-volatile vapour necessary for the growth of small nuclei. As a consequence, new particle formation is rather infrequent in urban areas (Alam et al., 2003). However, because the concentrations of non-volatile vapours are proportional to the concentration ratio between their precursor gases and pre-existing aerosol particles, nucleation can take place in both clean and polluted environments (Kulmala et al., 2004). This could be illustrated by observations at two rural British sites with different conditions of both gas phase precursor concentrations and particle surface area. New particle formation occurred in similar westerly maritime air mass situations at Weybourne, a rural site located on the east coast of England (Harrison et al., 2000a), and at Harwell, a rural site located in southern England (Charron et al., 2007). Westerly maritime air masses cross polluted areas of England before reaching Weybourne, while they cross uninhabited areas of Cornwall and Wales before reaching Harwell. As a consequence, at Weybourne nucleation occurred in air masses with large particle surface area and large concentrations of precursors such as sulfur dioxide, while at Harwell nucleation occurred in clean air masses. Kulmala et al. (2004) reviewed more than 100 investigations and concluded that formation rates are often in the range 0.01–10 cm−3 s−1 in the boundary layer and that they could exceed 100 cm−3 s−1 in urban areas, in industrial plumes or in coastal areas where high concentrations of condensable species are available. They also concluded that growth rates are generally in the range 1–20 nm h−1 except in coastal areas and polar areas where growth rates could be significantly higher for the former or lower for the latter. Particle growth occurs through condensation of supersaturated vapours on the surface of small nuclei. This requires a lower degree of supersaturation than nucleation, and condensation of semi-volatile compounds may reduce the rate of particle formation if both processes involve the same species. For this reason, Holmes

178

Environmental and Human Health Impacts of Nanotechnology

(2007) suggests that particle growth occurs through the condensation of gases other than those responsible for nucleation. Because observed growth rates of nucleated particles cannot usually be explained by the condensation of sulfuric acid, it is thought that semi-volatile organic compounds play a major role in new particle growth (Kulmala et al., 2004). The identity of these organic compounds is not yet well known and likely depends on the environment. Few organic compounds have been identified such as oxidation products of monoterpenes (Boy et al., 2003). Recently, Wehner et al. (2005) showed that sulfuric acid may explain a large part of the growth of the particles until their diameter is around 10–20 nm, after that secondary organic compounds would take part in the growth process. Many investigations have demonstrated that there is a meteorological control of secondary particle formation processes. Days with bursts of new particles are days with high solar radiation and lower cloud cover. They are also days that are on average warmer and dryer than average days (Birmili et al., 2001; Nilsson et al., 2001; Stanier et al., 2004; Vehkamäki et al., 2004; Rodriguez et al., 2005; Hamed et al., 2007). However, it was recently demonstrated (Charron et al., 2007) that the average higher temperature and lower relative humidity is likely the effect of the higher solar radiation (all these meteorological parameters are strongly correlated). Charron et al. (2007) showed that days of bursts were often associated with a drop in temperature after a warmer period, following a change of air mass influence and the passage of a cold front similar to observations at a remote site in Finland (Nilsson et al., 2001), Pittsburgh in USA (Stanier et al., 2004) and a rural site in northern Italy (Rodriguez et al., 2005). Such situations in the United Kingdom lead to a drop of temperature, bright weather with broken clouds, occasional showers and strong wind speed. In agreement with observations at Hyytiälä, Finland (Nilsson et al., 2001), clean and cool arctic or polar maritime air masses have been found to be the most favourable situation for nucleation and subsequent growth of new particles at Harwell, UK (Charron et al., 2007). These air masses are the coolest air masses arriving over the United Kingdom during the warmer season. This finding is in agreement with numerical calculations that showed that the mixing of two air parcels with different properties, such as precursor concentrations, temperature and relative humidity, generates more particles through nucleation than would be the case for the two air parcels separately (Nilsson and Kulmala, 1998; Jaenisch et al., 1997; Khosrawi and Konopka, 2003). The stronger wind speed on days with nucleation, also observed by other studies (Birmili et al., 2001; Rodriguez et al., 2005; Hamed et al., 2007), and the frequent high mixed layer height (at least in westerly air masses) favours atmospheric dispersion and also may enhance nucleation rates in agreement with Easter and Peters (1994), Nilsson and Kulmala (1998) and Uhrner et al. (2003).

5.3

Chemical Composition of Atmospheric Nanoparticles

The chemical composition of atmospheric NPs depends on the nature of local sources, and NPs measured in urban and rural locations may differ considerably as already discussed in Section 5.2.

Atmospheric Nanoparticles

179

Recent research using aerosol mass spectrometry and measurements of hygroscopicity and volatility (Smith et al., 2005; Sakurai et al., 2005) showed that new particles (6–15 nm) in Atlanta, USA, were entirely made of ammonium sulfate since no other compounds were detected in these small particles. This finding is in agreement with either ternary nucleation of particles or rapid neutralization of sulphuric acid by atmospheric ammonia. Evidence of new particle formation from semi-volatile organic compounds has also been acquired in various atmospheres (Section 5.2.2). Zhang et al. (2004a) identified the species that contributed to the condensational growth of new particles larger than 33 nm in Pittsburgh, USA, using an Aerodyne aerosol mass spectrometer. Sulfate was always the first species that contributed to the growth and NPs appeared to be acid during the initial stage of nucleation. They gradually become neutralized by ammonia (10–40 minutes after sulfate). Finally, the contribution of organics often commenced around 11:00 when the photochemistry was more intense and mainly oxygenated organic compounds were involved in the growth process during the latter part of the event. This finding is in agreement with that of Wehner et al. (2005), who showed that sulfuric acid largely explains the growth of NPs until their diameter is around 10–20 nm; before the condensation of secondary organic compounds. Oxidation products of terpenes have been identified as possible secondary organics involved in the growth process. Laaksonen et al. (2008) demonstrated that during and after nucleation in Hyytiälä, Finland, secondary organic compounds, likely oxidation products of terpenes, were a major constituent of NPs, in agreement with the finding of Boy et al. (2003). It is currently known that NPs from diesel exhaust emissions are of two modes, a nucleation mode below 30 nm (sulfate and/or hydrocarbons) and a soot mode around 60 nm (elemental carbon and adsorbed organic compounds, metals). The relative contribution of sulfate and hydrocarbons in the nucleation mode depends on many parameters (Section 5.2.1.1). Meyer and Ritovski, (2007) demonstrated, using a volatilization and humidification tandem differential mobility analyser (VH-TDMA), that ternary nucleation of sulfuric acid, ammonia and water played a role in the initial formation of diesel NPs produced at high loads, while at low loads diesel NPs were hydrophobic, suggesting that they were predominantly made of hydrocarbons. Schneider et al. (2005) showed that under specific conditions (high fuel sulfur content, high engine load) sulfate represented 90% in mass of nucleation particles (above 20 nm) emitted by a light duty diesel vehicle. On the other hand, much recent research has shown that lubricating oil hydrocarbons contribute greatly to the organic fraction of NPs emitted by heavy duty diesel vehicles. Tobias et al. (2001) and Sakurai et al. (2003) showed that branched alkanes and alkyl substituted cycloalkanes from lubricating oil and unburned fuel represented more than 95% in mass of particles ranging from 25 to 60 nm, while the sulfate was less than 5% in mass. A few studies have focused on nanoparticle emissions from petrol vehicles. Sodeman et al. (2005) examined the composition of individual particles ranging from 50 to 300 nm (aerodynamic diameters) emitted by 28 light duty petrol vehicles using an Ultrafine Aerosol Time-of-Flight Mass Spectrometer (UF-ATOFMS). Particles below 100 nm contained a large amount of elemental carbon and the

180

Environmental and Human Health Impacts of Nanotechnology

dominant class was that with elemental carbon internally mixed with calcium, phosphate, sulfate and a lower abundance of organic carbon. The relative abundance of organic carbon increased as the particle size increased (particles larger than 50 nm aerodynamic diameter) and calcium and phosphate are likely to be from lubricating oil additives. By number, the elemental carbon (combined with calcium, 2− PO3− 4 , SO4 , OC) dominated the emissions of the lowest emitting vehicles; while the highest emitting vehicles produced the highest number of particles, with the dominant class being organic compounds including substituted monoaromatic compounds and PAHs coupled with calcium and PO3− 4 . Considering both the work of Sodeman et al. and that soot particles would be progressively removed by diesel particle filters, it can be expected that the elemental carbon will soon no longer be a tracer of diesel emissions. Available research shows that the composition of NPs at urban locations is strongly influenced by vehicular emissions. Fushimi et al. (2008) sampled NPs (30–100 nm) with two low pressure impactors (Dekati) at a roadside site and at a background site in Kawasaki, Japan (analyses using a thermal/optical carbon analyser). The total carbon represented a significant fraction of particles smaller than 100 nm and was more than 80% of the mass of 30–60 nm (aerodynamic diameter) particles. The smaller the particles (down to 30 nm), sampled closer to the vehicular emissions, the larger the contribution of total carbon to the nanoparticle mass. At the roadside site, the elemental carbon represented more than 70% of the total carbon in the size fraction 30–60 nm and about 50% at the background site. At both sites, the contribution of the elemental carbon was larger for particles ranging from 60 to 100 nm (about 80% of total carbon at the roadside site and 63% of total carbon at the background site), in agreement with the contribution of soot mode particles from diesel emissions. Note that a larger relative contribution of the organic carbon is expected for particles smaller than 30 nm (not analysed at Kawasaki). Fushimi et al. (2008) compared mass chromatograms of diesel exhaust and roadside and urban background particles and concluded that the organic composition of roadside and urban background 30–60 nm particles was dominated by lubricating oils (hopanes: five-ring C17–C35 hydrocarbons) and slightly affected by unburned fuel (C18–C26 n-alkanes). Even though they make a much smaller contribution to the nanoparticle mass (less than 1%), a number of metals have been found in NPs in concentrations that may be relevant to health. Lin et al. (2005) made nanoparticle measurements with a nano-MOUDI impactor beside a heavily-trafficked road (10–56 nm and <100 nm, analyses by inductively coupled plasma-mass spectrometry). Respectively, 37%, 50%, 28%, 30%, 24%, 64%, 38% and 22% of the mass of silver, cadmiun, chromium, nickel, lead, antimony, vanadium and zinc were present in the NP range. Particles ranging from 10 to 56 nm contained more of the traffic-related metals (lead, cadmium, copper, zinc, barium and nickel) than particles of other sizes. Some of them, considered as toxic (silver, cadmium and antimony), were found in much larger proportions in particles ranging from 10 to 56 nm than in larger particles. Around 95% of the total metal mass of NPs was thought to be from vehicular emissions (exhaust and non-exhaust).

Atmospheric Nanoparticles

5.4

181

Fate and Behaviour of Atmospheric Nanoparticles

As with other particles, nanoparticle growth occurs through coagulation, condensation of low volatility components and surface reactions. However, because their dimensions are similar to those of gas molecules, the smallest ones show unusual physical properties (Kelvin effect, properties of the free molecular regime, Brownian motion for coagulation, etc.) that need to be taken into account (Lehtinen and Kulmala, 2003). On the other hand, nanoparticle shrinkage occurs through evaporation (because of their volatile nature). Condensation and evaporation lead to a change in the particle diameter and particle mass but not to the overall particle number concentration. On the other hand, coagulation and fragmentation lead to a change in the particle number concentration and change in the particle diameter but the particle mass is conserved. Parallel to these transformations, NPs may also be lost from the atmosphere by dry and wet deposition. Deposition processes lead to a reduction in particle number concentration and a shift in the size distribution to larger diameters, since both processes are more efficient for small particles than for coarser particles. All these processes may occur during the aging and transport of fresh emissions and generally lead to a shift in the particle size distribution to larger diameter, even though shifts to smaller diameters are also observed at urban scales. The fundamental theories of these processes are not the purpose of this chapter and the reader is referred elsewhere for details (Seinfeld and Pandis, 1998). Condensation corresponds to the surface deposition of gases of low volatility and this partitioning between gas and particle phases to reach the equilibrium concentrations at atmospheric conditions is reversible in the absence of surface reaction. This means that when the vapour phase concentrations are low, NPs can also shrink by evaporation. The formation of involatile oligomers from more volatile organic monomers in the aerosol phase after condensation was demonstrated by Vesterinen et al. (2007) and such reactions help in understanding observed growth rates in the atmosphere. The examination of condensation and evaporation processes for NPs requires consideration of the Kelvin effect (the saturation pressure goes to infinity as the particle radius goes to zero). Thus, saturation pressure increases as the particle size decreases and this means that the smaller the particles the higher the supersaturation ratios required for condensation. Similarly, the smaller the particles the faster the evaporation due to the Kelvin effect. Consequently, for the smallest NPs (below 10 nm in diameter), the Kelvin effect prevents condensation of any semi-volatile compounds in sufficient abundance for producing substantial growth, and heterogeneous reactions are thought to play a major role in the rapid growth of such small particles and to facilitate further condensation of organic vapours (Zhang and Wexler, 2002; Anttila and Kerminen, 2003). Coagulation is the process in which two particles collide and stick, which is slow except at high number concentrations. In a polydisperse aerosol, coagulation is more effective between particles of different sizes, since large particles provide a large surface for absorption and small particles have a greater diffusion rate. For accumulation mode particles, every collision would result in coalescence or

182

Environmental and Human Health Impacts of Nanotechnology

aggregation; however, that may not be the case for very small particles because of the possible competing effects of van der Waals attraction and Born repulsion (Slama Lighty et al., 2000). To explain the evolution of freshly emitted small particles near roadways, one of the most important processes is likely to be dilution with cleaner air. Thus, for example, particles from traffic mix upwards with less polluted air, leading to a reduction in the number concentrations and, generally, a shift of the size distribution towards larger size diameters simply because the dilution air contains coarser particles. Dilution of volatile gases at increased distances from the freeway may be responsible for the shrinking of the smaller particles by evaporation to reach new equilibrium, since the vapour phase concentrations are lower (Zhang et al. 2004b; Zhang and Wexler 2004). Different studies close to major highways or with models have drawn different conclusions concerning the predominant process. However, as a common trend it seems that condensation/evaporation and dilution are processes that have a more significant effect upon the particle size distribution than coagulation and deposition (Gidhagen et al., 2004a, 2004b; Zhang et al., 2004b; Zang and Wexler 2004), except in particular conditions such as low wind speed and in the absence of turbulent mixing when coagulation and deposition could contribute to, respectively, 10% and 50% of particles episodic losses (Gidhagen et al., 2005). Also, Kerminen et al. (2007) showed that coagulation may be significant in the high concentrations of the morning rush hour. Because condensation, evaporation, coagulation and dilution alone fail to explain the evolution pattern of nanoparticle size distributions near roadways, more complex scenarios involving the shrinking or fragmentation of NPs have been proposed. Jacobson et al. (2005) suggest that small (<15 nm) liquid NPs shed semi-volatile organics almost immediately upon emission. This shrinking enhances the coagulation rates of particles by over an order of magnitude. Measurements in the urban atmosphere of Kawasaki, Japan, supported this hypothesis (Fushimi et al., 2008).

5.5 Atmospheric Concentrations 5.5.1

Spatial Variations

Because of their nature (mostly volatile) and their small size, ultrafine particles have a shorter lifetime than coarser particles. As a consequence, particle number concentrations generally exhibit a much stronger variability than particle mass concentrations (Van Dingenen et al., 2004; Weijers et al., 2004). Van Dingenen et al. (2004) observed a gradual decrease in particle number concentrations of particles ranging from 10 to 30 nm moving from kerbside, to urban site, to rural site, to background site in Europe. When they compared the annual average particle number concentrations of these sites to their annual average PM2.5 concentrations, they observed a factor of 10 between the minimum concentrations and the maximum concentrations for particle number and a factor of three for PM2.5. Observations tend to show that the further the measurement site from vehicular emissions, the lower the concentrations.

Atmospheric Nanoparticles

183

The same conclusion could be drawn from other studies. As an example, Harrison and Jones (2005) reported number concentrations of particles larger than 7 nm at a number of British locations where the same instruments were deployed. Strong gradients were found between the industrial site of Port Talbot (not directly influenced by traffic emissions), the urban centres of Glasgow, London, Birmingham, Manchester and Belfast and the street canyon of Marylebone Road, London, which shows by far the highest particle number concentrations (Figure 5.6). Particle number concentrations at the background sites were typically around 20 000 cm−3, while they reach 117 000 cm−3 at Marylebone Road (Harrison and Jones, 2005). Note that for rough comparison (another instrument is used), average particle number concentrations of about 900 cm−3 were measured at the mountainous (free tropospheric) Jungfraujoch site, Switzerland (Van Dingenen et al., 2004). Weijers et al. (2004) observed a relationship between particle number concentrations and density of population in The Netherlands, a result that again highlights the strong contribution of traffic sources. Strong spatial variations of ultrafine particle number concentrations are also observed at the urban scale. Number concentrations showed strong peaks in the near vicinity of roads and highways with intensity depending on traffic characteristics (Weijers et al., 2004). Particle number concentrations exponentially decrease as the distance from the roadway increases (Zhu et al., 2002) to reach their upwind background concentration at distances that depend on the site characteristics (often

Figure 5.6 Seasonal influence on number concentration at various British sites. (Reprinted with permission from R. M. Harrison and A. M. Jones, Multisite study of particle number concentrations in urban air, Environmental Science and Technology, 39, 6063–70. Copyright 2005, American Chemical Society.)

184

Environmental and Human Health Impacts of Nanotechnology

from 50 to 300 m). Very recently, large studies of spatial variation of particle numbers concluded that particle number concentrations at various urban locations could not be reliably estimated from central site measurements (Puustinen et al., 2007; Hoek et al., 2008). Measurements of vertical profiles of ultrafine particles on urban roads has demonstrated the significant roles played by the microscale topography and the wind direction (parallel or perpendicular to the street) in the dispersion of vehicle emissions (Figure 5.7). Higher concentrations were found within the building envelope (Morawska et al., 1999b) than at rooftops, where particle counts were decreased by a factor five compared to street level concentrations (Väkevä et al., 1999) to reach levels close to urban background concentrations. The decrease of particle number concentrations with height is larger when the wind flow is parallel to the street than in case of a vortex within the street canyon; the latter situation favours the upwards advection of vehicle emissions (Longley et al., 2004). Detailed examinations of vertical profiles showed that particle number concentrations first increased to reach maximum concentrations around 2–7 m above the ground (Zhu and Hinds, 2005; Kumar et al., 2008) before decreasing to rooftop levels (Longley et al., 2004; Zhu and Hinds, 2005; Kumar et al., 2008). Dry deposition and diffusion (controlled by turbulence) are thought to be important processes in explaining vertical profiles within street canyons, unlike horizontal profiles which are largely explained by dilution by the wind speed (Zhu and Hinds, 2005; Kumar et al., 2008). Unlike primary NPs that are emitted from a source (mobile or stationary) and subsequently diluted (and transformed) in the atmosphere, secondary NPs are formed within the atmosphere and the spatial scale for nucleation will depend on local and regional atmospheric characteristics. Nucleation events may occur simultaneously over large geographical areas of hundreds of kilometres if favourable conditions for nucleation are homogeneously distributed over large distances (gaseous precursor concentrations, meteorological concentrations, etc.). The existence of these so-called regional or large scale nucleation events was demonstrated WIND

BUILDING

Figure 5.7

VORTEX

BUILDING

Schematic representation of a vortex situation in street canyon.

Atmospheric Nanoparticles

185

by simultaneous measurements of the same nucleation event at different sites, such as Melpitz/Leipzig in Germany (Birmili, 1999) and Värriö/Hyytiälä/Tahkuse in Finland and Estonia (Vana et al., 2004). During such events the growth of nucleated particles as observed at a measurement site continues throughout the day, regardless of the wind direction. On the other hand, some nucleation events occur in a relatively small domain in the atmosphere (a few kilometres) and are called small scale nucleation events. Different types have been reported. Particles may be photochemically produced in a previously dispersed small plume of a few kilometres in size, such as a sulfur dioxide rich plume from a power station (Brock et al., 2002, 2003), an urban plume (Kerminen and Wexler, 1996; Brock et al., 2003) or coastal plume of biogenic iodine emissions (Allen et al., 1999; O’Dowd et al., 2002a). Small scale nucleation events could also occur in a sunny region of the atmosphere surrounded by clouds (Charron et al., 2007). 5.5.2 Temporal Variations The diurnal profile of particle number concentrations at kerbside and urban centres reflects well that typical of traffic generated pollutants (Figure 5.1). Weekdays and weekends show different diurnal patterns with lower number concentrations at weekends (Morawska et al., 2002). Weekdays are characterized by a strong morning rush-hour peak that is followed by a decrease in concentrations until a second, much smaller, peak corresponding to the evening rush hour. Concentrations then decline slowly throughout the night to a minimum at around 4 a.m. Saturdays and Sundays do not exhibit the same diurnal pattern, even though the highest particle numbers are often observed on Friday and Saturday nights (Charron and Harrison, 2003; Hussein et al., 2005b). At Marylebone Road, London, a sharp peak that coincides with heavy duty (diesel) traffic is observed on Saturday mornings (Figure 5.1). On weekday mornings and weekend nights, high emissions from heavy traffic flows are enhanced by poor dispersion (stable nocturnal layer), low temperatures and possibly by a low pre-existing particles surface area. In unoccupied conditions, indoor concentrations correlate with the outdoor concentrations but at a lower concentration. Both indoor and outdoor concentrations were lowest during nighttime hours and showed a clear morning peak that coincides with morning rush hour and also smaller peaks at lunch time and in the evening (Hoek et al., 2008). However, in occupied conditions, the diurnal pattern of indoor nanoparticle concentrations is different from that of outdoor concentrations and is more closely related to human indoor activities (Morawska et al., 2003). At certain urban sites and/or during the warmer seasons, another peak of total particle number may be observed at midday due to photochemically produced particles (Woo et al., 2001). At remote/background sites such as Hyytiälä, Finland, where new particle formation is frequent (Mäkelä et al., 2000), a diurnal pattern is often observed with the highest concentrations at midday (around noon) and the lowest concentrations during the night. This diurnal pattern of particle numbers is associated with a diurnal pattern in particle size with the continuous and smooth growth of newly-formed particles during the afternoon. At the rural site of Harwell, UK, where new particle formation and local motorized traffic explain well the

186

Environmental and Human Health Impacts of Nanotechnology

overall observations of particle number size distributions (Charron et al., 2008a), the diurnal profile of the smallest particles (<30 nm) is different to that of larger particles (>30 nm) (Figure 5.8). The integrated particle counts between 30–100 and 100–450 nm have similar daily profiles with lower median numbers during daytimes and higher median numbers during nighttimes, as a result of stronger dispersion in the boundary layer during daytime (and the weak contribution of local emissions), while the integrated particle counts between 11–30 nm show higher concentrations during daytime (afternoon) and are anticorrelated with the daily variations of the modal diameter of particle number size distributions (Figure 5.8). At background sites and/or at sites where new particle formation is frequently observed, a seasonal pattern is generally observed, with higher frequency and intensity of new particle formation during the warmer season, while the precise seasonal variation differs at each individual site (Kulmala et al., 2004). However, it is often observed that spring and, to a lesser extent, autumn are more favourable seasons for nucleation events than summer. The spring maximum could be partly explained by the seasonal variation of solar radiation, with a peak in June and higher levels from May to July, and/or the possible role of higher biological activity that produces precursor gases during spring (Vehkamäki et al., 2004). Nilsson et al. (2001) have associated the seasonal cycle of nucleation events at Hyytiälä, Finland (spring and autumn maxima and a summer minimum), with the higher frequencies of arctic air masses and the larger latitudinal temperature difference responsible for higher cyclone activity at that time of the year. The findings of Charron et al. (2007) for Harwell, UK, are in agreement with the hypothesis of Nilsson et al. (2001) The annual distribution of nucleation events at Harwell during the 1999– 2001 period is represented in Figure 5.9. On the other hand, at urban British sites dominated by traffic emissions, a seasonal pattern with higher concentrations in the winter and lower concentrations in the summer is observed (Harrison and Jones, 2005) (Figure 5.6). This pattern

1200

3000

1200

1000

2500

1000

800

2000

600

1500

50

200

1000

500

40 35

800

30 25

600

20 N(11-30) modal diameter

400

15

Particle diameter (nm)

400

N(11-30), N(100-450) (cm-3)

N(11-30) N(100-450) N(30-100)

N(30-100) (cm-3)

N(11-30), N(100-450) (cm-3)

45

10

200

5 0

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

0 Time (h)

Time (h)

(a)

(b)

Figure 5.8 (a) Median daily cycles of integrated particle counts for the ranges 11–30 nm, 30–100 nm and 100–450 nm. (b) Median daily cycles of the 11–30 nm integrated particle counts and of the smaller mode of particle number size distribution. (Charron et al., 2008b)

Atmospheric Nanoparticles

Max N[11-30] (cm-3)

25000

187

SSNE LSNE

20000 15000 10000 5000 0 0

1

2

3

4

5

6

7

8

9 10 11 12

Month

Figure 5.9 Monthly maximum number concentration of particles ranging from 11 to 30 nm measured on days of new particle formation (1999–2001 dataset). SSNE: small scale nucleation events; LSNE: large scale nucleation events. (Adapted from Charron et al., 2008a)

reflects the poorer dispersion during the winter or the more efficient atmospheric mixing during the hotter months of the year (Harrison and Jones, 2005). The favourable nanoparticle formation from vehicle exhaust at low temperature (Charron and Harrison, 2003) and the shift of low volatility species to their gaseous phase during the warmer period (Kuhn et al., 2005) may also explain this urban seasonal pattern.

5.6

Measurement Methods for Atmospheric Nanoparticles

NPs present significant measurement challenges, although for some descriptors, such as number concentration, good quality methods have been available for many years. Firstly, the point needs to be made that, depending upon the reason for the measurements, NPs may be measured in terms of their number, surface area or mass (Harrison et al., 2000b). Largely because effective methods have been available for number measurement, this has been far the most commonly used means of characterisation, with surface area and mass commonly estimated by transformation of the number size distribution data. There are, however, methods available capable of directly measuring either surface area or mass. Since losses by diffusion or volatilization could occur when sampling NPs, sampling protocols, including the quantification of losses, should be well defined. In particular, the lengths and configurations of sample inlets and sampling tubes are important parameters. Recent experimental data showed that penetration efficiencies dramatically decreased for NPs smaller than 20 nm and losses increased when lengths of sampling tubes increased (Kumar et al., 2008). Theoretical calculations are given elsewhere (Hinds, 1999). A summary of methods prepared in relation to occupational hygiene exposure measurements appears in Table 5.2. A full section on measurement methods for airborne NPs is also available in Chapter 8.

188

Environmental and Human Health Impacts of Nanotechnology

Table 5.2 Instruments and techniques for monitoring nano-aerosol number, surface area and mass concentrations (adapted from Mark, 2007). Metric

Devices

Remarks

Number directly

CPC

CPCs provide real-time number concentration measurements between their particle diameter detection limits. Without a nanoparticle preseparator, they are not specific to the nanometre size range. P-Trak has diffusion screen to limit top size to 1 µm. Real-time size-selective (mobility diameter) detection of number concentration, giving number-based size distribution. Can select any upper cut point size. Off-line analysis of electron microscope samples can provide information on size-specific aerosol number concentration. The UFP consists of a Corona charger, a DMA and an electrometer. It provides real-time number concentrations within six size classes 20–30, 30–50, 50–70, 70–100, 100–200, >200 nm. Real-time size-selective (aerodynamic diameter) detection of active surface area concentration, giving aerosol size distribution. Data may be interpreted in terms of number concentration. Size-selected samples may be further analysed off-line. Real-time measurement of aerosol active surface area. Active surface area does not scale directly with geometric surface area above 100 nm. Note that not all commercially available diffusion chargers have a response that scales with particle active surface area below 100 nm. Diffusion chargers are only specific to nanoparticles if used with an appropriate inlet pre-separator. Real-time measurements of active surface area. Not specific to nanoparticles without an inlet pre-separator. Off-line analysis of electron microscope samples can provide information on particle surface area with respect to size. TEM analysis provides direct information on the projected area of collected particles, which may be related to geometric area for some particle shapes. Real-time size-selective (mobility diameter) detection of number concentration. Data may be interpreted in terms of aerosol surface area under certain circumstances. For instance, the mobility diameter of open agglomerates has been shown to correlate well with projected surface area. Differences in measured aerodynamic and mobility can be used to infer particle fractal dimension, which can be further used to estimate surface area.

SMPS

Electron microscopy UFP 330

Number by calculation

ELPI

Surface area directly

Diffusion charger

Epiphaniometer Surface area by calculation

Electron microscopy

SMPS

SMPS and ELPI used in parallel

Atmospheric Nanoparticles Table 5.2

189

(continued)

Metric

Devices

Remarks

Mass directly

Size selective static sampler

The only devices offering a cut-point around 100 nm are cascade impactors (Berner-type low-pressure impactors or Micro-orifice impactors). Allows gravimetric and chemical analysis of samples on stages below 100 nm. Sensitive real-time monitors such as the Tapered Element Oscillating Microbalance (TEOM) may be useable to measure nanoaerosol mass concentrations on-line, with a suitable size selective inlet. Real-time size-selective (aerodynamic diameter) detection of active surface area concentration giving aerosol size distribution. Mass concentration of aerosols can be calculated, only if particle charge and density are assumed or known. Size-selective samples may be further analysed off-line (as above). Real-time size-selective (mobility diameter) detection of number concentration, giving aerosol size distribution. Mass concentration of aerosols can be calculated, only if particle shape and density are known or assumed.

TEOM

Mass by calculation

ELPI

SMPS

.

5.6.1

Particle Number Concentration

Condensation particle counters are widely used to measure the total number count of particles in the atmosphere. The instruments operate by drawing particles through a zone within the instrument which is saturated with respect to a condensable vapour (Figure 5.10). Subsequent cooling of the flow leads to vapour condensation on the particles and consequent growth to a size where they can be measured by optical techniques. For many years, most condensation particle counters used butanol as the condensable vapour; more recently a number of instruments based upon water vapour have been developed. The lower size limit of such instruments is typically around three nanometres (although some have lower cut points larger than this), which is due to the limitations imposed by the Kelvin effect upon condensation on highly curved surfaces. The particles, which have grown by condensation, are measured optically either by counting pulses of scattered laser light as the particles exit the condensation chamber through an optical chamber or, alternatively, at higher concentrations in a photometric mode in which light obscuration by the particle stream is measured. As noted above, the technique is sensitive down to very small particle sizes and has no upper limit other than that imposed by the entry of particles into the sensing zone of the instrument. Consequently, when applied to the atmosphere, it typically measures particles between a few nanometres diameter and perhaps

190

Environmental and Human Health Impacts of Nanotechnology

Figure 5.10

Schematic of a condensation particle counter (CPC).

five micrometres diameter. However, measurements of particle size distributions indicate that typically around 80% of airborne particles by number in urban air are within the size range below 100 nanometres diameter and therefore the measurement of the condensation particle counter is an excellent reflection of the ultrafine particle number concentration. There are a number of aerosol size spectrometer instruments based upon particle separation followed by detection by the condensation particle counter. These generally go under the name of Differential Mobility Particle Sizer or Scanning Mobility Particle Sizer depending upon the precise mode of instrument operation. In the Scanning Mobility Particle Sizer, the condensation particle counter is preceded by a bipolar charger and differential mobility analyser (Figure 5.11). The bipolar charger typically comprises a radioactive source such as 85Kr, which leads to both negative and positive charging of aerosol particles. The particles are then passed into the differential mobility analyser, which involves them moving downwards within an annulus between a central rod and an earthed outer tube. The central rod and outer tube are held at different electrical voltages, creating a potential gradient within the annulus. The charged particles migrate horizontally within the electric field towards the central rod and by adjusting the electric field strength the instrument can be tuned to allow only one size of particles to exit through a slot in the base of the instrument. By ramping the voltage, a spectrum of particle sizes is obtained. Particles exiting through the slot are counted with a condensation particle counter and the count rate used to infer a number concentration for that size of particle in the inlet airstream, after allowing for the efficiency of particle charging and transmission through the differential mobility analyser. Smaller particles acquire only a single charge (or no charge at all) in the bipolar charger but larger particles can also acquire multiple charges. Doubly charged particles behave in the same way as singly charged particles of half the size and, therefore, provided the relative efficiencies of single and multiple charging are known, a correction can

Atmospheric Nanoparticles

Figure 5.11

191

Schematic of a Scanning Mobility Particle Sizer (SMPS).

be applied for multiply charged particles and is incorporated in the instrument software. In a recent variant on the Scanning Mobility Particle Sizer referred to as the SMPS+E, the condensation particle counter is replaced with a Faraday cup electrometer that measures the accumulated electric charge upon particles which enter it. Such a device is not subject to the lower size limitation imposed by the condensation particle counter and, therefore, has potential attractions in relation to measurement of nanometre size particles. The precise measurement range of a Scanning Mobility Particle Sizer is determined by the flow rates used within it but typically such instruments contain an impactor designed to remove larger particles and are limited to sizing particles of no greater than 1 µm diameter, although often they are operated in a size range of only up to around 500 nm diameter. Although rarely used in practice, off-line electron microscopy of sampled particles is capable of generating size-fractionated number concentrations. A major difficulty, however, lies with collecting suitable samples for microscopy. Filtration is unlikely to be very suitable as most aerosol filters are depth filters and NPs can get lost between the fibres. Nuclepore filters have a flat surface but are of limited efficiency for very small particles. Sampling devices based upon electrical charging and deposition by thermophoretic precipitation directly onto electron microscopy

192

Environmental and Human Health Impacts of Nanotechnology

grids are available (Maynard 1995), but size-specific knowledge of collection efficiencies is essential if absolute number concentrations are to be estimated. The electrical low pressure impactor (ELPI) provides another rather indirect means of estimating particle number concentrations (Figure 5.12). Air entering the ELPI first encounters a unipolar corona charger which imparts electrical charges upon the particles. The particles then enter a cascade impactor in which they are separated according to their aerodynamic size (see below). The smaller size fractions lie within the ultrafine particle size range. The impacted particles are collected on electrometers and the rate of charge deposition on those electrometers is related to the particle number concentration within that size range. Careful calibration of charging efficiencies as a function of particle size together with correction for diffusional deposition of ultrafine particles on the early (coarse particle) stages of the impactor allows an estimation of particle number size distribution. There are also instruments which are a kind of hybrid between the SMPS and ELPI. Within such instruments, particles are initially charged in a unipolar corona charger as with the ELPI. They then move into a differential mobility spectrometer within which they migrate in an electric field orthogonal to the flow rate of gas. However, instead of this instrument being tuned to sample different sizes of particles sequentially by ramping the voltage, the potential gradient remains constant but the walls of the differential mobility analyser are lined with electrometers,

Figure 5.12 (ELPI).

Diagram of the operating principle of an Electrical Low Pressure Impactor

Atmospheric Nanoparticles

193

which measure the charge associated with particles depositing at different points within the differential mobility analyser. This charge is converted to a current which can then be related to the number of particles depositing per unit time. Because such instruments are capable of measuring different size ranges simultaneously, as there is no change in the voltage with time, they are capable of very fast response measurements of airborne particle number size distributions. Although developed primarily for engine exhaust particle measurement applications, such instruments are applicable to the measurement of atmospheric particle size distributions. In particular, the UFP 330 instrument has been developed for routine monitoring stations since it is easy to use and needs little maintenance. The six size classes (20–30, 30–50, 50–70, 70–100, 100–200, >200 nm) measured by the UFP 330 and data from classical mobility spectrometers (SMPS/DMPS) have been compared. In general, a good agreement was found for particles ranging from 70 to 200 nm and the largest disagreements were found for particles below 30 nm and for those above 200 nm (Zschoppe et al., 2007; Hillemann et al., 2007; Wehner et al., 2007). 5.6.2

Surface Area

Although not intuitively obvious, there is more than one way of expressing the surface area of the particles within an aerosol. One means of expression is simply the total geometric surface of the particles. The meaning of surface area expressed in such a way is straightforward if particles have totally smooth surfaces but becomes far more complex when the surfaces are rough. As an analogy, consider measuring the coastline of the British Isles. If this were measured off a low resolution map, then it would appear much shorter than if it were measured using a much finer scale from much higher resolution maps. In other words, coarse resolution maps obscure a lot of the detail which can nonetheless be measured if the spatial resolution is increased. The same is likely to be true of airborne particles with uneven surfaces. Consequently, even when expressing a simple geometric surface area, there is always an underlying complexity unless the particles have entirely smooth surfaces. The other way of expressing surface area derives from its measurement by attaching species to the surface whose concentration can be determined. The rate of attachment of absorbing or condensing species is a simple function of particle surface area only for very small particles where the dominant process is molecular bombardment of their surface (Hinds, 1999). Larger particles, however, maintain a rather static boundary layer of gas molecules above their surface and access of adsorbing or condensing molecules depends upon diffusion through that laminar boundary layer prior to attachment to the particle surface. This diffusion process is not a simple function of particle surface area and, consequently, reflecting this change of physical processes, the mathematical expressions describing gas adsorption of condensation on particle surfaces is strongly particle size dependent. However, the situation is simpler for NPs as they lie within the regime described by molecular bombardment processes and consequently surface area, be it expressed geometrically or as inferred from molecular attachment processes, should be the same.

194

Environmental and Human Health Impacts of Nanotechnology

Probably the most frequently used method of estimating particle surface area is indirectly from the number size distribution. If it is assumed that all particles have simple spherical geometry, then given the number of particles in each size bin, it is a simple matter to calculate their surface area, which can then be summed across any specified range of particle sizes. Typically, measurements from the SMPS are treated in this way and it is perfectly feasibly to estimate a geometric surface area from SMPS data. If the geometry of the particle structures differs substantial from spherical, and particularly if the particles are comprised of fractal clusters, then simple calculations of surface area based upon assumed spherical geometry could be very considerably in error. The diffusion charger functions by transmitting the air sample through a corona charger containing a high concentration of unipolar air ions. The charged ions diffuse to the particle surface and attach at a rate determined for smaller particles by gas kinetic considerations, and for larger particles by the rate of diffusion of the ions through the laminar boundary layer at the particle surface. The resultant charged particles are collected on a filter, where the collected charge is quantified. In a similar device called an epiphaniometer (Baltensperger et al., 1988), the active (or Fuchs) surface area of the particles is measured by attaching radioactive lead atoms to the particle surface. The particles are subsequently collected on a filter which is assayed continuously for its alpha radioactivity. Lead atoms which do not attach the particle surface are not counted. Both the diffusion charge and epiphaniometer have the disadvantage of measuring active rather than geometric surface area except for very small particles, whilst for ultrafine particles this difficulty does not apply. It is, however, necessary to pre-separate larger particles from the gas stream before they enter the charging zone of the instrument. Making a size cut in the nanoparticle range can be achieved by impaction, but this becomes instrumentally quite complex because of the large pressure drops required. Measurements made with the epiphaniometer can relate well to those calculated from particle size distribution data (Shi et al., 2001b). 5.6.3

Mass Concentration

There are a number of cascade impactors which provide separation of particles within the nanoparticle range. Most notably amongst these are the Micro-Orifice Uniform Deposit Impactor (MOUDI) and the Berner-type low-pressure impactor. In addition to the standard MOUDI, there is nano-MOUDI with cut-points down to 10 nanometres. These devices separate particles on the basis of their inertial properties by accelerating particles through a fine jet below which is a flat plate on which particles of higher inertia deposit, whilst those of lower inertia are able to follow the gas streamlines and avoid impaction. By using progressively higher velocities on sequential stages, progressively smaller particles deposit on the impaction plates, from which they can be removed or weighed for chemical analysis. Low pressure impactors, as well as depending on very high velocities for particle impaction, use the fact that the Cunningham slip correction of the particle is reduced at lower pressure, causing particles to impact more easily than at atmospheric pressure (Hinds, 1999). The disadvantage of impactors for work with NPs is that for work in

Atmospheric Nanoparticles

195

ambient air it may be necessary to sample for several days in order to collect sufficient material for weighing and chemical analysis, whilst instruments such as the Scanning Mobility Particle Sizer, on the other hand, are able to scan an entire size distribution in a matter of around one minute. There are automated instruments available to determine the mass of airborne particles. One commonly used instrument is the Tapered Element Oscillating Microbalance (TEOM), within which particles are collected on a filter attached to the tip of a tapered element glass tube that forms part of an oscillation microbalance (Figure 5.13). The vibration frequency of the microbalance changes according to the mass of particles deposited on the filter, allowing a continuous readout of ambient concentrations with a time resolution generally of minutes. In order to determine the mass of a specific size fraction, the TEOM requires a size selective inlet, and whilst these are available for PM10, PM2.5 and PM1.0, they are not currently available within the nanoparticle range. There is also some evidence that whilst the majority of filters are very efficient for the collection of NPs, the filter used in the TEOM may allow some penetration which would lead to under-reading of the mass concentration (Wake, 2006).

Figure 5.13

Principle of operation of aTapered Element Oscillating Microbalance (TEOM).

196

Environmental and Human Health Impacts of Nanotechnology

5.6.4

Chemical Composition

As for total particulate matter, the measurement of nanoparticle chemical composition may be conducted using a sampling device such as an impactor for the collection of NPs, with subsequent extraction and chemical analysis in the laboratory, or may be achieved using new on-line techniques such as aerosol mass spectrometry. However, at the moment of writing, few aerosol mass spectrometers are able to measure the chemical composition of NPs, although latest results show promise. 5.6.4.1

Methods for Bulk Analysis

Size-segregated particles can be collected using impactors such as the MOUDI, the Berner or the ELPI (see description above) that first classify particles according to their aerodynamic diameters on a series of stages with decreasing cut size. The lowest cut-point ranges from 30 to 50 nm depending on the impactor type and particles smaller than the lowest cut-point are collected downstream on a back-up filter. This means that the last stages of impactors generally permit the collection of NPs. Impaction substrates are used in order to avoid particle bounce during collection. After collection, samples are transferred to a laboratory for solvent extraction and analysis using various analytical techniques, such as high performance liquid chromatography (HPLC), gas chromatography (GC) or inductively coupled plasma mass spectrometry (ICPMS) depending on the nature of compounds measured. The impaction substrate for collection and the solvent used for extraction also depend on the compounds measured. Non-destructive analytical methods (i.e. no solvent extraction) such as X-ray techniques may also be used. These methods require long sampling periods to acquire enough matter to be quantitatively measured and can be the subject of sampling artefacts, such as particle bounce, volatilization of semi-volatile compounds and adsorption of gases during particle collection, losses and contamination during transport and storage, and so on. Excellent comprehensive reviews on particulate measurement are available (Chow, 1995; McMurry, 2000) and the reader may refer to them for more details. 5.6.4.2 Aerosol Mass Spectrometry Overview of aerosol mass spectrometry The two past decades have seen the emergence of aerosol mass spectrometry instruments for on-line chemical analysis of aerosols, but further efforts are still required for the quantitative chemical analysis of NPs, especially for the smallest ones (below 50 nm). Aerosol mass spectrometers generally comprise an inlet to introduce the particles into vacuum and to concentrate them from the gas phase, a possible particle sizing unit, a vaporization unit and a mass spectrometry unit for the ionization of vaporized neutral molecules and the mass analysis. The principle of a mass spectrometer is to separate and count ions according to their mass to charge ratios (m/z). Mass spectrometry allows the measurement of a large range of inorganic and

Atmospheric Nanoparticles

197

organic species within aerosols without sample pre-treatment, with fast time resolution and high sensitivity, leading to reduced artefacts. A large number of techniques and principles have been implemented in the development of aerosol mass spectrometers and the basic concepts of the operation of these instruments are often classified according to the method by which particles are volatilized and the resulting gas phase compounds ionized. At the moment of writing this chapter, there are two main classes of instruments. The first class of instruments uses one or two lasers to both vaporize and ionize individual atmospheric particles into the mass spectrometer source region. This class of instruments provides analysis at the single particle level and hence gives information on the mixing state of particles. A few reviews are available (Noble and Prather, 2000; Sullivan and Prather, 2005). The second class of instruments uses thermal vaporization of individual or collected particles followed by various ionization techniques. The most commons ionization techniques are electron impact ionization and chemical ionization. The separation of the vaporization and ionization steps permits the quantitative detection of the aerosol chemical composition for an ensemble of particles. Basic information on the two classes of instruments is summarized in Table 5.3. Laser-based methods Instruments that use high-powered pulsed lasers to volatilize and ionize the compounds to detect individual particles, hit efficiently the particle in-flight, have a high duty factor for particle detection and allow the influence of the chemical composition on the efficiency of particle analysis to be easily assessed. However, particles smaller than 200–300 nm cannot be detected by most laser-based methods. Aerodynamic lenses are used to enhance the sampling efficiencies for NPs down to about 20 nm since they reduce the divergence of the particle beam. However, they are not very effective below this size because of beam broadening from Brownian motion. Also, the hit rates (defined as the ratio between the number of particles hit by the laser that generates detectable ions) are generally low for aerosols smaller than 50 nm. The Rapid Single-particle Mass Spectrometer (RSMS) is a laser ablation timeof-flight mass spectrometer. The most recent generation of the RSMS is described by Lake et al. (2003) and is called RSMS-III. An aerodynamic lens is used to focus the particles and a freely firing laser beam (193 nm) is used to irradiate the particle beam in order to generate positive and negative ions, which are analysed with dual drift tubes. Ions exiting the drift tubes are accelerated and detected with multichannel plate detectors. Particles between 45 and 1250 nm are measured in nine size bins determined by nine flow-limiting orifices. The optimum particle size range of the RSMS-III was found to be 50–770 nm in diameter (Lake et al., 2003). Recent modifications of the Aerosol Time-of-Flight Mass Spectrometer (ATOFMS) have increased its detection efficiency for ultrafine particles. The instrument is called the Ultrafine Aerosol Time-of-Flight Mass Spectrometer (UFATOFMS) (Su et al., 2004). An aerodynamic lens system replaces the converging nozzle inlet used on conventional ATOFMS instruments. In addition, the light scattering region has been modified to enhance the scattering signals for smaller

198

Environmental and Human Health Impacts of Nanotechnology

Table 5.3 Summary of the characteristics of the main two classes of aerosol mass spectrometer.

Principle

Main characteristics Species measured

Main advantages or disadvantages

Examples of Aerosol Mass Spectrometers used for the measurement of nanoparticles

Laser-based Methods

Thermal-based Methods

Use of one or two lasers to both vaporize and ionize individual atmospheric particles Single particle analysis, mixing states

Use of thermal vaporization of individual or collected particles followed by various ionization techniques One aspect of the chemical composition of a single particle, or, chemical composition of an ensemble of particles Suitable for semi-volatile aerosol components (NH4NO3, (NH4)2SO4, organic compounds) Not suitable for elemental carbon, sea salt and crustal material

Analysis of inorganic, organic compounds (organic carbon, sulfates, nitrates, sea salt, dust, metals etc.) Aliphatic organic compounds not detected Qualitative data, not quantitative Incomplete vaporisation (more sensitive to species on surface than in the core) Matrix effects (interaction between individual chemical components during the combined desorption and ionisation processes) UF-ATOFMS (Ultrafine Aerosol Time-of-Flight Mass Spectrometer): Su et al., 2004 RSMS (Rapid Single-Particle Mass Spectrometry): Lake et al., 2003

Quantitative detection of chemical composition Non-refractory species (see above) not measured Size of particles is not measured except if a DMA is used in series, or if the Aerodyne AMS is used in the Particle time-of-flight mode TDPBMS (Thermal Desorption Particle Beam Mass Spectrometer): Tobias et al., 2000 Aerodyne AMS (Aerodyne Aerosol Mass Spectrometer): Jayne et al., 2000; Drewnick et al., 2005; DeCarlo et al., 2006 TDCIMS (Thermal Desorption Chemical Ionization Mass Spectrometer): Voisin et al., 2003

particles. This allows single particles between 50 and 300 nm to be sized and detected. However, the ‘hit’ rate is higher than 80% for particles larger than 120 nm but dramatically drops to 55% for 95 nm particles (Su et al., 2004). Additionally, the particle detection efficiency was found to be 0.3% for 95 nm polystyrene latex spheres (Su et al., 2004). Thermal-based methods Particles are focused onto a heated element that continuously vaporizes the volatile fraction of the impacting particles that are subsequently ionized by electron ionization (ARI AMS; TDPBMS) or by chemical ionization (TDCIMS).

Atmospheric Nanoparticles

199

The initial version of the Aerodyne Research Inc. Aerosol Mass Spectrometer (ARI AMS) was designed to measure the real-time non-refractory chemical species of particles with aerodynamic diameters between 50 and 1000 nm as a function of particle size (Jayne et al., 2000). A review of measurements with the ARI AMS is available (Canagaratna et al., 2007). The AMS consists of an aerosol inlet, the particle sizing chamber and the particle composition detection section. The aerosol inlet samples sub-micron aerosol particles into the AMS through an aerodynamic lens. The particle beam is transmitted into the detection chamber where non-refractory components are vaporized upon impact onto a hot surface under a high vacuum. They are chemically analysed via electron impact ionization and mass spectrometry. The mass spectrometric detector is a quadrupole (Q) mass spectrometer for the Q-AMS (Jayne et al., 2000), while it is a ion time-of-flight (ToF) mass spectrometer for the ToF-AMS (Drewnick et al., 2005) and a high resolution ToF mass spectrometer for the HR-ToF-AMS (DeCarlo et al., 2006). Most of the information provided by the Q-AMS is representative of the ensemble averaged aerosol. The Q-AMS provides some single particle information but limited to one m/z per particle. On the contrary, the ToF-AMS produces complete mass spectra for single particles. The ToF-AMS alternates between two modes of operation, the particle time-of-flight mode (PToF) and the mass spectrum mode (MS). In the PToF mode the quadrupole MS is set to scan pre-selected fragment and measure their mass as a function of the particle size. In the MS mode the averaged chemical composition of the non-refractory aerosol components is determined by scanning the full mass spectrum. The mass collection efficiency is almost 100% for spherical particles with aerodynamic diameters between 60 and 600 nm. In a Thermal Desorption Particle Beam Mass Spectrometer coupled with a DMA (TDPBMS) (Tobias et al., 2000), particles are sampled into a differentiallypumped vacuum chamber, they are focused into a narrow, low-divergence particle beam using aerodynamic lenses, and then transported into a high vacuum region where they impact and evaporate onto a heated surface. The vapour is mass analyzed by a quadrupole mass spectrometer. Two modes of operation are available. Particles are either continuously vaporized for real-time analysis by resistively heating, or they are collected for temperature-programmed thermal desorption (TPTD) analysis by cooling the foil to −50 °C using an external liquid nitrogen bath. Desorbing molecules are ionized by 70 eV electrons and analyzed using a quadrupole mass spectrometer with a pulse counting detector. For TPTD analysis, particles were collected on the cold vaporizer and desorbed by heating as high as 400 °C using a linear ramp of 90 °C/min while scanning the mass spectrometer. The TDPBMS can analyse multi-component organic particles in the 20–500 nm size range and can be coupled to a Nano-DMA for particle size selection (Tobias et al., 2001). The Thermal Desorption Chemical Ionization Mass Spectrometer (TDCIMS; Voisin et al., 2003) is dedicated to measure semi-continuously the chemical composition of classified aerosol in the 4–25 nm range only. The instrument consists of an electrostatic precipitator for collecting charged particles, an evaporation–ionization chamber and a triple quadrupole mass spectrometer equipped with a collision-

200

Environmental and Human Health Impacts of Nanotechnology

induced dissociation chamber. A nano-DMA is used downstream for particle size selection. Aerosols are charged and are then collected by electrostic deposition onto a metal filament. Then the filament is slid from the collection position into the evaporation–ionization chamber where it is resistively heated to evaporate the collected aerosol. The desorbed molecules are ionized at atmospheric pressure by proton transfer with protonated water clusters or oxygen anions formed by reactions between α particles (emitted by a 231Am source) and the buffer gas mixture. Ions are then transferred to a triple quadrupole mass spectrometer for mass analysis. 5.6.4.3

Nanoparticle Properties Related to Chemical Composition

Recently, tandem differential mobility analysers (TDMA) have been developed in order to measure some nanoparticle properties that are related to their chemical nature, such as the volatility (volatility tandem differential mobility analyser or VTDMA) or their hygroscopicity (hygroscopicity tandem differential mobility analyser or HTDMA), or both for the volatility and humidification tandem differential mobility analysers (VH-TDMA). Basically, these instruments use two DMAs (or two SMPSs) in series. The first one selects a certain aerosol size. This monodisperse aerosol is then either heated (in the case of the VTDMA) or humidified (in the case of the HTDMA) before entering the second DMA that is scanned in order to measure the entire size distribution of processed aerosols. The concentration of the monodisperse aerosol is monitored using a condensation particle counter (CPC). These instruments give useful information on nanoparticle properties that may be climate and/or health-relevant. Information on aerosol mixing properties (Biswas et al., 2007; Philippin et al., 2004) and some chemical composition insights can be inferred by comparison with evaporation curves of known compounds (Sakurai et al., 2003; Meyer and Ristovski, 2007). Ultrafine Organic Tandem Differential Mobility Analysers (UFO-TDMA) also exist. These instruments measure the ethanol uptake of particles of a certain size (selected by a first DMA). These particles are introduced into air mixed with ethanol and their size changed is measured using a second DMA. The growth factor of particles depends on their chemical nature and size. Comparison with current knowledge on ethanol uptake permits inferences on the composition of atmospheric NPs (Joutsensaari et al., 2001).

5.7

Conclusions

The literature clearly indicates that road traffic emissions dominate NPs emissions, at least in the urban atmosphere. However, because of their small size, the current knowledge of atmospheric NPs is primarily of number concentrations, and far less is known of nanoparticle chemistry. This leads to difficulty to apportion the contributions of different sources. From the current knowledge, almost nothing is known on the contribution of engineered NPs to atmospheric NPs. Also, emissions from motor vehicles are not totally understood and further studies are clearly needed

Atmospheric Nanoparticles

201

for the reconciliation of the variability encountered between measurements from various laboratories for both exhaust and non-exhaust nanoparticle emissions. A better knowledge of atmospheric NPs will also help the understanding of their health effects and their atmospheric fate. Because of very small size and the fact that the smallest mainly comprise semivolatile material, atmospheric NPs behave differently to larger atmospheric particles. They have a shorter lifetime and measurements show strong spatial and temporal variability of their concentrations. Their formation in the atmosphere may depend on meteorological parameters and this leads to difficulty in quantifying emission sources. They may be involved partially or totally in gas–particle equilibria depending on atmospheric characteristics (concentration, temperature, etc.) and this may have implications for human exposure.

5.8

References

AQEG, Air Quality Expert Group (2005) Particulate matter, DEFRA, London. AQEG, Air Quality Expert Group (2006) Air quality and climate change: a UK perspective, DEFRA, London. Alam, A., J. P. Shi and R. M. Harrison (2003) Observations of new particle formation in urban air, J. Geophys. Res., 108, 4093, doi: 10.1029/2001JD001417. Allen, A. G., J. L. Grenfell R. M. Harrison et al. (1999) Nanoparticle formation in marine airmasses: Constrasting behaviour of the open ocean and coastal environments, Atmos. Res., 51, 1–14. Anttila, T. and V.-M. Kerminen (2003) Aerosol formation via aqueous-phase chemical reactions, J. Geophys. Res., 108 (D4), 4134, doi:10.1029/2002JD002764. Baltensperger, U., H. W. Gäggeler and D. T. Jost (1988) The epiphaniometer, a new device for continuous aerosol monitoring, J. Aerosol Sci., 19, 931. Beddows, C. S. and R. M. Harrison (2008) Comparison of average particle number emission factors for heavy and light duty vehicles derived from rolling chassis dynamometer and field studies, Atmos. Environ., 42, 7954–66. Birmili, W. (1999) Production of new ultrafine aerosol particles in continental air masses, Thesis, Leipzig University, VWF Verlag für Wissenschaft und Forschung GmbH, Berlin. Birmili, W., A. Wiedensohler, J. Heintzenberg and K. Lehmann (2001) Atmospheric particle number size distribution in Central Europe: Statistical relations to air masses and meteorology, J. Geophys. Res., 106, 32005–18. Birmili, W., H. Berresheim, C. Plass-Dülmer et al. (2003) The Hohenpeißenberg aerosol formation experiment (HAFEX): A long-term study including size-resolved aerosol, H2SO4, OH and monoterpenes measurements, Atmos. Chem. and Physics., 3, 361–76. Biswas, P. and C.-Y. Wu (2005) Nanoparticles and the environment, J. Air and Waste Manage. Assoc., 55, 708–46. Biswas, S., L. Ntziachristos, K. F. Moore and C. Sioutas (2007) Particle volatility in the vicinity of a freeway with heavy-duty diesel traffic, Atmos. Environ., 41, 3479–93. Bond, T. C., B. Wehner, A. Plewka et al. (2006) Climate-relevant properties of primary particulate emissions from oil and natural gas combustion, Atmos. Environ., 40, 3574–87. Boy, M., U. Rannik, K. E. J. Lehtinen et al. (2003) Nucleation events in the continental boundary layer: long-term statistical analyses of aerosol relevant characteristics, J. Geophys. Res., 108 (D21), 4667, doi:10.1029/2003JD003838.

202

Environmental and Human Health Impacts of Nanotechnology

Brock, C. A., R. A. Washenfelder, M. Trainer et al. (2002) Particle growth in the plumes of coal-fired power plants, J. Geophys. Res., 107 (D12), doi:10.1029/2001JD001062. Brock, C. A., M. Trainer, T. B. Ryerson et al. (2003) Particle growth in urban and industrial plumes in Texas, J. Geophys. Res., 108, doi:10.1029/2002JD002746. Canagaratna, M. R., J. T. Jayne, J. L. Jimenez et al. (2007) Chemical and microphysical characterization of ambient aerosols with the Aerodyne aerosol mass spectrometer, Mass Spectrometry reviews, 26, 185–222. Casati, R., V. Scheer, R. Vogt and T. Benter (2007) Measurement of nucleation and soot mode particle emission from a diesel passenger car in real world and laboratory in situ dilution, Atmos. Environ., 41, 2125–35. Charron, A. and R. M. Harrison (2003) Primary particle formation from vehicle emissions in the roadside atmosphere, Atmos. Environ., 37, 4109–19. Charron, A., W. Birmili and R. M. Harrison (2004) Sources and Processes that influence particle size, number and mass at a rural site in England (Harwell), Report to DEFRA, London. Charron, A., W. Birmili and R. M. Harrison (2007) Factors influencing new particle formation at the rural site, Harwell, United Kingdom, J. Geophys. Res., 112, D14210, doi: 10.1029/2007JD008425. Charron, A., W. Birmili and R. M. Harrison (2008a) Fingerprinting Particle Origins According to their Size Distribution at a UK Rural Site, J. Geophys. Res., 113, D07202, doi:10/1029/2007JD008562. Charron, A., W. Birmili and R. M. Harrison (2008b), Influence of Meteorological Factors on Various particle Metrics at Harwell, UK, in preparation. Chow, J. C. (1995) Critical review: measurement methods to determine compliance with ambient air quality standards for suspended particles, J. Air Waste Manage. Assoc., 45, 320–82. Clarke, A. D., S. R. Owens and Z. Jingchuan (2006) An ultrafine sea-salt flux from breaking waves: Implications for cloud condensation nuclei in the remote marine atmosphere, J. Geophys. Res., 111, 1–2. Dahl, A., A. Gharibi, E. Swietlicki et al. (2006) Traffic-generated emissions of ultrafine particles from pavement-tire interface, Atmos. environ., 40, 1314–23. DeCarlo, P. F., J. R. Kimmel, A. Trimborn et al. (2006) A Field-Deployable High-Resolution Time-of-Flight Aerosol Mass Spectrometer, Analytical Chemistry, 78, 8281–9. Diapouli, E., A. Chaloulakou and N. Spyrellis (2007) Levels of ultrafine particles in different microenvironments – Implications to children exposure, Sci. Total Environ., 388, 28–136. Dillner, A. M., J. J. Schauer, W. F. Christensen and G. R. Cass (2005) A quantitative method for clustering size distributions of elements, Atmos. Environ., 39, 1525–37. Donaldson, K. and V. Stone (2003) Current hypotheses on the mechanisms of toxicity of ultrafine particles, Ann. Ist. Super. Sanita., 39, 405–10. Drewnick, F., S. S. Hings, P. F. DeCarlo et al. (2005) A new time-of-flight aerosol mass spectrometer (ToF-AMS)-Instrument description and first field deployment, Aerosol Sci. Technol., 39, 637–58. Du, H. and F. Yu (2006) Role of the binary H2SO4–H2O homogeneous nucleation in the formation of volatile nanoparticles in the vehicular exhaust, Atmos. Environ., 40, 7579–88. Easter, R. C. and L. K. Peters (1994) Binary homogeneous nucleation: Temperature and relative humidity fluctuations, nonlinearity and aspects of new particle production in the atmosphere, J. of Appl. Meteorology, 33, 775–84. Fushimi, A., S. Hasegawa, K. Takahashi et al. (2008) Atmospheric fate of nuclei-mode particles estimated from the number concentrations and chemical composition of particles measured at roadside and background sites, Atmos. Environ., 42, 949–59. Garg, B. D., S. H. Cadle, P. A. Mulawa et al. (2000) Brake wear particulate emissions, Environmental science and technology 34, 21, 4463–9. Gidhagen, L., C. Johansson, G. Omstedt et al. (2004a), Model simulations of NOx and ultrafine particles close to a Swedish highway, Environ. Sci. Technol., 38, 6730–40.

Atmospheric Nanoparticles

203

Gidhagen, L., C. Johansson, J. Langner and G. Olivares (2004b) Simulation of NOx and ultrafine particles in a street canyon in Stockholm, Sweden, Atmos. Environ., 38, 2029–44. Gidhagen, L., C. Johansson, J. Langner and V. L. Foltescu (2005) Urban scale modelling of particle number concentrations in Stockholm, Atmos. Environ., 39, 1711–25. Giechaskiel, B., L. Ntziacristos, Z. Samaras et al. (2005) Formation potential of vehicle exhaust nucleation mode particles on-road and in the laboratory, Atmos. Environ., 39, 3191–8. Grose, M., H. Sakurai, J. Savstrom et al. (2006) Chemical and Physical properties of ultrafine diesel exhaust particles sampled downstream of a catalytic trap, Environ. Sci. Technol., 40, 5502–7. Guo, H., L. Morawska, C. He and D. Gilbert (2007) Impact of ventilation scenario on air exchange rates and on indoor particle number concentrations in an air-conditioned classroom, Atmos. Environ., 42, 757–68. Hamed, A., J. Joutsensaari, S. Mikkonen et al. (2007) Nucleation and growth of new particles in Po valley, Italy, Atmos. Chem. Phys., 7, 355–76. Harrison, R. M. (2007) Nanoparticles in the atmosphere, Issues in Environmental Science and Technology, No. 24, Nanothechnology: consequences for Human Health and the Environment, The Royal Society of Chemistry, 35–49. Harrison, R. M. and A. M. Jones (2005) Multisite study of particle number concentrations in urban air, Environ. Sci. Technol., 39, 6063–70. Harrison, R. M. and J. Yin (2000) Particulate matter in the atmosphere: which particle properties are important for its effects on health, Sci. Tot. Environ., 249, 85–101 Harrison, R. M., M. Jones and G. Collins (1999). Measurements of the Physical Properties of Particles in the Urban Atmosphere, Atmos. Environ., 33, 309–21. Harrison, R. M., J. L. Grenfell, N. Savage et al. (2000a), Observations of new particle production in the atmosphere of a moderately polluted site in Eastern England, J. Geophys. Res., 105, 17819–32. Harrison, R. M., J. P. Shi, S. Xi et al. (2000b), Measurement of number, mass and size distribution of particles in the atmosphere, Phil. Trans. R. Soc. Lond. A, 358, 2567–80. Harrison, R. M., C. Meddings, S. Thomas and J. M. Delgado Saborit (2004) Relationships between indoor/outdoor and central site concentrations of particulate matter determined as mass and ultrafine particle count within the RUPIOH epidemiological study, in Proceedings of the Eighth Annual UK Review Meeting on Outdoor and Indoor Air Pollution Research, No. 5.12, http://www.cranfield.ac.uk/health/researchareas/ environmenthealth/ieh/ieh%20publications/w18.pdf He, C., L. Morawska, J. Hitchins and D. Gilbert (2004) Contribution from indoor sources to particle number and mass concentrations in residential houses, Atmos. Environ., 38, 3405–15. He, C., L. Morawska, L. Taplin (2007) Particle emission characteristics of office printers, Environ. Sci. Technol., 41, 6039–45. Hillemann, L., A. Zschoppe, R. Caldow (2007) Characterization of environmental ultrafine particles with the UFP 330, system calibration and evaluation, Proceedings of UFIPOLNET Final Conference: Ultrafine Particle in Urban Air (ed. Gerwig, H.), Dresden, Germany, http://www.umwelt.sachsen.de/umwelt/en/5636.htm Hinds, W. C. (1999) Aerosol Technology: Properties, behaviour and measurement of airborne particles (2nd edn), John Wiley & Sons Inc., New York. Hoek, G., G. Kos, R. M. Harrison et al. (2008) Indoor–outdoor relationships of particle number and mass in four European cities, Atmos. Environ., 42, 156–69. Holmes, N. S. (2007) A review of particle formation events and growth in the atmosphere in the various environments and discussion of mechanistic implications, Atmos. Environ., 41, 2183–201. Hussein, T., K. Hämeri, M. S. A. Heikkinen and M. Kulmala (2005a) Indoor and outdoor particle size characterization at a family house in Espoo, Finland, Atmos. Environ., 39, 3697–709.

204

Environmental and Human Health Impacts of Nanotechnology

Hussein, T., K. Hämeri, P. P. Aalto et al. (2005b) Modal structure and spatial temporal variations of urban and suburban aerosols in Helsinki, Finland, Atmos. Environ., 39, 1655–68. Hussein, T., T. Glytsos, J. Ondrácˇ ek et al. (2006) Particle size characterization and emission rates during indoor activities in a house, Atmos. Environ., 40, 4285–307. Jacobson, M. Z., D. B. Kittelson, W. F. Watts (2005) Enhanced coagulation due to evaporation and its effect on nanoparticle evolution, Environ. Sci. Technol., 69, 9486–92. Jaenisch, V., F. Stratmann, P. H. Austin and D. A. Hegg (1997) Influence of turbulent mixing processes on new particle formation: first results, J. of Aerosol Sci., 28, S27–8. Janhäll, S., A. M. Jonsson, P. Molnàr et al. (2004) Size resolved traffic emission factors of submicrometer particles, Atmos. Environ., 38, 4331–40. Jayne, J. T., D. C. Leard, X. Zhang et al. (2000) Development of an aerosol mass spectrometer for size and composition analysis of submicron particles, Aerosol Sci. Technol., 33, 49–70. Joutsensaari, J., P. Vaattovaara, M. Vesterinen et al. (2001) A novel tandem differential mobility analyser with organic vapour treatment of aerosol particles, Atmos. Chem. Phys., 1, 51–60. Jung, H., D. B. Kittelson, M. R. Zachariah (2005) The influence of a cerium additive on ultrafine diesel particle emissions and kinetics of oxidation, Combust. Flame, 142, 276–88. Jung, J. G., P. J. Adams and S. N. Pandis (2006) Simulating the size distribution and chemical composition of ultrafine particles during nucleation events, Atmos. Environ., 40, 2248–59. Kavouras, I. G. and E. G. Stephanou (2002) Direct evidence of atmospheric secondary organic aerosol in forest atmosphere through heteromolecular nucleation, Environ. Sci. Technol., 36, 5083–91. Kerminen, V.-M. and A. S. Wexler (1996) The occurrence of sulfuric acid-water nucleation in plumes: urban environment, Tellus, 48B, 65–82. Kerminen, V.-M., T. A. Pakkanen, T. Mäkelä et al. (2007) Development of particle number size distribution near a major road in Helsinki during an episodic inversion situation, Atmos. Environ., 41, 1759–67. Khosrawi, F. and P. Konopka (2003) Enhanced particle formation and growth due to mixing processes in the tropopause region, Atmos. Environ., 37, 903–910. Kittelson, D. B. (1998). Engines and Nanoparticles: A Review. J. Aerosol Sci., 29, 575–88. Kittelson, D. B., W. F. Watts and J. P. Johnson (2006a) On-road and laboratory evaluation of combustion aerosols – Part 1: Summary of diesel engine results, J. Aerosol Sci., 37, 913–30. Kittelson, D. B., W. F. Watts, J. P. Johnson et al. (2006b) On-road and laboratory evaluation of combustion aerosols – Part 1: Summary of spark ignition engine results, J. Aerosol Sci., 37, 931–49. Kittelson, D. B., W. F. Watts, J. P. Johnson et al. (2006c) On-road evaluation of two diesel exhaust after treatment devices, J. Aerosol Sci., 37, 1140–51. Koponen, I. K., A. Asmi, P. Keronen et al. (2001) Indoor air measurement campaign in Helsinki, Finland 1999 – the effect of outdoor air pollution on indoor air, Atmos. Environ., 35, 1465–77. Korhonen, H., K. E. J. Lehtinene, L. Pirjola et al. (2003) Simulation of atmospheric nucleation mode: a comparison of nucleation models and size distributions representations, J. Geophys. Res., 108, D15, doi:10.1029/2002JD003305. Kuhn, T., M. Krudysz, Y. Zhu et al. (2005) Volatility of indoor and outdoor ultrafine particulate matter near a freeway, J. Aerosol Science, 36, 291–302. Kulmala, M., L. Pirjola, J. M. Mäkelä (2000) Stable sulfate clusters as a source of new atmospheric particles, Nature, 104, 66–9. Kulmala, M., H. Vehkamäki, T. Petäjä et al. (2004) Formation and growth rates of ultrafine atmospheric particles: A review of observations, J. of Aerosol Sci., 35, 143–76. Kulmala, M., I. Riipinen, M. Sipilä et al. (2007) Toward direct measurement of atmospheric nucleation, Science, 318, 89–92.

Atmospheric Nanoparticles

205

Kumar, P., P. Fennell, D. Langley and R. Britter (2008) Pseudo-simultaneous measurements for the vertical variation of coarse, fine and ultrafine particles in an urban street canyon, Atmos. Environ., 42, 4304–19. Laaksonen, A., M. Kulmala, C. D. O’Dowd et al. (2008) The role of VOC oxidation products in continental new particle formation, Atmos. Chem. Phys., 8, 2657–65. Lake, D. L., M. P. Tolocka, A. S. Wexler and M. V. Johnston (2003) Mass spectrometry of individual particles between 50 and 750 nm in diameter at the Baltimore Supersite, Environ. Sci. Technol., 37, 3268–74. Lee, C.-W. and D.-J. Hsu (2007) Measurements of fine and ultrafine particles formation in photocopy centers in Taiwan, Atmos. Environ., 41, 6598–609. Lehtinen, K. E. J. and M. Kulmala (2003) A model for particle formation and growth in the atmosphere with molecular resolution in size, Atmos. Chem. Phys., 3, 251–7. Lin, C.-C., S.-J. Chen, K.-L. Huang et al. (2005) Characteristics of metals in nano/ultrafine/ fine/coarse particles collected beside a heavily trafficked road., Environ. Sci. Technol., 39, 8113–21. Longley, I. D., M. W. Gallagher, J. R. Dorsey and M. Flynn (2004) A case-study of fine particle concentrations and fluxes measured in a busy street canyon in Manchester, UK, Atmos. Environ., 38, 3595–603. Lunden, M., M. Revzan, K. L. Fischer et al. (2003) The transformation of outdoor ammonium nitrate aerosols in the indoor environment, Atmos. Environ., 37, 5633–44. Mäkelä, J. M., P. Aalto, V. Jokinen et al. (1997) Observations of ultrafine aerosol particle formation and growth in boreal forest, Geophys. Res. Letts., 24, 1219–22. Mäkelä, J. M., I. K. Koponen, P. Aalto and M. Kulmala (2000) One-year data of submicron size modes of tropospheric background aerosol in southern Finland, J. Aerosol Sci., 31, 595–611. Maricq, M. M. (2007) Chemical characterization of particulate emissions from diesel engines: A review, J. Aerosol Sci., 38, 1079–118. Maricq, M. M., D. H. Podsiadlik and R. E. Chase (1999a) Gasoline vehicle particle size distributions: comparison of steady state, FTP, and US06 measurements, Environ. Sci. Technol., 33, 2007–15. Maricq, M. M., D. H. Podsiadlik and R. E. Chase (1999b) Examination of the size-resolved and transient nature of motor vehicle particle emissions, Environ. Sci. Technol., 33, 1618–26. Maricq, M. M., R. E. Chase, N. Xu and P. M. Laing (2002) The effects of the catalytic converter and fuel sulfur level on motor vehicle particulate matter emissions: light duty diesel vehicles, Environ. Sci. Technol., 36, 283–289. Mark, D. (2007) Occupational exposure to nanoparticles and nanotubes, in Issues in Environmental Science and Technology, Vol. 24, (ed. R. E. Hester and R. M. Harrison), Royal Society of Chemistry, Cambridge, 50–80. Mathis, U., M. Mohr and A.-M. Forss (2005) Comprehensive particle characterization of modern gasoline and diesel passenger cars at low ambient temperatures, Atmos. Environ., 39, 107–17. Maynard, A. D. (1995) The development of a new thermophoretic precipitator for scanning transmission electron microscope analysis of ultrafine aerosol partricles, Aerosol. Sci. Technol., 23, 521–33. McMurry, P. H. (2000) A review of atmospheric aerosol measurements, Atmos. Environ., 34, 1959–99. McMurry, P. H. and F. L. Eisele (2005) Preface to topical collection on new particle formation in Atlanta., J. Geophys. Res., 110, D22S01, doi:10.1029/2005JD006644. Merikanto, J., I. Napari, H. Vehkamäki et al. (2007) New parameterization of sulfuric acidammonia-water ternary nucleation rates at tropospheric conditions, J. Geophys. Res., 112, D15207, doi:10.1029/2006JD007977. Meyer, N. K. and Z. D. Ristovski (2007) Ternary nucleation as a mechanism for the production of diesel nanoparticles: experimental analysis of the volatile and hygroscopic proper-

206

Environmental and Human Health Impacts of Nanotechnology

ties of diesel exhaust using the volatilization and humidification tandem differential mobility analyzer, Environ. Sci. Technol., 41, 7309–14. Miller, A., G. Ahlstrand, D. Kittelson and M. Zachariah (2007) The fate of metal (Fe) during diesel combustion: morphology, chemistry, and formation pathways of nanoparticles, Combustion and Flame, 149, 129–43. Molnàr, P., S. Janhall and M. Hallquist (2002) Roadside Measurements of Fine and Ultrafine Particles at a Major Road North of Gothenburg, Atmos. Environ. 36, 4115–23. Morawska, L. and J. J. Zhang (2002) Combustion sources of particles. 1. Health relevance and source signatures, Chemosphere, 49, 1045–58. Morawska, L., S. Thomas, M. Jamriska, G. Johnson (1999a), The modality of particle size distributions of environmental aerosols, Atmos. Environ., 33, 4401–11. Morawska, L., S. Thomas, D. Gilbert et al. (1999b), A study of the horizontal and vertical profile of submicrometer particles in relation to a busy road, Atmos. Environ., 33, 1261–74. Morawska, L., E. R. Jayaratne, K. Mengersen et al. (2002) Differences in airborne particle and gaseous concentrations in urban air between weekdays and weekends, Atmos. Environ., 36, 4375–83. Morawska, L., C. He, J. Hitchins et al. (2003) Characteristics of particle number and mass concentrations in residential houses in Brisbane, Australia, Atmos. Environ., 37, 4195–203. Morawska L., M. Jamriska, S. Thomas et al. (2005) Quantification of particle number emission factors for motor vehicles from on-road measurements, Environ. Sci. Techcol., 39, 9130–39. Napari, I., M. Noppel, H. Vehkamäki and M. Kulmala (2002) Parametrization of ternary nucleation rates for H2SO4-NH3-H2O vapors, J. Geophys. Res., 107, no D19, 4381, doi:10.1029/2002JD002132. Nilsson, E. D. and M. Kulmala (1998) The potential for atmospheric mixing processes to enhance the binary nucleation rate, J. Geophys. Res., 103, 1381–9. Nilsson, E. D., J. Paatero and M. Boy (2001) Effects of air masses and synoptic weather on aerosol formation in the continental boundary layer, Tellus Series B Chem. and Phys. Met., 53, 462–78. Noble, C. A. and K. A. Prather (2000) Real-time single particle mass spectrometry: A historical review of a quarter century of the chemical analysis of aerosols, Mass Spectrom Rev., 19, 248–274. Ntziachristos, L., Z. Samarasa, E. Zervasb and P. Dorlhène (2005) Effects of a catalysed and an additized particle filter on the emissions of a diesel passenger car operating on low sulfur fuels, Atmos. Environ., 39, 4925–36. Oberdörster, G. (2000) Toxicology of ultrafine particles: in vivo studies, Philos. Trans. R. Soc. Lond. A., 358, 2719–40. O’Dowd, C., P. Aalto, K. Hämeri et al. (2002a) Marine aerosol formation from biogenic iodine emissions, Nature, 417, 632–6. O’Dowd, C. D., P. Aalto, K. Hämeri et al. (2002b) Atmospheric particles from organic vapours, Nature, 416, 497–8. O’Dowd, C. D., K. Hämeri, J. M. Mäkela et al. (2002c) A dedicated study of New Particle Formation and Fate in the Coastal Environment (PARFORCE): Overview of objectives and achievements, J. of Geophys. Res., 107, D19, 8107. Olivares, G, C. Johansson, J. Ström and H.-C. Hansson (2007) The role of ambient temperature for particle number concentrations in a street canyon, Atmos. Environ., 41, 2145–55. Penner, J. E., J. Quaas, T. Storelvmo et al. (2006) Model intercomparison of indirect aerosol effects, Atmos. Chem. Phys., 6, 3391–405. Philippin, S., A. Wiedensohler and F. Stratmann (2004) Measurements of non-volatile fractions of pollution aerosols with an eight-tube volatility tandem differential mobility analyzer (VTDMA-8), J. Aerosol Sci., 35, 185–203. Puustinen, A., K. Hämeri, J. Pekkanen et al. (2007) Spatial variation of particle number and mass over four European cities, Atmos. Environ., 41, 6622–36.

Atmospheric Nanoparticles

207

Riley, W. J., T. E. Mckone, A. C. K. Lai and W. W. Nazaroff (2002) Indoor particulate matter of outdoor origin: importance of size-dependent removal mechanisms, Environ. Sci. Technol., 36, 200–7. Ristovski, Z., L. Morawska, J. Hitchins (1998) Submicrometer and supermicrometer particulate emission from spark ignition vehicles, Environ. Sci. Technol., 32, 3845–52. Rodriguez, S., R. Van Dingenen, J.-P. Putaud et al. (2005) Nucleation and growth of new particles in the rural atmosphere of northern Italy – relation to air quality monitoring, Atmos. Environ., 39, 6734–46. Rohr, A. C., C. J. Weschler, P. Koutrakis and J. D. Spengler (2003) Generation and quantification of ultrafine particles through terpene/ozone reaction in a chamber stting, Aerosol Sci. technol., 37, 65–78. Rönkkö, T., A. Virtanen, K. Vaaraslahti et al. (2006) Effect of dilution conditions and driving parameters on nucleation mode particles in diesel exhaust: laboratory and on-road study, Atmos. Environ., 39, 6153. Rönkkö, T., A. Virtanen, J. Kannosto et al. (2007) Nucleation mode particles with a nonvolatile core in the exhaust of a heavy-duty diesel vehicle, Environ. Sci. Technol., 41, 6384–9. Sakurai, H., H. J. Tobias, K. Park et al. (2003) On-line measurements of diesel nanoparticle composition and volatility, Atmos. Environ., 37, 1199–210. Sakurai, H., M. A. Fink, P. H. McMurry et al. (2005) Hygroscopy and volatility of 4–10 nm particles during summertime atmospheric nucleation events in urban Atlanta, J. of Geophys. Res., 110, D22S04, doi:10.1029/2005JD005918. Sanders, P. G., N. Xu, T. M. Dalka and M. M. Maricq (2003) Airborne brake wear debris: size distributions, composition, and a comparison of dynamometer and vehicle tests, Environ. Sci. Technol., 37, 4060–9. Schneider, J., N. Hock, S. Weimer and S. Borrmann (2005) Nucleation particles in diesel exhaust: Composition inferred from in situ mass spectrometric analysis, Environ. Sci. Technol, 39, 6153–61. Seaton, A., J. Cherrie, M. Dennekamp et al. (2005) The London Underground: dust and hazards to health, Occup. Environ. Med., 62, 355–62. See, S. W. and R. Balasubramanian (2006) Risk assessment of exposure to indoor aerosols associated with Chinese cooking, Environ. Res., 102, 197–204. Seinfeld, J. H. and S. N. Pandis (1998) Atmospheric chemistry and physics, from air pollution to climate change, John Wiley& Sons Inc., New York. Shi, J. P. and R. M. Harrison (1999). Investigation of Ultrafine Particle Formation during Diesel Exhaust Dilution, Environ. Sci. Technol., 33, 3730–6. Shi, J. P., D. Mark and R. M. Harrison (2000) Characterization of Particles from a Current Technology Heavy-duty Diesel Engine, Environ. Sci. Technol., 34, 748–55. Shi, J. P., D. E. Evans, A. A. Khan and R. M. Harrison (2001a) Sources and concentration of nanoparticles (< 10 nm diameter) in the urban atmosphere, Atmos. Environ., 35, 1193–202. Shi, J. P., R. M. Harrison and D. Evans (2001b), Comparison of ambient particle surface area measurement by epiphaniometer and SMPS/APS, Atmos. Environ., 35, 6193–200. Skillas, G., Z. Qian, U. Baltensperger et al. (2000) The influence of additives on the size distribution and composition of particles produced by diesel engines, Combust. Sci. Technol., 154, 259–73. Slama Lighty, J., J. M. Veranth and A. F. Sarofim (2000) Combustion aerosols: factors governing their size and composition and implication to human health, J. Air and Waste Manage. Assoc., 50, 1565–18. Smith, J. N., K. F. Moore, F. L. Eisel et al. (2005) Chemical composition of atmospheric nanoparticules during nucleation events in Atlanta, J. Geophys. Res., 110, D22S03, doi:10.1029/2005JD005912. Sodeman, D. A., S. M. Toner and K. A. Prather (2005) Determination of single particle mass spectral signatures from light-duty vehicle emissions, Environ. Sci. Technol., 39, 4569–80.

208

Environmental and Human Health Impacts of Nanotechnology

Stanier, C. O., A. Y. Khlystov and S. N. Pandis (2004) Nucleation events during the Pittsburgh air quality study: description and relation to key meteorological, gas phase and aerosol parameters, Aerosol Sci. Technol., 38, 253–64 Stone, V. (2009) Human toxicology and effects of nanoparticles, in Environmental and human health effects of nanoparticles (eds Lead, J., and Smith, E.), Wiley-Blackwell, Oxford. Su, Y., M. F. Sipin, H. Furutani and K. A. Prather (2004) Development and characterization of an aerosol time-of-flight mass spectrometer with increased detection efficiency, Anal. Chem., 76, 712–9. Sullivan, R. C. and K. A. Prather (2005) Recent advances in our understanding of atmospheric chemistry and climate made possible by on-line aerosol analysis instrumentation, Anal. Chem., 77, 3861–86. Tobias, H. J., P. M. Kooiman, K. S. Docherty and P. J. Ziemann (2000) Real-Time Chemical Analysis of Organic Aerosols Using a Thermal Desorption Particle Beam Mass Spectrometer, Aerosol Sci. Technol., 33, 170–90. Tobias, H. J., D. E. Beving, P. J. Ziemann et al. (2001) Chemical analysis of diesel engine nanoparticles using a nano-DMA/thermal desorption particle beam mass spectrometer, Environ. Sci. Technol., 35, 2233–43. Tolocka, M. P., D. A. Lake, M. V. Johnston and A. S. Wexler (2004a) Number concentrations of fine and ultrafine particles containing metals, Atmos. Environ., 38, 3263–73. Tolocka, M. P., D. A. Lake, M. V. Johnston and A. S. Wexler (2004b) Ultrafine nitrate particle events in Baltimore observed by real-time single particle mass spectrometry, Atmos. Environ., 38, 3215–23. Uhrner, U., W. Birmili, F. Stratmann et al. (2003) Particle formation at a continental background site: comparison of model results with observations, Atmos. Chem. Phys., 3, 347–59. Uhrner, U., S. von Löwis, H. Vehkamäki et al. (2007) Dilution and aerosol dynamics within a diesel car exhaust plume – CFD simulations of on-road measurement conditions, Atmos. Environ., 41, 7440–61. Vaaraslahti, K., A. Virtanen, J. Ristimäki and J. Keskinen (2004) Nucleation mode formation in heavy-duty diesel exhaust with and without a particulate filter, Environ. Sci. Technol., 38, 4884–90. Vaaraslahti, K., J. Keskinen, B. Giechaskiel et al. (2005) Effect of lubricant on the formation of heavy-duty diesel exhaust nanoparticles, Environ. Sci. Technol., 39, 8497–504. Väkevä, M., K. Hämeri, M. Kulmala et al. (1999) Street level versus rooftop concentrations of submicron aerosol particles and gaseous pollutants in an urban street canyon, Atmos. Environ., 33, 1385–97. Vana, M., M. Kulmala, M. dal Maso et al. (2004) Comparative study of nucleation mode aerosol particle and intermediate air ions at three sites, J. Geophys. Res., 109, D17201, 1–10. Van Dingenen, R., F. Raes, J.-P. Putaud et al. (2004) A European Aerosol Phenomenology-1: Physical characteristics of particulate matter at kerbside, urban, rural and background sites in Europe, Atmos. Environ., 38, 2561–77. Vehkamäki, H., M. Kulmala, I. Napari et al. (2002) An improved parameterization for sulfuric acid-water nucleation rates for tropospheric and stratospheric conditions, J. Geophys. Res., 107 (D22), 4622, doi:10.1029/2002JD002184. Vehkamäki, H., M. Dal Maso, T. Hussein et al. (2004) Atmospheric particle formation events at Värriö measurement station in Finnish Lapland 1998–2002, Atmos. Chem. and Physics, 4, 2015–23. Vesterinen, M., K. E. J. Lehtinen, M. Kulmala and A. Laaksonen (2007) Effect of particle phase oligomer formation on aerosol growth, Atmos. Environ., 41, 1768–76. Vogt, R., V. Scheer, R. Casati and T. Benter (2003) On-road measurement of particle emission in the exhaust plume of a diesel passenger car, Environ. Sci. Technol., 37, 4070–6. Voisin, D., J. N. Smith, H. Sakurai et al. (2003) Thermal desorption chemical ionization mass spectrometer for ultrafine particle chemical composition, Aerosol Sci. Technol., 37, 471–5. Vouitsis, E., L. Ntziachristos and Z. Samaras (2005) Modelling of diesel exhaust aerosol during laboratory sampling, Atmos. Environ., 39, 1335–45.

Atmospheric Nanoparticles

209

Wake, D. (2006) The assessments of different metrics of the concentration of nano(ultrafine) particles in existing and new industries, Health & Safety Executive Report RR513, Buxton, UK. Wallace, L. (2005) Ultrafine particles from a vented gas clothes dryer, Atmos. Environ., 39, 5777–86. Wardoyo, A. Y. P., L. Morawska, Z. D. Ritovski et al. (2007) Size distribution of particles emitted from grass fires in the Northern Territory, Australia, Atmos. Environ., 41, 8609–19. Wehner, B., W. Birmili, T. Gnauk and A. Wiedensohler (2002) Particle Number Size Distributions in a Street Canyon and their Transformation into the Urban-Air Background: Measurements and a Simple Model Study, Atmos. Environ., 36, 2215–23. Wehner, B., T. Petäjä, M. Boy et al. (2005) The contribution of sulfuric acid and non-volatile compounds on the growth of freshly formed atmospheric aerosols, Geophys. Res. Lett., 32, L17810, doi:10.1029/2005GL023827. Wehner, B., T. Tuch, A. Wiedensohler et al. (2007) Ultrafine particles: comparisons UFP 330/ DMPS in Gerwig, H. (Ed.) Proceedings of UFIPOLNET Final Conference: Ultrafine Particle in Urban Air, Dresden, Germany, http://www.umwelt.sachsen.de/umwelt/en/5636. htm Weijers, E. P., A. Y. Khlystov, G. P. A. Kos and J. W. Erisman (2004) Variability of particulate matter concentrations along roads and motorways determined by a moving measurement unit, Atmos. Environ., 38, 2993–3002. Weschler, C. J. and H. C. Shields (1999) Indoor ozone/terpene reaction as a source of indoor particles, Atmos. Environ., 33, 2301–12. Woo, K., S. D. R. Chen, D. Y. H. Pui and P. H. McMurry (2001) Measurement of Atlanta Aerosol Size Distributions: Observations of Ultrafine Particles Events, Aerosol Sci. Technol., 34, 75–87. Young, L.-H., D. R. Benson, W. M. Montanaro et al. (2007) Enhanced new particle formation observed in the northern mid latitude tropopause region, J. Geophys. Re., 112, D10218, doi:10.1029/2006JD008109. Yu, F. and R. P. Turco (2000) Ultrafine aerosol formation via ion-mediated nucleation, Geophys. Res. Letters, 27, 883–6. Yu, H., Y. J. Kaufman, M. Chin et al. (2006) A review of measurement-based assessments of the aerosol direct radiative effect and forcing, Atmos. Chem. Phys., 6, 613–66. Zhang, K. M. and A. S. Wexler (2002) A hypothesis for growth of fresh atmospheric nuclei, J. Geophys. Res., 107 107, (D21), 4577, doi:10.1029/2002JD002180. Zhang, K. M. and A. S. Wexler (2004) Evolution of particle number distribution near roadways. Part I: analysis of aerosol dynamics and its implications for engine emission measurement, Atmos. Environ., 38, 6643–53. Zhang, Q., C. O. Stanier, M. R. Canagaratna et al. (2004a) Insights into the chemistry of new particle formation and growth events in Pittsburgh based on aerosol mass spectrometry, Environ. Sci. Technol., 38, 4797–809. Zhang, K. M., A. S. Wexler, Y. F. Zhu et al. (2004b) Evolution of particle number distribution near roadways. Part II: the ‘road-to-ambient’ process, Atmos. Environ., 38, 6655–65. Zhu, Y. and W. C. Hinds (2005) Predicting particle number concentrations near a highway based on vertical concentration profile, Atmos. Environ., 39, 1557–66. Zhu, Y., W. C. Hinds, S. Kim et al. (2002) Study of ultrafine particles near a major highway with heavy-duty diesel traffic, Atmos. Environ., 36, 4323–35. Zhu, Y., W. C. Hinds, M. Krudysz et al. (2005) Penetration of freeway ultrafine particles into indoor environments, J. Aerosol Sci., 36, 303–22. Zhu, Y., A. Eigurei-Fernandez, W. C. Hinds and A. H. Miguel (2007) In-cabin commuter exposure to ultrafine particles on Los Angeles freeways, Environ. Sci. Technol., 41, 2138–45. Zschoppe, A., L. Hillemann, A. Rudolph et al. (2007) A new particle measurement system for environmental ultrafine particles, Proceedings of UFIPOLNET Final Conference: Ultrafine Particle in Urban Air (ed. Gerwig, H.), Dresden, Germany, http://www.umwelt.sachsen.de/ umwelt/en/5636.htm

6 Analysis and Characterization of Manufactured Nanoparticles in Aquatic Environments Martin Hassellöv1 and Ralf Kaegi2 2

6.1

1 Department of Chemistry, University of Gothenburg, Sweden Swiss Federal Institute of Aquatic Science and Technology (Eawag), Dübendorf, Switzerland

Introduction

This chapter aims to discuss the minimum set of key properties that describe nanoparticles (NPs) in aquatic environments, and to review suitable analysis and characterization methods. Special focus is on characterization of manufactured NPs and associated nanoscale processes (Gilbert and Banfield, 2005). Manufactured NPs have recently raised concerns because of their potential effects on human health and the environment (Colvin, 2003; Donaldson et al., 2004; Handy et al., 2008; Nel et al., 2006; Oberdorster et al., 2005). Therefore, the analytical demands related to both exposure and effect assessment are discussed. Atmospheric NPs and airborne exposure are not covered in this chapter, but within Chapters 5 and 8, and also reviewed elsewhere (Burleson et al., 2004; Maynard, 2000); Powers et al., 2007). One aim has been to bring together concepts and knowledge from analytical chemistry, (e.g. quantitative determination of chemical composition in nanoparticle ensembles, such as size fractions or on a single NP), the mature field of particle size analysis (mainly developed for micron particles) (Barth and Flippen, 1995), environmental colloid science (Wilkinson and Lead, 2007) and the young field of nanometrology (Graham, 2007). Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

212

6.1.1

Environmental and Human Health Impacts of Nanotechnology

Nanoparticles in the Aquatic Environment

Nanoparticles are ubiquitous in freshwater (Wigginton et al., 2007a), seawater (Wells and Goldberg, 1991), groundwater (Banfield and Navrotsky, 2001), soils (Citeau et al., 2006), sediment (Waychunas et al., 2005; van der Zee et al., 2003) and ancient ice cores (Murr et al., 2004). In aquatic environments three main categories of NPs may coexist; natural NPs, incidental (adventitious) NPs and manufactured NPs. Sources of natural NPs can be both inorganic and organic (and are usually mixtures of both). Examples of inorganic natural NPs are iron oxyhydroxides (Davison and De Vitre, 1992; Perret et al., 2000; Waychunas et al., 2005), manganese oxides, aluminum hydroxides and alumina silicates (Filella, 2007). Examples of organic nanomaterials are humic substances (humic and fulvic acids) and acid polysaccharides (Buffle et al., 1998; Wilkinson et al., 1999). Sources of incidentally produced NPs in aquatic systems are, for example, mining activities, atmospheric deposition of combustion NPs, traffic related NPs (combustion, corrosion of vehicles, frictional wear of tires, road and brake system). While there is a high background of natural NPs in most environmental waters, the incidentally produced NPs are now becoming abundant in some polluted waters. It is likely that, with rapidly developing applications of nanotechnology, manufactured NPs will become an increasingly important component of the aquatic environment. These three categories not only coexist in the aquatic environment, but most likely also interact in complex ways which are poorly understood (Baalousha et al., 2008; Chen and Elimelech, 2007, 2008; Diegoli et al., 2008). Analysis of these mixtures is more challenging than the analysis of each single material separately. 6.1.2

Concepts and Definitions Relating to Analysis and Characterization

Analysis in a physicochemical sense generally means quantification of amounts, while characterization usually stands for determination of properties. However, the terms are overlapping since determination of a size distribution would be classified as a characterization of properties, but it actually requires quantification of the amounts (particle numbers or mass) in each size fraction. Many of the characterization techniques that determine distributions of properties are only semi-quantitative, or their concentration determinations are skewed in some way (e.g. light scattering that can be used to generate size distributions that are intensity weighted). In this chapter analysis deals with the quantitative determination of total numbers or mass (volume), while characterization is a broader term. It is often stated that very little is known about manufactured NPs in environmental samples, and that it is not known how to measure them (RS/RAEng, 2004). That is only partly true. Some manufactured NPs found in many new applications, for example titanium dioxide, have actually been synthesized and at least partly characterized in environmental waters for some time. Others, for example silica and iron oxides, have been used as model NPs in order to better understand their naturally produced analogues (Puls and Powell, 1992; Ryan et al., 2000). In addition, natural nanomaterials, often called colloids, have been extensively studied, quantified and characterized in almost all types of environmental compartments

Analysis and Characterization of Manufactured NPs in Aquatic Environments

213

(Buffle and van Leeuwen, 1992; Wilkinson and Lead, 2007). In addition to these fields of environmental science, the pure and applied physics (nanoscience) and physical chemistry (colloid and surface chemistry) can also contribute, to form the basis for the development of analytical and characterization methodologies for manufactured NPs in the environment. However, the instruments and methods often used in nanoscience and colloid chemistry rely on the assumption that samples are very simple systems with often only one type of NP, no other impurities, narrow size distributions (low polydispersity) and high concentrations. Most of these assumptions do not hold for environmental samples or ecotoxicological tests and caution should be taken when applying methods optimized for different operating conditions, even for the same type of manufactured NP. It is appropriate here to discuss briefly definitions of the terms NP, colloids and other terms from a measurement perspective; general definitions of NPs are discussed in Chapter 1. The fields of nanoscience and nanotechnology are young and rapidly developing. Therefore, rigorous definitions have not yet been formulated but extensive work are ongoing in various international and national standardization organizations, for example ISO, CEN, BSI, ANSI, and so on. Definitions for nanotechnology are that it applies to nanomaterials that have at least one dimension in the size range 1–100 nm plus that the material shall have an added functionality due to its size or that the nanotechnology includes manipulations of the materials at this length scale. But definitions of nanomaterials as such are often defined based only on their sizes. For example, a nanoparticle is defined to having all three dimensions in the range 1–100 nm (ISO, 2008). This is the definition of NP that will be applied here. Colloids or colloidal particles, were first defined in 1861 as particles ‘being immune to sedimentation’ (Graham, 1861). Colloids have been defined as being one continuous phase inside another continuous phase. One trivial example of a colloidal system is milk, where the fat droplets and proteins are dispersed in the water phase. The upper size range of colloids is controlled by the particle’s properties of being small enough to have Brownian motion (diffusion) that dominates over sedimentation and that holds up to about 1 µm (depending on density) for non-aggregating NPs. IUPAC has operationally defined colloids as particles having at least one dimension in the size range 1–1000 nm (Lyklema and van Olphen, 1979), and that is probably the most widely accepted definition (Lead and Wilkinson, 2006). However, other definitions are used. For instance, Gustafsson and Gschwend discussed a concept of a more functional definition of colloids (Gustafsson and Gschwend, 1997) in natural aquatic systems. In their ‘chemcentric’ view, colloids are defined as particles or macromolecules that can provide a molecular milieu which chemicals can be sorbed into and onto from the bulk phase. Such a milieu can be significantly different from the bulk phase, for example in terms of ionic composition, pH or charges. A colloid particle can be large enough to have an electrical double layer (EDL) around the particle with charges on the particle surface. However, this charge potential drops off exponentially as the distance from the surface increases and in the bulk phase. The chemcentric view on colloids is, therefore, not primarily based on size. Another important distinction/highlight of the chemcentric view is that colloids that fulfil the colloidal requirements in one water system may not do so in another, where water chemistry or physics are different (e.g different size cut-offs for sedimentation in a stagnant and turbulent water body).

214

Environmental and Human Health Impacts of Nanotechnology

Depending on the definition of colloids or NP used, different requirements are put on analysis and characterization methods. In reality, it can be seen in literature that operational definitions based on methods are often chosen to define the lower and upper size limit or other chemical boundaries of colloids. For example, it is often seen in literature that the colloids in a sample are defined as passing a 0.45 µm filter but being retained by a 1 kDa ultrafiltration membrane.

6.2

Nanoparticle Analysis and Characterization Methods

This section discusses briefly why certain nanoparticle characteristics are important in environmental transport, chemical reactivity and ecotoxicology. This is followed by a detailed discussion on a suite of applicable analysis and characterization methods and discussion on their analytical validation and use. 6.2.1

Important Nanoparticle Characteristics

Nanoparticles are often a heterogeneous mixture of particles of different sizes, shapes, chemical composition, crystal and amorphous structures and with coatings and surface chemistry. In particular, coatings and surface chemistry can, to some extent, control their behaviour and, more importantly, often change properties depending on their surrounding media (Banfield and Navrotsky, 2001; Filella, 2007). Therefore, the view of manufactured NPs as monodisperse, spherical particles with clean surfaces must be challenged, particularly when present in complex media such as environmental or biological samples. In Figure 6.1 the complexity of NPs is illustrated in a multidimensional array of properties, each of which likely exerts some control of both nanogeochemical and nanoecotoxicological processes. Consequently, to characterize most of the potentially significant physicochemical properties of NPs in environment is challenging from the methodological point of view. The specifications of each of the methods discussed (in Section 6.2.3 to 6.2.8) have been summarized for the respective physico-chemical property in Table 6.1. 6.2.1.1

Size and Size Distribution

Nanoparticle size is a governing parameter both for the environmental fate and behaviour (Gustafsson and Gschwend, 1997; Lead and Wilkinson, 2006) and for the bioavailability to cells and organisms (Chapter 7 and 9 (this volume); Donaldson et al., 2004). Jiang et al. (2008) showed that the cell signalling response was size specific with the highest effect for 40–50 nm. This was consistent with a study of gold nanoparticle uptake by mammalian cells that showed the fastest uptake for 50 nm particles, compared to both smaller (14 nm) and larger (74 nm) particles (Chithrani et al., 2006). Many manufactured NPs exhibit changes in structure and reactivity as a function of size (atomic clusters–NP–bulk particles). These reactivity changes are often due to shape, surface defects, crystal edges and corner, porosity, proportion of surface

Analysis and Characterization of Manufactured NPs in Aquatic Environments

Agglomeregation State

Concentration

215

Shape

Surface Speciation Size

Surface Charge

+ + + + +

Surface Functionality

Porosity / Surface Area

Size Distribution

Composition

Structure / Crystallinity

Figure 6.1 The important properties of manufactured nanoparticles in aqueous media are shown, indicating that the central concept of a homogeneous solid sphere with a clean surface is often an over-simplification. All or several of these properties are needed to understand the fate and behaviour of these nanoparticles in the environment or to characterize a certain ecotoxicology experiment. Therefore, a combination of analytical methods is required to obtain a complete characterization. (Figure partly adopted from Tinke et al., 2006.) (See colour plate section for a colour representation)

sites (Chapter 3 (this volume); Waychunas et al., 2005; Wigginton et al., 2007a; Madden and Hochella, 2005), and therefore it is essential to be able to characterize these properties as a function of size. Particle size is the most common property used to describe a particle but it is only for perfect, compact spheres that size is a trivial descriptor (only one value required). For irregularly shaped particles equivalent spherical diameters (ESD) have to be used (Jennings and Parslow, 1988) (Table 6.2). Different particle ESD could be of varying significance in different process studies. More importantly, the different methods discussed later in this section apply to specific ESD. Therefore, such methods can not provide directly comparable numbers. There is an extensive literature available on the fundamental aspects of particle size analysis (Barth and Flippen, 1995; Webb, 2008) and, although they cover only larger particles (∼0.1 µm and above), much can be learned from these and applied also to NPs.

Table 6.1 List of reviewed methods and the metrics (properties) that they are capable of delivering and to which extent. Symbols are taken from Figure 6.1. Method

Size (nm)

PSD capabilityA

ShapeB capability

Agglomeration state capabilityC

Concentr. rangeD

AFM

ppb – ppm

BET Centrifugation

powder det. dep.

Dialysis

det. Dep.

DLS

ppm

Electrophor.

ppm

EELS/EDX ESEM

ppm in sp ppb – ppm

Filtration

detect. dep.

Flow FFFSed FFF

UV: ppm, ICPMS: ppb

HDC

det. dep.

ICP-MS LIBD

ppt – ppb ppt

NTA

ppb – ppm

SEC

det dep

SEM

ppb – ppm

SLS

ppm

SAED Spectrometry

ppb – ppm

TEM

ppb – ppm

Turbidimetry

ppb – ppm

Ultrafiltration

det. dep.

XPS XRD

powder powder

A

Symbols indicates wether the method could only determine narrow PSD or also wide or multimodal particle size distributions (PSD). AFM has a shape capability but the lateral dimensions, especially in the scanning direction, are overestimated, which can lead to erroneous shapes for smaller nanoparticles. C Some methods capable of determining agglomeration state (e.g. TEM) also have a high degree of perturbation, so results should therefore be interpreted with caution. D The approximate concentration ranges applies to requirements for quantitative analysis of the complete sample (not only single particle analysis). B

Table 6.1 (continued) Method

Surface Chemistry/ Charge/Area

Structure/ Crystallinity

Single part./ population

Dynamics capabilityE

Level of perturbation

sp

medium

pp

high

Centrifugation

pp

low

Dialysis DLS

pp pp

low minimum

pp

minimum

EELS/EDX ESEM

sp sp

high medium

Filtration

pp

low-medium

Flow FFFSed FFF

pp

low

HDC

pp

low

ICP-MS LIBD

pp sp

destructive minimum

NTA

sp

minimum

SEC

pp

medium

SEM

sp

high

SLS

pp

minimum

SAED

sp

high

Spectrometry

pp

minimum

TEM

sp

high

Turbidimetry

pp

minimum

Ultrafiltration XPS

pp pp

medium high

XRD

pp

high

AFM +

+

+ +

+

SSA

BET

Electrophor. +

+

+ +

+

E The dynamics capability is either left blank if not possible to follow kinetics faster than time scale hours and a slow curve symbol for time scale ∼minutes-hour and fast decaying symbol for timescale seconds. PSD: particle size distribution det dep: detection dependent sp: single particle method pp: particle population method

Table 6.2

Different equivalent sizes measured by different methods.

Irregular particle

Equivalent spherical size measure

Applies to methodf

Hydrodynamic diametera

Flow-FFF, DLS, NTA

Equivalent spherical volume diameter

Sed-FFF (for known density), LIBD, Electrozone sensing, Centrifugation

Buoyant Mass Equivalent spherical mass diameter Equivalent spherical max length diameter

Sed-FFF ∝∆δ*V MSb Microscopyc

Equivalent spherical min length diameter

Equivalent spherical projected area diameter

This example value: 4

a

Equivalent spherical specific surface area diameter

BET with known density

Equivalent spherical pore size diameter

Particle filtrationd

Root mean square radius of gyratione

SLS

Aspect ratio: the longest dimension divided by the shortest

Microscopy, Combination of light scattering methods or different FFF methods

Calculated from the measured diffusion coefficient, using Stokes–Einstein equation To calculate size from mass assume a certain structure Can be derived manually or by image analysis softwares d Filter pore size often defined as maximum size (not average size) that penetrates filter e Mean square distances from center of mass of point masses within the particle f For method abbreviations, see text. b c

Analysis and Characterization of Manufactured NPs in Aquatic Environments

219

In addition to the fact that certain methods measure certain types of equivalent particle size measures (Table 6.2), these different methods also yield different types of averages or distributions. The three most important types of size averages are number average, mass/volume average and Z-average. The applicable methods and equations for each are shown in Table 6.3. Consequently, when comparing size measurements from different methods it is important both to keep in mind which type of equivalent diameters the methods report and also which type of averages apply. It has been pointed out that a particle size distribution is often not a material property but represents a temporary steady state of dispersive and agglomerative processes. If the analysis is describing the distribution of physical (hard sphere) sizes, and not agglomerate sizes, it is therefore better to use the term ‘primary’ particle size distribution. 6.2.1.2

Shape

Nanoparticle shapes can be kinetically controlled during synthesis with the aid of selective adhesion of capping agents to the different crystal facets (Yin and Alivisatos, 2005) in order to obtain certain reactivity properties. It is well known that microparticles with high aspect ratio (e.g. asbestos fibres) are highly inflammatory. Recently it was shown that long multi-walled carbon nanotubes resulted in asbestos-like, length dependent, pathogenic effects when introduced in mesothelial lining of mice body cavities (Poland et al., 2008). Silver nanoparticle shape influences the hazardous nature of the materials for bacteria (Pal et al., 2007), and for gold NPs rod-shaped particles (14 × 74 nm) had significantly lower uptake rates than spherical NPs of both 14 and 74 nm sizes (Chithrani et al., 2006). The authors state that the surface chemical difference of the capping agents between the spherical and rod-shaped particles may be part of the explanation. However, the lower

Table 6.3 Description of different types of size averages, with equations defining them and methods that are deriving such average sizes. Type of size average

Applies to method

Number average: size average of numbers of particles within a certain size class

Microscopy, LIBD, NTA

Equation

∑ n ∗d ∑n i

dn =

i

i

i

i

Mass or volume average: size average of volume of particles within a certain size class

FFF and SEC with most detection methods, CFF

∑V ∗d ∑V i

dv =

i

i

i

i

Z-average size: an intensity weighted average attributed to certain methods

Dynamic Light Scattering dn

∑ n ∗d = ∑ n ∗d i

6 i

i

5 i

i

i

220

Environmental and Human Health Impacts of Nanotechnology

aspect ratio rod has a greater uptake rate than a higher aspect ratio, therefore shape plays an important role. Many concepts defining and measuring shape factors and aspect ratios that are currently applied have been developed for larger particles and can be adopted for NPs. However, the methods to determine the properties are different (Xu et al., 2003; Barreiros et al., 2006; Jennings and Parslow, 1988). See also the discussion on image analysis and fractal dimension determinations in Section 6.2.1.4. 6.2.1.3

Number and Mass Concentrations

For ordinary chemicals the conversion from a mass concentration (e.g. mg l−1 or ppm) to a number based concentration (mol l−1) is simply a factor of the molar mass of the compound, but for particles an analogue conversion is only straight forward for perfect spheres of known size. Since many NPs are often heterogeneous in terms of the properties shown in Figure 6.1, this adds a certain level of uncertainty. Due to the small mass of individual NPs (or high numbers per mass), is it often stated that particle number concentration is more relevant than particle mass concentration (Aitken et al., 2007; Oberdorster et al., 2005). Consequently, if it is hypothesised that there is a size and shape dependence on uptake (mechanistic explanation not yet fully understood) and that the number of a certain NP is important, number based size distribution and shape are important metrics in characterizing uptake. In addition, it has been claimed that it is the total particle surface area that can be correlated with a toxic effect (Section 6.2.1.7). It should be noted, though, that although these metrics (particle numbers, mass, diameter and area) are geometrically dependent, for a heterogeneous system it is important to measure them independently in order to fully characterize the exposure conditions. Teeguarden et al. (2007) discuss that, especially for in vitro cell culture studies, it is insufficient to quantify the overall physico-chemical NP characteristics. The doses delivered to the specific sites of action must be considered, a challenging task to say the least. 6.2.1.4

Dispersion State and Agglomeration

The environmental transport, biological uptake and, to some extent, reactivity of NPs will be largely controlled by their dispersion and agglomeration behaviour. Nanoparticles dispersed in water are a colloidal system that is affected by weak physical forces of both attractive and repulsive character. If the particles are allowed to come into close proximity to each other then attractive van der Waals’ forces will most likely lead to agglomeration (O’Melia, 1987). But if the NPs are stabilized by either repulsive electrostatic forces or by means of steric stabilization, usually by hydrophilic polymer tails on the particles (Lourenco et al., 1996), then the system can be kinetically stable (O’Melia, 1990). The electrostatic stabilization can occur on particles with both a net negative or positive charge depending on the pH of the water and the point of zero net charge of the NP. Due to protonation of surface functional groups and sorption of natural organic matter (NOM), there is a close coupling between the particle surface chemistry, the water chemistry and the stability and agglomeration state.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

221

The degree of agglomeration, which often shifts the particle size distribution from the nanometre to the micrometre size range probably affects biological uptake. Agglomeration reduces the pool of particles that can be taken up by nonspecific diffusion (below a few nanometres) (Kloepfer et al., 2005) and perhaps endocytosis, and increases the pool of ingestible material that can be taken up through the gut. Agglomeration also changes the diffusion rates of the particles and hence decreases the transport rate of nanoparticle through the diffusive boundary layer and mucus to the organism surface. It is also important to determine agglomerate structure, in addition to agglomerate size. A porous, loosely assembled agglomerate has a lower effective density and a proportionally higher hydrodynamic friction drag than a more dense, compact agglomerate. Agglomerate structure is often an indicator of whether the agglomerate growth rate has been reaction limited (slow agglomeration) or diffusion limited (fast agglomeration). Reaction limited agglomeration produces loose, open structured agglomerates and diffusion limited agglomeration produces spherical, compact agglomerates. A determination of the fractal dimension is an objective method for quantifying agglomerate structure (Rizzi et al., 2004; Sterling et al., 2005; Limbach et al., 2005). Baalousha et al. (2008) used a fractal dimension analysis of TEM images to characterize iron oxide aggregates. Agglomeration may also partly make the nanoparticle surface less available for chemical reactions for several reasons. As dense agglomerates form, fractions of the particles merge together and the total exposed surface area of the sample decreases The diffusion of chemical reactants or products may be slower through the three dimensional structure of an agglomerate, which may impede catalytic reactions and, moreover, there may also be surface area less available for light (photocatalysis). 6.2.1.5

Structure and Crystallinity

Even though the structure activity relationships for NPs are still poorly studied, from both reactivity and toxicology points of view, there is evidence that crystallinity and crystal structure are important. For example, for titanium dioxide it is anatase, one of the crystalline polymorph structures, that has shown to be photoactive, while the others, rutile and brookite, are much less active, or not active at all (Augustynski, 1993). To complicate the picture it has been shown that a mixed phase of anatase and rutile is more photoactive than pure anatase, and it has been suggested that the interfaces between the two crystallinities are important for the formation of photoactive hotspots (Li et al., 2007) In toxicology an example is that crystalline quartz shows pulmonary toxic effects, while amorphous silica is much less harmful (Castranova, 2000). 6.2.1.6

Chemical Composition

The chemical composition of NPs is, of course, an important property in characterizing unknown materials or tracking the particles in certain experiments. It could either include an elemental analysis or chemical composition (e.g. distinguishing C60 and C70 fullerenes, or polystyrene and polyacrylate NPs). The elemental

222

Environmental and Human Health Impacts of Nanotechnology

composition may have a direct influence on the toxicity potential of the material. For example, certain metal-containing materials are made of heavy metals that are toxic at low concentrations of their dissolved ions. This has lead to the importance of measuring the dissolution rates of NPs (Section 6.2.1.11). But the chemical composition of NPs may not necessarily be homogeneously distributed, either in the particle population or even within each particle. Certain NPs (e.g. quantum dots) have different chemical composition in the core, shell and surface layer. The variations in chemical composition can lead to drastically different reactivities, so analysis of both the whole particle and the different parts may be important but difficult. Certain NPs, for example carbon nanotubes, have high levels of catalyst metal contamination which needs to be quantified. 6.2.1.7

Surface Area

It is often stated that NPs are important as catalysts because they have a much higher specific surface area (SSA) than their larger counterparts, and this is of course true, but not the full picture (Chapter 3). The importance of surface area for chemical reactivity can probably be translated to explain some of the nanotoxicology effects too. Still only a few studies exist that test this hypothesis. One example showed a correlation of human lung toxicity to SSA of the NP (Donaldson et al., 2004). 6.2.1.8

Surface Charge

When NPs are dispersed in water they rapidly interact with the protons and hydroxyl ions of the water and gain or lose protons depending on the nature of the material. This proton exchange leads to charged groups on the nanoparticle surface. The charged groups will attract oppositely charged ions (counter ions) that will associate strongly close to the surface, and these counter ions will dominate over other ions further away (∼3–100 nm from surface) in the so-called diffuse layer or electrical double layer (EDL) of the NP. The thickness of this layer is called the Debye length and is inversely proportional to the ionic strength of the water. Nanoparticles of the same surface charge will repel each other. When this happens, NPs are electrostatically stabilized. When ionic strength increases the Debye length decreases and the effective charge in the EDL decreases. This brings the nanoparticles into sufficiently close contact with each other that attractive van der Waals’ forces become dominant and aggregation occurs. The most appropriate metric when studying the specific binding of metals or other solutes to the surface is the surface proton charge (by acid–base titrations). But when studying nanoparticle stability and agglomeration it is more relevant to measure the charge in the EDL. This charge or potential at the hydrodynamic slipping plane is called zeta potential (Section 6.2.5.7). In environmental transport studies, surface charge is a key parameter (Tiller and O’Melia, 1993; Guzman et al., 2006). It has been shown that the coating of natural organic matter (especially humic substances) provides all surfaces with a negative surface charge (Hunter and Liss, 1979; Beckett and Lee, 1990; Ledin et al., 1993) that makes them more immune to agglomeration due to electrostatic stabilization

Analysis and Characterization of Manufactured NPs in Aquatic Environments

223

(Sholkowitz et al., 1978; Behrens et al., 1998; Wilkinson et al., 1997). Similar results have also been shown for manufactured NPs and NOM (Hyung et al., 2007). 6.2.1.9

Surface Functionality

The relative importance of nanoparticle surface functionality and nanoparticle core properties still needs to be researched but it is probably safe to say that surface functionality is a major factor in fate, behaviour and effects. In many nanoparticle applications the nanoparticle growth is hindered by a capping agent, which is often a polymer. The same polymer is sometimes used to provide the NP steric stabilization (Lourenco et al., 1996; Heijman and Stein, 1995). But there may also be a charge stabilization functionalization. Natural examples have been discussed in Section 6.2.1.8. Other types of NOM, mainly polysaccharides, can induce agglomeration due to a bridging flocculation mechanism. Both natural and synthetic organic functionalization can control the shape of NP formed (Banfield and Navrotsky, 2001; Yin and Alivisatos, 2005). In addition, it is known that quantum dot skin permeability changes with different surface functionalization (Ryman-Rasmussen et al., 2006). Further, a study of phytotoxicity of alumina NPs showed that the presence of organic coatings of certain organic compounds known to be free hydroxyl radical scavengers decreased the growth inhibition (Yang and Watts, 2005). 6.2.1.10

Surface Speciation

The surfaces of metal NPs often oxidize (corrode) in aqueous solutions. Consequently, these metal NPs may behave very similar to their pure metal oxide counterparts. The oxidation state of the material on the surface is often termed surface speciation. An example is the presence of Ag+ ions on the surface of a metal Ag(0) nanoparticle. The surface speciation is important for both reactivity (Stumm, 1993; Waychunas, 2001) and (eco)toxicology (Chapters 7 and 9). 6.2.1.11

Dissolution Rates and Desorption of Trace Constituents

Since many of the inorganic manufactured NPs contain heavy metals that are known to be toxic in their dissolved form, it is essential to determine the dissolution of metals from these NPs (Limbach et al., 2005; Borm et al., 2006; Franklin et al., 2007). These include, for example, metallic, metal oxide, metal sulfide NPs. For example, it has been reported that a significant part of the observed toxic effect of silver NPs is due to the dissolution of silver ions from the particles (Lok et al., 2006). In addition, it has been shown that desorption of contaminant metals or doping metals from, for example, fullerenes or carbon nanotubes is responsible for the generation of reactive oxygen species (ROS) once the nanoparticle has entered the cell (Limbach et al., 2005). Free metal ions can also be quenched in the ecotoxicology experiment by using chelating ligands, either free (e.g. EDTA for transition metals or cystein for silver) or matrix bound (e.g. ChelexTM resin). However, the use of chelators must be done with some caution, since these also bind essential di- and trivalent elements and, moreover, enhance the nanoparticle dissolution rates.

224

6.2.2

Environmental and Human Health Impacts of Nanotechnology

Sampling, NP Extraction, Sample Preparations

Environmental colloidal systems are dynamic and consist of continuous particle formation and dissolution, removal and adsorption of organic matter and microbial degradation (Chen and Buffle, 1996b). It is well known that natural NPs in the colloidal state are easily disturbed in water, groundwater, soil and sediment during sampling, handling and storage (Backhus et al., 1993; Puls, 1990; Citeau et al., 2006). Changes can include agglomeration, microbial growth, degradation or adhesion to containers. It is, therefore, desirable to use in situ techniques that remove the sampling step (Ledin et al., 1994; Kim and Walther, 2007; Maldiney and Mouchel, 1995), but there are almost none available that can provide more than very basic information. Instead, efforts have been put into sampling of the colloids as gently as possible, minimizing handling and storage to avoid perturbations (Chen and Buffle, 1996a; Lead et al., 1997; Chanudet and Filella, 2006). These procedures optimized for natural NPs should be suitable also for manufactured NPs in the same matrices. Most of the analytical techniques being discussed in this chapter are applicable to dispersed NPs. Although free NPs and suspended aggregates are the most important fractions in many applications, deposited or adhered NPs are also important in some matrices (e.g. soil and sediments). Therefore, it is important to consider methods to extract or detach NPs from solid matrices, as well as methods to characterize the attached NPs. Such detachment methods can include addition of dispersion agents (e.g. sodium pyrophosphate) and sonication prior to analysis. The same considerations also apply to the interaction of manufactured NPs as they do with the interactions of natural NPs, whether the manufactured NPs should be extracted/detached from natural materials or analysed as present. Sonication is often used to deliver energy to an agglomerated system to break apart reversibly agglomerated particles. It does not provide stabilization, but only the deagglomeration energy. Similarly, dispersion agents alone are usually not sufficient to de-agglomerate and stabilize agglomerated NPs. A drawback of sonication as a dispersion tool is that sonication increases drastically the collision rate between NPs and can thus lead to an induced agglomeration instead, depending on the particles, media and sonication conditions. Fullerenes can be chemically extracted with solvents (Fortner et al., 2005; Nowack and Bucheli, 2007). In hazard assessment of nanopowders it is important to take into account the representative sampling of the powders (Powers et al., 2007). 6.2.3

Light Scattering Methods

A coherent electromagnetic wave of laser light will interact with matter and induce an oscillating electric dipole in a particle. These induced dipoles will in turn re-radiate light. This is the basic principle of light scattering. Light scattering phenomena can be used in several different characterization techniques to determine different size related properties and also concentration. In addition to light scattering, interaction of light with the particles can also lead to absorption, fluorescence, refraction and diffraction, which are partly covered in the spectroscopy section.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

6.2.3.1

225

Dynamic Light Scattering

In dynamic light scattering (DLS), also called photon correlation spectroscopy or quasielastic light scattering, the fluctuations in the intensity of the scattered light, due to constructive or destructive interferences from point scatterers in the focal volume, are correlated to the diffusive motion of the NP. The light intensity is measured at very short time increments (nano- to microseconds) and autocorrelated to the initial reading. For the initial intensity readings the correlation is unity since the particles have hardly had any time to diffuse, but as time progress and particles move the intensity correlation decrease (Figure 6.2). The decay of the autocorrelation function is proportional to the diffusion coefficient, D, according to Equation (6.1), where A is an instrument constant and q is the scattering form factor and t is the time: g (2)( t ) − 1 ≈ A exp ( −2q2 Dt )

(6.1)

1.0

0.9

Correlation function (g(2) -1)

0.8

g ( 2 ) (t )− 1 ≈ Ae ( −2q

0.7

2

Dt )

0.6

0.5

0.4

0.3

0.2

0.1

0.0 0.1

1

10

100

1000

10000

100000

1000000

10000000

time (µs) Figure 6.2 Dynamic light scattering autocorrelation function over the decay time for a CuO nanopowder dispersed in distilled water. Since Equation (6.1) holds for small t, then the cumulant analysis to extract the z-average D or dH, Equation (6.2), is done in the first part of the autocorrelation function (dotted square). The obtained z-average diameter was 190 nm. The hump in the autocorrelation with a decay on much longer time scales indicate very large aggregates or dust contamination.

226

Environmental and Human Health Impacts of Nanotechnology

For fast diffusing particles, the light intensity loses its autocorrelation rapidly (microseconds), while for larger slow diffusing particles the intensity can be correlated to the initial for substantial timescales (many milliseconds). If the NPs are not affected by their adjacent neighbours (particle–particle interactions), then the scattered light fluctuations can be correlated to the self-diffusion of the particles (their diffusion coefficient) (Schurtenberger and Newman, 1993). For such dilute systems, the hydrodynamic diameter, dH, of a particle can be derived from the Stokes–Einstein relationship: dH =

kBT 3πηD

(6.2)

where D is the diffusion coefficient, η is the viscosity of the medium and kB is Boltzmann’s constant. But the scattering efficiency of the light varies as the sixth power of dH (or proportional to d 6H ) for particles smaller than one twentieth of the wavelength and as the square of dH (or proportional to d 2H ) for larger particles. This size dependence of the scattered light intensity will skew the result towards the larger particles in the measurement. DLS consequently delivers an intensityweighted correlation function, which is usually converted to an intensity weighted (z-average) diffusion coefficient (Table 6.3). If the size distribution of the sample contains several particle types or a broad size distribution, the deconvolution of several diffusion coefficients from the autocorrelation from such a sample is an ill posed mathematical problem and the obtained results are usually not very robust (Filella et al., 1997; Finsy, 1994; Schurtenberger and Newman, 1993). The advantage of DLS is rapid analysis time, along with simple operation, which can even be used in the field (Ledin et al., 1994), and is suitable to qualitatively monitor agglomeration (Viguie et al., 2007). The drawbacks are the complicated data interpretation for polydisperse samples and the difficulty of using intensity weighted results for any size measurements that are not very monodisperse samples (Filella et al., 1997). The conversion from diffusion coefficient to hydrodynamic diameter also involves a spherical particle assumption. 6.2.3.2

Static (Classical) Light Scattering

For particles much smaller than the wavelength (d < λ/20) of the incident light, the scattering intensity in the plane perpendicular to the polarization of the laser light is equal in all directions (the particles are isotropic scatterers). In the size range (λ/20 < d < λ) destructive interferences on all angles but forward favour forward scattering intensity, and the larger particles in this size range have the higher forward scattering intensity. When the particle sizes are in a similar size range as the wavelength of the laser light, then the electromagnetic interactions of the light with the matter within one particle will cause constructive or destructive light at certain angles which will be dependent on the particle size. When particle sizes are similar to or larger than the wavelength, the scattering angular dependency becomes complex with maxima and minima at certain angles, which is described by the Mie theory (van de Hulst, 1981). For these size ranges there is consequently an angular dependency of the scattered light. These phenomena are used in static light

Analysis and Characterization of Manufactured NPs in Aquatic Environments

227

scattering, where size related properties are derived from the intensity variations with detection angle (Schurtenberger and Newman, 1993; Wyatt, 1993). The full light scattering theory is quite complex but simplifications based on certain approximations have been developed. In its simplest forms, according to the Rayleigh–Gans–Debye and Guinier approximations, where it is assumed that the particles are much smaller than the incident lights wavelength, the refractive index is similar to that of the solvent and no light absorption occurs. Light scattering can be formulated as: 1 K *c ≈ + 2 A2 c Rθ Mw P (θ ) P (θ ) ≈ 1 − θ →0

16π n02 2 rg sin 2(θ 2) 3λ02

(6.3) (6.4)

where K* is the material constant ( 4π ( dn dc ) n 02 Na λ 04 ), Rθ the excess Rayleigh ratio (scattered intensity in excess of the scattering from pure solvent), Mw the molar mass of the particle/polymer, A2 the second virial coefficient, P(θ) the particle form factor, n0 the refractive index of the solvent, dn/dc the refractive index increment of the particles, λ0 the vacuum laser wavelength and θ the scattering angle (Schurtenberger and Newman, 1993;Wyatt, 1993). On plotting K*c/Rθ against sin2(θ/2) in a double extrapolation plot (or Zimmplot) the intercept yields molecular weight (Mw) at the zero concentration (c) and zero angle, and from the two slopes the root mean square radius of gyration rg2 and the second viral coefficient can be derived. The molar mass and second viral coefficient are mainly relevant for studies of polymers or proteins while rg2 is mainly relevant for NPs. rg2 describes how the mass of a particle is distributed from the centre of mass (Table 6.2). The sensitivity of light scattering is dependent on laser wavelength, particle refractive index increment and particle size. The sensitivity is inversely proportional to λ4, and inversely proportional to the d2 for larger particles and d6 for particles much smaller than the laser wavelength (Filella et al., 1997; Schurtenberger and Newman, 1993). This strong size dependency has major implication for characterization of NPs, since the detection limit increases to above the mg l−1 level for NPs below ∼50 nm. This means that the light scattering method has to be used in particle concentrations that may not be environmentally relevant. Even more important is that the strong sensitivity dependence on particle size means that a few larger components in the sample often completely skew the measured size. 2

6.2.3.3

Laser Diffraction

Laser diffraction methods build on the analysis of the diffraction patterns generated when a laser beam is interfered with by a particle (Figure 6.3). The understanding of laser diffraction is explained by the Mie theory, and mainly applicable to micron size particles (the Mie size region above the laser wavelength), but some manufacturers have incorporated backscatter detectors and multiple lasers in order to cover the submicron size range. Thus, the technique can be used in the size range

228

Environmental and Human Health Impacts of Nanotechnology

Figure 6.3 Principles of a laser diffraction instruments equipped also with a lower wavelength laser and additional forward and backscattering detectors to decrease the lower size limit of the technique. (Reprinted with permission from Malvern Instruments Ltd, UK.)

of ∼0.05–1000 µm, and it is thus very suitable to study flocculation processes. Laser diffraction measures an equivalent spherical volume diameter (Table 6.2). The major limitation of laser diffraction for NPs or their agglomerates is the poor sensitivity and the need for high detection angles for submicron particles. 6.2.3.4 Turbidimetry Turbidimetry (light transmission measurement) or nephelometry (scattering intensity measurement, typically at a right angle) can be used to measure particle concentration in a sample (Irache et al., 1993). Due to the simplicity of measurement and inexpensive instrumentation turbidimetry is often used in environmental studies and water quality monitoring (Peng et al., 2002). However, due to the reasons discussed in Section 6.2.3.2, the scattering intensity depends not only on particle concentration but also on the size, composition and shape. Consequently, for anything else than very well defined NPs turbidimetry should be used with caution. 6.2.3.5

Nanoparticle Tracking Analysis

Nanoparticle tracking analysis (NTA) is a recently developed method using tracking of the Brownian movement of individual point-scatterers illuminated by a laser in a flow cell under a conventional optical microscope. The detection is recorded in the microscope by a CCD camera as a high speed movie and the mean squared displacement between the frames in the sequence is determined for each individual nanoparticle trajectory. Even though diffusive movement takes place in three dimensions and the instrument only captures movement in two dimensions, there is a correction term to the Stokes–Einstein equation to be able to calculate diffusion coefficients or hydrodynamic diameter. Since the method relies on counting and measuring particles, the obtained size distributions are number based, compared with the related DLS technique that is

Analysis and Characterization of Manufactured NPs in Aquatic Environments

229

intensity weighted. The advantage of NTA compared with DLS is that the intensity of the scattered light should not bias the result. Consequently, some aggregates or larger particles are tracked and counted in a similar way as the small particles without skewing the results. The detection limit of the method is determined by the laser, optics and relative refractive index, and particle size. Currently available commercial instruments can detect nanoparticles as small as ∼10 nm for high refractive index nanoparticles such as silver but only down to ∼25 nm particles for titanium dioxide. Improving the optics with a lower noise CCD camera or lower laser wavelength can improve these lower limits of the detectable size window. The detection limits in terms of concentration (number of particles ml−1) are difficult to provide absolute numbers for, since the data acquisition (tracking) can be carried out for varying amount of time (typically 15–60 s) to compensate for low particle concentration. The optimum sample concentration is typically 107–109 particles ml−1. 6.2.4

Other Electromagnetic Scattering Methods

To overcome some of the limitations of light scattering, analogous methods with different wavelengths of electromagnetic radiation can be used. These generally are limited to large scale facilities to obtain the high photon fluxes needed for both X-ray and neutron experiments. In this chapter only a few of these techniques are mentioned. 6.2.4.1

Small Angle X-ray Scattering

Small angle X-ray scattering (SAXS) uses similar principles to static or dynamic light scattering but the particle sizes that can be probed are much lower (∼1–50 nm) than for these due to the lower wavelength of X-ray light (Waychunas et al., 2005; Megens et al., 1997; Mori et al., 2006). There are laboratory versions available but these generally provide too low X-ray intensity for characterization of nanoparticles. Generally, to obtain enough intensity of a narrow focused X-ray, it is necessary to perform the experiment at a synchrotron source (Megens et al., 1997). 6.2.4.2

Small Angle Neutron Scattering

In small angle neutron scattering experiments (SANS) the neutron beam (typical wavelength 0.01–3 nm) is excellent to interact with matter at the nanometer scale. While light and X-rays are scattered by the electrons, neutrons are scattered by the atom nucleus. In contrast to X-ray scattering, there is no sensitivity trend with increasing atomic number, and the neutron refractive index may differ even between isotopes of the same atom. The difference is most remarkable for hydrogen and deuterium, and this is often used in experiments where contrast matching is used to selectively probe the interactions between NPs, or between NPs and surfactants or other coatings (King, 1999). SANS has been used for environmental studies to some extent but there is great potential for future work (Diallo et al., 2005). One application where SANS may be important is to study NP surfactant interactions.

230

Environmental and Human Health Impacts of Nanotechnology

6.2.5

Fractionation and Separation Methods

Fractionation is defined by IUPAC to be classification according to physical, for example size or surface area, or chemical properties, for example hydrophobicity or chemical partitioning. In this section, discussion is limited to physical fractionation methods. 6.2.5.1

Microfiltration

Filtration methods are the most simple and common fractionation methods. They are distinguished based on the pore size or molecular weight cut-off (MWCO) of the membranes being used: for microfiltration 0.2–10 µm; ultrafiltration 1 kDa to 100 MDa; and for nanofiltration or reverse osmosis approximately 100–1000 Da. Common piston filtration, or frontal filtration, using microfiltration membranes (0.2–1 µm) is the most simple and common method, but filtration artefacts that often affect the results have been reported (Buffle et al., 1992; Morrison and Benoit, 2001). In piston filtration, particle concentration gradients are formed above the filter and this results in concentration polarization and pore clogging. This can lead to agglomeration and increased deposition of both oversized and undersized particles on the filter membranes. The presence of charged natural organic matter, for example humic substances, often decreases these artefacts since they stabilize the particles and also condition the membrane and provide a charge stabilization effect. Nanoparticles without significant stabilization, either natural or synthetic, are prone to much larger filtration artefacts. It has been observed that 100 nm CuO NP in distilled water can be effectively removed by a 1 µm filtration due to these effects (Hassellöv, unpublished data). There are many filter membrane types and pore sizes available, for example sieve-type filters with well defined pore size but very small porosity and depth filters with less well defined pores, which allow a larger amount of sample to be filtered before clogging artefacts start (Figure 6.4). Sieve-type filters are more suitable for analytical filtration but care has to be taken to avoid these artefacts (Morrison and Benoit, 2001). 6.2.5.2

Ultrafiltration

As been mentioned, ultrafiltration is based on membranes with very small pores. Ultrafiltration can also be used in a piston filtration mode in pressurised cells sometimes with stirring to reduce the concentration polarization. However, a more common method to reduce concentration polarization and colloid cake formation is to use a cross-flow in a tangential direction to the membrane in a recirculating fashion. The principle is called cross-flow filtration and can be used also for microfiltration but is mainly used for ultrafiltration (CFUF). The sample is typically pumped in a recirculating fashion repeatedly over the membrane and in each cycle a certain part of the sample permeates the membrane (Figure 6.5). Ultrafiltration membranes are given a nominal molecular weight cut-off (MWCO) based on typically 90% rejection of standard molecules, for example proteins. The most commonly used MWCO is 1 kDa, which is often used to operationally define the border

Analysis and Characterization of Manufactured NPs in Aquatic Environments

A

B

polyvinylidene fluoride

231

C

polysulfone membrane

cellulose membrane

E

D

track-etched membrane

honeycombe AlOx 20 nm

Figure 6.4 Detailed SEM pictures of different microparticle filter types. Upper row is showing different depth filters with high porosity, while the lower row is showing sieve type filters with well defined pore structure. (Images A–D reproduced with permission from Millipore Corporation. Image E kindly provided by Frank von der Kammer (2005), Characterization of Environmental Colloids applying Field-Flow Fractionation – Multi Detection Analysis with Emphasis on Light Scattering Techniques, Hamburg University of Technology.)

Recirc. pump

CFF-membrane Sample Feed pump

Permeate Retentate reservoir

Figure 6.5 A typical cross-flow filtration setup with a sample vessel, a recirculating retentate reservoir, pumps, ultrafiltration membrane and permeate vessel. Inset microscopy picture shows a surface of an ultrafiltration membrane with larger supporting pore structure underneath.

232

Environmental and Human Health Impacts of Nanotechnology

between truly dissolved solutes and colloids (or NPs). CFUF is also known to be prone to artefacts if the procedure and membrane selection are not optimized carefully. There are several studies that have examined both experimental setup and operating conditions to obtain good recovery and as precise a cut-off as possible in both seawater and freshwater (Guo et al., 2000; Larsson et al., 2002; Morrison and Benoit, 2004; Doucet et al., 2005). Ultrafiltration can not only be used to fractionate the samples, but also to obtain a concentration factor (1–100) but this increased concentration of natural colloids does change their physicochemical properties (e.g. agglomeration)(Liu and Lead, 2006). 6.2.5.3

Centrifugation

Centrifugation can be used as a pre-fractionation method or as the main nanoparticle fractionation method (ultracentrifugation). Centrifugation can induce agglomeration between particles due to differential settling velocities, but it has been shown to cause less perturbations than filtration for soil colloids (Gimbert et al., 2005). From the settling in the centrifuge gravitational field the fractionation of particles with a certain equivalent volumetric diameter (dV) can be calculated, providing that the particle density (ρpart) is known, using the following equation: 2 ( ρ part − ρ ) x (π v ) 2 ( t − t 0 ) ln   = dV2  x0  9η

(6.5)

where x0 and x are the starting and ending position distances from the centrifuge centre at time t0 and t, ν is the centrifuge rotational speed and η is the viscosity of the medium. Centrifugation can also be used to study dynamics of agglomeration and phase separation in analytical centrifugation with an integrated photometer by acquiring the particle mass profile in the centrifuge tube as a function of time (Lerche, 2002; Colfen et al., 2003) (Figure 6.6). 6.2.5.4

Field-Flow Fractionation

Field-Flow Fractionation (FFF) is a family of separation techniques for colloids, polymers and NPs that uses the combination of distribution of particles in some kind of field and a laminar flow in thin channels. The different FFF subtechniques are distinguished by the type of field, where the two most common are a second hydrodynamic field (Flow FFF) and a centrifugal field (Sedimentation FFF). The field forces the particles against one of the walls in the channel and the particle mean distribution is governed by a balance between the Brownian motion and on how the particles are affected by the field (Figure 6.7). The particle mean height is inversely proportional to the particle size and proportional to the transport velocity through the channel since the transport is faster close to the middle of the laminar flow (Giddings, 1993). Not only can FFF separate NPs but the elution time can be directly related to physical properties (diffusion coefficient or hydrodynamic diameter for Flow FFF and buoyant mass or equivalent spherical volume diameter for Sedimentation FFF) (Giddings, 1993)

Analysis and Characterization of Manufactured NPs in Aquatic Environments

233

NIR Light Source

6 ... 2300 g

Sample

Space

Time Colour Coded

Time

Transmission

∆t

Transmission Profiles CCD Sensor Radial Position

Figure 6.6 Analytical centrifuge showing the light source that is passed through the dispersed sample; the transmission of light is recorded as a function of time on a CCD sensor. Thus, settling rates of and settling profiles can be monitored at a higher gravity over time. (Reproduced with permission from L.U.M. GmbH.)

Field To Detector

Diffusion

Flow

Figure 6.7 The separating principle of Field-Flow Fractionation showing the laminar flow profile, and the steady state distribution of particles as a result of balanced transport by the field and concentration driven diffusion.

234

Environmental and Human Health Impacts of Nanotechnology

The retention ratio (R) in FFF is defined by: R=

t0 tr

(6.6)

where t0 is the void time and tr is the retention time. For highly retained components, R can be approximated by: R ≈ 6λ

(6.7)

while R can be estimated as follows for intermediate retention:

(

)

(6.8)

R = 6λ coth 1 2λ − 2λ 

The fundamental retention parameter (λ) is defined as the mean distance of the component from the wall (l) divided by the channel thickness (w):

λ=

l D = w Uw

(6.9)

where D is diffusion coefficient and U is the field induced force on the particles. Field-flow fractionation has been very valuable in environmental studies on natural colloids using both optical detectors (UV/Vis, Fluorescence and SLS/DLS) (v.d.Kammer et al., 2005) and ICP-MS (Beckett and Hart, 1993; Hassellöv et al., 2007). In addition, FFF has been applied for characterizing a wide range of manufactured NPs including silica, titania, metals, metal oxides, carbon black and carbon nanotubes (Schimpf et al., 2000; Siripinyanond and Barnes, 2002; Chen and Selegue, 2002; Fraunhofer et al., 2004; Gimbert et al., 2007; Moon et al., 2004). An example of FFF-ICPMS determination of size distribution in a copper(II) oxide nanopowder is shown in Figure 6.8. The obtained size distribution is a mass (or particle volume) distribution since the ICP-MS is analysing the mass concentration in each size fraction. Hydrodynamic diameter (nm)

ICPMS signal (cps)

0

100

200

300

400

40000

20000

500

1000

1500

2000

FFF Retention time (s)

Figure 6.8 An example of a FFF-ICPMS fractionation of engineered nanoparticles (CuO nanopowder), where the lower x axis is the raw data (retention time) and the upper is the converted size distribution. The y axis shows the copper signal from ICP-MS.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

235

In addition to size characterization, shape information has also been demonstrated for FFF in combination with microscopy or light scattering methods (Beckett et al., 1997; v.d.Kammer, 2005; Baalousha and Lead, 2007). The main advantages of FFF are the relatively mild fractionation, wide operating size range (1–800 nm for Flow FFF and 30–800 nm for Sedimentation FFF) and versatility in terms of the detector possibilities both on-line and off-line (Hassellöv and v.d.Kammer, 2005). The main limitation of FFF is that the versatility comes with a rather high complexity. The key to successful FFF analysis is proper optimization of the carrier composition for a new sample in order to minimize particle– particle and particle–wall interactions, while still allowing particles to approach the accumulation wall almost infinitely close. 6.2.5.5

Size Exclusion Chromatography

Size exclusion chromatography (SEC) is a size fractionation method where particles are separated through a column with a porous packing material that has a distribution of pores in the size range of particles to be fractionated (Barth and Boyes, 1992). The particles are separated based on to their hydrodynamic volume (size and shape) according to their ability to enter the porous structure of the packing materials. Larger particles enter pores to a lesser extent than smaller particles. Each SEC column has a certain operating size (or molar mass) window depending on its pore size distribution. The first eluting, oversized particles are those outside the operating window, then comes the fractionated particles and finally the ‘salt peak’ containing ions and molecules that have passed through the complete pore volume. SEC has been applied to carbon nanotubes and fullerenes and, as described in a later section, to natural organic and inorganic nanomaterials (Duesberg et al., 1998; Perminova et al., 2003; Vogl and Heumann, 1997; Jackson et al., 2005). 6.2.5.6

Hydrodynamic Chromatography

Hydrodynamic chromatography (HDC) is a size fractionation method where the separation takes place in narrow open capillaries, or in wider capillaries with nonporous packing materials that form capillary paths. Due to physical size restrictions, the centre of mass of a particle cannot come infinitely close to the walls, while a smaller particle can approach the wall closer than a larger one. Since the flow velocity is higher away from the wall, the elution order is the same as in SEC (larger particles first then small). The separation efficiency of HDC is very poor, but the operating size range is very good (∼5–300 nm). HDC has been successfully applied for the fractionation of NPs (Williams et al., 2002; Tiede, 2008). 6.2.5.7

Electrophoresis and Capillary Electrophoresis

Electrophoresis is a group of fractionation methods using the fact that charged particles will move in an electric field. The media can be a gel (Gel Electrophoresis or Capillary Gel Electrophoresis) or an electrolyte. The particles travel at a velocity (v), which is proportional to the applied electric field (E), and the coefficient is

236

Environmental and Human Health Impacts of Nanotechnology

defined as their electrophoretic mobility, µ = v/E. From measurements of electrophoretic mobility, the potential at the particles hydrodynamic slipping plane (or shear plane) called zeta potential (ζ), is calculated from:

ζ=

3µη 2ε f (κ r )

(6.10)

where η is the viscosity of the medium, ε is the dielectric constant and f(κa) is Henry’s function, which is simple and well defined for two extreme cases. Henry’s function is defined by the Smoluchowski approximation as 3/2 for situations where the diffuse layer thickness (Debye length, 1/κ) is much smaller than the particle radius (r), while it is defined by the Hűckel approximation as 1 for situations where the Debye length is larger than the particle radius. For commercial instruments either of the two approximations or customized values for Henry’s function can be selected. The Hűckel approximation is mainly valid in organic solvents or pure water. The Smoluchowski approximation is not valid for NPs in water with some ionic strength. Therefore, there have been several theories developed to more accurately parameterize the electrokinetics of small particles in ionic media, and these have been reviewed recently (Delgado et al., 2005). In addition to the old fashioned method of measuring particle velocity in an electrical field by means of a microscope and a stop-watch, modern instruments have been developed that use laser doppler velocimetry (e.g. light scattering phase analysis). Zeta potential measurements are very suitable to study dispersion properties and ionic strength and pH regions of electrostatic stabilization (Figure 6.9). As a rule of thumb a colloidal dispersion that has either a zeta potential that is less than −30 mV or above 30 mV can be considered electrostatically stable. For this particular case in Figure 6.9 that would apply to a pH below four or above 7.5. The maximum agglomeration usually occurs at the isoelectric point.

Zeta potential (mv)

60 Isoelectric point

40

Stable

20 Unstable

0 -20 -40

Stable

-60 2

4

6

8

10

12

pH

Figure 6.9 Illustration of a typical zeta potential pH titration curve. The pH where the zeta potential is zero is defined as the isoelectric point and is a characteristic of each nanomaterial.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

6.2.6

237

Microscopic Methods

Microscopic methods enable the investigation of physical and chemical properties at the level of individual particles as well as aggregates thereof. Due to the limited resolution of light microscopes, which is in the order of 0.1 µm (lambda/2 for uv light), individual NPs cannot be detected using conventional light microscopes. In the following discussion the focus is therefore on electron and atomic force microscopes. The extreme wealth of information that can be obtained on the individual particle level, however, is very time consuming. Thus, the number of particles that can be investigated is rather limited and also depends on the kind of information desired. Pure morphological information used to derive size distributions can be obtained by combination with image analysis tools. In this way, several microscopic images can be processed semi-automatically, resulting in several hundreds to thousands of particles that are evaluated (Mavrocordatos et al., 2000). If, however, chemical or crystallographic information is required, the number of particles that can be analysed within a reasonable time is reduced to a few tens. Microscopic investigations, therefore, are carried out on a very small subset of the particle ensemble; thus, they should be used in combination with other methods that can be performed on bulk samples and also deliver information on bulk properties of the sample. However, engineered NPs in the aqueous environment might occur only in very low amounts compared to the naturally occurring colloids, which questions the significance of data derived from bulk methods. On the other hand, it might be very challenging, if possible at all, to detect enough NPs within an overwhelming number of natural colloids of comparable size, to derived statistically meaningful results (Vigneau et al., 2000). 6.2.6.1

Electron Microscopy

Using electrons instead of light increases the resolution power of microscopes by several orders of magnitudes and enables the investigation of materials at the atomic level. There are two major electron microscope families: The scanning (SEM) and the transmission electron microscope (TEM) (Figure 6.10). It is possible to equip the TEM with a scanning unit, which then leads to a STEM (scanning transmission electron microscope), and more recently also TEM detectors are available for the SEM. The interaction of the electron beam with the specimen produces a variety of signals that can be used to obtain information such as chemical composition, morphology and structure of the specimen. Although the interaction of the electrons with matter is independent of the microscope, different sample properties (such as thickness) and different operational conditions of the microscopes (acceleration voltage) constrain the possible signals that can be detected. In the following, both SEM and TEM techniques are briefly reviewed, with a special focus on the particle characterization (both engineered and natural). The immense number of articles dealing with the analysis of nanosized materials documents the potential of the electron microscopic methods but at the same time renders an exhaustive literature review very challenging. Therefore, the focus here is on a selection of important articles, whereby the selection of the articles surely reflects in part the author’s view. Due to the rapid dilution of engineered NPs in

238

Environmental and Human Health Impacts of Nanotechnology Optical microscope Electron microscope (transmission) (transmission) Light bulb

Electron microscope (scanning)

Electron guns

High Voltage generator

Cathode ray tube (CRT) Electron gun

Condenser lens

Deflection coils

Specimen

Signal amplifier

Scan generator Objective lens

First image plane

Specimen

Fluorescent Deetector screen of CRT Final magn. <100000X

Projectior lens

Final image Binoculars or camera Final magn. <1000X

Fluorescent screen or camera Final magn. <1000000X

= Glass lens = Electromagnetic lens

Figure 6.10 Schematic views of transmission and scanning electron microscopes and optical microscope is shown for comparison. (Adapted from Wilkinson, K. J. and Lead, J. R. (2007), Environmental Colloids and Particles: Behaviour, Structure and Characterization, John Wiley & Sons Ltd, Chichester.)

the aquatic environment, such particles will always be outnumbered by naturally occurring nanosized materials. The challenge in the first place will be to selectively detect the engineered NPs. Thus, microscopic techniques that allow selective probing of specific properties of the engineered NPs will be most promising. Scanning electron microscopy. In the SEM, a focused electron probe is scanned over the surface of the sample and the detected signal is simultaneously projected on a monitor. Different signals can be recorded in parallel (such as secondary electrons (SE) and backscattered electrons (BSE)) providing different information about the investigated sample. There is no direct path between the electron beam and the image displayed and thus the signal can be electronically enhanced. For SEM a large selection of excellent textbooks is available and a good introduction into the topic can be found in Goldstein et al., 2003. For a brief

Analysis and Characterization of Manufactured NPs in Aquatic Environments

239

introduction to both SEM and TEM techniques, the textbook by Egerton (2007) is recommended. The conventional SEM requires high vacuum conditions, therefore the samples need to be conductive in order to prevent charging effects due to the electron bombardment. Newer devices equipped with field emission guns (FEG), however, can be operated at low acceleration voltages, which enable the investigation of uncoated, non-conductive samples. The resolution of the SEM can go down to ∼1 nm; however, this parameter has to be used cautiously. The resolving power on an SEM is determined either by using standard materials (gold particles on carbon) or by referring to the smallest beam diameter than can be achieved. The resolution can thus be used to compare the performance of different SEMs. In practice, however, the properties of the samples will limit the resolution that can be achieved. In addition, it has to be kept in mind that this resolution limit is only valid for the SE image. Due to the higher escape depth (which describes the maximum depth from where the electron can still reach the surface and leave the material) the resolution of the BSE signal is much lower than the SE signal. The resolution of the BSE signal can be increased by lowering the acceleration voltage and using a low voltage BSE detector. However, a certain acceleration voltage has to be applied to get enough BSE signal. The same holds true for the lateral resolution of the X-ray signal. SEMs that are capable of working at a low vacuum (so-called variable pressure or low vacuum SEMs where the pressure in the sample chamber can be a few millibar) can be used to investigate materials that are not stable under high vacuum conditions. The gas in the sample chamber can be chosen according to the specific applications, and most often H2O(g) is used. The working principle of the detectors is based on the gas ionisation, where negatively charged ions are accelerated towards the detector and the positive charges maintain the charge equilibrium of the sample (charge suppression). Specific detectors can be operated at even higher chamber pressures, up to about 20 mbar, which then refers to the environmental SEM (ESEM). In the ESEM the triple point of water (0.01 °C, 6.11 mbar) can be reached and, thus, water can be condensed on the sample surface by cooling the sample (for example with a peltier stage). This allows not only the investigation of non-conductive samples but also wet samples. An in depth description of the ESEM technology can be found in Danilatos (1988). For a shorter introduction into this topic Danilatos (1991) and Danilatos (1997) can be consulted. Due to the gas present in the sample chamber, the primary electron beam is scattered before it reaches the sample surface, which produces an electron ‘skirt’. However, for the image formation (SE image) only the signal resulting from the unscattered fraction of the electron beam is used. The SE resulting from the scattered primary electrons originate from a much larger surface and thus only provide a featureless noise to the image. The resolution in the ESEM is therefore primarily dependent on the electron optics used, and the presence of a gas in the sample chamber is of secondary importance. The resolution of the ESEM is degraded, when investigating organic materials, but this decreased resolution is not caused by the ESEM technique itself but rather by the properties/stability of the sample under the electron beam. As the same material cannot be investigated without

240

Environmental and Human Health Impacts of Nanotechnology

coating in the conventional SEM, a comparison of the resolution of ESEM and SEM is meaningless. Recently, TEM detectors that can be mounted underneath the sample have become available for the SEM. This basically turns the SEM into a STEM, and is a very good opportunity if no TEM is available or if a quick look at a sample is required. However, due to the rather low acceleration voltage of the SEM (30 kV) the penetration of the electrons within the material is rather limited, which requires very thin samples (see the section on (Scanning) Transmission Electron Microscopy). Applications and limits (SEM). Although the resolution of a modern SEM is at around 1 nm (topography), chemical information results from a much larger volume due to the larger excitation depth of X-rays. The same holds true for the BSE, which provide a chemical contrast, although the use of low voltage BSE detectors have improved the situation. However, if the NPs are well dispersed on the substrate and chemically distinct from the substrate then the backscattered signal can be used well to distinguish the NPs. An example is given in Figure 6.11, where mercury sulfide (HgS) particles have been located on different substrates (NaCl crystals, aluminium, carbon) in the ESEM. The SEM is most often used to investigate bulk materials to get a rough overview of the sample material. More detailed studies are then performed on TEM samples. For example, Labrenz et al. (2000) have shown an association between cells and micrometre-sized mineral aggregates based on SEM investigations. The mineral aggregates were then investigated in more detail using a TEM. Hochella et al. (2005) studied samples from riverbeds of the Clark Fork River Superfund complex (Montana) and reported iron oxide coatings on silicate grains down to less than 100 nm. Kaegi et al. (2008a) visualized the distribution of titanium dioxide particles on new and aged facades using SEM (low voltage BSE detector) and ESEM studies (Figures 6.12 and 6.13). Even if the particles are dispersed on a TEM grid and thus can be images with higher resolution in a TEM, the SE provides information about the sample topography which is complementary to the information obtained in the TEM. SEM investigations are powerful in situations, where a separation of NPs from the substrate is not feasible or not desired. Low voltage SEM and ESEM allow the investigation of non-conductive materials without coating which reduces the risk of artefacts. This allows, for example, the identification of NPs on a non-conductive substrate. ‘Bulk’ samples can be introduced into the sample chamber, which greatly reduces time and artefacts related to sample preparation, especially when compared to the TEM technology where the samples have to be small enough to be deposited on a TEM grid. (Scanning) Transmission Electron Microscopy. In the TEM the sample is irradiated with a parallel beam of electrons and the image of the transmitted electrons is visualized by either a fluorescent screen or recorded on a CCD camera. Recent TEMs are equipped with a CCD camera, which replaces the fluorescent screen, meaning that the operator no longer has to sit in a dark room. Crystalline materials can further be investigated using electron diffraction and related methods such as

Analysis and Characterization of Manufactured NPs in Aquatic Environments

241

(a)

(b)

Figure 6.11 Individual HgS nanoparticles dispersed on (a) NaCl crystals and aluminium and (b) on carbon recorded in an ESEM (XL30, FEG, FEI). Due to the strong elemental contrast of the particles compared to the background, the particles can clearly be resolved using the BSE signal.

dark field imaging or selected area electron diffraction (SAED). A comprehensive treatment of TEM is written in a very user friendly way by Williams and Carter (1996). In the TEM, electrons pass through the sample, and thus the samples have to be electron transparent, or in other words very thin. The maximal thickness of the sample depends on its composition (atomic weight) and on the energy of the electrons (acceleration voltage). The maximum sample thickness increases with increasing acceleration voltage and with decreasing atomic number of the sample. However, it also depends on the kind of analysis being performed. If only the morphology of the particles is of interest, then the particles can be up to a few µm. However, for analytical (electron energy loss spectroscopy (EELS) or energy dispersive X-ray

242

Environmental and Human Health Impacts of Nanotechnology

Figure 6.12 SEM image (low voltage BSE detector, acceleration voltage 6 kV, Hitachi S4800) of a new facade containing TiO2 pigments. The TiO2 particles are spread homogeneously over the whole facade. The average diameter of the TiO2 particles is around 150 nm. No sample treatment (coating) was applied. (Reprinted from R. Kaegi, A. Ulrich, B. Sinnet et al., Synthetic TiO2 nanoparticle emission from exterior facades into the aquatic environment, Environmental Pollution, 156, 233–9. Copyright 2008, with permission from Elsevier.)

Figure 6.13 SEM (BSE, low vacuum) image of a naturally aged facade recorded in an ESEM (XL30, FEG, FEI). The bright particles are TiO2 particles. (Reprinted from R. Kaegi, A. Ulrich, B. Sinnet et al., Synthetic TiO2 nanoparticle emission from exterior facades into the aquatic environment, Environmental Pollution, 156, 233–9. Copyright 2008, with permission from Elsevier.)

Analysis and Characterization of Manufactured NPs in Aquatic Environments

243

spectroscopy (EDX)) and high resolution applications, much thinner samples are required. As a rule of thumb morphological investigations can be carried out on micron sized particles and for analytical or high resolution applications the sample thickness should be <100 nm. Thus, investigating NPs within the TEM does in general not require any additional thinning, as the NPs are already thin enough. The conventional TEM can be equipped with a scanning unit, which leads to a scanning transmission electron microscope (STEM). In the STEM a focused electron probe is scanned over the sample and, thus, the STEM requires a convergent beam. Therefore, the operating principle of a STEM is fundamentally different from a conventional TEM, where a parallel beam is required, meaning STEM is more related to the principle of the SEM. A highly focused electron probe is required for the quantitative elemental analysis, which is discussed in the next section. Applications and limits. For a general investigation of the morphology of the particles, the TEM is operated in the conventional mode. Perret et al. (1994) separated particles from the River Rhine into different size fractions and then used TEM to visualize these size fractions. If the particles are homogeneously dispersed on the TEM grid, then the size distribution of the particles can be derived by applying image analysis tools to several TEM images. Absolute number concentrations can also be derived if the separation procedures are well documented (Kaegi et al., 2008b, 2008a). A comparison between images recorded in the SEM and in the TEM (conventional mode) is given in Figure 6.14. The SEM image is similar to a topographic image (surface) and the TEM image is a projection caused by a density and/or thickness contrast. The strong contrast of heavy elements can be used to detect nanoparticles consisting of elements such as silver, platinum or gold. In Figure 6.15 a TEM image of

Figure 6.14 Left: High resolution SEM (SE) image of a lepidocrocite recorded in a Hitachi S-4800 SEM. The low voltage (5 kV) allows the investigation of non (or poorly) conductive samples without treatment. Right: TEM image particles from the same sample recorded with a CM30 operated at 300 kV.

244

Environmental and Human Health Impacts of Nanotechnology

Figure 6.15 TEM image of silver nanoparticles collected in the waste of a laundry. The silver NPs are clearly visible due to high contrast. The inset shows an individual particle at high resolution.

silver chloride (AgCl) particles attached onto organic fibres is shown. The particles appear as black dots and are easily detected based on their strong contrast. The inset shows an individual particle at high resolution. Lattice fringes are clearly visible revealing the crystalline nature of the particle. Diffraction patterns (SAED) or Fast Fourier Transform (FFT) of high resolution images can be used to obtain information on the crystalline and the atomic structure of the particle (Labrenz et al., 2000; Hochella et al., 2005, 1999; Banfield et al., 2000). However, the TEM analysis is most powerful when combined with an elemental analysis; this is described in the next section. Elemental analysis within the electron microscope. The interaction of the electrons from the electron probe (primary electrons) with the specimen leads to the emission of various signals, which can be used to characterize the specimen. The composition of the sample can be determined by analysing the characteristic X-rays emitted from the sample due to the interaction with the electron beam. This technique is very widespread and can be used in the SEM as well as in the TEM. If the X-rays are only used in a qualitative manner, the requirements on the sample properties are not very stringent. However, if quantification is attempted, the samples need to fulfil certain criteria, which are discussed in the following section.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

245

X-ray analysis in the SEM: The basic assumption for a quantitative analysis of the X-rays within the SEM is that the electron beam loses all its energy within the specimen. This propagation of the electron probe within the specimen results in a so-called interaction volume, which describes the distribution of the primary electrons within the specimen. Further assumptions are that the sample is homogeneous, flat and infinitely thick with regard to the interaction volume. The interaction volume depends on the energy of the primary electron beam (acceleration voltage) and the atomic number of the specimen. The interaction volume ranges from about one to a few µm3 for typical acceleration voltages of 15 kV and increases with increasing acceleration voltage. If the above stated criteria are met, then a full quantification based on the X-ray signal is possible using correction algorithms. Mostly, ZAF correction methods are applied, where effects of atomic number (Z), absorption of X-rays within the sample (A) and the generation X-rays of a lower energy by X-rays of higher energy (secondary fluorescence) (F) are taken into account. EDX analysis of NPs in the SEM will be dominated by the contributions for the background (underlying substrate) and remains qualitative. X-ray analysis in the (S)TEM: The interaction volume in a TEM sample is essentially determined by the beam diameter and the sample thickness. In the TEM it is assumed that the electron beam penetrates the samples and the electron beam is only scattered once (single scattering regime). Specimens in the order of 100 nm and less fulfil these criteria, and for quantification the Cliff–Lorimer thin foil approximation can be used. This essentially means that absorption and fluorescence effects can be neglected and the ZAF correction procedure reduces to a Z correction procedure. This is done by relating elemental concentration ratios (of known standards) to measured intensity ratios of the standards. The obtained factors (k-factors) can then be used to convert measured intensity ratios of unknown samples into elemental concentrations. The procedure is clearly outlined in Williams and Carter (1996). The k-factors depend on parameters such as acceleration voltage and the beam size, and also on the optical setup of the microscope. Therefore, every TEM should be calibrated using respective standard materials in order to get reliable data. Applications and limits. The combination of conventional TEM with a qualitative elemental analysis (EDX) is straightforward, and thus most frequently used. These studies are generally performed on aquatic colloids with a lower size that extends well into the nanoscale size range (<100 nm). Examples include rather general descriptions of aquatic colloids from various environments (Leppard, 1992; Webb et al., 2000; Mondi et al., 2002; Lienemann et al., 1997; Chanudet and Filella, 2008) as well as detailed investigations on colloids such as iron colloids (Perret et al., 2000; Leppard et al., 1989; Lienemann et al., 1999; He et al., 1996; Tipping et al., 1981, 1982; Fortin et al., 1993) or marine snow (Leppard et al., 1996). Although the quantification of elemental abundances is rather straightforward, it requires adequate standards and puts more severe constraints on the maximal thickness of the particles, which explains why quantitative TEM-EDX data are only rarely found in the literature of environmental colloids. Buffle et al. (1989) report

246

Environmental and Human Health Impacts of Nanotechnology

elemental ratios of iron phosphates sampled in a eutrophic lake. However, it must be stressed that the Cliff–Lorimer method (thin foil approximation) can only be applied for ‘thin’ particles, which translates to roughly 100 nm and smaller depending on the sample composition and acceleration voltage. EELS/EF-TEM analysis: Elemental characterization of specimens in the TEM can also be performed using electron energy loss spectroscopy (EELS). In the TEM, electrons can lose energy when they travel trough the specimen due to inelastic scattering (mostly due to electron–electron interactions). This energy loss of the primary electrons can be used to gain elemental information on the specimen. Useful data can be obtained from samples as thick as 100 nm, but for quantification samples in the order of only a few tens of a nanometre are generally required. A short introduction into the EELS technique is given in Brydson (2001) and Williams and Carter (1996). For a rigorous treatment of the EELS technique the textbook by Egerton (1996) can be consulted. The EELS spectrum can be divided into three distinct parts: the zero loss peak, the low loss spectrum (<50 eV) and the high loss spectrum (>50 eV). The zero loss peak is caused by the unscattered and elastically scattered electrons and is the most intense feature in the EELS spectrum. It can be used together with the low loss spectrum to determine the specimen thickness, but carries no useful elemental or structural information. The low loss spectrum is dominated by electrons that have set up plasmon oscillations and is, therefore, often referred to as the plasmon peak. However, electrons that have also generated intra- or interband transitions are contained in the low loss spectrum. The high loss spectrum is caused by removal of inner or core shell electrons from an atom leading to ionization of the atom. This part of the spectrum carries the elemental information of the sample and thus is most important for environmental analysis. The decay of the ionized atoms can lead to the production of either x-rays or auger electrons. Thus, EELS and EDX are caused by the same phenomena but analysed in a different way, which explains the highly complementary nature of these two techniques. In the energy filtered technique (EF-TEM) the energy loss of the electrons is used for imaging purposes. In this case, only electrons that experience a certain energy loss, which is determined by the elements present in the specimen, are used for image formation. Images then represent elemental maps of the specimen, which can be recorded very fast (seconds to minutes), depending on the concentration of the elements of interest. Applications and limits. EELS and EF-TEM analysis are frequently used in material science to characterize engineered NPs. In environmental sciences, however, only a few studies have been published using these techniques (Perret et al., 1991; Mavrocordatos et al., 1994; Mavrocordatos et al., 2000). Quantification of iron(hydr) oxides has been performed in an outstanding study by Mavrocordatos et al. (2002). High Angle Annular Dark Field (HAADF) imaging: Although not strictly an elemental analysis, the HAADF technique is briefly described in this section, since this technique provides some elemental information and can be extremely useful when

Analysis and Characterization of Manufactured NPs in Aquatic Environments

247

investigating environmental colloids. In this technique, the image is formed only from very high angle, incoherently scattered electrons. The signal shows a strong dependence on the atomic weight of the specimen and the resulting images are therefore also referred to as z-contrast images. In addition, the specimen thickness and the acceleration voltage also influence the signal intensity. The thicker the specimen and the lower the acceleration voltage, the more the electrons are scattered at high angles and thus the higher the signal intensity. Applied to the analysis of (engineered) NPs this means that particles of comparable size will appear brighter the heavier they are, which makes the quest for metal particles much easier. Utunomiya and Ewing (2003) outline the potential of the technique for the analysis of NPs in the environment in general. Using this technique, nanoscale particles in rock samples (Utsunomiya and Ewing, 2003) and within ambient fine particles (Utsunomiya et al., 2002, 2004) were discovered. In an experimental study Morones et al. (2005) localized individual silver NPs inside E. coli using the HAADF technique. Sample preparation. In conventional SEM technology, samples have to be electrically conductive to avoid charging of the sample due to the electron bombardment. This requires that non-conductive samples are coated with a conducting layer, such as carbon or platinum. Newer SEMs can be operated at low acceleration voltages (a few kV or less), which also allows investigation of non-conductive samples without coating. In addition, SEMs that can be operated at low vacuum conditions (including the ESEM technology) do not require conductive samples anymore. Thus, depending on the capabilities of the respective SEM, the samples can be directly investigated without any treatment or need to be coated with a conductive layer. However, for TEM investigations, a proper sample preparation procedure is of the highest importance and often determines the quality of the analysis obtained afterwards. Therefore, the sample preparation procedure critically depends on the goal of the respective investigation. Generally, the preparation methods developed in materials science and biology can be used with slight modifications. A special technique for the investigation of organic materials has been developed and optimized (Leppard, 1995; Santschi et al., 1998; Wilkinson et al., 1999; Perret et al., 1991; Mavrocordatos and Perret, 1995; Lienemann et al., 1998). To minimize artefacts related to the drying process in the ultrahigh vacuum in the TEM these authors strongly recommend the use of ‘nanoplast’, a non-plastified melamine resin. Other issues such as optimized centrifugation deposition and optimum coverage of the sample grids are well reviewed by Mavrocordatos et al. (2007). 6.2.6.2 Atomic Force Microscopy and Other Scanning Probe Microscopy Atomic force microscopy (AFM) is a scanning probe microscope (SPM) method that uses an oscillating nanometre structured fine tip in the end of a cantilever that is being scanned over the surface of the sample substrate. Both the oscillating movement of the cantilever (z axis) and the scanning over the surface (x axis and y axis) are controlled by piezo-electric actuators. The oscillating movements of the cantilever are measured by correlating the deflection of a laser reflection that falls onto a quadrant photo diode (Figure 6.16). The laser-balance is very sensitive to

248

Environmental and Human Health Impacts of Nanotechnology

Photodiode

Laser

Feedback Loop

Cantilever tip

Piezo-electric scanner

Figure 6.16 Schematic principle of atomic force microscope, showing the cantilever with enlarged view of a cantilever tip, the laser balance with laser deflection detection photodiode and piezo-electric actuator scanner, and an example of obtained AFM force topography image.

deviations in the cantilever oscillations and thus acts as a balance that can measure both repulsive (Pauli principle) and attractive (van der Waals’) forces in the range from 10−7 to 10−12 N. These forces cause the cantilever to deflect or change its oscillating frequency if set to oscillate. There are different ways of obtaining topographical imaging in AFM but they are all based on coordinating the cantilever oscillation and the x–y–z scanning. In contact mode, the cantilever tip comes in close contact with the sample and the x–y–z scanner is set to scan over the sample in order to keep the tip deflection constant. The measured z-scanner positions are then used to make an image. It is also possible to keep the scanning height constant and measure the deflections as a function of separating distance. The drawback of contact mode is that loosely bound particles will be moved by the tip while it scans over them. Therefore, a tapping mode (oscillating and briefly touching the sample) and a non-contact mode (where the tip is deflected by the force interactions but not allowed to touch the sample) are often selected instead. AFM can be used in air and in water. There is

Analysis and Characterization of Manufactured NPs in Aquatic Environments

249

less experience of AFM in water due to difficulties of scanning over the sample without moving around the particles when probing them, and difficulties to be overcome since organics will adhere to tip and change behaviour. Sample preparation is also important in AFM, although not as critical as in electron microscopy, since less perturbing preparations can be used (e.g. only drying to 100% humidity). Nanoparticle deposition on the atomically flat substrate (e.g. mica or silicon wafer) has been studied for AFM analysis in air (Balnois and Wilkinson, 2002; Lead et al., 2005; Balnois et al., 2007). The particle height measurement has a lower size limit around 0.1 nm and the accuracy is generally good, but the lateral dimension (x–y) measurements are not accurate for NPs in AFM. This problem results from the tip having a radius of curvature of about 10 nm. For NPs of this size or smaller, the AFM cannot probe the distances close to the particle edge, but rather starts measuring the deflections from the particle before the tip centre is actually over the particle (Balnois et al., 2007), and thereby overestimates the lateral dimension. AFM can, in addition to image sample topography, measure forces between the tip and the sample. By varying the tip material, nanoparticle interactions with other materials can be quantitatively measured. There are other scanning probe microscopy techniques that can be used to measure magnetic or electron images of the particles. For example, electron tunnelling properties of a bacterial surface have been measured by Wigginton et al. (2007b). 6.2.6.3

Image Analysis and Automation

Due to the inherent single particle counting nature of microscopic measurements the statistical representativeness is often low. To obtain enough counting statistics to derive a number-based average from a distribution of particles it is necessary to count and measure many thousands of particles (Vigneau et al., 2000). Therefore, further development of automation of data acquisition and computerized image analysis algorithms will be important for the application of microscopy in the characterization of NPs in environmental or other complex samples (Bowen, 2002; Li et al., 2006; Bao et al., 2004; Jose et al., 2005; Reetz et al., 2000). 6.2.7

Spectroscopic Methods

Spectroscopic methods can be used to probe chemical entities on whole samples and on the nanoparticle ensembles in fractionated samples. 6.2.7.1

Optical Spectroscopy

UV-Vis spectroscopy can, of course, be used for quantitative analysis of light absorption in NPs according to the Beer–Lambert law. However, NPs are more complicated than molecules, since the light scattering contribution may contribute significantly to the extinction coefficient. The scattering contribution is especially important for larger NPs, aggregates and particles with a high refractive index. Since there is no simple relationship between the scattering contribution to the

250

Environmental and Human Health Impacts of Nanotechnology

extinction and concentration, the method is less suitable for complex nanoparticle mixtures of different compositions and sizes. Zattoni et al. (2003) have developed algorithms to quantitatively convert the extinction coefficient (actually turbidity) to concentration. Nevertheless, due to the minimum perturbation, simple operation and, most importantly, the ability to measure certain desired properties, optical spectroscopy has found use in nanoparticle characterization (Yu et al., 2003) In addition to the scattering contributions, small metal NPs possess the special features of surface plasmon effects (Noguez, 2005), which are special absorption bands that do not show up in the larger metal particle dispersions nor in metal salt solutions. The surface plasmon effects come from oscillations in the electronic clouds at the metal–water interface, and UV/Vis absorption and sometimes fluorescence due to quantum confinement effects have been shown for these small particles. The position of the absorption bands is dependent on both the particle size and shape and can be used in characterization of the metal nanoparticle dispersion (Noguez, 2005; Haiss et al., 2007; Akthakul et al., 2005; Panacek et al., 2006; Link and El-Sayed, 1999). It can also be used to study the first steps in agglomeration, since agglomeration from primary particles to dimers increases the aspect ratio of the particle (Aryal et al., 2006). Since the plasmon absorption bands for silver and gold are in the 400–550 nm range, and thus are separated from the absorption maxima for humic substances (250–350 nm), this is a good way to study their interactions (Diegoli et al., 2008). 6.2.7.2

X-ray Photoelectron Spectroscopy

X-ray photoelectron spectroscopy (XPS) is based on X-ray irradiation of the particulate substrate (in ultra high vaccum) and measure the electron kinetic energy and electron numbers that dissipate from the surface at an angle. From the difference between the energy of the X-rays used and the kinetic energy the electron binding energy can be calculated. An XPS spectrum is then plotted as the number of electrons detected as a function of binding energy. XPS is a surface analysis technique that can be used to study the surface (top 1–10 nm) composition and redox state of NPs. XPS is hence very suitable for studying surface chemical composition that is different from the bulk chemical composition. A study comparing the surface redox state speciation for ceria NPs found that XPS measured a more reduced surface than another X-ray spectroscopy method (X-ray absorption near edges spectroscopy) (Feng et al., 2004). XPS can also be used to study the adsorption of capping agents on metal nanoparticle surfaces. However, XPS is sensitive for contamination of the particle surfaces. Important aspects that need to be further investigated are sample preparation procedures for nanoparticles. 6.2.7.3

Fluorescence

Fluorescence spectroscopy is of course limited to particles that are either self fluorescent based on their chemical structure and surrounding, or to particles that have been labelled with a fluorescing molecules. The best example of NPs with native fluorescence are semiconductor quantum dots, for example CdSe, CdS or CdTe NPs (Yu et al., 2003). The fluorescence spectra are selective to nanoparticle

Analysis and Characterization of Manufactured NPs in Aquatic Environments

251

size and can thus be used to characterize the quantum dots during synthesis and, also, in risk assessment experiments (Bailey et al., 2004). For example, agglomeration of quantum dots can be studied by following changes in their fluorescence spectra. 6.2.8

Surface Area Measurements with Nitrogen Gas Adsorption

As discussed in Section 6.2.1.7, the specific surface area (m2 g−1) is very important for understanding nanoparticle reactivity and toxicity. In microscopy, a projected area can be determined, but that has limitations in terms of three dimensional area and porosity. The most commonly used specific surface area methodology that is commercially available is gas adsorption (mainly nitrogen) analysis. The particulate sample first has to be dried or freeze dried to a powder, and further pre-treatment to remove surface contamination may be necessary. The drying and vacuum treatments induce morphological changes to the sample that may also lead to a less accessible surface. Nitrogen gas is allowed to equilibrate with the particle powder, often at sequential incremental additions of gas. After each nitrogen addition the pressure is allowed to stabilize and the amount of adsorbed gas is calculated. From the adsorption isotherms different models can be applied to interpret the surface area, pore size, pore volume and pore area (Gregg and Sing, 1982). The BET model from Brunauer, Emmet and Teller (1938), is the most common one applied for both synthetic materials and natural particles (Marsh et al., 1984). The BET theory extends the Langmuir monolayer adsorption to multilayer thereby enabling the extraction of the surface area. 6.2.9

Method Validation

The corner stones of all analytical measurements are quality control, measurement uncertainties and method validation. A new method needs to be validated (benchmarked) with either reference materials that have been certified for a specific metric, or by comparing a set of unknown samples with known measurements using validated methods. The ideal and most independent benchmarking method is an inter-laboratory comparison (round-robin) where unknown samples are distributed to the participating laboratories. These kinds of exercises are much less common for nanometrology than in analytical chemistry in general, but the advantage can not be emphasised enough. In between the method validations, it is considered good laboratory practise (and a requirement in all accredited analyses) to routinely run a quality control sample (stable dispersion) and plot the measured QC value in a control diagram so any deviations can be easily spotted. A general review on the development of particulate reference materials can be found in Mitchell (1992). There exist only a few reference materials for NPs. These are certified reference materials (CRM) of gold NPs at sizes 10, 30 and 60 nm from NIST, certified for particle size, and a non-certified reference material from IRMM, a ~40 nm silica nanoparticle in aqueous dispersion. In addition there exists a range of non-certified size standards of polystyrene latex from 20 nm and upwards. There is also a reference iron oxide dispersion from NIST, available where positive zeta potential is certified. The procedure and requirements for development of CRM

252

Environmental and Human Health Impacts of Nanotechnology

from a measurement point of view are described in ISO Guides 34 and 35. The procedures for developing reference substances for toxicological testing are described in the OECD GD34 (OECD, 2005).

6.3 Analytical Test Strategy in NP Exposure Assessment The following three-tier analytical test strategy reflects the authors view but is in line with initial discussions at international standardization organizations, such as the OECD working party on manufactured nanomaterials and the ISO Nanotechnologies Technical Committee 229 (ISO/IEC/NIST/OECD, 2008). 6.3.1

Initial Material Characterization

Nanomaterials for risk assessment are best characterized in a tiered approach where first the material as produced (as a powder or dispersion in the reagents) is characterized for its material properties and dynamic properties as supplied. The list of important physicochemical properties provided by both OECD’s Working Party on Manufactured Nanomaterials (2008) and the ISO Technical Committee 229 on Nanotechnology, WG3 Environment, Health and Security are almost identical to those in Figure 6.1, but there are a few exceptions; for example, the view of the authors is that water solubility is a misleading term to use, since NPs are actually not dissolved in water but dispersed, and different chemical processes govern solubility and dispersability. Also, representative TEM picture(s) are mentioned as a property even though this is actually a method output that can be used to obtain certain physico-chemical information but it is not in itself a property. The raison d’être for this probably lies in the old saying ‘a picture says more than a thousand words’. However, see the discussion of representativeness in the microscopy section (Section 6.2.6). 6.3.2

Fate and Behaviour Assessment

At the second characterization tier in environmental fate and behaviour studies many of the physico-chemical properties mentioned above (or in Figure 6.1) are relevant, but characterization needs to be performed under a range of relevant environmental media conditions in order to generate process understanding and parameters for deriving predictive models (Chapter 4). In addition, the OECD’s Working Party on Manufactured Nanomaterials (2008) suggests the following list, which is largely adopted from fate test strategies of conventional chemicals, and still need to be scrutinized in the near future for relevance and applicability for fate testing of nanomaterials: • • • • • •

Dispersion stability in water Biotic degradability Ready biodegradability Simulation testing on ultimate degradation in surface water Soil simulation testing Sediment simulation testing

Analysis and Characterization of Manufactured NPs in Aquatic Environments

• • • • • • • •

253

Sewage treatment simulation testing Identification of degradation product(s) Further testing of degradation product(s) as required Abiotic degradability and fate Hydrolysis, for surface modified nanomaterials Adsorption–desorption Adsorption to soil or sediment Bioaccumulation potential.

Several of these properties are not directly applicable or relevant for NPs, while others are extremely challenging to accurately determine. To accommodate all analytical requirements, strategies need to be developed and methods optimised for characterization of NPs (free, deposited, aggregated) and dissolution products in a range of complex matrixes (soil, sediment, sewage sludge, and biological tissue). Hence, it is not only advanced analytical instrumentation that is required but, as importantly, sampling, handling, storage protocols, extraction and digestion methods also need to be optimized and validated. Again, this list reflects the driving force to accommodate nanomaterials under existing regulation rather than primarily based on scientific grounds on which NP-specific properties need to be determined. For example, most NPs are not inherently biodegradable, although to some extent they can dissolve, and the dissolution rate may be enhanced or decreased by external factors. More importantly, many of the tests build on the assumption that the nanomaterials are driven by chemical equilibrium processes (e.g. the sorption tests and octanol–water partitioning), and this is scientifically unfounded for NPs. Nanoparticle dispersions are inherently unstable (although maybe meta-stable) and the kinetically controlled processes are chemically irreversible. 6.3.3

Exposure Characterization in Effect Assessment Experiments

Analysis of the actual exposure situation in the effect experiments is necessary as a third characterization tier due to the often unstable state of engineered NPs in aqueous dispersions in general, and in the water chemistry of ecotoxicology media in particular. But since the effects experiments require many replications and the analytical capabilities are not necessarily at hand in the biological laboratories, it is important to carefully consider the requirements for appropriate metrics during the course of the experiments. Powers et al. (2007) suggest that physico-chemical characterization should be ideally carried out on the nanomaterials as-received, as-dosed (in the exposure medium), as interacting with the organisms with possible changes induced by mucus and excudates, and, finally, as present in the organisms (post mortem analysis). This corresponds to the strategy being outlined here and is also similar to some of the standardization working group discussions. There may also be a concentration dependence on some of the properties; therefore, dose– response experiments should be individually characterized to determine such dependency. Although the need was not recognized in the pioneering studies of nanoecotoxicology, characterization is now starting to be implemented in most effects

254

Environmental and Human Health Impacts of Nanotechnology

experiments. In a recent review (Foss Hansen et al., 2007) it was shown that although size determinations are becoming more common (17–96% of exposure and effects studies) other relevant properties are rarely characterized (e.g. surface area in 6–33% of studies). Assuming that all relevant physico-chemical characteristics are available from tier 1 studies (material characterization), and that colloidal stability/agglomeration rates, deposition rates and dissolution rates are determined in the fate studies (tier 2) in the same test media as in effects studies, then the following additional confirmation analysis are recommended as a minimum set: • Determination of total concentration, filterable (0.22 µm filter nominal pore size) and centrifugable (equivalent diameter ∼100 nm) concentration of the major composition of the NPs after an appropriate digestion method, during the course of experiments. • Tracking of changes in size distribution or size average in the aquatic experiments, in order to trace agglomeration kinetics. • Confirmation of enhanced depositition onto organisms or test system induced by mucus or excudates. The recommendation of methods for the above analyses based on minimum perturbation, general availability, ease of use, cost, analysis time are: • Chemical composition by standard methods (e.g. AAS, ICPMS) after appropriate digestions/separations. • Dynamic light scattering (bearing in mind its limitations). • Nanoparticle tracking analysis is very promising but needs validation and is rarely available yet. • Turbidity is a simple, inexpensive but crude method to follow agglomeration. • Samples for electron microscopy can be prepared in any laboratory for subsequent analysis in specialized lab. 6.3.4

Monitoring Nanopollution

The emissions of manufactured NPs will most likely follow the trends of the development of nanotechnology (Roco, 2005), and the increasing incorporation of nanomaterials in consumer products (http://www.nanotechproject.org/inventories/ consumer/), medicine and industrial catalysts, and so on. There are very few studies on the prediction of environmental concentrations of manufactured NPs, but one attempt predicted µg l−1 of titanium dioxide, when assuming most current products would be shifted to nano-sized titanium dioxide(Boxall et al., 2007). Another study based on usage patterns in Switzerland used a life-cycle perspective (substance flow analysis) to derive predicted environmental concentrations for nano-sized titanium dioxide as 0.7–16 µg l−1 (Mueller and Nowack, 2008). There are still many unknown parameters in these models that could be improved by systematic studies on leaching patterns in various product life stages, and results from a few such studies are becoming available. One paper simulating regular wash procedures of nano-sized silver coated socks studied leaching of nano-sized silver and silver ions using total digestions, electron microscopy, and a silver ion selective electrode (Benn and

Analysis and Characterization of Manufactured NPs in Aquatic Environments

255

Westerhoff, 2008). Another studied the weathering of ordinary façade paints (containing pigment grade titanium dioxide) and measured the leaching of NPs of titanium dioxide using centrifugation sample preparation for electron microscopy (Kaegi et al., 2008a). However, more work on both leaching and environmental fate is needed in order to facilitate such modelling efforts. As a complementary approach to predictive modelling, direct measurements of environmental concentrations of relevant NPs in relevant media should be carried out, as the standard procedure for validating the modelling approaches. There are some difficulties involved in performing this for NPs but also some potential solutions. 6.3.4.1

Challenges

In the relatively early phase of mass production of manufactured NPs in products, there will be very low environmental concentrations to be detected. Characterization of the physicochemical properties of the manufactured NPs will of course be even more challenging when they are present in barely detectable levels. The other challenge is that these extremely low concentration of manufactured NPs, are accompanied by and interact with a large amount of natural and incidentally produced NPs (Banfield and Navrotsky, 2001; Wigginton et al., 2007a; Nowack and Bucheli, 2007). 6.3.4.2

Possibilities

Development of measurement and characterization methods for manufactured NPs in both air and water is one of the main objectives identified to support a sustainable development of nanotechnology (Maynard et al., 2006). There are two general approaches that the author would like to highlight due to their potential, but it should be noted that this task of developing and validating these or other methods for detection of environmental concentration of manufactured NPs still needs to be undertaken. • The use of single nanoparticle analysis methods with screening capability. An example could be TEM with elemental mapping coupled with automated scanning over a large proportion of the specimen and utilizing image analysis. • The use of highly sensitive and selective detection methods following some sort of fractionation method. The detection should pick up chemical features that are specific for the manufactured NPs compared to the natural and incidentally produced, as well as the dissolved solutes. One example that we have started to optimize for this purpose is FlFFF-ICP-MS, but other methods using the same concept could be explored.

6.4

Conclusions

There is a multitude of methods in our analytical toolbox, each of which is applicable to varying extents for characterization and analysis of manufactured NPs. Not

256

Environmental and Human Health Impacts of Nanotechnology

all methods are validated or even tested for manufactured NPs in relevant environmental media but many have shown good potential for NPs in simple media or for natural NPs in natural waters and soils. It cannot be emphasized enough that it is necessary to use several methods to obtain an appropriate level of characterization. But the different methods often measure different metrics, and sometimes a different variant of the same metric (Table 6.2). All are equally correct, but not necessarily directly comparable, and a level of understanding is required by relevant regulatory, industrial and scientific communities; a degree of knowledge exchange as well as fundamental research is required, too. Further, different methods often have their results skewed to some extent dependent on the underlying principles or the operating conditions. This should all be taken into consideration when comparing the results for different methods. In this chapter, a three-tier physico-chemical characterisation procedure is suggested for assessing exposure of NPs: based on initial material characterization, fate and behaviour characterization in relevant environmental media, and confirmation analysis of the exposure conditions during the actual effect experiments. For detection and analysis of manufactured NPs in the environment there is still a major challenge to develop suitable methods. Two promising approaches have been indicated. One builds on coupling of fractionation methods (e.g. Field-Flow Fractionation or centrifugation) to a selective and sensitive detection method that is probing the homogeneous properties of these NPs (e.g. elemental composition, fluorescence, shape, etc). The other is to use electron microscopy, but then the samples need to be pre-fractionated and image analysis would need to be developed.

6.5 Acknowledgements The Swedish Environmental Research Council FORMAS, and University of Gothenburg Nanoparticle platform was funding Hassellöv for this work. I.-M. Hassellöv is acknowledged for producing illustrations to this chapter.

6.6

References

Aitken, R. J., S. M. Hankin, C. L. Tran, et al. (2007) REFNANO: Reference materials for engineered nanoparticle toxicology and metrology, DEFRA, London, UK. Akthakul, A., A. I. Hochbaum, F. Stellacci and A. M. Mayes, (2005) Size fractionation of metal nanoparticles by membrane filtration. Advanced Materials, 17, 532–5. Aryal, S., K. C. R. Bahadur, N. Bhattard, et al. (2006) Study of electrolyte induced aggregation of gold nanoparticles capped by amino acids, Journal of Colloid and Interface Science, 299, 191–197. Augustynski, J. (1993) The role of the surface intermediates in the photoelectrochemical behaviour of anatase and rutile TiO2, Electrochimica Acta, 38, 43–6. Baalousha, M. and J. R. Lead, (2007) Characterization of natural aquatic colloids (<5 nm) by flow-field flow fractionation and atomic force microscopy. Environmental Science & Technology, 41, 1111–7.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

257

Baalousha, M., A. Manciulea, S. Cumberland, et al. (2008) Aggregation and surface properties of iron oxide nanoparticles: Influence of pH and natural organic matter. Environmental Toxicology And Chemistry, 27, 1875–82. Backhus, D. A., J. N. Ryan, D. M. Groher, et al. (1993) Sampling colloids and colloidassociated contaminants in ground water. Groundwater, 31, 466–79. Bailey, R. E., A. M. Smith and S. Nie, (2004) Quantum dots in biology and medicine. Physica E, 25, 1–12. Balnois, E. and K. J. Wilkinson, (2002) Sample preparation techniques for the observation of environmental biopolymers by atomic force microscopy. Colloids And Surfaces A-Physicochemical And Engineering Aspects, 207, 229–42. Balnois, E., G. Papastavrou and K. J. Wilkinson, (2007) Force Microscopy and Force Measurements of Environmental Colloids, in Environmental Colloids and Particles: Behaviour, Structure and Characterization (eds Wilkinson, K. J. and Lead, J. R.), John Wiley and Sons Ltd, Chichester. Banfield, J. F. and A. Navrotsky, (eds.) (2001) Nanoparticles and the environment, The Mineralogy Society of America, Washington, DC. Banfield, J. F., S. A. Welch, H. Zhang, et al. (2000) Aggregation-Based Crystal Growth and Microstructure Development in Natural Iron Oxyhydroxide Biomineralization Products. Science, 289, 751–4. Bao, N. Z., L. M. Shen, X. Feng, et al. (2004) Shape and size characterization of potassium titanate fibers by image analysis. Journal of Materials Science, 39, 469–76. Barreiros, F. M., P. J. Ferreira and M. M. Figueiredo, (2006) Calculating Shape Factors from Particle Sizing Data. Particle and Particle Systems Characterization, 13, 368–73. Barth, H. G. and B. E. Boyes, (1992) Size Exclusion Chromatography. Analytical Chemistry, 64, 428R–442R. Barth, H. G. and R. B. Flippen, (1995) Particle Size Analysis. Analytical Chemistry, 67, 257R–272R. Beckett, R. and B. T. Hart, (1993) Use of field-flow fractionation techniques to characterise aquatic particles, colloids, and macromolecules, in Environmental Particles, 2nd volume (eds Buffle, J. and Van Leeuwen, H. P.), Lewis Publishers, Boca Raton, Florida. Beckett, R. and N. P. Lee, (1990) The Role of Organic Matter and Ionic Composition Determining the Surface Charge of Suspended Particles in Natural Waters. Colloids and Surfaces, 44, 35–49. Beckett, R., D. Murphy, S.Tadjiki, D. J. Chittleborough and J. C. Giddings, (1997) Determination of thickness, aspect ratio and size distributions for platey particles using sedimentation field-flow fractionation and electron microscopy. Colloids and Surfaces a-Physicochemical and Engineering Aspects, 120, 17–26. Behrens, S. H., M. Borkovec and P. Schurtenberger, (1998) Aggregation in charge-stabilized colloidal suspensions revisited. Langmuir, 14, 1951–4. Benn, T. B. and P. Westerhoff, (2008) Nanoparticle Silver Released into Water from Commercially Available Sock Fabrics. Environ. Sci. Technol., 42 (18), 7025–6. Borm, P., F. C. Klaessig, T. D. Landry, et al. (2006) Research Strategies for Safety Evaluation of Nanomaterials, Part V: Role of Dissolution in Biological Fate and Effects of Nanoscale Particles. Toxicological Sciences, 90, 23–32. Bowen, P. (2002) Particle size distribution measurement from millimeters to nanometers, and from rods to platelets. Journal of Dispersion Science and Technology, 23, 631–62. Boxall, A., K. Tiede, Q. Chaudhry, et al. (2007) Current and Future Predicted Exposure to Engineered Nanoparticles. http://www.defra.gov.uk/environment/chemicals/achs/070605/ annex-c.pdf, DEFRA. Brunauer, S., P. H. Emmet and E. Teller, (1938) Adsoprtion of gases in multimolecular layers. Journal of the American Chemical Society, 60, 309–19. Brydson, R. (2001) Electron Energy Loss Spectroscopy, BIOS Scientific Publishers Ltd, Guildford, UK. Buffle, J. and H. P. Van Leeuwen, (eds.) (1992) Environmental Particles I, Lewis Publishers, Boca Raton, Florida.

258

Environmental and Human Health Impacts of Nanotechnology

Buffle, J., R. R. Devitre, D. Perret and G. G. Leppard, (1989) Physicochemical Characteristics Of A Colloidal Iron Phosphate Species Formed At The Oxic-Anoxic Interface Of A Eutrophic Lake. Geochimica Et Cosmochimica Acta, 53, 399–408. Buffle, J., J. Perret and J. Newman, (1992) The use of filtration and ultrafiltration for size fractionation of aquatic particles, colloids and macromolecules, in Environmental Particles 1 (eds Buffle, J. and Van Leeuwen, H. P.), Lewis Publishers, Boca Raton, Florida. Buffle, J., K. J. Wilkinson, S. Stoll, et al. (1998) A Generalized Description of Aquatic Colloidal Interactions: The Three-colloidal Component Approach. Environ. Sci. Technol., 32, 2887–99. Burleson, D. J., M. D. Driessen and R. Lee Penn, (2004) On the Characterization of Environmental Nanoparticles. Journal of Environmental Science and Health, Part A, A39, 2707–53. Castranova, V. (2000) From coal mine dust to quartz: Mechanisms of pulmonary pathogenicity. Inhalation Toxicology, 12, 7–14. Chanudet, V. and M. Filella, (2006) A non-perturbing scheme for the mineralogical characterization and quantification of inorganic colloids in natural waters. Environmental Science & Technology, 40, 5045–51. Chanudet, V. and V. Filella, (2008) Size and composition of inorganic colloids in a peri-alpine, glacial flour-rich lake. Geochimica Et Cosmochimica Acta, 72, 1466–79. Chen, B. L. and J. P. Selegue, (2002) Separation and characterization of single-walled and multiwalled carbon nanotubes by using flow field-flow fractionation. Analytical Chemistry, 74, 4774–80. Chen, K. L. and M. Elimelech, (2007) Influence of humic acid on the aggregation kinetics of fullerene (C-60) nanoparticles in monovalent and divalent electrolyte solutions. Journal Of Colloid And Interface Science, 309, 126–34. Chen, K. L. and M. Elimelech, (2008) Interaction of Fullerene (C-60) Nanoparticles with Humic Acid and Alginate Coated Silica Surfaces: Measurements, Mechanisms, and Environmental Implications. Environmental Science & Technology, 42, 7607–14. Chen, Y. W. and J. Buffle, (1996a) Physicochemical and microbial preservation of colloid characteristics of natural water samples. 1. Experimental conditions. Water Research, 30, 2178–84. Chen, Y. W. and J. Buffle, (1996b) Physicochemical and microbial preservation of colloid characteristics of natural water samples. 2. Physical and microbial evolution. Water Research, 30, 2185–92. Chithrani, B. D., A. A. Ghazani and W. C. W. Chan, (2006) Determining the size and shape dependence of gold nanoparticle uptake into mammalian cells. Nano Letters, 6, 662–8. Citeau, L., F. Gaboriaud, F. Elsass, et al. (2006) Investigation of physico-chemical features of soil colloidal suspensions. Colloids And Surfaces A-Physicochemical And Engineering Aspects, 287, 94–105. Colfen, H., S. Tirosh and A. Zaban, (2003) Nanocrystal surface structure analysis by analytical ultracentrifugation. Langmuir, 19, 10654–9. Colvin, V. L. (2003) The potential environmental impact of engineered nanomaterials. Nature Biotechnology, 21, 1166–70. Danilatos, G. D. (1988) Foundations of environmental scanning electron microscopy. Advances in electronics and electron physics, 71, 109–250. Danilatos, G. D. (1991) Review and Outline of Environmental Sem at Present. Journal of Microscopy-Oxford, 162, 391–402. Danilatos, G. D. (1997) Evironmental Scanning Electron Microscopy, in In-Situ Microscopy in Materials Research (ed. Gai, P.L.) Dordrecht, Kluwer Academic Publishers, Dordrecht, The Netherlands. Davison, W. and R. R. De Vitre, (1992) Iron particles in freshwater, in Environmental Particles (eds Buffle, J. and Van Leeuwen, H. P.), Lewis Publishers, Boca Raton, Florida. Delgado, A. V., F. Gonzalez-Cabballero, R. J. Hunter, et al. (2005) Measurement and interpretation of electrokinetic phenomena. Pure and Applied Chemistry, 77, 1753–1805.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

259

Diallo, M. S., C. J. Glinka, W. A. Goddard and J. H. Johnson, (2005) Characterization of nanoparticles and colloids in aquatic systems 1. Small angle neutron scattering investigations of Suwannee River fulvic acid aggregates in aqueous solutions. Journal Of Nanoparticle Research, 7, 435–48. Diegoli, S., A. L. Manciulea, S. Begum, et al. (2008) Interaction between manufactured gold nanoparticles and naturally occurring organic macromolecules. Science Of The Total Environment, 402, 51–61. Donaldson, K., V. Stone, C. L. Tran, et al. (2004) Nanotoxicology. Occupational and Environmental Medicine, 61, 727–8. Doucet, F. J., L. Maguire and J. R. Lead, (2005) Assessment of cross-flow filtration for the size fractionation of freshwater colloids and particles. Talanta, 67, 144–54. Duesberg, G. S., M. Burghard, J. Muster, et al. (1998) Separation of carbon nanotubes by size exclusion chromatography. Chem. Commun., 3, 435–6. Egerton, R. F. (1996) Electron Energy Loss Spectroscopy in the Electron Microscope, Plenum Press, New York. Egerton, R. F. (2007) Physical Principles of Electron Microscopy, Springer + Business Media, New York. Feng, Z., W. Peng, J. Koberstein, et al. (2004) Cerium oxidation state in ceria nanoparticles studied with X-ray photoelectron spectroscopy and absorption near edge spectroscopy. Surface Science Reports, 563, 74–82. Filella, M. (2007) Colloidal Properties of Submicron Particles in Natural Waters, in Environmental Colloids and Particles: Behaviour, Structure and Characterization (eds Wilkinson, K. J. & Lead, J. R.), John Wiley & Sons Ltd, Chichester. Filella, M., J. Zhang, M. E. Newman and J. Buffle, (1997) Analytical applications of photon correlation spectroscopy for size distribution measurements of natural colloidal suspensions. Colloids and Surfaces A, 120, 27–46. Finsy, R. (1994) Particle Sizing by Quasi-Elastic Light-Scattering. Advances in Colloid and Interface Science, 52, 79–143. Fortin, D., G. G. Leppard and A. Tessier, (1993) Characteristics of Lacustrine Diagenetic Iron Oxyhydroxides. Geochimica Et Cosmochimica Acta, 57, 4391–404. Fortner, J. D., D. Y. Lyon, C. M. Sayes, et al. (2005) C60 in water: nanocrystal formation and microbial response. Environ Sci Technol 39, 4307–16. Foss Hansen, S., B. H. Larsen, S. I. Olsen and A. Baun, (2007) Categorization framework to aid hazard identification of nanomaterials. Nanotoxicology, 1, 243–50. Franklin, N. M., N. J. Rogers, S. C. Apte, et al. (2007) Comparative toxicity of nanoparticulate ZnO, bulk ZnO, and ZnCl2 to a freshwater microalga (Pseudokirchneriella subcapitata): The importance of particle solubility. Environmental Science & Technology, 41, 8484–90. Fraunhofer, W., G. Winter and C. Coester, (2004) Asymmetrical flow field-flow fractionation and multiangle light scattering for analysis of gelatin nanoparticle drug carrier systems. Analytical Chemistry, 76, 1909–20. Giddings, J. C. (1993) Field-flow Fractionation: Analysis of Macromolecular, Colloidal, and Particulate Matter. Science, 260, 1456–65. Gilbert, B. and J. F. Banfield, (2005) Molecular-scale processes involving nanoparticulate minerals in biogeochemical systems. Molecular Geomicrobiology, 58 (1), 109–55. Gimbert, L. J., P. M. Haygarth, R. Beckett and P. J. Worsfold, (2005) Comparison of centrifugation and filtration techniques for the size fractionation of colloidal material in soil suspensions using sedimentation field-flow fractionation. Environmental Science & Technology, 39, 1731–5. Gimbert, L. J., R. E. Hamon, P. S. Casey and P. J. Worsfold, (2007) Partitioning and stability of engineered ZnO nanoparticles in soil suspensions using flow field-flow fractionation. Environmental Chemistry, 4, 8–10. Goldstein, J., D. Newbury, D. Joy, et al. (2003) Scanning Electron Microscopy and X-Ray Microanalysis, Kluwer Academic Press, New York. Graham, D. (2007) Nanometrology – is it the next big thing in measurement? Analyst, 132, 95–6.

260

Environmental and Human Health Impacts of Nanotechnology

Graham, T. (1861) Liquid diffusion applied to analysis. Trans. Roy. Soc. (Lond.), 151, 183–224. Gregg, S. J. and K. S. W. Sing, (1982) Adsorption, Surface Area and Porosity, Academic Press, London. Guo, L., L.-S. Wen, D. Tang and P. H. Santschi, (2000) Re-examination of cross-flow ultrafiltration for sampling marine colloids: evidence from molecular probes. Marine Chemistry, 69, 75–90. Gustafsson, Ö. and P. M. Gschwend, (1997) Aquatic colloids: Concepts, definitions and current challenges. Limnology and Oceanography, 42, 519–28. Guzman, K. A. D., M. P. Finnegan and J. F. Banfield, (2006) Influence of surface potential on aggregation and transport of titania nanoparticles. Environmental Science & Technology, 40, 7688v93. Haiss, W., N. T. K. Thanh, J. Aveyard and D. G. Ferning, (2007) Determination of size and concentration of gold nanoparticles from UV-Vis Spectra. Analytical Chemistry, 79, 4215–21. Handy, R. D., F. v.d.Kammer, J. R. Lead, et al. (2008) The ecotoxicology and chemistry of manufactured nanoparticles. Ecotoxicology, 17, 287–314. Hassellöv, M. and F. v.d.Kammer, (2005) Detection methods in FFF, in Encyclopedia of Chromatography, 2nd edn (ed. Cazes, J.), Marcel Dekker, Inc, New York. Hassellöv, M., F. v.d.Kammer and R. Beckett, (2007) Characterisation Of Aquatic Colloids And Macromolecules By Field-Flow Fractionation, in Environmental Colloids and Particles: Behaviour, Structure and Characterization (eds Wilkinson, K. J. & Lead, J. R.), John Wiley & Sons Ltd, Chichester. He, Q. H., G. G. Leppard, C. R. Paige and W. J. Snodgrass, (1996) Transmission electron microscopy of a phosphate effect on the colloid structure of iron hydroxide. Water Research, 30, 1345–52. Heijman, S. G. J. and H. N. Stein, (1995) Electrostatic and Sterical Stabilization of TiO2 Dispersions. Langmuir, 11, 422–427. Hochella, M. F., J. N. Moore, U. Golla and A. Putnis, (1999) A TEM study of samples from acid mine drainage systems: Metal-mineral association with implications for transport. Geochimica Et Cosmochimica Acta, 63, 3395–406. Hochella, M. F., T. Kasama, A. Putnis, et al. (2005) Environmentally important, poorly crystalline Fe/Mn hydrous oxides: Ferrihydrite and a possibly new vernadite-like mineral from the Clark Fork River Superfund Complex. American Mineralogist, 90, 718–24. Hunter, K. A. and P. S. Liss, (1979) The surface charge of suspended particles in estuarine and coastal water. Nature, 282, 823–5. Hyung, H., J. D. Fortner, J. B. Hughes and J. H. Kim, (2007) Natural organic matter stabilizes carbon nanotubes in the aqueous phase. Environmental Science & Technology, 41, 179–84. Irache, J. M., C. Durrer, G. Ponchel and D. Duchene, (1993) Determination of particle concentration in latexes by turbidimetery. International Journal of Pharmaceutics, 90, R9–12. ISO, I.O.F.S. (2008) ISO/TS 27687:2008 Nanotechnologies – Terminology and definitions for nano-objects – Nanoparticle, nanofibre and nanoplate. ISO/IEC/NIST/OECD(2008) International workshop on documentary standards for measurement and characterization for nanotechnologies, NIST, Gaithersburg, Maryland, USA, 26–28 February 2008. Jackson, B. P., J. F. Ranville, P. M. Bertsch and A. G. Sowder, (2005) Characterization of colloidal and humic-bound Ni and U in the ‘dissolved’ fraction of contaminated sediment extracts. Environmental Science & Technology, 39, 2478–85. Jennings, B. R. and K. Parslow, (1988) Particle-Size Measurement – The Equivalent Spherical Diameter. Proceedings Of The Royal Society Of London Series A-Mathematical Physical And Engineering Sciences, 419, 137–49. Jiang, W., B. Y. S. Kim, J. T. Rutka and W. C. W. Chan, (2008) Nanoparticle-mediated cellular response is size-dependent. Nature Nanotechnology, 3, 145–50.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

261

Jose, A. J., L. S. Wong, J. Merrington and M. Bradley, (2005) Automated image analysis of polymer beads and size distribution. Industrial & Engineering Chemistry Research, 44, 8659–62. Kaegi, R., A. Ulrich, B. Sinnet, et al. (2008a) Release of Synthetic TiO2-Nanoparticles into the Aquatic Environment. Environmental Pollution, 156 (2), 233–9. Kaegi, R., T. Wagner, B. Hetzer, et al. (2008b) Size, number and chemical composition of nanosized particles in drinking water determined by analytical microscopy and LIBD. Water Research, 42, 2778–86. Kim, J. I. and C. Walther, (2007) Laser Induced Breakdown Detection (LIBD), in Environmental Colloids and Particles: Behaviour, Structure and Characterization (eds Wilkinson, K. J. & Lead, J. R.), John Wiley & Sons Ltd, Chichester. King, S. M. (1999) Small Angle Neutron Scattering, in Modern Techniques for Polymer Characterisation (eds Pethrick, R. A. and Dawkins, J. V.), New York, John Wiley & Sons Inc., New York. Kloepfer, J. A., R. E. Mielke and J. L. Nadeau, (2005) Uptake of CdSe and CdSe/ZnS Quantum Dots into Bacteria via Purine-Dependant Mechanisms. Applied and Environmental Microbiology, 71, 2548–57. Labrenz, M., G. K. Druschel, T. Thomsen-Ebert, et al. (2000) Formation of sphalerite (ZnS) deposits in natural biofilms of sulfate-reducing bacteria. Science, 290, 1744–7. Larsson, J., O. Gustafsson and J. Ingri, (2002) Evaluation and optimization of two complementary cross-flow ultrafiltration systems toward isolation of coastal surface water colloids. Environmental Science and Technology, 36, 2236–41. Lead, J. R. and K. J. Wilkinson, (2006) Aquatic colloids and nanoparticles: Current knowledge and future trends. Environmental Chemistry, 3, 159–71. Lead, J. R., W. Davison, J. Hamilton-Taylor and J. Buffle, (1997) Characterizing colloidal material in natural waters. Aquatic Geochemistry, 3, 212–32. Lead, J. R., D. Muirhead and C. T. Gibson, (2005) Characterization of freshwater natural aquatic colloids by atomic force microscopy (AFM). Environmental Science & Technology, 39, 6930–6. Ledin, A., S. Karlsson and B. Allard, (1993) Effects Of Ph, Ionic-Strength And A Fulvic-Acid On Size Distribution And Surface-Charge Of Colloidal Quartz And Hematite. Applied Geochemistry, 8, 409–14. Ledin, A., S. Karlsson, A. Duker and B. Allard, (1994) Measurements in-situ of concentration and size distribution of colloidal matter in deep groundwaters by photon-correlation spectroscopy. Water Research, 28, 1539–45. Leppard, G. G. (1992) Size, Morphology and Composition of Particulates in Aquatic Ecosystems – Solving Speciation Problems by Correlative Electron-Microscopy. Analyst, 117, 595–603. Leppard, G. G. (1995) The Characterization Of Algal And Microbial Mucilages And Their Aggregates In Aquatic Ecosystems. Science Of The Total Environment, 165, 103–31. Leppard, G. G., R. R. Devitre, D. Perret and J. Buffle, (1989) Colloidal Iron OxyhydroxyPhosphate – the Sizing and Morphology of an Amorphous Species in Relation to Partitioning Phenomena. Science of the Total Environment, 87–8, 345–54. Leppard, G. G., A. Heissenberger and G. J. Herndl, (1996) Ultrastructure of marine snow. 1. Transmission electron microscopy methodology. Marine Ecology-Progress Series, 135, 289–98. Lerche, D. (2002) Dispersion stability and particle characterization by sedimentation kinetics in a centrifugal field. Journal of Dispersion Science and Technology, 23, 699–709. Li, G., L. Chen, M. E. Graham and K. A. Gray, (2007) A comparison of mixed phase titania photocatalysts prepared by physical and chemical methods: The importance of the solid– solid interface. Journal of Molecular Catalysis A: Chemical, 275, 30–35. Li, Z. M., J. H. Yang and L. Wang, (2006) Determination of latex particle size distribution by transmission electron microscopy (TEM) and image analysis: Standardized procedures and key factors. International Journal of Polymer Analysis and Characterization, 11, 337–51.

262

Environmental and Human Health Impacts of Nanotechnology

Lienemann, C. P., M. Taillefert, D. Perret and J. F. Gaillard, (1997) Association of cobalt and manganese in aquatic systems: Chemical and microscopic evidence. Geochimica Et Cosmochimica Acta, 61, 1437–46. Lienemann, C. P., A. Heissenberger, G. G. Leppard and D. Perret, (1998) Optimal preparation of water samples for the examination of colloidal material by transmission electron microscopy. Aquatic Microbial Ecology, 14, 205–13. Lienemann, C. P., M. Monnerat, J. Dominik and D. Perret, (1999) Identification of stoichiometric iron-phosphorus colloids produced in a eutrophic lake. Aquatic Sciences, 61, 133–49. Limbach, L. K., L. Yuchun, R. N. Grass, et al. (2005) Oxide Nanoparticle Uptake in Human Lung Fibroblasts: Effects of Particle Size, Agglomeration, and Diffusion at Low Concentrations. Environmental Science & Technology, 39, 9370–6. Link, S. and M. A. El-Sayed, (1999) Spectral Properties and Relaxation Dynamics of Surface Plasmon Electronic Oscillations in Gold and Silver Nanodots and Nanorods. J Phys Chem B, 103, 8410–26. Liu, R. and J. R. Lead, (2006) Partial validation of cross flow ultrafiltration by atomic force microscopy. Analytical Chemistry, 78, 8105–12. Lok, C.-N., C.-M. Ho, R. Chen, et al. (2006) Proteomic Analysis of the Mode of Antibacterial Action of Silver Nanoparticles. Journal of Proteome Research, 5, 916–24. Lourenco, C., M. Teixeira, S. Simoes and R. Gaspar, (1996) Steric stabilization of nanoparticles: Size and surface properties. International Journal of Pharmaceutics, 138, 1–12. Lyklema, J. and H. Van Olphen, (1979) Terminology and symbols in colloid and surface chemistry – part 1.13: definitions, terminology and symbols for rheological properties. Pure Appl. Chem., 51, 1213–8. Madden, A. S. and M. F. Hochella, (2005) A test of geochemical reactivity as a function of mineral size: Manganese oxidation promoted by hematite nanoparticles. Geochimica Et Cosmochimica Acta, 69, 389–98. Maldiney, M. A. and J. M. Mouchel, (1995) In-situ observation of suspended solid aggregates in rivers. Hydrobiologia, 301, 365–73. Marsh, J. G., S. A. Crosby, D. R. Glasson and G. E. Millward, (1984) BET nitrogen adsorption studies of iron oxides from natural and synthetic sources. Thermochimica Acta, 82, 221. Mavrocordatos, D. and D. Perret, (1995) Non-Artifacted Specimen Preparation for Transmission Electron-Microscopy of Submicron Soil Particles. Communications in Soil Science and Plant Analysis, 26, 2593–602. Mavrocordatos, D., C. P. Lienemann and D. Perret, (1994) Energy-Filtered Transmission Electron-Microscopy for the Physicochemical Characterization of Aquatic Submicron Colloids. Mikrochimica Acta, 117, 39–47. Mavrocordatos, D., C. Mondi-Couture, O. Atteia, et al. (2000) Formation of a distinct class of Fe-Ca(-C-org)-rich particles in a complex peat-karst system. Journal of Hydrology, 237, 234–47. Mavrocordatos, D., R. Kaegi and V. Schmatloch, (2002) Fractal analysis of wood combustion aggregates by contact mode atomic force microscopy. Atmospheric Environment, 36, 5653–60. Mavrocordatos, D., D. Perret and G. G. Leppard, (2007) Strategies and advances in the characterization of environmental colloids by electron microscopy, in Environmental Colloids and Particles: Behaviour, Structure and Characterization (eds Wilkinson, K. J. & Lead, J. R.), John Wiley & Sons Ltd, Chichester. Maynard, A. D. (2000) Overview of methods for analysing single ultrafine particles. Philosophical Transactions of the Royal Society of London, Series A, 358, 2593–609. Maynard, A. D., R. J. Aitken, T. Butz, et al. (2006) Safe handling of nanotechnology. Nature, 444, 267–9. Megens, M., C. M. Van Kats, P. Bosecke and W. L. Vos, (1997) In situ characterization of colloidal spheres by synchrotron small-angle x-ray scattering. Langmuir, 13, 6120–9. Mitchell, J. P. (1992) Aerosol Standards. Journal of Aerosol Science, 23, 289–92.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

263

Mondi, C., K. Leifer, D. Mavrocordatos and D. Perret, (2002) Analytical electron microscopy as a tool for accessing colloid formation process in natural waters. Journal of MicroscopyOxford, 207, 180–90. Moon, M. H., D. J. Kang, J. H. Jung and J. M. Kim, (2004) Separation of carbon nanotubes by frit inlet asymmetrical flow field-flow fractionation. Journal of Separation Science, 27, 710–17. Mori, Y., M. Furukawa, T. Hayashi and K. Nakamura, (2006) Size distribution of gold nanoparticles used by small angle x-ray scattering. Particulate Science and Technology, 24, 97–103. Morones, J. R., J. L. Elechiguerra, A. Camacho, et al. (2005) The bactericidal effect of silver nanoparticles. Nanotechnology, 16, 2346–53. Morrison, M. A. and G. Benoit, (2001) Filtration artifacts caused by overloading membrane filters. Environmental Science and Technology, 35, 3774–9. Morrison, M. A. and G. Benoit, (2004) Investigation of conventional membrane and tangential flow ultrafiltration artifacts and their application to the characterization of freshwater colloids. Environmental Science & Technology, 38, 6817–23. Mueller, N. C. and B. Nowack, (2008) Exposure modeling of engineered nanoparticles in the environment. Environmental Science & Technology, 42, 4447–53. Murr, L. E., E. V. Esquivel, J. J. Bang, et al. (2004) Chemistry and nanoparticulate compositions of a 10,000 year-old ice core melt water. Water Res 38, 4282–96. Nel, A., T. Xia, L. Madler and N. Li, (2006) Toxic potential of materials at the nanolevel. Science, 311, 622–7. Noguez, C. (2005) Optical properties of isolated and supported metal nanoparticles. Optical Materials, 27, 1204–11. Nowack, B. and T. D. Bucheli, (2007) Occurrence, behavior and effects of nanoparticles in the environment. Environ Pollut, 150, 5–22. O’melia, C. R. (1987) Particle-particle interactions, in Aquatic Surface Chemistry, (ed. W. Stumm), John Wiley& Sons Inc., New York. O’melia, C. R. (1990) Kinetics of colloid chemical processes in aquatic systems, in Aquatic Chemical Kinetics, Reaction Rates of Processes in Natural Waters, (ed. W. Stumm), John Wiley & Sons Inc., New York. Oberdorster, G., E. Oberdorster and J. Oberdorster, (2005) Nanotoxicology: An emerging discipline evolving from studies of ultrafine particles. Environmental Health Perspectives, 113, 823–39. OECD Working Party On Manufactured Nanomaterials (2008) List Of Manufactured Nanomaterials And List Of Endpoints For Phase One Of The OECD Testing Programme. Series of Safety of Manufactured Nanomaterials, OECD Environment Directorate, Paris. OECD (2005) Guidance document on the validation and international acceptance of new or updated test methods for hazard assessment. OECD Environment, Health and Safety Publications Series on Testing and Assessment, OECD Environment Directorate, Paris. Pal, S., Y. K. Tak and J. M. Song, (2007) Does the antibacterial activity of silver nanoparticles depend on the shape of the nanoparticle? A study of the gram-negative bacterium Escherichia coli. Applied And Environmental Microbiology, 73, 1712–20. Panacek, A., L. Kvitek, R. Prucek, et al. (2006) Silver colloid nanoparticles: Synthesis, characterization, and their antibacterial activity. Journal of Physical Chemistry B, 110, 16248–53. Peng, F., D. L. Johnson and S. W. Effler, (2002) Suspensoids in New York City’s drinking water reservoirs: Turbidity apportionment. Journal of the American Water Resources Association, 38, 1453–65. Perminova, I. V., F. H. Frimmel, A. V. Kudryavtsev, et al. (2003) Molecular weight characteristics of humic substances from different environments as determined by size exclusion chromatography and their statistical evaluation. Environmental Science & Technology, 37, 2477–85. Perret, D., G. G. Leppard, M. Muller, et al. (1991) Electron-Microscopy of Aquatic Colloids – Non-Perturbing Preparation of Specimens in the Field. Water Research, 25, 1333–43.

264

Environmental and Human Health Impacts of Nanotechnology

Perret, D., M. E. Newman, J. C. Negre, et al. (1994) Submicron Particles in the Rhine River. 1. Physicochemical Characterization. Water Research, 28, 91–106. Perret, D., J. F. Gaillard, J. Dominik and O. Atteia, (2000) The Diversity of Natural Hydrous Iron Oxides. Environ. Sci. Technol., 34, 3540–6. Poland, C. A., R. Duffin, I. Kinloch, et al. (2008) Carbon nanotubes introduced into the abdominal cavity of mice show asbestos-like pathogenicity in a pilot study. Nature Nanotechnology, 3, 423–8. Powers, K. W., M. Palazuelos, B. M. Moudgil and S. M. Roberts, (2007) Characterization of the size, shape, and state of dispersion of nanoparticles for toxicological studies. Nanotoxicology, 1, 42–51. Puls, R. W. (1990) Colloidal considerations in groundwater sampling and contaminant transport predictions. Nuclear Safety, 31, 58–65. Puls, R. W. and R. M. Powell, (1992) Transport of Inorganic Colloids through Natural Aquifer Material: Implications for Contaminant Transport. Environmental Science & Technology, 26, 614–21. Reetz, M. T., M. O. Maase, T. Schilling and B. Tesche, (2000) Computer Image Processing of Transmission Electron Micrograph Pictures as a Fast and Reliable Tool to Analyze the Size of Nanoparticles. Journal of Physical Chemistry B, 104 (37), 8779–81. Rizzi, F. R., S. Stoll, N. Senesi and J. Buffle, (2004) A transmission electron microscopy study of the fractal properties and aggregation processes of humic acids. Soil Science, 169, 765–75. Roco, M. C. (2005) International perspective on government nanotechnology funding in 2005. Journal Of Nanoparticle Research, 7, 707–12. RS/RAEng (2004) Nanoscience and nanotechnologies: opportunities and uncertainties. The Royal Society and The Royal Academy of Engineering, UK. Ryan, J. N., M. Elimelech, J. L. Baeseman and R. D. Magelky, (2000) Silica-coated titania and zirconia colloids for subsurface transport field experiments. Environmental Science & Technology, 34, 2000–5. Ryman-Rasmussen, J. P., J. E. Riviere and N. A. Monteiro-Riviere, (2006) Penetration of Intact Skin by Quantum Dots with Diverse Physicochemical Properties. Toxicological Sciences, 91, 159–65. Santschi, P. H., E. Balnois, K. J. Wilkinson, et al. (1998) Fibrillar polysacharides in marine macromolecular organic matter as imaged by atomic force microscopy. Limnology and Oceanography, 43, 896–908. Schimpf, M., K. Caldwell and J. C. Giddings, (eds) (2000) Field-Flow Fractionation Handbook, John Wiley & Sons, Inc., New York. Schurtenberger, P. and M. E. Newman, (1993) Characterization of biological and environmental particles using static and dynamic light scattering, in Environmental Particles, 2nd volume (eds Buffle, J. and Van Leeuwen, H. P.), Lewis Publishers, Boca Raton, Florida. Sholkowitz, E. R., E. A. Boyle and N. B. Price, (1978) The removal of dissolved humic acids and iron during estuarine mixing. Earth and Planetary Science Letters, 40. Siripinyanond, A. and R. M. Barnes, (2002) Flow field-flow fractionation-inductively coupled plasma mass spectrometry of chemical mechanical polishing slurries. Spectrochimica Acta Part B-Atomic Spectroscopy, 57, 1885–96. Sterling, M. C. J., J. S. Bonnerb, A. N. S. Ernestc, et al. (2005) Application of fractal flocculation and vertical transport model to aquatic sol–sediment systems. Water Research, 39, 1818–30. Stumm, W. (1993) Aquatic colloids as chemical reactants: surface structure and reactivity. Colloids and Surfaces A, 73, 1–18. Teeguarden, J. G., P. M. Hinderliter, O. Gayler, et al. (2007) Particokinetics In Vitro: Dosimetry Considerations for In Vitro Nanoparticle Toxicity Assessments. Toxicological Sciences, 3, 300–12. Tiede, K. (2008) Detection and fate of engineered nanoparticles in aquatic systems. Environment Department, University of York, UK.

Analysis and Characterization of Manufactured NPs in Aquatic Environments

265

Tiller, C. L. and C. R. O’melia, (1993) Natural Organic Matter and Colloidal Stability – Models and Measurements. Colloids and surfaces A – Physicochemical and Engineering Aspects, 73, 89–102. Tipping, E., C. Woof and D. Cooke, (1981) Iron-Oxide from a Seasonally Anoxic Lake. Geochimica Et Cosmochimica Acta, 45, 1411–9. Tipping, E., C. Woof and M. Ohnstad, (1982) Forms of Iron in the Oxygenated Waters of Esthwaite Water, UK. Hydrobiologia, 91–2, 383–393. Utsunomiya, S. and R. C. Ewing, (2003) Application of high-angle annular dark field scanning transmission electron microscopy, scanning transmission electron microscopy-energy dispersive X-ray spectrometry, and energy-filtered transmission electron microscopy to the characterization of nanoparticles in the environment. Environmental Science & Technology, 37, 786–91. Utsunomiya, S., K. A. Jensen, G. J. Keeler and R. C. Ewing, (2002) Uraninite and fullerene in atmospheric particulates. Environmental Science & Technology, 36, 4943–7. Utsunomiya, S., K. A. Jensen, G. J. Keeler and R. C. Ewing, (2004) Direct identification of trace metals in fine and ultrafine particles in the Detroit urban atmosphere. Environmental Science & Technology, 38, 2289–97. v.d.Kammer, F. (2005) Characterization of Environmental Colloids applying Field-Flow Fractionation – Multi Detection Analysis with Emphasis on Light Scattering Techniques. Hamburg University of Technology, Germany. v.d.Kammer, F., M. and Baborowski, K. Friese, (2005) Field-Flow Fractionation coupled to Multi-Angle Laser Light Scattering Detectors: Applicability and Analytical Benefits for the Analysis of Environmental Colloids. Analytica Chimica Acta, 552, 166–74. Van De Hulst, H. C. (1981) Light scattering by small particles, Dover, New York. Van Der Zee, C., D. R. Roberts, D. G. Rancourt and C. P. Slomp, (2003) Nanogoethite is the dominant reactive oxyhydroxide phase in lake and marine sediments. Geology, 31, 993–6. Vigneau, E., C. Loisel, M. F. Devaux and P. Cantoni, (2000) Number of particles for the determination of size distribution from microscopic images. Powder Technology, 107, 243–50. Viguie, J. R., J. Sukmanowski, B. Nolting and F. X. Royer, (2007) Study of agglomeration of alumina nanoparticles by atomic force microscopy (AFM) and photon correlation spectroscopy (PCS). Colloids And Surfaces A-Physicochemical And Engineering Aspects, 302, 269–75. Vogl, J. and K. G. Heumann, (1997) Determination of heavy metal complexes with humic substances by HPLC/ICP-MS coupling using on-line isotope dilution technique. Fresenius Journal of Analytical Chemistry, 359, 438–41. Waychunas, G. A. (2001) Structure, aggregation, and characterization of nanoparticles, in Nanoparticles and the environment (eds Banfield, J. F. & Navrotsky, A.) The Mineralogy Society of America, Washington, DC. Waychunas, G. A., C. S. Kim and J. F. Banfield, (2005) Nanoparticulate iron oxide minerals in soils and sediments: unique properties and contaminant scavenging mechanisms. Journal Of Nanoparticle Research, 7, 409–33. Webb, P. (2008) Interpretation of Particle Size Reported by Different Analytical Techniques. Micromeritics Corporation: application note. Webb, S. M., G. G. Leppard and J. F. Gaillard, (2000) Zinc speciation in a contaminated aquatic environment: Characterization of environmental particles by analytical electron microscopy. Environmental Science & Technology, 34, 1926–33. Wells, M. L. and E. D. Goldberg, (1991) Occurrence Of Small Colloids In Sea-Water. Nature, 353, 342–4. Wigginton, N. S., K. L. Haus and M. F. Hochella, (2007a) Aquatic environmental nanoparticles. Journal Of Environmental Monitoring, 9, 1306–16. Wigginton, N. S., K. M. Rosso, B. H. Lower, et al. (2007b) Electron tunneling properties of outer-membrane decaheme cytochromes from Shewanella oneidensis. Geochimica Et Cosmochimica Acta, 71, 543–55.

266

Environmental and Human Health Impacts of Nanotechnology

Wilkinson, K. J. and J. R. Lead, (eds) (2007) Environmental Colloids and Particles: Behaviour, Structure and Characterization, John Wiley & Sons Ltd, Chichester. Wilkinson, K. J., A. Joz-Roland and J. Buffle, (1997) Different roles of pedogenic fulvic acids and aquagenic biopolymers on colloid aggregation and stability in freshwaters. Limnology And Oceanography, 42, 1714–24. Wilkinson, K. J., E. Balnois, G. G. Leppard and J. Buffle, (1999) Characteristic features of the major components of freshwater colloidal organic matter revealed by transmission electron and atomic force microscopy. Colloids And Surfaces A-Physicochemical And Engineering Aspects, 155, 287–310. Williams, D. B. and C. B. Carter, (1996) Transmission Electron Microscopy: A textbook for materials science, Plenum Press, New York. Williams, A., E. Varela, E. Meehan and K. Tribe, (2002) Characterisation of nanoparticulate systems by hydrodynamic chromatography. International Journal of Pharmaceutics, 242, 295–9. Wyatt, P. J. (1993) Light Scattering and absolute characterization of macromolecules. Analytica Chimica Acta, 272, 1–40. Xu, H., F. Yan, E. E. Monson and R. Kopelman, (2003) Room-temperature preparation and characterization of poly(ethylene glycol)-coated silica nanoparticles for biomedical applications. Journal of Biomedical Materials Research Part A, 66A, 870–9. Yang, L. and D. J. Watts, (2005) Particle surface characteristics may play an important role in phytotoxicity of alumina nanoparticles. Toxicology Letters, 158, 122–32. Yin, Y. and A. P. Alivisatos, (2005) Colloidal nanocrystal synthesis and the organic-inorganic interface. Nature, 437, 664–70. Yu, W. W., L. Qu, W. Guo and X. Peng, (2003) Experimental determination of the extinction coefficient of CdTe, CdSe, and CdS nanocrystals. Chem. Mater., 15, 2854–60. Zattoni, A., E. L. Piccolomini, G. Torsi and P. Reschiglian, (2003) Turbidimetric detection method in flow-assisted separation of dispersed samples. Analytical Chemistry, 75, 6469–77.

7 Ecotoxicology of Manufactured Nanoparticles Simon C. Apte, Nicola J. Rogers and Graeme E. Batley Centre for Environmental Contaminants Research, CSIRO Land and Water, Bangor, NSW, Australia

7.1

Introduction

Ecotoxicology is a relatively young branch of science and has evolved significantly over the last 30 years. Forbes and Forbes (1994) define ecotoxicology as ‘the field of study which integrates the ecological and toxicological effects of chemical pollutants in populations, communities and ecosystems with the fate (transport, transformation and breakdown) of such pollutants in the environment’. Ecotoxicologists acquire information by both laboratory testing (bioassays) and field studies, which may involve individual organisms, communities and populations. Both approaches have a role to play in understanding the effects of contaminants on living organisms. Laboratory bioassays may examine acute or chronic organism responses and may use various endpoints such as lethality or suppression of growth. Biomarkers of organism stress may also be employed to measure the sub-lethal effects of contaminants on organisms. Ecotoxicology plays a central role in the testing and regulation of new substances. Safety evaluations carried out in most developed countries include ecotoxicity tests to enable some assessment of the risk to the environment (e.g. acute toxicity to fish, invertebrates and algae). A number of standard test procedures have been adopted internationally and are used for the testing of new substances (OECD, 2006). A

Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

268

Environmental and Human Health Impacts of Nanotechnology

Percent surface molecules

constant challenge to this area of ecotoxicology is the characterisation of the risks posed by emerging contaminants which may pose new, or hitherto unforseen threats to our environment. Examples of recent emerging contaminants include endocrine disrupting chemicals, pharmaceuticals and the breakdown products of fluorinated polymers. Clearly, manufactured nanoparticles, which are finding increasing applications in industry and consumer products (Klaine et al. 2008), fit into the category of emerging contaminants. As noted in other chapters in this book, manufactured nanoparticles have dimensions, usually ranging from 1 to 100 nanometres (nm), which compares to the dimensions of living cells which are typically 1–10 micron in diameter. The nanoscale confers different properties compared to the bulk material including larger surface area, enhanced reactivity and changed physical properties (e.g. quantum effects). For example, as illustrated in Figure 7.1, the percentage of surface molecules increases exponentially as particle diameter decreases below 100 nm. For a given mass concentration of a nanomaterial, the actual number of molecules in potential contact with living cells will therefore be higher than encountered for the same concentration of micron sized (or greater) particle dimensions of the same material. This is potentially significant for materials where toxicity is related to surface area (e.g. catalytic generation of reactive oxygen species, release of toxic species by dissolution). Studies examining the toxicity of engineered nanoparticles in cell cultures and animals have shown that size, surface area, surface chemistry, solubility and possibly shape all play a role in determining the potential for engineered nanoparticles to cause harm (Maynard et al., 2006). A key question is whether nanoparticles have toxicity relating to their size or nanostructure that is different to the bulk material.

60

40

20

0 1

10

100 Diameter (nm)

1,000

10,000

Figure 7.1 Percent surface molecules as a function of particle size. Note that the percentage of surface molecules increases exponentially when particle size decreases <100 nm. For instance, a particle of 30 nm diameter has about 10% surface molecules, whereas a particle of 3 nm size has 50%. This emphasises the importance of surface area for increased chemical and potentially biological activity. (Reproduced with permission from Oberdorster, G., Oberdorster, E., Oberdorster, J. (2005) Nanotoxicology: An emerging discipline evolving from studies of ultrafine particles, Environmental Health Perspectives, 113, 2005, 823–39.)

Ecotoxicology of Manufactured Nanoparticles

269

Atmospheric deposition

Soil application

Effluent discharge

Surface runoff

Road runoff

Groundwater discharge Accidental spillage

Figure 7.2

Potential sources of manufactured nanoparticles in the environment.

The exposure of organisms to nanoparticles will, of course, depend on their release into the environment and their environmental concentrations. The potential sources of manufactured nanoparticles to the environment are illustrated schematically in Figure 7.2. Like other industrially-derived chemicals, manufactured nanoparticles may be discharged to the environment at several stages in their life cycle. This includes during manufacture, transport (accidental spillage), use in industry or by consumers, and during disposal. In this chapter, scientific literature to date on the ecotoxicology of manufactured nanoparticles is reviewed and critical recommendations made for further studies and the development of ecotoxicological test procedures. Background chemical and physical information on manufactured nanoparticles of relevance to ecotoxicologists is also provided.

7.2

Physico-Chemical Transformation of Nanoparticles

The detailed chemistry of manufactured nanomaterials and analytical methods to characterise the properties manufactured nanoparticles (e.g. particle size, zeta potential) are covered in chapters 5, 6 and 8. A summary is given here on important physical and chemical processes that may modulate nanoparticle toxicity, both in the environment following release of the nanoparticulates and also over the duration of the ecotoxicity tests performed in the laboratory. The basic processes affecting the fate and transport of nanoparticles in aquatic systems are illustrated in Figure 7.3.

270

Environmental and Human Health Impacts of Nanotechnology Stabilization of nanoparticles by surfactants Binding to suspended particles

Binding to NOM

Aggregation

and other colloids

and other colloids

Dissolution

Mn+

Binding to NOM

Biological degradation, hydrolysis, photolysis

Mn+ n+

M

Sedimentation

Figure 7.3

7.2.1

Processes affecting the fate and transport of nanoparticles in aquatic systems.

Particle Dispersion and Aggregation

The dispersion of nanoparticles to form stable suspensions is a desirable step prior to toxicity testing. Unfortunately, some areas of the ecotoxicological literature have confused this process with that of dissolution. Dissolution is the process by which solids (including nanoparticles) dissolve in a solvent to form a solution containing ions or solvated molecules. This process results in a reduction in size of the particle and ultimately its disappearance. Conversely, dispersion results in the formation of a liquid phase suspension in which the individual nanoparticles are present as separate entities surrounded by solvent molecules. Solubilising agents assist the dissolution of particles whereas dispersing agents assist the formation of a nanoparticle suspension. Nanoparticles are a sub-set of colloids and are mainly assemblages of molecules or atoms. For such nanoparticles it is therefore appropriate to use the term disperse rather than dissolve. Some potential exceptions include fullerenes and dendrimers, which are molecules in their own right. Dissolution may be correctly applied to these nanomaterials when referring to the solvation of individual molecules by a solvent to form a solution. In practice, however, most fullerenes are poorly soluble (especially in aqueous media) and are more likely to form semi-stable suspensions of nanoparticles by the aggregation of many fullerene molecules. The stability of nanoparticles in aqueous media has been covered in detail in chapters 2 and 4. To briefly summarise, nanoparticles may form stable suspensions in solution if their surface charge is sufficient to repel other particles or if there are steric factors (such as the presence of long chain polymers attached to the particle surface) that inhibit particle aggregation. The Derjaguin, Landau, Verwey and Overbeek (DLVO) theory of colloid stability in aqueous electrolyte solution is

Ecotoxicology of Manufactured Nanoparticles

271

based on the interactions of two forces: electrostatic repulsion, due to the compression of the electrical double layer of surfaces with the same charge, and van der Waals’ forces of attraction (Hunter, 2001). This theory can be applied to explain the stability of a dispersion of charged nanoparticles in an aqueous solution. Zeta potential measures the accumulated charge of the particle surface resulting from its movement within a fluid environment and gives an indication of the stability of the particle dispersion (Hunter, 2001). For charged particles, a zeta potential of above +30 mV or below −30 mV is generally considered as stable and unlikely to aggregate (Malvern Instruments, 2004). From a practical perspective, the dispersion in aqueous solutions of nanoparticles that are supplied in powder form can be very difficult. Nanopowders may not fully disperse even following sonication or addition of surfactants (Franklin et al. 2007). Filtration may be required to separate the non-dispersed material prior to use. Nanoparticle stability may be achieved by adding a surface coating referred to as a stabiliser. The purpose of a stabiliser is to prevent aggregation by either electrostatic repulsion, in the form of a charged ligand, or steric hindrance, often in the form of a polymer preventing the particles from coming into close contact. For instance, citrate is used to stabilise gold and silver nanoparticles in solution. Examples of steric stabilisation are given in chapter 4 of this book. Nanoparticle suspensions may be perfectly stable under optimum conditions, but as soon as they are transferred to other aqueous environments, such as the test solutions required to conduct ecotoxicity bioassays, this may change. Factors affecting aggregation include changes in pH, ionic strength and major ion concentrations. Particle aggregation over the course of toxicity tests appears to be commonplace. For example, Franklin et al. (2007) found extensive aggregation of metal oxide nanoparticles in algal growth bioassay media (Figure 7.4). Griffitt et al. (2007) also described the aggregation of copper nanoparticles during the execution of fish toxicity tests. Aggregates may well represent the stable form of nanoparticles in environmental systems and it is therefore appropriate to expose organisms to nanoparticles in this form. Clearly, it is necessary to understand how aggregation affects the bioavailability of nanoparticles. This may be linked to the physical stability of the aggregates. In natural waters it is highly likely that nanoparticles will interact with natural organic matter. Coatings of humic and fulvic acids on the outer surfaces of the nanoparticles will influence their surface charge and ability to aggregate. It appears that the interaction of certain types of manufactured nanoparticle with natural organic matter may actually increase particle stability. Hyung et al. (2007) showed that the addition of standard Suwannee River humic acid greatly enhanced the dispersion of multi-walled carbon nanotubes in deionised water. The same effects were also seen in suspensions in Suwannee River water samples. The stabilising effect was greater than that observed with the surfactant sodium dodecylsulfate, which commonly used as a dispersant for carbon nanotubes. As noted in Figure 7.3, in aquatic systems nanoparticles may interact with natural colloids, suspended particulate matter and natural organic matter, leading to aggregation and potentially sedimentation from solution. Aggregation and sedimentation may constitute a pathway for the transport of nanoparticles from the water column to benthic sediments. In this environment the nanoparticles may be

272

Environmental and Human Health Impacts of Nanotechnology

(a)

(b)

Figure 7.4 TEM images of metal oxide nanoparticles (100 mg/l) in algal growth medium (pH 7.5) (a) ZnO with higher magnification image of the circled aggregate. (b) CeO2 nanoaggregates.

bioaccumulated by deposit- and filter-feeding organisms. To date, such interactions have not been fully evaluated but may significantly affect nanoparticle fate and toxicity. 7.2.2

Nanoparticle Dissolution

The terms soluble, insoluble and sparingly soluble are used quite frequently in the chemical sciences but without rigorous definition. From a materials science perspective, many nanomaterials have low solubilities, but have significant solubility (e.g. mg/l range) from an environmental and toxicological perspective. For instance, in chemical data handbooks and material safety data sheets, zinc oxide is often

Ecotoxicology of Manufactured Nanoparticles

273

classified as insoluble in water (CRC Handbook, 2007), whereas its aqueous solubility ranges from several thousand milligrams per litre at pH 6 to around one milligram per litre at pH 8 (Stumm and Morgan, 1995). For example, one gram of zinc oxide nanoparticles added to one litre of water at pH 7.8 will generate a dissolved zinc concentration of about 10 mg/l. Clearly less than 2% of the zinc oxide has dissolved but the resulting solution zinc concentration is sufficiently toxic to kill most aquatic organisms. The rate of nanoparticle dissolution is also important, particularly if it is fast compared to the duration of the toxicity test. Franklin et al. (2007) demonstrated the rapid dissolution of zinc oxide nanoparticles in algal test media at pH 7.5 (Figure 7.5). Within six hours, sufficient zinc had dissolved to become toxic. From a theoretical perspective, the solubility of a particle increases as size decreases. This is known as the Gibbs–Thompson, or Kelvin, effect and is a thermodynamic effect that results from the increased chemical potential found at curved surfaces (Borm et al., 2006). The rate of dissolution is dependent on particle surface area and, consequently, nanoparticulate materials should dissolve faster than larger sized bulk materials, for the same mass, on surface area considerations alone (Borm et al., 2006). Ostwald ripening is a potentially important kinetic effect which is a consequence of the preferential dissolution of small particles followed by reprecipitation of larger particles. Particle size therefore changes with time tending toward larger particles (Borm et al., 2006). It is interesting to consider the interaction of slowly dissolving nanoparticles (dissolving over the time course of days) with living cells. A schematic diagram showing the concentration gradient of a dissolved species with distance from the surface of the nanoparticle is shown in Figure 7.6. The highest concentrations are found close to the particle surface. Clearly, organisms in intimate contact with such particles will receive a higher dose of the dissolving material than is predicted by measurement of the bulk solution concentration. The measurement of nanoparticle solubility is not a trivial exercise, as it necessitates the sampling of dissolved species after an appropriate equilibration time. This can be done by use of equilibrium dialysis (Figure 7.5) or potentially ultrafiltration. Ion selective electrodes, where available, are also suited for such applications as they can measure the concentration of the dissolving species without the need for separation. 7.2.3

Oxidation

The major oxidants in environmental systems are oxygen (dissolved or atmospheric), iron(III) and manganese dioxide (MnO2). Iron(III) becomes an important oxidant in aqueous solutions at low pH as its solubility increases markedly. Manganese dioxide is found in the solid phase in soils and sediments and takes part in various heterogeneous oxidation processes. A number of nanoparticles may be composed of, or contain, constituents that may be subject to oxidation both in aquatic and terrestrial environments. These include elemental metal nanoparticles such as silver (Henglein, 1998; Lok et al., 2007) and iron (Nurmi et al., 2005). Metal sulfides and selenides, which are major components of quantum dots, are also

12

Zn(NO3)2

Zn (mg/L)

10 8 6 4 2 Dialysed Zn 0 0

1

2

3

4

5

6

7

Time (hours) 20

Bulk ZnO

Zn (mg/L)

15

10

5 Dialysed Zn 0.1 µm filterable Zn 0 0

20

20

40 60 Time (hours)

80

Nano-ZnO

Zn (mg/L)

15

10

5 Dialysed Zn 0.1 µm filterable Zn 0 0

20

40 60 Time (hours)

80

Figure 7.5 Determination of bulk ZnO and nanoparticulate ZnO dissolution rates by equilibrium dialysis. The kinetics of Zn2+ diffusion into the dialysis cell is also shown. The nominal total zinc concentration was 10 mg/l for Zn(NO3)2 and 100 mg/l for bulk and nanoparticulate ZnO. Values represent the means of duplicate determinations. The pH values for the Zn(NO3)2 experiment ranged from 7.40 to 7.50, for the bulk ZnO and nanoparticulate ZnO experiments from 7.50 to 7.65. (Reprinted with permission from Franklin, N. M.; Rogers, N. J.; Apte, S. C. et al. Comparative toxicity of nanoparticulate ZnO, bulk ZnO, and ZnCl2 to a freshwater microalga (Pseudokirchneriella subcapitata): the importance of particle solubility, Environmental Science & Technology, 41, 8484–90. Copyright 2007 American Chemical Society.)

Ecotoxicology of Manufactured Nanoparticles

275

Dissolved Metal Concentration DBL Bulk solution concentration

Distance from particle surface

Figure 7.6 Idealised diagram showing the concentration gradient that develops around a slowly dissolving spherical metal oxide nanoparticle in aqueous solution. DBL refers to the diffusive boundary layer. Note the high localised dissolved metal concentrations close to the nanoparticle surface.

susceptible to oxidation, which may result in the release of soluble metal ions that are potentially toxic, such as Cd2+ (Derfus et al., 2004). The high surface area of nanoparticles may increase oxidation rates beyond those observed for bulk materials. Oxidation may result in the build up of an oxide coating on the surface of the nanoparticle; this can act as an impermeable barrier reducing further oxidation. However, the oxidation products themselves may also be toxic. 7.2.4 Adsorption Reactions The high surface area and chemical composition of some nanoparticles may lead to high adsorptive affinity for various dissolved molecules, including essential nutrients, cations and trace metals. Iron nanoparticles rapidly form outer iron hydroxide layers, which are effective in adsorbing both anions and cations (Waychunas et al., 2005). This may be problematic in algal growth bioassays where nutrient phosphate and other micronutrients are required to ensure adequate growth. Carbon-based nanomaterials may also adsorb hydrophobic organic molecules including many contaminants. In this respect, nanoparticles may be considered as nanovectors. Contaminants may be preconcentrated by adsorption onto nanoparticles only to be released inside organisms or in environments where the physico-chemical conditions favour desorption.

7.3 7.3.1

Mechanisms of Nanoparticle Toxicity in the Environment Exposure Routes

Mechanisms by which nanoparticles may have deleterious effects on the environment are likely to be more diverse than those for human health and will be influ-

276

Environmental and Human Health Impacts of Nanotechnology

enced by factors including the type of organism (e.g. trophic level, uni- or multicellular, aquatic or terrestrial) and the interactions of a particular organism with its environment. For terrestrial organisms, exposure routes will be much the same as those studied in human toxicology, namely via inhalation, ingestion or dermal absorption (Oberdorster et al., 2005). Exposure to atmospheric contaminants is covered in the area of environmental toxicology and usually relies on human exposure data. In aquatic animals, however, uptake across the gill and other external surface epithelia is also possible, and interactions with plants may include adsorption onto the root surface, incorporation into the cell wall or diffusion into the intercellular space (Nowack and Bucheli, 2007). The possible mechanisms by which nanoparticles may interact with biological systems are illustrated schematically in Figure 7.7. At the cellular level, most internalisation of nanoparticles by eukaryotic organisms will occur via endocytosis (Moore, 2006). Endocytosis is a process by which particulate material may enter a cell without passing through the cell membrane. The membrane folds around material outside the cell, resulting in the formation of a sac-like vesicle into which the material is incorporated. Bacteria are not able to endocytose and there are three possible mechanisms through which nanoparticles

UV

O2 O2–

Example: HO OH O

h+ UV activation of electron hole pairs leading to bond splitting and radical formation

HO C R O

Coating may protect the surface, change cellular uptake or could lead to release 01 toxic chemicals

O2

e–

C C R

Dissolution

Electron-donor/acceptor active groups

Media interactions by particle dissolution, coating, passivation and hydrophobicity/ hydrophilicity

Passivation H2O H2O Hydrophobicity→interactions with cell membranes, determining uptake Hydrophilicity→water suspendability

Material e– composition, e.g., discontinuous crystal planes and defects, generating active electonic conligurations Redox cycling and catalytic chemistry via coating metals (e.g., Fe) and organics (e.g., quinones) Fe++

O2–

e– Q

O2– H2O2

Q–

O2 Q = quinone Q– = semiquinone

OH Fenton chemistry

Figure 7.7 Possible mechanisms by which nanoparticles may interact with biological systems. (From A. Nel, T. Xia, L. Madler and N. Li, Toxic potential of materials at the nanolevel, Science, 311, 2006, 622–7. Reprinted with permission from AAAS.)

Ecotoxicology of Manufactured Nanoparticles

277

could pass through bacterial cell walls and membranes: nonspecific diffusion, nonspecific membrane damage, and specific uptake. It should be noted that the largest globular proteins observed to pass through intact bacterial cell walls are typically 4 nm diameter (Demchick and Koch 1996). 7.3.2

Nanoparticle Interactions with Cells: Cellular Uptake

Many classical models of toxicity assume that a toxicant must be taken up by the cell (usually via some kind of receptor) before a toxic response may be observed (Paquin et al., 2002). Because of their small size, nanoparticles have generally been considered to be more bioavailable than corresponding bulk materials (Colvin, 2003). The most likely and widespread mechanism by which nanoparticles may be taken up by eukaryotic cells is by receptor-mediated endocytosis (Colvin, 2003; Moore, 2006; Nowack and Bucheli, 2007), leading to their deposition in the cytoplasm and association with intracellular organelles. While work on the physiology of nanoparticle uptake in relation to the study of ecotoxicology is limited, it is instructive to consider studies involving mammalian cell lines and bacteria. Nanoparticle uptake has been demonstrated with a number of mammalian cell lines (Shukla et al., 2005; Chithrani et al., 2006), although this is not necessarily associated with enhanced cytotoxicity (Shukla et al., 2005). Nanoparticles may also be internalised by mechanisms other than endocytosis, for example passive diffusion or specific transport. Xu et al. (2004) used localised surface plasmon resonance spectroscopy (LSPRS) and dark-field optical microscopy to show silver nanoparticles up to 80 nm transporting in and out of bacterial cells (Pseudomonas aeruginosa). This was surprising as there is no evidence that prokaryotes possess mechanisms of endocytosis, and the largest pore sizes known for specific transport mechanisms in bacteria are generally considered to be less than 6 nm (Kloepfer et al., 2005). However, membrane transport of nanoparticles by a mechanism other than endocytosis is supported by both in vivo and in vitro studies with mammalian cells (Geiser et al., 2005; Rothen-Rutishauser et al., 2006). Geiser et al. (2005) found inhaled ultrafine titanium dioxide particles distributed among different lung compartments and intracellularly localised mainly in the cytoplasm, but particles within cells were not membrane bound indicating endocytosis was not the mechanism of uptake. Further studies using red blood cells as a model for non-phagocytic cells showed fluorescent spheres, titanium dioxide and gold nanoparticles ≤200 nm were able to penetrate into the cells by diffusion or adhesive interaction (Rothen-Rutishauser et al., 2006). Attention should therefore be paid to the mechanism of nanoparticle uptake, as intracellular particles which are not membrane bound may have direct access to intracellular proteins, organelles and DNA and hence enhanced toxic potential (Geiser et al., 2005). A few studies have demonstrated nanoparticle uptake by aquatic organisms (Table 7.1). In addition, a number of nanoparticles have been visualised in the cytoplasm of bacterial cells by TEM, including silver (Figure 7.8a), (Morones et al., 2005; Xu et al., 2004), metal oxides (Makhluf et al., 2005; Brayner et al., 2006) and quantum dots (Kloepfer et al., 2005). As yet, few data exist from ecotoxicological studies to demonstrate intracellular nanoparticle uptake by endocytosis or other

Whole cell

Gut

Gill

Escherichia coli

Escherichia coli Staphylococcus aureus Escherichia coli

Escherichia coli Bacillus subtilis Pseudomonas aeruginosa

Escherichia coli

Daphnids Daphnia magna

Fish Danio rerio

Danio rerio

Oncorhynchus mykiss

SWCNT

MgO

CdSe/Cd/Se/ ZnS Ag

Ag

SWCNT

Cu

Ag

SWCNT

ZnO

Membrane

Bacteria Escherichia coli Bacillus subtilis

MgO

Gill Gut

8-cell embryo

Whole cell

Whole cell

Whole cell

Whole Cell

Membrane

Target Organ

Organism

Summary of particle uptake studies.

Nanoparticle

Table 7.1

Histopathological analyses reveal damage to gill lamellae by proliferation of epithelial cells and oedema of gill filaments. Unclear if effects mediated by particle uptake. Single silver nanoparticles moved across the chorion and made their way into the inner mass of the embryo. The single silver nanoparticles were retained by the embryo as it matured into the adult and were found embedded in different tissues including the retina, brain, heart, gill arches and tail. Histopathological changes to the gill and gut and liver. Aggregated SWCNTs observed in the gut lumen.

Rapid (45 min) ingestion and presence of lipid-coated SWCNT in the gut track observed on time-course micrographs.

TEM, fluorescence spectroscopy show adenine- conjugated QDs < 5 nm are internalised. EDS used to confirm intracellular Cd and Se concentrations. Localised surface plasmon resonance spectroscopy (LSPRS) and dark-field optical microscopy used to show particles up to 80 nm transporting in and out of cells. TEM images confirm electron dense areas in the cytoplasm. TEM images showing electron dense intracellular areas. EDS elemental mapping confirms Ag distribution throughout the cell. 1–10 nm particles interact preferentially with the cell.

Atomic Force Microscopy used to show membrane on contact with halogenated adducts of MgO nanoparticle. TEM images confirm damage and leajkage of cell contents. Fluorescent dyes (PI and DAPI) show increased membrane permeability in cells in direct contact with SWCNT. Physical damage to the membrane and leakage of cell contents is proposed as the antibacterial mechanism. TEM shows ultrastructural changes on exposure to 8 ± 1 and 11 ± 1 nm particles. Electron dense intracellular areas observed for MgO nanoparticle treated cells. Elevated intracellular Mg confirmed using X-ray microanalysis. TEM reveal electron dense areas in the cytoplasm. No elemental analysis.

Evidence

Smith et al., 2007

Lee et al., 2007b

Griffitt et al., 2007

Roberts et al., 2007

Morones et al., 2005

Brayner et al., 2006 Kloepfer et al., 2005 Xu et al., 2004

Makhluf et al., 2005

Kang et al., 2007

Stoimenov, et al., 2002

Ref

Ecotoxicology of Manufactured Nanoparticles

279

250 nm

250 nm (a)

3.0

3.0

2.0

2.0

1.0

1.0

0

1.0

0

3.0 µm

2.0

nm 200

nm 250

0

0

1.0

2.0

3.0 µm

–250

–200 0

1.0

2.0

3.0

µm

0

1.0

2.0

3.0 µm

(b)

Figure 7.8 (a) TEM images of Escherichia coli cells Left: untreated. Right: treated with halogenated MgO nanoparticles for 60 minutes. (b) Tapping mode AFM images of E. coli cells with the corresponding cross sections below. Left: Untreated (z-height 0–920 nm). Right: treated with halogenated MgO nanoparticles for 20 minutes (z-height 0–450 nm). Note the changes in smoothness and height of the cell indicating damage to the E. coli cell envelope upon nanoparticle treatment. (Reprinted with permission from P. K. Stoimenov, R. L. Klinger, G. L. Marchin and K. J. Klabunde, Metal oxide nanoparticles as bactericidal agents, Langmuir, 18, 6679–86. Copyright 2002 American Chemical Society.) (See colour plate section for a colour representation)

280

Environmental and Human Health Impacts of Nanotechnology

mechanisms in eukaryotes. However, nanoparticles may also be taken into the tissues of aquatic organisms by ingestion or across the gill surface epithelia. Roberts et al. (2007) showed that water fleas (Daphnia magna) rapidly ingested lipid-coated nanotubes via normal feeding behaviour (Figure 7.8b), metabolising the lipid coating as a food source. In fish, single-walled carbon nanotubes (SWCNTs) have been observed in the gut lumen of animals exposed to sub-lethal (0.1 mg/l) concentrations for 10 days, which was associated with an increase in oxidative stress markers and ionoregulatory disturbance (Smith et al., 2007). 7.3.3 Toxicity Mechanisms Nanoparticles may cause cellular toxicity via a number of different mechanisms including, but not limited to, physical damage, dissolution of toxic ions or species and generation of reactive oxygen species (Figure 7.7). As discussed in the preceding section, the large surface area and abrasive nature of some nanoparticles may be sufficient to inflict sufficient non-specific physical damage to cell membranes and/or organelles to cause toxicity or cell death. 7.3.3.1

Membrane Damage

Nanoparticulate materials may elicit an acute toxic response either directly at the cell membrane or by gaining uptake into the cell. Nanocrystals can be prepared with very high surface areas and possess crystal morphologies with numerous edges, corners, defects and other reactive sites (Klabunde et al., 1996; Stoimenov et al., 2002). The abrasive nature of such particles may bestow the potential to inflict non-specific physical damage to cell membranes. Stoimenov et al. (2002) reported that halogenated adducts of magnesium oxide nanoparticles induced major damage to bacterial cell membranes resulting in leakage of the cell contents and cell death (Figure 7.9). Both silver (Morones et al., 2005) and zinc oxide (Brayner et al., 2006) nanoparticles have also been shown damage to bacterial membranes when exposed in the high mg/l range. Kang et al. (2007) exposed Escherichia coli cells to pristine SWCNTs (5 mg/l) with a narrow diameter distribution (0.75–1.2 nm) and demonstrated that direct contact between bacterial cells and the SWCNTs caused severe membrane damage and subsequent cell inactivation. Membrane damage may be an especially important toxicity mechanism for unicellular organisms with limited capacity to recover from massive physical damage to the cell envelope; however, microbes can also serve as useful models in elucidating cytotoxicity mechanisms that can be extrapolated to eukaryotic cells (Wiesner et al., 2006). Damage to the membranes of human cell lines has been demonstrated for nano-C60 colloidal suspensions (Sayes et al., 2005). Given the outcomes of these studies, membrane damage should be considered as a possible mechanism of nanoparticle toxicity in ecotoxicological studies on eukaryotes and also a potentially significant process for both aquatic and terrestrial bacteria that come into contact with nanoparticles. Histopathological evidence indicates that the membranes of fish gills are damaged by exposure SWCNTs (Smith et al., 2007), nano-TiO2 (Federici et al., 2007) and

Ecotoxicology of Manufactured Nanoparticles

281

(a)

(b)

Figure 7.9 Uptake of nanoparticles by aquatic organisms. (a) Left: Silver nanoparticles in the membrane and inside of an Escherichia coli cell; right: EDS elemental mapping showing silver distribution through the sample. (b) Daphnia magna exposed to 5 mg/l of lipid coated single-walled nanotubes showing large numbers of tubes filling the gut track (1 h exposure) and clumps of precipitated tubes around the daphnid (20 h) (bar = 200 µm). ((a) Reprinted with permission from J. R. Morones, J. L. Elechiguerra, A. Camacho et al. (2005) The bactericidal effect of silver nanoparticles, Nanotechnology, 16, 2346–53. Copyright 2005 Institute of Physics. (b) Reproduced with permission from A. P. Roberts, A. S. Mount, B. Seda et al. (2007) In vivo biomodification of lipid-coated carbon nanotubes by Daphnia magna, Environmental Science & Technology, 41, 3025–9. Copyright 2007 American Chemical Society.) (See colour plate section for a colour representation)

copper nanoparticles (Griffitt et al., 2007) but it is unclear if these effects are related to the direct accumulation of nanoparticles in the gill tissue. 7.3.3.2

Release of Toxic Dissolved Species

As noted in Section 7.2.2, metal-containing nanoparticles may exert toxic effects by dissolution of their component metals. Slowly dissolving nanoparticles may present a localised ‘point source’ of high concentrations of dissolved metals. Cells in close proximity to such particles will experience a much higher dose of metal than is suggested by the bulk solution dissolved metal concentration (Figure 7.6).

282

Environmental and Human Health Impacts of Nanotechnology

Metals such as copper and zinc are extremely toxic to aquatic organisms and effects on aquatic biota start to become apparent at dissolved metal concentrations of only a few µg/l. Franklin et al., (2007) compared the toxic effects of nanoparticulate ZnO, bulk ZnO and ZnCl2 to the freshwater alga Pseudokirchneriella subcapitata. Chemical investigations using equilibrium dialysis demonstrated rapid dissolution of the ZnO nanoparticles in a freshwater medium (pH 7.6), with a saturation solubility in the milligramme per litre range, similar to that of bulk ZnO. Growth inhibition experiments revealed comparable toxicity for nanoparticulate ZnO, bulk ZnO, and ZnCl2, with a 72 h IC50 value around 60 µg/l zinc, which was attributable solely to dissolved zinc. Consideration of published species sensitivity distributions (Bodar et al., 2005) indicates that 5 mg/l zinc is sufficient to cause adverse effects to a majority of aquatic species, including algae, invertebrates and fish. Thus, the most likely cause of nano (or bulk) ZnO aquatic toxicity is via dissolution and not necessarily through any specific particulate effects. Brunner et al. (2006) also noted the importance of rapid ZnO dissolution and the toxicity of Zn2+ in cytotoxicity studies using human mesothelioma and rodent fibroblast cell lines. Conversely, whilst 80 nm copper particles have been shown to be acutely toxic to zebra fish (Danio rerio), with a 48 h LC50 concentration of 1.5 mg/l in dechlorinated tap water (pH 8.2) (Griffitt et al., 2007), dissolution of the nanoparticles was insufficient to explain the observed mortality compared to a dissolved metal control. Furthermore, sub-lethal concentrations of nanocopper produced different morphological effects and gene expression patterns in the gill than soluble copper, demonstrating that the toxic effects were not mediated solely by dissolution. Thus, dissolution is a potentially important mechanism of nanoparticle toxicity which must be considered on a case by case basis, with appropriate controls included in every study of sparingly soluble nanoparticulate materials. 7.3.3.4

Generation of Reactive Oxygen Species (ROS)

Due to their large surface area to volume ratio and high chemical reactivity, nanoparticles have a greater propensity to generate reactive oxygen species (ROS: oxygen ions, peroxides and free radicals) than bulk materials (Oberdorster et al., 2005; Nel et al., 2006). Reactive oxygen species (ROS) may be generated by a number of mechanisms (Figure 7.7) which are not fully understood (Oberdorster et al., 2005; Nel 2006et al.,) but may include: (i) material composition, for example discontinuous crystal planes and defects generate active electronic configurations on the particle surface which can react with molecular oxygen (O2) to generate superoxide ( O
2 ); (ii) interaction with particular environmental conditions, for example UV activation; or (iii) the presence of redox active chemicals, either as impurities on the particle surface or of environmental or biological origin (e.g. transition metals or quinones), may lead to radical formation (Nel et al., 2006). ROS generation has been demonstrated for materials as diverse as fullerenes, carbon nanotubes, quantum dots and metal oxides (Oberdorster et al., 2005; Sayes et al., 2005, Derfus et al., 2004; Sawai et al., 1996) and has been shown to occur both in cell free systems (Brown et al., 2001, Sayes et al., 2004, Foucaud et al., 2007) and in vivo (Li et al., 2003).

Ecotoxicology of Manufactured Nanoparticles

283

To date, the generation of ROS is the most discussed and best developed model for nanoparticle toxicity and has primarily developed from studies with ultrafine particles (UFPs) in human health toxicology (Chapter 9). Reactive oxygen species may react with cells via a number of mechanisms and if present at sufficiently high concentrations will lead to oxidative stress. Intracellularly an excess of ROS may cause damage to DNA, proteins and organelles. Mitochondria are considered to be an intracellular target for nanoparticle deposition (Li et al., 2003; Oberdorster et al., 2005) and mitochondrion can themselves be a source of free electrons leaking from the electron transport chain (Unfried et al., 2007). This may potentiate the effects of redox active nanoparticles leading to further production of O
2 or other free radical species (Nel et al., 2006; Unfried et al., 2007). In prokaryotes, the electron transport chain is located in the cell envelope, thus a similar mechanism may contribute to bacterial membrane damage. In human cell lines, the cytotoxicity of water soluble fullerene species has been shown to be mediated via reactive oxygen species causing lipid peroxidation and membrane damage (Sayes et al., 2004, 2005). However, other studies have shown that soluble nC60 species may act as antioxidants and protect against radical initiated lipid peroxidation (Wang et al., 1999; Nel et al., 2006). A number of ecotoxicological investigations have considered oxidative stress as a mechanism for nanoparticle toxicity. Rainbow trout (Oncorhynchus mykiss) exposed to nanoparticulate titanium dioxide showed an increase in oxidative stress markers in the gill, intestine and brain (Federici et al., 2007) and water solubilised nC60 elevated lipid peroxidation in the brain and gill tissue of adult fathead minnow (Pimephales promelas) (Zhu et al., 2006). However, none of these studies directly measured ROS generation in the experimental media, so it is unclear if oxidative stress was a due to ROS generation directly from the nanoparticles or was a result of an immunological response. Adams et al. (2006a, 2006b) showed that illumination (direct sunlight, 6 h) increased the antibacterial activity of TiO2 and SiO2 to E. coli and B. subtilis, supporting the view that the toxicity of these metal oxides is related to ROS production. However, there was no effect of pre-illumination (250 watts, 30 min) on the toxicity of TiO2 to the green alga Desmodesmus subspicatus, suggesting ROS generation was not a major mechanism of toxicity (Hunde-Rinke and Simon, 2006). 7.3.4

Bioaccumulation

The uptake and gradual accumulation of nanoparticles by living organisms may lead to cumulative toxicity effects. These effects may of course be counteracted by the organism’s ability to detoxify accumulated particles either by excretion, passivation or storage in a benign form. It is also conceivable that the trophic transfer of bioaccumulated nanoparticles may occur. Nanoparticles also have the potential to accumulate (or adsorb) other toxins in the environment, such as heavy metals or persistent organics, and to facilitate their transport into living organisms, thus enhancing their bioavailability and/or bioaccumulation. Titanium dioxide nanoparticles have been demonstrated to enhance the bioaccumulation of heavy metals such as arsenic and cadmium in freshwater fish (Sun et al., 2007; Zhang et al., 2007).

284

Environmental and Human Health Impacts of Nanotechnology

7.4

Development of Valid/Realistic Toxicity Testing Protocols

The screening of new chemicals for ecotoxicity by using a battery of standardised tests is a cornerstone of our current approach to hazard assessment. Organisations such as the US EPA and the OECD play a central role in developing standardised testing protocols that are used internationally. Research studies which deviate from the standard toxicity testing approaches are a vital complimentary facet of ecotoxicology, as these provide information on the effects of environmental variables on toxicity and also the effects of toxicants on other organisms that are not standardised test species. Typically, important environmental variables that influence toxicity include pH, dissolved organic mater concentration, ionic strength and the concentration of inorganic ions such as calcium and magnesium. Such information is not accessed from routine toxicity testing and provides vital information for the development of predictive models of toxicity. An immediate problem is how relevant standardised testing protocols are for nanoparticle testing. Key concerns are for the chemical and physical form of the nanoparticles (size, aggregation, solubility), the use of appropriate dose metrics, compounding effects from other contaminants (natural organic matter, surfactants) and photocatalytic effects. In research studies appropriate controls comprise micron-sized versions of the nanomaterial (e.g. particulate zinc oxide, gold, silver, etc.) and where possible a dissolved control (e.g. ionic silver for silver nanoparticle toxicity studies). For routine toxicity testing, control solutions comprising a ‘standard’ nanoparticle with known toxicity would be highly desirable. As already discussed, nanoparticle aggregation in some cases can be minimised by the addition of compounds that change the surface properties, by changing solution pH, by coating or by altering surface charge. In practice, this may not be entirely feasible as many toxicity tests may be affected by the additives used to disperse the nanoparticles. Any toxicity testing in the presence of such compounds must include controls that determine the toxicity of the additives alone. Even so, the role of the additives in facilitating interactions of nanoparticles with cell membranes will not be distinguished by such controls, so data interpretation will be difficult. Similarly, as will be discussed later in relation to carbon nanotubes, additives in the form of solvents used in the preparation of nano-sized particles can have toxic effects that can be interpreted as being due to the nanoparticles (Zhu et al., 2006). A number of chemical measurements are required in order to fully understand the dynamics of nanoparticle interactions during a toxicity test. The determination of particle size at the start and finish of the test is highly desirable, as this will characterise any aggregation that may occur over the time course of the toxicity test. It will be important in assessing nanoparticle toxicity to determine the extent to which dissolution of particles contributes to any observed effects. Size-based separation methods such as ultrafiltration and dialysis are appropriate procedures to measure the soluble fractions. Franklin et al. (2007) showed that nanoparticulate zinc oxide had appreciable solubility in water, despite literature reports that it was

Ecotoxicology of Manufactured Nanoparticles

285

‘insoluble’. A dialysed suspension of nanoparticulate zinc oxide was as toxic to the freshwater alga Pseudokirchnerilla subcapitata as a suspension of bulk zinc oxide or an equivalent concentration of zinc chloride in water. Failure to test for solubility effects might have led to the erroneous conclusion that a suspension of nanoparticulate zinc oxide was toxic to the alga. The photocatalytic activity of nanoparticles can be an additional contributor to their toxicity. Adams et al. (2006a, 2006b) showed that the antibacterial activity (to Bacillus subtilis and Escherichia coli) of both TiO2 and SiO2 increased significantly in the presence of UV light. Laboratory tests should incorporate some assessment of photolytic effects so that they can be related to real effects in the environment. In some cases it is necessary to characterise effects due to impurities introduced during the manufacturing process. As will be discussed later, the presence of impurities in carbon nanotubes (Templeton et al., 2006) was found to result in enhanced toxicity compared to purified samples. Whilst it is appropriate for toxicity testing to determine the concentrations of nanoparticles causing effects on organisms, which can often be at concentrations far higher than ever encountered in the environment, the effect of long-term exposure to low concentrations of nanoparticles should not be neglected. Chronic, low dose exposure may invoke mechanisms of toxicity that are not observed in shortterm laboratory toxicity tests.

7.5 7.5.1

Review of Ecotoxicity Studies Overview

In this section, the aquatic ecotoxicology studies (algae, invertebrates and fish) conducted to date are summarised. A comprehensive tabulation of ecotoxicity data for nanoparticles is provided in Table 7.2. It is clear that there are many data gaps and a complete absence of ecotoxicity data for nanomaterials such as gold, silver, cerium oxide, nanoclays and quantum dots. The effects of nanoparticles on a range of bacterial cultures are also discussed in the ensuing sections. These data have been included as bacteria are important constituents of ecosystems carrying out many vital processes. In our opinion, the effects of nanoparticles on such organisms should not be overlooked. This view was also expressed by Klaine et al. (2008), who noted the importance of bacteria in ecosystems as providers of key environmental services such as primary productivity, waste decomposition and nutrient cycling. An understanding of nanoparticle toxicity to bacteria is therefore an important consideration when evaluating nanoparticle impacts on environmental systems. By far the most used aquatic invertebrate test organism is the water flea, Daphnia magna. Aside from acute toxicity tests, behaviour tests with filtered fullerenes have also been conducted. Daphnia exhibited behavioural responses, with juveniles showing an apparent inability to swim down from the surface and adults demonstrating sporadic swimming and disorientation (Lovern and Klaper, 2006). Exposure

Size Fraction (nm)

Nominally 10–200 Nominally 10–200

Average diameter 30

10–20

Nominal 10–200

Nominally 10–200 Nominally 10–200 Nominally 30–100 Nominally 10–200 Nominally 10–200 100 nm aggregates Not reported

Nanomaterial

n-C60 water solubilised n-C60 THF extract

n-C60 water solubilised

n-C60 THF extract

n-C60 water solubilised,

n-C60 THF extract n-C60 water solubilised, n-C60 THF extract n-C60 water solubilised, n-C60 water solubilised, n-C60 THF extract SWCNT purified

Standard US EPA medium Standard US EPA medium Synthetic hard water Synthetic hard water Synthetic hard water Synthetic hard water Seawater

Standard US EPA medium Moderately hard freshwater US EPA protocol Moderately hard freshwater US EPA protocol Moderately hard freshwater US EPA protocol Synthetic hard water

Test Medium

Oryzias latipes Danio rerio Amphiascus tenuiremis

Zebrafish embryos Meiobenthic copepods

Pimephales promelas Pimephales promelas Mycropterus salmoides Hyalella azteca

Daphnia magna

Daphnia magna

Juvenile largemouth bass Freshwater crustacea Japanese medaka

Fathead minnow

Fathead minnow

Water flea

Water flea

Daphnia magna

Daphnia magna

Water flea Water flea

Daphnia magna

Water flea

Test species

Table 7.2 Summary of toxicity testing results for manufactured nanomaterials.

No effects at 10 mg/l. Evidence of ingestion and aggregation

No acute toxicity at 0.5 mg/l for 96 h 1.5 mg/l was toxic

0.8 mg/l 100% mortality in 6–18 h No toxicity below 7 mg/l

40% mortality at 2.5 mg/l over 21 days. No acute toxicity up to 35 mg/l 0.5 mg/l 100% mortality in 6–18 h 0.5 mg/l no effects after 48 h

48-h LC50 0.46 mg/l; NOEC 150 µg/l

48-h LC50 5.5 mg/l

Templeton et al., 2006

Oberdorster, 2004 Oberdorster et al., 2006a Oberdorster et al., 2006a Zhu et al., 2007

Zhu et al., 2006

Zhu et al., 2006

Oberdorster et al., 2006a

Lovern and Klaper. 2006

Lovern and Klaper, 2006

Zhu et al., 2006

Zhu et al., 2006

48-h LC50>35 mg/l 48-h LC50 0.8 mg/l

Reference

End point

Seawater Freshwater with up to 0.15 mg/l SDS Freshwater and seawater Moderately hard water OECD 201 protocol Moderately hard water OECD 201 protocol Moderately hard water OECD 202 protocol Moderately hard freshwater US EPA protocol Moderately hard freshwater US EPA protocol Moderately hard water OECD 202 protocol

Not reported

Not reported

Aggregates

Nominal 25 (small); 100 (large)

Average 140

Nominal 25 (small); 100 (large)

30 THF; 100–500 sonicated 30 THF; 100–500 sonicated Average 140

SWCNT as prepared SWCNT

SWCNT

TiO2

TiO2

TiO2

TiO2 THF dispersed; sonicated TiO2 THF dispersed; sonicated TiO2

Test Medium

Size Fraction (nm)

Nanomaterial

Daphnia magna Daphnia magna Daphnia magna

Water flea Water flea

Daphnia magna

Pseudokirchneriella subcapitata

Desmodesmus subspicatus

Danio rerio

Amphiascus tenuiremis Oncorhynchus mykiss

Water flea

Water flea

Algae

Algae

Zebrafish embryos

Meiobenthic copepods Rainbow trout

Test species

Warheit et al., 2007

48-h EC50>100 mg/l

(continued overleaf)

Lovern et al., 2007

Lovern and Klaper, 2006

Hund-Rinke and Simon, 2006

Warheit et al., 2007

Cheng et al., 2007 Hund-Rinke and Simon, 2006

Templeton et al., 2006 Smith et al., 2007

Reference

No significant behavioural changes LOEC 2.0 mg/l

48-h EC50 THF 5.5 mg/l; Sonicated > 500 mg/l

No concentration-effect curve observed up to 3 m/l

72-h EC50 16–21 mg/L

Chlorophyll fluorescence 72-h EC50 44 mg/l small; no dose response large

No effect at 1.6 mg/l; effects at 10 mg/l. Effects on ventilation rate, gill pathologies and gill mucus secretion at 0.5 mg/l Hatching delay at 150 mg/l

End point

Moderately hard water OECD 201 protocol De-chlorinated tap water De-chlorinated tap water

140

TEM: 50–400

Nominal 19

Average 178–361 Mixed

Mixed

Nominally 80

Average diameter 70 Average diameter 12

TiO2

TiO2

TiO2

ZnO

ZnO

SiO2

Cu

Fe

Ag

De-chlorinated tap water

24

TiO2

Dilute NaCl

Spring water + food pellets Spring water + food pellets De-chlorinated tap water USEPA protocol

USEPA, pH 7.5

Test Medium

Size Fraction (nm)

Nanomaterial

Table 7.2 (continued)

Zebrafish

Water flea

Zebra fish

Water flea

Water flea

Algae

Carp

Carp

Rainbow trout

Rainbow trout

Test species

Danio rerio

Daphnia magna

Danio rerio

Daphnia magna

Pseudokirchneriella subcapitata Daphnia magna

Cyprinus carpio

Cyprinus carpio

Oncorhynchus mykiss

Oncorhynchus mykiss

Griffit et al., 2007 Oberdorster et al., 2006b Lee et al., 2007b

48-h LC50 1.5 mg/l

Embryo abnormalities EC50 = 10–20 ng/l

48-h LC50 55 mg/l

Adams et al., 2006b

Franklin et al., 2007 Adams et al., 2006b

Zhang et al., 2007

Sun et al., 2007

Warheit et al., 2007

Federici et al., 2007

Reference

8-day EC50<10 mg/l

No mortality during 25 day exposure to 10 mg/l. No mortality with 25 day exposure to 10 mg/l TiO2. Increased Cd accumulation 72-h EC50 68 µg/l due to dissolved Zn 8-day EC50 0.2–0.5 mg/l, possibly dissolved Zn

No mortality during 14-day exposure up to 1.0 mg/l. Sublethal effects including gill damage, observed. 96-h EC50>100 mg/l

End point

Ecotoxicology of Manufactured Nanoparticles

289

of negatively charged fluorescent polystyrene nanoparticle beads to aquatic crustaceans (e.g. D. magna) resulted in rapid ingestion within 30 minutes of exposure and also adsorption to the exoskeleton of the organism (Fernandes et al., 2006). Stone et al. (2006) studied marine and freshwater crustaceans (Artemia, Daphnia, Gammarids) exposed to sonicated suspensions of titanium dioxide, ultrafine carbon black and nC60. Results indicated that particles are ingested resulting in accumulation in the gastrointestinal tract. The particles also adhere to the exoskeleton surfaces of the and exposed organisms, suggesting multiple routes of exposure and absorption. LC50 (48 h) values ranged between 5 and 20 mg/l. Non-rodent vertebrate studies include several publications on fish toxicity (e.g. Griffitt et al., 2007; Oberdorster, 2004). Selective transport of nanoparticles to the brain of rodents has been observed in other studies (Oberdorster et al., 2005) and the authors suggest that this, along with the lack of neural antioxidant defences, could explain the enhanced brain lipid peroxidation levels observed. Hardly any work has been done on the effects of nanoparticles on plants and other terrestrial organisms. Possible interactions of nanoparticles with plant roots include adsorption onto the root surface, incorporation into the cell wall and uptake into the cell. The nanoparticle could also diffuse into the intercellular space, the apoplast and be adsorbed or incorporated into membranes there. Yang and Watts (2005) investigated the phytotoxicity of 13 nm alumina nanoparticles on root growth by the seeds of five different plant species in hydroponic studies. Species tested included commercially important species used in ecological risk assessments of pesticides: corn (Zea mays), cucumber (Cucumis sativus), soybean (Glycine max), cabbage (Brassica oleracea), and carrot (Daucus corota). The alumina nanoparticles inhibited root growth at high concentrations (2 mg/ml), while larger alumina particles of 200–300 nm had no effect, indicating that the alumina itself was not causing the toxicity. The inhibiting effects on root growth were decreased by the addition of dimethyl sulfoxide, a molecule that scavenges free radicals such as hydroxyl radicals, suggesting that oxidative stress may play a role in the effects of the nanoparticles on root elongation. A slight reduction in root elongation was found in the presence of uncoated alumina nanoparticles but not with nanoparticles coated with phenanthrene. The authors hypothesised that the surface charge on the alumina nanoparticles may have played a role in the decreased plant root growth; however, they did not measure the concentrations of soluble Al3+ in their studies which is a potent root toxicant and known to inhibit root growth (Murashov, 2006). The solubility of aluminium oxide increases with decreasing particle size and modification of the surface by adsorbed compounds is known to affect the dissolution rate. 7.5.2

Carbon-Based Nanoparticles

Carbon-based nanoparticles comprise amorphous carbon, graphite and fullerenebased compounds including carbon nanotubes. The fullerene-based materials are a rapidly expanding class of engineered nanoparticles whose production is expected to soon exceed several thousand tonnes per annum and whose applications include plastics, catalysts, battery and fuel cell electrodes, super-capacitors, water purifica-

290

Environmental and Human Health Impacts of Nanotechnology

tion systems, orthopaedic implants, conductive coatings, adhesives and composites, components in electronics, aircraft, aerospace and automotive industries, sensors, and for drug and vaccine delivery. Buckminsterfullerene (C60) is a spherical cage (buckyball) of 60 bonded carbon atoms. It has negligible solubility in water but forms stable colloidal aggregates (nano-C60) when fullerene in organic solvents such as tetrahydrofuran (THF) is rapidly mixed with water and the solvent is subsequently evaporated off (Deguchi et al., 2001; Fortner et al., 2005). Nanotubes can be water dispersed in the absence of solvents by stirring over a long period (weeks) (Zhu et al., 2006). Carbon nanotubes (CNTs) are allotropes of carbon that are manufactured both as cylindrical, single-walled one-atom thick (SWCNT) or multi-walled (MWCNT) varieties. Their diameter is several nanometres, but they can be several millimetres in length. There have been a number of studies that have examined the toxicity of C60 and CNTs and their derivatives in the environment to bacteria, daphnids, fish embryos and other aquatic and soil biota. The presence of impurities in both C60 and CNTs has been a major issue in toxicity studies. For nano-C60, residual THF used in some preparations has been shown to have toxic effects (Zhu et al., 2006), while for CNTs, the concern has been for the effects of residual metal catalysts (Oberdorster et al., 2006a). In examining the effects on the environment, it could be argued that THF-containing C60 and metal-free SWCNTs are both unnatural forms and, therefore, are not environmentally relevant. Nevertheless, it is important to be able to also understand the potential interactions of impurity-free nanoparticles. Mammalian studies give some indication of potential mechanisms of toxicity. Sayes et al. (2005) showed that a water soluble nano-C60 colloidal suspension disrupted normal cellular function (human dermal fibroblasts, human liver carcinoma cells) through lipid peroxidation, increasing membrane permeability. Antioxidant addition prevented this damage. Lee et al. (2007a) demonstrated the photochemical production of reactive oxygen species with C60 that was absent with aggregated C60 particles or with C60 associated with natural organic matter. Nano-C60 particles are crystalline aggregates 5–500 nm in size that are highly stable and carry a net negative charge (Fortner et al., 2005). The size of the aggregates is highly dependent on the preparation method (Lyon et al., 2006). Studies with nano-C60 have shown toxicity to the bacteria Escherichia coli and Bacillus subtilis (Lyon et al., 2005). Cell damage is most likely associated with reactive oxygen species. The inhibitory concentrations were reasonably high at 0.5–3 mg/l and there was evidence of physical association possibly exacerbated in the case of E. coli by the sticky lipopolysaccharide layer outside the cell well. The presence of high protein or salt concentrations induced aggregation and removed the toxic effect. Experiments looking at the effects of fullerene derivatives, C60-OH, C60-COOH and C60-NH2, on the bacteria E. coli and Shewanella oneidensis showed that the positively charged amino derivative inhibited growth more than the neutral C60 or its hydroxyl derivative, while the negatively charged carboxylate derivative had no effect (Tang et al., 2007). The negative charge on the bacteria was responsible for

Ecotoxicology of Manufactured Nanoparticles

291

greater adsorption of the positively charged C60 derivative. Inhibitory concentrations for this derivative were relatively high at 10 mg/l. The likelihood of reaching such concentrations in an aquatic system needs to be considered in assessing the risk of impacts. Fang et al. (2007) demonstrated that the Gram-negative bacterium Pseudomonas putida responded to oxidative stress from nano-C60 by changing membrane composition, decreasing the amount of unsaturated fatty acids in favour of cyclopropane fatty acids. Similar adaptive changes in membrane composition and behaviour were observed for the Gram-positive B. subtilis at concentrations of C60 as low as 0.01 mg/l. In soil systems, the addition of an aqueous suspension of nano-C60 (1 mg/kg) or as C60 in granular form (1000 mg/kg) was shown to have little impact on soil microbial communities or on microbial processes (Tong et al., 2007). This is possibly not surprising given the high concentrations needed to have effects in aquatic systems. The environmental impacts of SWCNTs in effluents reaching estuarine systems were examined using the estuarine meiobenthic copepods Amphiascus tenuiremis (Templeton et al., 2006). The impacts were critically dependent on the presence of impurities, as found elsewhere and discussed earlier. Purified SWCNTs showed no effects, whereas unpurified nanotubes showed increased effects on life cycle mortality in microplate exposures conducted over 35 days. These effects included increased mortality, reduced development rate and success, and reduced fertilisation success at 10 mg/l, but no deleterious effects at 1.6 mg/l. Significantly, a manufacturing byproduct, a soluble fluorescent fraction, showed chronic toxicity at sub-mg/l concentrations. The mechanism of toxicity was not examined but it was postulated that ingestion of nanoparticles may have resulted in physical disruption of feeding appendages, penetration of the gut wall, and/or active uptake followed by oxidative stress. One of the most cited studies on C60 impacts on aquatic organisms is that by Oberdorster (2004) on juvenile large-mouth bass. The study found oxidative damage in the form of lipid peroxidation in the brains of the juveniles after a 48-h exposure to 0.5 mg/l of uncoated C60 nanoparticles. Total glutathione levels were also depleted in the gills. These findings followed the work of Foley et al. (2002) showing that a water soluble fullerene derivative (C61(COOH)2) was able to cross cell membranes and reinforced the fears that engineered nanoparticles, many of which are redox active, represented a major environmental threat (Colvin, 2003). As noted earlier, ‘water-soluble’ nano-C60 can be prepared from THF soluble fractions mixed with water. Even after the THF is notionally removed by evaporation, some can remain. It was shown that fathead minnow (Pimephales promelas) exposed to aqueous solutions of 0.5 mg/l of nano-C60 prepared from THF solutions showed 100% mortality within 18 hours (Zhu et al., 2006). Nano-C60 prepared by water stirring showed no toxic effects, although there was increased lipid peroxidation in the brain tissue and gills, together with increased expression of CYP2 family isozymes in the liver compared to controls. Similarly, an order of magnitude greater toxicity of the THF-derived C60 was observed with Daphnia magna. At the same time, a second study of Daphnia magna by Lovern and Klaper (2006) found that for C60 samples sonicated in water for 30 minutes, the LC50 was 7.9 mg/l compared

292

Environmental and Human Health Impacts of Nanotechnology

to 0.5 mg/l for a THF-derived product. Sub-lethal effects such as immobilisation were often seen in a relatively short period (1 h). Experiments using filtered titanium dioxide in the presence and absence of initial THF showed similar toxicities, implying no effects due to residual THF. No data were reported for THF-treated water in the absence of titanium dioxide. In a study conducted using larval zebrafish, Danio rerio, Henry et al. (2007) investigated changes in survival and gene expression after exposure to aggregates of C60 prepared by two methods: stirring and sonication of C60 in water (C60–water); and suspension of C60 in THF followed by evaporation under reduced pressure, resuspension in water and sparging with nitrogen gas (THF–C60). Survival of larval zebrafish was reduced in THF–C60 and THF–water but not in C60–water. The greatest differences in gene expression were observed in fish exposed to THF–C60 and most of these genes were similarly expressed in fish exposed to THF–water. Analyses of THF–C60 and THF–water by gas chromatography–mass spectrometry did not detect THF but found THF oxidation products γ-butyrolactone and tetrahydro-2furanol. The toxicity of γ-butyrolactone (72-hr LC50 concentration 47 mg/l) indicated that the toxicity effects in observed in the THF–C60 treatment could have resulted from γ-butyrolactone toxicity. The use of THF as a dispersing agent is a controversial issue. For additional information on this topic, the review by Klaine et al. (2008) is recommended. The most definitive toxicity study to date, by Oberdorster et al. (2006a), examined daphnids, the freshwater crustacean Hyalella, marine copepods and fathead minnows using THF-free nano-C60 preparations. Daphnia magna showed no acute toxicity up to 35 mg/l. However, there was uptake of nanoparticles and sub-lethal effects such as delays in moulting, reduced ability to produce offspring and some mortality at concentrations as low as 2.5 mg/l. Acute toxic effects were absent with Hyalella azteca at concentrations below 7 mg/l. Similarly, the marine copepod tested showed no toxicity below 22.5 mg/l. In fathead minnow, no acute toxicity was seen at 0.5 mg/l as already observed (Zhu et al., 2006), and also no significant biomarker effects, based on traditional biomarkers of lipophillic exposure. Japanese medaka, Oryzias latipes were less sensitive to nano-C60 than fathead minnows. A comparison of the responses of nano-C60 and SWCNTs to fathead minnows indicated that the latter show lesser biochemical or gene expression changes (Oberdorster et al., 2006a). It has been reported that bioaccumulation of nano-C60 had been observed in the gastrointestinal tract of Daphnia magna and other body compartments, with particles taken up as food. This uptake was enhanced with SWCNTs by coating the nanotubes with lysophosphatidylcholine to make them dispersible in aqueous media (Roberts et al., 2007). Daphnids were shown to not only ingest the coated nanotubes but to egest insoluble nanotubes, free of the coating, that subsequently aggregated and precipitated. The need to disperse CNTs for toxicity studies again raises the issue of environmental relevance. Smith et al. (2007) chose to use a surfactant, sodium dodecyl sulfate, to disperse SWCNTs for studying their toxicity to rainbow trout (Oncorhynchus mykiss). Their comprehensive study of physiological effects, showed an absence of oxidative stress indicators, but rather the nanoparticles precipitated

Ecotoxicology of Manufactured Nanoparticles

293

in the gill mucous. The implication was that long term exposure (1–2 weeks) effects on gills and the associated respiratory distress would result in mortality. Cheng et al. (2007) suspended SWCNTs by stirring in the absence of dispersants in their study of the impact on embryos of zebrafish, Danio rerio. Not surprisingly, there was considerable aggregation with clumps greater than 3 µm in size, so clearly not nanoparticles. While there was no mortality or embryonic developmental malfunctions up to 360 mg/l, there was a hatching delay that may be attributable to the presence of cobalt and nickel catalysts. Tests on zebrafish embryos using 1.5 mg/l of THF-derived nano-C60 showed only 45% survival after 96 hours compared to solvent controls. Interestingly, the addition of glutathione significantly reduced toxicity, as a consequence of its antioxidant properties alleviating oxidative stress (Zhu et al., 2007). Average aggregate size here was considerably smaller than that of SWCNTs above, at 100 nm. 7.5.3

Metal Oxides

Metal oxide nanoparticles are finding increasing application in a wide range of applications and represent about one-third of the consumer products nanotechnology market (Maynard, 2006). Applications for these materials include pigments in paints (TiO2), as sunscreens (TiO2, ZnO), in ceramics and catalysts (CeO2) and as antimicrobial agents (MgO). Concern over the toxicity of metal oxide nanoparticles is related to their assumed persistence as small particles in the environment but in fact their solubility is highly pH dependant and at pH 7.6 ranges from moderate (>80 mg/l) for MgO to extremely insoluble (<0.15 mg/l) for TiO2 and CeO2 (Rogers et al., 2007). Additionally, they can exist in different crystalline forms, for example anatase or rutile TiO2, which may affect their toxic potential. Thus any studies concerned with the toxicity of metal oxide nanoparticles must ensure that appropriate controls for particle size and solubility are included in the experimental design and, ideally, include information on the crystal structure and surface coating of the particles. A number of studies have examined the toxic effects of metal oxide nanoparticles to bacteria, algae, daphnids and fish, although toxicity data for terrestrial species is currently lacking or very limited (Table 7.2). The need for comprehensive characterisation of the nanoparticle being tested and, more especially, for appropriate controls is an issue for some of these studies. However, as standard ecotoxicological test protocols are developed and measurement techniques become more available these difficulties should be overcome. To date, most ecotoxicological studies have concerned TiO2 and ZnO, due to the widespread use of these materials as pigments and sunscreens. Additionally, the bacterial toxicity of SiO2, CeO2 and MgO has been explored by a number of authors (Adams et al., 2006a, 2006b; Thill et al., 2006; Makhluf et al., 2005), due to the potential application of these nanoparticles as bactericides. Warheit et al. (2007) undertook to develop a base set of toxicity test for fine (380 nm) and ultrafine (140 nm) TiO2 in the rutile phase with an average size of 140 nm measured by dynamic light scattering (DLS). These tests included the aquatic test species Pseudokirchneriella subcapitata (alga), Daphnia magna (water

294

Environmental and Human Health Impacts of Nanotechnology

flea) and Oncorhynchus mykiss (rainbow trout) using accepted OECD/US EPA test protocols. The EC50 values were >100 mg/l for O. mykiss (96-h mortality) and D. magna (48-h immobility) and 21 mg/l for P. subcapitata (72-h growth inhibition), indicating that the toxicity of ultrafine TiO2 is of low to medium concern for aquatic species. However, due to the size of the primary particles used in this study (140 nm) they may not be considered truly nanoparticulate. Hund-Rinke and Simon (2006) considered the effects of 25 nm (mainly anatase) and 100 nm (100% anatase) TiO2 nanoparticles to the green alga Desmodesmus subspicatus and to D. magna. The smaller (25 nm) particles produced a 72-h EC50 of 44 mg/l for algal chlorophyll fluorescence, whereas no clear concentration effect curve could be determined for the larger particles. The extent of aggregation was not determined in the test for either size particle, making it unclear which properties contribute to toxicity. Both products caused a toxic response in the daphnid but no clear concentration effect was observed with concentrations up to 3 mg/l; higher TiO2 concentrations were not tested. To determine if the nanoparticulate TiO2 effects could be enhanced photocatalytically these tests were also repeated under UV illumination. No effects of pre-illumination of the particles were observed for the alga and the effects were not concentration dependent for the daphnid, suggesting that the photocatalytic activity of TiO2 was not a major contributor to toxicity under the conditions studied. Another study on the effects of nano-sized TiO2 on D. magna in US EPA medium indicated the 48-h EC50 was 5.5 mg/l for TiO2 dispersed in THF and filtered to remove the residual solvent (Lovern and Klaper, 2006). Conversely, no effect was observed for TiO2 dispersed by sonication alone (no THF). The filtered particles had an average diameter of 30 nm compared to 100–500 nm for the sonicated particles. The potential toxic effects of THF were investigated by comparing filtered TiO2 with and without the solvent and no statistical differences in toxicity were observed. Furthermore, D. magna survival was not affected in THF-treated and evaporated water alone (Lovern and Klaper, 2006). Sub-lethal concentrations, 2 mg/l (LOEC), of TiO2 nanoparticle showed no statistically significant effects on D. magna behaviour (hopping rate, heart rate, appendage beat and abdominal curls) over a 60minute exposure period (Lovern et al., 2007). Federici et al. (2007) exposed rainbow trout (Oncorhynchus mykiss) to a sublethal concentration (1 mg/l) of nanoparticulate TiO2 with an average particle diameter of 24 nm for 14 days. This comprehensive study of physiological effects showed gill damage, disruption of copper and zinc homeostasis and increased oxidative stress markers in the fish but no mortality was observed over this exposure period. No controls for particle size were included in this study however, making it difficult to conclude that the effects were specifically due to the properties of nanoparticulate TiO2. Other studies with the freshwater fish species Cyprinus carpio have demonstrated the potential for TiO2 nanoparticles to facilitate the transport of other metal contaminants (Sun et al., 2007; Zhang et al., 2007). In these studies, C. carpio were exposed to 10 mg/l TiO2 nanoparticles for 25 days in the presence or absence of 200 µg/l arsenic(V) (Sun et al., 2007) or 100 µg/l cadmium (Zhang et al., 2007). TEM characterisation of the nanoparticles revealed small 50–400 nm aggregates (nominally 21 nm particles) (Sun et al., 2007) and adsorption isotherms indicated that the metals were rapidly adsorbed onto the particles with equilibrium

Ecotoxicology of Manufactured Nanoparticles

295

reached within 30 minutes. When exposed to metal contaminated water in the presence of TiO2 nanoparticles metal accumulation in the fish rose sharply to 132% and 146% for arsenic(V) and cadmium, respectively, compared to metal contaminated waters alone. Furthermore, Zhang et al. (2007) showed TiO2 nanoparticles adsorbed 65% more cadmium than 19 µm sediment particles and that the presence of sediment particles did not have a significant influence on the accumulation of cadmium in the fish over the 25-day exposure period. A comparison of the toxic effects of TiO2 and ZnO nanoparticles to D. magna showed that ZnO was considerably more toxic than TiO2 (Adams et al., 2006a, 2006b). Only 40% mortality was observed in D. magna exposed to TiO2 in spring water for eight days but ZnO concentrations of 0.5 and 0.2 mg/l resulted in 100% and 73% mortality, respectively. In this study no effects of primary particle size were observed using particles nominally ranging from 66 to 44 µm and DLS and optical microscopy characterisation confirmed significant aggregation of the particles in water. Franklin et al. (2007) compared the toxic effects of nanoparticulate ZnO, bulk ZnO and ZnCl2 to the freshwater alga Pseudokirchneriella subcapitata. Particle characterisation using transmission electron microscopy and DLS techniques showed that particle aggregation was significant in a freshwater system, resulting in flocs ranging from several hundred nanometers to several microns (Figure 7.4). As discussed previously in Section 7.3.3.2, rapid dissolution of both nanoparticulate and bulk ZnO was observed and all the observed toxicity could be attributed solely to dissolved zinc. As the primary agents driving biogeochemical change, microorganisms are potential mediators of nanoparticle transformations which could affect their mobility and toxicity (Wiesner et al., 2006). Furthermore, microbial ecotoxicology studies may play an important role in elucidating cytotoxicity mechanisms that can be extrapolated to eukaryotic cells (Wiesner et al., 2006). Metal oxides are well known for their antibacterial activity (Sawai, 2003). A number of investigators have considered the potential for enhanced bactericidal effects from nanoparticulate metal oxides, with zinc oxide being the most widely studied. Yamamoto (2001) studied the antibacterial effects of ZnO particles ranging from 100 to 800 nm (0.85–26.0 m2/g) to both Gram-negative and Gram-positive bacteria (Escherichia coli and Staphylococcus aureus respectively). Very high concentrations (400–1000 mg/l) of ZnO were used in these experiments; the aim being to assess their use as potential biocides rather than any detrimental environmental effects, but nevertheless antibacterial effects were shown to increase with both increasing ZnO concentration and decreasing particle size. The effects were similar for both E. coli and S. aureus with the most pronounced effect observed for E. coli. The mechanism of antibacterial activity was assumed to be generation of hydrogen peroxide (H2O2) from the ZnO surface (Yamamoto et al., 1998). Adams et al. (2006a, 2006b) also presented evidence that the antibacterial effects of nanoparticulate metal oxides were due, at least in part, to the generation of reactive oxygen species by UV activation. The toxicity of TiO2 and SiO2 to both E. coli and Bacillus subtilus was greater in the presence of light than in the dark; however, toxicity was observed under dark conditions indicating that a mechanism other than ROS generation must contribute to the antibacterial activity of these materials.

296

Environmental and Human Health Impacts of Nanotechnology

Further evidence for the bactericidal effects of ZnO nanoparticles was obtained by Brayner et al. (2006). These authors synthesised ZnO in di(ethylene glycol) (DEG) medium by forced hydrolysis of zinc acetate (Feldmann, 2003) with a diameter of 10.8 ± 2.2 nm in that medium. The growth of E. coli was inhibited by 85% on Luria-Bertani (LB) medium agar plates containing 125 mg/l ZnO nanoparticles, although no particle characterisation was performed for the particles suspended in LB medium. TEM studies of E. coli cells exposed to 80 mg/l ZnO nanoparticles indicated sub-lethal changes in the cell morphology, cell membrane damage and changes to the appearance of the cytoplasm. The latter was attributed to the internalisation of the nanoparticles but no chemical characterisation was undertaken to confirm the composition of the electron dense areas of cytoplasm. Reddy et al. (2007) showed that 13 nm ZnO particles synthesised and characterised in DEG medium (Feldmann, 2003) completely inhibited the growth of E. coli at 275 mg/l and S. aureus at 80 mg/l on LB plates. Flow cytometry assays were used in parallel to demonstrate a substantial loss in cell viability and membrane integrity in nanoparticle exposed cells. Makhluf et al. (2005) investigated the inhibitory activity of MgO nanoparticles ranging from 8 ± 1 nm to 23 ± 2 nm to both E. coli and S. aureus in LB broth. These particles were aggregated in the test medium with DLS measurements revealing an agglomerate size of 350 nm, but antibacterial activity was still size dependent with the smaller 8 nm particles being most toxic. A mechanism was proposed involving the break up of soft agglomerates at the cell surface and subsequent internalisation of individual nanoparticles (Makhluf et al., 2005). E. coli was more sensitive to MgO nanoparticles than S. aureus but both organisms were more inhibited in physiological saline than in LB. In saline, even the largest 23 nm particles caused >99% reduction in viability within 2 h indicating that protein in the growth medium alleviated the antibacterial activity of the nanoparticle. These authors suggest that the toxic mechanism is via the production of H2O2 either intracellularly (Sawai et al., 1996, 1998) or at the cell surface causing membrane damage, or both. Evidence for both mechanisms is presented; the former by elevated intracellular magnesium concentrations measured by X-ray microanalysis and the latter via a biomimetic membrane assay showing interaction of the nanoparticles with the lipid bilayer. Thill et al. (2006) showed toxicity of nominally 7 nm CeO2 nanoparticles to E. coli with 5 mg/l CeO2 causing a 50% reduction in cell viability in 3 h. Three types of interaction with the bacterial cells were observed, adsorption, oxidation/reduction and toxicity. TEM images of E. coli cells exposed to CeO2 nanoparticles (12 mg/m2 bacterial surface) showed an outer shell of high electron density corresponding to a layer of nanoparticles around the membrane, but this study did not determine if the nanoparticles were internalised. It was suggested that the mechanism of toxicity arises from intimate contact of the nanoparticles with the cell surface and biotic reduction of Ce(IV) to Ce(III) causing oxidative stress to the cells. 7.5.4

Silver

The antibacterial properties of silver and its salts have been known since ancient times. Silver was used as an aid in wound dressings until the early twentieth century

Ecotoxicology of Manufactured Nanoparticles

297

when its use then declined following the introduction of modern antibiotics. Silver nanoparticles are now finding applications as antimicrobial agents in a wide range of products, including wound dressings, coated medical devices and various ‘sterile’ domestic products such as vacuum cleaners, washing machines, antibacterial kitchenware, socks and air filters. As of February 2008, there were 143 known consumer products that claimed to use nanoparticulate silver (Woodrow Wilson Institute, 2008). However, it is not clear whether the nanoparticulate form enhances the observed antimicrobial effects in many of these products. Sondi and Salopek-Sondi (2004) provided evidence of silver nanoparticle accumulation in the membrane of the bacterium E. coli, causing cell wall pits leading to cell permeability changes and ultimate death. Morones et al. (2005) showed that silver nanoparticles act primarily in three ways against Gram-negative bacteria: (i) nanoparticles mainly in the range 1–10 nm attach to the surface of the cell membrane and drastically disturb its proper function, for example transport and respiration; (ii) they are able to penetrate inside the bacteria and cause further damage by possibly interacting with sulfur- and phosphorus-containing compounds such as DNA; (iii) release of silver ions, which will have an additional contribution to the bactericidal effects. The bactericidal effects of silver nanoparticles have been shown to markedly increase when the surface of the nanoparticle has been oxidised producing chemisorbed ionic silver (Lok et al., 2007). Zero-valent silver nanoparticles which were not subject to partial oxidation were not toxic to E. coli. This study suggests that silver nanoparticle toxicity could be associated with the direct transfer of sorbed Ag+ on the nanoparticle surface to the bacterial membrane (Lok et al., 2007) Pal et al. (2007) investigated the antibacterial properties of differently shaped silver nanoparticles against E. coli, both in liquid systems and on agar plates. Truncated triangular silver nanoplates displayed the strongest biocidal action, compared with spherical and rod shaped nanoparticles and Ag+ (in the form of silver nitrate (AgNO3)). It was proposed that the nanoscale size and the presence of a {111} plane combine to promote this biocidal property. It should be noted, however, that nanoparticles were dosed in the experiments as a fixed mass of silver (typically 1–100 µg). The effect of surface area on toxicity, which will vary between the different shaped particles, was not evaluated. To date, there is only one published report of an ecotoxicity study using higher organisms. This is surprising considering the use of nanoparticulate silver in commercially-available products. Lee et al. (2007b) studied nanosilver toxicity to the embryos of zebrafish. This work was critically reviewed by Luoma (2008). The formulation of the silver nanoparticle was stable; the particles stayed as the ∼11 nm particle throughout the experiments. The tests were conducted on a sensitive developmental stage of the organism and for the entire developmental period from the embryo to adult (which is short in zebrafish; 120 hours). Zebrafish embryos, at the eight cell cleavage stage, were immersed in water of the same salinity as occurs naturally in the egg (1.2 nM sodium chloride) with different concentrations of silver nanoparticles. The cleavage stage is the most sensitive of several stages in the development of the embryo because its eight cells proliferate into many cells which ultimately form the functioning organ systems in the adult organism. This is an extremely sensitive system because disruption of one cell has great implications for

298

Environmental and Human Health Impacts of Nanotechnology

normal development of the embryo into an adult. Single silver nanoparticles were observed to move across the chorion, the protective membranous tissue that protects the embryo from the external environment. The chorionic pores were large: 500–700 nm in diameter. Lee et al. (2007b) also showed that single silver nanoparticles made their way into the inner mass of the embryo. The single silver nanoparticles were retained by the embryo as it matured into the adult and were found embedded in different tissues as they developed, including the retina, brain, heart, gill arches and tail. Maturation of the zebrafish embryo was normal at silver concentrations of 8 ng/l but was affected at higher silver doses. Different abnormalities also occurred as dose increased. Fin abnormalities and spinal chord deformities occurred at 19 ng/l. This was followed by malformation of the heart and swelling of the yolk sac (edema) occurred at the next highest doses. At silver concentrations of between 44–66 ng/l, swelling of the head and eye abnormalities occurred. Both quickly resulted in death. 7.5.5

Copper

Griffitt et al. (2007) studied the acute toxicity of soluble copper and 80 nm copper nanoparticle suspensions (nanocopper) to the zebrafish Danio rerio. Nanocopper was acutely toxic to zebrafish, with a 48-h LC50 concentration of 1.5 mg/l. Rapid aggregation of the copper nanoparticles occurred after suspension in water, resulting in 50–60% of the added mass leaving the water column. The study quantified the release of dissolved copper (which is highly toxic to fish) from the added nanoparticles. The measured dissolved copper concentrations alone were insufficient to explain the mortality observed in the toxicity tests, clearly indicating the direct involvement of the nanoparticles. Histological and biochemical analysis revealed that the gill was the primary target organ for nanoparticulate copper. To further investigate the effects of nanocopper on the gill, zebrafish were exposed to 100 µg/l of nanocopper or to the concentration of soluble copper released by dissolution of the nanoparticles. Under these conditions, nanocopper produced different morphological effects and global gene expression patterns in the gill than dissolved copper, clearly demonstrating that the effects of nanocopper on gill are not mediated solely by the particle dissolution and toxicity of dissolved copper. The authors did not include a bulk particulate copper control, so the importance of particle size remains to be evaluated. The chemical composition of the nanocopper was not defined. Clearly, release of dissolved copper from the nanoparticles would either involve oxidation of elemental copper followed by subsequent dissolution of the ionic copper-containing oxidation products, or release of ionic copper that may be sorbed to the surface of the nanoparticles. 7.5.6

Quantum Dots

Quantum dots (QDs) are semiconductor nanocrystals ranging from 2 to 100 nm diameter with unique optical and electrical properties. Structurally, QDs consist of a metalloid crystalline core and a ‘cap’ or ‘shell’ that shields the core and renders the QD bioavailable. QD cores consist of a variety of metal and metalloid complexes, for instance, indium phosphitle, indium arsenide, gallium arsenide, zinc–

Ecotoxicology of Manufactured Nanoparticles

299

selenium, cadmium–selenium and cadmium–tellurium (Hardman, 2006). One of the most valuable properties of QDs is their intense fluorescence, which renders them useful for biomedical imaging. Fluorescent QDs can be conjugated with bioactive moieties (e.g. antibodies, receptor ligands) to target specific biologic cellular structures, such as DNA and cell membrane receptors (Hardman, 2006). Concern has been raised over the potential cytotoxicity of QDs owing to the toxic nature of the parent materials (Hardman, 2006). For instance, the cytotoxicity of bulk cadmium selenide (CdSe) is well documented (Derfus et al., 2004). Toxic metal ions such as Cd2+ may be released upon oxidation of the quantum dot. Zinc sulfide coatings, which are often applied to QDs, may also oxidise in oxygenated environments, leading to the release of zinc ions and then the oxidation products of the QD core. QD oxidation and release Cd2+ may also be enhanced by illumination with UV light (Derfus et al., 2004). A number of studies have investigated the cytotoxicity of QDs (Derfus et al., 2004; Kirchner et al., 2005; Hardman, 2006). Derfus et al. (2004) have reported that CdSe QDs release Cd2+ ions in aqueous solution and that the concentration of the Cd2+ ions directly correlates with cytotoxic effects to liver hepatocytes. Surface coatings such as zinc sulfide and bovine serum albumin were shown to significantly reduce, but not eliminate, cytotoxicity. To date, no ectoxicity information is available for QDs. Given the high toxicity of cadmium to aquatic and terrestrial organisms it is likely that QDs will exhibit substantial toxicity to test organisms. 7.5.7

Iron

Nanoscale Fe(0) iron particles (commonly referred to as zero-valent iron) represent a new generation of environmental remediation technologies. Nanoscale zerovalent iron particles have large surface areas and readily undergo oxidation. They are effective for the dechlorination and detoxification of a wide variety of common environmental contaminants, such as chlorinated organic solvents, organochlorine pesticides and polychlorinated biphenyls (PCBs) (Zhang, 2003). Zero-valent iron nanoparticles have been used at more than 20 sites for the in situ remediation of groundwater contaminants in pilot or full scale operations (Wiesner et al., 2006). The particles can be injected directly into groundwater or used to detoxify contaminated water in above-ground tanks. These particles are of potential ecotoxicological interest (particularly to soil organisms) given their potential release into the environment. Natural iron hydroxyoxide colloids are ubiquitous in water and soil systems; however, the high reactivity and oxidative potential of the zero-valent iron nanoparticles may pose some issues of toxicity. Oberdorster et al. (2006b) investigated the toxicity of iron nanoparticles with an average particle size of 70 nm to the water flea (Daphnia magna) using standard US EPA test methodology. It was reported that the particles were composed of an iron oxide shell and an elemental iron core. The 48-h LC50 for Daphnia magna was ∼55 mg/l, which was approximately the same as that for bulk iron. Further studies are clearly needed to evaluate the toxicity of zero-valent iron, particularly in soils.

300

Environmental and Human Health Impacts of Nanotechnology

7.6

General Conclusions and Future Directions

As has been shown in this chapter, to date, very few ecotoxicity studies on manufactured nanoparticles have been conducted. With the exception of TiO2, even the most basic aquatic ecotoxicity testing data set recommended for the regulation of new materials (algae, invertebrate and fish) is not available in the scientific literature. In particular, there is a notable absence of terrestrial ecotoxicity data including phytotoxicity. Clearly a lot more work is required to address these deficiencies. Firstly, the development and application of validated toxicity testing protocols that can be applied to new nanomaterials and formulations that enter the marketplace is required. This will lead to the generation of basic toxicity testing data for a range of manufactured nanomaterials allowing some assessment of hazards to be made. Secondly, targeted research to identify the ecotoxicological properties of manufactured nanoparticles needs to be carried out. This work should encompass: i. Research studies that address fundamental questions of how nanoparticle size, shape and reactivity influence ecotoxicity. To address this important question it is essential to conduct comparative studies that have appropriate bulk material controls and also take into account the potential for nanoparticle dissolution to contribute to toxicity. Nanoparticles do not form simple dispersions in solution and may aggregate, dissolve or oxidise. This raises fundamental questions about how we should interpret data from current toxicity tests with nanoparticles and what is the most appropriate metric of dose. ii. Given the widespread observations of nanoparticle aggregation in aqueous solutions, understanding the toxicity of nanoaggregates and how they might differ from stabilised nanoparticles is also important. Interactions between nanoparticles and natural colloids leading to the formation of heterogeneous structures may also occur in real waters and soil profiles which may well change toxicity. iii. The long term effects of low concentrations of manufactured nanoparticles on organisms need to be studied, as they are not characterised by routine toxicity testing, which often use test concentrations that are much higher than typical environmental concentrations and short exposure durations. Long periods of very low dose may be important. Sub-lethal effects studies and an understanding of exposure pathways (e.g. importance of gut absorption of nanoparticles), bioaccumulation and trophic transfer are also needed. iv. Our focus is currently on relatively simple manufactured nanoparticles. As nanotechnology evolves, the challenge will be to understand how the varied range of composite materials and surface modified particles affect toxicity. Ecotoxicologists will need to keep pace with these exciting developments. In our opinion, these challenges are best addressed by multidisciplinary teams in which ecotoxicologists work alongside chemists, physicists, physiologists, life cycle assessment specialists and the nanotechnologists designing these new materials.

Ecotoxicology of Manufactured Nanoparticles

301

Beyond the goals outlined above, which relate to toxicity testing and understanding the responses to manufactured nanomaterials at the cellular level, there lies the need to understand the broader ecosystem effects. For example, what are the potential impacts on primary production, decomposition, nutrient cycling and energy flow and the biota that perform these vital functions. Some of this work is already in progress and no doubt will expand our understanding of the ecotoxicological effects of manufactured nanoparticles.

7.7

References

Adams, L. K., D. Y. Lyon and P. J. J. Alavarez (2006a) Comparative eco-toxicity of nanoscale TiO2, SiO2, and ZnO water suspensions; Water Res., 40, 3527–32. Adams, L. K., D. Y. Lyon, A. McIntosh and P. J. J. Alvarez (2006b) Comparative toxicity of nanoscale TiO2, SiO2, and ZnO water suspensions; Water Sci. Technol., 54, 327–34. Bodar, C. W. M., M. E. J. Pronk and D. T. H. M. Sijm (2005) The European Union risk assessment on zinc and zinc compounds: the process and the facts; Integr Env Assess Manage, 1, 301–19. Borm, P., F. C. Klaessig, T. D. Landry, et al. (2006) Research strategies for safety evalulation of nanomaterials, Part V: dissolution in biological fate and effects of nanoscale particles; Toxicol. Sci., 90, 23–32. Brayner, R., R. Ferrari-Iliou, N. Brivois, et al. (2006) Toxicological impact studies based on Escherichia coli bacteria in ultrafine ZnO nanoparticles colloidal medium; Nano Lett., 6, 866–70. Brown, D. M., M. R. Wilson, W. MacNee, et al. (2001) Size-dependent proinflammatory effects of ultrafine polystyrene particles: a role for surface area and oxidative stress in the enhanced activity of ultrafines; Toxicol App Pharmacol., 175, 191–9. Brunner, T. J., P. Wick, P. Manser, et al. (2006) In vitro cytotoxicity of oxide nanoparticles: comparison to asbestos, silica, and the effect of particle solubility; Environ. Sci. Technol., 40, 4374–81. Cheng, J., E. Flahaut and S. H. Cheng (2007) Effect of carbon nanotubes on developing zebrafish (Danio rerio) embryos; Environ. Toxicol. Chem., 26, 708–16. Chithrani, B. D., A. A. Ghazani and W. C. Chan (2006) Determining the size and shape dependence of gold nanoparticle uptake into mammalian cells; Nano. Lett., 6, 662–8. Colvin, V. L. (2003) The potential environmental impact of engineered nanomaterials; Nature Biotech., 21, 1166–70. CRC (2006–2007) Handbook of Chemistry and Physics Online, 87th edn CRC Press; Boca Raton, FL. Deguchi, S., R. Alargova and K. Tsujii (2001) Stable dispersions of fullerenes, C-60 and C-70, in water. Preparation and characterization; Langmuir, 17, 6013–7. Demchick, P. and A. L. Koch (1996) The permeability of the wall fabric of Escherichia coli and Bacillus subtilis; J. Bacteriol., 178, 768–73. Derfus, A. M., W. C. W. Chan and S. N. Bhatia (2004) Probing the cytotoxicity of semiconductor quantum dots; Nano. Lett., 4, 11–18. Fang, J. S., D. Y. Lyon, M. R. Wiesner, et al. (2007) Effect of a fullerene water suspension on bacterial phospholipids and membrane phase behaviour; Environ. Sci. Technol., 41, 2636–42. Federici, G., B. J. Shaw and R. D. Handy (2007) Toxicity of titanium dioxide nanoparticles to rainbow trout (Oncorhynchus mykiss): gill injury, oxidative stress, and other physiological effects; Aquat Toxicol, 84, 415–30. Feldmann, C. (2003) Polyol-mediated synthesis of nanoscale functional materials; Adv. Func. Mater., 13, 101–7.

302

Environmental and Human Health Impacts of Nanotechnology

Fernandes, T. F., P. Rosenkranz, A. Ford, et al. (2006) Ecotoxicology of nanoparticles (NPs); SETAC Globe, 43–5. Foley. S., C. Crowley, M. Smaihi, et al. (2002) Cellular localisation of a water-soluble fullerene derivative; Biochem. Biophys. Res. Commun., 294, 116–9. Forbes, V. E. and T. L. Forbes (1994) Ecotoxicology in theory and practice. Chapman and Hall, London, 247. Fortner, J. D., D. Y. Lyon, C. M. Sayes, et al. (2005) C60 in water; nanocrystal formation and microbial response; Environ. Sci. Technol., 39, 4307–16. Foucaud, L., M. R. Wilson, D. M. Brown and V. Stone (2007) Measurement of reactive species production by nanoparticle prepared in biologically relevant media; Toxicol. Lett., 174, 1–9. Franklin, N. M., N. J. Rogers, S. C. Apte, et al. (2007) Comparative toxicity of nanoparticulate ZnO, bulk ZnO, and ZnCl2 to a freshwater microalga (Pseudokirchneriella subcapitata): the importance of particle solubility; Environ. Sci. Technol., 41, 8484– 90. Geiser, M., B. Rothen-Rutishauser, N. Kapp, et al. (2005) Ultrafine particles cross cellular membranes by nonphagocytic mechanism in lungs and cultured cells; Environ. Health Persp., 113, 1555–60. Griffitt, R. J., R. Weil, K. A. Hyndman, et al. (2007) Exposure to copper nanoparticles causes gill injury and acute lethality in zebrafish (Danio rerio); Environ. Sci. Technol., 41, 8178–86. Hardman, R. (2006) A toxicologic review of quantum dots: Toxicity depends on physicochemical and environmental factors; Environ. Health Persp., 114, 165–72. Henglein, A. (1998) Colloidal silver nanoparticles: photochemical preparation and interaction with O2, CCl4, and some metal ions; Chem. Mater., 10, 444–50. Hund-Rinke, K. and M. Simon (2006) Ecotoxic effect of photocatalytic active nanoparticles (TiO2) on algae and daphnids; Environ. Sci. Pollut. Res., 1–8. Henry, T. B., F-M. Menn, J. T. Fleming, et al. (2007) Attributing Effects of Aqueous C60 Nano-Aggregates to Tetrahydrofuran Decomposition Products in Larval Zebrafish by Assessment of Gene Expression; Environ. Health Persp., 115, 1059–65. Hunter, R. (2001) Foundations of colloid science; 2nd edn; Oxford University Press; Oxford. Hyung, H., J. D. Fortner, J. B. Hughes and J. H. Kim (2007) Natural organic matter stabilizes carbon nanotubes in the aqueous phase; Environ. Sci. Technol., 41, 179–84. Kang, S., M. Pinault, L. D. Pfefferle and M. Elimelech (2007) Single-walled carbon nanotubes exhibit strong antimicrobial activity; Langmuir, 23, 8670–3. Kirchner, C., T. Liedl, S. Kudera, et al. (2005) Cytotoxicity of colloidal CdSe and CdSe/ZnS nanoparticles; Nano. Lett., 5, 331–8. Klabunde, K. J., J. Stark, O. Koper, et al. (1996) Nanocrystals as stochiometric reagents with unique surface chemistry; J. Physi. Chem., 100, 12142–53. Klaine, S. J., P. J. J, Alvarez, G.E Batley, et al. (2008) Nanomaterials in the environment: behaviour, fate, bioavailability and effects. Environ. Toxicol. Chem., 27, 1825–51. Kloepfer, J. A., R. E. Mielke and J. L. Nadeau (2005). Uptake of CdSe and CdSe/ZnS quantum dots into bacteria via purine-dependent mechanisms; App. Environ. Microbiol., 71, 2548–57. Lee, J. D., J. D. Fortner, J. B. Hughes and J. H. Kim (2007a) Photochemical production of reactive oxygen species by C60 in the aqueous phase during UV activation; Environ. Sci. Technol., 41, 2529–35. Lee, K. J., P. D. Nallathamby, L. M. Browning, et al. (2007b) In vivo imaging of transport and biocompatibility of single silver nanoparticles in early development of zebrafish embryos; ACS Nano, 1, 133–43. Li, N., C. Sioutas, A. Cho, et al. (2003) Ultrafine particulate pollutants induce oxidative stress and mitochondrial damage; Environ. Health Persp., 111, 455–60. Lok, C. N., C. M. Ho, R. Chen, et al. (2007) Silver nanoparticles: partial oxidation and antibacterial activities; J. Biol. Inorg. Chem., 12, 527–534.

Ecotoxicology of Manufactured Nanoparticles

303

Lovern, S. B. and R. Klaper (2006) Daphnia magna mortality when exposed to titanium dioxide and fullerene (C60) nanoparticles; Environ. Toxicol. Chem., 25, 1132–7. Lovern, S. B., J. R. Strickler and R. Klaper (2007) Behavioural and physiological changes in Daphnia magna when exposed to nanoparticle suspensions (titanium dioxide, nano-C60, and C60HxC70Hx); Environ. Sci. & Technol., 41, 4465–70. Luoma, S. N. (2008) Silver nanotechnologies and the environment: old problems or new challenges; Woodrow Wilson Institute; Washington; Report PEN 15, 68, pages. Lyon, D. Y., J. D. Fortner, C. M. Sayes, et al. (2005) Bacterial cell association and antimicrobial activity of a C-60 water suspension; Environ. Toxicol. Chem., 24, 2757–62. Lyon, D. Y., L. K. Adams, J. C. Falkner and P. J. J. Alvarez (2006) Antibacterial activity of fullerene water suspensions: effects of preparation method and particle size; Environ. Sci. Technol., 40, 4360–6. Makhluf, S., R. Dror, Y. Nitzan, et al. (2005) Microwave-assisted synthesis of nanocrystalline MgO and its use as a bacteriocide; Adv. Funct. Mater., 15, 1708–1715. Malvern Instruments (2004) Zeta Potential theory; in Zetasizer Nano Series User Manual, 2.1 edn; Worcestershire, UK. Maynard, A. D. (2006) Nanotechnology: A research strategy for addressing risk. Project on Emerging Nanotechnologies; Woodrow Wilson International Center for Scholars; Washington, DC. Maynard, A. D., R. J. Aitken, T. Butz, et al. (2006) Safe handling of nanotechnology; Nature, 444, 267–9. Moore, M. N. (2006) Do nanoparticles present ecotoxicological risks for the health of the aquatic environment? Environ. International, 32, 967–76. Morones, J. R., J. L. Elechiguerra, A. Camacho, et al. (2005) The bactericidal effect of silver nanoparticles; Nanotech, 16, 2346–53. Murashov, V. (2006) Comments on ‘Particle surface characteristics may play an important role in phytotoxicity of alumina nanoparticles’ by Yang, L. and Watts, D. J., (Toxicology Letters, 2005, 158, 122–32); Toxicol. Lett., 164, 185–7. Nel, A., T. Xia, L. Madler and N. Li (2006) Toxic potential of materials at the nanolevel; Science, 311, 622–7. Nowack, B. and T. D. Bucheli (2007) Occurrence, behaviour and effects of nanoparticles in the environment; Environ. Pollut., 150, 5–22. Nurmi, J. T., P. G. Tratnyek, V. Sarathy, et al. (2005) Characterization and properties of metallic iron nanoparticles: spectroscopy, electrochemistry, and kinetics; Environ. Sci. & Technol., 39, 1221–30. Oberdorster, E. (2004) Manufactured nanoparticles (fullerenes, C60) induce oxidative stress in the brain of juvenile largemouth bass; Environ. Health Perspect., 112, 1058–62. Oberdorster, G., E. Oberdorster and J. Oberdorster (2005) Nanotoxicology: an emerging discipline evolving from studies of ultrafine particles; Environ. Health Perspect., 113, 823–39. Oberdorster, E., S. Zhu, T. M. Blickley, et al. (2006a) Ecotoxicology of carbon-based engineered NPs: effects of fullerene (C60) on aquatic organisms; Carbon, 44, 1112–20. Oberdorster, E., P. Larkin and J. Rogers (2006b) Rapid environmental impact screening for engineered nanomaterials: a case study using microarray technology; Project on Emerging Nanotechnologies; Woodrow Wilson International Center for Scholars Report; Washington, DC, USA; 22 pages. OECD (2006) Guidance Document for the Development of OECD Guidelines for Testing of Chemicals, 2nd edn; OECD Environmental Health and Safety Publications, Series on Testing and Assessment No. 1; Organisation for Economic Co-operation and Development; Paris, France. Pal, S., Y. K. Tak and J. M. Song (2007) Does the antibacterial activity of silver nanoparticles depend on the shape of the nanoparticle? A study of the gram-negative bacterium Escherichia coli; App. Environ. Microbiol., 73, 1712–20. Paquin, P. R., J. W. Gorsuch, S. Apte, et al. (2002) The biotic ligand model: a historical overview; Comp. Biochem. Physiol. C-Toxicol. Pharmacol., 133, 3–35.

304

Environmental and Human Health Impacts of Nanotechnology

Reddy, K. M., K. Feris, J. Bell, et al. (2007) Selective toxicity of zinc oxide nanoparticles to prokaryotic and eukaryotic systems; Appl. Phys. Lett. 90, 213902, 1–3. Roberts, A. P., A. S. Mount, B. Seda, et al. (2007) In vivo biomodification of lipid-coated carbon nanotubes by Daphnia magna; Environ. Sci. Technol., 41, 3025–9. Rogers, N., N. Franklin, S. Apte and G. Batley (2007) The importance of physical and chemical characterization in nanoparticle toxicity studies; Integr. Environ. Assess. Manage., 3, 303–4. Rothen-Rutishauser, B. M., S. Schürch, B. Haenni, et al. (2006) Interaction of fine particles and nanoparticle with red blood cells visualized with advanced microscope techniques; Environmental Science & Technology, 40, 4353–9. Sawai, J. (2003) Quantitative evaluation of antibacterial activities of metallic oxide powders (ZnO, MgO and CaO) by conductimetric assay; J Microbiol Meth, 54, 177–82. Sawai, J., E. Kawada, F. Kanou, et al. (1996) Detection of active oxygen generated from ceramic powders having antibacterial activity; J. Chem. Eng. Japan, 29, 627–33. Sawai, J., S. Shoji, H. Igarashi, et al. (1998) Hydrogen peroxide as an antibacterial factor in zinc oxide powder slurry; J Ferment. Engi., 86, 521–2. Sayes, C. M., J. D. Fortner, W. Guo, et al. (2004) The differential cytotoxicity of watersoluble fullerenes; Nano. Letter, 4, 1881–7. Sayes, C. M., A. M. Gobin. K. D. Ausman, et al. (2005) Nano-C60 cytotoxicity is due to lipid peroxidation; Biomaterials, 26, 7587–95. Shukla, R., V. Bansal, M. Chaudhary, et al. (2005) Biocompatibility of gold nanoparticle and their endocytotic fate inside the cellular compartment: a microscopic overview; Langmuir, 21, 10644–54. Smith, C. J., B. J. Shaw and R. D. Handy (2007) Toxicity of single walled carbon nanotubes to rainbow trout, (Oncorhynchus mykiss): respiratory toxicity, organ pathologies, and other physiological effects; Aquat. Toxicol., 82, 94–109. Sondi, I. and B. Salopek-Sondi (2004) Silver nanoparticles as antimicrobial agent: a case study on E coli as a model for Gram-negative bacteria; J. Colloid Interface Sci., 275, 177–82. Stoimenov, P. K., R. L. Klinger, G. L. Marchin and K. J. Klabunde (2002) Metal oxide nanoparticles as bactericidal agents; Langmuir, 18, 6679–86. Stone, V., T. F. Fernandes, A. T. Ford and N. Christofi (2006) Suggested strategies for the ecotoxicology testing of nanoparticles; Mater. Res. Soc. Symp. Proc., 895, 3.1–6. Stumm, W. and J. J. Morgan (1995) Aquatic Chemistry. Chemical Equilibria and Rates in Natural Waters, 3rd edn., John Wiley & Sons Inc., New York. Sun, H. W., X. Z. Zhang, Q. Niu, et al. (2007) Enhanced accumulation of arsenate in carp in the presence of titanium dioxide nanoparticles; Water Air Soil Pollut, 178, 245–54. Tang, Y. J., J. M. Ashcroft, D. Chen, et al. (2007) Charge associated effects of fullerene on microbial structural integrity and central metabolism; Nano Lett., 7, 745–60. Templeton, R. C., P. L. Ferguson, K. M. Washburn, et al. (2006) Life-cycle effects of singlewalled carbon nanotubes (SWCNTs) on an estuarine meiobenthic copepod; Environ. Sci. Technol., 40, 7387–93. Thill, A., O. Zeyons, O. Spalla, et al. (2006) Cytotoxicity of CeO2 Nanoparticle for Escherichia coli; Physico-chemical insight of the cytotoxicity mechanism; Environ. Sci. Technol, 40, 6151–6. Tong, Z. H., M. Bischoff, L. Nies, et al. (2007) Impact of fullerene (C-60) on a soil microbial community; Environ. Sci. Technol., 41, 2985–91. Unfried, K., C. Albrecht, L. Klotz, et al. (2007) Cellular responses to nanoparticles: target structures and mechanisms; Nanotox., 1, 52–71. Wang, I. C., L. A. Tai, D. D. Lee, et al. (1999) C60 and water-soluble fullerene derivatives as antioxidants against radical-imitated lipid peroxidation; Journal of Medicinal Chemistry, 42, 4614–20. Warheit, D. B., R. A. Hoke, C. Finlay, et al. (2007) Development of a base set of toxicity tests using ultrafine TiO2 particles as a component of nanoparticle risk management; Toxicol. Lett., 171, 99–110.

Ecotoxicology of Manufactured Nanoparticles

305

Waychunas, G. A., C. S. Kim and J. F. Banfield (2005) Nanoparticulate iron oxide minerals in soils and sediments: unique properties and contaminant scavenging mechanisms; J. Nanopart. Res., 7, 409–33. Wiesner, M. R., G. V. Lowry, P. Alvarez, et al. (2006) Assessing the risks of manufactured nanomaterials; Environ. Sci. Technol., 40, 4336–45. Woodrow Wilson International Centre for Scholars; Pew Charitable Trusts (2008) Project on emerging nanotechnology: analysis of nanotechnology-based consumer products based on major materials, http://www.nanotechproject.org/inventories/consumer/ analysis_draft/ Xu, X-H. N., W. J. Brownlow, S. V. Kyriacou, et al. (2004) Real-time probing of membrane transport in living microbial cells using single nanoparticle optics and living cell imaging; Biochem., 43, 10400–13. Yamamoto, O. (2001) Influence of particle size on the antibacterial activity of zinc oxide; Intern. J. Inorg. Materials, 3, 643–6. Yamamoto, O., M. Hotta, J. Sawai and H. Kojima (1998) Influence of powder characteristic of ZnO on antibacterial activity – Effect of specific surface area; J. Ceram. Soc. Japan, 106, 1007–11. Yang, L. and D. J. Watts (2005) Particle surface characteristics may play an important role in phytotoxicity of alumina nanoparticles; Toxicology Letters, 158, 122–32. Zhang, W. X. (2003) Nanoscale iron particles for environmental remediation: an overview; J. Nanopart. Res., 5, 323–32. Zhang, X. Z., H. W. Sun, Z. Y. Zhang, et al. (2007) Enhanced bioaccumulation of cadmium in carp in the presence of titanium dioxide nanoparticles; Chemosphere, 67, 160–6. Zhu, S., E. Oberdörster and M. L. Haasch (2006) Toxicity of an engineered nanoparticle (fullerene, C60) in two aquatic species, Daphnia and fathead minnow; Mar. Environ. Res., 62, S5–9. Zhu, X., L. Zhu, Y. Li, et al. (2007) Developmental toxicity in zebrafish (Danio rerio) embryos after exposure to manufactured nanomaterials: buckminsterfullerene aggregates (nC60) and fullerol; Environ Toxicol Chem, 26, 976–9.

8 Exposure to Nanoparticles Robert J. Aitken, Karen S. Galea, C. Lang Tran and John W. Cherrie Institute of Occupational Medicine, Edinburgh, United Kingdom

8.1

Introduction

Nanotechnology is a rapidly evolving and expanding discipline and has aroused growing media and public interest. Articles appear daily in the scientific and popular press and on a host of websites dedicated to the field. New companies, often spin outs from university research departments, are being formed and are finding no shortage of investors willing to back their ideas and products. New materials are being discovered or produced and astonishing claims are being made concerning their properties, behaviours and applications. While much of the current ‘hype’ is highly speculative, there is no doubt that worldwide, governments and major industrial companies are committing significant resources for research into the development of nanometre scale processes, materials and products. In Europe, the Seventh Framework Programme has nanotechnology as one of its seven main thematic programmes (www.cordis.lu). The programme, ‘Nanotechnology and nanosciences, knowledge-based multifunctional materials and new production processes and devices’ has a budget of some €4500 million for the period 2006–2013 (http://cordis.europa.eu/fp7/cooperation/nanotechnology_en. html). A similar large scale programme, the National Nanotechnology Initiative (NNI), is running in the United States with a budget of approximately $1000 million for 2005 (http://www.nano.gov/) and similar in future years. In the United Kingdom, the Department of Trade and Industry (DTI) has run the ‘Micro and nanotechnology manufacturing initiative’ with a budget of more than £90 million (http://www.dti.gov.uk/nanotechnology/). Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

308

Environmental and Human Health Impacts of Nanotechnology

Development of new nanomaterials is a major theme of all of these programmes. Ordinary materials such as carbon or silicon, when reduced to the nanoscale, often exhibit novel and unpredictable characteristics, such as extraordinary strength, chemical reactivity, electrical conductivity or other characteristics that the same material does not possess at the micro or macro-scale. A huge range of materials has already been produced including nanotubes, nanowires, fullerene derivatives and other nanoscale materials. Nanotechnologies are gaining in commercial application. Nanoscale materials are currently being used in electronic, magnetic and optoelectronic, biomedical, pharmaceutical, cosmetic, energy, catalytic and materials applications. Areas producing the greatest revenue for nanoparticles are reportedly chemical-mechanical polishing, magnetic recording tapes, sunscreens, automotive catalyst supports, biolabelling, electro-conductive coatings and optical fibres (http://www.nano.gov/html/ facts/appsprod.html). Many well known industrial processes produce materials that have dimensions in the nanometre size range. One example is the synthesis of carbon black by flame pyrolysis, which produces a powdered form of carbon with a very high surface to mass ratio. While primary particles are generally in the 10–300 nanometer range, carbon black products are placed into commerce (the final product) as agglomerates, which are much larger in size (100–1000 nanometers in diameter (ICBA, 2008a). Worldwide production of carbon black was approximately 8.1 million tonnes in 2005 (ICBA, 2008b). Other common materials produced by flame pyrolysis or similar thermal processes include fumed silica (silicon dioxide), ultrafine titanium dioxide (TiO2) and other ultrafine metals such as nickel. Other industrial processes create and use nanoparticles as part of the process. An example of this is thermal spraying and coating, where a coating material (usually metal) is vaporised in a gas flame or plasma and deposited as a thin film onto a surface to improve its hardness or corrosion resistance. Elsewhere, nanoparticles are an undesirable by-product of industrial processes. The most obvious example of this is welding, which can generate large quantities of nanoparticles usually in the form of a well defined plume of aggregated nanoparticles. Particles in the nanometre size range are also produced in large quantities from diesel engines and from domestic activities such as gas cooking. Nanometre sized particles are also found in the atmosphere where they originate from combustion sources (e.g. traffic, forest fires), volcanic activity and from atmospheric gas to particle conversion processes such as photochemically driven nucleation. In fact, nanoparticles are the end product of a wide variety of physical, chemical and biological processes, some of which are novel and radically different, while others are quite commonplace. It is widely acknowledged that there is a lack of information concerning the human health and environmental implications of manufactured nanomaterials and concerns have been expressed regarding potential risks to health that might arise during their manufacture, use and disposal (European Commission, 2004). There is a view that the biological activity of nanoparticles, which have both beneficial and potentially adverse effects, tends to increase as their size decreases. Evidence for other particle types (e.g. ‘low toxicity dusts’) clearly shows that the toxicity of these

Exposure to Nanoparticles

309

materials is often dependent on their specific surface area rather than their composition (Tran et al., 2000). Epidemiological evidence from industrial processes, such as the manufacture of carbon black, where workers may potentially be exposed to nanoparticles, also indicates potential respiratory health issues (Gardiner et al., 1992). The large specific surface area, crystalline structure, reactivity and exotic properties of some nanoparticles, coupled with what appears to be an imminent shift away from laboratory based development to industrial manufacture, strongly indicates a need for a clearer understanding of the risks associated with these materials. Since the publication of the UK Royal Society/Royal Academy of Engineering report in 2004 (RS/RAEng, 2004), there has been a significant increase in research activity relating to the potential risks to health and the environment from nanomaterials. Much of this work has been concerned with the hazardous nature of these materials and this is discussed in detail in other chapters in this book. Up to this point, very little of this activity has addressed the question of exposure. What is meant by exposure in this context? Exposure is generally considered in terms of contact between a potential stressor, in this case nanoparticles, and an individual population or environment. Exposure is usually measured in terms of its intensity and duration (or frequency). As is well known, an understanding of exposure is critical in risk assessment and management, along with an understanding of hazard, which is considered elsewhere in this book. Knowledge of plausible exposure levels and duration enables realistic interpretation of those response relationships. Control of exposure (to zero) effectively removes the risks from nanomaterials. Without exposure there is no risk. Given the importance of exposure in understanding and controlling risk, it is both surprising and disappointing that until now such little attention has been given to understanding, quantification and assessment of exposure to nanoparticles. Reviewed in this chapter is the potential for exposure to aerosols from processes involving the deliberate development and manufacture of nanoparticle products. In particular the following have been considered: • • • • • •

terminology and characteristics of nanoparticles; potential routes for human exposure; industrial sources of occupational exposure; levels of exposure; means of, and effectiveness of, control measures; potential numbers of people exposed.

8.2

Physical Characteristics and Properties of Nanoparticles

8.2.1 Terminology and Definitions In occupational health the importance of particle size has been well understood for some time. In relation to the transport of aerosols in a workplace, the probability that the aerosol will be inhaled, the region of the respiratory tract that they will reach and whether or not they will deposit, all depend on particle size. This has led

310

Environmental and Human Health Impacts of Nanotechnology

to established criteria for the measurement of occupational exposure to airborne particles, based on measurement of particle size. International standards (CEN, 1993; ISO, 1992; ACGIH, 1993) have been developed which define the size of particles capable of penetrating to the various regions of the respiratory tract. The three most commonly used standards relating to occupational exposure are the inhalable fraction (representing that fraction of aerosol which can enter the respiratory tract), the thoracic fraction (particles capable of penetrating to the airways below the larynx defined according to a selection curve with a 50% cut-off at 10 µm (10 000 nm)) and the respirable fraction (particles, defined according to a selection curve with a 50% cut-off at 4 µm (4000 nm), which can penetrate beyond the ciliated airways to the gas exchange region of the lung). These conventions represent size fractions which are much larger than that normally considered relevant to nanometre-diameter particles. These standards define sampling conventions for particle size fractions that are to be used in assessing the possible health effects resulting from inhalation of airborne particles in the workplace. In principle, they are derived from experimental inhalation data for healthy adults. These specifications are stated in terms of mass fractions, but they may also be used when the intention is to evaluate the total specific surface area or the number of particles in the collected material. Although in common use, the terms ultrafine particle, ultrafine aerosol, nanoparticle, nanoparticle aerosols and nanostructured particles, have not been so rigorously defined leading to potential confusion. In relation to environmental pollution studies it has been common to use the term ‘ultrafine particles’, although often the term is not defined. Where definitions are provided there has been broad agreement that ultrafine particles were those with a diameter ‘less than’ 100 nm (Preining, 1998). Generally this was taken to imply the physical diameter of the particles, although it could imply a diffusion diameter, as instruments to measure particles in this size range often use diffusion as a classifying mechanism. Particles are seldom present as a single size (monodisperse) but rather can be represented by a distribution of sizes, which is commonly characterised by median (in terms of mass or number) and a geometric standard deviation. This simplistic definition (less than 100 nm) fails to take account of size distribution. It is not clear, for example, whether the definition implies all particles less than 100 nm, 95% of particles less than 100 nm, a mean (or median) of less than 100 nm or any particles less than 100 nm. Sometimes, the term ‘nominal diameter’ has been used to provide an effective dimension, where information about particle shape was not available or the particles were known not to be spherical. The terminology used by industry can be even less consistent. For example, the website of a company involved in powder handling and processing defines micronising as the ‘production of an ultrafine powder with a top size of 20 microns’ (i.e. equivalent to 20 000 nm). A recent attempt to develop a more structured approach has been published by the British Standards Institution (BSI) (BSI, 2007a). In its ‘Terminology for nanomaterials’ (BSI, 2007a) it defines nanoscale as ‘size range from approximately 1 nm to 100 nm’, a nano-object as a ‘discrete piece of material with one or more external

Exposure to Nanoparticles

311

dimensions in the nanoscale’ and a nanoparticle as a ‘nano-object with all three external dimensions in the nanoscale’. A nanomaterial is a ‘material having one or more external dimensions in the nanoscale or which is nanostructured’, with nanostructured being defined as ‘possessing a structure comprising contiguous elements with one or more dimensions in the nanoscale’. Definitions are also provided for nanorods (‘nano-object with two similar external dimensions in the nanoscale and the third dimension significantly larger than the other two external dimensions’), nanofibre (‘flexible nanorods’) and nanotubes (‘hollow nanorods’). A further key issue to consider is the role of agglomerates. It is useful here to distinguish between aggregates and agglomerates. These terms are not used rigorously in the literature but in ISO (2007) the following definitions are proposed: • Agglomerate: A group of particles held together by relatively weak forces, including Van der Waals’ forces, electrostatic forces and surface tension. • Aggregate: A heterogeneous particle in which the various components are not easily broken apart. Hence, while an aggregate may be considered to be permanent, agglomerates may break up under certain circumstances. The issue is whether an aerosol of primary particles in the nanometre scale that exists as loosely bound agglomerates in the micrometre scale can be considered a nanoparticle aerosol? A strong argument for including these agglomerated aerosols as nanoparticle aerosols is that in the lung they may break up, resulting in primary nanometre scale particles becoming distributed in the respiratory tract. Oberdorster et al. (1994) found that particles in the form of agglomerated TiO2 aerosol had increased toxicity when compared with an aerosol of larger TiO2 particles, even though both aerosols were similar in aerodynamic size distribution. This has important implications for how nanoparticle aerosols are measured or even detected. All of these types of objects are potentially of interest in relation to occupational and other types of exposure. 8.2.2

Nanoparticle Types

The development of new nanomaterials is a rapidly progressing science and it is beyond the scope of this chapter to track all of these developments. However, several excellent summaries are available. A special edition of the Journal of Materials Chemistry (Rao, 2004; Rao et al., 2004) published 47 papers concerning the development of new nanomaterials, including metallic nanoparticles, germanium, ceramic and aluminium oxide nanowires, carbon, silicon and germanium nanotubes, zinc oxide nanocrystals, gold nanowafers and copper oxide nanocubes. Table 8.1, adapted from Jortner and Rao (2002), summarises the main categories of nanoparticle according to their morphologies, material from which they are composed and the type of application in which they may be used. 8.2.2.1

Fullerenes

Fullerenes are a family of carbon allotropes, molecules composed entirely of carbon, in the form of a hollow sphere, ellipsoid, tube or plane. Spherical fullerenes are also

312

Environmental and Human Health Impacts of Nanotechnology

Table 8.1 Nanoparticles – categories and applications (adapted from Jortner and Rao (2002) Nanostructure

Example Material or Application

Fullerenes (incl. nanotubes) Nanowires Quantum dots Other nanoparticles

Carbon metals, semiconductors, oxides, sulfides, nitrides insulators, semiconductors, metals, magnetic materials ceramic oxides, metals

called buckyballs and cylindrical ones are called carbon nanotubes or buckytubes. Fullerenes are similar in structure to graphite, which is composed of stacked sheets of linked hexagonal rings. C60 was the first fullerene discovered (Kroto et al., 1985) and is known as buckminsterfullerene, after architect Buckminster Fuller’s geodesic domes, which the molecule resembles. Another fairly common buckminsterfullerene is C70 but fullerenes with 72, 76, 84 and 100 carbon atoms can also be found. 8.2.2.2

Nanotubes

Carbon nanotubes (CNT), first discovered by Iijima (1991), are a specific type of fullerene. They are similar in structure to C60 but are elongated to form tubular structures, 1–2 nm in diameter. They can be produced with very large aspect ratios (length/diameter) and can be more than 1 mm in length. In their simplest form, nanotubes comprise a single layer of carbon atoms (single molecule) arranged in a cylinder. These are known as single-wall carbon nanotubes (SWCNTs). They can also be formed as multiple concentric tubes (multi-wall carbon nanotubes, MWCNTs) having diameters significantly greater, up to 20 nm, and length greater than 1 mm. CNTs have great tensile strength and are considered to be 100 times stronger than steel whilst being only one sixth of its weight, thus making them potentially the strongest, smallest fibre known. They also exhibit high conductivity, high specific surface area, unique electronic properties and potentially high molecular adsorption capacity (Maynard et al., 2004). Applications that are currently being investigated include polymer composites (conductive and structural filler), electromagnetic shielding, electron field emitters (flat panel displays), super capacitors, batteries, hydrogen storage and structural composites. From an occupational health perspective, thinking of nanotubes as a single group of materials is overly simplistic. In practice they can exist in a wide variety of forms including single and multi-walled, they may appear straight and relatively rigid, or they may become twisted into ropes and balls. They may also be functionalised with a wide range of molecules. However, the large aspect ratios of CNTs, their durability and the desire to produce bulk quantities make them of particular interest. These properties provide parallels to other durable inorganic fibres, such as asbestos, that are known to cause significant adverse health effects when inhaled.

Exposure to Nanoparticles

8.2.2.3

313

Nanowires

Nanowires are small conducting or semi-conducting nanoparticles with a single crystal structure and a typical diameter of a few nanometres and a large aspect ratio. They are composed of metals such as silver, gold or iron, or semi-conductors such as silicon, zinc oxide or germanium. Applications include use as inter-connectors for the transport of electrons in nanoelectronic devices. Most approaches to the fabrication of nanowires are derived from methods currently used in the semi-conductor industry for the fabrication of microchips. Van Zant (2000) provides a comprehensive review of microchip fabrication that makes useful background reading. Typically, they involve the manufacture of a template followed by the deposition of a vapour to fill the template and grow the nanowire. Deposition processes include Electrochemical Deposition and Chemical Vapour Deposition (CVD). The template may be formed by various processes including etching or the use of other nanoparticles, particularly nanotubes. 8.2.2.4

Quantum Dots

Quantum dots, also known as nanocrystals or qdots, are a special class of semiconductor crystal. They may be composed of various types of semi-conductor material including zinc sulfide, lead sulfide, cadmium selenide and indium phosphide. The number of atoms in a quantum dot, which range from 1000 to 100 000, makes it neither an extended solid structure nor a single molecular entity leading to novel electronic, optical, magnetic and catalytic properties. Quantum dots, exhibit distinct ‘quantum size effects’. The light emitted from a quantum dot can be tuned to the desired wavelength by altering the particle size through careful control of the growth steps. Various methods can be employed to make quantum dots, but the most common is the wet chemical colloidal process. 8.2.2.5

Other Nanoparticles

This catch-all category includes a wide range of primarily spherical or aggregated dendritic forms of nanoparticles. Dendritic forms are where spherical or other compact forms of primary particles aggregate together to form chain like or branching structures. Welding fume is the best known example of this type of nanoparticle, plus other materials such as ultrafine carbon black and fumed silica, which are synthesised in bulk form through flame pyrolysis methods. Nanoparticles of this type may be formed from many materials including metals, oxides, ceramics, semiconductors and organic materials. The particles may be composites having, for example, a metal core with an oxide shell or alloys in which mixtures of metals are present. Many of the production processes involve the direct generation of aerosols through gas phase synthesis, similar to flame pyrolysis, but other production processes including wet chemical and attrition methods may be used. This group of particles may be categorised as being less well defined in terms of size and shape, generally larger (although still within what could be considered nanoparticles), and likely to be produced in larger bulk quantities than other forms of nanoparticles.

314

Environmental and Human Health Impacts of Nanotechnology

From an occupational health perspective, the likelihood of aerosol generation and their availability in bulk quantities makes these nanoparticles of particular interest. 8.2.3

Nanoparticle Production Processes

Nanoparticles, even from the same material, can be synthesised using a variety of methods. Different methods are used in order to optimise specific properties of the materials. These properties include, but are not limited to, size (diameter, length, volume), size distribution, symmetry, surface properties, surface coating, purity, ease of manipulation, yield and suitability for scaling up. Methods used for the commercial or deliberate manufacture of nanoparticles may be divided into four main groups. These are: • Gas phase processes including flame pyrolysis, high temperature evaporation and plasma synthesis; • Vapour deposition synthesis; • Colloidal or liquid phase methods, in which chemical reactions in solvents lead to the formation of colloids; • Attrition methods including grinding, milling and alloying. For all of these processes the recovery stage may be quite similar and is likely to comprise mainly powder or slurry handling techniques. 8.2.3.1

Gas Phase Synthesis Methods

Gas phase processes may be used to produce a wide range of materials. Most, but not all, nanoparticle synthesis methods in the gas phase are based on homogeneous nucleation of a supersaturated vapour and subsequent particle growth by condensation, coagulation and capture. Gas phase synthesis methods have been reviewed by several authors, including Kruis and Fissan (1998) and Swihart (2003). Multiple approaches can be used to generate the super-saturated vapour, dependent on the materials (precursors) used and the form of materials to be produced. In general, the formation of the vapour occurs within an aerosol reactor at elevated temperatures. The most straightforward method of achieving super-saturation is to heat a solid and evaporate it into a background gas. This method is well suited for the production of metal nanoparticles in particular. By including a reactive gas such as oxygen, oxides or other compounds of the evaporated material can be produced. This method has also been used to prepare composite nanoparticles and to control the morphology of single component nanoparticle. The precursor materials are introduced into the reactor as solids, powders, liquids or gases. In some cases, the precursors are nanoparticles, produced by a separate process. In the reactor, the precursors are heated and mixed with a carrier gas. The super-saturated vapour is produced by cooling or by chemical reaction or by some combination of these. Cooling may be induced by expansion, mixing with a cooler gas or by heat transfer to the surroundings. Chemical reactions used are usually decomposition reactions.

Exposure to Nanoparticles

315

8.2.3.2 Vapour Deposition Methods Chemical Vapour Deposition (CVD) methods are based on established methods for the manufacture of semiconductors (Van Zant, 2000). These systems have conventionally been used to deposit thin films of silicon and other semi-conductors on to semi-conductor wafers. Vapour is formed in a reaction chamber by pyrolysis, reduction, oxidation and nitridation. Deposited film growth, in several stages beginning with nucleation as the first few atoms, deposit on the surface. These first atoms form islands that spread and coalesce into a continuous film. After this transition film is formed, growth continues until thicker film develops. Areas of growth on the wafer are controlled using various patterning processes (also known as photolithography or photomasking) in which deposition patterns are etched on to the surface layers of the wafers. CVD methods have been used to produce nanoparticles from many different materials including TiO2 and ZnO (Nakaso et al., 2002). However, the most important application is the synthesis of carbon nanotubes where CVD is considered to offer one of the most effective routes for scaling up to industrial production (Singh et al., 2003). 8.2.3.3

Colloidal Methods

The third major group is the colloidal methods. These are well established ‘conventional’ wet chemistry precipitation processes in which solutions of the different ions are mixed under controlled conditions of temperature and pressure to form insoluble precipitates. Colloidal methods provide a simple route to the synthesis of nanoparticles. This approach enables the relatively straightforward production of significant quantities of nanoparticle material at modest capital cost. As with other approaches, much of the recent emphasis has been on the development of more monodisperse particles with better defined shape. The earliest reported colloids were metals. Preparation of metallic colloids dates back several centuries, but scientific investigation of their preparation or properties was first reported by Faraday (1857) in his experiments with gold. Development of colloidal theory, processes and methods have been ongoing since that time. Many comprehensive reviews are available, including Hiemenz and Rajagopalan (1997) and Holmberg (2002), which describe this science in great detail. Nanomaterials produced by colloidal process include metals, metal oxides, organics, and pharmaceuticals. 8.2.3.4 Attrition Methods The final group of methods is that containing the mechanical attrition methods. In contrast to the previous three groups where nanoparticles were built ‘bottom-up’ from individual molecules, in attrition methods nanoparticles are produced top down from larger particles. Size reduction by grinding and milling is a very well established industrial process used to produce progressively finer forms of materials, including minerals such as clay, coal and metals. Production rates of materials can be of the order of tonnes

316

Environmental and Human Health Impacts of Nanotechnology

per hour. Production of the finest grades of material was previously referred to as micronising. Production of particles in the nanometre size range is referred to as ultrafine grinding (Mende et al., 2003) or nanosizing (Merisko-Liversidge et al., 2003). The process involves wet milling in high shear media mills. Mende et al. (2003) used a stirred media mill to produce suspensions of fused carborundum with a median per hole size of 50 nm. The milling chamber comprised rotating perforated plates. Alumina suspensions were also produced. Due to increasing particle–particle interactions in this process, it was necessary to stabilise the suspension by adjustment of the pH to prevent particle recombination. Merisko-Liversidge et al. (2003) used a similar process (described as Nanocrystal Technology) to produce nanometre size drug particles of poorly-water-soluble compounds. Again stabilisation was required to prevent recombination of the particles. Milligramme quantities of the drug were produced although it is understood that higher production rates can be achieved in commercial systems (www.elan. com). Particles with a diameter of 147 nm were produced by this method. 8.2.4

Nanoparticle Behaviour

In general, occupational hygiene has largely focussed on exposure from the inhalable route based on the general belief that this was generally the highest in terms of risk. Hence an understanding of aerosol behaviour is necessary. Aerosol science is a well understood and described field of science that has been investigated and described over the last 100 years or so. Many excellent textbooks are available, including Davies (1966), Fuchs (1964) and Hinds (1999), which describe in great detail the fundamental properties of aerosols, their behaviour, their measurement and their applications; the reader is also referred to Chapter 5 in this volume. Particle size is the principal parameter governing the behaviour of aerosols. Aerosol behaviour is governed primarily by inertial, gravitational and diffusional forces. For particles in the micrometre scale, inertial and gravitational forces dominate. As particle size decreases into the nanometre scale, diffusional forces dominate and particle behaviour is more like a gas or a vapour. In considering how nanoparticle aerosols may differ in behavioural terms from larger aerosols, important aspects include the rates at which particles diffuse, agglomerate, deposit and re-suspend. It is useful to consider the differences between large (inertial) and small (diffusional) particles in relation to aspects of exposure and control. 8.2.4.1

Diffusion

As particle size decreases towards the molecular level, their behaviour is more like that of a gas or vapour (ICRP, 1994). The kinetic behaviour of nanoparticles follows basic laws of gaseous diffusion. Particle diffusion (Brownian motion) occurs because particles suspended in a gas are bombarded by gas molecules causing the particle to move in a random fashion. The rate at which particles diffuse is determined by the diffusion coefficient, which is inversely proportional to the particle diameter. Particles with high diffusion coef-

Exposure to Nanoparticles Table 8.2

317

Coagulation half-life (from Preining, 1998)

Particle diameter (nm)

1 2 5 10 20

Half life 1 g/m3

1 mg/m3

1 mg/m3

1 ng/m3

2.2 ms 12 ms 0.12 ms 0.7 ms 3.8 ms

2.20 ms 12.00 ms 0.12 s 0.7 s 3.8 s

2.2 s 12 s 2 min 11.67 min 63.34 min

36.67 min 3.34 h 33.34 h 8.1 d 43.98 d

ficient have high mobility and mix rapidly in aerosol systems. Nanometre size particles will have much higher mobility than particles in the micrometer scale. This has implications for the ease in which they can be enclosed in systems and the ease in which they can be controlled. In an enclosed system which has a leak, nanometre sized particles would be much more likely to escape than larger particles because of their higher mobility. Therefore, for nanometre particle systems, an enclosure system design needs to provide a higher level of integrity than for micrometer sized aerosol systems. Systems normally used to contain gaseous emissions would be appropriate. Where particles are released into the workplace, atmosphere diffusion will cause migration from a higher concentration to a lower one. In this case nanometre size particles escaping will mix rapidly through the workplace air and will be quickly dispersed. This has both positive and negative aspects. Nanoparticle aerosols will not remain localised so the concentration at the site of the leak will fall rapidly. However, leaking nanoparticles could end up at great distance from the source, potentially leading to larger numbers of individuals being exposed. 8.2.4.2 Agglomeration As a result of diffusion, particles will undergo multiple collisions leading to coagulation, agglomeration/aggregation and growth in size. The rate at which agglomeration occurs depends primarily on the particle number concentration and their mobility; both of these factors increase as particle size decreases. The agglomeration half-life of different concentrations of nanoparticles of various sizes is shown in Table 8.2. It can be seen from this that very small nanoparticles (e.g. 1 nm) coagulate rapidly even at low mass concentrations. However, the outcome of this coagulation is still a nanoparticle, albeit a slightly large one. In turn this will coagulate and grow further. For larger nanoparticles the coagulation half-life is much longer and therefore the growth is slower. While this mechanism will lead to rapid coagulation and, therefore, very short lifetimes of very small nanoparticles, larger nanoparticles will persist for longer times.

318

Environmental and Human Health Impacts of Nanotechnology

8.2.4.3

Deposition

Particles may be removed from the atmosphere by depositing onto floors, walls and other surfaces. The gravitational settling velocity of the particle is proportional to its diameter. Therefore, airborne nanoparticles will fall out much more slowly than larger particles and gravitation settling will not be an effective removal process. This, potentially, would have the implication of causing higher and longer exposures for more workers. However, deposition of particles will occur primarily because of diffusion. The high mobility of nanoparticles means that they will migrate quickly to and become trapped in the boundary layer on all surfaces including walls, ceilings and floors. Deposition of nanoparticles will therefore be independent of orientation but will be diffusion controlled and not sedimentation controlled. Hence, a leakage of nanoparticles would end up widely dispersed and deposited on all surfaces throughout the workplace. This contrasts with larger particles in which any leakage would tend to be more localised and produce relatively localised contamination. Decontamination after a leakage of nanoparticles would therefore be much more difficult than after a leakage of larger particles and so the clean up processes may not be as effective. This could mean that small, widely dispersed deposits of nanoparticle material could remain attached to the surfaces for much longer periods, leading to possible chronic exposure resulting from other routes such as dermal and ingestion exposure. These behavioural aspects all relate to an airborne release of nanometre size particles. In principle, this event could occur within the synthesis process of a gas phase or vapour phase production system. There would be much less likelihood of a release of this type in a liquid phase process. 8.2.4.4

Re-suspension

An alternative scenario is the re-suspension of nanoparticle material during powder handling or powder mixing activities in the recovery stage of the process. In this case, the issue is whether or not a powdered nanoparticle material, which has been aggregated into bulk form, is likely to become re-suspended as a nanoparticle aerosol. Re-suspension of aerosols from bulk powders is extremely complex and not easily predicted on a theoretical basis. Many factors can influence the possibility of re-suspension, including particle size, particle shape, particle charge, the energy used in the activity which causes the re-suspension and the moisture content of the powder. Aerosol particles that contact one another generally adhere and form aggregates or agglomerates. These particles tend to stick together because of attractive Van der Waals’ forces, which act over very short distances relative to particle dimensions. Van der Waals’ forces would also act to keep a particle attached to a surface. The implication of this is that once attached (or agglomerated) small particles would be much more difficult to split up or re-suspend than large particles. In the United Kingdom, the concept of dustiness has been applied to powders. Dustiness is an index of the relative ease by which powder materials can become re-suspended (HSE, 1996) and is assessed by agitating bulk materials and measuring the aerosol which is produced. Lyons and Mark (1994) assessed the dustiness

Exposure to Nanoparticles

319

of several materials including ‘fine’ carbon black and silica flour. Results were assessed in terms of mass in the respirable and inhalable fractions. Both of these were found to be similar in dustiness to other common materials such as animal feeds and plaster. Information on dustiness as a function of particle number would be a more useful basis for comparing nanoparticle powders. Work to extend this approach into the nanoparticle size range has begun but is at an early stage. Maynard et al. (2004) investigated the relative ease with which carbon nanotubes could be re-suspended. They used a fluidised bed to try to generate aerosols from bulk material. They concluded that it was very difficult to generate an aerosol from these materials. This study is discussed in more detail in Section 8.3.4. Schneider and Jensen (2008) investigated the dustiness of nine powders in which there was thought to be a nanoparticle component. These included ‘pigment grade’ TiO2, ‘ultrafine’ TiO2 and fused silica. They used techniques based on the ‘rotating drum’ and ‘single drop’ methods in EN 15051 (CEN, 2006), but adapted to use smaller quantities of test material and measured the particle size of the generated aerosol. For all of the powders they tested they found, for the rotating drum method, a bimodal distribution with one peak in the range 101–219 nm and the other ranging from 1360 to 2564 nm. An exception was pigment grade TiO2, where no particles were detected in the smaller particle mode. The results from the single drum method were very similar. Interestingly, the total number of particles in each mode was similar for all cases, at least within one order of magnitude. Dustiness as quantified by particle number and by the mass-based dustiness index had a large range. The lowest dustiness was found to be for pigment grade TiO2 and the largest for ultrafine TiO2. In the large particle mode, the difference was about a factor 350 based on particle number mode and a factor 280 based on mass. In the small particle mode, ultrafine TiO2 had the highest dustiness while none could be detected for the pigment grade. These findings suggest a correspondingly large difference in exposure potential. In particular, they suggest that preventive measures would have to be much stricter if pigment grade TiO2 were to be replaced by the ultrafine version. Taken together, these two studies suggest that high aspect ratio nanomaterials such as CNT may be relatively difficult to re-suspend, whereas materials such as TiO2 might be relatively easy and be more ‘dusty’ in ultrafine form. However, given the complexity of the area it is too early to make such generalisations with confidence.

8.3 8.3.1

Nanoparticle Exposure Exposure Scenarios

There are multiple scenarios through which humans could become exposed to engineered nanoparticles. Figure 8.1 is taken from the Royal Society/Royal Academy of Engineering report (RS/RAEng 2004) and illustrates a range of possible occupational, environmental and consumer exposure scenarios.

320

Environmental and Human Health Impacts of Nanotechnology

CONSUMER PRODUCT

STORAGE discharge/leakage

WORKERS TRANSPORT discharge/leakage

PRODUCTION Lab/Factory discharge/leakage

WASTE discharge/leakage

Release of fixed nanoparticles/ nanotubes during product lifecycle?

Transport/Diffusion? AIR WATER Transformation/Degradation?

Potential use of nanoparticles in environmental applications (eg remediation of polluted groundwater)

DIET

Transport/Diffusion?

Figure 8.1 Some possible exposure routes for nanoscale materials based on current and potential future applications. (Reproduced from (RS/RAEng 2004) with permission from the Royal Society.)

In an occupational context, exposure to nanomaterials can occur for workers at all phases of material life cycle. During the development of a new material, it is probable that material will be produced under tightly controlled conditions, in typically very small quantities. Nevertheless, the possibility of release into the workplace air exists if appropriate measures for handling are not implemented. Accidental releases due to spills and accidents are also a possibility. Once the material moves into commercial production, exposures can potentially occur during synthesis of the material or in downstream activities such as recovery, packaging, transport, and storage. In these circumstances, the quantities of materials being handled will typically be much larger. Depending on the specific properties of the new material, it may be incorporated subsequently in a range of other products or may be used in other processes as a feedstock material. Again, these scenarios have the potential to result in exposure to workers involved in them. Nanomaterials may also be incorporated, for example, into a composite material which may be subsequently re-engineered or reprocessed (e.g. by cutting, sawing or finishing). Again in these circumstances, the potential for exposure exists. Finally, end-of-life scenarios where the material is disposed of,

Exposure to Nanoparticles

321

perhaps by incineration or some other process such as shredding or grinding, should be considered. Again in these circumstances the potential for exposure to those carrying out these procedures does exist. Hence, it may be easily seen that for a single material there are multiple exposure scenarios, which may or may not occur depending on the details of manufacture, use and disposal of that material. Throughout these scenarios the population exposed, the levels of exposure, the duration of exposure and the nature of the material to which people are exposed are all different. Consumers may therefore become exposed as a result of nanomaterial contamination in air, water or the food chain, or through the use of consumer products containing nanomaterials. Discharge of materials into the environment is feasible as waste or industrial pollution, directly into the air or water systems or due to deliberate release in applications such as remediation of contaminated land. In assessing human exposure in all of these scenarios it is necessary to consider the route of entry into the human body. In occupational exposure most emphasis has (rightly) been placed on exposure by inhalation. More recently, however, emphasis has been put on dermal exposure and exposure by ingestion. For nanomaterials, given their mobility and potential for translocation, it is highly appropriate to consider these other routes. In this section potential occupational, environmental and consumer exposure scenarios are examined in more detail. 8.3.1.1

Exposures in Occupational Scenarios

In all nanoparticle production processes there is a potential for exposure to occur at both the synthesis and recovery phase of the process (Aitken et al., 2004). The nature of the exposure, the likely level and the probability of exposure will differ according to the specific process and the stage of the process. Similarly, the optimum strategy to control exposure, and the efficacy of the control methods used, will differ depending on the specific process. In gas phase processes, nanoparticles are formed as an aerosol inside a reactor vessel. Hence, there are potential risks of inhalation exposure in the event of leakage of product from the vessel, particularly if the system operates at positive pressure. The nature of the aerosol released would be dependent on the point in the process at which leakage occurs. In the initial stages of the process, primary nanoparticles could be released. Later in the synthesis process a more aggregated aerosol (still of nanoparticles) could be released. Ultimately, particles in the form of loosely bound agglomerates could be released into the workplace air. Exposure by inhalation could also occur during product recovery. The recovery method will differ depending on the process. In some gas phase processes, product nanoparticles are collected in bag filters. These are reverse pulsed and the product nanoparticles recollected into a hopper or other receptacle. Ineffective filter systems could result in escape of product into the working environment, particularly where recirculation into the workplace air occurs. During the recovery stage and in any further processing or packing the particles are likely to be in the form of an agglomerated bulk

322

Environmental and Human Health Impacts of Nanotechnology

powder. The likelihood of exposure by inhalation will be dependent on the details of the process and the characteristics of the product. In gas phase processes, there are potential risks of dermal (and ingestion) exposure by touching surfaces contaminated by airborne releases, handling of product during recovery and processing or packing and during maintenance or cleaning of the plant and workplace. In vapour phase processes, for example CVD, nanoparticle formation is on a substrate, so direct release of particles during synthesis into the workplace air is unlikely. Product recovery is likely to involve mechanical removal of particles from deposition substrates. This could be manual or automated. Depending on the level of energy used, this could result in re-suspension of product into the air. While it unlikely that this would be in the form of discrete nanoparticles, agglomerations of nanoparticles, possibly in the respirable size range and so able to enter the respiratory system, may be produced. Any post-recovery processing and packing is likely to be similar to gas phase processing, with similar risks of exposure. Scenarios for dermal and ingestion exposure will be similar to gas phase processes. In colloidal and other chemical processes particle formation is in liquid suspension, so direct exposure by inhalation is unlikely during the synthesis stage. In some of these processes, product recovery is by spray drying in which the product is sprayed into an evaporation chamber. In these situations, airborne exposure is possible in the event of leakage, although it is unlikely to be to primary single nanoparticles but rather to be agglomerates of nanoparticles. Spillage of liquid product in the workplace, followed by evaporation and cleaning, could result in airborne dispersion and inhalation exposure. Again, this is more likely to be to agglomerated materials. In these processes, dermal (and ingestion) exposure could occur as a result of spillage in the workplace. Exposure could be due directly to the suspension or to dried material. Handling of the product during recovery/packing and maintenance and cleaning could also result in dermal exposure. Once potential applications of nanomaterials begin to be considered, the number of exposure scenarios and the likely differences in nature, intensity and level of exposure increase dramatically. In a general sense, use of nanomaterials as a feedstock in producing further materials would result in exposure scenarios typical of powder handling activities. These scenarios would include activities such as bag opening, powder transfer, mixing and so on. Exposures in these scenarios would depend very much on the dustiness of the material, the quantities involved, and the type of control systems in place. It is unlikely that the energy involved in these processes would be sufficient for significant concentrations of single nanoparticles to be released from agglomerated powders (at least on a mass basis). Nevertheless, all of these types of activities give the possibility for respirable dust to be released into the air, which in these circumstances may contain strongly or loosely bound agglomerations of nanoparticles. When nanomaterials become part of a downstream product or formulation, the types of exposure scenarios are then highly dependent on the purpose and use of that product. These may be dramatically different dependent on the product type

Exposure to Nanoparticles

323

and its application. This can be easily imagined by considering two potential applications, both of which are already identified applications for various classes of nanomaterials. The first example is the incorporation of nanomaterials into paint. Nanomaterials used in paints include metals and metal oxides (Aitken et al., 2006). Exposures resulting from use of paint containing nanoparticles within the formulation will be dependent on the method of use (painting). Application of paints is commonly done though a spraying process. This would result in the generation of an aerosol of fine droplets of the solvent material in which the nanomaterial component would be suspended along with other components (e.g. binders, thinners, additives, etc.). Exposure could therefore be to these droplets or to matrices in which the solvent has evaporated leaving a residual aggregate formed from the other components of the paints including any nanomaterial component. The particle size of these objects could be highly variable and may or may not be in a size range considered to be a nanoparticle. While it is unlikely that this scenario will lead to exposure to individual nanoparticles of the component nanomaterial, it is likely that aerosols will be in a size range which can be inhaled into the respiratory tract. Whilst exposure by inhalation would be a principle concern, exposure to the skin and by ingestion through hand-to-mouth contact are also plausible. Exposure could be reduced by various control methods, including use of a ventilated spray booth or by using personal Respiratory Protective Equipment (RPE). The levels of exposure would be entirely dependent on the details of the process (equipment used, quantity of material, object being coated), on the individual applying the coating and the working practices used. So for the same paint applied in a different way, for example by brushing, the exposure characteristics are likely to be very different. In this case the likelihood of generating an inhalable or respirable aerosol would be much lower. Hence, exposure by inhalation is likely to be much less. However, exposure to the skin, for example from holding the paintbrush without wearing a suitable glove, is likely to be much higher. At the current time there are no published data relating to exposures to nanomaterials in scenarios of this type. Nanomaterials can be used in similar applications including powder coating, antimicrobial coatings and dirt repellent coatings. The exposure scenarios in these applications would be similar but the nature, level and duration of exposure could differ greatly. A second example is the use of nanomaterials incorporated into a composite material. Naturally formed nanoparticles such as nanoclays have been used as fillers in composite materials for many years. Now manufacturers are increasingly turning to nanomaterials such as carbon nanotubes, seeking to use their properties of strength and lightness to create stronger lighter composites. For composites, exposures could occur during the manufacturing process or from the powder handling processes described earlier. It is plausible that that following manufacture, the composite material would be cut, shaped, drilled or surface processed prior to its distribution. These finishing activities may also result in exposure due to the release of particles into the air. In these circumstances, it is unlikely that single nanoparticles of the component nanomaterials will be released. It is more likely that aggregates of the nanoparticles, perhaps in the presence of a binder, will be the

324

Environmental and Human Health Impacts of Nanotechnology

predominate form. The size of these aggregates will depend upon the energy used in the finishing process. At this point in time there are no published data from scenarios of this type. Once composites of this type are released into wider usage, for example as a construction material, the potential of exposure to a much larger number of people becomes possible. As a common building material the composite could easily be cut, shaped, heated or otherwise worked, all of which processes give the potential for the release of particles into the air. In all of these circumstances it is very likely that the nature of dust released into the air will be of an agglomerated form, which may or may not have the potential to disassociate when or indeed if it enters the human lung. For occupational exposure, there is wide potential for exposure by inhalation, by the skin or by ingestion for nanomaterials in widely differing forms. However, at the current time there is almost no information about the exposure levels (occupational), the nature of the material to which people are exposed or the duration of exposures. 8.3.1.2

Exposures in Consumer Scenarios

One of the surprising aspects of the development of nanotechnology has been the extent to which nanomaterials have already found their way into consumer products. There is already a wide range of consumer products which have the potential to lead to exposures of consumers to nanoparticles by inhalation, ingestion and through the skin (Chaudhry et al., 2008; Aitken et al., 2006). There is also the possibility of direct exposure to members of the general public through the use of nanomaterials in nanomedicine. However, the regulatory landscape for medicines is rather different and so these applications will not be considered further in this chapter. The Woodrow Wilson Project on Emerging Nanotechnology (PEN) has developed the first publicly available on-line inventory of nanotechnology-based consumer products. This is available as an on-line database (PEN, 2008). As of late 2007, the nanotechnology consumer products inventory contained 580 products or product lines. For each entry, the information is provided on product name, company, manufacturer or supplier, and so on, as well as a hyperlink to the product webpage. Products are grouped according to relevant main categories that are loosely based on publicly available (US) consumer product classification systems. The largest main category is Health and Fitness, with a total of 356 products. The subcategories associated with this main category are illustrated in Figure 8.2. It includes Cosmetics (89 products), Clothing (92), Personal Care (85), Sporting Goods (59), Sunscreen (27) and Filtration (20). Again, products with relevance to multiple categories have been accounted for multiple times. The Cosmetics, Clothing and Personal Care sub-categories are now the largest in the inventory. Other categories include Home and Garden (69), which includes a range of cleaning, coating, sealing and air freshening/deodorising products, and Food and Beverage (68), containing products such as food supplements and additives, and food packaging.

Exposure to Nanoparticles

325

100 80 60 40 20

filtration

sunscreen

sporting goods

personal care

clothing

cosmetics

0

Figure 8.2 Number of products in various categories in the Woodrow Wilson inventory.

Table 8.3

Potential exposure routs for consumer products.

Route

Product Type

Inhalation

Skin care (spray) Sunscreen (spray) Air fresheners Sealants Paints and coatings

Dermal

Skin care (cream) Sunscreen (cream) Sealants Paints and coatings

Ingestion

Food supplements Food additives and colourings Food packaging Health supplements

For many of the product types identified in this database there is the possibility for consumer exposure by a variety of routes. These have been summarised in Table 8.3. From this it may be seen that exposure by inhalation, dermal and ingestion is possible for a wide range of consumer products already in the marketplace. The level of exposure in each case will be dependent on the nature of the product, the way it is used and the frequency of use. In a review carried out on the behalf of DEFRA, Boxall et al. (2007) used a simple modelling approach to estimate possible exposure by inhalation from

326

Environmental and Human Health Impacts of Nanotechnology

spray-on sunscreen. Based on information obtained about the quantity of nanomaterials present in the sunscreen, the suppliers’ instructions regarding usage of the product and making an estimate of the fugitive spray emission being 10% of the total spray (indicating that 90% of the material actually deposits on the surface of the skin), they estimated that an exposure concentration of 3.5 mg/m3 was plausible. In an occupational context this would be a non-trivial exposure concentration, albeit that the duration of exposure to this concentration will be very short. In an earlier part of their report they also estimated approximately 3 g might be applied daily, which would give a skin coverage of about 1 mg/cm2 of skin. Of this, approximately 5% is the nanoparticle ingredient. Although very few materials have an occupational dermal exposure limit, this again would be a non-trivial exposure. These two examples illustrate that apparently non-trivial inhalation and dermal exposure situations can occur in the use of relatively common place products which can now contain nanomaterials. Exposure by ingestion can also occur in case of food supplements and additives and non-intentional exposure by ingestion may arise from food packaging products (Chaudhry et al., 2008). At the current time no coherent attempt has been made to derive estimates of consumer exposure for nanoparticle-containing products. However, the above examples give sufficient cause to suggest that this is now overdue. 8.3.1.3

Environmental Exposure Scenarios

It is clearly conceivable that fugitive emission from processes in which nanomaterials are produced could potentially lead to increased air concentration of these nanomaterials. As well as environmental exposure in these circumstances, it is plausible that the general public would become exposed. It is likely that such emissions would result in plume type dispersion. One interesting application that could lead to the increased exposure of the general population is the use of nanomaterials as a fuel additive. It has been widely reported that cerium oxide has been used as an additive in diesel fuel and that this has been used in a number cities in the UK. Boxall et al. (2007) have estimated potential concentrations of cerium oxide in air, based on assumptions of about the quantity of cerium oxide present in fuel, the uptake of fuel containing cerium oxide and using dispersion models developed and validated and used by the UK Highways Agency. In their assessment, using a mix of traffic type with traffic flow at 40 km/h and at 1000 vehicles per day, they estimated a cerium oxide concentration at a distance of five metres from the road of 0.6 µg/m−3. In relation to occupational exposure this is very low. Their estimate, however, did not take into account standing traffic in congested city centre streets. In these circumstances, it is likely that concentrations would be perhaps several orders of magnitude higher that that reported. Clearly, however, the potential for exposure scenarios of this type to result in significant exposures needs to be further evaluated.

Exposure to Nanoparticles

8.3.2 8.3.2.1

327

Exposure Metrics Inhalation

In most early studies that looked at the health effects of inhaled particles, dust (particle) samples were collected by drawing air through a filter or other media and subsequently analysed off-line to define estimates of exposure, expressed as a concentration in air. For example, in some of the early studies in the coal industry the samples were analysed by counting particles on the filter under a light microscope (Walton and Vincent, 1998). This resulted in an estimate of exposure in terms of particle number concentration, expressed as number of particles per cm3 or per m3 of air. Epidemiology studies in the coal industry later showed a good correlation between pneumoconiosis and mass concentration, expressed typically as mg/m3. Subsequently, the use of Workplace Exposure Limits (WEL), based on mass concentrations, have become the norm for measuring or regulating exposure for most hazardous chemicals or particles (HSE, 2005). A further refinement of this approach is that sampling of particles is based on collection of a biologically relevant aerosol fraction. In this context, biological relevance is characterised by where in the respiratory tract a particle can potentially deposit, and determined as a function of particle size, measured in terms of aerodynamic diameter. The conventions which are used to determine biological relevance are the Inhalable, Thoracic and Respirable fractions defined in Section 8.2.1. The one class of aerosols that does not fall into this category is fibrous aerosols. Fibrous particles such as asbestos or glass fibre are interesting in three respects: they have an extreme shape (aspect ratios); their physical behaviour in the lungs differs substantially from many more compact particles; and they persist for long periods in the lungs following deposition. Although some of the toxic mechanisms associated with asbestos exposure remain unclear, it is known that ill health following exposure is associated with physicochemical properties such as fibre length and surface chemistry, and that the significance of these properties is exacerbated by persistence of the fibres in the lungs. As a result, exposure is not characterised in terms of averaged mass and composition, but rather by the number (concentration) of fibres in the air with a specific shape and composition. What then is an appropriate metric for measurement of exposure by inhalation? As in the earlier coal industry example, an ideal approach is to choose a metric which is correlated with the heath effect of concern and can be relatively easily measured. At the current time, this is not universally agreed. From the toxicology, there are strong indications that possible health effects may be better correlated with specific surface area, rather than with mass concentration. A number of recent studies have shown that the inflammatory response depends on the specific surface area of particles deposited in the proximal alveolar region of the lungs (Tran et al., 2000; Faux et al., 2003). On this basis, particle specific surface area is a better metric than mass for relating the particle dose to the inflammatory reaction. These issues are discussed in more detail in Chapter 9. However, the estimate of specific surface area does depend on the measurement technique used. Depending on this and other factors, measurement can increase

328

Environmental and Human Health Impacts of Nanotechnology

significantly. For example, if a prune is seen as approximately a sphere then it has approximately the surface area of a smooth sphere of the same diameter; but if the surface wrinkles are included then its surface area becomes notably larger. Similarly, with techniques that use adsorption of gas molecules on to the surface of the particle to measure the surface, the smaller the gas molecule chosen, the more it can enter the wrinkles on the particle surface. Nitrogen as the gas molecule for adsorption and measurement of specific surface area may be a good choice, but other gases (or other surface measurement techniques) are worth considering.‘Biologically relevant’ specific surface area is likely to be that accessible to biologically relevant molecules in the body, although this is not a physically meaningful measurement and is, therefore, difficult to measure and standardise. Likely, this hypothetical property will vary between organism and between organ/cell type. While a strong case may be made for using aerosol specific surface area as a dose metric, this may not be universally so. For example, for fibrous aerosols such as asbestos and glass fibre and carbon nanotubes, the currently accepted best approach is to count fibres (i.e. the metric is particle number concentration). Some types of nanoparticle could be considered fibres. Carbon nanotubes, for example, have high aspect ratios and are highly durable. They can be manufactured with lengths well within the range conventionally considered to be a fibre (although their diameters are much smaller than could be detected with current standard counting methods, which are based on optical microscopy). There are also practical issues to consider in selection of a metric. For example, the concentration of nanoparticles might be very small in terms of mass, quite large in terms of specific surface area and huge in terms of particle number. If nanoparticles were present in the air with larger particles then, depending on the size distribution, the mass concentration would be totally dominated by the larger particles and the contribution of the nanoparticles would be almost entirely undetected. In contrast, the number concentration would be totally dominated by the nanoparticle component. Thus, it is apparent that measuring exposures against a mass concentration alone is unlikely to be sufficient for nanoparticles. It is probable that there is a need to consider characterising exposures against specific surface area and number concentration until further information and improved methods are available. For each of these exposure metrics, but particularly in the case of mass concentration, sizeselective sampling will need to be employed to ensure only particles within the relevant size range are sampled. 8.3.2.2

Dermal Exposure

The importance of dermal exposure to hazardous substances continues to increase. The UK. Health and Safety Executive (HSE) estimates that there are around 85 000 cases of work-related skin disease at any one time, with people employed as bricklayers, laboratory technicians and hairdressers being at particular risk (Hodgson et al., 1993). Substances that are considered to be potentially harmful via dermal exposure include pesticides such as methyl parathion and solvents such as carbon tetrachloride. Harmful effects arising from skin exposure may either occur locally

Exposure to Nanoparticles

329

within the skin or alternatively the substance may be absorbed through the skin and disseminated via the bloodstream, possibly causing systemic effects. Because of this there have been considerable efforts made to develop quantitative methods for monitoring skin exposure. All of these techniques essentially measure the mass of material deposited onto the skin, either per body part or per unit area of skin exposed. Difficulties in this area are compounded by a multitude of measuring methods, which are not easily comparable. The development of a conceptual model of dermal exposure has been beneficial in clarifying some of these issues (Schneider et al., 1999). Systemic effects from dermal exposure are, in particular, unlikely to depend only on the mass on the skin but are more likely to be related to uptake through the skin. Uptake is the flux through the skin and depends on the concentration of the substance on the skin, the area exposed and the duration of exposure. Because of this, some authors have suggested that uptake is a more appropriate metric than the mass on the skin (Robertson and Cherrie, 1995). A review of dermal exposure issues carried out by the HSE concluded that there was no evidence to indicate that specific health problems are currently arising from dermal penetration of ultrafine particles (nanoparticles) (HSE, 2000). However, the review conceded that dermal absorption of ultrafine particles has not been well investigated and suggested that ultrafine particles may penetrate into hair follicles, where constituents of the particles could dissolve in the aqueous conditions and enter the skin. Direct penetration of the skin has been reported by Tinkle et al. (2003) for particles with a diameter of 1000 nm, much larger than nanoparticles. It is reasonable to postulate that nanoparticles are more likely to penetrate through the outer layers of the skin – the stratum corneum – but this has not yet been demonstrated. Several pharmaceutical companies are believed to be working on dermal penetration of nanoparticles as a drug delivery route. Any metric proposed to assess dermal exposure to nanoparticles should be biologically relevant and should relate to potential health effects. Further work, including workplace studies and in vitro assessment of penetration, is required. Based on the current level of knowledge, measurement approaches should include assessment of mass and particle number concentration, area exposed and duration of exposure. 8.3.2.3

Ingestion Exposure

In occupational settings very little work has been done up to now on ingestion exposure. Lead is one of the few materials where the ingestion route has received some attention. Lead paint removal activity can produce high ingestion exposures via hand-to-mouth contact and food contamination in certain workplaces. A study has shown that workers involved in the supply and removal of scaffolding can have particularly high blood lead levels as a result of hand contamination and subsequent ingestion (Sen et al., 2002). No research work has been undertaken in relation to ingestion exposure of nanoparticles. Metrics to assess ingestion exposure to nanoparticles should be biologically relevant and (probably) in the first instance should be measures of mass uptake.

330

Environmental and Human Health Impacts of Nanotechnology

In conclusion, scientific evidence, so far, has demonstrated that particle specific surface area and surface reactivity is likely to be the metric of choice to describe the inflammatory reaction to deposited particles in the proximal alveolar region of the lung. For nanoparticles, their potential dispersal to other organs as well as the possibility of exposure by other routes such as dermal or ingestion mean that possible health risks beyond the lung can not be ruled out. Further research to generate vital data on the possible mode action of nanoparticles in the extra-pulmonary system is needed in order to assess realistically the health risks to nanoparticle exposure.

8.3.3 8.3.3.1

Methods of Measuring and Characterising Exposure to Nanoparticles Rationale for Measurement

Good information about levels of exposure in the workplace or other scenarios requires measurements to be made. The types of measurements which are appropriate depend on the purpose of the measurement programme. In PD 6699-2:2007, ‘Guide to safe handling and disposal of manufactured nanomaterials’ (BSI, 2007b), the following are identified as typical activities which a sampling programme may support: • identification of sources of nanoparticle emissions; • assessment of the effectiveness of any control measure implemented; • ensuring compliance with any workplace exposure limit or self-imposed (inhouse) exposure standard; • identifying any failures or deterioration of the control measures which could result in a serious health effect. In practice, each of these tasks will require specific and often different types of instrumentation, and an appropriate study design and strategy. A range of instrumentation is available; this is discussed in Section 8.3.3.2. Possible strategies are discussed in Section 8.3.3.3. 8.3.3.2

Measurement Methods

Current best practice to measure the exposure of an individual to a chemical or other material present as an aerosol is to use a personal sampling device to collect a sample biologically relevant fraction of the aerosol (HSE, 2000). In occupational hygiene, it is common practice to use samplers conforming to the inhalable or respirable convention. Samples collected in this way, usually over a full working shift, are then subsequently assessed either gravimetrically or via chemical analysis to determine the mass and, hence, average concentration over the defined period. These samplers provide an estimate of mass concentration, from which personal exposure may be derived. A range of sampling instruments is available to collect samples for analysis. Usually these are small devices, comprising a selection stage, a filter (the combination often referred to as the sampling head) and a sampling pump mounted on the

Exposure to Nanoparticles

331

torso of a wearer. The sampling head is typically positioned in the ‘breathing zone’, normally on the upper chest, of the wearer and the samples collected referred to as personal samples. The use of personal sampling is widespread, since aerosol concentrations in workplaces can have wide spatial variability and a personal sample represents the closest approximation to actual exposures. One important exception is the class of fibrous particles such as asbestos or glass fibre. Although some of the toxic mechanisms associated with asbestos exposure remain unclear, it is known that ill health following exposure is associated with physicochemical properties such as fibre length and surface chemistry and the persistence of the fibres in the lungs. As a result, exposure is not characterised in terms of averaged mass and composition, but rather by the number (concentration) of fibres in the air with a specific shape and composition (WHO, 1997). This method relies on collection of fibres in air onto a filter and manual counting by optical microscopy to determine exposure and hence concentration. A key issue in methods to assess exposure to such materials is the definition of a fibre. The World Health Organisation (WHO) method defines a fibre as an object with lengths greater than five microns, a width less than three microns and a length to width to ratio (aspect ratio) greater than 3 : 1 (WHO, 1997). The method is based on determination of the airborne fibre number concentrations by phase contrast optical microscopy. Air samples are collected on a membrane filter by means of a sampling pump, the filter is mounted on a microscope slide and is rendered transparent. Fibres on a measured area of the filter are counted visually using phase contrast optical microscopy and the number concentration in the volume of air is calculated. The analysis relies on manual fibre counting of a relatively small number of fibres and the method is recognised as one of the least precise analytical techniques used in the occupational environment (HSE, 1998). The count is subject to a number of systematic and random errors as well as individual counter bias including limit of detection issues. The minimum visible width depends on the resolving power of the microscope, the difference in the refracted index between the fibre and the medium and visual acuity of the analyst. The WHO method states that with a good correctly adjusted microscope conforming to the specification of the WHO method, the limit of visibility is about 0.13 to 0.15 × 10−6 m. However, in practice, the smallest visible fibres will be about 0.2 to 0.25 microns wide. Since some fibres will fall below the limit of visibility, the count represents only a certain proportion of the total number of fibres present. Thus, the count represents only an index of the numerical concentration of fibres and is not an absolute measure of the number of fibres present. It is clear from the discussion of exposure metrics that for nanometre size aerosols measurement of mass is not sufficient. Particle number would be a more appropriate metric than mass, though ideally the preferred metric may be particle specific surface area. Hence, an ideal sampler to measure biologically relevant exposure to nanoparticle aerosols would be a personal sampling device which collects a relevant size fraction and provides either an instantaneous measure of the sample specific surface area or which facilitates the off-line analysis of the sample to provide a measure of specific surface area. Unfortunately, there are no systems that currently provide a solution of this type.

332

Environmental and Human Health Impacts of Nanotechnology

One important issue to consider is the upper size limit of the particles to be collected or measured. The actual size cut that should be used for assessing exposure which relates to potential human health impact will depend on particle behaviour and subsequent biological interactions. A simplistic view would be to collect only nanoparticles, which would imply that exclusion of all particles larger than 100 nanometres be appropriate. Two factors, however, mitigate against this simplistic approach. Firstly, it seems highly unlikely that particles which are just larger than the 100 nanometre limit are likely to have any significant differences in terms of potential health impact from particles which are just smaller than 100 nanometres. Selection of 100 nanometres is therefore somewhat arbitrary. Secondly, in many of these scenarios, which are described in Section 8.3.2, it is clear that nanoparticles are unlikely to be present in the air as single discrete particles. More likely is that these particles will be formed in agglomerates or attached to other materials or matrices which are likely to result in individual entities which are greater than 100 nanometres. At the current time it is unclear whether the biological impact of discrete particles depositing within the respiratory tract is similar to or different from the impact of larger conglomerate of the same particles. In terms of specific specific surface area, three or four nanoparticles clumped together to form an aggregate are likely to have a similar specific specific surface area to three or four individual nanoparticles. In addition, it is conceivable that aggregates of nanomaterials will disaggregate following deposition in the lung resulting to an identical dose to the lung as it would result from the inhalation of four discreet nanoparticles. Taking these issues into account, it seems that any decision at this time to exclude particles greater than 100 nanometres is premature. Methods of measurement of nanoparticles in air have been reviewed by Maynard and Aitken (2007) amongst others. Table 8.4, which is adapted from Maynard and Aitken (2007), provides a summary of the various types of devices and approaches that can used to provide measurement information on nanometre sized aerosols for the estimation of exposure. Additional information on methods is available in in PD ISO/TR 27628 (ISO, 2007) and is discussed in more depth in Chapter 5. Measurement of number concentration. Simple optical particle counters such as the Grimm 1.104 Work-Check or SKC 3886 Handheld Laser Particle Counter (SKC Inc.) have a lower detection size limit that is governed by the wavelength of their light source. In most cases this means that they cannot ‘see’ or count particles less than 300 nm. As a result, these devices have little value in the measurement of nanoparticles. There are, however, several more sophisticated devices which can provide estimates of particle number concentration, some of which have been used in studies investigating nanoparticle exposure levels (Wake, 2001; Maynard et al., 2004). None of these are personal sampling devices. The most widely used type of instrument for detecting and counting nanoparticle aerosols is the Condensation Particle Counter (CPC). These devices operate by condensing a vapour (typically isopropanol, butanol or water) onto particles which have been drawn into the instrument, to grow them to a size range that can be detected optically by a standard optical counter. These devices are produced

Table 8.4

Summary of methods for measuring exposure to nanoparticles.

Metric

Devices

Notes

Number

Condensation Particle Counter (CPC) Scanning Mobility Particle Sizer (SMPS) Electrical Low Pressure Impactor (ELPI) Optical Particle Counter (OPC)

Real-time particle number concentration measurements between >10 nm and 1 µm.

Mass

Specific surface area

Real-time (mobility diameter) measurement of aerosol size distribution – interpreted as number concentration. Real-time size selective (aerodynamic diameter) measurement of size distribution. Size selected samples may be further analysed off-line. Real-time number concentration measurement of particles larger than ∼300 nm in diameter. When used with a CPC, OPCs may offer insight into the number concentration of particles smaller than ∼300 nm. Electron Off-line analysis of aerosol number concentration (and Microscopy other parameters). Size selective Off-line gravimetric or chemical analysis of filter or personal sampler impactor samples. Some cascade impactors have lower stages which sample particles <100 nm. Size selective static Off-line gravimetric or chemical analysis of filter or sampler impactor samples. Some cascade impactors have lower stages which sample particles <100 nm aerodynamic diameter. Tapered Element Sensitive real-time monitors such as the TEOM may be Oscillating useable to measure nanoaerosol mass concentration Microbalance on-line, with a suitable size selector inlet. (TEOM) Scanning Mobility Real-time (mobility diameter) measurement of aerosol Particle Sizer size distribution. Data may be interpreted in terms of (SMPS) aerosol mass concentration, only if particle shape and density are known. Electrical Low Real-time size selective (aerodynamic diameter) measurePressure ment of size distribution. Data may be interpreted in Impactor (ELPI) terms of mass concentration if particle charge and density are known. Size selected samples may be further analysed off-line. Scanning Mobility Real-time (mobility diameter) measurement of size Particle Sizer distribution – interpreted as specific specific surface (SMPS) area concentration (Rogak et al., 1993). Electrical Low Real-time size selective (aerodynamic diameter) measurePressure ment of active specific specific surface area. Size Impactor (ELPI) selected samples may be further analysed off-line. SMPS and ELPI Differences in measured aerodynamic and mobility used in parallel diameter can be further used to estimate surface area. Diffusion Charger Real-time measurement of aerosol active surface area. Active specific surface area scales as particle diameter squared below ∼100 nm, and particle diameter to the power 1 above ∼1 µm. Some commercial devices tailor the response to particle specific surface area depositing in selected regions of the lungs. Electron Off-line analysis of aerosol specific surface area concenMicroscopy tration (and other parameters). Specific surface area is derived form a two-dimensional projection when using transmission electron microscopy.

334

Environmental and Human Health Impacts of Nanotechnology

commercially by several manufacturers. Generally they are small, battery operated and hand-held with some cases claiming lower detection limits of down to 3 nm. One example of this type of device is the TSI Model 3007 Condensation Particle Counter which is a hand-held device, highly portable, with a claimed size range window of 10–1000 nm and a concentration range of 0–100 000 particles per ml (www.tsi.com). Claims of manufacturers for instrument measurement ranges must be treated with caution and validated prior to use. If this type of validation and calibration is not performed adequately, collected data is likely to be of poor quality. One of the difficulties in using CPC to measure nanoparticles in occupational scenarios is that they cannot discriminate between particles produced by a workplace source and those particles arising from combustion processes or natural sources which are generally present in the ambient atmosphere. Unless the workplace is isolated from the external environment and all air entering the working is filtered, for example in a clean room, then particle counts in indoor workplaces are likely to be similar to those in the outside atmosphere. Other sources on the workplace, for example heaters, may also contribute towards the overall count. These difficulties may be overcome to some extent by the choice on an appropriate sampling strategy (Section 8.3.4). Measuring particle number concentration in isolation can, however, be misleading. In all particle number concentration measurements, the integration limits over which a particular instrument operates are critical in interpreting the reported results. CPC instruments become increasingly insensitive to particles smaller than 10–20 nm. Concentrations measured with instruments with different sensitivities might therefore differ substantially, particularly if the particle count median diameter is close to or in this range. A further limitation of CPC devices is the lack of size information. Instruments that provide both size and number information are, not surprisingly, larger, more complex and more expensive. The most commonly used instrument of this type is the scanning mobility particle sizer (SMPS). Devices of this type (www.tsi.com) are apparently capable of measuring aerosol size distribution from 3 to 800 nm, although not simultaneously over the complete range. The size distribution is expressed in terms of particle mobility diameter. The SMPS operates by charging particles and separating them based on their mobility passing between electrodes. Separated particles are then counted to give size range of mobilities. Application of these devices in occupational hygiene investigations has been limited due to their lack of mobility, expense and complications in use. Measurement of particle number concentration for high aspect ratio nanoparticles (HARN). In the same way, CPCs do not provide any information about particle shape and indeed may give biased data if (nano)particles are permeable or nonspherical. CPCs, therefore, cannot discriminate between HARN and compact nanoparticle forms and indeed may give misleading data. Off-line assessment of nanoparticle number concentrations is feasible by collecting particles on to suitable media and analysis (counting) by electron microscopy in an analogous approach to that used for counting asbestos fibres (WHO, 1997), although this approach has to

Exposure to Nanoparticles

335

not been validated. Despite the difficulties, this is currently the only practicable approach for assessing HARN number concentration. Measurement of mass concentration. An assessment of mass could be based on a size selective personal sampler with a pre-determined cut-off point, either 100 nm, or at some other point in the nanometre size range considered to be more appropriate. The sample could be analysed by weighing or by chemical analysis. There are no commercial devices of this type currently available, but there is no reason in principle why they could not be developed. Particle size selective stages based on diffusion would make appropriate pre-selectors for devices of this type. However, at flow rates typical for current personal samplers (up to five litres per minute) one limitation could be the small mass that would be collected. Given a typical limit of detection for a gravimetric sample collected on a filter of 0.01 mg, the lowest measurable concentration based on a full shift collection would be 0.02 mg/m3. An important issue here relates to what amount of nanoparticle material it would be necessary to measure by this or any other method. To be useful, a measurement method needs to have a limit of detection that is lower than that at which it is considered that health effects might occur based on either toxicological or epidemiological studies. For all new nanoparticles and for many existing nanoparticles, a health effects level has not yet been established. Health effect levels for any nanoparticle are likely to be highly dependent on the specific particle, its morphology, surface, composition and size. Provided that a satisfactory judgement about an effective level could be made, the limitation of detection issue could be overcome by using a high volume static device with an appropriate cut-off point. However, at this point no devices of this type are in common use nor have they been used to assess concentration levels associated with nanoparticle production. A more recent device is the Electronic Low Pressure Impactor (ELPI) sold in commercial form by Dekati (www.dekati.com). It is claimed that this device can measure particle size distribution and concentration in the size range 7–10 nm. In this device, sampled particles are charged and then passed into a low pressure impactor with electrically isolated collection stages. The electrical current carried by the charged particles onto each impactor stage is measured in real time by a sensitive multi-channel electrometer. The particle collection into each impactor stage is dependent on the aerodynamic size of the particles. After the collection period, collected particles can be removed from the impactor stages for further analysis. Other low pressure impactors such as the nanoMoudie (http://appliedphysicsusa.com/moudi.asp) are also available. A major limitation of all of these measurement methods is that they cannot discriminate agglomerates of nanoparticles from single larger particles. As evidence does indicate that agglomerated nanoparticles can have increased toxicity compared to larger particles (Oberdorster et al., 1994), this limits the usefulness of these methods to assess exposure to aerosols of this type. (It is very probable that most nanoparticle aerosols will be agglomerated to some extent.) Measurement of specific surface area concentration. This issue of agglomeration further indicates that measurement of specific surface area is preferable, although

336

Environmental and Human Health Impacts of Nanotechnology

it may be that specific surface area is independent of agglomeration/aggregation. One early example of such a device is the Epiphaniometer (Gaggeler et al.; 1989). In this device the aerosol is passed through a charging chamber where lead isotopes created from a decaying actinium source are attached to the particle surfaces. Once collected on a filter, the amount of radioactivity measured is proportional to the particle specific surface area. This is a complex and difficult instrument which has not found widespread acceptance and use. New specific surface area measurement systems are becoming available however. Wilson et al. (2007) described the use of an Electrical Aerosol Detector (EAD) as an indicator for the total particle specific surface area deposited in the lung. In this device, sampled aerosol is passed through a diffusion charger in which the attachment of ions to particles relates to the total specific surface area of an aerosol. By capturing and measuring this charge in an electrometer, an estimate of the specific surface area can be derived. Shin et al. (2007) showed that the response function of this instrument can be further modified by altering the voltage of an ion trap within the instrument; this selectively removes a portion of the aerosol such that the output provides a measure of the deposited specific surface area in the lung. A commercial device, based on this approach is available from TSI. Measurements of specific surface area are more commonly carried out using bulk methods such as BET (Brauner–Emmett–Teller) analysis. This widely used process depends on gas adsorption using nitrogen, krypton, argon or carbon dioxide gas. The sample sizes used are typically greater than that which might be expected based on occupational hygiene sampling. Nevertheless, this is an important measurement process by which bulk materials may be characterised and, as such, could be used as part of an overall strategy for assessment of exposure. In addition to these processes, there are a number of imaging processes which may be used along with scanning (SEM) or transmission (TEM) electron microscopes to obtain size, shape, structure and, with appropriate detectors and imaging software, quantitative chemical, compositional and morphological information from single aerosol particles or their agglomerates. Conventional SEMs typically have a spatial resolution of 5–10 nm whereas TEMs can resolve sizes below 1 nm. Samples can be collected directly onto filters, filter substrates or impaction substrates. Filters can collect particles smaller than their pore size; however, it is preferable to use filters with a pore size comparable to the smallest particles of interest. Samples may also be collected directly on to SEM supports using electrostatic precipitation or diffusion. Prior to imaging, samples are generally coated with gold or carbon and a commercial sputter coating device that deposits a layer of atoms a few nanometres thick under vacuum over the sample. Various commercial imaging packages are available to facilitate the analysis process. Although it has higher definition, TEM often requires more complex arrangements for sample collection than the SEM. For both SEM and TEM there is an important need to calibrate the loading of the particles onto the sample filter correctly. It is necessary to provide sufficient coverage to allow analysis, but to ensure the particles do not touch or overlap. This can be particularly difficult in cases where samples comprise a wide range of particle sizes.

Exposure to Nanoparticles

337

Another microscopy technique, atomic force microscopy (AFM), also provides sub-nanometre resolution and, unlike EMs, can image nanoparticles in air or water without loss of resolution. AFMs offer significant advantages and some limitations and are a useful additional technique, but are little used in this area. Chapter 6 contains more information about FM and EM methods. 8.3.3.3

Sampling Strategy

All of the measurement methods that may be used clearly fall short of what would be an ideal sampling and measurement system for nanoparticles. However, all have some potential to provide useful information about particular aspects of nanometre size particles, particularly when they are used in combination. The key issue is to develop an appropriate exposure assessment strategy, particular to the process and materials under consideration, which optimises the information available from the various sources. By using suitable combinations of instruments and methods, identification of appropriate surrogate measures and using appropriate models, taking account of determinants of exposure in a structured way and recognising the limitations of all of these, good information concerning the nature of the aerosol and, in particular, how it may change can be obtained. However, there is a pressing need for more research into the development of new and improved measurement methods, combination approaches and the development of generic strategies to provide reliable assessments of exposure to nanoparticles. The US National Institute for Occupational Safety and Health (NIOSH) provides advice on a possible sampling strategy (NIOSH, 2008). Since, at the current time, there is no single sampling method that can be used universally to characterise exposure to nano-sized aerosols, it recommends a multifaceted approach incorporating many of the sampling techniques mentioned above. This follows the approach suggested by Brouwer et al. (2004), which recommends that all relevant characteristics of nanoparticle exposure be measured. The NIOSH approach recommends a step-by-step process, using different instruments at each step so as to maximise the information gained. NIOSH recommends that the first step would involve identifying the source(s) of nanoparticle emissions in a workplace. A CPC would provide acceptable capability for this purpose, being portable and simple to use. It is necessary to determine ambient or background particle counts before measuring particle counts during the process of interest. This may be possible by taking measurements before and after a process is turned on and comparing these with measurements during operation of the process. For identification of specific nanoparticles of interest, area sampling with a filter suitable for analysis by electron microscopy can also be employed. Transmission electron microscopy (TEM) can then be used to identify specific particles and to estimate the size distribution of the particles. NIOSH recommends that once the source of emissions is identified, more extensive investigations may be carried out. This could include aerosol specific surface area measurements conducted with a portable diffusion charger and aerosol size distributions determined with an SMPS or ELPI using static (area) monitoring. It suggests that a small portable specific surface area instrument could be adapted to

338

Environmental and Human Health Impacts of Nanotechnology

be worn by a worker, for example in a backpack, although depending on the nature of the work, this is likely to be cumbersome. The location of these area monitoring instruments would need to be considered carefully. Ideally, they should be placed close to the work areas of the workers but other factors, such as size of the instrumentation, power source, an so on, will need to be considered. As a final stage, NIOSH recommends that personal sampling using filters or grids suitable for analysis by electron microscopy or chemical identification should be employed, particularly if measuring exposures to specific nanoparticles is of interest. Electron microscopy can be used to identify the particles and can provide an estimate of the size distribution of the particle of interest. The use of a personal cascade impactor or a respirable cyclone sampler with a filter, though limited, will help to remove larger particles that may be of limited interest and allow a more definitive determination of particle size. Standard analytical chemical methodologies should be employed to analyse the filters. NIOSH (2008) states, ‘By using a combination of these techniques, an assessment of worker exposure to nanoparticles can be conducted. This approach will allow a determination of the presence and identification of nanoparticles and the characterization of the important aerosol metrics. However, since this approach relies primarily on static or area sampling some uncertainty will exist in estimating worker exposures.’ Clearly, even such a detailed assessment is limited and more detailed schemes are required for research purposes, which might be streamlined for routine monitoring. 8.3.4 8.3.4.1

Studies Investigating Nanoparticle Exposure Exposures in New Nanoparticle Processes

Few studies have as yet directly investigated exposure in new nanoparticle processes. The study by Maynard et al. (2004) is perhaps the single most thoroughly described study available at this time. Maynard carried out a laboratory based study to evaluate the physical nature of the aerosol formed from single-walled carbon nanotubes (SWCNT) during mechanical agitation. This was complemented by a field study in the United States in which airborne and dermal exposure to SWCNT was investigated while handling unrefined material. Details of the field study are provided below. Two techniques for producing SWCNT were investigated. These were the laser ablation and high pressure carbon monoxide (HiPCO) processes. Both of these processes lead to the production of a very low density material comprised of nanometre-diameter catalyst metal particles, CNTs and other forms of elemental carbon which are manually recovered prior to further processing. Measurements of unprocessed airborne nanotube exposures were made at four plants where SWCNT material was removed from production vessels and handled prior to processing. Instruments used included CPCs and SMPSs. Filter samples for electron microscopy analysis were also collected. The filter samples were taken over the time period the workers spent in the enclosure, which was typically about half an hour.

Exposure to Nanoparticles

339

The sample scenarios were as follows: • Laser ablation facility: sampling carried out during material removal and clean up. • HiPCO process removal simulation: removal of SWCNT was simulated by pouring previously generated material between two buckets normally used for collecting nanotubes. Sampling occurred during filling, pouring and clean up activities. • Laser ablation process removal simulation: due to space constraints the collection chamber for the laser process was removed from the production system and placed into the clean air enclosure for powder removal. • HiPCO process: Collection chamber was removed from the production system and placed into the clean air enclosure for powder removal. The samples were analysed for iron and nickel as surrogates for total nanotube product mass, thus providing a low limit of detection while discriminating between SWCNT and other airborne material. SWCNT mass was estimated for each sample assuming a combination of nickel and iron particles constituted 30% of the mass of the material. Estimates of nanotube concentrations ranged from 0.7 µg/m3 in the ablation facility to 53 µg/m3 in the HiPCO process. SEM analysis of filter samples indicated that many of the particles appeared compact, rather than having an open, low density structure more generally associated with unprocessed SWCNT. Some open structures were observed, including some large (non-respirable) clumps. Estimates of the SWCNT material on the individual gloves ranged from 217 to 6020 µg, with most of the material appearing on the parts of the gloves in direct contact with surfaces (inner surfaces of fingers and palms). Although the use of gloves and personal protective equipment (PPE) will minimise dermal exposure during handling of this material, the possibility for large clumps to become airborne and remain so for long periods may lead to dermal exposures in less well protected regions. Measurements indicated higher air and glove SWCNT concentrations for HiPCO material. These higher levels may have been associated with the lower density, ‘fluffier’ HiPCO material becoming more easily airborne as large clumps of material. Inspection of samples showed relatively few particles. Samples from HiPCO SWCNT contained a small number of particles, in the order of 100 µm to 1 mm in diameter, with relatively open ‘nanorope’ structures. However, most micrometresized particles in the analysed HiPCO sample appeared to have a compact structure, with very few nanotubes apparent. In contrast, micrometre-sized particles from the laser ablation process were more clearly comprised of nanoropes. No evidence of millimetre-sized nanotube material clumps were found in aerosol samples from laser ablated material. While laboratory studies have indicated that with sufficient agitation SWCNT material can release fine particles into the air, the aerosol concentrations generated while handling unrefined material in the field at the workloads and rates observed were low in mass terms. In none of the field studies is there any indication that handling the nanotube material leads to an increase in the number concentra-

340

Environmental and Human Health Impacts of Nanotechnology

tion of fine particles, suggesting that released particles tend to be larger then 1 µm or so in diameter. The findings agreed with laboratory data indicating that sub-micrometre particles are not readily released at low levels of agitation. More energetic processes would be expected to release higher concentrations of nanotube aerosol. There is some question over whether the use of a vacuum cleaner during the clean up led to large releases of small nanotube clusters, or whether the observed particles originated from the device itself. This is clearly an area requiring further investigation but at this stage it would seem prudent to use High Efficiency Particulate Arrester (HEPA) filtered cleaners with correctly fitted filter units to prevent the potential release of large nanotube number concentrations. This study provided a first indication of the propensity with which unprocessed SWCNT forms an aerosol during handling. However, given the difficulty of identifying CNTs by SEM and the nature of the surrogate analysis (metals rather than CNTs directly), much work still needs to be done. A recent study (Han et al. 2008) is the first published attempt to measure exposure to CNT using methods similar to those used to measure asbestos. The study measured exposures in the post-production recovery of MWCNT and in a blending activity, which is part of a composite formulation process. An SMPS, which consisted of an electrostatic classifier (model 3080, TSI) equipped with a long-differential mobility analyser (LDMA, model 3081, TSI) and ultrafine CPC (UCPC, model 3025, TSI), was used to monitor the particle size distribution, from 14 to 630 nm. An APS (model 3321, TSI) was used to observe the particle size distribution, ranging from 0.5 to 20 µm in aerodynamic diameter. In addition, a portable aethalometer (model AE42-7-ER-MC, Magee Scientific) was used to measure the mass concentration of carbon particles based on an optical absorption analysis. Air samples were taken by drawing air through mixed cellulose ester filters in sampling cassettes (35 mm diameter, 0.8 µm nominal pore size and 2-inch cowl). The samples were collected in the breathing zone using SKC-117 battery operated sampling pumps at a flow rate of 1.5–2.0 l/min. After mounting on a suitable substrate, the samples were counted using a TEM. All objects, identified as MWCNT with an aspect ratio greater than three were counted. The diameter and length were measured and were 52–56 nm and 1473–1760 nm, respectively. Tests were carried out before and after control measures (essentially isolating the process) were put in place. Most of the laboratory MWCNT exposure levels (maximum 0.43 mg/m3) were lower than the current TLVs for carbon black (ACGIH 3 mg/m3 and Particles Not Otherwise Specified (PNOS) (3.5 mg/m3), although the number of tubular structures (maximum 194 tubes/cm3) was over the current fibre TLVs (asbestos 0.1/cm3; glass wool 1/cm3; rock wool 1/cm3; refractory ceramic fibre 0.1/cm3; etc.). The MWCNT lengths that were shorter than 5 µm may differentiate MWCNTs from asbestos and other fibre structures. Health effects of durable shorter fibres remain controversial and the durability of MWCNTs is not clearly known. Other main biological determinants known for fibres, such as aspect ratio, dimension and deposition, may differentiate MWCNTs from asbestos and other fibre structures.

Exposure to Nanoparticles

341

These exposure results provide strong indications that: (i) conventional exposure monitoring methods, such as personal and area sampling, combined with newly emerging nanoparticle measurement techniques, may be effective in measuring MWCNT exposure concentrations; (ii) MWCNTs can be counted like asbestos and measured for the number of tubes per millilitre using TEM, despite the unavailability of supplementary light microscopy, as suggested by Donaldson et al. (2006), which would be inadequate because of its lack of size resolution; and (iii) conventional engineering control measures work well for MWCNT exposure control. Yeganeh et al. (2008) carried out a study to characterise airborne particle concentrations during the production of carbonaceous nanomaterials, such as fullerenes and carbon nanotubes, in a ‘small’ commercial nanotechnology facility in the United States. They measured fine particle mass concentrations (PM2.5), submicron size distributions and photoionisation potential, an indicator of the particles’ carbonaceous content. There was no attempt to characterise the particles in terms of morphology. Fine particulate matter mass concentrations (aerodynamic diameter <2.5 µm, PM2.5) were measured using a light-scattering aerosol photometer (TSI DustTrak 8520); submicrometer particle number concentrations and size distributions were measured using a scanning mobility particle sizer (TSI SMPS 39301 and CPC 3025A) with long and nano differential mobility analysers (LDMA and NDMA) that measure particles between 14 and 673 nm and 4–160 nm, respectively; photoionisation potential, an indicator of surface chemical composition, was measured using a photoelectric aerosol sensor. The PAS’ response has been shown to be proportional to polycyclic aromatic hydrocarbon, elemental carbon, and/or sooty content of the particles and is used here as an indicator of carbonaceous particle composition. Measurements were made at three locations, inside the facility inside the fume hood where nanomaterials were produced, just outside the fume hood and in the background. There was no specific attempt to discriminate between manufactured nanomaterials and naturally occurring or incidental particles. Average PM2.5 and particle number concentrations were not significantly different inside the facility or outdoors. However, large, short term increases in PM2.5 and particle number concentrations were associated with physical handling of nanomaterials and other production activities. In many cases, an increase in the number of sub-100 nm particles accounted for the majority of the increase in total number concentrations. Photoionisation results indicate that the particles suspended during nanomaterial handling inside the fume hood were carbonaceous and therefore likely to include engineered nanoparticles, whereas those suspended by other production activities taking place outside the fume hood were not. Measurements made within the fume hood clearly showed that ultrafine particles were aerosolised during handling. However, the engineering controls at the facility appear to be effective at limiting exposure to nanomaterials. Fujitani et al. (2008) measured the physical properties, number concentrations and number size distributions of aerosols in a fullerene factory in Japan. In this factory, the mixed fullerene is extracted by solvent from soot generated by the combustion of hydrocarbon–oxygen mixtures. Mixed fullerenes include C60, approximately 60%, C70, approximately 25%, and other higher fullerenes. The molecular

342

Environmental and Human Health Impacts of Nanotechnology

diameter of C60 is about 0.7 nm; however, the diameter of the mixed fullerene particles is about 20 µm (sic). The factory has a production capacity of 40 000 kg per year. Fullerenes are produced in a closed system, minimising the potential for exposure during production. After the drying process, the fullerene is transported to, and kept in, a storage tank until it is removed by workers and bagged. During this procedure, a vacuum cleaner facilitates the bagging operation. After a certain amount of fullerene is bagged, it is weighed in the same room. The authors used an SMPS and an optical particle counter to measure the particle number size distributions of particles ranging in diameter (Dp) from 10 to >5000 nm in the fullerene factory and used SEM to examine the morphology of the particles. Results were presented as number and volume distributions, including size differentiated times series, enabling comparison of results for a series of activities, non-work, bagging, use of a vacuum cleaner and moderate agitation. For the time series, the results were grouped into four size bands, 10–50 nm, 50–100 nm, 100–2000 nm and >2000 nm. The source, workers and measuring instruments were all within a 1.5 m range. A modal diameter of 25 nm was found in the working area during the non-work period, which the authors considered was probably due to the ingress of outdoor air. The particle number concentration for Dp < 50 nm and the particle volume concentration for Dp > 2000 nm increased during the bagging operation (peaking around 20 000 particles ml−1, compared with a background of around 10 000 particles ml−1). Afterward, both of these concentrations decreased when no workers were present in the room. Peaks were also seen during vacuuming and during agitation (30 000 particles ml−1). The authors considered one possible source of nanoparticles to be an operating vacuum cleaner (in the carbon brushes of the motor). SEM revealed that the coarse particles emitted during bagging and/or weighing were aggregates/agglomerates of fullerenes; although the origin of particles with Dp < 50 nm is unclear. These few studies represent the totality of knowledge on exposures in manufacturing and use of new nanoparticles. 8.3.4.2

Exposures in Existing Nanoparticle Processes

Given the absence of real data on exposure to new nanomaterials it is pertinent to examine data from existing processes where nanomaterials might be used or where nanomaterials are an unwanted by-product of the process. Kuhlbusch and Fissan (2006) measured physical and chemical characteristics of airborne particles (ultrafine, PM1, PM2.5 and PM10) in reactor and pelletting areas during carbon black production. Particle number and mass concentration measurements were conducted in these work areas and at ambient comparison sites at each of the three carbon black plants. The carbon black production line had three process steps: reaction–collection, pelletising–drying and packing. In the first process phase, chemical reactions took place in the reactor to produce the primary carbon black particles during partial combustion or thermal decomposition of hydrocarbons by gas-to-particle conversion. Primary particle sizes can range from 1 to 500 nm, with most produced in the 10–100 nm range. A few to many tens of primary particles

Exposure to Nanoparticles

343

immediately formed highly branched aggregate structures. During the process of collection, these ‘aggregates’ agglomerated further in cyclones and bag houses to much larger entities. Carbon black was packed for shipment either as pellets or as ‘fluffy’ agglomerates. Measurements were conducted for PM10 mass concentrations and for sub-micrometre number size distributions simultaneously within the work areas and outside at the comparison site. Ratios were developed for the simultaneous measurements. Particle size distributions were measured with a SMPS (Platform 3080 series; Differential Mobility Analyzer 3081; Condensation Particle Counter 3025; and model 3077 Aerosol Neutralizer; TSI Inc., Shoreview, Minn.), which determined the particle number size distributions in the size range 15–734 nm. An aerodynamic particle sizer (APS, model 3310; TSI Inc.) was used to measure particle size distributions in the range 0.5–15 µm (dae, aerodynamic diameter). Results were expressed in terms of levels of ultrafine particle number concentrations (UFP, <100 nm). No elevated ultrafine particle number concentrations with respect to ambient were determined in the work areas of Plant 1, intermittently elevated concentrations at Plant 2, and permanently elevated concentrations at Plant 3. The intermittently elevated UFP concentrations in the pelletiser and reactor areas of Plant 2 were considered to be related to nearby traffic emissions. Both work areas of Plant 3 showed elevated UFP concentrations in the pelletiser and reactor areas. In the case of the reactor, which was the only enclosed reactor area investigated among the three facilities, the source of the elevated UFP number concentration was considered by the authors to be most likely attributable to grease and oil fumes from maintenance activities, a conclusion supported by carbon fractionation analysis. The elevated UFP number concentrations in the pelletting area in this same plant were identified by the authors as resulting from leaks in the production line, which allowed particulate matter to escape into the work area. The cause of the high number concentrations and the high mass concentrations was due to emissions stemming from one or more leaks in the production line, specifically, seals on dryers immediate adjacent to the pelletting area, as observed by one of the authors. At the peak size, at 90 nm, the ratio of counts in the pelletting area to the comparison site was 1800, a very significant increase. While the authors rightly state that ‘no carbon black is released in the reactor and pelletting areas from the closed production lines under normal operating conditions’, it is interesting that this significant leakage had not been detected in the plant, prior to this measurement programme. Aitken et al. (2004) reviewed data from ‘existing nanoparticle processes’. This study included data from established industries where materials in the nanoparticle size range have been manufactured for a number of years, such as carbon black and TiO2, nanoparticle by-product processes such as welding and soldering and relevant powder handling processes such as in the pharmaceutical industry. Their data are summarised in Table 8.5. A key issue here is discrimination of the manufactured nanoparticle from the background aerosol. It is not clear whether ambient aerosol counts (e.g. from combustion or traffic) have been subtracted in the studies where number counts have

344

Environmental and Human Health Impacts of Nanotechnology

Table 8.5

Summary of aerosol exposure by industry, process and activity.

Industry

Measurements (range)

Activity

Ref

Carbon Black

3470–49 900 particles/cm3 inhalable
Bagging Manufacture

respirable
Manufacture

3670–21 200 particles/cm3 23 106–70 745 particles/cm3

Bagging Sieving

Wake (2001) Gardiner (1992, 1996) van Tongeren (2000, 2001) Wake (2001) Wake (2001)

4150–16 615 particles/cm3 4998–21 167 particles/cm3 0.1–2.7 mg/m3 3590–11 200 particles/cm3 2820–11 300 particles/cm3 11 700–25 900 particles/cm3 12 000–54 000 particles/cm3 42 000–680 000 particles/cm3 56 000–100 000 particles/cm3 800–3700 µg/m3

Bagging Dust prep Powder handling Wire coat Auto Hand Sintering Refinery top Refinery floor Rotary furnace

Wake (2001)

117 649–>500 000 particles/cm3 65 512–278 170 particles/cm3 11 752–>500 000 particles/cm3

MIG MMA Tinning

Wake (2001)

Nickel powder Precious metal blacks Titanium dioxide Pharmaceutical Thermal coating Zinc refining Aluminium smelter Welding Hand soldering

Naumann (1997) Wake (2001) Wake (2001) Healy (2001)

Wake (2001)

LoD = limit of detection.

been provided. It is interesting to note that the number concentrations from powder handling processes are of a similar magnitude as those which might be considered to be ‘bad’ environmental pollution. Much higher number concentrations were seen in activities such as welding that in powder handling processes.

8.3.5

Numbers of People Potentially Exposed

Nanotechnology is not a single technology. It is a multidisciplinary grouping of physical, chemical, biological engineering and electronic processes, materials, and so on, in which the unifying characteristic is one of size. Up until now, there has been little attempt to estimate the numbers of people who may be exposed occupationally to these materials. For example in the United States, NIOSH is ‘unaware’ of any comprehensive statistics on the number of people employed in all occupations or industries in which they might be exposed to engineered, nano-diameter particles in the production or use of nanomaterials (NIOSH, 2007).

Exposure to Nanoparticles

345

Aitken et al. (2004) attempted to make a first estimate of workers involved in United Kingdom nanotechnology industries. They derived best-estimates based on information available in the public domain coupled with assumptions about links between economic value and employment which they conceded had not been tested. They identified several limitations to their work including that: • Nanotechnology is ‘a multifaceted and malleable group of technologies’ and it is difficult to associate it with specific areas of application. • There is a growing ability to manipulate materials on the nanoscale, most of which will be in combination with non-nanotechnologies. Whilst most companies involved in nanotechnology will market themselves as such it is possible that some companies may not do so to the same extent as others and so might be overlooked. • There are few readily available, comprehensive and up-to-date databases which contain summary company information. • The numbers of companies involved in nanotechnology are increasing constantly; therefore, any estimates will be soon redundant. • Information on numbers of employees per company are not always available in the public domain. • Even when employee information is available, it is often unclear whether the figures cited are for total company employees or those specifically involved in manufacturing / research activities. • It is not always obvious which country company manufacturing sites are located or indeed how many sites a company has. Because of these limitations they indicated that the numbers provided no more than a very broad estimate of the potential and predicted numbers of employees involved in nanotechnology and should not be taken as a definitive figure. Their estimates, which have been widely quoted, were at that time (2004): • Approximately 2000 people employed in the university/research sector and in new nanoparticle companies in activities in which they may potentially be exposed to nanoparticles in some form. • Based on the information available, a maximum of 500 workers are considered to potentially be exposed to nanoparticles through existing ultrafine manufacturing processes. Most of these are involved in the manufacture of carbon black. • Nearly 100 000 individuals may potentially be exposed to fine powders through various powder handling processes, the majority of which being employed by the pharmaceutical industry. It is not possible to say what proportion of these may be exposed to nanoparticles. The authors also emphasised that most exposure to nanoparticles in the United Kingdom will be via the incidental production of nanoparticles through processes such as welding and refining. Their estimate was that over a million workers are potentially exposed to nanoparticle by-products due to these processes. Given the level of investment, it is highly like that the number of people in the university/research sectors and in new nanoparticle companies has grown substan-

346

Environmental and Human Health Impacts of Nanotechnology

tially since that estimate. The authors suggested that it may double over the five years to 2009. The proportion of those involved in existing chemical and pharmaceutical companies and in other powder handling activities exposed to nanoparticles is likely to increase substantially as the use of nanoparticle materials increases. It is reasonable to expect that the balance between numbers exposed in research and production and applications will shift, with greater emphasis and exposures being found during the secondary manufacturing of products. It is also clear that potential occupational and public exposure to manufactured nanoparticles will increase dramatically in the near future due to the ability of nanomaterial to improve the quality and performance of many consumer products the public employs daily, as well as the development of medical therapies and tests which will use manufactured nanoparticles.

8.4 8.4.1

Control of Exposure Introduction

Using chemicals or other hazardous substances at work can potentially put people’s health at risk. In the United Kingdom, the law requires employers to control exposure to hazardous substances to prevent ill health. The framework by which compliance can be achieved and demonstrated is the Control of Substances Hazardous to Health Regulations 2002 (COSHH) (HSE, 2002). This is a scheme of good health and safety management involving eight basic measures that set out in a simple step-by-step approach to help assess risks, implement any measures needed to control exposure and establish good working practices. The major elements in this approach are: • • • •

Identify the hazard Assess the risk Prevent or control the risk Evaluate the effectiveness of control measures.

At the current time there are some major uncertainties that limit the extent to which adequate assessment of the risks arising from exposure to nanoparticles can be made. Risk assessment requires an understanding of the toxic potential of a material and the levels of exposure that are likely to arise in various scenarios where it is used. It is clear from the foregoing that nanoparticles are not a single group of objects but a multiplicity of shapes, sizes and compounds. A unifying feature of nanoparticles is that they will be smaller than the materials that they replace and will have larger specific surface area per unit mass. These parameters are known to increase the toxic potential of a material (Tran et al., 2000). It is clear from this review that little is known concerning the exposures of those working with these materials. There is some reassurance that the limited studies that have been carried out do not suggest that there are high airborne concentrations of nanoparticles in workplaces where they are manufactured, although these

Exposure to Nanoparticles

347

tend not to take account of agglomerated particles. However, it is not easy to be confident that in all workplaces where these materials are being manufactured or used, an adequate assessment of the risks has or will be made. Nevertheless, assessment of risks has to be made on the best available information and control strategies will have to be developed based on these assessments. Strategies to control exposure to nanoparticles may include: • • • • • • • • •

Total enclosure of the process Partial enclosure with local exhaust ventilation Local exhaust ventilation General ventilation Limitation of numbers of workers and exclusion of others Reduction in periods of exposure Regular cleaning of wall and other surfaces Use of suitable PPE Prohibition of eating and drinking in contaminated areas.

Choices for each of these will need to be made based on some understanding of the differences between nanoparticles and larger particles. 8.4.2 8.4.2.1

Inhalation Exposure Engineering Control

For air velocities prevailing in workplaces, airborne nanoparticles can be considered as having no inertia. They will, therefore, behave in a similar way to a gas and if not fully enclosed, will diffuse rapidly and will remain airborne for a long time. Because of their high diffusion velocity, these particles will readily find leakage paths in systems in which the containment is not complete. Engineering control systems designed for use to control nanoparticles, such as enclosures, local ventilation or general ventilation, therefore need to be of similar quality and specification to that which is normally used for gases rather than for particulate challenges. These systems do exist and are in common use in the chemicals and other industries. Like all such systems, however, effective performance is highly dependent on appropriate use and maintenance. No relevant research has been identified that has specifically sought to evaluate the effectiveness of engineering control systems against new nanoparticle challenges. While most of these systems can in principle be used to control exposure, they do not always do so. There is no reason to expect that application of these methods to new nanoparticle generation processes will result in better control than that previously demonstrated in micro-scale powders and in gases. 8.4.2.2

Filtration

Filtration plays an important role in the control of exposure to airborne particles. High Efficiency Particulate Arrester (HEPA) filters are used in engineering control systems to clean the air before returning it to the workplace. These filters are usually referred to as mechanical filters.

348

Environmental and Human Health Impacts of Nanotechnology

Filtration theory is well understood and has been extensively described by several authors (e.g. Brown, 1993). As an aerosol penetrates through a filter, the trajectories of the particles deviate from the streamline due to various well understood mechanisms. As a result, particles may collide with the filter elements (fibres) and become deposited on them. The mechanisms include diffusion, interception, initial impaction and gravitational settling. Electrostatic forces can also play a role in some filter types. For particles less than 100 nm, Brownian diffusion is the dominant mechanism (Lee and Mukund, 2001). Filtration efficiency due to Brownian diffusion increases as particle size decreases. Brownian diffusion is caused by collisions between particles and the air molecules to create random paths that the particles follow. The random motion increases the probability of a particle contacting one of the filter elements. Once the particle is collected onto a surface it will adhere to it due to the Van der Waals’ forces. Therefore, filters are likely to be good collectors of nanoparticles. Current methods for certification of HEPA filters and for respirator filters do not routinely require testing at particle sizes in the nanometre size range. Internationally recognised standards for HEPA filters (DOE, 1998) require that the filter is challenged with an aerosol with a mass median diameter of 300 nm and that the particle collection efficiency is greater than 99.97%. Three hundred nanometres is considered to be a much more penetrating aerosol for these filters than nanometre size particles due to the decrease of Brownian diffusion at this particle size. Similarly, European Standards for respirator filter cartridges (CEN, 2000) and for filtering face pieces (CEN, 2001) require that these systems are tested against sodium chloride aerosols with a mass median diameter of 300 nm. Again, this is based on an expectation that this would be the most penetrating size. However, some authors have suggested that penetration of nanometre particles through wire screens (filters) can deviate from the classical penetration models if the effect of thermal rebound is significant. Wang (1996) used a modified penetration model including the effect of thermal rebound to compare with the experimental results of Ichitsubo et al. (1996). He found good agreement using model parameters derived from literature. Both experiment and theory suggested significant thermal rebound and increased penetration for particles smaller than 2 nm. Similar results have been reported by Otani (2002), who found that the particle rebound may increase the penetration for platinum nanoparticles through circular diffusion tubes when the particles are smaller than 2 nm. 8.4.2.3

Use of Personal Protective Equipment

Use of PPE such as respirators and air-fed devices may be used (as a final option) as a method of control for any airborne hazard. All of these devices depend on filtration as a means of cleaning the air prior to it being breathed by the worker. The discussion relating to filtration applies equally here. It should be expected that, for all but the smallest nanoparticles (<2 nm), the filtration efficiency will be high.

Exposure to Nanoparticles

349

It is well recognised, however, that the determining factor which governs the effectiveness of respiratory protection equipment (RPE) against particulate challenges is not absolute penetration through the filter, but rather face seal leakage which bypasses the device. Face seal leakage is dependent on many factors, including the fit of the mask to the face, duration of wearing, work activity, and so on. Since it is expected that nanoparticle aerosols will have high mobility, it is possible that enhanced leakage will occur although no more than might be expected for a gas. No relevant research to quantify this has been identified. 8.4.3

Dermal Exposure

Based on our understanding of the various processes by which nanoparticles can be synthesised there seems to be a strong possibility that dermal exposure may occur, most likely in the later stages of the process, that is recovery, or resulting from surface contamination. There is some evidence that dermal exposure to nanoparticles may lead to direct penetration of nanoparticles into the epidermis and possibly beyond into the blood stream. Therefore, it may be necessary to introduce control to exclude or limit the level of dermal exposure likely to occur. As with control of exposure by inhalation, the first approach is enclosure of the process. This should certainly be achievable as powder handling processes can be enclosed successfully. However, in practice, particularly with products or processes that are in development, the main emphasis is on investment and expenditure at the synthesis end of the process. This is likely to limit the expenditure on sophisticated control and automation processes to deal with what will be perceived as relatively mundane tasks, such as harvesting and packing of nanomaterials. In any case, even where such processes are in place, the requirements for attention to breakdowns, maintenance and so on means that the possibility of dermal exposure cannot be excluded at all times. In these and other instances protection against dermal exposure typically consists of the use of Skin Protective Equipment (SPE), that is suits, gloves and other items of protective clothing. Even for powders in the macroscale, it is recognised that SPE is very limited in its effectiveness to reduce or control dermal exposure. Based on current understanding (Schnieder et al., 1999) multiple processes contribute to dermal exposure and the relative ineffectiveness of SPE. In addition to the classical view that the failure of SPE results from direct penetration or permeation of an agent through the material from which the equipment is constructed, other failure processes include transfer of substances by direct contact between surface, skin and outer/ inner layers of clothing or gloves, and redistribution of substances between compartments of the same type, e.g. redistribution of contaminants from one part of the skin contaminant layer to another as a result of touching the face with contaminated fingers. Current European testing for certification of PPE against dermal exposure only takes account of permeation or penetration, although new tests have been proposed which take account of the other human factors based on simulations (Brouwer et al., 2005). Since it is likely that nanoparticles which escape into the workplace will become widely dispersed and will have high specific surface area, it is likely

350

Environmental and Human Health Impacts of Nanotechnology

that the human factor element will be even more critical than for larger particles (e.g. particles greater than 500 nm). In this case, it is quite likely that SPE will be less effective against nanoparticles than against larger particles. It is also quite likely that direct penetration of nanoparticles through the material from which the protective clothing is made will also be higher than for larger particles. Preliminary work published by the NANOSAFE project indicates that this is indeed the case. Few details of their experimental approach are available but their recommendation is that penetration is higher and that two sets of gloves should be worn (NANOSAFE, 2008). It has not been possible to identify any relevant research work which has looked at the impact of nanoparticles on the probability of failure of SPE due to human factors. No information on these issues was available from any of the manufacturer’s websites. 8.4.4

Ingestion Exposure

Exposure by ingestion in the workplace is currently not well understood. It is considered that ingestion exposure in the workplace results primarily from hand-tomouth contact. It follows that strategies that tend to reduce dermal exposure in the workplace will also tend to reduce exposure by ingestion. The converse of this is also true. At this point in time no relevant research has been identified that has successfully quantifies exposure to nanoparticles by ingestion in the workplace or the effectiveness of strategies to reduce this exposure.

8.5

Discussion

The development and application of nanoparticles represents a major portion of nanotechnology activity and the number of nanoparticle products continues to grow. The purpose of producing these new materials and products is that their behaviour is expected (and has been demonstrated) to be different in the nanometre scale than in the microscale. Nanoparticles are produced using a wide range of synthesis methods by university and other research groups, small emerging companies and by established major international organisations. The main processes by which nanoparticles are manufactured are gas phase synthesis, vapour deposition, colloidal methods and attrition methods. Although these processes are capable of producing materials with strikingly different properties from bulk forms of the same material, from an occupational hygiene perspective the processes themselves are not dissimilar to conventional well established chemical and engineering processes. In principle, all of these production methods may give rise to exposure by inhalation, through the skin and by ingestion. Exposure by inhalation caused by direct leakage of nanoparticles into the workplace air during synthesis only seems possible in gas phase processes. In vapour deposition, particle growth is on a substrate and in the colloidal and attrition methods particle formation is in the liquid phase, therefore aerosols will not be formed during synthesis. However, for all production methods, product recovery

Exposure to Nanoparticles

351

and subsequent processing are powder handling activities which, if sufficient energy is used, may result in the generation of respirable or inhalable concentrations of agglomerated nanoparticles. It is, however, very probable that these powder handling activities will not generate separated nanoparticles, due to the relatively high energies which would be necessary to break the (Van der Waals’) forces which keep particles agglomerated. Even generation of larger agglomerations of nanoparticles as aerosols may be quite difficult. All of the production processes described could potentially result in dermal exposure, particularly at the powder handling, packaging and bagging stages. Dermal exposure is likely to result in ingestion exposure from hand-to-mouth contact. It has been postulated that nanoparticle exposure to the skin can result in direct penetration through the skin. At least one pharmaceutical company is developing drug delivery systems based on topical application of lipid nanoparticles. It is, therefore, reasonable to conclude that nanoparticles depositing on the skin could potentially penetrate into the epidermis and possibly beyond. As far as the authors are aware toxicological effects arising from dermal exposure to nanoparticles, or from ingestion, have not as yet been investigated or identified. There is evidence to suggest that specific surface area is the most appropriate metric to use to assess exposure by inhalation. This appears to fit best with current toxicological evidence and would deal directly with the issue of agglomeration. Ideally, a personal sampler would be available which could assess this metric. Although there are now methods by which exposure in terms of specific surface area may be measured, these are largely untried. While a strong case may be made for using surface area as a metric, this may not be universally so. For high aspect ratio nanoparticles (HARN), particle number may be more appropriate, as is the case currently for other high aspect ratio particles, such as asbestos fibres. In any case, it is also necessary to consider characterising exposures against aerosol mass and number concentration until further information and improved methods are available. For each of these exposure metrics, but particularly in the case of mass concentration, size selective sampling will need to be employed to ensure only particles within the relevant size range are sampled. That in itself presents a problem, however, since at this point it is not clear at what size the selection should be made. The use of 100 nm is overly simplistic, since there is no reason to expect that a particle with a size of 105 nm would have a significantly different potential to cause harm compared with a particle with a size of 90 nm. A 100 nm size selection could also exclude aggregates of sub-100 nm particles. Again, from a hazard perspective, there is no reason to do this. For dermal exposure and for ingestion exposure, measurements should also be biologically relevant. At this stage there is insufficient evidence to indicate whether mass, number or specific surface area is the most appropriate metric. For dermal exposure, measurement approaches should ideally also consider the skin area exposed and the duration of exposure. At the current time, information about the exposure of workers engaged in the manufacture and use of new nanoparticles is very limited. Only a handful of studies have been published thus far and it is too early to draw any solid conclusions from these. For high aspect ratio nanomaterials, Maynard et al (2004) demonstrated that

352

Environmental and Human Health Impacts of Nanotechnology

handling activities with CNT resulted in measurable airborne exposure concentrations. They did not, however, attempt to quantify these in terms of ‘fibre count concentration’, as would be used for other high aspect ratio particles. They did not conclusively demonstrate that fibre type aerosols were produced in this activity. Indeed, most of the aerosol which they observed was amorphous carbon, catalyst particles or aggregates of CNT. On the other hand, Han et al. (2008) showed clearly that during handling of MWCNT in the production of composites, MWCNT were released into the atmosphere in a form that enabled them to be counted using a method based on current techniques for fibrous aerosols. They also showed, however, when good ventilation control was applied to the process, that release was effectively contained and exposures were negligible. Yeganeh et al. (2008) showed effective control of the release of nanoparticles in a small commercial nanotechnology facility producing fullerenes. No attempt was made to assess the morphology of the particles. Fujitani et al. (2008), however, did observe increases in the airborne concentrations associated with bagging in a fullerene factory, even though the primary product was fullerene aggregates/agglomerates with a diameter of 20 000 nm. Interestingly, Kuhlbusch and Fissan (2006) reported release of sub-100 nm particles in a carbon black production facility, which they attributed to leaks in the ventilation seals. In seems clear that, based on theory and on limited experimental evidence, there are conventional control strategies and systems available which may be used to control exposures in nanoparticle processes. Engineering controls should be able to be designed to provide sufficient levels of containment. Filtration systems including respiratory protective equipment should be effective providing that they are used and maintained correctly. Until relatively recently, less emphasis has been placed on dermal exposure as a route of exposure. One consequence of this is that the effectiveness of control measures to prevent dermal exposures are poorly understood and almost certainly not as effective as approaches to control exposure by inhalation. Studies evaluating the efficacy of gloves and suits to control dermal exposure are beginning to emerge. At the present time there is very little available information about the number of people who are occupationally exposed to engineered nanoparticles. Probably, the number is relatively small. Aitken et al. (2004) estimated that approximately 2000 people are currently employed in the United Kingdom in the university/ research sector and in new nanoparticle companies in activities in which they may potentially be exposed to nanoparticles in some form. It is certainly the case that the number of people in the university/research sectors and in new nanoparticle companies may increase substantially over the next few years. The proportion of those involved in existing chemical and pharmaceutical companies and in other powder handling activities who are exposed to nanoparticles is likely to increase substantially as the use of nanoparticle materials increases. Already, there is significant potential for consumers to become exposed to nanoparticles from a whole series of products, from personal care products to food additives. This exposure can be by inhalation, through the skin or by ingestion, dependent on the product. There are no good estimates of current levels of expo-

Exposure to Nanoparticles

353

sure but the numbers of people potentially exposed is likely to increase dramatically as the number of products increases. Proper assessment and management of the risks associated with exposure to nanoparticles requires good understanding of the toxicological hazards associated with these materials and of the levels of exposure, expressed in an appropriate metric, which are likely to occur. Given the substantial gaps in the exposure information available, it is difficult to see how these risk assessments can be adequate at this time.

8.6

References

Aitken, R. J., K. S. Creely and C. L. Tran (2004) Nanoparticles: An Occupational Hygiene Review, Health and Safety Executive, Sudbury, UK. Aitken, R. J., M. Chaudhry, A. Boxall and M. Hull (2006) Manufacture and use of nanomaterials: current status in the UK and global trends. Occupational Medicine, 56, 300–6. ACGIH (1993) Threshold limit values of chemical substances and physical agents, American Conference of Governmental Occupational Hygienists, Cincinatti, Ohio. Boxall, A., Q. Chaudhry, C. Sinclair, et al. (2007) Current and future predicted environmental exposure to engineered nanoparticles, Central Science Laboratory, Sand Hutton, York. Brouwer, D. H., J. H. Gijsbers, and M. W. Lurvink (2004) Personal exposure to ultrafine particles in the workplace: exploring sampling techniques and strategies. The Annals of Occupational Hygiene, 48, 439–53. Brouwer, D. H., R. J. Aitken, R. Oppl and J. W. Cherrie (2005) Concepts of skin protection: considerations for the evaluation and terminology of the performance of skin protective equipment, Journal of Occupational and Environmental Hygiene, 2, (9), 425–34. Brown, R. J. (1993) Air filtration: an integrated approach to the theory and applications of fibrous filters, Pergamon Press, Oxford. BSI (2007a) PAS 136:2007 Terminology for nanomaterials, British Standards Institution (BSI). BSI (2007b) PD 6699-2:2007 Nanotechnologies – part 2: guide to safe handling and disposal of manufactured nanomaterials, British Standards Institution (BSI). CEN (1993) Workplace atmospheres: size fraction definitions for measurement of airborne particles in the workplace, CEN standard EN 481, CEN, Brussels, Belgium. CEN (2000) EN 141:2000 – Respiratory protective devices. Gas filters and combined filters. Requirements, testing, marking Respiratory protective devices, British Standards Institution. CEN (2001) EN 149:2001 – Respiratory protective devices. Filtering half masks to protect against particles. Requirements, testing, marking, British Standards Institution. CEN (2006) EN 15051 Workplace atmospheres – Measurement of the dustiness of bulk materials – Requirements and reference test methods, British Standards Institution. Chaudhry, Q., M. Scotter, J. Blackburn, et al. (2008) Applications and implications of nanotechnologies for the food sector, Food additives and contaminants, 25 (3), 241–58. Davies, C. N. (1966) Aerosol Science, Academic Press, Oxford. Donaldson K., R. Aitken, L. Tran, et al. (2006) Carbon nanotubes: a review of their properties in relation to pulmonary toxicology and workplace safety. Toxicol Sci., 92 (1), 5–22. European Commission (2004) Nanotechnologies: A preliminary risk analysis on the basis of a workshop organized in Brussels on 1–2 March 2004 by the health and consumer protection directorate general of the European commission. DOE (1998) HEPA filter test program (DOE-STD-3022-98), US Department of Commerce, Technology Administration, Springfield.

354

Environmental and Human Health Impacts of Nanotechnology

Faraday, M. (1857) The Bakerian Lecture: Experimental Relations of Gold (and Other Metals) to Light, Philosophical Transactions of the Royal Society of London (1776–1886), 147 (1), 145. Faux, S. P., C. L. Tran, B. G. Miller, et al. (2003) In vitro determinants of particulate toxicity: the dose metric for poorly soluble dusts, HSE Research Report 154, Health and Safety Executive, Sudbury, UK. Fuchs, N. A. (1964) The mechanics of aerosols, MacMillan, New York. Fujitani, Y., T. Kobayashi, K. Arashidani, et al. (2008) Measurement of the physical properties of aerosols in a fullerene factory for inhalation exposure assessment, Journal of Occupational and Environmental Hygiene, 5 (6), 380–9. Gaggeler, H. W., U. Baltensperger, M. Emmenegger, et al. (1989) The Epiphaniometer, a new device for continuous aerosol monitoring, Journal of Aerosol Science, 20 (5), 557. Gardiner, K., W. N. Trethowan, J. M. Harrington, et al. (1992) Occupational exposure to carbon black in its manufacture, The Annals of Occupational Hygiene, 36 (5), 477–96. Gardiner, K., Calvert, I. A., van Tongeren, M. J. A. and Harrington, JM. (1996) Occupational exposure to carbon black in its manufacture: Data from 1987 to 1992. Annals of Occupational Hygiene, 40: 65–77. Han, J. H., E. J. Lee, J. H. Lee, et al. (2008) Monitoring multiwalled carbon nanotube exposure in carbon nanotube research facility, Inhalation toxicology, 20 (8), 741–9. Healy, J., S. D. Bradley, C. Northage and E. Scobbie (2001) Inhalation exposure in secondary aluminium smelting, The Annals of Occupational Hygiene, 45 (3), 217–25. Hiemenz, P. C. and R. Rajagopalan (eds) (1997) Principles of colloid and surface chemistry, 3rd edn, Marcel Dekker, New York. Hinds, W. (1999) Aerosol technology: properties, behavior, and measurement of airborne particles, 2nd edn, John Wiley & Sons Ltd, Chichester. Hodgson, J. T., J. R. Jones, R. C. Elliott and J. Osman (1993) Self-reported work-related illness in Research Paper 33 HSE Books, Health and Safety Executive, Sudbury, UK. Holmberg, K. (2002) Handbook of applied surface and colloid chemistry, John Wiley & Sons Ltd, Chichester. HSE (1996) Dustiness of powders and materials, MDHS 81, HSE Books, Health and Safety Executive, Sudbury, UK. HSE (1998) Methods for determination of hazardous substances; Fibres in air (MDHS87/1998). HMSO, London. HSE (2000) General methods for sampling and gravimetric analysis of respirable and total inhalable dust, MDHS 14/3, HSE Books, Health and Safety Executive, Sudbury, UK. HSE (2002) Control of substances hazardous to health regulations 2002, 4th edn, HSE Books, Health and Safety Executive, Sudbury, UK. HSE (2005) Workplace exposure limits (EH40/2005), HMSO, London. Ichitsubo, H., T. Hasimoto, M. Alonso and Y. Kousaka (1996) Penetration of ultrafine particles and ion clusters through wire screens, Aerosol Science Technology, 24, 119–28. ICBA (International Carbon Black Association) (2008a) Carbon black, frequently asked questions [Homepage of International Carbon Black Association], Available: http://www. carbon-black.org/faq.html ICBA (International Carbon Black Association) (2008b) What is Carbon Black [Homepage of International Carbon Black Association], Available: http://www.carbon-black.org/ what_is.html ICRP (1994) Human respiratory tract model for radiological protection, ICRP Publication 66, Annals of the ICRP, 24, 1–3. Iijima, S. (1991) Helical microtubules of graphitic carbon, Nature, 354, 56–8. ISO (1992) Air quality: particle size fraction definitions for health-related sampling, CD7708, International Standards Organisation, Geneva. ISO (2007) Workplace atmospheres – ultrafine, nanoparticle and nano-structured aerosols – inhalation exposure characterization and assessment. PD ISO/TR 27628, International Standards Organisation, Geneva. Jortner, J. and C. N. R. Rao (2002) Nanostructured advanced materials. Perspectives and Directions, Pure and Applied Chemistry, 74, 1491–506.

Exposure to Nanoparticles

355

Kuhlbusch, T. A. and H. Fissan (2006) Particle characteristics in the reactor and pelletting areas of carbon black production, Journal of Occupational and Environmental Hygiene, 3 (10), 558–67. Kroto, H. W., J. R. Heath, S. C. O’Brian, et al. (1985) C60: Buckminsterfullerene, Nature, 318 (6042), 162–3. Kruis, F. E. and H. Fissan (1998) Synthesis of nanoparticles in the gas phase for electronic, optical and magnetic applications – a review, Journal of Aerosol Science, 29, 511–35. Lee, K. and R. Mukund (2001) Filter Collection in Aerosol measurement and principles, techniques and applications. John Wiley & Sons Inc., New York, 197–228. Lyons, C. P. and D. Mark (1994) Development and testing of a procedure to evaluate the dustiness of powders and dusts in industrial use. (HSE Contract Research Report No 62/1994), HSE Books, Health and Safety Executive, Sudbury, UK. Maynard, A. D. and R. J. Aitken (2007) Assessing exposure to airborne nanomaterials: Current abilities and future requirements, Nanotoxicology, 1 (1), 26–41. Maynard, A. D., P. A. Baron, M. Foley, et al. (2004) Exposure to carbon nanotube material: aerosol release during the handling of unrefined single-walled carbon nanotube material, Journal of Toxicology and Environmental Health A, 67 (1), 87–107. Mende, S., F. Stenger, J. Peukert and J. Schwedes (2003) Mechanical production and stabilization of submicron particles in stirred media mills, Powder Technology, 132, 64. Merisko-Liversidge, E., G. G. Liversidge and E. R. Cooper (2003) Nanosizing: a formulation approach for poorly-water-soluble compounds, European Journal of Pharmaceutical Sciences, 18 (2), 113. Nakaso, K., M. Shimada, K. Okuyama and K. Deppert (2002) Evaluation of the change in the morphology of gold nanoparticles during sintering, Journal of Aerosol Science, 33 (7), 1061. NANOSAFE (2008) Safe production and use of nanomaterials. http://www.nanosafe.org/ home/liblocal/docs/Dissemination%20report/DR1_s.pdf. Naumann, B. D. and E. V. Sargent (1997) Setting occupational exposure limits for pharmaceuticals, Occupational Medicine (Philadelphia, Pa.), 12 (1), 67–80. NIOSH (2007) Frequently asked questions. Available: http://www.cdc.gov/niosh/topics/ nanotech/faq.html#howmany NIOSH (2008) Approaches to Safe Nanotechnology: An Information Exchange with NIOSH. Available: http://www.cdc.gov/niosh/topics/nanotech/safenano/assessment.html Oberdorster, G., J. Ferin and B. E. Lehnert (1994) Correlation between particle size, in vivo particle persistence, and lung injury, Environmental Health Perspectives, 102 (5), 173–9. Otani, Y. (2002) Generation of nanometer size particles and their removal from air, Advanced Powder Technology, 4, 255. Preining, O. (1998) The physical nature of very, very small particles and its impact on their behaviour, Journal of Aerosol Science, 29 (5–6), 481. PEN (2008) 21/8/08-last update, Project on Emerging Nanotechnologies. Available: http:// www.nanotechproject.org/inventories/consumer Rao, C. N. R. (2004) New developments in nanomaterials, Journal of Materials Chemistry, 14 (4), E4. Rao, C. N. R., A. Müller and A. Cheetham (2004) The Chemistry of Nanomaterials: Synthesis, Properties and Applications, John Wiley & Sons Ltd, Chichester. Robertson, A. and J. W. Cherrie (1995) Biologically Relevant Assessment of Dermal Exposure, Annals of Occupational Hygiene, 39 (3), 387. Rogak S. N., R. C. Flagan and H. V. Nguyen (1993) The mobility and structure of aerosol agglomerates, Aerosol Science and Technology, 18 (1). RS/RAEng (2004) Nanoscience and nanotechnologies: Opportunities and Uncertainties, The Royal Society and the Royal Academy of Engineering, UK. Schneider, T., R. Vermeulen, D. H. Brouwer, et al. (1999) Conceptual model for assessment of dermal exposure, Occup. Environ. Med., 56 (11), 765–73.

356

Environmental and Human Health Impacts of Nanotechnology

Schneider, T. and K. A. Jensen (2008) Combined single-drop and rotating drum dustiness test of fine to nanosize powders using a small drum, The Annals of Occupational Hygiene, 52 (1), 23–34. Sen, D., H. Wolfson and M. Dilworth (2002) Lead exposure in scaffolders during refurbishment construction activity – an observational study, Occupational Medicine (Oxford, England), 52 (1), 49–54. Shin, W., D. Pui, H. Fissan, et al. (2007) Calibration and numerical simulation of Nanoparticle Specific surface area Monitor (TSI Model 3550 NSAM), Journal of Nanoparticle Research, 9 (1), 61–9. Singh, C., A. H. Shaffer and A. H. Windle (2003) Production of controlled architectures of aligned carbon nanotubes by an injection chemical vapour deposition method, Carbon, 41 (2), 359. Swihart, M. T. (2003) Vapour phase synthesis of nanoparticles, Current Opinion in Colloid & Interface Science, 8 (1), 127. Tinkle, S. S., J. M. Antonini, B. A. Rich, et al. (2003) Skin as a route of exposure and sensitization in chronic beryllium disease, Environmental Health Perspectives, 111 (9), 1202–8. Tran, C. L., D. Buchanan, R. T. Cullen, et al. (2000) Inhalation of poorly soluble particles. II. Influence Of particle specific surface area on inflammation and clearance, InhalationT, 12 (12), 1113–26. van Tongeren, M. J. and K. Gardiner (2001) Determinants of inhalable dust exposure in the European carbon black manufacturing industry, Applied Occupational and Environmental Hygiene, 16 (2), 237–45. van Tongeren, M. J., H. Kromhout and K. Gardiner (2000) Trends in levels of inhalable dust exposure, exceedance and overexposure in the European carbon black manufacturing industry, The Annals of Occupational Hygiene, 44 (4), 271–80. Van Zant, P. (2000) A practical guide to semiconductor processing, 4th edn, McGraw Hill, New York. Walton, W. H. and J. H. Vincent (1998) Aerosol instrumentation in occupational hygiene: An historical perspective, Aerosol Science and Technology, 28, 417–38. Wake, D. (2001) Ultrafine aerosols in the workplace, IR/ECO/00/18. Health and Safety Laboratory, Sheffield. Wang, H. C. (1996) Comparison of thermal rebound theory with penetration measurements of nanometer particles through wire screens, Aerosol Science And Technology, 24, 129–34. WHO (1997) Determination of airborne fibre number concentrations: a recommended method by phase constrast optical microscopy, World Health Organisation, Geneva. Wilson, W. E., J. Stanek, H. S. Han, T. Johnson, H. Sakurai, D. Y. Pui, J. Turner, D. R. Chen and S. Duthie (2007) Use of the electrical aerosol detector as an indicator of the specific surface area of fine particles deposited in the lung, Journal of the Air & Waste Management Association, 57 (2), 211–20. Yeganeh, B., C. M. Kull, M. S. Hull and L. C. Marr (2008) Characterization of airborne particles during production of carbonaceous nanomaterials, Environmental Science & Technology, 42 (12), 4600–6.

9 Human Toxicology and Effects of Nanoparticles Vicki Stone*, Martin J. D. Clift† and Helinor Johnston* *Applied Research Centre for Health, Environment and Society, Edinburgh Napier University, Edinburgh, United Kingdom

†Institute for Anatomy, Division of Histology, University of Bern, Bern, Switzerland

9.1

Introduction

9.1.1 Toxicology – What Is It? Toxicology is often defined as the study of the noxious effects of chemical substances on living systems (Timbrell, 2002). The noxious effects of chemicals have been recognised for many centuries, for example Paracelsus (1493–1541), a Swiss alchemist and physician, is quoted as saying: ‘All things are poison and nothing is without poison, only the dose permits something not to be poisonous’ (Klaassen, 2001). This suggests that any chemical substance has the potential to be toxic if it reaches a sufficiently high concentration in the body of an organism; therefore, the relevance of dose is of prime importance when considering toxicity. 9.1.2

Particle Toxicology

It is worth noting that toxicology is not limited to dissolved chemicals, but that it also includes particles that enter the body. For example, there are many publications describing the noxious effects of respirable particles such as asbestos and crystalline silica (alpha-quartz). Both of these particle types are defined as carcinogens by the International Agency for Research on Cancer (IARC) due to their ability to induce cancer following inhalation exposure (IARC, 1997, 1987). Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

358

Environmental and Human Health Impacts of Nanotechnology

In the last five years much of the attention of particle toxicologists has turned to nanoparticles, previously termed ultrafine particles (see below). As outlined in Chapter 8, the British Standards Institution (BSI) has recently defined nano-objects as ‘discrete pieces of material with one or more external dimensions in the nanoscale’ and nanoparticles as a ‘nano-object with all three external dimensions in the nanoscale’ (BSI, PAS 136: 2007). Engineered nanoparticles are diverse in their chemistry, shape and applications. Many are engineered due to changes in their properties that occur at the nanoscale, and that impart useful properties such as catalytic activity, electrical conductivity and relative strength. The interest of particle toxicologists in nanoparticles is driven by a number of factors. Firstly, exposure of humans and the environment to combustion derived nanoparticles via air pollution already occurs and is suggested to be linked to a number of health effects (see below, Chapter 5 and Chapter 8). In addition, exposure to manufactured nanoparticles already exists through the use of a wide variety of consumer products ranging from cosmetics, suntan lotions, food additives and clothing, to medicines and diagnostics (http://www.nanotechproject.org/44, observed 04/01/07). With the rapid expansion of nanotechnology and the ever increasing applications for which nanoparticles are developed, such exposure is likely to increase further (Aitken et al., 2006). At present the consequences of such exposure are poorly understood. While many nanoparticles may prove to be relatively harmless, it is important to understand the risks associated with such new materials, especially when exposures are likely to be from multiple sources and to a variety of nanoparticle types. In addition, the physico-chemical properties that make nanoparticles so exciting for the development of new products could also be the same properties that drive interactions with biological systems. 9.1.3

Risk Assessment

Toxicology is part of the risk assessment procedure: Risk = Hazard ( toxicity) × Exposure ( dose) An understanding of toxicity and dose therefore allows companies, regulators and individuals to assess the risk associated with the use of a particular substance (Figure 9.1). At this time something is known about the hazard of a small number of manufactured nanoparticles (summarised below) but very little about exposure, or even how to measure exposure (Maynard and Aitken, 2007). In the current climate it is therefore difficult to generate clear evidence-based advice to companies and consumers regarding the safe use of nanoparticles. To bridge this current knowledge gap, particle toxicologists are now working to assess the physico-chemical factors that determine the fate and behaviour of different types of nanoparticles in the body of humans (Figure 9.2) and in the environment. An understanding of the relationship between particle properties and toxicity will allow improved approaches to testing and risk management, as well as improved safety design of nanoparticles for future use and applications (Figure 9.3).

Human Toxicology and Effects of Nanoparticles

359

Physico-chemical characteristics Behaviour of particles in the environment

Behaviour of particles in the biological systems

Exposure routes and concentrations

Hazard in biological systems

Risk Figure 9.1 Understanding Risk. Risk is estimated from a combination of the exposure and hazard (toxicity) of a substance. The physico-chemical characteristics of nanoparticles influence how they behave in the environment and, therefore, what the route and quantity of exposure might be, as well as their toxicity.

e.g. fibre shape effects clearance Entry route and efficiency

Inherent toxicity e.g. low TiO2 high Cd SHAPE

Translocation SIZE/ SURFACE AREA

Electrochemical gradients

ELECTRICAL CONDUCTANCE

COMPOSITION

NANOPARTICLE Physico-chemical characteristics

Membrane potential Durability Biopersistence

CHARGE

Molecular interactions

CRYSTAL STRUCTURE

STRENGTH

e.g. crystalline silica vs. amorphous silica

SOLUBILITY

Clearance Biopersistence

Clearance

Release of toxic components

Figure 9.2 Linking the physico-chemical characteristics of particles to their biological activity. These properties include published information for particles (quartz, asbestos, air pollution) and nanoparticles, as well as purely hypothetical ideas (e.g. electrical conductance).

360

Environmental and Human Health Impacts of Nanotechnology Understanding hazard

Understanding exposure

Understanding risk Manage hazard

Manage exposure Manage risk

Maximise potential of nanotechnology with minimum risk

Figure 9.3 Understanding hazard and exposure allows risk to be understood, which in turn allows hazard, exposure and therefore risk to be managed. If risk can be managed then the new technologies and industries using nanoparticles can reach their full potential with minimum risk.

9.2

Ultrafine Particle Toxicology

9.2.1 Air Pollution In the mid 1990s evidence accumulated to support the suggestion that ambient air particulate pollution was associated with increased ill health and deaths due to both respiratory and cardiovascular diseases (Dockery et al., 1993; Schwartz, 1994). Ambient air particulate pollution is monitored in many cities around the world as PM10 (approximately defined as particles collected through a filter collecting 10 µm diameter particles with an efficiency of 50%) and consists of particles of 10 µm diameter and smaller. The figure of 10 µm represents the largest particles that can be inhaled into the human respiratory system. These particles include carbonaceous particles (contaminated with metals and organic material), coarse wind blown dust and soluble secondary particles made by photochemical reactions as well as organic particles such as pollen (Donaldson et al., 2003) (for a more detailed description of the composition of PM10 see Chapter 5). In fact, in towns and cities a large majority of the particles in PM10 are carbonaceous particles derived from traffic and other combustion processes; these particles are below 100 nm in diameter and in the past have been termed ultrafine particles (Stone et al., 2000a; Donaldson et al., 2005). Seaton et al. (1995) published the ‘Ultrafine hypothesis’ in the Lancet. This paper suggested that it was the ultrafine particles in PM10 which are responsible for driving the adverse cardiovascular health effects caused by this pollutant. This hypothesis was based upon the observations by Oberdorster’s group that ultrafine TiO2 induced a greater inflammatory response in the rat lung than larger respirable TiO2 particles (Ferin et al., 1990; Oberdoerster et al., 1990). Seaton et al., (1995) therefore hypothesised that the ultrafine particles in PM10 induce an inflammatory response, that in susceptible individuals led to an exacerbation of their pre-existing cardiovascular disease symptoms, generating an increase in hospital admissions and deaths.

Human Toxicology and Effects of Nanoparticles

361

9.2.2 Testing the Ultrafine Particle Hypothesis To assess the ultrafine particle hypothesis it was necessary for toxicologists to obtain substantial quantities of well defined ultrafine particles, as well as larger respirable particles for comparison. Obtaining such samples from ambient air was difficult as collection devices were limited in terms of the volume of air that could be sampled. This meant that samples were low in mass, often below that required for a comprehensive toxicology study and physicochemical analysis. In addition, such sampling devices often required that the particles be collected onto a filter, from which they are difficult to retrieve without modifying their chemistry and aggregation status. For this reason a number of studies used surrogate particles such as carbon black (CB) (Li et al., 1999; Brown et al., 2000; Stoeger et al., 2006), TiO2 (Ferin et al., 1990; Oberdorster et al., 1990) and polystyrene beads (Brown et al., 2001). These particles all have a low aqueous solubility, suggesting that when they deposit in the lung they do not simply dissolve in the lung lining fluid. Such particles would therefore have to be physically removed to allow clearance from the respiratory system (Figure 9.4).

mucus blood capillary

alveloar space

airways

Clearance via mucociliary escalator

Epithelial cells

particle macrophage

lymphatic vessel

Clearance via lymphatics

Figure 9.4 The lung contains a number of defence mechanisms to remove inhaled particles. These defence mechanisms include cilia and mucus in the airways that work together to generate a mucociliary escalator allowing deposited particles to be blown from the nose or swallowed into the stomach. Particles depositing in the peripheral, or deeper parts of the lungs known as the alveoli, are cleared by immune cells such as macrophages. Macrophages identify the foreign particle, ingest it by phagocytosis and then remove the particle from the lung surface, either by moving to the lymphatic system or via joining the mucociliary escalator.

362

Environmental and Human Health Impacts of Nanotechnology

It is worth noting that many or most humans deal well with particle inhalation due to evolutionary adaptations which have resulted in protective mechanisms that clear particles from the lung surface. Physical removal from the respiratory system involves two possible mechanisms. The lung consists of a series of branching tubes (airways) that terminate in sacs (alveoli). The airways are lined with epithelial cells, which are covered in hair-like cilia and thick, sticky mucus. The cilia continually beat and therefore move the mucus up and out of the lung to either be blown from the nose or swallowed into the stomach. This process is known as the mucociliary escalator. Particles depositing in the airways become trapped in the mucus and are therefore cleared. The epithelium of the alveoli is not ciliated and particles depositing in this deeper part of the lung have to be cleared by cells from the immune system. Such cells include macrophages, which have the ability to move from blood into the lung tissue and then into the airspaces. Once in the alveoli the macrophage cell identifies the ‘foreign’ particle and ingests it by a process known as phagocytosis. If the particle cannot be digested by the cell, then it is physically removed by movement of the macrophage, either on to the mucociliary escalator or by transport to the lymphatic system (Donaldson et al., 2001a). The consequence of movement of different particle types to the lymph nodes is not well understood. Particle size influences the site of deposition of particles within the respiratory system. Particles between about 2.5 and 10 µm in size deposit in the airways and are therefore cleared by the mucociliary escalator. Particles below 2.5 µm can deposit throughout the entire respiratory system and are therefore cleared by a combination of macrophage activity and the mucociliary escalator. This means that ultrafine or nanoparticles when inhaled can deposit anywhere in the respiratory system. It is important to realise that the enhanced respiratory toxicity of ultrafine or nanoparticles is not solely because they reach the alveolar regions of the lung. For example, in the study by Oberdorster’s group, 250 nm TiO2 particles did not induce inflammation, while the 25 nm particles did; both particles deposited in the same region of the lung as they both possessed a mean aerodynamic diameter of 1 µm. This means that there is some factor other than site of deposition that makes ultrafine or nanoparticles more reactive in the lung than larger particles. The same is true for carbon black and polystyrene beads. Since carbon black is made by the combustion of organic material, trace contaminants of polyaromatic hydrocarbons or transition metals can be found in some samples. For the studies in which nanoparticle carbon black was used to induce inflammation in the rat lung, the potential involvement of contaminants was examined by making an aqueous extract of the particles. This extract failed to induce any significant inflammation in the rat lung; furthermore, in vitro it did not exhibit reactive oxygen species (ROS; oxygen containing molecules that have unpaired electrons) production or induce calcium signalling in macrophages (Brown et al., 2000), as had been shown for carbon black (Stone et al., 2000b) (see below). This was clear evidence that the activity of these particles did not lie in a soluble component of the carbon black particle preparation.

Human Toxicology and Effects of Nanoparticles

9.2.3

363

Reactive Oxygen Species and Oxidative Stress

To provide evidence to support the ultrafine particle hypothesis, a number of studies have been conducted to identify a mechanism by which such particles could be more reactive in biological systems. Such studies have demonstrated that these low solubility nanoparticles can generate ROS to a greater extent than larger respirable particles. For example, the plasmid assay has been used to demonstrate the ability of carbon nanoparticles (Stone et al., 1998) and a panel of metal nanoparticles (Dick et al., 2003) to induce ROS production in a cell-free system. The dye 2,7-dichlorofluorescin-diacetate (DCFH-DA) has also been used to show that nanoparticle carbon black (Wilson et al., 2002) and polystyrene beads (Brown et al., 2001) are more potent in a cell free environment, than larger particles at generating ROS, therefore confirming the results of the plasmid assay. The same fluorescent dye DCFH-DA has also been used to measure nanoparticle induced intracellular ROS. For example, Wilson et al. (2002) demonstrated that 14 nm carbon black, but not 260 nm carbon black, stimulated an increase in intracellular ROS in a macrophage cell line. Due to the propensity of nanoparticles to aggregate in suspension, Foucaud et al. (2007) suspended carbon black particles in saline supplemented with either bovine serum albumin (BSA) (1%) or dipalmitoyl phosphatidyl choline (DPPC) (0.025%) and investigated the impact of these solutions on particle aggregation and ROS production. Both BSA and DPPC significantly improved the stability of the nanoparticle suspension, decreasing the extent of particle agglomeration and settling over time. Foucaud et al. (2007) used the DCFH assay to measure ROS and found that ROS production by 14 nm carbon black was greater in the DPPC suspension than in either saline alone or the BSA solution. Dispersion of the 14 nm carbon black particles with either BSA or DPPC also enhanced intracellular ROS detection in the MonoMac6 cell line, suggesting that enhanced dispersion also enhances ROS production for the carbon black nanoparticles. Additional evidence for ROS production by nanoparticles is outlined in Chapter 3. ROS are of interest because their unpaired electron makes them very reactive, resulting in damage to many biological molecules including lipids, proteins and DNA. The body contains a number of antioxidant defence mechanisms to protect against the damaging effects of ROS, however these defence mechanisms are not always effective. Oxidative stress is an imbalance between oxidants and antioxidants and can be caused by excessive production of oxidants/ROS or depletion of antioxidants (Li et al., 1999; MacNee and Rahman, 2001). The evidence that ultrafine particles can generate ROS in a cell-free environment suggests that such particles could themselves generate ROS that have the ability to induce oxidative stress in cells. Furthermore, oxidants can be generated by activated phagocytic cells such as macrophages (MacNee and Rahman, 2001), contributing to the pro-oxidant status of a particle exposed organ. In order to identify whether nanoparticles could induce oxidative stress, Stone et al. (1998) investigated the effect of 14 nm carbon black on the antioxidant glutathione in an alveolar epithelial cell line. This study identified that 14 nm carbon black induced depletion of reduced glutathione and accumulation of oxidised glutathione, an effect which was

364

Environmental and Human Health Impacts of Nanotechnology

not observed for the 260 nm particles. At higher concentrations (125 µg/ml) cytotoxicity was also observed with the smaller, but not the larger, particles, and this cytotoxicity could be prevented by antioxidants, suggesting that it is mediated via oxidative stress. In animal studies, instillation into the lung of rats of nanoparticle carbon black also led to a depletion of glutathione in the lung tissue (Li et al., 1999). These studies therefore provide evidence that carbon black nanoparticles, but not larger particles induce oxidative stress in lung cells. Nanoparticle (25 nm) TiO2 has also been shown to induce ROS production by BV2 microglia cells (Long et al., 2007). This ROS production was associated with up-regulation of an inflammatory response, apoptosis and cell cycling pathways, while energy metabolism was down-regulated. The same study indicated that nanoparticle TiO2 did not induce toxicity to N27 neuronal cells, indicating a cell selective effect. More recently, Sayes et al. (2005) have also demonstrated that C60 nanoparticles elicit oxidative stress mediated toxicity in cells in vitro, suggesting that oxidative stress is not limited to carbon black. Many of the respiratory and cardiovascular diseases associated with ambient ultrafine particles and PM10 are driven by inflammation. Inflammation is an activation of the immune system, including activation of the white blood cells known as macrophages. Inflammation is essential to fight infection, but when it is inappropriate in amplitude or duration then it can lead to disease. Acute effects might include exacerbation of disease symptoms, such as asthma, while chronic and more severe effects might include fibrosis (scar tissue formation) and cancer. Oxidative stress activates inflammation via a number of intracellular signalling pathways. These pathways transmit a signal to transcription factor proteins that enter the nucleus of the cell and control the expression of genes involved in inflammation. Such transcription factors include nuclear factor kappa B (NFκB) and activator protein 1 (AP1), both of which are sensitive to oxidative stress. Prior to activation NFκB is located in the cytoplasm of the cell where it is bound to an inhibitor subunit called IκB. IκB is activated by the enzyme IκB kinase which phosphorylates IkB, resulting in dissociation and degradation, allowing active NFκB to translocate to the nucleus. Once in the nucleus, NFκB binds specific promoter motifs and initiates transcription of the genes driving inflammation (e.g. tumour necrosis factor α(TNFα) and interleukin 8) (Christman et al., 2000). Using immunofluorescence and confocal microscopy, Brown et al. (2004) demonstrated that 14 nm carbon black induced nuclear translocation of NFκB in primary human macrophages in vitro. This effect could be blocked by antioxidants, suggesting that the oxidative stress induced by the particles was responsible for activating this transcription factor. In addition, the same study also assessed the effect of 14 nm carbon black on AP1 DNA binding; this marker of activation was also enhanced by the particles and prevented by the antioxidant. Treatment of the macrophages with 260 nm carbon black had no significant impact on either NFκB or AP1 activation in this study, indicating that the effect was size specific. Oxidative stress also influences intracellular calcium signalling (Orrenius et al., 1992; Brown et al., 2007a). Intracellular calcium signalling is very important in the regulation of many essential cellular functions, including enzymes that control transcription factor activation (Berridge, 2004). ROS are thought to affect calcium

Human Toxicology and Effects of Nanoparticles

365

signalling by oxidation of thiol groups, for example in the Ca2+ ATPase pumps. These pumps are responsible for sequestering calcium into stores such as the endoplasmic reticulum and for pumping calcium out of the cell via the plasma membrane (Barnes et al., 2000; Kourie, 1998). In addition, there is evidence that ROS can impact upon membrane calcium channels and calcium binding proteins (Kourie, 1998). Dysfunction of these calcium controlling mechanisms due to ROS results in an influx of calcium into the cytoplasm and potentially uncontrolled calcium signalling. In a monocytic (immature macrophage) cell line known as Monomac 6, and in primary rat alveolar macrophages, 14 nm carbon black has been found to induce a significant calcium influx (Stone et al., 2000a,b). This effect was not observed with 260 nm carbon black particles. Similar results were also demonstrated for polystyrene particles of different sizes (Brown et al., 2001). The calcium signalling induced by carbon black nanoparticles was shown to be important in activating the transcription factor NFκB and in stimulating the production of the pro-inflammatory cytokine TNFα. Addition of antioxidants, to block the ROS effects, prevented the 14 nm carbon black calcium signal NFκB activation (as outlined above) and TNFα production, suggesting a strong link between the ability of these nanoparticles to generate ROS and inflammation. 9.2.4

Uptake of Nanoparticles into Cells

Each cell is surrounded by a plasma membrane, which segregates the internal and external environments of a cell and is responsible for regulating the entry and exit of substances into and out of the cell (Conner and Schmid, 2003). There are a number of clearly defined mechanisms for crossing the plasma membrane that include: • • • •

diffusion facilitated diffusion active transport endocytosis.

The plasma membrane consists of a phospholipids bilayer and proteins and is therefore a lipophyllic environment that allows selective permeability to low molecular weight lipophillic substances by passive diffusion (e.g. alcohol). In contrast, hydrophilic or high molecular weight substances must gain access via channels, transporter proteins or endocytosis, all of which are active processes. The extracellular surface of the plasma membrane consists of the negatively charged phosphate heads of the phospholipids molecules, resulting in a negatively charged surface that could influence the way in which charged particles interact with the cell membrane. In addition, there is an overall intracellular negative charge, caused by unequal distribution of charged ions across the plasma membrane. This unequal distribution of ions is driven by the activity of ion channels and pumps such as the sodium/potassium (Na/K ATPase) pump, which transports two potassium ions into the cell for every three sodium ions pumped out. The cell actually uses this negative electrical (and chemical) membrane potential to drive the transport of substances, often against a concentration gradient, into or out of the

366

Environmental and Human Health Impacts of Nanotechnology

cell. For example, charge influences the uptake of cationic molecules, which are strongly attracted to the cell surface due to the non-specific electrostatic interactions that occur with the negative charge of the plasma membrane interior (Patel et al., 2007). It is worth noting that charged molecules cannot pass through the plasma membrane simply via diffusion. It is conceivable that this membrane potential could provide a driving force to drive positively charged particles to move into the cell. Whether nanoparticles have the ability to move into cells via diffusion is currently unclear. A study by Geiser et al. (2005) found that TiO2 particles were located inside lung cells following inhalation. The EM images used to image the TiO2 suggested that plasma membrane was not clearly distinguishable around the nanoparticles, from which the authors concluded that uptake could be via a nonendocytic pathway such as diffusion. However, much more evidence is required before diffusion is widely accepted as a viable uptake route for nanoparticles into cells. Facilitated diffusion allows substances to pass from an area of high to low concentration through selective membrane protein channels; for this process energy is not directly required, as it is usually driven by a chemical concentration gradient. The carrier proteins and channels can be opened or closed depending on the cells needs, such examples include ligand and voltage gated ion channels. Movement of nanoparticles into the cell via such channels would require the particles to be as small as the channel pore size, often in the region of 100–300 Å (10–30 nm). Active transport of substances across the membrane is conducted by proteins and occurs against a concentration gradient using energy in the form of ATP. Transporter structure is very specific for the molecules to be actively transported, and therefore it is unlikely that nanoparticles could use this as a route of entry into the cell. The most likely route of uptake of nanoparticles (and larger particles) into cells is via endocytosis. This is a route of cell entry used for larger molecules and particles. Endocytosis involves incorporating molecules or particles into membrane bound vesicles derived from the invagination or pinching-off of the plasma membrane (Conner and Schmid, 2003; Watts and Marsh, 1992). Endocytosis is in fact a collective term that includes phagocytosis, clathrin mediated endocytosis, caveolae mediated endocytosis as well as clathrin and caveolae independent endocytosis (Conner and Schmid, 2003). The size of the vesicles formed for each pathway differs, for example clathrin coated pits are approximately 120 nm in diameter while caveolae are 50–80 nm and micropinosomes are generally 1–5 µm (Patel et al., 2007). Although these ranges are not definitive, it is likely that size limits exist to restrict the cargo dimensions internalised (Patel et al., 2007). 9.2.5

Interaction of Nanoparticles with Defence Mechanisms

Phagocytosis is, of course, the process of cell uptake used by defensive cells of the immune system such as macrophages and neutrophils, also known as professional phagocytes (Conner and Schmid, 2003). These cells take up relatively large materials (greater than 0.5 µm) such as bacteria or cell debris (Khalil et al. 2006). During phagocytosis, the cell recognises ligands via cell surface receptors. Receptor binding

Human Toxicology and Effects of Nanoparticles

367

then triggers the polymerisation and rearrangement of the actin cytoskeleton to form membrane extensions, so that the plasma membrane surrounds the material to be internalised (Liu and Shapiro 2003; Perret et al., 2005; Khalil et al., 2006). The phagosome that is formed will fuse with lysosomes, so that the cargo can be degraded if possible (Perret et al. 2005). As described in the introduction, macrophages play a key role in the defence of the respiratory system. However, such mechanisms are not unique to the lung. Inflammatory cells such as macrophages can penetrate almost all tissues of the body, providing a useful defence mechanism against inhaled, ingested, injected, implanted and dermally absorbed particles. Some studies suggest that relatively low concentrations of TiO2 and carbon black nanoparticles increase the phagocytic ability of macrophages (Renwick et al., 2001; Hoet and Nemery, 2001), while at higher concentrations the phagocytic function of these cells is diminished (Renwick et al., 2001; Lundborg et al., 2006). The mechanism or cause of this inhibition remains unknown. Macrophages have the ability to ingest large volumes of particles and, therefore, inhibition of phagocytosis could be as a result of filling of the cell volume or a decrease in membrane area to allow continued uptake. One research group has suggested that nanoparticle uptake cells does not require membrane invagination, phagocytosis or pinocytosis (Geiser et al., 2005), suggesting uptake via diffusion across the cell membrane. Further studies are required to verify this result with other types of nanoparticles and other cell types. Epithelial cells also play an important role in defence by providing a barrier across which particles must pass to gain access to the body. As mentioned previously, carbon black nanoparticles induce glutathione depletion, indicative of oxidative stress, in a human lung epithelial cell line (Stone et al., 1998). Epithelial cells when exposed to damaging or infective agents can generate chemotactic factors; these chemotactic factors stimulate inflammatory cell recruitment and, therefore, inflammation. For example, some nanoparticles have been shown in vitro to stimulate production of the chemoattractant interleukin-8 by epithelial cells (Montellier et al., 2006). As mentioned previously, carbon black, TiO2 and polystyrene beads are all low solubility materials. It is also very important to note that they are also relatively low toxicity materials. For newly engineered nanoparticles made from more toxic materials (e.g. the cadmium used in some quantum dots) and biodegradable materials (e.g. the polymers used in drug delivery), the toxicity may be very different. Therefore, new studies will have to focus on whether the nanoparticulate nature of these materials impacts upon, or interacts with their chemical mechanism of action resulting in any altered or unexpected toxicity. 9.2.6

Nanoparticle Interactions with Other Pollutants and Molecules

Due to the presence of transition metals in particulate air pollution, Wilson et al. (2002) investigated the effect of mixing carbon nanoparticles with metal salts. Transition metals such as iron are known to generate hydroxyl radicals via Fenton chemistry (Stohs and Bagchi, 1995), and since 14 nm carbon black particles have

368

Environmental and Human Health Impacts of Nanotechnology

also been shown to generate ROS (Stone et al., 1998), it was hypothesised that the two might interact. Mixing 14 nm carbon black with FeCl3, FeSO4 or CuSO4 in a cell free environment, resulted in potentiation of ROS production, as assessed by the DCFH assay. Wilson et al. (2002) also investigated the interaction in vivo by instilling a mixture of the 14 nm carbon black and FeCl3 solution into the rat lung. The iron salt was found to potentiate the carbon black induced lung inflammation, showing that the particles and metal salts clearly interact, both in vitro and in vivo. At the time of this publication, the discussion of these results was limited to the relevance to air pollution research, but with many nanoparticles, including soluble metal ions, such findings are also useful to inform the potential mechanism of toxicity of manufactured nanoparticles. Wilson et al. (2007) also investigated the interaction between zinc chloride and carbon black. In this study, the authors found that zinc could not potentiate ROS production by the carbon black particles, but that it did induce a large synergistic activation of macrophages in vitro, as indicated by production of the proinflammatory cytokine TNFα. Again, this data is highly relevant to manufactured nanoparticles, some of which (e.g. quantum dots) are coated in zinc sulphide. There have also been suggestions that diesel exhaust air pollution particles (DEP), which contain substantial amounts of combustion derived nanoparticles can interact with airborne allergens, inducing an adjuvant effect. This means that when inhaled in the presence of pollen, or dust mite allergens, the resultant allergic response is enhanced (Takano et al., 1998). The adjuvant effects of nanoparticles are actually being exploited in nanomedicine in the development of vaccines (Peek et al., 2008).

9.3

Engineered Nanoparticles

As mentioned previously there is a vast array of engineered nanoparticles, many of which are described in previous chapters, which need to be considered in terms of their toxicology. While inhalation is still important as a route of exposure, especially in an occupational environment where powders are employed, other routes of exposure, such as ingestion, dermal adsorption and injection/implantation, also need to be considered. Therefore, if every new nanoparticle was to be tested for toxicity using a protocol relevant to multiple doses, and multiple routes of exposure, thousands of tests would be required. Add to this, variations in formulations, interactions with other toxins and variability in susceptibility between individuals, and the number of experiments required is almost beyond comprehension, and certainly beyond budget. Instead of testing every nanoparticle for every relevant scenario, in the near future it will be necessary to try to generate some general rules regarding the physicochemical factors which drive the mechanism by which particles interact with biological systems, and therefore their toxicity. For this reason, the following section is broken down by nanoparticle type. While it is still unclear which factors are most important in determining toxicity, classification of nanoparticle by material is a first useful step in the absence of more useful information.

Human Toxicology and Effects of Nanoparticles

9.3.1

369

Fullerenes

The Buckminster fullerene, also known as C60, was first discovered by Kroto et al. (1985). This nanoparticle consists of 60 linked carbon atoms in a highly stable icosahedron structure, with 60 vertices and 32 faces (12 pentagonal and 20 hexagonal). The term fullerene now applies to molecules composed entirely of carbon that form spheres or tubes. The use of C60 is being investigated for use in optics and superconductors (Da Ros and Prato, 1999) and for drug delivery (Vogelson, 2001). Cosmetic products such as face creams containing C60 nanoparticles have also entered the market (Halford, 2006), with the suggestion that they contain antioxidant properties (see below). C60 particles are made on a large scale; for example Mitsubishi has opened its first fullerene plant in Japan which aims to produce fullerenes by the ton (Tremblay, 2003). A number of studies provide evidence that C60 can act as an antioxidant. For example, Gharbi et al. (2005) found that C60 was able to prevent carbon tetrachloride (CCl4) induced liver toxicity in the rat. In a different study, lipid and water soluble C60 derivatives prevented superoxide and hydroxyl radical initiated lipid peroxidation to a greater extent than the natural antioxidant, vitamin E (Wang et al., 1999). Using electron spin resonance (ESR), Xiao et al. (2006) also demonstrated that chemically generated hydroxyl radicals were quenched by polyethylene glycol (PEG)-modified and hydroxyl fullerenes. C60 fullerenes and single-wall carbon nanotubes (SWCNT) have also been shown not to induce measurable production of the reactive nitrogen species, nitric oxide, by a mouse macrophage cell line (Fiorito et al., 2006). These studies together suggest that C60 and its derivatives could actually be of beneficial health effect, rather than induce toxicity. However, toxicity depends upon the dose (as stated in the introduction) as well as the environment in which the ‘antioxidant’ is investigated. Many antioxidants have the capacity to act as oxidants in the right conditions (e.g. vitamin C). Therefore, it is not surprising that other studies have presented data suggesting that C60 and its derivatives have pro-oxidant and toxic effects. Sayes et al., (2005) looked at the effects of nano-C60 prepared in tetrahydrofuran (THF) on three cell lines. The nano-C60 was cytotoxic to all three cell types at doses above 50 ppb, with LC50 ranging from 2 to 50 ppb depending on the cell type. In this study nano-C60 induced membrane damage prevented by the addition of an antioxidant, confirming a role for ROS. Note that a number of ecotoxicology studies have also used THF as a solvent to disperse C60 (Oberdorster, 2004), but it has been suggested that the THF may influence toxicity (Brant et al., 2005), and so this data requires careful consideration. Markovic et al. (2007) compared C60 prepared in THF, ethanol and water in terms of their ability to generate ROS, cause mitochondrial depolarisation and necrotic cell death in a variety of cell lines. The THF preparation was found to be most potent, followed by the ethanol preparation, while the water suspension was least potent. Some researchers have suggested that the concentration of THF remaining in such a C60 preparation would deliver a final dose in an organism that is below concentrations normally associated with THF toxicity. However, this suggests that the concentration of THF is distributed throughout the body as if it were

370

Environmental and Human Health Impacts of Nanotechnology

introduced as a pure chemical. Instead, the THF is trapped within C60 clusters, so that the fate and distribution of the C60 cluster determines the THF fate and distribution. Such delivery mechanisms are being used by the pharmaceutical industry to target drugs to specific locations, reducing delivery to non-target sites and, therefore, reducing the required dose and reducing side effects. While this has not been demonstrated for THF in C60 preparations it needs to be considered as a potential complicating factor. Furthermore, there could be potential for C60 and THF to interact (antagonistically or synergistically) to alter toxicity making such experiments difficult to interpret or control. 9.3.2

Nanotubes and Other Fibre-Like Nanostructures

High aspect ratio nanoparticles (HARN) are long fibre-like nanoparticles including nanotubes, nanowires, nanorods and nanofilaments (Oberdoerster et al., 2007). The bulk manufacture and use of such nanoparticles has raised concerns because of their resemblance to asbestos in terms of dimensions and durability; both factors which determine fibre toxicity leading to fibrosis and the cancer mesothelioma (Donaldson et al., 2006). Single- and multi-walled carbon nanotubes have received relatively more attention than other high aspect ration particles in relation to toxicology studies, probably because of the greater production volumes leading to greater exposure risk. Toxicology studies relating to carbon nanotubes (CNTs) appear to fall into a number of catagories: (i) respiratory effects (ii) dermal effects (iii) biomedical applications. In addition, CNTs might have effects as particles or, because of their high aspect ratio, they may exhibit effects as fibres; two different paradigms dictate these two different effects. Paradigm I: CNT as particles (i) Respiratory effects of carbon nanotubes The respiratory studies are generally conducted in order to ascertain the potential for disease resulting from occupational exposure to airborne carbon nanotubes, since this was the most relevant route of exposure leading to asbestos induced disease in the asbestos industry. A study published by Maynard et al. (2004) has been widely cited with respect to the potential for nanotubes to become airborne. This study investigated single-walled carbon nanotubes (SWCNT) but not multiwalled carbon nanotubes (MWCNT). They found that agitation in a controlled laboratory setting of unrefined SWCNT could result in fine particle release into air. However, in general, airborne concentrations in the region of SWCNT production equipment were interpreted as low (less than 53 µg/m3). The relevance of these exposure concentrations is discussed further in Chapter 8. There is no information currently available regarding the exposure to MWCNT in a laboratory, experimental or industrial setting. In our own experience, SWCNT and MWCNT vary greatly

Human Toxicology and Effects of Nanoparticles

371

in terms of their properties and quality. Not only do they differ greatly between different sources, but even the same laboratory/manufacturing process can generate vastly different samples from day to day. Add to this the fact that composition, length, wall number and catalyst can all be manipulated, then it is clear that it is still too early to disregard inhalation as a feasible route of exposure. Some of the first studies to investigate SWCNT respiratory toxicology used pharyngeal aspiration as a means of delivery to mice (up to 40 µg/mouse) (Shvedova et al., 2005). The SWCNT generated an acute inflammatory response with increased neutrophils at day 1 and macrophages at day 7. In addition, fibrosis of two forms was generated within 28 days at a dose of 20 or 40 µg/mouse. The first presented as granulomas mainly associated with hypertrophied epithelial ells surrounding SWCNT aggregates. The second consisted of diffuse fibrosis and alveolar wall thickening. Biochemical markers of lung damage included increased bronchoalveolar lavage protein, lactate dehydrogenase (LDH) and g-glutamyl transferase activity. Induction of oxidative stress was indicated by increased 4-hydroxynonenal and glutathione depletion of lung tissue. Another study delivered MWCNT to rats via instillation into the lung (0.5, 2 or 5 mg/animal) (Muller et al., 2005). In this study MWCNT were compared with ground MWCNT. Both samples persisted in the lung tissue for 60 days, induced inflammation and resulted in fibrosis (2 mg/rat). The pathology developed into a collagen rich granuloma by two months, caused by the accumulation of agglomerates of MWCNT in the airways. Ground MWCNT were better dispersed but still caused inflammation and fibrosis. Interestingly, this study compared the MWCNT with nanoparticle carbon black; they found the MWCNT to be more pathogenic. Other studies comparing SWCNT with nanoparticle carbon black have made similar observations (Warheit et al., 2004; Lam et al., 2004). The reason for the relative potency of single- and multi-walled carbon nanotubes compared to nanoparticle carbon black remains unclear, but could include chemical differences in the graphitic surface of the particles or the physical size of the aggregates produced in the studies, as suggested by Warheit et al. (2004). In the study by Lam et al. (2004) the SWCNT was found to be more potent at inducing granulomas and fibrotic lesions than quartz, a well known fibrotic and carcinogenic respirable particle. The different studies quoted here vary in terms of the source of the carbon nanotubes, metal contamination and particle dimensions. A systematic approach is required to identify the physicochemical characteristics that drive the pathogenic responses observed in these studies. All of the studies described above have involved delivery of a bolus dose of aggregated or agglomerated nanoparticles. To date, inhalation studies of airborne multi-walled or single-walled carbon nanotubes have not been published. As mentioned previously one of the main defence mechanisms from the lung involves clearance by macrophages. Kagan et al. (2006) compared SWCNT that either contained iron (26 wt-% of iron) or were iron depleted (0.23 wt-% of iron). In a cell-free system the iron content determined the ability of the SWCNT to generate free radicals. In the same study, the authors treated the RAW 264.7 macrophage cell line in vitro with SWCNT and found that the iron rich sample induced markers of oxidative stress such as glutathione depletion and lipid peroxidation,

372

Environmental and Human Health Impacts of Nanotechnology

suggesting iron content is important in determining redox dependent effects in macrophages. The bioavailability of iron in SWCNT was investigated by Guo et al. (2007). These authors were able to mobilise redox active iron from a diverse array of commercially available nanotubes. The amount of iron that could be mobilised could not be predicted from the total amount of iron present in the sample, suggesting that they differ in terms of the amount trapped within the nanotube structure. With particulate air pollution there is a clear link between the respiratory exposure and subsequent cardiovascular effects, although the mechanism remains unclear (Donaldson et al., 2001c; Mills et al., 2005). Li et al. (2007) exposed mice to SWCNT (10 and 40 µg/mouse) and found induction of oxidative markers in the lung, aorta and heart tissue. In addition, the SWCNT accelerated plaque formation in ApoE–/– mice fed an atherogenic diet. This requires further investigation in order to prevent increased risk of cardiovascular disease as a result of occupational exposure and use of nanotubes. (ii) Dermal effects of carbon nanotubes The studies conducted to investigate the potential dermal toxicity of nanotubes have been conducted using either in vitro cell line models, or implantation into the skin of rodents. The main reason for conducting such studies is due to the potential for graphite and other carbon materials (e.g. carbon fibres) to induce dermatitis. Shvedova et al. (2003) found that treatment of a human keratinocyte cell line with SWCNT (30% iron) (0.6 mg/ml for 18 hours) led to oxidative stress and cytotoxicity. Monteiro-Riviere et al. (2005) exposed human epidermal keratinocytes to MWCNT for up to 48 hours at relatively high concentrations of 100, 200 and 400 µg/ml. The MWCNT were observed by transmission electron microscopy (TEM) within cytoplasmic vacuoles as early as one hour after exposure. All concentrations resulted in IL8 (interleukin 8) production by eight hours and significant cytotoxicity at 24 hours. In animal studies, Koyama et al. (2006) implanted both SWCNT and MWCNT into the subcutaneous tissue of mice for up to three months. The authors noted that the CNT induced toxicity, but that this subcutaneous toxicity was relatively low in comparison to asbestos. The toxicity took the form of activation of antigen/antibody response systems, in that after one week the SWCNT induced activation of major histocompatibility complex (MHC) class 1 of CD4+/DC8+ T-cells. After two weeks all CNT samples activated MHC class II markers in CD4+/DC8+ cells and in CD4+ cells. Granulomatous tissue encapsulating nanotube agglomerates was observed in both the SWCNT and MWCNT treated animals. A similar study was conducted by Yokoyama et al. (2005) using hat-stacked carbon nanofibres instead of carbon nanotubes. In this study the implantation was into rats; inflammatory response generated was described by the authors as mild, in that it did not generate widespread necrosis, but it did induce the appearance of granuloma like structures. Paradigm II: CNT and the fibre paradigm Our own group has found that the physical structure of MWCNT greatly influences their interaction with macrophages. Long (50 µm) rigid MWCNT formed fibre-like aggregates that were too large to be taken up by macrophages in vitro, resulting in ROS production, frustrated phagocytosis and TNFα production. While entangled

Human Toxicology and Effects of Nanoparticles

373

MWCNT formed aggregates that were easily ingested by the macrophages resulting in little ROS or TNFα production. If this observation translates to the in vivo situation, this would suggest that carbon nanotubes behave like other pathogenic fibres such as asbestos. As a consequence of the observation that MWCNT have the potential to behave like pathogenic fibres in vitro, an in vivo study was conducted to determine the consequence of MWCNT interaction with the mesothelium. As a pilot study, Poland et al. (2008) instilled a selection of entangled and relatively straight MWCNT into the peritoneal cavity of the mouse, and compared these with carbon nanoparticles, long fibre amosite (LFA) (pathogenic asbestos sample) and short fibre amosite (SFA) (non-pathogenic). As hypothesised, the long relatively straight MWCNT behaved similarly to LFA in that they generated an inflammatory response that was associated with pathological changes in the mesothelial layer lining the peritoneal surface of the diaphragm. In contrast, entangled and relatively short MWCNT, as well as the nanoparticle carbon black and SFA, did not induce any significant pathological change. This data suggests that MWCNT, if long and straight conform to the fibre paradigm of toxicity, but this needs to be confirmed in the pleural cavity following exposure via inhalation. (iii) Biomedical applications There are many studies which have been conducted with the aim to investigate the potential for use of carbon nanotubes as nanomedicines, including drug delivery, gene delivery and nanocomposites for implantation. A full review of these applications is beyond the scope of this chapter and the reader is referred elsewhere for this information (Lacerda et al., 2006). Instead, this section focuses on publications in this area that provide toxicologically relevant information. For example, Singh et al. (2006) injected radio-labelled SWCNT into mice to identify their biodistribution and clearance. The SWCNT had been modified by functionalisation in order to make them more water soluble and, therefore, more useful in biomedical applications. Intravenous administration of the functionalised SWCNT resulted in rapid excretion of the radioactivity from the blood with a half-life of three hours. Clearance was via the kidney, as verified by electron microscopy analysis of urine. Functionalisation appeared to prevent uptake by the reticuloendothelial system, including the liver and spleen. Such functionalisation is obviously useful in preventing accumulation in the body, facilitating clearance and thereby reducing the potential for toxicity. However, it is also a disadavangate if the nanotube is cleared too rapidly from the body to allow drug or gene delivery to a specific target. Another study injected intravenously unmodified SWCNT into rabbits and identified their biodistribution using near-infrared fluorescence. Again, the nanotubes were found to clear from the blood, this time with a half-life of one hour. The nanotubes were found accumulated in the liver 24 hours after administration, the consequences of which remain unknown. 9.3.3

Metals

There is a wide range of metal and metal oxide nanoparticles, in fact too many to be covered comprehensively by this chapter. Instead, the focus is on some

374

Environmental and Human Health Impacts of Nanotechnology

nanoparticles that have attracted the attention of toxicologists because of their elemental components or because of their widespread use. Silver nanoparticles are of interest due to the widespread availability to the public via the internet. A quick search on the internet finds that colloidal silver is purported to be a cure for colds and infections, AIDS and even cancer! There is little doubt that silver, and nanoparticles of silver, have antimicrobial properties (Alt et al., 2004; Baker et al., 2005; Edwards-Jones, 2006). These antimicrobial properties have been harnessed for medical purposes to provide an additional tool to fight antibiotic resistant infections (Furno et al., 2004) and to provide medical workers with a more effective barrier to prevent infection (Li et al., 2006). However, it has been known for many decades that ingestion of silver into the body results in argyria, pigmentation of skin due to removal of silver from the body in general into the skin (Chen and Schluesener, 2008). There is also some evidence of association with seizures and altered mental status (Brandt et al., 2005; Gulbranson et al., 2000; White et al., 2003). Much of the published toxicology of silver nanoparticles relates to effects on microorganisms, but there are small number of studies that have addressed effects in toxicology and ecotoxicology models. For example, Hussain et al. (2005) compared the effects of a panel of nanoparticles (silver, molybdenum, aluminium, titanium dioxide and iron oxide (Fe3O4) on the rat liver cell line BRL 3A. The silver nanoparticles (15 nm diameter) decreased mitochondrial function and increased cytotoxicity (5–50 µg/ml), while molybdenum (30 nm), aluminium (30 nm), TiO2 (40 nm) and iron oxide (30 nm) had no significant effect at these concentrations. The authors suggested that the effects of silver were due to oxidative stress since glutathione was also depleted. Ji et al. (2007) exposed rats to an aerosol of silver nanoparticles for up to four weeks (five days per week). The endpoints measured were limited to body weight and markers of haematology and blood biochemistry. No significant effects were observed, but studies of the impacts on the lung and lung inflammation were not investigated. Even gold has been shown to alter its properties when generated at the nanoscale. Gold nanoparticles demonstrate catalytic properties including oxidation of carbon monoxid and hydrocarbons (reviewed in van Bokhoven et al., 2006). Gold nanoparticles have been used for toxicokinetic studies due to the availability of relatively easy and sensitive quantitative detection techniques (Takenaka et al., 2006). Such studies have demonstrated that gold nanoparticles injected, or ingested, have the potential to translocate to many organs, including the liver, spleen, kidney and brain. Furthermore, in pregnant female rats, accumulation of gold nanoparticles in the foetus has been observed (Semmler-Behnke et al., 2007). 9.3.4

Metal Oxides

A variety of metal oxides have been developed as nanoparticles. Perhaps one of the most widely studied metal oxide nanoparticles is TiO2, which has been manufactured and marketed for many years. TiO2 nanoparticles are being used widely in sunscreens and in cosmetics, where they act to reflect UV light. Probably the first toxicology study to demonstrate the enhanced toxicity of nanoparticles compared

Human Toxicology and Effects of Nanoparticles

375

to larger particles was conducted using TiO2 (Ferin et al., 1992; Oberdorster et al., 1994). As described previously, these studies indicated that nanoparticle TiO2 induced a greater pulmonary inflammatory response than fine TiO2. However, it is worth noting that when compared to other nanoparticles, such as nickel, cobalt or carbon black, the TiO2 nanoparticles are less potent at inducing lung inflammation, and this seems to be related to their lower capacity to generate free radicals (Dick et al., 2003). TiO2 can be obtained in two crystal forms, rutile and anatase. Warheit et al. (2007) recently reported that a nanoparticle TiO2 preparation consisting of both anatase and rutile (80/20) was significantly more potent in the rat lung than those containing rutile alone, suggesting that the crystal form is important in determining toxicity. Due to the fact that TiO2 is used as a sunscreen and in cosmetics, dermal penetration and toxicology are of great interest. There are many studies which have investigated the ability of TiO2 to penetrate skin. The techniques used include tape stripping, placing and removing adhesive tape onto and from the skin repeatedly and subsequently analysing the components striped from the skin surface (Mavon et al., 2007). In vitro techniques use skin explants from humans (Pflucker et al., 2001) and pigs (Gamer et al., 2006) cultured in a tissue bath. All of these studies conclude that there is no significant skin penetration according to these models. However, questions have been raised in some reports as to the appropriateness of intact healthy skin models, since sunscreen is often applied to burnt, diseased or damaged skin (Borm et al., 2006). Much work is required to address such questions. The potential toxicity of zinc oxide nanoparticles has also been studied due to their inclusion in suntan lotions and cosmetics. Much of this work has focused on skin, but will not be reviewed here as it is beyond the scope of this chapter. Other studies have investigated effects on the lung due to the potential for occupational exposure to dust and consumer exposure to aerosols. For example, Sayes et al. (2007) exposed rats by instillation to a range of particles including nano and microsized zinc oxide particles (1–5 mg/kg). Both sizes induced an inflammatory response which was resolved by one month post-exposure. This was reflected in vitro by cytotoxicity to L2 epithelial cells as assessed by the LDH assay. Interestingly, the MTT cytotoxicity assay and markers of inflammation (e.g. cytokine expression) did not concur with the in vivo results, suggesting that the in vitro endpoints and time points used were not good indicators of the in vivo response. This study did not generate a clear difference between the micro and nano-sized zinc oxide particles. The solubility of zinc oxide may be an issue relating to its toxicity. Our own work has demonstrated that soluble zinc salts interact synergistically with carbon black nanoparticles to activate pro-inflammatory gene expression by macrophages in vitro (Wilson et al., 2007) The metal oxide cerium dioxide (CeO2) has been developed as a fuel additive; for example Oxonica (Kidlington, UK) markets ENVIROX™ diesel fuel borne catalyst to reduce fuel consumption and exhaust emissions (http://www.oxonica. com/, accessed 26 March 2008). A small number of studies have investigated the potential toxicity of CeO2 nanoparticles. A number of studies demonstrate a relatively low toxicity associated with exposure to this material. For example, Fall et al.

376

Environmental and Human Health Impacts of Nanotechnology

(2007) investigated the effect on the pulmonary toxicity of diesel exhaust fumes generated from CeO2 supplemented diesel. This study used lung slices exposed to the freshly generated fumes, and no impact on viability (ATP and intracellular glutathione), pro-inflammatory mediator expression (TNFα) and antioxidant enzyme activity (glutathione peroxidase, Mn Superoxide Dismutase) could be detected. The antioxidant enzyme catalase was elevated, probably as a defence response. The authors concluded, on the basis of these results and data relating to exposure concentrations, that there is very little risk associated with exposure to diesel fume generated from CeO2 supplemented diesel fuel. Park et al. (2007) also screened the toxicity of CeO2 using a battery of in vitro protocols including a skin irritant test, cytotoxicity assay (BS EN ISO 10993–5) and a test of mutagenicity (Ames Test). The CeO2 was found not to have the potential to be a skin irritant, showed no evidence of cytotoxicity and did not possess any detectable mutagenic activity. In addition, Park et al. (2007) also investigated the effects of CeO2 on the water flea, Daphnia magna, and no toxic effects were observed during a 48-hour exposure. Our own studies confirm this observation (unpublished data), but also suggest that longer exposures (greater than five days) to very high concentrations (10 mg/l) of CeO2 nanoparticles (< 25 nm) result in significant lethality to Daphnia magna. Concentrations of 3 mg/l had no impact on lethality for up to 21 days (unpublished data). Sub-lethal effects have yet to be investigated in this organism. Other studies have identified CeO2 induced toxicity. A study by Lin et al. (2006b) investigated the effects of CeO2 (20 nm diameter) on a human lung caner cell line. Cells were treated at concentrations up to 23 µg/ml and for times of up to 72 hours. The authors observed a dose and time dependent increase in cytotoxicity, oxidative stress (glutathione and α-tocopherol depletion) and lipid peroxidation. The differences in results between studies could be accounted for by the source and characteristics of the nanoparticles used, which although they are both CeO2 could differ by as yet unknown physico-chemical characteristics or contaminants. Nanoparticle silica (SiO2) is often assumed or reported to be amorphous rather than crystalline. The crystallinity of silica is very important in relation to its toxicity. Crystalline respirable silica (alpha-quartz) is classified by IARC as a class I carcinogen (IARC, 1997). There is much evidence that crystalline silica induces lung cancer and silicosis (fibrosis) of the lung (Donaldson et al., 2001b). Until recently, amorphous silica has considered to be relatively benign when inhaled, although much of the evidence used to draw this conclusion is based upon micron sized particles, with little knowledge for silica nanoparticles. One study by Kaewamatawong et al. (2006) has investigated the effects of colloidal nanoparticle silica particles on the mouse lung. Particles were suspended in saline and instilled into the lungs via the trachea, at doses ranging from 0.3 to 100 µg per mouse. At the two highest doses (30 and 100 ug) the colloidal silica nanoparticles induced a transient acute and moderate inflammatory reaction. In addition, there was evidence of increased apoptosis as well as oxidative DNA damage. Lin et al. (2006a) compared the toxicity of silica nanoparticles (15 and 46 nm diameter) with a well characterised and studied form of crystalline silica (Min-U-Sil

Human Toxicology and Effects of Nanoparticles

377

5) in vitro. The bronchoalveolar carcinoma cell line was more sensitive to the silica nanoparticle induced cytotoxicity than to the Min-U-Sil 5. These nanoparticles were found to generate ROS, to induce oxidative stress (glutathione depletion) and lipid peroxidation. These studies suggests that the silica nanoparticles have the potential to induce acute effects both in vitro and in vivo, but much more work is needed before risk can be determined. A small number of studies have also investigated the toxicity of iron oxide nanoparticles. Such particles are being developed for a number of medical applications including magnetic resonance contrast agents for MRI (Choi et al., 2004). Since the route of entry into the body will be injection there is little doubt about exposure to such nanoparticles. Gojova et al. (2007) exposed endothelial cells in vitro to iron oxide (Fe2O3) or zinc oxide (ZnO) nanoparticles. Both nanoparticles were taken up by the endothelial cells. The iron oxide particles had no effect on the endothelial cell pro-inflammatory gene expression, while zinc oxide particles generated a pro-inflammatory response at concentrations greater than 10 µg/ml. A number of studies have investigated neuronal effects of iron oxide, due to increased use in imaging the central nervous system (CNS) by magnetic resonance imaging (MRI). In a study using the neuronal cell line PC12, Pisanic et al. (2007) found iron oxide (Fe2O3) particles to induce a dose dependent cytotoxicity, as well as decreased neurite generation. In a study by Muldoon et al. (2005) rats were exposed to commercial preparations of iron oxide, which were either injected directly into the cerebellum, or into the blood of rats with an incomplete blood– brain barrier (BBB), or into rats with intracerebral tumour xenografts. MRI imaging revealed that MRI signal declined over weeks to months following the intracerebral injection, suggesting that the particles were cleared from the CNS. For the nanoparticles injected into rats with a defective BBB, a CNS MRI signal could be detected transiently for some samples (three days) and more long term for others (28 days). These exposures were not associated with pathological changes to the CNS of normal rats. However, in one of the three tumour models investigated, some tumour enhancement was observed in one animal. The authors suggest that these results demonstrate the safety of the commercially available iron oxide preparations used for MRI. Further work is required to verify this result. Hussain et al. (2005) compared the impact of a range of nanoparticles, including iron oxide (Fe3O4), TiO2 and silver on the rat liver derived cell line BRL 3A. As described previously, the silver (5–50 µg/ml) nanoparticles decreased mitochondrial function (MTT assay) and cell death (LDH assay), while Fe3O4 and TiO2 at the same concentration had no significant impact on either endpoint. Concentrations had to be increased to 100–250 µg/ml before significant effects of these two metal oxides could be detected in these two assays. 9.3.5

Quantum Dots

Quantum dots are semi-conducting nanoparticles often made from cadmium selenide (CdSe) or cadmium telluride (CdTe). These nanoparticles generate light, the

378

Environmental and Human Health Impacts of Nanotechnology

emission spectra of which varies according to particle size. This makes quantum dots a potentially useful alternative to the relatively unstable fluorescent organic dies currently used for immunofluorescence (Xing et al., 2007). Perhaps more interestingly, quantum dots have also been developed for use in medical applications since they are relatively photostable and their light emission can be detected from a whole organism using digital imaging devices (Giepmans et al., 2005; Jaiswal et al., 2003). Targeting of the quantum dot to specific locations in the body, for example a tumour, may be possible through the manipulation of particle size and surface modifications (e.g. antibodies), therefore allowing diagnosis of disease without the need for surgery (Ballou et al., 2004; Michalet et al., 2005). The advantages of using quantum dots as diagnostic tools have been demonstrated in a number of studies (Ballou et al., 2004; Xing et al., 2007). For example, one study by Gao and Nie (2004) reported that both subcutaneous and systemic injection resulted in the accumulation of multifunctional quantum dots probes within human prostate tumours grown in mice. Accumulation was mediated due to the enhanced permeability of blood vessel walls in the tumour site, as well as the antibody binding to cancer-specific cell surface biomarkers. Quantum dots are often coated with a shell consisting of zinc sulfide. This shell has been introduced to improve the stability of the cadmium containing core, with the aim to reduce leaching of this toxic component (Derfus et al., 2004). In addition, further surface modifications include organic (neutral charge and hydrophobic), amino or amine (positively charged and hydrophilic) or carboxyl (negatively charged and hydrophilic) chemically bonded groups. The polymer (PEG) can also be attached to the particle surface, with the aim that PEG provides a steric repulsive barrier resulting in lack of recognition by blood proteins and phagocytic cells (Porter et al., 1992; Gref et al., 1994). The consequence of avoiding uptake by macrophages is that the particles increase their residence time in the blood and increase their chance of reaching the required target. Although quantum dots appear to be an effective bio-imaging diagnostic tool, knowledge relating to their toxicity is fairly limited. Unlike the low toxicity, low solubility carbon black and TiO2 nanoparticles, these nanoparticles include constituents which are inherently toxic. For example, cadmium is associated with renal toxicity. A limited number of studies have investigated quantum dot toxicity. Derfus et al. (2004) identified that uncoated cadmium selenide quantum dots induced a concentration dependent toxicity in hepatocytes, which was related to a release of free cadmium ions (Cd2+) from the quantum dot core. Derfus et al. (2004) hypothesised that the addition of a shell of zinc sulfide, dihydrolipoic acid or a polymer based material to quantum dots would reduce their toxicity. However, other studies have shown cadmium ions to be released from quantum dots both with and without a shell layer (Kloepfer et al., 2003). It has been suggested that the shell might in fact decrease the stability of the quantum dots and subsequently cause a greater toxic response (Michalet, et al. 2005). The impact of coating on quantum dot stability and toxicity varies substantially between different studies from different groups. The cause of this variation is unclear, but is likely to stem from differences in the properties of the quantum dots used, as well as their quality. Quantum dots with a negatively charged, carboxylated

Human Toxicology and Effects of Nanoparticles

379

chemical surface coating have been demonstrated to be less toxic than positively charged, amine coated quantum dots. The amine coated quantum dots appeared to be less stable over time than the carboxylated quantum dots (Hoshino et al., 2004b). Shiohara et al. (2004) also found that positively charged CdSe/ZnS quantum dots exhibited reduced stability. It was further observed by Shiohara et al. (2004) that CdSe/ZnS quantum dots were highly toxic irrespective of coating and induced cell death even at low concentrations. Lovric et al. (2005b) also investigated unmodified quantum dots with a cadmium telluride (CdTe) core. In a breast cancer cell line, they identified damage to the plasma membrane, mitochondria and nucleus. Pretreatment of the cells with antioxidants prevented the quantum dot induced toxicity, suggesting a role for ROS. The same group also identified that the localisation of the CdTe quantum dots in neuronal cell lines was dependent upon particle size (Lovric et al., 2005a). In addition, the smaller quantum dots were more toxic as associated by membrane blebbing and chromatin condensation. Again antioxidants provided protection against the quantum dot induced cytotoxicity (Lovric et al. 2005a).

9.4 Relating Physico-Chemical Properties to Toxicity: Structure-Activity Relationships The question ‘what is so special about nanoparticles that makes them toxic?’ is one that toxicologists working in this area have heard many times. Nanoparticles are manufactured because they possess properties that are not present in the bulk material or larger particles. Such properties include surface reactivity, light emission and electrical conductivity. These properties occur for a number of reasons, including the high proportion of atoms at the particle surface (Chapter 2). The more atoms there are at the particle surface, the greater the number to participate in reactions with other molecules in the environment or in the body. For example, a 100 µm particle has 0.0001% of its atoms at the surface, while a particle of the same chemical composition but just 10 nm in size has 10% of its atoms at the particle surface (Stone et al., 2007). Furthermore, as the particle size becomes smaller there is less room for electrons to orbit around each atom within the particle; restriction of this movement means that the atoms change their behaviour and therefore the particle characteristics. In addition, bond angles between atoms within a nanoparticle may be restricted, so that they are not optimum for stability, especially below 5 nm in diameter. This again results in instability and the potential for increased reactivity (Stone et al. 2007). Therefore, the very properties for which nanoparticles are manufactured and exploited are also the factors that are likely to drive their biological reactivity. In order to determine the relationship between physico-chemical characteristics it is essential that particles used and published in toxicology studies are well characterised in terms of size, surface area and composition. Our own work has shown that surface area is important in driving the ability of low solubility, low toxicity nanoparticles to induce inflammation in the rat lung. Nanoparticles and larger particles of carbon black, polystyrene beads and TiO2

380

Environmental and Human Health Impacts of Nanotechnology

were instilled into the rat lung at different mass doses in order to generate a range of surface area doses. The data generated suggested a straight-line relationship between surface area instilled and the inflammation generated (Duffin et al., 2002, 2007). Furthermore, we also included alpha-quartz, a particle with a highly reactive surface, known to be very potent at inducing pulmonary inflammation leading to fibrosis (silicosis) and lung cancer. This pathogenic particle induced a much larger inflammatory reaction that was above the straight line for the low toxicity low solubility materials, suggesting that ability to induce inflammation is a function of both surface area and surface reactivity (Duffin et al., 2002, 2007). Solubility is also likely to be very important in determining toxicity of particles. If a nanoparticle is soluble then it will obviously release its resultant components. If these components are not toxic and the dissolution is relatively rapid (up to a few hours) then such a particle is unlikely to induce any significant toxicity or health effects. However, if the particle contains toxic materials, such as the cadmium in quantum dots, then toxicity could be related to the release of toxic solutes. The small size of the particles may mean that the particles dissolve quickly releasing a bolus dose of the toxin. However, note that smaller size and larger surface area do not always mean greater rate of dissolution (Chapter 3). In addition, if a nanoparticle is soluble, the removal of solutes by blood or tissue fluid may be sufficiently high to prevent concentrations reaching a toxic dose within any particular region of the body. However, it is also worth noting that many particles are ingested into cells by phagocytosis, allowing them to accumulate within intracellular structures such as endosomes or lysosomes, resulting in confined release of solutes allowing concentrations of solutes to accumulate. This could be further enhanced by the low pH of cellular components such as the lysosome. All of this is currently supposition, but it illustrates that solubility is likely to play a complex role in determining nanoparticle toxicity. As described for asbestos and high aspect ratio nanoparticles, the shape of nanoparticles is also likely to be important in influencing toxicity. However, in addition to fibre-like dimensions and length, rigidity is also important. A long but flexible nanotube may be relatively easy to clear if it will fold up inside a cell, whereas a rigid nanotube is more likely to exhibit fibre-like properties (Brown et al., 2007b). It is now important to determine which of the physico-chemical characteristics of nanoparticles are most associated with their biological effects (Figure 9.2). As outlined previously, an understanding of this relationship can be used to predict toxicity and prevent the requirement for toxicity testing of all types of nanoparticles. This process would be enhanced by the access of toxicologists to a bank of well characterised reference materials. This would allow comparisons to be made between laboratories and, once a database of information is generated in relation to the reference material toxicity, comparison between particles. A list of suggested reference materials has been proposed by the UK Government Department of the Environment, Food and Rural Affairs (Defra) funded project, REFNANO (Aitken et al., 2008). The particles included in this list were chosen because of their relevance to industry and therefore potential for exposure.

Human Toxicology and Effects of Nanoparticles

9.5

381

Suggestions for Future Study Designs

There are many considerations when designing a toxicity study for nanoparticles. This chapter will not outline all of these considerations but will finish by highlighting some key design factors that should be considered. (i)

The use of appropriate controls. It is obvious that all studies should include controls such as a negative control (untreated animal or cells), but other controls are also worth considering. For example, is there a positive control that is known to induce the effect that you are looking for? Better still, is there a particle that is known to induce this effect? Regarding negative controls, it might also be worth considering a particle as a negative control, perhaps a larger particle of the same chemical composition if this is available. (ii) Particle preparation. This is a big problem, and is anticipated to be an active area of research over the next few years. Our current understanding of routes of exposure is relatively poor, and this impacts on our ability to design appropriate exposure protocols. We may not be exposed to monodispersed versions of the nanoparticle tested and so it may not be necessary to go to extravagant lengths to generate a monodispersed suspension. When adding particle to organisms or cells, should we used a reagent to aid dispersion? Proteins such as those found in serum seem to be pretty effective in aiding particle dispersion and, of course, this is relevant for any nanoparticle gaining access to the body, especially the blood (Sager et al., 2007, Foucaud et al., 2007). If a particle is prepared in a formulation prior to use then it might also be more appropriate to check the particle toxicity dispersed in the formulation rather than the pristine particles. Whatever the final decision, it is important to make these as relevant to real life as is possible. (iii) Susceptibility. As shown for air pollution particles, individuals with pre-existing inflammatory diseases are more susceptible to induction of morbidity and mortality following exposure. Many of the toxicology models used currently employ healthy models, and so it is important to consider more appropriate models in the future. This is of course difficult, but models of cardiovascular disease, such as the ApoE mouse, are now available. (iv) Dose. The dose of nanoparticles that should be used in toxicology studies is difficult in the absence of exposure information. However, we do have information about the respiratory toxicity of particles such as PM10, quartz and asbestos and these can, until such time exposure information is available, be used as guidance to allow benchmarking of particle potency.

9.6

Conclusions

This rapidly evolving field is still in its infancy. Over the next decade it can be anticipated that significant advances will be made in terms of understanding the factors that influence and drive nanoparticle toxicity or safety. However,

382

Environmental and Human Health Impacts of Nanotechnology

nanotechnology is also developing at a rapid race and the particles that we are concerned about today may be just a small part of a much bigger array of nanostructures available for interaction with the body. There is no doubt that this will be a challenging but exciting opportunity for scientists engaged in this research area.

9.7 Abbreviations activator protein 1 British Standards Institute blood brain barrier bovine serum albumin carbon black carbon nanotubes central nervous system Department of the Environment, Food and Rural Affairs 2,7-dichlorofluorescin-diacetate dipalmitoyl phosphatidyl choline electron spin resonance high aspect ratio nanoparticles International Agency for Research on Cancer lactate dehydrogenase long fibre amosite magnetic resonance imaging major histocompatibility complex multiwalled carbon nanotubes. nuclear factor kappa B polyethylene glycol reactive oxygen species short fibre amosite single wall carbon nanotubes tetrahydrofuran transmission electron microscopy tumour necrosis factorα

9.8

(AP1) (BSI) (BBB) (BSA) (CB) (CNT) (CNS) (Defra) (DCFH-DA) (DPPC) (ESR), (HARN) (IARC) (LDH) (LFA) (MRI) (MHC) (MWCNT) (NFκB) (PEG) (ROS (SFA) (SWNT) (THF) (TEM) (TNFα)

References

Aitken, R. J., M. Q. Chaudhry, A. B. Boxall and M. Hull (2006) Manufacture and use of nanomaterials: current status in the UK and global trends. Occup. Med. (Lond), 56, 300–6. Aitken, R. J., S. M. Hankin, C. L. Tran, et al. (2008) A multidisciplinary approach to the identification of reference materials for engineered nanoparticle toxicology. Nanotoxicology, 2, 71–8. Alt, V., T. Bechert, P. Steinrucke, et al. (2004) An in vitro assessment of the antibacterial properties and cytotoxicity of nanoparticulate silver bone cement. Biomaterials, 25, 4383–91.

Human Toxicology and Effects of Nanoparticles

383

Baker, C., A. Pradhan, L. Pakstis, et al. (2005) Synthesis and antibacterial properties of silver nanoparticles. J. Nanosci. Nanotechnol., 5, 244–9. Ballou, B., B. C. Lagerholm, L. A. Ernst, et al. (2004) Noninvasive imaging of quantum dots in mice. Bioconjug. Chem., 15, 79–86. Barnes, K. A., S. E. Samson and A. K. Grover (2000) Sarco/endoplasmic reticulum Ca2+-pump isoform SERCA3a is more resistant to superoxide damage than SERCA2b. Mol. Cell Biochem., 203, 17–21. Berridge, M. J. (2004) Calcium signal transduction and cellular control mechanisms. Biochim. Biophys. Acta, 1742, 3–7. Borm, P. J., D. Robbins, S. Haubold, et al. (2006) The potential risks of nanomaterials: a review carried out for ECETOC. Part Fibre. Toxicol, 3, 11. Brandt, D., B. Park, M. Hoang and H. T. Jacobe (2005) Argyria secondary to ingestion of homemade silver solution. J. Am. Acad. Dermatol., 53, S105–7. Brant, J., H. Lecoanet, M. Hotze and M. Wiesner (2005) Comparison of electrokinetic properties of colloidal fullerenes (n-C60) formed using two procedures. Environ. Sci. Technol., 39, 6343–51. BSI (2007) Terminology for nanomaterials. PAS 136. British Standards Institution. http:// www.standardsuk.com/contact.php Brown, D. M., V. Stone, P. Findlay, et al. (2000) Increased inflammation and intracellular calcium caused by ultrafine carbon black is independent of transition metals or other soluble components. Occup. Environ. Med., 57, 685–91. Brown, D. M., M. R. Wilson, W. MacNee, et al. (2001) Size-dependent proinflammatory effects of ultrafine polystyrene particles: a role for surface area and oxidative stress in the enhanced activity of ultrafines. Toxicol. Appl. Pharmacol., 175, 191–9. Brown, D. M., K. Donaldson, P. J. Borm, R. P. Schins, M. Denhart, P. Gilmour, L. A. Jimenez and V. Stone (2004). Calcium and reactive oxygen species-mediated activation of transcription factors and TNFa cytokine gene expression in macrophages exposed to ultrafine particles. Am. J. Physiol Lung cell Mol. Physiol, 286, L344–L353. Brown, D. M., L. Hutchison, K. Donaldson, et al. (2007a). The effect of oxidative stress on macrophages and lung epithelial cells: the role of phosphodiesterases 1 and 4. Toxicol. Lett., 168, 1–6. Brown, D. M., I. A. Kinloch, U. Bangert, et al. (2007b). An in vitro study of the potential of carbon nanotubes and nanofibres to induce inflammation mediators and frustrated phagocytosis. Carbon, 45, 1743–56. Chen, X. and H. J. Schluesener (2008) Nanosilver: A nanoproduct in medical application. Toxicol. Lett., 176, 1–12. Choi, H., S. R. Choi, R. Zhou, et al. (2004) Iron oxide nanoparticles as magnetic resonance contrast agent for tumor imaging via folate receptor-targeted delivery. Acad. Radiol., 11, 996–1004. Christman, J. W., T. S. Blackwell and B. H. Juurlink (2000) Redox regulation of nuclear factor kappa B: therapeutic potential for attenuating inflammatory responses. Brain Pathol., 10, 153–62. Conner, S. D. and S. L. Schmid (2003) Regulated portals of entry into the cell. Nature, 422, 37–44. Da Ros, T. and M. Prato (1999) Medicinal chemistry with fullerenes and fullerene derivatives. Chemical Communications, 663–9. Derfus, A. M., W. C. W. Chan and S. N. Bhatia (2004) Probing the cytotoxicity of semiconductor quantum dots. Nano Letters, 1, 11–8. Dick, C. A., D. M. Brown, K. Donaldson and V. Stone (2003) The role of free radicals in the toxic and inflammatory effects of four different ultrafine particle types. Inhal. Toxicol., 15, 39–52. Dockery, D. W., C. A. III, Pope, X. Xu, et al. (1993) An association between air pollution and mortality in six U. S. cities. N. Engl. J. Med., 329, 1753–9. Donaldson, K., V. Stone, A. Clouter, et al. (2001a). Ultrafine particles. Occup. Environ. Med., 58, 211–6, 199.

384

Environmental and Human Health Impacts of Nanotechnology

Donaldson, K., V. Stone, R. Duffin, et al. (2001b). The quartz hazard: effects of surface and matrix on inflammogenic activity. J. Environ. Pathol. Toxicol. Oncol., 20 (Suppl 1), 109–18. Donaldson, K., V. Stone, A. Seaton and W. MacNee (2001c). Ambient particle inhalation and the cardiovascular system: potential mechanisms. Environ. Health Perspect., 109 (Suppl 4), 523–7. Donaldson, K., V. Stone, P. J. Borm, et al. (2003) Oxidative stress and calcium signalling in the adverse effects of environmental particles (PM(10)). Free Radic. Biol. Med., 34, 1369–82. Donaldson, K., C. L. Tran, L. A. Jimenez, et al. (2005) Combustion-derived nanoparticles: A critical review of their toxicology following inhalation exposure. Particle Fibre Toxicol., 2, 1–14. Donaldson, K., R. Aitken, C. L. Tran, et al. (2006) Carbon Nanotubes: A review of their properties in relation to pulmonary toxicology and workplace safety. Toxicol. Sci. Epub. Duffin, R., C. L. Tran, A. Clouter, et al. (2002) The importance of surface area and specific reactivity in the acute pulmonary inflammatory response to particles. Ann. Occup. Hyg., 46 (Suppl. 1), 242–5. Duffin, R., L. Tran, D. Brown, et al. (2007) Proinflammogenic effects of low-toxicity and metal nanoparticles in vivo and in vitro: highlighting the role of particle surface area and surface reactivity. Inhal. Toxicol., 19, 849–56. Edwards-Jones, V. (2006) Antimicrobial and barrier effects of silver against methicillinresistant Staphylococcus aureus. J. Wound. Care, 15, 285–90. Fall, M., M. Guerbet, B. Park, et al. (2007) Evaluation nof cerium oxide and cerium oxide based fuel additive safety on organotypic cultures of lung slices. Nanotoxicology, 1, 226–33. Ferin, J., G. Oberdorster and D. P. Penney (1992) Pulmonary retention of ultrafine and fine particles in rats. Am. J. Respir. Cell Mol. Biol., 6, 535–42. Ferin, J., G. Oberdoster, D. P. Penney, et al. (1990) Increased pulmonary toxicity of ultrafine particles? I. Particle clearance, translocation, morphology. J. Aerosol. Sci., 21, 381–4. Fiorito, S., A. Serafino, F. Andreola, et al. (2006) Toxicity and biocompatibility of carbon nanoparticies. Journal of Nanoscience and Nanotechnology, 6, 591–9. Foucaud, L., M. R. Wilson, D. M. Brown and V. Stone (2007) Measurement of reactive species production by nanoparticles prepared in biologically relevant media. Toxicol. Lett., 174, 1–9. Furno, F., K. S. Morley, B. Wong, et al. (2004) Silver nanoparticles and polymeric medical devices: a new approach to prevention of infection? J. Antimicrob. Chemother., 54, 1019–24. Gamer, A. O., E. Leibold and R. B. van (2006) The in vitro absorption of microfine zinc oxide and titanium dioxide through porcine skin. Toxicol. In Vitro, 20, 301–7. Gao, X. and S. Nie (2004) Quantum dot-encoded mesoporous beads with high brightness and uniformity: rapid readout using flow cytometry. Anal. Chem., 76, 2406–10. Geiser, M., N. Kapp, S. Schurch, et al. (2005) Ultrafine particles cross celullar membranes by non-phagocytic mechanisms in lungs and in cultured cells. Environ. Health Perspect, 113, 1555–1560. Gharbi, N., M. Pressac, M. Hadchouel, et al. (2005) [60]Fullerene is a powerful antioxidant in vivo with no acute or subacute toxicity. Nano Letters, 5, 2578–85. Giepmans, B. N., T. J. Deerinck, B. L. Smarr, et al. (2005) Correlated light and electron microscopic imaging of multiple endogenous proteins using Quantum dots. Nat. Methods, 2, 743–9. Gojova, A., B. Guo, R. S. Kota, et al. (2007) Induction of inflammation in vascular endothelial cells by metal oxide nanoparticles: effect of particle composition. Environ. Health Perspect., 115, 403–9. Gref, R., Y. Minamitake, M. T. Peracchia, et al. (1994) Biodegradable long-circulating polymeric nanospheres. Science, 263, 1600–3.

Human Toxicology and Effects of Nanoparticles

385

Gulbranson, S. H., J. A. Hud and R. C. Hansen (2000) Argyria following the use of dietary supplements containing colloidal silver protein. Cutis, 66, 373–4. Guo, L., D. G. Morris, X. Liu, et al. (2007) Iron bioavailability and redox activity in diverse carbon nanotube samples. Chemistry of Materials, 19, 3472–8. Halford, B. (2006) Fullerene for the face. Chemical & Engineering News, 84, 47. Hoet, P. H. and B. Nemery (2001) Stimulation of phagocytosis by ultrafine particles. Toxicol. Appl. Pharmacol., 176, 203. Hoshino, A., K., Fujioka, T. Oku, S. Masakazu, Y. F. Sasaki, T. Ohta, M. Yasuhara, K. Suzuki and K. Yamamoto (2004b). Physicochemical properties and cellular toxicity of nanocrystal quantum dots depend on their surface modification. Nano Letters, 4 (11), 2163–2169. Hussain, S. M., K. L. Hess, J. M. Gearhart, et al. (2005). In vitro toxicity of nanoparticles in BRL 3A rat liver cells. Toxicol. In Vitro, 19, 975–83. IARC (1987) Summaries & Evaluations. Supplement 7. ASBESTOS (Actinolite, amosite, anthophyllite, chrysotile, crocidolite, tremolite) (Group 1), World Health Organisation, International Agency for Research on Cancer, Lyon, France. IARC (1997) Monographs on the Evaluation of Carcinogenic Risks to Humans, Silica, Some Silicates, Coal Dist and para-Aramid Fibrils. World Health Organisation, International Agency for Research on Cancer Lyon. 68, Lyon, France. Jaiswal, J. K., H. Mattoussi, J. M. Mauro and S. M. Simon (2003) Long-term multiple color imaging of live cells using quantum dot bioconjugates. Nat. Biotechnol., 21, 47–51. Ji, J. H., J. H. Jung, S. S. Kim, et al. (2007) Twenty-eight-day inhalation toxicity study of silver nanoparticles in Sprague-Dawley rats. Inhal. Toxicol., 19, 857–71. Kaewamatawong, T., A. Shimada, M. Okajima, et al. (2006) Acute and subacute pulmonary toxicity of low dose of ultrafine colloidal silica particles in mice after intratracheal instillation. Toxicol. Pathol., 34, 958–65. Kagan, V. E., Y. Y. Tyurina, V. A. Tyurin, et al. (2006) Direct and indirect effects of single walled carbon nanotubes on RAW 264.7 macrophages: Role of iron. Toxicol. Lett., 165, 88–100. Khalil, I. A., K. Kogure, H. Akita and H. Harashima (2006). Uptake pathways and subsequent intracellular trafficking in nonviral gene delivery. Pharmacol. Rev., 58, 32–45. Klaassen, C. D. (2001) Casarett & Doull’s Toxicology: The Basic Science of Poisons. McGrawHill, United States of America. Kloepfer, J. A., R. E. Mielke, M. S. Wong, et al. (2003) Quantum dots as strain- and metabolism-specific microbiological labels. Appl. Environ. Microbiol., 69, 4205–13. Kourie, J. I. (1998) Interaction of reactive oxygen species with ion transport mechanisms. Am. J. Physiol, 275, C1–24. Koyama S., M. Endo, Y.-A. Kim, et al. (2006) Role of systemic T-cells and histopathological aspects after subcutaneous implantation of various carbon nanotubes in mice. Carbon, 44, 1079–92. Kroto, H. W., J. R. Heath, S. C. Obrien, et al. (1985) C-60 – Buckminsterfullerene. Nature, 318, 162–3. Lacerda, L., A. Bianco, M. Prato and K. Kostarelos (2006) Carbon nanotubes as nanomedicines: from toxicology to pharmacology. Adv. Drug Deliv. Rev., 58, 1460–70. Lam, C. W., J. T. James, R. McCluskey and R. L. Hunter (2004). Pulmonary toxicity of singlewall carbon nanotubes in mice 7 and 90 days after intratracheal instillation. Toxicol. Sci., 77, 126–134. Li, X. Y., D. Brown, S. Smith, et al. (1999) Short-term inflammatory responses following intratracheal instillation of fine and ultrafine carbon black in rats. Inhal. Toxicol., 11, 709–31. Li, Y., P. Leung, L. Yao, et al. (2006) Antimicrobial effect of surgical masks coated with nanoparticles. J. Hosp. Infect., 62, 58–63. Li, Z., T. Hulderman, R. Salmen, et al. (2007) Cardiovascular effects of pulmonary exposure to single-wall carbon nanotubes. Environ. Health Perspect., 115, 377–82. Lin, W., Y. W. Huang, X. D. Zhou and Y. Ma (2006a). In vitro toxicity of silica nanoparticles in human lung cancer cells. Toxicol. Appl. Pharmacol, 217, 252–9.

386

Environmental and Human Health Impacts of Nanotechnology

Lin, W., Y. W. Huang, X. D. Zhou and Y. Ma (2006b). Toxicity of cerium oxide nanoparticles in human lung cancer cells. Int. J. Toxicol., 25, 451–7. Liu, J. and J. I. Shapiro (2003). Endocytosis and signal transduction: basic science update. Biol. Res. Nurs., 5, 117–128. Long, T. C., J. Tajuba, P. Sama, et al. (2007) Nanosize titanium dioxide stimulates reactive oxygen species in brain microglia and damages neurons in vitro. Environ. Health Perspect., 115, 1631–7. Lovric, J., H. S. Bazzi, Y. Cuie, et al. (2005a). Differences in subcellular distribution and toxicity of green and red emitting CdTe quantum dots. J. Mol. Med., 83, 377–385. Lovric, J., S. J. Cho, F. M. Winnik and D. Maysinger (2005b). Unmodified cadmium telluride quantum dots induce reactive oxygen species formation leading to multiple organelle damage and cell death. Chem. Biol., 12, 1227–34. Lundborg, M., S. E. Dahlen, U. Johard, et al. (2006) Aggregates of ultrafine particles impair phagocytosis of microorganisms by human alveolar macrophages. Environ. Res., 100, 197–204. MacNee, W. and I. Rahman (2001) Is oxidative stress central to the pathogenesis of chronic obstructive pulmonary disease? Trends in Molecular Medicine, 7, 55–62. Markovic, Z., B. Todorovic-Markovic, D. Kleut, et al. (2007) The mechanism of cell-damaging reactive oxygen generation by colloidal fullerenes. Biomaterials, 28, 5437–48. Mavon, A., C. Miquel, O. Lejeune, et al. (2007) In vitro percutaneous absorption and in vivo stratum corneum distribution of an organic and a mineral sunscreen. Skin Pharmacol. Physiol, 20, 10–20. Maynard, A. D. and R. J. Aitken (2007) Assessing exposure to airborne nanomaterials: Current abilities and future requirements. Nanotoxicology, 1, 26–41. Maynard, A. D., P. A. Baron, M. Foley, et al. (2004) Exposure to carbon nanotube material: aerosol release during the handling of unrefined single-walled carbon nanotube material. J. Toxicol. Environ. Health A, 67, 87–107. Michalet, X., F. F. Pinaud, L. A. Bentolila, et al. (2005) Quantum dots for live cells, in vivo imaging, and diagnostics. Science, 307, 538–44. Mills, N. L., H. Tornqvist, S. D. Robinson, et al. (2005) Diesel exhaust inhalation causes vascular dysfunction and impaired endogenous fibrinolysis. Circulation, 112, 3930–6. Monteiro-Riviere, N. A., R. J. Nemanich, A. O. Inman, et al. (2005) Multi-walled carbon nanotube interactions with human epidermal keratinocytes. Toxicol. Lett., 155, 377–84. Montellier, C., C. L. Tran, W. MacNee, et al. (2006) The pro-inflammatory effects of low solublity low toxicity particles, nanoparticles and fine particles on epithelial cells in vitro: the role of surface area and surface reactivity. Occup. Environ. Med., 64, 609–15. Muldoon, L. L., M. Sandor, K. E. Pinkston and E. A. Neuwelt (2005) Imaging, distribution, and toxicity of superparamagnetic iron oxide magnetic resonance nanoparticles in the rat brain and intracerebral tumor. Neurosurgery, 57, 785–96. Muller, J., F. Huaux, N. Moreau, et al. (2005) Respiratory toxicity of multi-wall carbon nanotubes. Toxicol. Appl. Pharmacol., 207, 221–31. Oberdoerster, E. (2004) Manufactured nanomaterials (Fullerenes, C60) induce oxidative stress in the brain of juvenile largemouth Bass. Environ. Health Perspect., 112, 1058–62. Oberdoerster, G., J. Ferin, J. Finkelstein, et al. (1990) Increased pulmonary toxicity of ultrafine particles? II. Lung lavage studies. J. Aerosol. Sci., 21, 384–7. Oberdoerster, G., J. Ferin and B. E. Lehnert (1994) Correlation between particle size, in vivo particle persistence, and lung injury. Environ. Health Perspect., 102 (Suppl 5), 173–9. Oberdoerster, G., V. Stone and K. Donaldson (2007) Toxicology of nanoparticles: A historical perspective. Nanotoxicology, 1, 2–25. Orrenius, S., M. J. Burkitt, G. E. Kass, et al. (1992) Calcium ions and oxidative cell injury. Ann. Neurol., 32 Suppl, S33–42. Park, B., P. Martin, C. Harris, et al. (2007) Initial in vitro screening approach to investigate the potential health and environmental hazards of Envirox™ – a nanoparticulate cerium oxide diesel fuel additive. Particle Fibre Toxicol., 4, 12.

Human Toxicology and Effects of Nanoparticles

387

Patel, L. N., J. L. Zaro, and W. C. Shen (2007). Cell penetrating peptides: intracellular pathways and pharmaceutical perspectives. Pharm. Res., 24, 1977–1992. Peek, L. J., C. R. Middaugh and C. Berkland (2008) Nanotechnology in vaccine delivery. Adv. Drug Deliv. Rev., 60, 915–928. Perret, E., A. Lakkaraju, S. Deborde, R. Schreiner and E. Rodriguez-Boulan (2005). Evolving endosomes: how many varieties and why? Curr. Opin. Cell Biol., 17, 423–434. Pflucker, F., V. Wendel, H. Hohenberg, et al. (2001) The human stratum corneum layer: an effective barrier against dermal uptake of different forms of topically applied micronised titanium dioxide. Skin Pharmacol. Appl. Skin Physiol, 14 (Suppl 1), 92–7. Pisanic, T. R., J. D. Blackwell, V. I. Shubayev, et al. (2007) Nanotoxicity of iron oxide nanoparticle internalization in growing neurons. Biomaterials, 28, 2572–81. Poland, C. A., R. Duffin, I. A. Kinloch, et al. (2008) Carbon nanotubes introduced into the abdominal cavity ofmice show asbestos-like pathogenicity in a pilot study. Nature Nanotechnology, 3, 423–428. Porter, C. J., S. M. Moghimi, L. Illum and S. S. Davis (1992) The polyoxyethylene/polyoxypropylene block co-polymer poloxamer-407 selectively redirects intravenously injected microspheres to sinusoidal endothelial cells of rabbit bone marrow. FEBS Lett., 305, 62–6. Renwick, L. C., K. Donaldson and A. Clouter (2001) Impairment of alveolar macrophage phagocytosis by ultrafine particles. Toxicol. Appl. Pharmacol., 172, 119–27. Sager, T. M., D. W. Porter, V. A. Robinson, et al. (2007) Improved method to disperse nanoparticles for in vitro and in vivo investigation of toxicity. Nanotoxicology, 1, 118–29. Sayes, C. M., A. M. Gobin, K. D. Ausman, et al. (2005) Nano-C60 cytotoxicity is due to lipid peroxidation. Biomaterials, 26, 7587–95. Sayes, C. M., K. L. Reed and D. B. Warheit (2007) Assessing toxicity of fine and nanoparticles: comparing in vitro measurements to in vivo pulmonary toxicity profiles. Toxicol. Sci., 97, 163–80. Schwartz, J. (1994) Air pollution and daily mortality: a review and meta analysis. Environ. Res., 64, 36–52. Seaton, A., W. MacNee, K. Donaldson and D. Godden (1995) Particulate air pollution and acute health effects. Lancet, 345, 176–8. Semmler-Behnke, M., S. Takenaka, S. Fertsch, et al. (2007) Efficient elimination of inhaled nanoparticles from the alveolar region: evidence for interstitial uptake and subsequent reentrainment onto airways epithelium. Environ. Health Perspect., 115, 728–33. Shiohara, A., A. Hoshino, K. Hanaki, et al. (2004) On the cyto-toxicity caused by quantum dots. Microbiol. Immunol., 48, 669–75. Shvedova, A. A., V. Castranova, E. R. Kisin, et al. (2003) Exposure to carbon nanotube material: assessment of nanotube cytotoxicity using human keratinocyte cells. J. Toxicol. Environ. Health A, 66, 1909–26. Shvedova, A. A., E. R. Kisin, R. Mercer, et al. (2005) Unusual inflammatory and fibrogenic pulmonary responses to single-walled carbon nanotubes in mice. Am. J. Physiol Lung Cell Mol. Physiol, 289, L698–708. Singh, R., D. Pantarotto, L. Lacerda, et al. (2006) Tissue biodistribution and blood clearance rates of intravenously administered carbon nanotube radiotracers. Proc. Natl. Acad. Sci. USA, 103, 3357–62. Stoeger, T., C. Reinhard, S. Takenaka, A. Schroeppel, E. Karg, B. Ritter, J. Heyder and H. Schulz (2006). Instillation of six different ultrafine carbon particles indicates a surface area threshold dose for acute lung inflammation in mice. Environ. Health Perspect., 114, 328–333. Stohs, S. J. and D. Bagchi (1995) Oxidative mechanisms in the toxicity of metal ions. Free Radic. Biol. Med., 18, 321–36. Stone, V., J. Shaw, D. M. Brown, et al. (1998) The role of oxidative stress in the prolonged inhibitory effect of ultrafine carbon black on epithelial cell function. Toxicol. In Vitro, 12, 649–59.

388

Environmental and Human Health Impacts of Nanotechnology

Stone, V., D. M. Brown, N. Watt, et al. (2000a). Ultrafine particle-mediated activation of macrophages: Intracellular calcium signalling and oxidative stress. Inhal. Toxicol., 12(Suppl.3), 345–51. Stone, V., M. Tuinman, J. E. Vamvakopoulos, et al. (2000b). Increased calcium influx in a monocytic cell line on exposure to ultrafine carbon black. Eur. Respir. J., 15, 297–303. Stone, V. and I. A. Kinloch (2007) Nanoparticle interactions with biological systems and subsequent activation of intracellular signalling mechanisms. In Nanotoxicology (eds N. A. Monteiro-Riviere, and C. L. Tran), CRC Press, London. Takano, H., T. Ichinose, Y. Miyabara, et al. (1998) Inhalation of diesel exhaust enhances allergen-related eosinophil recruitment and airway hyperresponsiveness in mice. Toxicol. Appl. Pharmacol., 150, 328–37. Takenaka, S., E. Karg, W. G. Kreyling, et al. (2006) Distribution pattern of inhaled ultrafine gold particles in the rat lung. Inhal. Toxicol., 18, 733–40. Timbrell, J. (2002) Introduction to Toxicology. CRC Press, London. Tremblay, J. F. (2003) Fullerenes by the ton. Chemical & Engineering News, 81, 13–4. van Bokhoven, J. A., C. Louis, J. T. Miller, et al. (2006) Activation of oxygen on gold/alumina catalysts: in situ high-energy-resolution fluorescence and time-resolved X-ray spectroscopy. Angew. Chem. Int. Ed Engl., 45, 4651–4. Vogelson, C. T. (2001) Advances in drug delivery systems. Mod. Drug. Discov., 4, 49–50. Wang, I. C., L. A. Tai, D. D. Lee, et al. (1999) C-60 and water-soluble fullerene derivatives as antioxidants against radical-initiated lipid peroxidation. Journal of Medicinal Chemistry, 42, 4614–20. Warheit, D. B., B. R. Laurence, K. L. Reed, et al. (2004). Comparative Pulmonary Toxicity Assessment of Single-wall Carbon Nanotubes in Rats. Toxicol. Sci., 77, 117–25. Warheit, D. B., T. R. Webb, K. L. Reed, et al. (2007) Pulmonary toxicity study in rats with three forms of ultrafine-TiO2 particles: differential responses related to surface properties. Toxicology, 230, 90–104. Watts, C. and M. Marsh (1992) Endocytosis: what goes in and how? J. Cell Sci., 103 (Pt 1), 1–8. White, J. M., A. M. Powell, K. Brady and R. Russell-Jones (2003) Severe generalized argyria secondary to ingestion of colloidal silver protein. Clin. Exp. Dermatol., 28, 254–6. Wilson, M. R., J. H. Lightbody, K. Donaldson, et al. (2002) Interactions between ultrafine particles and transition metals in vivo and in vitro. Toxicol. Appl. Pharmacol., 184, 172–9. Wilson, M. R., L. Foucaud, P. G. Barlow, et al. (2007). Nanoparticle interactions with zinc and iron: Implications for toxicology and inflammation. Toxicol. Appl. Pharmacol., 225, 80–9. Xiao, L., H. Takada, X. Gan and N. Miwa (2006) The water-soluble fullerene derivative ‘Radical Sponge’ exerts cytoprotective action against UVA irradiation but not visiblelight-catalyzed cytotoxicity in human skin keratinocytes. Bioorg. Med. Chem Lett., 16, 1590–5. Xing, Y., Q. Chaudry, C. Shen, et al. (2007) Bioconjugated quantum dots for multiplexed and quantitative immunohistochemistry. Nat. Protoc., 2, 1152–65. Yokoyama, A., Y. Sato, Y. Nodasaka, et al. (2005) Biological behavior of hat-stacked carbon nanofibers in the subcutaneous tissue in rats. Nano Lett., 5, 157–61.

10 Risk Assessment of Manufactured Nanomaterials Sophie A. Rocks1, Simon J. Pollard1, Robert A. Dorey2, Paul T.C. Harrison3, Leonard S. Levy3, Richard D. Handy4, John F. Garrod5 and Richard Owen6 1

Collaborative Centre of Excellence in Understanding and Managing Natural and Environmental Risks, Cranfield University, United Kingdom 2 Microsystems & Nanotechnology Centre, Cranfield University, United Kingdom 3 Institute of Environment and Health, Cranfield University, United Kingdom 4 School of Biological Sciences, University of Plymouth, United Kingdom 5 Department for Environment, Food and Rural Affairs, London, United Kingdom 6 School of Biosciences, University of Westminster, United Kingdom

10.1

Introduction

Risk assessment is the qualitative or quantitative estimation of the likelihood of adverse effects that may result from exposure to a substance or hazardous situation. Its purpose is to inform decision makers on how best to manage risk preventatively so as to avoid harm. This is achieved through understanding the magnitude and significance of the risk and, by virtue of the risk assessment, through a thorough evaluation of the features of an exposure situation (scenario) that influence the risk most strongly. Considered alongside issues of cost, social issues, management arrangements, practicality, legality and the availability of technology, the key contributors to significant risk then become the focus for intervention (Pollard, 2006). Risk assessment has been applied extensively to chemicals in the context of environmental protection (Pollard et al., 2002; Ashmore and Nathanail, 2008), in

Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

390

Environmental and Human Health Impacts of Nanotechnology

public health protection and food safety (Keerotipibul and Lekroengsin, 2008; Rowbotham et al., 2000), and for the management of occupational exposure in the workplace (Duhayon et al., 2008; van Leeuwen and Vermeire, 2007). Risk assessments are routinely performed on individual chemicals or groups of chemicals (e.g. polyaromatic hydrocarbons, pesticides and metals) and specific exposure scenarios subject to multiple hazards (e.g. catchment-based risk assessments). Risk assessments are inevitably technically detailed. One challenge for risk assessors and policy makers is to ensure that the process of risk assessment remains evidence based and transparent, so that the logic of the subsequent decisions is clearly articulated, with the prospect that stakeholders may have confidence in the risk management decisions that follow. There has been considerable debate as to whether manufactured nanomaterials pose significant risks to the environment and human health (RS/RAEng, 2004; Owen and Depledge, 2005; SCENIHR, 2005; Owen and Handy, 2007). As with many emerging risks, the evidence base is currently incomplete, though important features are emerging. Current knowledge gaps include a lack of data on exposure and biological effects, which are essential data inputs to most risk assessment processes. Addressing these gaps is critical for the development of proportionate, evidence-based regulatory controls and for facilitating the responsible development of nanotechnologies. In response to the identified knowledge gaps, and given that many nanomaterials are already in wide use, the international approach to policy development has been to initiate: • the development of an evidence base to support appropriate controls through the commissioning of research (e.g. United States Environmental Protection Agency (US EPA) National Centre for Environmental Research (NCER) programme (http://es.epa.gov/ncer/nano/), European Union (EU) Framework Programme (http://cordis.europa.eu/nanotechnology/) and UK Environmental Nanoscience Initiative (http://www.nerc.ac.uk/research/ programmes/nanoscience/); • voluntary reporting and stewardship schemes for industry to provide information on the types of nanomaterials being manufactured and their applications (e.g. Nanotechnology Industries Association at http://www.nanotechia.co.uk). These schemes aim to complement the scientific research programmes, together providing the evidence basis; and • comprehensive regulatory reviews (e.g. in the UK by Defra (Frater et al., 2006) and in the EU by ECC (ECC, 2008)) analysing how, and to what extent, current regulations cover nanomaterials and their safety. Nanomaterials are materials of a specific size (considered as materials with at least one dimension of 1–100 nm; SCENIHR, 2006), and comprise of many different physico–chemical forms. Therefore, one approach to their regulatory control is to bring them within existing international chemicals legislation (EP, 2006) where a sound international consensus exists on how risks should be assessed and managed. However, the development of a regulatory strategy for nanomaterials requires structure. Two key questions are emerging:

Risk Assessment of Manufactured Nanomaterials

391

i. Whether current risk assessment methodologies for chemicals are ‘fit for purpose’ given the unique issues presented by nanomaterials. ii. Whether the current legislation that requires risk assessments (e.g. REACH in the European Union) covers nanomaterials sufficiently. On the international stage, the Organisation for Economic Co-operation and Development (OECD) is taking a leading role in exchanging information on risk assessment approaches for nanomaterials through its programme of work under the Working Party for Manufactured Nanomaterials (http://www.oecd.org/environment/nanosafety). The primary question is a matter of how risk assessments should be configured for nanomaterials. The EU Commission’s Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR) has offered ‘opinions’ on this (SCENIHR, 2005, 2007; see Section 10.5). The United Kingdom has also examined the fitness for purpose of ecotoxicological assessments for nanomaterials (Crane and Handy, 2007; Crane et al., 2008). A further evaluation of test methods using reference nanomaterials is being undertaken by the OECD (Working Party on Manufactured Nanomaterials, Steering Group 4; http://www.oecd.org/ department/0,3355,en_2649_37015404_1_1_1_1_1,00.html).

10.2

Risk Assessment Process

Chemical risk assessment has been developed over many years through the processes of national and international consensus (Risk Assessment and Toxicology Steering Committee, 1999). Conventionally, it comprises: (i) Hazard identification identifies possible adverse biological effects on organisms (the ‘receptor’) using data published in the peer reviewed literature, the collection of new toxicological data from in vitro and/or in vivo studies, or chemical modelling (e.g. quantitative structure activity relationships, QSAR). Hazard identification may include hazards from metabolites or reactive intermediates, not just the parent compound, and includes a consideration of tolerant and sensitive organisms. This is important in ecosystems, where the sensitive organism is a keystone or protected species, and in human health, where it may be unacceptable to not protect sensitive individuals (e.g. slow metabolisers, or patients with history of hypersensitivity). (ii) Hazard assessment establishes the existence of exposure pathways and quantitatively evaluates the observed adverse effects including dose-response assessment, species differences or sensitivity distributions as well as mechanisms of action. Here, quantitative estimates of hazard (e.g. predicted no effects concentrations, lethal or sub-lethal effects concentrations) may be estimated. (iii) Risk estimation addresses the potential risk to the identified receptors via each of the identified exposure pathways and involves an estimation of intake or exposure to a chemical (magnitude, duration and frequency) for the general population, sub-groups or individuals. (iv) Risk evaluation combines the stages above to draw conclusions on the significance of the risk posed.

392

Environmental and Human Health Impacts of Nanotechnology

Regulatory approaches to risk assessment have been developed internationally for assessing existing and new high volume production chemicals (NRC, 1983; IPCS, 1999; REACH technical guidance documents, accessed at http://echa.europa.eu/ reach_en.asp; EC JRC, 2003). Notwithstanding the value of these approaches and their systematic processes, much of the fundamental science on which risk assessments are based have a degree of uncertainty attached. Given this, it is essential to recall the purpose of the assessment – to judge the significance of the risk and inform decisions on how best risks should be preventatively managed. It is essential to ensure that risk estimates do not infer a level of precision that the underpinning evidence base cannot support. In practice, risk assessments incorporate both qualitative and quantitative elements (Harrison and Holmes, 2006) and are applied using tiered approaches that allow for early screening and prioritisation with the defensible and judicious use of quantitative techniques for key features of the problem under study (DETR et al., 2000; Figure 10.1). Risk assessors have learnt, given the complexity and data needs involved, that maintaining a focus on the essential questions of ‘what/who is at risk?’, and ‘what is it/are they at risk from?’ is critical for sound decisions on managing risk. Good problem formulation, conducted in a consultative and participative fashion with inputs from a range of stakeholders, guides the assessment towards the key exposure routes and biological effects of concern (Pollard, 2006; Owen and Handy, 2007).

Problem formulation Stages within each tier Risk prioritisation

Tiered Risk Assessment Hazard identification Tier 1 Qualitative

Hazard assessment

Tier 2 Generic quantitative

Risk estimation

Tier 3 Detailed quantitative

Risk evaluation

Options Appraisal Economics

Technology

Social issues

Management

Risk management Continuing data collection, monitoring and process iteration

Figure 10.1 Design of risk assessment and risk management framework showing qualitative and quantitative risk assessment and the stages required for each tier (after DETR, et al., 2000).

Risk Assessment of Manufactured Nanomaterials

393

Despite its importance, there has been little systematic consideration of problem formulation for manufactured nanomaterials, either generically within risk assessment methodologies or, more specifically, for one substance or a mixture of substances (Owen and Handy, 2007). Risk assessments may be requested by regulators as the basis for coordinating dispersed or incomplete evidence on risks. In these situations, whilst a risk assessment framework may provide a valuable ‘route map’ for decision makers, risk assessors still have to assemble and structure the evidence to support the safe use of a chemical. This includes identifying individual lines and strength of evidence and presenting the overall weight so as to inform the characterisation of risks and its management to acceptable levels of residual risk. The triggers for initiating a risk assessment are likely to include: • • • •

situations where significant consequences are suspected following exposure; where evidence on the adverse consequences of exposure is emerging; the presence of contradictory or incomplete evidence on hazards; situations where the mechanisms and priorities for risk management are unclear; or • where there are opportunities to act in a precautionary fashion to prevent future exposures.

Even for complex hazards that give rise to a wide range of endpoints within occupational, consumer, public health and environmental settings, an analysis of the sensitivities of the risk estimates to specific aspects of exposure (e.g., release rates, distance to exposure points, exposure factors, critical receptor groups) can provide the basis for preventative interventions that dramatically reduce risk. As a prerequisite to a discussion of nanomaterials, it is useful to set out some of the key themes for regulatory risk assessment, illustrated here by reference to industrial chemicals and occupational exposure. This sets the scene for a discussion below on scoping out needs for risk assessment for nanomaterials. The majority of regulatory risk assessment guidelines have ‘triggers’ that determine whether a risk assessment is required by a manufacturer or operator and the quantity and type of information required within it. One important initial trigger for risk assessment is the mass of chemical manufactured or imported per year (i.e. 12 calendar months), as shown in Table 10.1. The initial risk assessment trigger in the majority of countries shown in Table 10.1 is 1 tonne/year (except New Zealand where the act of importing or manufacturing the chemical is the trigger). The amount of information required at that point differs. Indeed, within the European Union, substances manufactured or imported at levels exceeding 1 tonne/year must undergo registration with further information required at sequential triggers. The chemical categories excluded from the risk assessment frameworks for industrial chemicals include polymers, radioactive materials, medicinal products and food stuffs, which are considered under other legislation. Non-isolated intermediates, by-products and chemicals produced incidentally during production are not considered as separate compounds but will be included in the assessment of the initial compound. Naturally occurring biological

394

Environmental and Human Health Impacts of Nanotechnology

Table 10.1 Trigger points for chemical risk assessment from example countries, amounts have been converted to metric tonnes for ease of comparison.

Initial trigger for (limited) regulation Futher triggers for information

Australia

European Community

Canada

United States of America

Japan

New Zealand

1 tonne/ year

1 tonne/ year

1 tonne/ year

10 tonnes/ year

1 tonne/ year

–

10 tonnes/ year

–

100 tonnes/ year

–

1000 tonnes/ year

11 tonnes/ year (4 year IUR) 136 tonnes/ year (4 year IUR) –

10 tonnes/ year –

–

10 tonnes/ year 100 tonnes/ year –

Act of manufacture or importing –

–

–

IUR = Inventory Update Rule.

chemicals (e.g. hormones, natural antioxidants) are also generally excluded from risk assessment frameworks and are considered separately by regulators. International risk assessment frameworks for chemicals consider both the physico-chemical characteristics of the chemical as well as the toxicological and environmental effects. Although the requirements differ slightly between countries, all expect hazard identification and assessment to take account of: • physico-chemical properties – detailing melting/boiling point, relative density, vapour pressure, water solubility, flammability, partition coefficient (n-octanol/ water), state (e.g. solid, liquid); • toxicological information – evaluation of skin irritation/corrosion, eye irritation, skin sensitisation, mutagenicity (bacterial and mammalian cell studies), acute toxicity (route dependant on physical state of chemical), short term repeated dose toxicity, reproductive toxicity, developmental toxicity, toxicokinetics; • ecotoxicological information – short term toxicity testing (Daphnia and fish), growth inhibition study on algae, long term toxicity testing (Daphnia and fish), effects on terrestrial organisms and micro-organisms; and • fate and behaviour – degradation (biodegradability), bioavailability (fish), hydrolysis (as function of pH) and adsorption/desorption screening. Other relevant physico-chemical, toxicological and ecotoxicological information may also be requested. In general, the technical guidance and experimental methods followed are those typically validated and/or adopted by the OECD, although other tests are available and may be used. The international chemicals risk assessment process consists of a number of broadly similar steps that follow the initial weight trigger (1 tonne/year; Figure 10.2).

Risk Assessment of Manufactured Nanomaterials

STOP

Substance

General exclusions: non-isolated intermediates, under customs supervision, radioactive, or polymer?

Y

395

> 1 tonne/year?

N

N

STOP

Y

> 10 tonnes/year?

Y

N

Toxicological information

Ecotoxicological information

Acute toxicity • Oral/dermal/inhalation route • Short-term repeated dose toxicity • Sub-chronic toxicity study • Reproductive toxicity screening • Developmental toxicity study • Toxicokinetics Skin irritation/corrosion Eye irritation Mutagenicity

Aquatic toxicity • Short-term toxicity testing (Daphnia) • Long-term toxicity testing (Daphnia) • Long-term toxicity testing (fish) Algal growth inhibition Biodegradation/degradation

Physiochemical properties

Substance state Melting/freezing point Boiling point Relative density Vapour Pressure Surface tension Water solubility Flash-point Flammability Explosive properties Self-ignition temperature Oxidising properties Granulometry Partition coefficient n-octanol/water Viscosity Chemical name Molecular structure

All other available information on chemical properties, toxicological information, ecotoxicological information

All available data must be presented

N > 100 tonnes/year?

> 10 tonnes/year?

Y

N

> 100 tonnes/year? Y

N

Low Quantity Risk Assessment Report

Y Long term toxicity testing 2-generation reproductive study Carcinogenicity study In vivo testing

Adsorption/desorption screening test Acute fish/daphnia/algae toxicity test Bio-concentration Effects on soil micro-organisms Long term toxicity tests on soil invertebrates/plants

Full Risk Assessment Report

Figure 10.2 Schematic for the generalised risk assessment of a chemical substance being manufactured or imported.

Following the initial trigger, further information is required for chemicals manufactured or imported at quantities above 10 tonnes/year (Table 10.1). In contrast to the above, where the emphasis is on assessing the risks of a chemical prior to marketing and use (i.e. ‘data before market’), retrospective risk assessment may be undertaken after manufacture and is triggered for a number of reasons and under a number of regulatory regimes. Biological impacts observed in the field may be one such trigger; for example in the case of endocrine disrupting chemicals where a quantitative environmental risk assessment was undertaken after impacts were observed in fish (Jobling et al., 2006) and marine snails (Vasconcelos et al., 2006). Retrospective risk assessments are supported by the

396

Environmental and Human Health Impacts of Nanotechnology

same conceptual approach outlined above. Frequently, through a weight of evidence approach, they seek to understand the risks posed by sources of specific chemicals released to the environment in the context of biological impacts observed in the field (at individual, community or ecosystem levels). There are some manufactured nanomaterials (including silver, titanium dioxide and iron oxide) that have been shown to exceed the 1 tonne/year registration trigger under REACH (Schmid and Riediker, 2008). However, the community most likely to be exposed to nanomaterials will be those individuals specifically involved in their manufacture and handling, that is the occupational production and use of nanomaterials. The control of occupational health risks from harmful substances in the workplace is, arguably, the most developed system for the control of chemical exposure. It has been developed as a result of the long history of the industrial use of chemicals/materials and the resulting incidence of occupational diseases and illnesses, for example silicosis from the inhalation of crystalline quartz (Altree-Williams and Clapp, 2002; Nij et al., 2003; Nij and Heederik, 2005) and lead poisoning from the inhalation of the dust and fumes from lead and lead-containing compounds (Grimsley and Adams-Mount, 1994; Pierre et al., 2002; Sen et al., 2002). Nowadays, it is mandatory to carry out a risk assessment before allowing any worker to be exposed to any substance in the workplace. In the European Union, this takes place through the Chemical Agents Directive (Chemical Agents Directive, 98/24/EC), which is implemented within the United Kingdom through the Control of Substances Hazardous to Health Regulations (COSHH, 1988 and last consolidated in 2002) enforced by the Health and Safety Executive (HSE). In the United States, the Occupational Safety and Health act (1970, last amended 2004) regulates the occupational use of chemicals, for which there are two coordinating bodies: the Occupational Safety and Health Administration (OSHA), which develops and regulates workplace health and safety regulations, and the National Institute for Occupational Safety and Health (NIOSH), which recommends health and safety standards and provides information on hazards and prevention (Thorne, 2001). The routes of exposure for individuals in the occupational environment are normally inhalation, dermal contact or ingestion. Dermal and inhalation exposure monitoring, as well as biomarker monitoring, can be used to characterise the exposure of specific workers, for example farm workers exposed to pesticides (FAO WHO, 1986; US EPA, 1988). Although it is unlikely that exposure would be limited to one chemical species, the toxicity of individual substances must be considered initially (EEC, 1988; EC, 1999). There is also considerable experience of risk assessment and occupational health monitoring in industries where dusts or particulate matter are a concern. These include milling and baking industries (e.g. fine flour dust), welding and ship building (metal vapours and particulates), as well as industries that produce building materials (e.g. cement dust and sawdust). The monitoring often includes annual respiratory health checks that are useful to ensure that risk management is effective. Some of these industries may generate dusts that contain ultrafine (nanoparticle) fractions, and risk assessors therefore have incidental historic experience with the nanoscale. All occupational risk assessments require the employer to assess the risks associated with a particular work activity or procedure. The steps include gathering

Risk Assessment of Manufactured Nanomaterials

397

information on: the toxicity of the substances (e.g. from material safety data sheets; MSDS); the likelihood of exposure of the workers performing the procedure and of other individuals in the vicinity; and the likely consequence of accidental exposure for the individuals (e.g. minor injury, hospitalisation, severe/permanent injury). These factors are used to estimate the overall risk, and then consideration is given to how exposure can be prevented or controlled so as to avoid/minimise the risks. For airborne exposures, the risk assessment must also include compliance with exposure limits for individual chemicals. Occupational exposure limits (OELs) are airborne standards designed to protect workers from the acute or chronic effects of inhaling hazardous agents. Compliance with OELs is mandatory and OELs are normally defined as an average over a reference period (e.g. eight hours; also referred to as the time weighted average; TWA). OELs have been used since the 1930s for specific substances (e.g. cotton dust) (Topping, 2001). Threshold limit values (TLVs) are the maximum concentration that a worker can be exposed to daily in a working lifetime without experiencing adverse health effects. These are advisory limits set by the American Conference of Governmental Industrial Hygienists (ACGIH) and are airborne standard guidelines for occupational risk assessment (http://www.acgih.org). In the case of particulate materials, OELs have not always been scientifically based. Historically, many particles were regarded as ‘nuisance’ or ‘low toxicity’ dusts, which meant that little attention was given to scientifically defining the precise OEL. Few of these dusts produced any systemic toxicity and the control of exposure was difficult (e.g. in construction, mines and welding sectors). As a consequence, a generic approach to standard setting was taken for many particulates, resulting in a generic inhalable OEL of 10 mg/m3 and a respirable OEL of 4 mg/m3 for many substances (Table 10.2). These were not suitable for particles with a

Table 10.2 Particles with generic occupational exposure limits (OELs) of 10 mg/m3 (inhalable) and 4 mg/m3 (respirable). Particles with generic inhalable and respirable OEL Aluminium metal Aluminium oxides Calcium carbonate Calciypum silicate Cellulose Graphite Gypsum Limestone Magnesite Magnesium oxide

Marble Pentaerythritol Plaster of Paris Platinum metal Portland cement Rouge Silicon Silicon carbide Starch Titanium dioxide Zinc distearate Zinc oxide fume

398

Environmental and Human Health Impacts of Nanotechnology

known inhalation or systemic toxicity (e.g. asbestos and lead, respectively) for which specific OELs were determined. Currently, OELs are set using all available data (e.g. human, experimental, in vivo and in vitro, physico-chemical, mechanistic understanding of pathogenicity, inter-species differences and cellular responses). However, over time, epidemiological research (with improved health surveillance) has shown links between exposure to ‘low toxicity’ dusts and long term illness. For example, crystalline silica is known to cause silicosis but only recently have its links to increased lung cancer have been recognised (Nij and Heederik, 2005), suggesting that the ‘low toxicity’ determination may be over optimistic (IEH, 1996; Fairhurst, 2003) and that a chemical’s species-specific long term effects must be considered when setting OELs. Indeed, the US ACGIH has recently defined ‘low toxicity’ particles as Particles Not Otherwise Specified (PNOS; which have no TLV, poor water solubility and low toxicity). As a result, the ACGIH currently recommends that PNOS have a TLV (TWA 8 hours) of 10 mg/m3 (inhalable) and 3 mg/m3 (respirable), whereas other countries (e.g. MAK Commission in Germany) have lower respirable limits based on extrapolation from large human occupational groups. These limits reinforce the notion that a simple generic dust standard, based on a premise that the main effect is that of nuisance, is no longer defensible in view of in vitro and in vivo investigations demonstrating the importance of particle size, especially surface area, in determining many factors in lung pathogenicity (Oberdörster et al., 1992; Penn et al., 2005). The establishment of airborne standards for European Union workplaces currently comes under the Chemical Agents Directive (specifically COSHH, which implements the Chemical Agents Directive within the United Kingdom) and REACH (WATCH, 2006). As part of COSHH, legal bases have been given to the occupational exposure standard (OES) and the maximum exposure limit (MEL) (Figure 10.3). The OEL ensures a minimum level of safety, which can be exceeded as long as steps are taken to reduce exposure as far as reasonably practical; whereas MEL is used to maintain safety levels for workers (the MEL must not be exceeded) (Topping, 2001). REACH (discussed in Appendix 10.A) requires that the manufacturer or importer of substances must determine the safe operating conditions and appropriate risk management for the substance, whereas in COSHH it is the employer which must assess the risk of a substance and cover all work activities at that site (e.g. production, application and disposal). REACH (a directly-acting European Union Regulation) applies, without prejudice, to workplace health and safety legislation. The occupational exposure of workers to ultrafine particles is a well studied area (IEH, 1996) and ultrafine particles are considered an aerosol particle in the nanoscale range (e.g. diesel exhaust particulates). However, the occupational risk assessment of manufactured nanomaterials is less well characterised (Balbus et al., 2007a; Boccuni et al., 2008). The UK Health and Safety Executive (HSE) considers there to be three main sources of industrial activities likely to cause exposure to nanoparticles: nanotechnology research and development (in universities, research centres and companies), existing ultrafine manufacturing processes (carbon black, titanium dioxide, alumina manufacturing) and powder handling processes (e.g.

Risk Assessment of Manufactured Nanomaterials

399

Define exposure for negligible risk

Is industry able to comply with exposure level? Y Set OES

N Define exposure level for maximum tolerable risk Is industry able to comply with exposure level? Y Set MEL

Figure 10.3

N Ban

Decision procedure for setting OEL/MEL in the workplace.

manufacture of dyes, pigments and pharmaceuticals; HSE, 2004). The added complications of particle size, surface modification, particle morphology and the possibility of translocation within the body has concerned scientists, engineers, risk assessors and regulators alike (RS/RAEng, 2004), with many calling for the development of risk assessment strategies for novel particles and, in particular, nanomaterials where surface area and surface properties may be important factors (Aitken et al., 2004). The scheme set out in Table 10.3 is an amalgam of a number of proposed strategies. If the above testing strategy is used in risk assessment application for particles in the occupational setting, it can also be used in the development of risk assessment methodology for airborne particles in the environment and for nanoparticulates in consumer products to protect public health. Requirements for the latter two scenarios include likely human exposure (measured or modelled) and the choice of uncertainty factor (also known as a ‘safety’ or ‘assessment’ factor) if an airborne standard is to be set as part of any risk management strategy.

10.3

Nanomaterials – Issues for Risk Assessment

Current regulations in the European Union consider manufactured nanomaterials as chemicals and, therefore, in the European Union, they will be registered, evaluated and authorised under the REACH legislation that came into force in June 2007 when the nanomaterial production reaches manufacturer-specific weight

400

Environmental and Human Health Impacts of Nanotechnology

Table 10.3

Possible testing strategies for new particles.

1 Preliminary assessment

• physical, chemical data • particle size characterisation

2 In vitro studies

• macrophages and epithelial cells • test for cell toxicity (cell membrane, impaired electron transport) • test for migratory ability, phagocytosis and release of inflammatory mediators

3 Short-term In vitro studies

• comparison to ‘benchmark’ particles • inhalation studies of 1–5 days • analysis of bronchioalveolar lavage (BAL) fluid for cells and inflammatory markers • histopathology

4 Medium-term studies • if indicated from 3 above • up to 90 days with similar parameters • includes 2–3 month recovery period 5 Long-term studies

• if indicated from 4 above • all above end-points plus cancer

limits or other registration triggers (NIA, 2007; discussed in more detail in Appendix 10.A). There are no specific regulatory requirements for nanomaterials under REACH, although substances need to be risk assessed in the form in which they are placed on the market. This will include nanomaterials defined as existing substances (present on the EU EINECS chemicals database), with registration deadlines depending on the volume of production of the substance. Concern has been raised that nanomaterials made of the same material as an existing bulk chemical may escape appropriate triggers for risk assessment. This could be resolved by adding an extra trigger to the hazard evaluation that would require evidence that the nanomaterial has the same chemical or biological properties as the bulk compound (Crane and Handy, 2007; Crane et al., 2008). If the materials behave in the same way, this could be treated as an existing material (i.e. already considered under REACH) from the viewpoint of hazard. Alternatively, if the nanomaterial showed some differences from the existing bulk substance, this would trigger a full risk assessment as if it was a previously unregistered chemical and subject to REACH registration as described above. On 17 June 2008, the European Commission issued a communication on the regulatory aspects of nanomaterials stating, ‘Nanotechnologies are enabling technologies, with high potential benefits for consumers, workers, patients, and the environment, as well as the creation of jobs. On the other hand, nanotechnologies and nanomaterials may expose humans and the environment to new risks, possibly involving quite different mechanisms of interference with the physiology of human and environmental species’ (ECC, 2008). Overall, it can be concluded that current legislation covers, to a large extent, the relative risks of nanomaterials and that the risks can be dealt with under the current regulatory framework. However, current legislation may have to be modified in the light of new information becoming avail-

Risk Assessment of Manufactured Nanomaterials

401

able, for example, as regards thresholds used in some legislation (as reported by Royal Commission as Environmental Pollution, http://www.rcep.org.uk/reports/27novel%20materials/27-novelmaterials.htm). The implementation of legislation and the use of regulatory instruments created by legislation remains a particular challenge. Documents that support implementation, particularly in relation to risk assessment, adopted within the context of current legislation will have to be reviewed in order to ensure that they effectively address the risks associated with nanomaterials and make best use of the information becoming available. Similarly, authorities and agencies will have to pay special attention to risks in relation to nanomaterials where protection and marketing are subject to pre-market control. To properly develop, modify or, in particular, implement legislation, the scientific knowledge base needs to be improved (ECC, 2008). In principle, the REACH risk assessment process for chemicals has been judged by the EU Commission SCENIHR (2005, 2006, 2007) as being appropriate to nanomaterials, although it may need to be amended in the future as knowledge gaps are identified and resolved (ECC, 2008; Crane et al., 2008; Section 10.6). In principle, the major steps in risk assessment appear fit for purpose, although the technical details in each step require some consideration (see below). The most over-arching concern is that chemical risk assessment assumes that the chemical toxicology ultimately drives the adverse biological effects. This might hold true for some nanomaterials (e.g. nanoparticulate TiO2 as an oxidising chemical causes oxidative stress in organisms (Handy et al., 2008c). However, it is also possible that physical properties alone (e.g. shape, size, aspect ratio) are a key driver of biological effects. Indeed, experience with materials such as asbestos has shown that it is the aspect ratio (ratio of length to width) and biopersistence that are important mediators of pathogenicity rather than the inherent chemical composition of the material. This can be resolved by ensuring that additional physical parameters are added to the physico-chemical properties measured in the hazard identification step. These additions for nanomaterials, and the relevance of the existing list of physico-chemical measurements for new substances, are discussed at length elsewhere (Handy et al., 2008a; Crane et al., 2008). The toxicity of different sizes and shapes of the same nanomaterial should be measured in the hazard identification step so that risks associated with chemical reactivity can be separated from other physical effects. If these risks change with size/shape, then it may be necessary to risk assess individual size ranges of the same nanomaterial. Pragmatically, this is something best avoided if possible, but at the very least it may be necessary to consider each particle size range as a separate strand of evidence in the risk analysis. Hazard identification requires collection of existing data on known toxic effects (e.g. reviews by Handy et al., 2008a, 2008b). New experimental evidence is rapidly emerging. One fundamental concern about the hazard identification step is that novel materials may impart novel biological effects that have not yet been determined. This is a source of uncertainty, but the review process expected of all risk assessments can reduce this as new data emerges. In principle, this is no different to any other substance, as any new substance has the theoretical potential to show new toxic effects. The consensus view is that the standard endpoints used in hazard

402

Environmental and Human Health Impacts of Nanotechnology

assessment should continue to be used (Crane and Handy, 2007; Crane et al., 2008), but additional measurements may need to be adopted to account for novel properties as they emerge. The exposure assessment is perhaps the most problematic, mainly because of a lack of data on measured levels of nanomaterials in the environment, although this is less of a problem in the workplace where the source of the material is easily identified. Environmental fate and behaviour studies are needed for nanomaterials in the environment so that exposure point concentrations can be estimated. There is a massive background of natural nanomaterials already present in the environment and manufactured nanomaterials would account for a trivial proportion of the total (Handy et al., 2008b). There are also technical issues to overcome, such as measurement methods for nanomaterials in complex environmental matrices, but this is not fundamentally different to any other new substance that would similarly require a measurement technique to be developed and approved. The risk analysis step should also consider the physical form of the product and what it is used for (Hansen et al., 2008). Much of the current hazard data on nanomaterials relates to nanoparticles in a free rather than fixed (or embedded) form, and it is important to provide this context at the outset (Ozin and Arsenault, 2006). There is a concern that the risk will change with the product life cycle. For example, a product coated with carbon nanotubes in a resin matrix will presumably eventually degrade to release individual particles. The context of the risk would depend on how the product wears and where it is eventually disposed; a carbon nanotube coated tennis racket would presumably be disposed of in landfill, but there are uncertainties about the long term erosion of nanomaterials in landfill. This risk might be offset by the increased strength and durability that might be imparted by making the product from nanomaterials leading to a longer product life. Similar arguments apply to the deterioration of nanoproducts in any occupational setting and the risk of long term, low level exposure may then be an issue. Clearly, the complexity of behaviour and toxic effects of nanoparticles generate uncertainties which have significant implications for the risk analysis. Nonetheless, the uncertainties are regarded as manageable through collecting data from new experiments to reduce the level of uncertainty of the hazard and exposure issues. A further need is to consider the actual fate of nanomaterials in environmental matrices. Many nanomaterials are likely to form aggregates in a wide range of natural waters and the question arises as to whether materials should be dispersed in laboratory tests or allowed to aggregate (Crane and Handy, 2007; Crane et al., 2008). It might be technically feasible to develop a lowest observed effect concentration (LOEC) or no observed adverse effect level (NOAEL) using one or more standard test methods, but these may have little relevance for actual exposures in complex natural systems. This is not a new problem and the inevitable difference between standardised laboratory tests and the real environment is just another source of uncertainty. However, there is no reason to believe this aspect is any better or worse than that of other new chemicals. High uncertainty in hazard assessment might mean the requirement for higher assessment factors when calculating, for example, Predicted No Effect Concentrations (PNEC) from LOEC data. In addition, the tendency of nanomaterials to aggregate as mass concentration increases

Risk Assessment of Manufactured Nanomaterials

403

raises the concern that the use of concentration as the dose metric may be inappropriate (Crane et al., 2008; Wardak et al., 2007). This might be resolved by reporting hazard as a function of particle number or surface area. Nonetheless, the uncertainties are cumulative in the risk analysis and uncertainty in predicted environmental concentrations would compound uncertainty about predicted effects concentrations. These issues require further research to understand, quantify and model. The diversity of nanomaterials (discussed in Chapter 9) includes different particle sizes, morphologies, surface chemistries, surface charge heterogeneities, capping formulations, surface ligands (functionalities) and impurities for any one given nanomaterial (Gogotsi, 2006; Ozin and Arsenault, 2006; Roduner, 2006). There are also nanocomposites made of layers of different chemicals (e.g. quantum dots) and nanomaterials incorporated into a matrix with other chemicals in product manufacturing processes (Hansen et al., 2008). The issue of nanomaterial diversity, problem formulation and prioritisation in risk assessment of nanomaterials has been considered by SCENIHR, where the description of the material (including the size of particle, exposure, homogeneity, chemical composition) are determinants of the extent to which risk assessment is required and the component methods that should be used (Figure 10.4). The diversity of functionalities (chemical ligands on the surface of the nanomaterial), geometries, crystal structures and impurities of individual nanomaterials, and the diversity of nanomaterials as a ‘group’ of substances, must be reconciled with the challenges of understanding and quantifying the complexity of behaviour for any one material and its variants. Ideally, detailed and validated models for each type of nanomaterial are needed to minimise uncertainty. It is clear that this will not be achievable for every nanomaterial or its variants, nor should such detailed studies be advocated for every nanomaterial. Instead, attention should focus on how nanomaterials are prioritised for quantitative risk assessment, and in particular on the essential phase of problem formulation within risk assessment outlined in the section above (Owen and Handy 2007).

10.4 Assembling Evidence for Safety and Intervention It is clear from the above discussion that risk assessors need to assemble evidence of varying quality for nanomaterials risk assessments. One context in which risk assessors collate evidence is that of the ‘safety case’. A case is made by an operator to a regulator for the safety of an activity, such as the operation of a large integrated refinery, chlorine production facility or radioactive waste repository. Safety case legislation is widely used across the transport, aerospace and chemical process sectors, and the risk assessment that support safety cases must draw together evidence for operational safety across a plethora of fields, integrating qualitative and quantitative data with the objective of presenting an overall case for safety based on the full set of evidence. These situations are analogous to that currently faced by the manufacturers of nanomaterials. Partial, contradictory and equivalent evidence must be collated, assessed and weighed in its entirety, so as to draw conclusions on the significance of risks posed by the release of nanomaterials in

404

Environmental and Human Health Impacts of Nanotechnology

Human or environmental exposure likely?

Nanomaterial

N

Reassess if use, manufacture or disposal changes

N

Full risk assessment required

Y Y Particles less than 100nm

Toxicity of substance known?

N Y Existing data may be sufficient

Do particles need to be assessed separately

Y N

Homogeneous particles? Y Do particles rapidly agglomerate or coalesce?

Y

Existing exposure data may be sufficient

N Are other chemicals absorbed into the particle?

Y

High priority for study

N

Existing exposure data may be sufficient

Significantly increased reactivity?

Y

Figure 10.4 Decision flow chart for the risk assessment of nanomaterials and nanoparticles.

occupational, consumer, public health and environmental settings (Linkov et al., 2007). Criteria already exist for evaluating the plausibility of associations in disease causation. For instance, Bradford Hill’s criteria (Hill, 1965) attempted to separate causal from non-causal explanations of observed associations by reference to a number of criteria (defined by Hofler, 2005) including: • strength of association (a strong association is likely to have a causal component) • consistency (reproducibility)

Risk Assessment of Manufactured Nanomaterials

• • • • • • •

405

specificity temporality (effect succeeds action or factor) biological gradient (dose response) plausibility (biological explanations) coherence (agrees with current knowledge) randomised experiments (good study design) analogy (effect has already been shown)

The use of Hill’s criteria to determine whether observations of effect are causally linked to specific experimental conditions is supported by good study design. For example, reproducibility of results is required to ensure that artefacts are not considered indicative of a true result; this is also true for the analysis of data collected which would be covered by the criteria of biological gradient, plausibility, temporality, specificity and analogy. It can be argued that a similar version of Bradford Hill’s criteria should be used to assess all scientific studies for use in risk assessments. Klimisch and colleagues (1997) developed a systematic approach for the evaluation of the quality of data, establishing the following scoring system for reliability: 1 = reliable without restrictions. Where the data were generated according to internationally accepted (or validated) testing guidelines (e.g. OECD) and preferably performed according to good laboratory practice (GLP) or where the test parameters are closely related to a guideline method; 2 = reliable with restrictions. Where the data were mostly not performed according to GLP and where the test parameters do not totally comply with the testing guideline, but are considered sufficient to accept the data; 3 = not reliable. Where there were interferences between the measuring system and the test substance, or in which the test systems were used which are not relevant in relation to the exposure, or were generated using an unacceptable method; 4 = not assignable. The experimental details provided were not sufficient and were only listed in short abstracts or secondary literature. These evaluation criteria have been applied to many risk assessment methods previously and have proved to be an acceptable method of determining the quality of the data to be assessed. As long as the initial requirements for information (and the quality of that information) are established in defined technical guidance documents, and adopted by bodies such as OECD, then the application of the Klimisch score to collected data is acceptable. However, the information requirements and the analysis of the collected data must be robust and transparent to ensure that the Klimisch score is applied correctly and robustly. The hierarchy of relevance for toxicological data (stated as individual study > human > animal > in vitro) is generally accepted. The Klimsch score may also be taken into consideration. In doing so, experts need experience in evaluating toxicological endpoints as well as materials characterisation to ensure that the knowledge presented is reliable from both toxicology and materials view points. Individual expertise is also required to determine the relevance and association of cause and effect. For nanomaterials, there exists the possibility of benchmarking certain lines of evidence by reference to well characterised materials (as analogues)

406

Environmental and Human Health Impacts of Nanotechnology

to indicate which areas of evidence should be considered further. This is already possible for some chemicals to a certain extent with QSAR. However, its suitability for use with nanomaterials remains uncertain. If the size or shape of a nanomaterial determines its toxicity, then the relationship may be clear and the outcome only affected if the environment transforms the nanomaterial (either by physical or biological methods). If there are other factors that contribute to the inherent toxicity of the material, for example the surface chemistry, then the need for specific information becomes clear (including information on the aerodynamic diameter, route into body and potential for agglomeration). Recommendations have been made on the base set of characterisation data required for a good toxicological experiment (Handy et al., 2008a, 2008b; Crane and Handy, 2007; Crane et al., 2008). However, not all the published studies on biological effects provide this information at present. An agreement has not been reached on standard methodology to measure exposure in the environment and, until such agreement is reached, the quality of each measurement should be considered on the basis of its scientific merit (e.g. use of calibrations, standard addition tests, spike recovery tests, matrix effects and other controls to prove the methodology). Overall, there is sufficient peer reviewed laboratory data to show that, in principle, nanomaterials can have toxic effects (e.g. Handy et al., 2008a). However, the lines of evidence for environmental exposure are few, and currently this is derived by modelling information on product usage and manufacture for environmental risk assessments (Boxall et al., 2007). The evidence base for workplace exposure is more robust, but there remains a requirement to measure nanomaterials in workplace air to quantify exposure concentrations, and correct this for the benefit of any protective clothing or other controls (e.g. use of dust masks or ventilation). Risk analysts consider both positive and negative evidence to balance the weight of evidence. At this early stage in the science it is imperative that scientific reports of ‘no effect’ or even ‘beneficial effects’ are published, so that this information can be balanced against some of the more high profile reports on adverse effects. Researchers should also consider, and look for, positive effects of nanomaterial chemistry. For example, aggregation may limit the long range transport of nanomaterials in the environment in the short term. Some nanomaterials may chelate or bind other organic chemicals to reduce their toxicity in mixtures. These ideas need to be explored, and the data collected made available for the risk analysis.

10.5

International Case Studies

International bodies, including OECD, ECHA and EPA, recognise that engineered nanomaterials are currently in use but that more information is required to conduct a complete risk assessment of these materials (as discussed in reports including RS/RAEng, 2004 and SCENIHR, 2005). Whilst fundamental research on the toxicity and actions of manufactured nanomaterials is ongoing, risk assessment case studies of existing nanomaterials are also being used to identify knowledge gaps and prioritise further research; the OECD Working Party on Manufactured Nanomaterials (accessed at http://www.oecd.org/department/0,3355,en_2649_ 37015404_1_1_1_1_1,00.html) has identified the need to consider the exposure and

Risk Assessment of Manufactured Nanomaterials

407

dose-response relationships for a robust risk assessment of nanomaterials. The representative nanomaterials chosen for case studies include single- and multiwalled carbon nanotubes, silver nanoparticles, cerium oxide nanoparticles, titanium dioxide nanoparticles, iron nanoparticles and carbon black. This work is ongoing; however, industrial bodies such as the DuPont/Environmental Defense Nano Risk Framework have considered risk assessment of individual nanomaterials using a life cycle approach (accessed at http://nanoriskframework.org/page. cfm?tagID=1326), which suggests that a specific base data set is collected with the need for further information identified by defined criteria (Table 10.4). The ‘triggers’ that require further information to be gathered include high exposure potential, high production volume, magnitude of environmental release, high potential for chronic human/environmental exposure by repeated or continuous release, uncertain or high inherent hazard potential and an incomplete base set of either hazard or exposure data. To date, the Nano Risk Framework has been applied to determine the risks related to specific forms of titanium dioxide nanoparticles and zero-valent iron nanoparticles (nZVI) by defining the substance and application, the lifecycles (properties, hazards and exposure), evaluating the risks, assessing options, recording the decisions and actions, and continuously reviewing and adapting with new information. Titanium dioxide (TiO2), found in paints, plastics and cosmetics is produced in large amounts (e.g. over one million tonnes per annum) with a median particle size between 250 and 350 nm (Gogotsi, 2006). The Nano Risk Framework has assessed TiO2 used as inorganic light stabilisers for polymers (particle size range 130–140 nm, with approximately 20% of particles with particles sizes less than 100 nm), which is found in a wide range of products including sporting goods, outdoor furniture, fabrics and toys. Toxicological studies suggested that there was no significant variation between the effects of light stabilising TiO2 and pigmentation TiO2, and an acceptable exposure limit (AEL) was set by DuPont (manufacturer) as 2 mg m−3. No discernible risks were associated with TiO2 exposure here ( http://www.environmentaldefense.org/documents/6552_TiO2_Summary.pdf ), although other studies have shown hazard or potential hazard from TiO2 nanomaterials. The nZVI assessment (primary particle size of <100 nm), used as a reagent to destroy organic contaminants in ground water, suggested that whilst many manufacturers provided some hazard data this may not have been specific to nZVI and the extent of environmental and human exposure must be taken into consideration as well as the ultimate fate of nZVI. The Nano Risk Framework also considered carbon nanotubes which are not discussed here.

10.6

Data Gaps in Risk Assessment of Nanomaterials

The knowledge gaps in nanomaterial risk assessment are significant at this early stage in the science. The issue of problem formulation is hindered by a lack of data on exposure and biological effects (Handy et al., 2008a), and the current focus is on test organisms used in the regulatory arena, rather than the diverse range of organisms in the natural environment. So while data are being collected on fish species and Daphnia magna, there are few data on many other organisms; especially

Additional Data Biological fate and behaviour Chronic inhalation/ingestion

Chronic dermal irritation/sensitisation Developmental and reproductive Neurotoxicity Extensive genotoxicity studies

Dispersability Bulk density Agglomeration state

Porosity Surface charge Surface reactivity

Focused toxicity studies (susceptibility studies; allergencity and immunotoxicity organ-function bioassays) Endocrine disruption studies

Genetic toxicity tests

Skin penetration

Skin sensitisation/irritation

Single dose instillation study (90 day observation period) 28 day repeated dose oral study (90 day observation period)

Solubility (in water and biologically relevant fluids)

Physical form/shape (at room temperature and pressure) Particle size, size distribution and surface area Particle density

Chemical composition (including surface coating) Molecular and crystal structure

Short term toxicity 28 day inhalation study

Technical and commercial name Common form

(90 day observation period)

Health Hazard Date

Physicochem Data

Biodegradability (organic based) Photodegradability/photo transformation Stability in water Bioaccumulation Additional Data ADME on aquatic organisms Chronic toxicity (aquatic, terrestrial, microorganisms) Additional species Avian toxicity testing Population/ecosystem-level studies Activated sludge respirationinhibition test Microorganism toxicity Persistence potential Transformations via oxidation/ reduction reactions

Persistence potential screen

Acute aquatic toxicity Fish Invertebrates Aquatic plants Terrestrial toxicity Invertebrates Plants Physicochemical properties (including list) Adsorption-desorption coefficients Aggregation or disaggregation in media

Environmental Hazard/Fate Data

Polymerisation Photoactivity

Additional Data Stability decomposition

Environmental releases Post-use management

Distribution/ storage

Reactivity

Corrosivity

Use

Processing

Explosivity Incompatibility

Manufacture

Exposure Data

Flammability

Safety Hazard Data

Table 10.4 Summary of recommended data set suggested for Nano Risk Framework (as discussed in Environmental Defense and DuPont, 2007).

Risk Assessment of Manufactured Nanomaterials

409

algae, plants, bacteria and the higher vertebrates. Most of the existing toxicological studies have used mg L1 concentrations for exposures. Data at low µg L1 concentrations and over long timescales (weeks, months) are lacking. Information on exposure concentration is currently hampered by technical barriers in our ability to measure engineered nanomaterials in the natural environment (Handy et al., 2008a), and it is most likely that a combination of separation and quantification methods will be needed to measure nanomaterials in natural waters, sediments, and soils. These are tractable problems but, with current technology, measurements are laborious and require high levels of expertise, with perhaps only a few samples being analysed each day at best. The development of rapid techniques that might be employed in routine environmental monitoring programmes to protect the environment or public health is a long way off. An alternative approach is to model predicted worse case concentrations on the basis of product usage and information on likely occurence in effluents (Boxall et al., 2007). This is a useful interim approach while measurement technology is being developed. Data for modelling are currently based on product information volunteered by users/manufacturers. The responses to voluntary reporting schemes in the United Kingdom have been limited, and one way to strengthen the data set would be to implement a mandatory reporting scheme (as reported by Royal Commission as Environmental Pollution, http://www.rcep.org.uk/reports/27novel%20materials/27-novelmaterials.htm). There are also levels of complexity in the chemistry that leave knowledge gaps. Hazard data on the abiotic factors that influence aggregation chemistry and, therefore, toxicity are lacking (discussed above). The aim in the short term should be to identify whether toxicity increases or decreases with changes in pH, ionic strength, Ca2+ concentration, presence/absence of humic substances and other conditions, with a selected set of organisms that are used in regulatory tests. This will provide a basis for some risk calculations and, similar to the situations with metals, many decades of research will be needed to add the fine detail and mechanistic explanations of abiotic effects. There is also a significant lack of knowledge on mixtures and concerns exist that some hydrophobic nanomaterials may act as delivery vehicles for other organic chemicals (Baun et al., 2008). There are a number of knowledge gaps that relate to the product life cycle of nanomaterials. The manufacture of nanomaterials will presumably generate waste material and wastewater effluents. Of course, some nanomaterials are expensive to produce and manufacturers will be seeking to minimise losses during production. However, without chemical monitoring methods, we cannot quantify these losses to the environment directly. Data might be collected from the manufacturers on aspects of production efficiency so that losses may be estimated, but many of the companies that produce nanomaterials are small manufacturers that do not have the automated plant engineering that could provide such data. Exposure in the work place must be controlled and, while individual users will have made risk assessments (e.g. general risk assessments of work procedures, COSHH assessments) and local arrangements for occupational health will be in place, there are not enough data to determine if these measures are being effective. In the short term, at least to our knowledge, no clinical conditions have arisen from

410

Environmental and Human Health Impacts of Nanotechnology

exposure to nanomaterials in the work place whilst using approved safety procedures. However, the long term risk and the effectiveness of these measures will only be shown by long term monitoring of the workplace. Waste disposal of nanomaterials is also an area with many knowledge gaps. Knowledge on the suitability of incineration as a disposal method for carbon-based and other non-metallic nanomaterials is lacking. It is unclear whether nanomaterials would be completely destroyed during incineration or whether they will contribute to air pollution. Some aspects are well known, such as the combustion chemistry of soot particles (Elmquist et al., 2004), but it is unclear as to how much of this existing knowledge could be directly transferred to C60 or other carbon-based nanomaterials. Household products containing nanomaterials are likely to be disposed of at landfill. The chemical fate and behaviour of nanomaterials in landfill is unknown and there are concerns that nanoscale materials may penetrate the clay lining materials of the landfill, releasing nanomaterial leachate into the water table. Managing uncertainty in the risk assessment of manufactured nanomaterials is likely to be critical because the relationship between environmentally relevant doses and the potential toxicological responses in human and environmental receptors is not well characterised (Nel et al., 2006; Renn and Roco, 2006; Balbus et al., 2007a). Practitioners can expect to have to assemble and weigh the evidence (discussed further in Mayo and Hollander, 1991; Forbes and Calow, 2002; Hrudey and Leiss, 2003; Weed, 2005; Balbus et al., 2007b) from various research studies and apply precaution (discussed in ILGRA, 2001; Harrison and Holmes, 2006), especially where irresolvable uncertainties in the assessment suggest that significant consequences might occur.

10.7

Summary

The current consensus is that existing risk assessment frameworks and their components are, in principle, appropriate for the risk assessment of nanoparticles and nanomaterials, but that implementation may be associated with significant new issues. These issues relate in part to two features of nanomaterials, their diversity and the complexity of their behaviour in natural systems and the uncertainty this may introduce to risk assessments, notably in the quantification of hazard and quantitative predictions of exposure. Many aspects of diversity and complexity are covered in other chapters of this book. Reducing uncertainty in hazard and exposure assessment will require significant and potentially lengthy research into behaviour, interaction and bioavailability in natural systems for any one nanomaterial, let alone any variants of those materials associated with the addition of active chemical groups onto the surface of the material. This must be considered alongside the ever growing diversity of nanomaterials and their functionalities associated with rapid technological development. Commitment to reduce uncertainty in hazard, exposure and risk assessment must be justified by robust problem formulation, underpinning what risk assessors call ‘justifying the intent’ to undertake a risk assessment. To date, while considerations of the technical aspects of exposure and hazard continue, far less consideration has been given to this critical phase of risk assessment.

Risk Assessment of Manufactured Nanomaterials

411

Appendix 10.A Chemicals Risk Assessment Under REACH and TSCA 10.A.1

REACH

The European Commission (EC) regulates industrial chemicals under the Registration, Evaluation, Authorisation and restriction of CHemical substances regulation (REACH) (EC No 1907/2006), which came into force on 1 June 2007. The REACH regulation applies to chemicals that are manufactured or imported in amounts greater than 1 tonne/year. Substances that are excluded from REACH are covered by more specific legislation (e.g. human medicines, radioactive substances, non-isolated intermediates and waste). Although this regulation is specific to the EC, similar regulation of industrial chemicals occurs world wide (e.g. Morgenstern et al., 2000; Thanawalla, 2002; NICNAS, 2004; ECHA, 2008). REACH provides an over-arching legislation applying to the manufacture, placing on the market and use of substances on their own, in preparations or in articles. REACH is based on the principle that manufacturers, importers and downstream users have to ensure that they manufacture, place on the market or use such substances that do not adversely affect human health or the environment. There are no specific provisions in REACH referring to nanomaterials, but REACH applies to manufactured nanomaterials on their own, in preparations or in articles when they are considered as chemical substances. Under REACH, manufacturers and importers have to submit a registration dossier for substances manufactured at or above 1 tonne/year. If substances are imported or manufactured at levels greater than 10 tonnes/year, the registrant will have to perform a chemicals safety assessment (CSA) and produce a chemical safety report (CSR; as shown in Figure 10.A.1). The CSA will have to cover the risks of the nanomaterials including the properties of the nanomaterials, its classification and labelling of hazardous properties, its risks and risk management measures. The risk management measures have to be communicated to the supply chain. Data generated under REACH serves as input to other regulations, for example worker protection, cosmetics and environmental protection. Regulation of chemical substances under REACH is based on the (precautionary) principle ‘that industry should manufacture, import or use substances or place them on the market in a way that, under reasonably foreseeable conditions, human health and the environment are not adversely affected’ (ECC, 2000; ILGRA, 2001; ECHA, 2007). Therefore, the emphasis is on the manufacturer or importer to collect or generate the data on substances and to assess the risks involved. Within REACH, there are several triggers (pieces of information which force the regulator or manufacturer to supply further information) for specific information requirements (Figure 10.A.1). The CSA, for substances imported or manufactured in amounts over 1 tonne/year, contains information about the physicochemical properties of the chemical and some toxicological information. Further information triggers at 10, 100 and 1000 tonnes/annum require a CSR to be submitted with more detailed information on toxicological, ecotoxicological and carcinogenicity required as well as the physico-chemical properties of the substance.

Y

Y

Figure 10.A.1

Chemical Safety Assessment

N

> 10 tonnes/year?

Stability in organic solvents Dissociated constant Viscosity

Y

> 100 tonne/year?

N

Carcinogenicity study

Y

> 1000 tonnes/year?

Y 2-generation reproductive study

Positive result or > 100 tonnes/year?

Long-term toxicity (sediment organisms) Long-term/reproductive toxicity (birds)

N

Y In vivo mutagencity studies

Positive result or > 100 tonnes/year?

Relevant in vivo study

Y

Corrosive, strong acid/ base, flammable, very toxic, or skin irritant; and >10 tonne/year?

Y

> 1 tonne/year?

Substance

Ecotoxicological information

> 100 tonnes/year?

N

N

> 10 tonnes/year?

Aquatic toxicity Short-term toxicity testing (Daphnia) Long-term toxicity testing (Daphnia) Long-term toxicity testing (fish)

STOP

Y

> 1000 tonnes/year?

N

Bioconcentration (fish) Further studies on adsorption/desorption Y Short-term toxicity to earthworms/plants Effects on soil microorganisms Long-term toxicity on earthworms/plants/soil invertebrates

Y Identification of degradation products Further degradation testing

N

Hydrolysis as a function of pH

Ready biodegradability Degradation in surface water Soil simulation testing Sediment simulation testing

Growth inhibition study (algae) Short-term toxicity testing (fish) Activated sludge respiration inhibition testing

< 10 tonnes/years and insoluble or does not cross biological membranes?

Chemical Safety Report

N

> 100 tonnes/year?

Environmental fate/behaviour Adsorption/desorption

Y

N

Methods of detection and analysis

Other available physiochemical, toxicological and ecotoxicological information

Overview of risk assessment by REACH showing information triggers (taken from Rocks et al., 2008).

Acute toxicity Oral/dermal/inhalation route Short-term repeated dose toxicity Sub-chronic toxicity study Reproductive toxicity screening Developmental toxicity study Toxicokinetics

Y

> 10 tonnes/year?

Mutagenicity In vitro gene mutation In vitro cytogenicity In vitro gene mutation

Eye irritation Assessment of: human/animal data acid or alkaline reaction

Skin irritation/corrosion Assessment of: human/animal data acid or alkaline reaction

Toxicological information

N

Is substance a non-isolated intermediate, under customs supervision, radioactive, or a polymer?

Substance state Melting/freezing point Boiling point Relative density Vapour Pressure Surface tension Water solubility Flash-point Flammability Explosive properties Self-ignition temperature Oxidising properties Granulometry Partition coefficient n-octanol/water

Physiochemical properties

STOP

412 Environmental and Human Health Impacts of Nanotechnology

Risk Assessment of Manufactured Nanomaterials

413

Physico-Chemical Properties There are many published sources of physico-chemical data, including the Merck Index and IUPAC Solubility Data Series, that can be considered and used within risk assessments rather than experimental results. However, the data should be considered carefully and the state of the substance and range of values must be evaluated. Included in the physico-chemical properties of a chemical substance is the determination of the particle size distribution, which would account for nanomaterials and particles (Figure 10.A.2). Toxicological Information The toxicological information required under REACH can be split further into eight information groups. While these have been considered separately due to the ethics of animal work, the experiments must be carefully designed with crossovers between groups. The groups are: 1. Skin Irritation/Corrosion. These tests are not necessary if data show the substance is corrosive/irritating. In vitro tests are required for CSA, whilst in vivo tests are required for unclassified substances manufactured/imported in amounts greater than 10 tonnes/year. 2. Eye Irritation. In vitro tests are not necessary if the substance is considered irritant/corrosive to skin (and is classified as irritating to the eye). For substances produced or imported in amounts greater than 10 tonnes/year an in vivo test is required, unless substance has been determined as irritating. 3. Skin Sensitisation. In vitro tests are required, and if there is not enough information to classify the substance as a skin sensitiser in vivo tests necessary. 4. Mutagencity. In vitro and in vivo tests in somatic and germ cells are used to determine whether the substance is genotoxic in somatic and/or germ cells. 5. Acute Toxicity. Physico-chemical data previously collected are used to determine the route of administration of the substance (indicative of the common route of human exposure). Oral in vivo acute toxicity tests are required unless oral exposure is not possible, in which case an inhalation study is necessary. Further in vivo studies for another (different) exposure route are required for substances imported or manufactured in amounts greater than 10 tonnes/year. 6. Reproductive and Developmental Toxicity. If the substance has already been classed as a genotoxic carcinogen or a germ cell mutagen and the appropriate risk management measures are in place then further testing is not necessary. A two-generation reproductive toxicity study is required for substances manufactured or imported in amounts ≥100 tonnes/year, and for those ≥10 tonnes/year if the reproductive and developmental toxicity screening study is positive or if repeated dose toxicity study indicates potential reproductive toxicity. 7. Repeat Dose Toxicity. These tests are only required if there were indications in the acute toxicity testing or if the chemical is over the 10 tonnes/year weight trigger. The tests determine the NOAEL for the substance over 28 days. If the amount of substance is greater than 100 tonnes/year, or data suggest accumulation, then a sub-chronic or chronic repeated dose study (90 days or 12 months) is required.

Methods: Cascade, Laser, Rotating drum and Continuous drop method

Y

N

Water insoluble

Stop testing

Y

Can mass median aerodynamic diameter demonstrate no inhalation risk?

Y

Is inhalation risk study required?

Are particles soluble?

Y

N

N

Inhalation route for toxicity testing

Stop testing

Investigate particles using microscopy, sedimentation, and laser doppler

Water soluble

Determine relative density and water solubility

Particles <100 µm

Are particles <100 µm?

Light microscope examination or 100 µm sieve

powders

Determination of particle size distribution and morphology under REACH.

Oral route for toxicity testing

N Virtually no particles <100 µm

granulates

Investigate particles using microscopy, sedimentation, electrical sensing, and laser doppler velocimetry

Image analysis

Figure 10.A.2

SEM TEM

Light microscope examination

fibres

substance

414 Environmental and Human Health Impacts of Nanotechnology

Risk Assessment of Manufactured Nanomaterials

415

8. Carcinogenicity Studies. These studies are only required for chemicals in amounts greater than 1000 tonnes/year and carcinogenicity indicators are normally incorporated into other toxicity tests. Ecotoxicological Information The ecotoxicological information required under REACH has previously been considered (Defra, 2000; Crane and Handy, 2007; Crane et al., 2008). Depending on the ecotoxicological data already available and mitigating factors, hazard assessment is performed and refined. The initial ecotoxicological information required includes short term toxicity tests (with Daphnia and fish as target species) and growth inhibition studies on algae. Further, chronic (long term) testing is required to refine the risk assessment process if the substance is classified as PBT (Persistent, Bioaccummulative, Toxic) or vPvB (very Persistent, very Bioaccumulative). Further terrestrial hazard assessments are required schemes (Crane and Handy, 2007). To address the specific properties, hazards and risks associated with nanoscale particles, additional testing or information may be required. To determine specific hazards associated with nanoscale materials, current test guidelines may need to be modified. Until specific test guidelines for nanomaterials exist, testing will have to be carried out according to already existing guidelines. Substances (and therefore also nanomaterials) may also be subject to dossier evaluation or, when there are grounds to consider that they may be of risk, to substance evaluation. For substances of very high concern, an authorisation will be required for their use and their placing on the market. The restrictions procedures enable the restriction of manufacturing, placing on market and/or use of nanomaterials where there is an unacceptable risk. Authorisation and restriction schemes apply regardless of quantities manufactured or placed on the market. 10.A.2 TSCA Whilst, within the European Union, risk assessment is applied only after manufacture, within the United States notification is required before manufacture occurs, which allows for potential concerns to be addressed early (Denison, 2007). The statutory risk assessment controlling the importation and manufacture of new chemical substances in the United States of America is controlled under the Toxic Substances Control Act (TSCA, 1976). TSCA requires the Environmental Protection Agency (EPA) to assess and regulate risks to human health and the environment before a new chemical substance is introduced into the market. Any available data on a new chemical substance (specifically including chemical structure and name) must be submitted as a Pre-Manufacture Notification (PMN) to the EPA. However, there are no specific requirements for the toxicity testing data that must be submitted and the EPA can only require the testing of chemicals if unreasonable risk to health or environment is suspected within a period of 90 days after submission (Lynch et al., 1991; Thanawalla, 2002). EPA classifies chemical substances as either ‘new’ chemicals or ‘existing’ chemicals, which are listed in the

416

Environmental and Human Health Impacts of Nanotechnology

TSCA Chemical Substances Inventory. Some chemicals are excluded under certain conditions: • • • • • •

products of incidental reactions and end-use reactions; naturally occurring materials; mixtures (but not their components); substances manufactured for export only; non-isolated intermediates; and by-products and products formed during manufacture of an article.

There are six different reporting levels for new chemicals; these are Low Volume Exemption (LVW; less than 10 tonnes/year), Research and Development (small quantities only), Low Releases and Low Exposures (LoREX), Test Marketing (TME; manufacture or importation), Polymer (specific criteria only) and PreManufacture Notice (all chemicals not previously excluded). After submission of the PMN to the EPA, the initial determination is to ensure that chemical identity is clear and the substance will undergo a Chemistry Review Phase to determine the chemical name and structure, CAS registry number, molecular weight, impurities and by-products, physical and chemical properties, and molecular structure (Francis and Farland, 1987). The second phase of the PMN review process is the Hazard Evaluation phase where the human and ecological hazards are identified, the environmental fate is established and the structure– activity relationship is considered. The third phase is the Exposure Evaluation which considers human and environmental exposure. The final phase of the PMN review process is the Risk Assessment/Management review. After the four phases are complete, the EPA may issue a PMN exemption or request that a standard review of the chemical is completed (when the chemical presents a significant risk but either the hazard or exposure information is not adequate to characterise the risk), where the EPA can issue a TCSA Section 5(e) order to prohibit or limit activities associated with the substances if: • there is insufficient information to evaluate the human health and environmental effects of the substance; • there is an unreasonable risk of injury to human health or the environment; or • the substance will be produced in substantial quantities that will enter the environment or that there may be significant or substantial human exposure. These trigger the PMN submitter to develop and submit certain toxicity or environmental fate tests before exceeding a specified production volume. This method allows for sales of the chemical to generate enough revenue to pay for the testing. Other Section 5(e) orders include (i) exposure-based Section 5(e) orders, where submitters are required to conduct triggered testing (plus record keeping and ‘risk notification’ in case the test data indicates a risk), and (ii) risk-based TSCA Section 5(e) orders that require exposure controls such as gloves, goggles, respirators, specified disposal technologies or restrictions on releases to water, and hazard communication such as MSDS, labels and work training. The PMN submission information is required to be updated every four years by the Inventory Update Rule (IUR). However, chemicals of low toxicity or low

Risk Assessment of Manufactured Nanomaterials

417

current interest are exempt from the IUR, as are chemicals imported or produced in quantities below 25 000 pounds/year (equivalent to 11 tonnes/year; Romano, 2003).

10.8

References

Aitken, R. J., K. S. Creely and C. L. Tran (2004) Nanoparticles: an occupational hygiene review. RR 274, Health and Safety Executive, UK. Altree-Williams, S. and R. Clapp (2002) Specific toxicity and crystallinity of α-quartz in respirable dust samples, Am. Ind. Hyg. Assoc. J., 63, 348–53. Ashmore, M. H. and C. P. Nathanail (2008) A critical evaluation of the implications for risk based land management of the environmental chemistry of Sulphur Mustard, Environment International, 34, 1192-1203. Balbus, J. M., K. Florini, R. A. Denison and S. A. Walsh (2007a) Protecting workers and the environment: An environmental NGO’s perspective on nanotechnology, J. Nanoparticle Research, 9, 11–22. Balbus, J. M., A. D. Maynard, V. L. Colvin, et al. (2007b) Meeting report: hazard assessment for nanoparticles – report from an interdisciplinary workshop, Environ. Health Persp., 115, 1654–9. Baun, A., S. N. Sorensen, R. F. Rasmussen, et al. (2008) Toxicity and bioaccumulation of xenobiotic organic compounds in the presence of aqueous suspensions of aggregates of nano-C-60, Aquatic Toxicol, 86, 379–87. Boccuni, F., B. Rondinone, C. Petyx and S. Lavicoli (2008) Potential occupational exposure to manufactured nanoparticles in Italy, J. Clean. Product., 16, 949–56. Boxall, A. B. A., K. Tiede and Q. Chaudhry (2007) Engineered nanomaterials in soils and water: How do they behave and could they pose a risk to human health? Nanomedicine, 2 (6), 919–27. COSHH (2002) The Control of Substances Hazardous toHealth Regulations, The Stationery Office, London, accessed at http://www.opsi.gov.uk/si/si2002/20022677.htm. Crane, M. and R. D. Handy (2007) An assessment of regulatory testing strategies and methods for characterizing the ecotoxicological hazards of nanomaterials, Report for Defra, London, UK. Available at: http://randd.defra.gov.uk/Document. aspx?DocumentID=2270/ Crane, M., R. D. Handy, J. Garrod and R. Owen (2008) Ecotoxicity test methods and environmental hazard assessment for engineered nanoparticles. Ecotoxicology, 17, 421–37. Denison, R. A. (2007) Not that innocent: a comparative analysis of Canadian, European Union, and United States policies on industrial chemicals. Environmental Defense, Washington, DC, USA. Defra (2000) Guidelines for Environmental Risk Assessment and Management, Department for Environment, Food and Rural Affairs, London, UK. DETR et al. (2000) Guidelines for Environmental Risk Assessment and Management, Revised Departmental Guidance, Department of the Environment Transport and the Regions, Environment Agency and Institute for Environment and Health, The Stationery Office, London, UK. Duhayon, S., P. Hoet, G. Maele-Fabry and D. Lison (2008) Carcinogenic potential of formaldehyde in occupational settings: A critical assessment and possible impact on occupational exposure levels. Int. Arch. Occ. Env. Hea., 81, 695–710. ECHA, Guidance on Registration, June 2007 accessed http://reach.jrc.it/docs/guidance document/registration_en.htm on 14th October 2007. Elmquist, M., O. Gustafsson and P. Andersson (2004) Quantification of sedimentary black carbon using the chemothermal oxidation method: an evaluation of ex situ pretreatments and standard additions approaches. Limnol. Oceanogr.-Methods 2, 417–27.

418

Environmental and Human Health Impacts of Nanotechnology

Environmental Defense – DuPont Nano Risk Framework, www.nanoriskframework.com accessed on 9th November 2007. European Chemicals Agency (2008) Guidance on Registration accessed at http://guidance. echa.europa.eu/guidance_en.htm. EC (European Communities) (1999) Commission of European Communities Regulation No. 2232/96. EC JCR (European Communities JRC) (2003) Technical guidance document in support of Commission Directive 93/67/EEC on risk assessment for new notified substances and Commission Regulation (EC) 1488/94 on risk assessment for existing substances. European Commission. ECC (European Commission’s Communication) (2000) Communication from the commission on the precautionary principle, accessed at http://europa.eu.int/comm/dgs/health_ consumer/library/pub/pub07_en.pdf. European Commission’s Communication (2008) Communication from the commission to the European parliament, the council and the European economic and social committee on the regulatory aspects of nanomaterials, accessed at http://eur-lex.europa.eu/ LexUriServ/LexUriServ.do?uri=COM:2008:0366:FIN:EN:PDF. EEC (European Economic Community) (1988) Methods for the determination of toxicity. Official Journal of the European Communities, L133. EP (2006) Report on nanosciences and nanotechnologies: an action plan for 2005–2009. European Parliament Report A6-0216/06. Fairhurst, S. (2003) Hazard and risk assessment of industrial chemicals in the occupational context in Europe: some current issues, Food Chem. Toxicol., 41, 1453–62. FAO WHO (1986) Pesticide residues in food. Report of the 1985 Joint Meeting of the FAO Working Party of Experts on Pesticide Residues and the WHO Expert Group on Pesticide Residues, FAO Plant Production and Protection Paper 68, Food and Agriculture Organization of the United Nations World Health Organisation, Rome, Italy. Forbes, V. E. and P. Calow (2002) Applying weight of evidence in retrospective ecological risk assessment when quantitative data are limited. Hum. Ecol. Risk Assess., 8, 1625–39. Francis, E. Z. and W. H. Farland (1987) Application of the preliminary developmental toxicity screen for chemical hazard identification under the toxic substances control act, Teratogen. Carcin. Mut., 7, 107–17. Frater, L., E. Stokes, R. Less and T. Oriola (2006) An overview of the frame work of current regulation affecting the development and marketing of nanomaterials. DTI BRASS report Department of Trade and Industry, Cardiff University, Cardiff, UK, accessed at www.berr. gov.uk/files/file36167.pdf. Gogotsi, Y. (2006) Nanomaterials Handbook, Taylor & Francis, Boca Raton, FL, USA. Grimsley, E. W. and L. Adams-Mount (1994) Occupational lead intoxication: Report of four cases. South. Med. J., 87, 689–91. Handy, R. D., F. V. D. Kammer, J. R. Lead, et al. (2008a) The ecotoxicology and chemistry of manufactured nanoparticles. Ecotoxicology, 17, 287–314. Handy, R. D., R. Owen and E. Valsami-Jones (2008b) The ecotoxicology of nanoparticles and nanomaterials: Current status, knowledge gaps, challenges, and future needs. Ecotoxicology, 17, 315–25. Handy, R. D., T. B. Henry, T. M. Scown, et al. (2008c) Manufactured nanoparticles: Their uptake and effects on fish – A mechanistic analysis. Ecotoxicology, 17, 396–409. Hansen, S. F., E. S. Michelson, A. Kamper, et al. (2008) Categorization framework to aid exposure assessment of nanomaterials in consumer products. Ecotoxicology, 17, 438–47. Harrison P. T. C. and P. Holmes (2006) Assessing risks to human health. In Issues in Environmental Science and Technology, No.22, Chemicals in the Environment: Assessing and Managing Risk (eds Hester R. E. and Harrison, R. M.), Royal Society of Chemistry, Cambridge, UK. Hill, A. B. (1965) The environment and disease: association or causation? Proc. Roy. Soc. Med. – London, 58, 295–300.

Risk Assessment of Manufactured Nanomaterials

419

Hofler, M. (2005) The Bradford Hill considerations on causality: a counterfactual perspective. Emerg. Themes Epidemiol., 2, 11–20. Hrudey, S. E. and W. Leiss (2003) Risk management and precaution: Insights on the cautious use of evidence. Environment. Health Perspect., 111, 1577–81. HSE (2004) Nanoparticles: an occupational hygiene review, Institute of Occupational Medicine for the Health and Safety Executive, Sudbury, UK. IEH (1999) Approaches to predicting toxicity from occupational exposure to dusts, Medical Research Council, Institute for Environment and Health, Leicester, UK. ILGRA (2001) The precautionary principle: policy and application, Interdepartmental Liason Group on Risk Assessment http://www.hse.gov.uk/aboutus/meetings/ilgra/pppa. htm. IPCS (1999) Environmental Health Criteria 210: Principles for the assessment of risks to human health from exposure to chemicals, World Health Organisation, Geneva. Jobling, S., R. Williams, A. Johnson, et al. (2006) Predicted Exposures to Steroid Estrogens in U.K. Rivers Correlate with Widespread Sexual Disruption in Wild Fish Populations. Environ Health Perspect., 114, 32–9. Keeratipibul, S. and S. Lekroengsin, (2008) Risk assessment of Listeria spp. contamination in the production line of ready-to-eat chicken meat products. J. Food Protection, 71, 946–52. Klimisch, H. J., M. Andreae and U. Tillmann (1997) A systematic approach for evaluating the quality of experimental toxicological and ecotoxicological data. Reg Toxicol Pharmacol, 25, 1–5. Linkov, I., K. Satterstrom, J. Steevens, et al. (2007) Multicriteria decision analysis and nanotechnology. J. Nanoparticle Res., 9, 543–54. Lynch, D. G., N. F. Tirado, R. S. Boethling, et al. (1991) Performance of on-line chemical property estimation methods with TSCA premanufacture notice chemicals. Sci. Tot. Environ, 109/110, 643–8. Mayo, D. G. and R. Hollander (eds) (1991) Acceptable Evidence: Science and Values in Risk Management. Oxford University Press. Morgenstern, R. D., J. S. Shih and S. L. Sessions (2000) Comparative risk assessment: an international comparison of methodologies and results, J. Hazard. Mats., 78, 19–39. Nel, A., T. Xia, L. Mälder and N. Li (2006) Toxic potential of materials at the nanolevel. Science, 311, 622–7. NIA (2007) Micromega: REACH for nanomaterials, Synthesis of a NIA masterclass, 9 May, 2007, Nanotechnology Industries Association, London, UK. NICNAS (2004) Handbook for Notifiers, National Industrial Chemicals Notification and Assessment Scheme, Commonwealth of Australia. Accessed at http://www.nicnas.gov.au/ Publications/NICNAS_Handbook.asp, Canberra, Australia. Nij, E. T. and D. Heederik (2005) Risk assessment of silicosis and lung cancer among construction workers exposed to respirable quartz. Scand. J. Work Environ. Health, 31, 49–56. Nij, E. T., S. Hilhorst, T. Spee, et al. (2003) Dust control measures in the construction industry. Annals Occup. Hyg., 47, 211–8. NRC (National Research Council) (1983) Risk assessment in the federal government: managing the process. National Academy Research, Washington DC, USA. Oberdörster, G., J. Ferin, R. Gelein, et al. (1992) Role of the alveolar macrophage in lung injury, studies with ultra fine particles, Environ. Health Persp., 97, 193–7. Owen, R. and M. H. Depledge (2005). Nanotechnology, and the environment: risks and rewards. Marine Poll. Bull, 50, 609–12. Owen, R. and R. Handy (2007) Formulating the problems for environmental risk assessment of nanomaterials , Environ. Sci. Tech., 41, 5582–8. Ozin, G. A. and A. C. Arsenault (2006) Nanochemistry: A chemical approach to nanomaterials, Royal Society of Chemistry, Cambridge, UK. Penn, A., G. Murphy, S. Barker, et al. (2005) Combustion-derived ultrafine particles transport organic toxicants to target respiratory cells., Environ. Health Perspect., 113, 956–63.

420

Environmental and Human Health Impacts of Nanotechnology

Pierre, F., C. Vallayer, F. Baruthio, et al. (2002) Specific relationship between blood lead and air lead in the crystal industry Int. Arch. Occup. Environ. Health, 75, 217–23. Pollard, S. J. T. (2006) Risk management for the environmental practitioner, IEMA Practitioner No. 7, Best practice series, Institute of Environmental Management & Assessment, Lincoln, UK. Pollard, S. J. T., R. Yearsley, N. Reynard, et al. (2002) Current Directions in the Practice of Environmental Risk Assessment in the United Kingdom, Environ. Sci. Technol., 36, 530–8. Renn, O. and M. Roco (2006) Nanotechnology and the need for governance. J Nanoparticle Research, 8, 153–91. Risk assessment and Toxicology Sterering Committee (1999), Risk assessment approaches used by UK government for evaluating human health effects of chemicals (cr2). Leicester, Institute for environment and Health. Rocks, S. A., S. J. T. Pollard, R. A. Dorey, et al. (2008) Comparison of risk assessment approaches for manufacture nanomaterials, report for Defra, available at http://randd. defra.gov.uk/Document.aspx?Document=CB0403_7306_ABS.doc. Roduner, E. (2006) Nanoscopic materials size-dependent phenomena. Royal Society of Chemistry, Cambridge, UK. Romano, A. M. (2003) Recent amendments to EPA’s TSCA inventory update rule. Environ. Qual. Managem, 1, 35–44. Rowbotham, A. L., L. S. Levy and L. K. Shuker (2000) Chromium in the environment: An evaluation of exposure of the UK general population and possible adverse health effects J. Toxicol. Environ. Health – Part B 3, 145–78. RS/RAEng (2004) Nanoscience and nanotechnologies: opportunities and uncertainties, Royal Society and Royal Academy of Engineering, London, UK. SCENIHR (Scientific Committee on Emerging and Newly Identified Health Risks) (2005) Opinion of the appropriateness of existing methodologies to assess the potential risk associated with engineered and adventitious products of nanotechnologies, 002/05, European Commission Health and Consumer Protection Directorate-General , Brussels. SCENIHR (Scientific Committee on Emerging and Newly Identified Health Risks) (2006) Opinion of the appropriateness of existing methodologies to assess the potential risks associated with engineered and adventitious products of nanotechnologies, 10 March 2006, European Commission Health and Consumer Protection Directorate-General , Brussels. SCENIHR (Scientific Committee on Emerging and Newly Identified Health Risks) (2007) The appropriateness of the risk assessment methodology in accordance with the Technical Guidance Documents for new and existing substances for assessing the risks of nanomaterials. European Commission , Brussels. Schmid, M. and M. Riediker (2008) Use of nanoparticles in Swiss industry: a targeted survey. Environ. Sci. Technol., 42, 2253–60. Sen, D., H. Wolfson and M. Dilworth (2002) Lead exposure in scaffolders during refurbishment construction activity – An observational study. Occup. Med., 52, 49–54. Thanawalla, C. B. (2002) Complying with TSCA Inventory Requirements. John Wiley and Sons Inc., New York, USA. Thorne, P. S. (2001) Occupational Toxicology, in Casarrett and Doull’s Toxicology (ed. Klaaassen C. D.), 6th edn, McGraw Hill, London, UK. Topping, M. (2001) Occupational exposure limits for chemicals. Occup Environ Med., 58, 138–44. US EPA (1988) Technical Support Document on Risk Assessment of Chemical Mixtures. EPA-600/8-90/064. US Environmental Protection Agency, Office of Research and Development, Washington, DC, USA. Van Leeuwen, K. and T. Vermeire (Eds) (2007) Risk assessment of Chemicals – an introduction, 2nd edn, Springer, Berlin. Vasconcelos, P., M. B. Gaspar and M. Castro (2006) Imposex in Hexaplex (Trunculariopsis) trunculus (Gastropoda: Muricidae) from the Ria Formosa lagoon (Algarve coast–southern Portugal) Marine Poll. Bull., 52, 337–41.

Risk Assessment of Manufactured Nanomaterials

421

Wardak, A., M. E. Gorman, N. Swami and D. Rejeski (2007) Environmental regulation of nanotechnology and the TSCA. IEEE Tech. Soc., 26, 48–56. WATCH Committee Paper (2006) COSHH 2002 Communicating the changes, accessed at www.hse.gov.uk/aboutus/hsc/iacs/acts/watch/091106/p6.pdf. Weed, D. L. (2005) Weight of evidence: A review of concept and methods. Risk Analysis. 25, 1545–57.

Index

abrasives 72 acceptable exposure limits (AEL) 407 active nanostructures 6–7 active transport 366 adsorption reactions 275 AEL see acceptable exposure limits aerosols characterization 36, 179–80, 196–200, 336 exposure to nanoparticles 310, 313–18, 327–8, 336, 341, 348 see also atmospheric nanoparticles; ultrafine particles AFM see atomic force microscopy agglomeration aquatic nanoparticles 220–1, 224, 232, 236 exposure to nanoparticles 311, 317–18, 321–2, 332, 335–6, 343 aggregation chemical properties 98–9 environmental fate 270–2 exposure to nanoparticles 311, 318, 332, 343 kinetics 129–30, 136–40 natural colloids 129–30, 136–40 size/shape–property relationships 98–9 air pollution see atmospheric nanoparticles allergens 368 aluminium nanoparticles ecotoxicology 289 human toxicology 374 natural colloids 114–16, 121–2, 132 aluminium phyllosilicates 114–15 American National Standards Institute (ANSI) 23 American Society for Testing and Materials (ASTM) 4, 23 ammonium nitrate 175–6 analytical procedures aquatic nanoparticles 211–66

atmospheric nanoparticles 187–200 chemical composition 196–200 chemical properties 221–2, 223 colloid formation 212–14, 224, 232, 245–7 crystalline structure 221, 237–47 definitions 212–14 detection and characterization 21, 23 dispersion/agglomeration 220–1, 224, 232, 236 dissolution rates 223 ecotoxicology 253–4 electromagnetic scattering methods 229 elemental analysis 244–7 environmental fate 252–3 exposure to nanoparticles 330–8 fractionation methods 230–6 future developments 23–4 human toxicology 381 initial material characterization 252 light scattering methods 224–9 mass concentration 189, 194–5 method validation 251–2 microscopic methods 237–49 monitoring nanopollution 254–5 number/mass concentration 220, 227–8 overview 2, 4 particle number concentration 188–93 particle size distribution 214–19, 226–9, 234–5, 243 physical properties 33–4 sampling considerations 212–14, 224, 240, 247, 330–1, 337–8 shape 219–20, 228 spectroscopic methods 249–51 surface characteristics 188, 193–4, 222–3, 250–1 test strategy 252–5 trace constituents 223

Environmental and Human Health Impacts of Nanotechnology Edited by Jamie R. Lead and Emma Smith © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17634-7

424

Index

see also individual techniques ANSI see American National Standards Institute aquatic nanoparticles analytical procedures 211–66 bioavailability 111 chemical properties 221–2, 223 colloid formation 1–2, 109–12, 212–14, 224, 232, 245–7 crystalline structure 221, 237–47 dispersion/agglomeration 220–1, 224, 232, 236 dissolution rates 223 ecotoxicology 253–4, 277–8 electromagnetic scattering methods 229 elemental analysis 244–7 environmental fate 17–18, 136–41, 252–3 fractionation methods 230–6 initial material characterization 252 inorganic colloids 114–17 interaction forces 126–36 light scattering methods 224–9 method validation 251–2 microscopic methods 237–49 monitoring nanopollution 254–5 natural colloids 109–41 number/mass concentration 220, 227–8 organic macromolecules 117–24, 133–5, 137–8, 140–1 particle size distribution 214–19, 226–9, 234–5, 243 physical properties 121–36, 214–23 sampling considerations 212–14, 224, 240, 247 shape 219–20, 228 sources 212 spectroscopic methods 249–51 surface characteristics 222–3, 250–1 test strategy 252–5 trace constituents 223 transport processes 110–11 architecture see physical properties arrested precipitation 58–9 asbestos applications 66–7 human toxicology 357, 370, 373, 380 preparative methods 54 risk assessment 401 aspect ratio 16 aquatic nanoparticles 220 exposure to nanoparticles 319, 334–5, 351–2 human toxicology 370–3, 380 risk assessment 401 asthma 364

ASTM see American Society for Testing and Materials atmospheric nanoparticles 163–209 analytical procedures 187–200 chemical composition 178–80, 196–200 coagulation scavenging 177, 181–2 concentrations 182–7 condensation mechanisms 168, 170, 176–8, 181–2 environmental fate 17, 181–2 exposure to nanoparticles 321–2 human toxicology 360, 367–8, 370–2 indoor sources 171–4 industrial sources 174–5 mass concentration 189, 194–5 meteorological factors 175–8 natural sources 175, 177 nucleation mechanisms 166–70, 175–9, 184–7 occupational exposure 187–9 overview 2, 163–4 particle number concentration 188–93 primary sources 164–75 risk assessment 398–9, 410 secondary sources 175–8 sources 164–78 spatial variations 182–5 surface area 188, 193–4 temporal variations 185–7 vehicle emissions 164–71, 179–80, 182–7, 368, 375–6 atomic force microscopy (AFM) aquatic nanoparticles 247–9 exposure to nanoparticles 337 natural colloids 120, 124, 130, 135 overview 4 attrition methods 315–16 bacteria 277–81, 290–1, 295–7 bactericides 70–1 baking dust 396 BET see Brauner–Emmett–Teller bioaccumulation 283–4, 292, 378, 415 bioavailability 111, 372 biologically relevant aerosol fraction 327 biomass burning 175 biopolymers 137 biotic ligand model (BLM) 111 block copolymers 43, 64–5, 71 bottom up approach 55–9 Bradford Hill’s criteria 404–5 brake-wear emissions 171–2 Brauner–Emmett–Teller (BET) analysis 336

Index British Standards Institution (BSI) 4, 23, 310–11, 330, 358 buckminsterfullerene 312, 369 cadmium nanoparticles analytical procedures 250–1 ecotoxicology 299 physical properties 47, 52–3 size/shape–property relationships 83, 96–8 cadmium selendie nanoparticles 96–8 cadmium selenide/telluride see quantum dots calcium sulfate 55–6 capillary electrophoresis 235–6 capping agents 223, 250 carbon black applications 66, 67 exposure to nanoparticles 308, 319, 342, 344 human toxicology 361–4, 367–8, 371 preparative methods 62 carbon nanoparticles ecotoxicology 10–11, 289–93 environmental fate 19–20 physical properties 39 preparative methods 60–3 see also carbon black; carbon nanotubes; fullerenes carbon nanotubes (CNTs) analytical procedures 235 definitions 4 ecotoxicology 280–2, 290–3 environmental fate 18 exposure to nanoparticles 308, 312, 319, 323, 328, 338–41, 352 human toxicology 369, 370–3 physical properties 39, 48–9 preparative methods 60–1 carborundum 316 catalysis 46 cation exchange reactions 96–8 causality 404–5 centrifugation 232–3 ceramic nanoparticles 11 cerium nanoparticles applications 9 atmospheric nanoparticles 170 ecotoxicology 272, 293, 296 exposure to nanoparticles 326 human toxicology 375–6 overview 3 preparative methods 63 certified reference materials (CRM) 251–2 CFT see classical filtration theory CFUF see cross-flow ultrafiltration

425

charge stabilisation 41–2 Chemical Agents Directive 396, 398 chemical properties aggregation 98–9 analytical procedures 221–2, 223 aquatic nanoparticles 221–2, 223 atmospheric nanoparticles 178–80, 196–200 cation exchange reactions 96–8 classification 12–13 human toxicology 358–9, 379–80 overview 1 preparative methods 53 redox chemistry 81–7, 100 risk assessment 394, 401, 403, 408, 413–14 size/shape–property relationships 81–92, 94–99 surface functionality 223, 250 surface speciation 223 chemical vapour deposition (CVD) 61, 313, 315, 322 chemicals safety assessments (CSA) 411–13 classic light scattering 226–7 classical filtration theory (CFT) 143–7 classification of nanoparticles 12–14, 21–3 clay particle aggregation 114 cleaning activities 172–3 cmc see critical micelle concentration CNTs see carbon nanotubes coagulation scavenging 177, 181–2 coatings 92–3, 123–4 cobalt nanoparticles 96, 98 colloid formation analytical procedures 212–14, 224, 232, 245–7 aquatic nanoparticles 1–2, 109–12, 212–14, 224, 232, 245–7 ecotoxicology 270 exposure to nanoparticles 315, 322 physical properties 37, 41–5 terrestrial nanoparticles 109–12 see also natural colloids colloidal organic matter (COM) 117 combustion of hydrocarbons atmospheric nanoparticles 164–72, 174–5, 179–80, 182–7 human toxicology 368, 375–6 composite materials 323–4 condensation mechanisms 168, 170, 176–8, 181–2 condensation particle counters (CPC) 188–90, 332–4, 338 conductivity 50–2 consumer exposure 321, 324–6 contaminated land remediation 69–70, 321

426

Index

Control of Substances Hazardous to Health (COSHH) Regulations 346, 396, 398, 409 cooking activities 171–2 copolymers 43, 64–5, 71 copper nanoparticles 52, 281–2, 288, 298 COSHH see Control of Substances Hazardous to Health cosmetics 71–2, 358 cost factors 54 CPC see condensation particle counters critical micelle concentration (cmc) 44–5 CRM see certified reference materials cross-flow ultrafiltration (CFUF) 230–2 crystalline silica 357, 376–7, 398 CSA see chemicals safety assessments CVD see chemical vapour deposition daphnids ecotoxicology 277–8, 280–1, 285–9, 291–5, 299 exposure to nanoparticles 376 risk assessment 409 dark-field optical microscopy 277 DDL see diffuse double layers DEFRA see Department for the Environment, Food and Rural Affairs degradation pathways 89–98 dendritic nanoparticles 37–8, 313 Department of the Environment, Food and Rural Affairs (DEFRA) 325–6, 380, 390 Department of Trade and Industry (DTI) 307 deposition 318 dermal exposure 325, 328–9, 349–50 human toxicology 372, 376 risk assessment 413 diatoms 116 diesel emissions 164–70, 172, 179, 368, 375–6 diffuse double layers (DDL) 122–3 diffusion barriers 50–1 chargers 188, 194 exposure to nanoparticles 316–17 human toxicology 365–6 rates 50 diffusion limited aggregation (DLA) 138–9 dilution ratios 166–7 dimensional scale of the universe 4–5 disaggregation 140–1 dispersing agents 270

dispersion aquatic nanoparticles 220–1, 224, 232, 236 environmental fate 270–2 preparative methods 53–4 dissolution aquatic nanoparticles 223 ecotoxicology 270, 272–4 size/shape–property relationships 89–91 dissolved organic matter (DOM) 117 DLA see diffusion limited aggregation DLS see dynamic light scattering DLVO theory 126–30, 146, 270–1 DOM see dissolved organic matter dosimetry 23 double layer interactions 128 drug delivery 9, 71 exposure to nanoparticles 329 human toxicology 369 DSSC see dye sensitised solar cells DTI see Department of Trade and Industry dumbbell nanoparticles 37–8 dye sensitised solar cells (DSSC) 52–3, 69–70 dynamic light scattering (DLS) 225–6, 293 EAD see electrical aerosol detectors economic factors 6 ecotoxicology 267–305 adsorption reactions 275 analytical procedures 253–4 aquatic nanoparticles 253–4 bioaccumulation 283–4, 292 carbon-based nanoparticles 289–93 carriers of coexisting contaminants 11 cellular uptake 276–80 colloid formation 270 copper nanoparticles 281–2, 288, 298 direct toxicity 10–11 dissolution 270, 272–4 effects on micro-ecosystems 11–12 exposure routes 275–7 future investigations 23–4 iron nanoparticles 273, 275, 288, 299 membrane damage 280–1 metal oxides 272–4, 277–8, 282, 293–6 overview 1, 2, 267–9 physical properties 72, 268, 273 quantum dots 277–8, 282, 298–9 reactive oxygen species 282–3 redox reactions 273–5 release of toxic species 281–2 risk assessment 389, 394, 405–6, 415 silver nanoparticles 280, 288, 296–8 sources of nanoparticles 269

Index sources/pathways of nanomaterials 14 study review 285–9 testing protocols 284–5 toxicity mechanisms 280–3 see also environmental fate EDCs see endocrine disrupting chemicals EDX see energy dispersive X-ray spectroscopy EELS see electron energy loss spectroscopy electrical aerosol detectors (EAD) 336 electrical low pressure impactors (ELPI) 188–9, 192–3, 196, 333, 335, 337–8 electrochemical deposition 313 electromagnetic scattering methods 229 electron energy loss spectroscopy (EELS) 241, 243, 246 electron microscopy 188, 191–2, 237–47 electrophoresis 235–6 ELPI see electrical low pressure impactors endocrine disrupting chemicals (EDCs) 110 endocytosis 276–7, 366 energy dispersive X-ray spectroscopy (EDX) 241–3, 245–6 environmental applications 8–9 environmental colloids see natural colloids environmental fate aggregation/dispersion 270–2 analytical procedures 252–3 aquatic nanoparticles 17–18, 136–41, 252–3 atmospheric nanoparticles 17, 181–2 dissolution 270, 272–4 ecotoxicology 23–4, 269–75 exposure to nanoparticles 316–19, 321, 326 natural colloids 136–47 redox reactions 273–5 risk assessment 389, 394, 402–3, 409 size/shape–property relationships 79–80, 99–101 terrestrial nanoparticles 19–20, 141–7 Environmental Nanoscience Initiative 390 Environmental Protection Agency (USEPA) applications 66 ecotoxicology 284, 294, 299 risk assessment 390, 396, 415–17 environmental SEM (ESEM) see scanning electron microscopy enzymes 86 EPA see Environmental Protection Agency epiphaniometry 188, 194, 336 EPS see extracellular polymeric substances

427

equivalent spherical diameters (ESD) 215, 218 ESEM see scanning electron microscopy eutrophication 117, 246 exhaust emissions 164–72, 179–80, 182–7, 368, 375–6 exposure limits 397–9, 407 exposure to nanoparticles 307–56 aerosols 310, 313–18, 327–8, 336, 341, 348 analytical procedures 330–8 atmospheric nanoparticles 321–2 carbon nanotubes 308, 312, 319, 323, 328, 338–41, 352 composite materials 323–4 consumers 321, 324–6 control measures 346–50 definitions 309–11 dendritic nanoparticles 313 dermal 325, 328–9, 349–50, 372, 376, 413 emerging processes 338–42 environmental fate 316–19, 321, 326 existing processes 342–4 exposure metrics 327–30 exposure scenarios 319–26 fullerenes 308, 311–12 ingestion exposure 325–6, 329–30, 350 inhalation 321–3, 325–8, 330–2, 347–9, 361–2, 370–2, 376–7 mass concentration 333, 335 media coverage 307 nanowires 308, 313 number concentration 332–5 numbers of people at risk 344–6, 352–3 overview 307–9 physical properties 309–19, 320–1, 332–7 production growth 307–8 production processes 313, 314–16 quantum dots 313 risk assessment 346–7, 353, 358–60, 402, 413 sampling strategy 330–1, 337–8 specific surface area concentration 335–7 studies 338–44 see also human toxicology; occupational exposure external substances 92–3 extracellular polymeric substances (EPS) 120 facilitated diffusion 366 ferritin 33 FFF see field-flow fractionation FIAM see free ion activity model fibrillar polysaccharides 120–1

428

Index

field-flow fractionation (FFF) 232–5, 255 fish 277–8, 282–3, 291–8 flame pyrolysis 57, 58–9 flocculation processes 228 flow cytometry 296 fluorescence spectroscopy 250–1 fluorescent nanoparticles 68 food safety 390 fractal dimension analytical procedures 221 natural colloids 124–6, 138–9, 141 Framework Programme (EU) 390 free ion activity model (FIAM) 111 fuel additives 70, 170, 375–6 fullerenes 10–11, 19–20 analytical procedures 221–2, 235 ecotoxicology 270, 280, 282–3, 286, 289–93 exposure to nanoparticles 308, 311–12, 341–2 human toxicology 369–70 preparative methods 60 fulvic acids 119, 124–5, 271 gas chromatography (GC) 196 gas phase synthesis 314 GC see gas chromatography GLP see good laboratory practice gold nanoparticles 3 applications 65–6 human toxicology 374 physical properties 46 size/shape–property relationships 84–5 good laboratory practice (GLP) 405 graphene sheets 60–1, 62 groundwater remediation 8 HAADF see high angle annular dark field HARN see high aspect ratio nanoparticles hazard identification/characterisation 391–2, 401–2 HDC see hydrodynamic chromatography Health and Safety Executive (HSE) 327–9, 330, 346, 396, 398 HEPA see high efficiency particulate arresters high angle annular dark field (HAADF) imaging 246–7 high aspect ratio nanoparticles (HARN) exposure to nanoparticles 334–5, 351–2 human toxicology 370–3, 380 high efficiency particulate arresters (HEPA) 340, 347–8 high performance liquid chromatography (HPLC) 196

high pressure carbon monoxide (HiPCO) 338–9 high temperature methods 57, 58–9 Hill’s criteria 404–5 HiPCO see high pressure carbon monoxide HOCs see hydrophobic organic contaminants household emissions 171–4 HPLC see high performance liquid chromatography HSE see Health and Safety Executive human health applications 9–10 human toxicology 357–88 adverse effects 12 analytical procedures 381 atmospheric nanoparticles 360, 367–8 cellular uptake of nanoparticles 365–6 chemical properties 358–9, 379–80 definitions 357 dermal exposure 372, 376 endocytosis 366 exposure potential 20–1 exposure routes 275–7 fullerenes 369–70 future study designs 381 high aspect ratio nanoparticles 370–3 inflammatory responses 360–2, 364–7 inhalation exposure 361–2, 370–2, 376–7 metallic nanoparticles 373–4 occupational exposure 370–2 overview 2 oxides 374–7 particle toxicology 357–8 phagocytosis 361–2, 366–7 physical properties 358–9, 362, 379–80 quantum dots 377–9 reactive oxygen species 362–5, 373, 379 risk assessment 358–60, 390, 405–6, 413–15 structure: activity relationships 379–80 structure–toxicity relationship 15–16 ultrafine particles 360–8 see also exposure to nanoparticles humic acids 119, 124–5, 271 hydration effect 130–2 hydrocarbon combustion atmospheric nanoparticles 164–72, 174–5, 179–80, 182–7 human toxicology 368, 375–6 hydrodynamic chromatography (HDC) 235 hydrophobic interactions 132–3 hydrophobic organic contaminants (HOCs) 111 hydrous oxides 115–17 hydroxyl radicals 82, 175

Index IARC see International Agency for Research on Cancer IEC see International Electrotechnical Commission immunolabelling 68 indium nanoparticles 68 indoor emissions 171–4 industrial emissions 174–5 ingestion exposure 325–6, 329–30, 350 inhalation exposure 321–3, 325–8 analytical procedures 330–2 control measures 347–9 human toxicology 361–2, 370–2, 376–7 inorganic nanoparticles anatomy 80–1 aquatic nanoparticles 212 chemical properties 81–7 colloid formation 114–17 fate and degradation pathways 89–98 physical properties 39 preparative methods 63–4 size/shape–property relationships 79–108 sorption processes 87–9 interaction forces aggregation kinetics 129–30, 136–40 disaggregation 140–1 DLVO theory 126–30, 146 double layer interactions 128 hydration effect 130–2 hydrophobic interactions 132–3 natural colloids 126–36 non-DLVO interactions 130–6 polymer bridging 135–6 stability criteria 129 steric interactions 133–5 van der Waals’ forces 127–8 interfacial properties 49–50 International Agency for Research on Cancer (IARC) 357 International Electrotechnical Commission (IEC) 23 International Organisation for Standardisation (ISO) 4, 23 exposure to nanoparticles 310–11, 332 method validation 251–2 Inventory Update Rule (IUR) 416–17 investment 6 ionic strength 138–40 iron nanoparticles 19 analytical procedures 246 applications 69–70 atmospheric nanoparticles 170 ecotoxicology 273, 275, 288, 299 human toxicology 374, 377

429

natural colloids 109–10, 115–17 risk assessment 407 size/shape–property relationships 85–7, 88, 99–100 ISO see International Organisation for Standardisation IUR see Inventory Update Rule Kelvin equation 89, 95 Klimisch scores 405 lactate dehydrogenase (LDH) assays 371, 375, 377 lamp black 65–6 landfill 402, 410 Langmuir adsorption equation 87–9 laser ablation 338–9 laser diffraction 227–8 LDH see lactate dehydrogenase lead nanoparticles exposure to nanoparticles 329 natural colloids 116 size/shape–property relationships 94–6 lepidocrocite 243 LFA see long fibre amosite localised surface plasmon spectroscopy (LSPRS) 277 LOEC see low observed effect concentrations long fibre amosite (LFA) 373 low observed effect concentrations (LOEC) 402 low temperature processing 51 LSPRS see localised surface plasmon spectroscopy magnesium nanoparticles 278–9 magnetic properties 49 magnetic resonance imaging (MRI) 377 manganese nanoparticles 3, 115–17 mass average size 219 mass concentration 220, 227–8, 333, 335 mass spectrometry (MS) aquatic nanoparticles 234, 255 atmospheric nanoparticles 179–80, 196–200 material properties 14–15 material safety data sheets (MSDS) 416 maximum exposure limits (MEL) 398 mechanical attrition methods 315–16 mechanical properties 34, 48–9 media coverage 307 MEL see maximum exposure limits mercury nanoparticles 241 metal oxides see oxides

430

Index

metallic nanoparticles human toxicology 373–4 physical properties 39 preparative methods 59–60 size/shape–property relationships 84–5 see also individual metals meteorological factors 175–8 method validation 251–2 micelle encapsulation 57–8 micro-orifice uniform deposit impactors (MOUDI) 180, 194, 196 microemulsions 57–8 microfiltration 230 milling dust 396 molecular beam epitaxy 62 molecular nanosystems 6–7 molybdenum nanoparticles 82, 109–10, 374 monitoring nanopollution 254–5 morphological effects 37–8, 91–2, 94, 96 MOUDI see micro-orifice uniform deposit impactors MRI see magnetic resonance imaging MS see mass spectrometry MSDS see material safety data sheets multi-walled carbon nanotubes (MWCNTs) ecotoxicology 290 environmental fate 18 exposure to nanoparticles 340–1, 352 human toxicology 370–3 physical properties 48–9 preparative methods 60–1 Nano Risk Framework 407–8 nanoclays 54–5, 68, 323 nanocomposites 54–5, 68 nanomachines 34 nanomaterials definitions 4 exposure to nanoparticles 311, 341 risk assessment 399–403 nano-MOUDI impactors 180, 194 nanoparticle tracking analysis (NTA) 228–9 nanorods 4, 37–8 NANOSAFE project 350 nanosensors 8–9 nanowires 308, 313 narrow band gap semiconductors 63–4 National Centre for Environmental Research (NCER) 390 National Institute for Occupational Safety and Health (NIOSH) 337–8, 344, 396 National Nanotechnology Initiative (NNI) 3

natural colloids 109–61 aggregation kinetics 129–30, 136–40 aluminium phyllosilicates 114–15 aquatic nanoparticles 109–41 behavioural processes 117–18 classification 112–21 definitions 112 disaggregation 140–1 DLVO theory 126–30, 146 environmental fate 136–47 fibrillar polysaccharides 120–1 formation 109–12 fractal dimension 124–6, 138–9, 141 humic substances 119 hydrous oxides 115–17 inorganic colloids 114–17 interaction forces 126–36 non-DLVO interactions 130–6 organic macromolecules 117–24, 133–5, 137–8, 140–1 overview 1–2 oxides 115–17 particle filtration 142–7 physical properties 121–36 saturated porous media 143–6 sedimentation 141–2 size distribution 121 stability criteria 129 surface charge 121–3 surface coatings 123–4 terrestrial nanoparticles 109–36, 141–7 transport processes 110–11 unsaturated porous media 146–7 natural nanoparticles 308, 323 natural organic matter (NOM) aquatic nanoparticles 220–1, 223, 230 colloid formation 117–24, 133–5, 137–8 ecotoxicology 270–2 NCER see National Centre for Environmental Research nickel nanoparticles 308, 344 NIOSH see National Institute for Occupational Safety and Health NNI see National Nanotechnology Initiative no observed adverse effect levels (NOAEL) 402, 413 NOM see natural organic matter nomenclature of nanoparticles 21–3 NTA see nanoparticle tracking analysis nucleation mechanisms 166–70, 175–9, 184–7 number average size 219 number concentration 220, 227–8, 332–5 nZVI see zero-valent iron

Index occupational exposure 2 analytical methods 330–2, 337–8 atmospheric nanoparticles 187–9 carbon nanoparticles 312–13, 352 control measures 346–50 definitions 309–10 exposure metrics 327–30, 350–3 exposure scenarios 320–4 human toxicology 370–2 numbers of people at risk 344–6 risk assessment 390, 396–9, 409–10 studies 338–44 occupational exposure limits (OELs) 397–9 occupational exposure standard (OES) 398 Occupational Safety and Health Administration (OSHA) 396 OECD see Organisation for Economic Co-operation and Development OELs see occupational exposure limits OES see occupational exposure standard OPC see optical particle counters optical microscopy 277 optical nanoparticles 67 optical particle counters (OPC) 332–3 optical spectroscopy 249–50 organic macromolecules 117–24 aggregation kinetics 137–8 fibrillar polysaccharides 120–1 humic substances 119 interaction forces 133–5 surface charge 122 organic nanoparticles 212 see also natural organic matter Organisation for Economic Co-operation and Development (OECD) 4, 6 ecotoxicology 284, 294 method validation 252 risk assessment 391, 394, 405–7 OSHA see Occupational Safety and Health Administration Ostwald ripening 59, 273 oxidation catalysts 169–70 oxidative stress 362–5, 374 oxides analytical procedures 246 ecotoxicology 272–4, 277–8, 282, 293–6 human toxicology 374–7 natural colloids 109–10, 115–17 preparative methods 63 size/shape–property relationships 86–7, 88 PAA see poly(acrylic acid) PAHs see polyaromatic hydrocarbons paint 329

431

palladium catalysis 46 palladium nanoparticles 84–5 particle filtration 142–7 particle size distribution (PSD) aquatic nanoparticles 214–19, 226–9, 234–5, 243 ecotoxicology 273 risk assessment 414 particle toxicology 357–8 particles not otherwise specified (PNOS) 398 particulate organic matter (POM) 117 passive nanoparticles 6–7, 34 PCBs see polychlorinated biphenyls PCDDs see polychlorinated dibenzo-p-dioxins PCDFs see polychlorinated dibenzofurans PEN see Project on Emerging Nanotechnology peroxidases 86 personal protective equipment (PPE) 339, 348–50 pesticides 289, 299, 396 petrol emissions 164, 170–2 phagocytosis 361–2, 366–7 pharmaceuticals 110, 329, 344 phosphides 64 photo-oxidants 175 photo-resistant packaging 72 photoredox chemistry 81–4, 100 photovoltaics 52–3, 68–70 physical properties 31–53 analytical procedures 33–4, 45–6 applications 71–2 aquatic nanoparticles 214–23 architecture 1, 35–45 catalysis 46 charge stabilisation 41–2 classification 13–14 colloid formation 37, 41–5 conductivity 50–2 crystalline structure 221, 237–47 definitions 35–6 diffusion barriers 50–1 diffusion rates 50 dispersion/agglomeration 220–1, 224, 232, 236 dissolution rates 223 ecotoxicology 72, 268, 273 exposure to nanoparticles 309–19, 320–1, 332–7 human toxicology 358–9, 362, 379–80 interaction forces 126–36 interfacial properties 49–50 low temperature processing 51

432

Index

magnetic properties 49 mechanical properties 34, 48–9 morphologies 37–8 natural colloids 121–36 number/mass concentration 220, 227–8 overview 1, 31–5 particle properties 45–53 particle size distribution 121, 214–19, 226–9, 234–5, 243, 273, 414 quantum confinement 47–8 risk assessment 394, 401–3, 408, 413–14 shape 219–20, 228 solar cells 52–3 steric stabilisation 42–5 surface characteristics 38–41, 121–4, 193–4, 222–3, 250–1, 335–7 see also size/shape–property relationships Pickering emulsions 49–50 platinum catalysis 46 platinum nanoparticles 84–5 PMN see Pre-Manufacture Notifications PNEC see predicted no effect concentrations PNOS see particles not otherwise specified pollen fragments 175 poly(acrylic acid) (PAA) 19 polyaromatic hydrocarbons (PAHs) 62, 67, 110 polychlorinated biphenyls (PCBs) 299 polychlorinated dibenzo-p-dioxins (PCDDs) 110 polychlorinated dibenzofurans (PCDFs) 110 polymer bridging 135–6 polymer nanoparticles 64–5 polysaccharides 120–1 polystyrene beads 361–3, 367 POM see particulate organic matter power stations 174–5 PPE see personal protective equipment Pre-Manufacture Notifications (PMN) 415–16 predicted no effect concentrations (PNEC) 402 preparative methods 53–65 arrested precipitation 58–9 bottom up approach 55–9 carbon black 62 carbon nanoparticles 60–3 chemical properties 53 cost factors 54 dispersability 53–4 graphene sheets 60–1, 62 high temperature methods 57, 58–9 inorganic nanoparticles 63–4

metallic nanoparticles 59–60 micelle encapsulation 57–8 narrow band gap semiconductors 63–4 natural sources 54–5 oxides 63 polymer nanoparticles 64–5 scale up 53 top down approach 55 wet methods 57–9 Project on Emerging Nanotechnology (PEN) 324–5 PSD see particle size distribution quantum confinement 47–8 quantum dots 9, 48 analytical procedures 250–1 ecotoxicology 277–8, 282, 298–9 exposure to nanoparticles 313 human toxicology 377–9 quantum size effects 81–2 radio frequency identification (RFID) tags 50–1, 52 rapid single-particle mass spectrometry (RSMS) 197 re-suspension 318–19, 322 REACH guidelines 392, 396, 398–401, 411–15 reaction limited aggregation (RLA) 138–9 reactive oxygen species (ROS) aquatic nanoparticles 223 ecotoxicology 282–3 human toxicology 362–5, 373, 379 redox chemistry 81–7, 100, 273–5 respiratory protective equipment (RPE) 323, 349 retrospective risk assessment 395–6 RFID see radio frequency identification rhodium catalysis 46 risk assessment 389–421 causality 404–5 chemical properties 394, 401, 403, 408, 413–14 data gaps 407–10 ecotoxicology 389, 394, 405–6, 415 environmental fate 389, 394, 402–3, 409 evidence assembly 403–6 exposure limits 397–9, 407 exposure to nanoparticles 346–7, 353, 358–60, 402, 413 framework design 392 hazard identification/ characterisation 391–2, 401–2 human toxicology 390, 405–6, 413–15 International case studies 406–7

Index nanomaterials 399–403 occupational exposure 390, 396–9, 409–10 overview 2, 23, 389–91 physical properties 394, 401–3, 408, 413–14 process 391–9 regulatory guidelines 392–5, 397–401, 411–17 retrospective 395–6 risk estimation/evaluation 391–2 testing strategies 399–400 trigger points 393–5, 399, 411–12 RLA see reaction limited aggregation road–tyre interface 171 rolling chassis dynamometers 166 ROS see reactive oxygen species rotaxanes 34–5 Royal Society/Royal Academy of Engineering (RS/RAEng) 3–4 exposure to nanoparticles 309, 319–20 risk assessment 390, 399, 406 RPE see respiratory protective equipment RSMS see rapid single-particle mass spectrometry SAED see selected area electron diffraction safety cases 403–4 sampling strategies aquatic nanoparticles 212–14, 224, 240, 247 exposure to nanoparticles 330–1, 337–8 SANS see small angle neutron scattering saturated porous media 143–6 SAXS see small angle X-ray scattering scale up 53 scanning electron microscopy (SEM) 237–47 exposure to nanoparticles 333, 336–9, 342 historical development 33 overview 4 scanning mobility particle sizers (SMPS) 188–91, 194, 333–4, 337–8, 340–3 scanning probe microscopy (SPM) aquatic nanoparticles 247–9 historical development 33 particle properties 45–6 Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR) 391, 401, 403, 406 SEC see size exclusion chromatography sedimentation 41, 141–2, 271

433

selected area electron diffraction (SAED) 241 selenium nanoparticles 83, 96–8 SEM see scanning electron microscopy semi-volatile gases 175–8 semiconducting nanoparticles analytical procedures 250–1 applications 8–9 photoredox chemistry 81–4, 100 physical properties 47–8, 52–3 preparative methods 63–4 Seventh Framework Programme 307 short fibre amosite (SFA) 373 silica nanoparticles ecotoxicology 283, 285, 288, 293, 295 exposure to nanoparticles 308, 319 human toxicology 357, 376–7 natural colloids 114–16 physical properties 39 preparative methods 54–5 risk assessment 398 silver ink 50–2 silver nanoparticles chemical properties 223 ecotoxicology 280, 288, 296–8 human toxicology 374 natural colloids 121–2 physical properties 219 size/shape–property relationships 84–5, 100 single-walled carbon nanotubes (SWCNTs) ecotoxicology 280–1, 290–3 exposure to nanoparticles 312, 338–40 human toxicology 369, 370–3 preparative methods 60 size averages 219 size exclusion chromatography (SEC) 235 size selective counters 333, 335 size/shape–property relationships 79–108 aggregation 98–9 anatomy 80–1 coatings and external substances 92–3 dissolution 89–91 environmental fate 79–80, 99–101 fate and degradation pathways 89–98 inorganic nanoparticles 79–108 metallic nanoparticles 84–5 morphological effects 91–2, 94, 96 redox chemistry 81–7, 100 semiconducting nanoparticles 81–4 solid state cation movement 96–8 sorption processes 87–9 skin protective equipment (SPE) 349 small angle neutron scattering (SANS) 229 small angle X-ray scattering (SAXS) 229

434

Index

smoking 171–2 SMPS see scanning mobility particle sizers solar cells see photovoltaics soldering 344 solid state cation movement 96–8 solubilising agents 270 soot mode 166, 169, 179 sorption processes 87–9, 110 spatial variations 182–5 SPE see skin protective equipment specific surface area concentration 335–7 SPM see scanning probe microscopy SPR see surface plasmon resonance stability criteria 129 state classification 14 static light scattering 226–7 STEM see transmission electron microscopy steric interactions 133–5 steric stabilisation 42–5 Stern layers 122–3 structure–toxicity relationship 15–16 sulfides 82 natural colloids 110 preparative methods 63–4 redox properties 82 size/shape–property relationships 94–6 sulfur-rich fuels 168–70 sunscreens 71–2, 79, 325–6, 358 supersaturation 56 surface characteristics 38–41 analytical procedures 188, 193–4, 222–3, 250–1 aquatic nanoparticles 222–3, 250–1 atmospheric nanoparticles 188, 193–4 charge 121–3, 222–3 exposure to nanoparticles 335–7 functionality 223, 250 natural colloids 121–4 speciation 223 surface area 188, 193–4, 222, 251, 335–7 surface plasmon effects 250 surface plasmon resonance (SPR) 65 surfactants 44–5, 60, 271 SWCNTs see single-walled carbon nanotubes systems of nanosystems 6–7 tandem differential mobility analysers (TDMA) 200 tapered element oscillating microbalances (TEOM) 189, 195, 333 TDCIMS see thermal desorption chemical ionization mass spectrometry TDMA see tandem differential mobility analysers

TDPBMS see thermal desorption particle beam mass spectrometry tear drop nanoparticles 37–8 TEM see transmission electron microscopy temporal variations 185–7 TEOM see tapered element oscillating microbalances terpenes 177 terrestrial nanoparticles bioavailability 111 colloid formation 1–2, 109–12 environmental fate 19–20, 141–7 inorganic colloids 114–17 interaction forces 126–36 natural colloids 109–36, 141–7 organic macromolecules 117–24, 133–5 physical properties 121–36 transport processes 110–11 tetrapod nanoparticles 37–8 thermal desorption chemical ionization mass spectrometry (TDCIMS) 199–200 thermal desorption particle beam mass spectrometry (TDPBMS) 199 threshold limit values (TLVs) 397–8 tin nanoparticles 8–9 titanium dioxide analytical procedures 221, 229, 240, 242 applications 70, 71–2 ecotoxicology 277, 280–1, 283–4, 287–8, 292–5 exposure to nanoparticles 308, 311, 315, 319, 343–4 human toxicology 360, 364, 366–7, 374–5, 377 risk assessment 407 size/shape–property relationships 79, 88, 90–1, 100 TLVs see threshold limit values top down approach 55 Toxic Substances Control Act (TSCA) 415–17 toxicology see ecotoxicology; human toxicology trace constituents 223 transmission electron microscopy (TEM) aquatic nanoparticles 237, 240–7, 252, 255 exposure to nanoparticles 333, 336–8, 341 historical development 33 human toxicology 372 natural colloids 120, 124 overview 4, 22 size/shape–property relationships 94–5

Index trigger points 393–5, 399, 411–12 TSCA see Toxic Substances Control Act turbidimetry 228 UF-ATOFMS see ultrafine aerosol time-offlight mass spectrometry UFPs see ultrafine particles ultracentrifugation 232–3 ultrafiltration 230–2 ultrafine aerosol time-of-flight mass spectrometry (UF-ATOFMS) 179–80, 197–8 ultrafine particles (UFPs) 12, 17 atmospheric nanoparticles 163–4, 170–6, 182–4 characterization 36 definitions 310 ecotoxicology 277, 283 exposure to nanoparticles 310, 340, 343 fuel additives 70 human toxicology 360–8 risk assessment 398–9 ultraviolet (UV) spectroscopy 249–50 unsaturated porous media 146–7 USEPA see Environmental Protection Agency van der Waals’ forces 127–8, 271, 311, 318 vapour–liquid–solid (VLS) method 61 vehicle emissions 164–72, 179–80, 182–7, 368, 375–6 VH-TDMA see volatilization and humidification tandem differential mobility analysers viruses 175

435

visible spectroscopy 249–50 VLS see vapour–liquid–solid volatilization and humidification tandem differential mobility analysers (VH-TDMA) 179 volume average size 219 voluntary reporting 390 waste disposal 402, 410 WEL see workplace exposure limits welding 344, 396 wet preparation methods 57–9 WHO see World Health Organization Woodrow Wilson inventory 324–5 workplace exposure limits (WEL) 327 World Health Organization (WHO) 331, 334, 396 X-ray absorption spectroscopy (XAS) 99 X-ray diffraction (XRD) 94, 97 X-ray photoelectron spectroscopy (XPS) 94, 250 Z-average size 219 zeolites 32 zero-valent iron (nZVI) 85–6, 100, 407 zinc nanoparticles applications 8–9, 71–2 ecotoxicology 272–4, 280–2, 285, 288, 293–6 exposure to nanoparticles 315, 344 human toxicology 375 preparative methods 63 size/shape–property relationships 99 ZVI see zero-valent iron