Self-Cleaning Materials and Surfaces: a Nanotechnology Approach 9781119991779, 1711721751, 2052052062, 1119991773

Citation preview

Editor: Walid A. Daoud

Self-Cleaning Materials and Surfaces A Nanotechnology Approach

Self-Cleaning Materials and Surfaces A Nanotechnology Approach

Self-Cleaning Materials and Surfaces A Nanotechnology Approach

Edited by WALID A. DAOUD School of Energy and Environment, City University of Hong Kong, Hong Kong

This edition first published 2013  C 2013 John Wiley & Sons, 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.

Library of Congress Cataloging-in-Publication Data Self-cleaning materials and surfaces : a nanotechnology approach / edited by Walid A. Daoud. pages cm Includes bibliographical references and index. ISBN 978-1-119-99177-9 (cloth) 1. Coatings. 2. Surface active agents. 3. Materials–Cleaning. 4. Nanostructured materials. TA418.9.C57S45 2013 667 .9–dc23

I. Daoud, Walid A.

2013016955 A catalogue record for this book is available from the British Library. ISBN: 9781119991779 Set in 10/12pt Times by Aptara Inc., New Delhi, India. 1 2013

Contents

List of Contributors Preface

xiii xv

PART I CONCEPTS OF SELF-CLEANING SURFACES 1

Superhydrophobicity and Self-Cleaning Paul Roach and Neil Shirtcliffe 1.1

Superhydrophobicity 1.1.1 Introducing Superhydrophobicity 1.1.2 Contact Angles and Wetting 1.1.3 Contact Angle Hysteresis 1.1.4 The Effect of Roughness on Contact Angles 1.1.5 Where the Equations Come From 1.1.6 Which State Does a Drop Move Into? 1.2 Self-Cleaning on Superhydrophobic Surfaces 1.2.1 Mechanisms of Self-Cleaning on Superhydrophobic Surfaces 1.2.2 Other Factors 1.2.3 Nature’s Answers 1.2.4 Superhydrophilic Self-Cleaning Surfaces 1.2.5 Functional Properties of Superhydrophobic Surfaces 1.3 Materials and Fabrication 1.4 Future Perspectives References

3 3 3 4 4 6 8 11 12 12 15 17 19 20 25 27 28

PART II APPLICATIONS OF SELF-CLEANING SURFACES 2

Recent Development on Self-Cleaning Cementitious Coatings Daniele Enea

35

2.1 2.2

35 36 36 37 37 38

2.3

Introduction Atmospheric Pollution: Substances and Laws 2.2.1 Nitrogen Oxides 2.2.2 Particulate Matter 2.2.3 Volatile Organic Compounds Heterogeneous Photocatalysis

vi

Contents

2.4

3

Self-Cleaning Surfaces 2.4.1 Mechanisms of Photo-Reduction of Air Pollutants 2.4.2 Some Experimental Evidences 2.5 Main Applications 2.6 Test Methods 2.6.1 Colour 2.6.2 Photocatalytic Degradation of Nitrogen Oxides 2.6.3 Photocatalytic Degradation of Micro-Pollutants in Air 2.6.4 Photocatalytic Degradation of Rhodamine B 2.6.5 Spectroscopic Techniques 2.7 Future Developments References

39 41 41 44 46 46 47 49 51 53 53 54

Recent Progress on Self-Cleaning Glasses and Integration with Other Functions Baoshun Liu, Qingnan Zhao and Xiujian Zhao

57

3.1 3.2

4

Introduction Theoretical Fundamentals for Self-Cleaning Glasses 3.2.1 Wettability 3.2.2 Photoinduced Hydrophilicity 3.2.3 Heterogeneous Photocatalysis 3.3 Self-Cleaning Glasses Based on Photocatalysis and Photoinduced Hydrophilicity 3.3.1 Self-Cleaning Glasses with Pores 3.3.2 Doping to Realize Visible-Light-Induced Self-Cleaning Glasses 3.3.3 The Use of Hole Transfer to Realize Self-Cleaning 3.3.4 The Effect of Temperature and Atmosphere on the Photoinduced Hydrophilicity 3.3.5 The Effect of Soda Ions on the Properties of Self-Cleaning Glasses 3.3.6 The Anti-Bacterial Effect and Anti-Fogging Effect 3.3.7 The Composite SiO2 Films for Self-Cleaning Glasses with High Antireflection 3.4 Inorganic Hydrophobic Self-Cleaning Glasses 3.4.1 Modifying The TiO2 Film by Low-Electronegativity Elements 3.4.2 The Application of ZnO Material in a Superhydrophobic Material 3.5 Self-Cleaning Glasses Modified by Organic Molecules 3.6 The Functionality of Self-Cleaning Glasses References

57 58 58 59 62

Self-Cleaning Surface of Clay Roofing Tiles Jonjaua Ranogajec and Miroslava Radeka

89

4.1

89 89

Clay Roofing Tiles and Their Deterioration Phenomena 4.1.1 Raw Material Composition and Firing Process

63 63 65 67 67 69 70 72 75 75 77 79 80 84

Contents

4.1.2 4.1.3

Surface Characteristics of Clay Roofing Tiles Frost, Chemical and Biocorrosion Deterioration of Clay Roofing Tiles 4.1.4 Simulation of Weathering of Clay Roofing Tiles in Laboratory Conditions 4.2 Protective and Self-Cleaning Materials for Clay Roofing Tiles 4.2.1 Design of Protective and Self-Cleaning Coatings 4.2.2 Monitoring the Characteristics of Coated Clay Roofing Tiles References 5

91 96 97 105 107 113 123

Self-Cleaning Fibers and Fabrics Wing Sze Tung and Walid A. Daoud

129

5.1 5.2

129 130 131 132

Introduction Photocatalysis 5.2.1 Mechanisms 5.2.2 Titanium Dioxide Photocatalyst 5.3 Photocatalytic Self-Cleaning Surface Functionalization of Fibrous Materials 5.3.1 Self-Cleaning Cellulosic Fibers 5.3.2 Self-Cleaning Keratin Fibers 5.3.3 Self-Cleaning Synthetic Fibers 5.4 Application of Photocatalytic Self-Cleaning Fibers 5.4.1 Protective Clothing 5.4.2 Household Appliances and Interior Furnishing 5.5 Limitations 5.5.1 Environmental Concerns 5.5.2 Human Safety Concerns 5.5.3 Photocatalytic Efficiency and Stability 5.6 Future Prospects 5.6.1 Visible Light Activation 5.6.2 Remote Photocatalytic Effect 5.6.3 Process Modification 5.6.4 Empirical Measurements 5.7 Conclusions References 6

vii

134 134 139 140 142 142 143 144 144 144 145 146 146 146 146 147 147 147

Self-Cleaning Materials for Plastic and Plastic-Containing Substrates Houman Yaghoubi

153

6.1 6.2

153

Introduction TiO2 Thin Films on Polymers: Sol–Gel-Based Wet Coating Techniques 6.2.1 Wet Coating Techniques: History and Advantages 6.2.2 TiO2 Photocatalytic Thin Films on PC and PMMA 6.2.3 SiO2 Incorporation into TiO2 - SiO2 as an Interfacial Layer for TiO2 6.2.4 TiO2 Photocatalytic Thin Films on PET and HDPE

155 155 156 162 167

viii

Contents

6.2.5 6.2.6

TiO2 Photocatalytic Thin Films on PS Modified Hybrid TiO2 Sols on Plastics: ABS, Polystyrene, and PVC 6.2.7 TiO2 on Paints and Self-Cleaning Paints 6.2.8 MW Irradiation–Assisted Dip Coating for Low-Temperature TiO2 Deposition on Polymers 6.2.9 Nanomechanical Properties of Dipped TiO2 Granular Thin Films on Polymer Substrates 6.3 TiO2 –Polymer Nanocomposites Review: Casting (Mixing) Techniques 6.3.1 Short History and Advantages 6.3.2 Ag/Polyethylene Glycol (PEG)–Polyurethane (PU)–TiO2 Nanocomposite Films by Solution Casting Techniques 6.3.3 Antimicrobial Activity of TiO2 -Isotactic Polypropylene (iPP) Composites 6.3.4 TiO2 Immobilized Biodegradable Polymers 6.4 TiO2 Sputter-Coated Films on Polymer Substrates 6.4.1 DC Reactive Magnetron Sputtering of Photocatalytic TiO2 Films on PC 6.4.2 Reactive Radio-Frequency [RF] Magnetron Sputtering of Photocatalytic TiO2 Films on PET 6.5 TiO2 Thin Films on PET and PMMA by Nanoparticle Deposition Systems (NPDS) 6.6 Photo-Responsive Discharging Effect of Static Electricity on TiO2 -Coated Plastic Films 6.7 Recent Achievements 6.7.1 Commercialized Products: Ube-Nitto Kasei Co. and the University of Tokyo 6.7.2 Patents: University of Wisconsin Acknowledgements References

171 172 175 178 179 181 181 182 183 184 187 187 189 190 192 192 192 193 194 194

PART III ADVANCES IN SELF-CLEANING SURFACES 7

Self-Cleaning Textiles Modified by TiO2 and Bactericide Textiles Modified by Ag and Cu John Kiwi and Cesar Pulgarin 7.1 7.2 7.3 7.4

Introduction Self-Cleaning Textiles: RF-Plasma Pretreatment to Increase the Binding of TiO2 Self-Cleaning Mechanism for Colorless and Colored Stains on Textiles Self-Cleaning Textiles: Vacuum-UVC Pretreatment to Increase the Binding of TiO2

205 205 206 208 209

Contents

XPS to Follow Stain Discoloration on Cotton Modified with TiO2 and Characterization of the TiO2 Coating 7.6 Bactericide/Ag/Textiles Prepared by Pretreatment with Vacuum-UVC 7.7 DC-Magnetron Sputtering of Textiles with Ag Inactivating Airborne Bacteria 7.8 Inactivation of E. coli by CuO in Suspension in the Dark and Under Visible Light 7.9 Inactivation of E. coli by Pretreated Cotton Textiles Modified with Cu/CuO at the Solid/Air Interface 7.10 Direct Current Magnetron Sputtering (DC and DCP) of Nanoparticulate Continuous Cu-Coatings on Cotton Textile Inducing Bacterial Inactivation in the Dark and Under Light Irradiation 7.11 Future Trends References

ix

7.5

8

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings with Easy-to-Clean Properties Mikko Aromaa, Joe A. Pimenoff and Jyrki M. M¨akel¨a 8.1 8.2

Gas-Phase Synthesis of Nanoparticles Aerosol Reactors 8.2.1 Hot Wall Reactors 8.2.2 Laser Reactors 8.2.3 Plasma Reactors 8.2.4 Flame Reactors 8.2.5 Spray Pyrolysis 8.3 Liquid Flame Spray 8.3.1 Synthesis of Nanoparticles via LFS 8.3.2 Multicomponent Nanoparticles 8.3.3 Synthesis and Deposition of Nanoparticle Coatings 8.4 Liquid Flame Spray in Synthesis of Easy-to-Clean Antimicrobial Coatings 8.4.1 Synthesis of Titanium Dioxide 8.4.2 Deposition of the Titania Coatings 8.4.3 Doping of the Coatings 8.4.4 Performance of the Antimicrobial Easy-to-Clean Coatings 8.5 Summary References 9

212 214 217 218 220

220 223 224

229 229 233 233 234 234 235 236 237 237 238 240 243 243 244 246 247 249 249

Pulsed Laser Deposition of Surfaces with Tunable Wettability Evie L. Papadopoulou

253

9.1 9.2

253 254 254 255

Introduction Basic Theory of Wetting Properties of Surfaces 9.2.1 Planar Surfaces 9.2.2 Rough Surfaces

x

Contents

9.3

10

Roughening a Flat Surface 9.3.1 PLD Technique Overview 9.3.2 Nanostructures Grown by PLD 9.4 Switchable Wettability 9.4.1 Photoinduced Wettability on PLD Structures 9.4.2 Electrowetting on PLD Structures 9.5 Concluding Remarks References

256 257 257 263 263 267 270 271

Fabrication of Antireflective Self-Cleaning Surfaces Using Layer-by-Layer Assembly Techniques Yu-Min Yang

277

10.1 Introduction 10.2 Antireflective Coatings 10.2.1 Interference Multiple Layers 10.2.2 Inhomogeneous Layer with Gradient Refractive Index 10.3 Solution-Based Layer-by-Layer (LbL) Assembly Techniques 10.3.1 Electrostatic Assembly 10.3.2 Langmuir–Blodgett (LB) Assembly 10.3.3 Self-Assembly 10.4 Mechanisms of Self-Cleaning 10.4.1 Hydrophilic Surfaces 10.4.2 Hydrophobic Surfaces 10.5 Fabrication of Antireflective Self-Cleaning Surfaces Using Electrostatic Layer-by-Layer (ELbL) Assembly of Nanoparticles 10.5.1 Superhydrophilic Self-Cleaning Surfaces with Antireflective Properties 10.5.2 Superhydrophobic Self-Cleaning Surfaces with Antireflective Properties 10.6 Fabrication of Superhydrophobic Self-Cleaning Surfaces Using LB Assembly of Micro-/Nanoparticles 10.7 Characterization of As-Fabricated Surfaces 10.7.1 Surface Morphology and Roughness 10.7.2 Thickness, Porosity, and Refractive Index 10.7.3 Transmittance 10.7.4 Photocatalytic Properties 10.7.5 Contact Angle and Contact Angle Hysteresis 10.7.6 Mechanical Stability 10.8 Challenges and Future Development 10.9 Conclusion References

277 278 278 279 280 280 281 282 283 283 284

285 285 291 297 300 300 301 302 303 304 305 306 307 307

Contents

xi

PART IV POTENTIAL HAZARDS AND LIMITATIONS

OF SELF-CLEANING SURFACES 11

The Environmental Impact of a Nanoparticle-Based Reduced Need of Cleaning Product and the Limitation Thereof L. Reijnders 11.1 Introduction 11.1.1 Outline 11.1.2 Nanoparticle-Based Reduced Need of Cleaning Surfaces 11.2 Titania and Amorphous Silica Nanoparticles and Carbon Nanotubes Can Be Hazardous and May Pose a Risk 11.2.1 Molecular Mechanisms 11.2.2 Risk Caused by Nanoparticles 11.3 Environmental Impact of a Reduced Need of Cleaning Product 11.3.1 Direct Environmental Effects of a Nanoparticle-Based Reduced Need of Cleaning Product 11.3.2 Net Direct Environmental Benefits 11.3.3 Indirect Environmental Effects of a Nanoparticle-Based Reduced Need of Cleaning Product 11.4 Limiting the Direct Environmental Impact of a Nanoparticle-Based Reduced Need of Cleaning Product, Including Limitation of Risks Following from Exposure to Nanoparticles 11.4.1 Limiting the Direct Environmental Impact 11.4.2 Limitation of Risks Following from Exposure to Nanoparticles 11.5 Conclusion References

Index

315 315 315 316 319 322 322 323 324 328 329

330 330 330 331 331 347

List of Contributors

Mikko Aromaa, Aerosol Physics Laboratory, Department of Physics, Tampere University of Technology, Finland Walid A. Daoud, School of Energy and Environment, City University of Hong Kong, Hong Kong Daniele Enea, Department of Architecture, University of Palermo, Italy John Kiwi, Institute of Chemical Sciences and Engineering, Swiss Federal Institute of Technology Lausanne (EPFL), Switzerland Baoshun Liu, State Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, PR China and School of Material Science and Engineering, Wuhan University of Technology, PR China Jyrki M. M¨akel¨a, Aerosol Physics Laboratory, Department of Physics, Tampere University of Technology, Finland Evie L. Papadopoulou, Institute of Electronic Structures and Lasers, Foundation for Research and Technology-Hellas, Greece. Current address: Istituto Italiano di Tecnologia, Genova, Italy Joe A. Pimenoff, Beneq Oy, Vantaa, Finland Cesar Pulgarin, Institute of Chemical Sciences and Engineering, Swiss Federal Institute of Technology Lausanne (EPFL), Switzerland Miroslava Radeka, Faculty of Technical Sciences, University of Novi Sad, Serbia Jonjaua Ranogajec, Faculty of Technology, University of Novi Sad, Serbia L. Reijnders, Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, The Netherlands Paul Roach, Institute for Science and Technology in Medicine, Guy Hilton Research Centre, Keele University, UK

xiv

List of Contributors

Neil Shirtcliffe, Faculty of Technology and Bionics, Hochschule Rhein-Waal, Germany Wing Sze Tung, School of Applied Sciences and Engineering, Monash University, Australia Houman Yaghoubi, Department of Mechanical Engineering/Department of Electrical Engineering, University of South Florida, USA Yu-Min Yang, Department of Chemical Engineering, National Cheng Kung University, Taiwan Qingnan Zhao, State Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, PR China Xiujian Zhao, State Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, PR China

Preface

With increasing demand for hygienic, self-disinfecting, and contamination-free surfaces, interest in developing efficient self-cleaning, protective surfaces and materials has grown. Due to rising population density, the spreading of antibiotic-resistant pathogens remains a growing global concern. The ability of microorganisms to survive on environmental surfaces makes infection transmission a critical issue, and studies have shown that some infectious bacteria can survive on the surface of various polymeric and textile materials for more than 90 days. Self-cleaning surfaces not only provide protection against infectious diseases but also against odor, staining, deterioration and allergies. Advances in nanotechnologies could make dirt-free (or no-wash) surfaces a reality. This would improve the environment through reduced use of water, energy and petroleum-derived detergents. Having been an active researcher in self-cleaning nanotechnology since 2002, witnessing a rapidly growing interest in the field of self-cleaning coatings, surfaces and materials from the media, industry, and academia, I felt a compelling need for a book that describes the recent developments and provides a timely account of this topic. Following an invitation from Wiley, I have approached fellow researchers from across the globe, renowned experts in the field, to contribute to this book with their fascinating achievements covering all areas from the basic and fundamental knowledge of the concepts, potential applications, and recent and future development of self-cleaning nanotechnologies, to their potential hazards and environmental impact. The book is divided into four parts, starting with the general concepts of self-cleaning mechanisms covering both hydrophobic and hydrophilic surfaces. This is followed by specific applications of self-cleaning surfaces and coatings, such as cementitious materials, glasses, clay roof tiles, textiles and plastics. The third part describes recent achievements in self-cleaning surfaces, using advanced materials and technologies, such as liquid flame spray, pulsed laser deposition, and layer-by-layer assembly. In the last part, the potential hazards, environmental impact, and limitations of self-cleaning surfaces are discussed toward further development. Many aspects of this book can be used for general public education, further research and development, as well as in the curriculum development of courses in the areas of materials science and engineering, nanotechnology, and textile finishing. I would like to take this opportunity to express my sincere gratitude to all the authors, my PhD student, Dr Wing Sze Tung, and my research assistant, Ms Stephanie Kung. Special thanks are also due to Wiley editorial staff, Ms Emma Strickland, Ms Sarah Tilley, and the editing team. Walid A. Daoud

Part I Concepts of Self-Cleaning Surfaces

1 Superhydrophobicity and Self-Cleaning Paul Roach1 and Neil Shirtcliffe2 1

Institute for Science and Technology in Medicine, Guy Hilton Research Centre, Keele University, UK 2 Faculty of Technology and Bionics, Hochschule Rhein-Waal, Germany

One of the ways that surfaces can be self-cleaning is by repelling water so effectively that water-borne contaminants cannot attach – by being superhydrophobic. This is demonstrated particularly well by the Indian Lotus, Nelumbo nucifera, which has leaves that remain clean in muddy water. The leaves can be cleaned of most things by drops of water, an effect that has been patented and used in technical systems [1].

1.1 1.1.1

Superhydrophobicity Introducing Superhydrophobicity

Superhydrophobicity is where a surface repels water more effectively than any flat surface, R including one of PTFE (Teflon ). This is possible if the surface of a hydrophobic solid is roughened; the liquid/solid interfacial area is increased and the surface energy cost increases. If the roughness is made very large, water drops bounce off the surface and it can become self-cleaning when it is periodically wetted. To understand more about this type of self-cleaning it is necessary to consider how normal surfaces become wetted and become dirty. The effect has been a focus of much recent research and has been reviewed recently [2– 7]. A good mathematical explanation can be found in a recent book chapter by Extrand [8]. Self-Cleaning Materials and Surfaces: A Nanotechnology Approach, First Edition. Edited by Walid A. Daoud. © 2013 John Wiley & Sons, Ltd. Published 2013 by John Wiley & Sons, Ltd.

4

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 1.1 Cross-section of a drop on a flat surface with the contact angle θ . Contact angles also form at the edge of larger pools of water, in tubes, at bubbles on underwater surfaces and any other configuration where a liquid interface meets a solid.

1.1.2

Contact Angles and Wetting

When a liquid rests on a surface the “contact angle” is measured through the droplet between the solid/liquid and liquid/air interfaces. The equilibrium angle that forms is known as Young’s angle after a theory proposed by Young, but not actually formulated in his work [9]. Young’s equation can be considered as a force balance of lateral forces on a contact line. In a perfect system the contact line cannot sustain any lateral force, so will always move to a position where the forces balance. This is achieved mathematically by taking the components of each force in the plane of the surface, at right angles to the contact line, as shown in Figure 1.1. γSG = γSL + γLG cos θ

(1.1)

where γ is the interfacial tension and the subscripts refer to solid, liquid and gas, for example, γ SL is the interfacial tension between solid and liquid. Young’s equation can also be derived from the surface and interfacial energies and their changes. The contact angle is an important measure of the interaction between the three phases, one solid, a liquid and another fluid, which may be a liquid or a gas. For small drops on a flat surface the drops form spherical caps, spheres intersecting the surface. External factors, such as electric fields, may also influence the drop shape, with gravity playing a role in distorting larger droplets. At the contact line the angle tends to the Young angle except when the contact line is moving relatively rapidly. In most systems there is a certain uncertainty in contact angle known as contact angle hysteresis. 1.1.3

Contact Angle Hysteresis

In practice the equilibrium angle is often difficult to measure because there are a small range of angles on every surface that are stable. These are often described as local energy minima close to the global energy minimum. In practice the contact line therefore often behaves as though it were fixed over a small range of angles close to the equilibrium angle [10]. Traditionally, the equilibrium contact angle was approached by vibrating the surface to supply the energy for the drop to escape the local minima. Although the static angle can

Superhydrophobicity and Self-Cleaning

5

Figure 1.2 A drop on a vertical surface sliding slowly with advancing angle at the front and receding angle at the back, in practice geometrical factors and speed of movement will change the angles away from the actual advancing and receding angles.

vary, the contact line begins to move at a certain angle when the liquid front is advanced and at a different angle when it recedes. These values are simpler to measure so it is often the greatest stable angle and the lowest stable angle that are measured, known as the advancing and receding angles. The angles commonly quoted are those measured at a very low speed as the measured angles are affected by the speed of motion of the contact line. This is usually carried out by injecting liquid slowly into a drop and removing it again. Often the advancing and receding angles are of more practical use than the equilibrium angle, although the equilibrium value can be used to derive surface energies. It is sometimes possible to determine the equilibrium angle if both advancing and receding angles are measured. This still assumes that hysteresis is not very large and the surface is reasonably flat [11]. The difference between the advancing and receding angles, or rather the difference between the cosines of the angles governs whether liquids will stick to a surface or slide or fall off. A drop on a vertical sheet can have the advancing angle at the bottom and the receding angle at the top without moving (Figure 1.2). Surfaces with low hysteresis allow drops to slide over them whatever the equilibrium contact angle. The energy required for a drop to move can be calculated as [12], F=

1 γLG (cos θrec − cos θadv ) 2π r

(1.2)

6

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

where r is the base radius of the drop. The contact angle itself enters the equation in two ways: first the cosine function enhances differences near 90◦ ; secondly the value of the contact radius r, for a given volume depends upon the contact angle. Furmidge calculated the angle of tilt, α, required in order for a drop to slide [13], mg sin α (1.3) = γLG (cos θrec − cos θadv ) w where w is the width of the drop. Measurement of the force required to remove drops from surfaces and tilting angles shows the general trend is correct but some differences can be measured, particularly for softer surfaces. Going back to Young’s equation, if the force balance approach is used, the surface tension components in the plane of the solid are balanced to give the contact angle, but this leaves a vertical force on the surface, depending upon contact angle. Theories by de Gennes and Shanahan [14] and experiments on soft materials suggest that this force distorts the surface, generating a ring like an atoll around the base of the drop and increasing the force restraining the drop from sliding on the surface. Of course the drop profile is also far from a circle if hysteresis is significant, particularly for large drops (for example that shown in Figure 1.2). The receding angle (and liquid properties) controls whether a drop falls off an inverted surface, the advancing angle is not involved as it is never reached in this case. The work needed to pull a liquid from a surface has been reported to be determined by [15, 16]. W = γLG (1 + cos θ R) 1.1.4 1.1.4.1

(1.4)

The Effect of Roughness on Contact Angles Fully Wet Surfaces; Wenzel’s Equation

As the roughness is increased the water initially wets the entire surface, as shown in Figure 1.3b, the increasing surface area of the interface means that the advancing contact angle on a surface with a flat contact angle of greater than 90◦ increases, whereas that of one below 90◦ decreases. A surface with exactly 90◦ contact angle would show no effect of roughness. This type of wetting can, therefore, be considered to be an amplification of the properties of the surface by the roughness. The contact angle of a rough surface of this type can be calculated using Wenzel’s equation [17], which modifies the cosine of the angle by the specific surface area, r , the amount of times the surface is larger than a flat surface of the same size. The subscript e has been used to highlight that usually the equilibrium contact angle is considered as opposed to the receding or advancing contact angles introduced in Section 1.1.3. cos θrough = r cos θe

(1.5)

The amplification of both hydrophilicity and hydrophobicity arises from the change in sign of cosθ at 90◦ . 1.1.4.2

Bridging the Roughness; Cassie and Baxter’s Equation

If the surface is roughened it eventually becomes energetically favourable for the liquid to sit on the top of the roughness and reduce the area of the interface, as shown in Figure 1.3c.

Superhydrophobicity and Self-Cleaning

7

Figure 1.3 Wetting on flat and rough surfaces: (a) flat, (b) rough, Wenzel case; (c) Cassie and Baxter case.

In this case the state approaches that of a liquid on a flat surface with domains of different contact angles but where one of the materials is the second fluid (in this example air). The simplest expression for the contact angle on a surface of this type was formulated in 1944 by Cassie and Baxter [18]. This considers the cosine of the angle to be the mean of the cosines of both contributing surfaces weighted by their relative areas, denoted by f, the fraction of the interface that is solid. cos θrough = f cos θe + ( f − 1)

(1.6)

This equation considers both the solid/liquid and the liquid/gas interfaces to be planar, which is only the case if the surface consists of equal height flat-topped pillars. The original Cassie and Baxter paper allowed for deviations from this by effectively using Wenzel’s equation for the wetted part and allowing changes in the effective roughness with penetration. The main problem with this approach is that it is often difficult to determine where the liquid/solid interface lies.

8

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

1.1.5

Where the Equations Come From

Both Wenzel and Cassie–Baxter equations can be derived from forces at the contact line or from interfacial areas. Using the interfacial areas effectively considers a minimisation of the surface energy of the system. A force balance argument is equivalent, but considers the surface energy from the forces it generates and creates conceptual difficulties when sharp corners are considered [19]. Because of the increased interfacial area in the Wenzel case and the decrease in interfacial area in the Cassie–Baxter case the hysteresis observed increases in the Wenzel state and decreases in the Cassie–Baxter state, giving rise to low water adhesion in the Cassie–Baxter state [20]. 1.1.5.1

Flat Surfaces

Consider a liquid on a surface with a contact line at a contact angle; if we allow this line to move by an infinitesimal amount and assume that it will move in this manner until it reaches an energy minimum the energy minimum can be defined as the position where moving the contact line by a small amount does not change the interfacial energy. This does assume that there is a single minimum in the energy profile – a reasonable assumption for a flat surface. The energy change for moving forward a small amount is illustrated in Figure 1.4. The area of the liquid/fluid interface changes by A cos θ , the solid liquid interface changes area by A and replaces or is replaced by the same amount of solid surface (depending on the direction of motion). The total change in surface free energy, F, accompanying an advance of the contact line is therefore, F = (γSL − γSG ) A + AγLG cos θ

(1.7)

If we set the change in free energy to zero we will find the minimum or maximum of the expression, in this case because we are starting close to the global minimum we will approach that. The result can be rearranged to form Young’s equation.

Figure 1.4 Contact angle and surface free energy.

Superhydrophobicity and Self-Cleaning

9

Figure 1.5 Wenzel wetting.

On a flat surface this treatment is equivalent to a force balance, but on rough surfaces this surface free energy treatment averages over a period of the roughness or a representative area. Unlike a force balance there are no difficulties when the contact line meets the corner of a feature and the intrinsic assumption that the contact line is always on a representative proportion of the surface is slightly more obvious. In cases where this is not true, for patterns that are large compared with the size of the drop, when the contact line can sit on one part of the pattern or when the pattern is anisotropic (e.g., parallel grooves) the approach cannot be applied without some modification. 1.1.5.2

Wenzel Case

For a rough surface where the liquid wets into the rough features (Figure 1.5), the treatment is the same as the flat surface but the surface areas of both the solid/liquid and the solid/vapour interfaces associated with the advance of the contact line are increased by a factor, r, the specific surface area of the rough surface at the contact line. In other words the number of times larger the area is than if it were flat. The roughness factor compares the rough surface to a two-dimensional surface of the same size and is, therefore, better served by this surface energy treatment. When the new values of the surface energies are treated in the same way as before the following expressions result, F = (γSL − γSG ) r A + AγLG cos θ If F = 0 then cos θ =

(γSL − γSG )r A AγLG

(1.8) (1.9)

This can be substituted into Young’s equation to give Wenzel’s equation. 1.1.5.3

Cassie–Baxter Case

To consider only bridging wetting we can imagine flat-topped pillars with water bridging the gaps between with horizontal menisci, as shown in Figure 1.6. In this particular

10

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 1.6 The Cassie and Baxter case.

configuration the surface area of the base of the water is the same as it would be on a flat surface. Again the air/liquid interface at the top of the drop is unaffected by the roughness, the lower part advances over a combination of fluid (air) and solid, the interfacial area, A, can be divided into two components and these assigned to the solid or the fluid interface. The proportions of these two components are determined by the shape of the surface, in this case the relative areas of the tops of the pillars to the gaps. The surface free energy can be minimised as before giving: F = (γSL − γSG ) f A + (1 − f ) AγLG + γLG A cos θ

(1.10)

and, again, with substitution into Young’s equation it becomes reduced to the form of Cassie and Baxter’s equation; cos θrough = f cos θe + ( f − 1)

(1.11)

It can be seen that the observed contact angle on this type of surface is intermediate between the liquid/solid contact angle and the liquid/fluid contact angle. If the second fluid is air or another gas the contact angle will always increase, even if the surface is hydrophilic. The reverse situation can be imagined where the pores at the surface are pre-filled with the same liquid as the drop, in this case the contact angle will decrease, even if the surface is repellent to the liquid. On hydrophilic surfaces this situation can arise when a film of liquid spreads through the roughness of the surface before the macroscopic drop spreads. The same equation can be used for flat surfaces with areas of different contact angle as long as they are distributed well. As mentioned above the original Cassie and Baxter paper considered the combined effect of these two situations.

Superhydrophobicity and Self-Cleaning

1.1.5.4

11

Important Considerations

There has been some criticism of these equations, but these can also be interpreted as criticism of their misuse [21–24]. Both equations require a set of assumptions to be true (or at least locally true) for them to apply. First, there is a requirement that the pattern of roughness or chemistry is arranged so that the contact line is always on the average of all parts of the structures. This is implicit in the treatments above where always an entire cycle of roughness (or chemical pattern) is taken. This requirement is broken if the pattern allows the contact line to arrange itself so that it is mostly on one type of the surface. This is particularly evident in grooved surfaces where the contact angles parallel and perpendicular to the grooves are different. Perpendicular to the grooves the expected angles form, whilst parallel to the groove direction a cyclic change is observed as the contact line moves over the peaks and troughs. Similar problems arise from other pattern geometries. Another way this requirement can be broken is if the size of the patterned features becomes large enough such that the contact line bends to reduce the interfacial energy of the liquid.  γLG (1.12) κ −1 = ρg The capillary length (Eq. 1.12) describes the general size where gravity will have a larger effect than surface tension on a drop of liquid. As can be seen the quantities compared are the surface tension (γLG ) and the effect of gravity on the liquid through density(ρ) and gravity (g). For water on our planet this critical length is 2.73 mm; drops of radius much smaller than this, typically a tenth of this size, are almost spherical. In the same way the meniscus bridging two features will be distorted by gravity and this can be considered to become significant as the gaps reach a tenth of the capillary length. Structures larger than this can distort the contact line as they influence it via interfacial tension. Secondly, as the thought experiment that generates the equations considers small movements from the equilibrium position the state of a liquid is only determined by the surface near the contact line. This is a long-winded way of stating that a drop on a hydrophobic surface will not spontaneously jump to a hydrophilic surface unless the contact line intersects both surfaces. It means that the solid/liquid interfacial area under the drop but away from the contact line is largely irrelevant when determining the contact angle, but if there are differences these will be revealed if the contact line moves over the surface – if a drop slides over the surface for example. 1.1.6

Which State Does a Drop Move Into?

As the Wenzel type of wetting is very different from Cassie and Baxter bridging wetting it is important to know which surface will end up in which state. Initial attempts to predict which state a surface would go into from Cassie–Baxter and Wenzel’s equations met with mixed success. Even when the surface allows this type of comparison it is only possible from the equations above to find which of the states has the lowest energy minimum. Some theoretical treatments of the transition do exist and have shown success predicting experiments. In the simplest the energy levels of both states and those of intermediate states are calculated to determine which states are lowest in energy. More complicated ones attempt to discover when a water drop on the surface can become trapped in one

12

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

state or the other. In many experimental cases the small-scale roughness of the surface is difficult to measure, preventing this type of detailed calculation [25–27]. If a drop is placed onto the surface it is likely to start in the Cassie–Baxer state and may become trapped there even if the Wenzel state has a lower energy. Conversely, if water condenses onto a superhydrophobic surface it initially wets inside the roughness so generally starts in the Wenzel state and almost always becomes trapped there [28].

1.2 1.2.1

Self-Cleaning on Superhydrophobic Surfaces Mechanisms of Self-Cleaning on Superhydrophobic Surfaces

Self-cleaning superhydrophobic surfaces first received attention when a paper was published on the Lotus leaf [29]. Lotus leaves remain clean in muddy water because of the way their surfaces are structured and water repellent. The leaves are strongly superhydrophobic and, although they collect particles of dust, they are fully cleaned by rain. One of the mechanisms for self-cleaning, and that initially suggested for the lotus leaf, depends on how the water drop moves. A drop on a surface with high contact angle and low contact angle hysteresis, usually a bridging super-hydrophobic surface, can roll instead of sliding. This type of motion allows the drop to collect more of the particles at the surface of the solid compared to the usual sliding mechanism. The question that then arises is why a rolling drop should collect hydrophobic particles from a superhydrophobic surface. Particles, even hydrophobic ones, are strongly attached to a liquid/gas interface. If the particle is modelled as a sphere its lowest energy configuration is when it is located in the interface; partially immersed so that the local contact angle can be the equilibrium contact angle. The energy of attachment of a particle on a liquid interface can be calculated by comparing the surface energies of three possibilities, the particle away from the liquid, the particle at its equilibrium position in the interface and the particle inside the liquid. For a hydrophobic particle and water the third case will not be the lowest in energy so we can consider the energy change from the particle resting in air to being held at the interface. When a spherical particle of radius R and contact angle θe attaches to a liquid interface the angle between the surfaces is the contact angle (Figure 1.7). The area of the sphere that becomes wetted can be described by Eq. (1.13), this is both the solid/gas interface that is lost and the solid/liquid interface gained. The liquid also loses some interface, the circular

Figure 1.7 A hydrophobic, spherical particle moving from the air to a position in a water interface where it has its contact angle with the liquid.

Superhydrophobicity and Self-Cleaning

13

patch that is now covered by the particle, given by Eq. (1.14) (Rs being the radius of the sphere). 2π R2S (1 + cos θe )

(1.13)

π R2S sin2 θe

(1.14)

The change in surface energy can therefore be calculated as: F = 2π R2S (1 + cos θe ) (γ SL − γSV ) − π R2S sin2 θe γLG

(1.15)

Substituting with Young’s equation, Eq. (1.1) gives F = −γLG π R2S (1 + cos θe )2

(1.16)

As can be seen from the equation unless the equilibrium contact angle is 180◦ or 0◦ for a particle moving into the liquid it is always energetically favourable to attach a sphere to the interface. Very small particles may obtain enough energy from Brownian motion to escape. As the mass increases with radius cubed and the surface energy with radius squared, large particles can eventually become heavy enough to detach by gravity. This explains why most particles should adhere to a passing droplet, but not why a rolling drop should be more efficient at removing them. Examination of the rear edge of the drop as it pulls off the surface reveals some of the possible mechanisms for self-cleaning. On a flat or a rough Wenzel-type surface the liquid wets the whole surface and the contact line slides over it as it retreats. As the line reaches a particle at the surface it moves over the particle, exerting little or no upward force as it is pinned on the surface to both sides of the particle (Figure 1.8a). On a bridging Cassie–Baxter surface when the contact line reaches the particle it can detach from the features around the particle but remain attached to it as it is a little higher. This allows considerable upward force to be exerted on the particle by the liquid, which could dislodge it (Figure 1.8b). If we consider that a thin film may be left on the surface after the drop has passed this alters the situation a little. On a flat or Wenzel surface the interfaces of the drop are being lost at its rear and regenerated at the front, like a slug leaving a trail (Figure 1.8c). In this case any particles in the upper or lower interface will be dumped back onto the surface when the film evaporates, unless there is a very large flow carrying the particles away. Therefore, nearly all particles will return to their starting positions after the drop has passed. For the Cassie–Baxter bridging case leaving a water film, each interacting peak will spawn a tiny droplet as the main drop passes. This means that particles close to the peaks may not be carried away, but those further away will be plucked out of the surface as in the previous example [30, 31]. In most cases the hydrophobicity of the particle and the surface will prevent water from penetrating between the particle and surface. As the contact line recedes these particles can also be removed by attachment to the drop as liquid will in this case not be left on the surface (Figure 1.8d). The adhesion between the particle and the solid surface can be a direct adhesion, in which case the surface energies of the two solids are high so bringing them together reduces the

14

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 1.8 Contact lines receding over different surface types and self-cleaning: (a) Receding into the plane of the page on a flat surface, dirt is trapped; (b) receding to left on Cassie–Baxter surface, surface tension removing particle; (c) receding to left, flat with precursor film, particles in both interfaces remain; (d) receding to left, Cassie–Baxter with film, locally same as (a) but drops/small particles left on protrusions.

global surface energy. Alternatively, two hydrophobic surfaces can adhere by weaker van der Waals interactions, but if water is present they will be held together by hydrophobic interactions because separation requires wetting of the two interfaces, which would cost surface energy. For a typical superhydrophobic surface the base material is hydrophobic, meaning that hydrophobic interactions will be important. In this case, as shown in Figure 1.8b, the water would not be expected to wet the crack between the particle and the surface, making removal by the contact line most effective for rolling drops. A second factor in the removal of particulate material from a roughened surface is the reduction in solid/solid interfacial area. The particles sit on top of small-scale roughness and are not bound strongly because they do not contact a large surface area. The multilayer roughness of the Lotus leaf is important here, the smaller scale roughness prevents particles nesting into crevices and having larger contact areas than on a flat surface. The third factor is impacting drops – if a surface has different scales of roughness and the instantaneous pressure of the drop impact is only sufficient for it to enter the larger scale of roughness it can collect particles from the crevices of the larger scale roughness.

Superhydrophobicity and Self-Cleaning

1.2.2 1.2.2.1

15

Other Factors Water Impact

In most cases the water impinging on a self-cleaning superhydrophobic surface will have some momentum. It will either be raindrops or will come from a spray of some kind. In this case there will be a short-lived pressure wave that will push the water into the surface. This could convert the wetting state from bridging Cassie–Baxter to fully wetting Wenzel and, therefore, allow the water to adhere strongly. If we model the meniscus between some pillars as a vertical capillary we can calculate the pressure required to force liquid to the base as the Laplace pressure.   1 1 (1.17) p = γLG + R1 R2 The radii required are those of a sphere that forms the advancing angle at the surface of the capillary. For a circular capillary (pore) R = R1 = R2 and can be calculated from, R=

r cos θe

(1.18)

where r is the pore radius, giving p =

2γLG cos θe r

(1.19)

For a superhydrophobic surface with vertical pillars this penetration resistance pressure can be calculated using the size of the gaps. For multiple layers of roughness the equations above can be used to calculate the effective contact angle on the sides of the pillars. As expected, and as can be seen from the formula, increasing the contact angle and decreasing the pattern size improves the pressure resistance of superhydrophobic surfaces. A raindrop falling on the surface of a roof will be around 5 mm in diameter and hit the surface at around 9 m s−1 . The instantaneous impact pressure has been shown to be a maximum around the periphery of the impact zone, the worst possible case if this area transitions to Wenzel wetting and becomes high hysteresis as the whole drop will then become stuck, the peak pressure is of the order of 4 MPa. For a circular capillary with a contact angle of 180◦ , using water at standard temperature, the capillary would have to be less than 350 nm in diameter to prevent water from being forced inside. An equivalent triangular lattice of pillars would have a separation of around 300 nm. Using a more realistic contact angle of 100◦ reduces the critical size to 50 nm and allowing a safety margin reduces it still further. This suggests that the minimum feature size for practical self-cleaning surfaces is quite small and the features must, of course, be able to withstand high impact pressures without damage. 1.2.2.2

Condensation

A further complication is that of condensation inside the roughness. As the surface becomes colder than the air, water condenses directly onto it. In this case it starts in a fully wetting state and often nucleates initially at the base of any roughness, leading to filled patterns, Wenzel wetting and high contact angle hysteresis [32, 33].

16

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

This problem is difficult to avoid; theoretically increasing the roughness above a certain level can make the Wenzel wetting state energetically unfavourable, but partial wetting states can still occur, making water adhere strongly to the surface. Very high structures of very small size are fragile. Even with very small features and high aspect ratios it is still possible to become trapped in the Wenzel state. Despite this it has been shown that hierarchically structured surfaces are relatively stable against condensation. Surfaces consisting of layers of fibres are particularly effective as the heat transfer to the substrate is low and water nucleates at the top leaving bridging so the drop can recover. This type of surface is, however, not very good at self-cleaning if the dirt particles penetrate between the fibres and is more suited to water repellence applications. Typically, the Wenzel state and the Cassie–Baxter state of a very rough surface will represent separate energy minima separated by an energy barrier. The energy barrier is present because most partially filled states are higher in energy. The energy barrier can become large, trapping liquid in one state or the other, which becomes a problem if any ever enters the Wenzel fully wetting state as it is then difficult to remove. Also worth mentioning at this juncture is the Stenocara beetle, which uses local hydrophilic patches to direct condensation, allowing the superhydrophobic part of the surface to remain dry, causing drops of water to grow until the patches cannot hold them, and they then roll to the beetle’s mouth [34]. 1.2.2.3

Oil Contamination

Surfactants and oils are a serious problem for superhydrophobic self-cleaning surfaces. Oils have relatively low surface tension and are, therefore, more difficult to suspend in a bridging state than water. Likewise, the addition of surfactants to water can reduce the surface tension and therefore the pressure required to penetrate between the features of the roughness, and many surfactants and oils have a high vapour pressure so will not evaporate under normal conditions, making them very difficult to remove [35, 36]. On alkane-based supehydrophobic surfaces oils will super-spread. Their contact angles on flat alkane surfaces are low, so when roughened they decrease to zero; it becomes energetically favourable to cover them with a layer of oil. This is particularly challenging for surfaces that are expected to come into contact with oil, such as vehicle parts and kitchen surfaces. The only solution to this is to use fluorocarbon surfaces with very low surface energies, therefore generating reasonable contact angles with both oils and surfactants, and to enhance this by using highly undercut features to allow liquid to become suspended in the Cassie–Baxter bridging state for contact angles below 90◦ . This is very successful, but is unlikely to allow technical surfaces to repel oils very effectively due to their cost of fabrication and the fact that once wet with oil through pressure or heat such surfaces are very difficult to clean as the Wenzel state on these surfaces still has a lower energy than the bridging state. 1.2.2.4

Multiple Scale Roughness (Hierarchical Roughness)

The most common way to generate an effective superhydrophobic surface is to use multiscale roughness. If the smaller scale is small enough to prevent pressure-related penetration

Superhydrophobicity and Self-Cleaning

17

and the larger scale aids rolling, then the surface becomes much more effective than one with a single roughness scale but higher roughness. There are many benefits of using multiple scales. First, low levels of roughness on several length scales affect each other. Several levels of low angle slopes combined generate overhanging structures, which are particularly effective at promoting water bridging and Cassie–Baxter superhydrophobicity. In this case low levels of roughness can be added together to produce effective superhydrophobic surfaces [37]. This is useful because lower peak sharpness improves the resistance of the structure against friction. Secondly, in the case of waviness the levels of roughness interact to generate steeper pitches for the liquid to interact with. The obvious example of this was proposed by Herminghaus [38] where a sine wave in two dimensions generates a wavy surface but two overlaid sine waves of very different frequencies generate very steep roughness indeed, even if the amplitude of each one is not that great. In fact the use of multiple layers of roughness is a simple way of generating overhanging roughness. Overhanging roughness can suspend liquids even for contact angles below 90◦ in a local energy minimum [39]. Thirdly, as mentioned above, multiple level roughness is particularly resistant to pressure wave wetting as well as being more resistant to low surface tension liquids, such as alcohols, and to dew formation [40, 41]. 1.2.3

Nature’s Answers

Superhydrophobic self-cleaning was first observed on the leaves of the (Indian) Lotus, Nelumbo nucifera, which show small wax crystals on the top of waxy bumps. These are highly efficient at self-cleaning and repel a range of water-based liquids, but are sensitive to condensation and physical damage. The wax self-organises on the surface to form nanostructures and microstructures on top of microstructures formed from other components of the leaves in different ways. As described by Koch et al. the typical morphologies of these waxes are tubes, prisms and flakes [42]. The self-organised growth of waxes on the leaves means that the plant only needs to exude the waxy mixture and damage to the structures will tend to repair itself [43]. This is a major advantage over artificial surfaces that wear away and become less effective over time. Both the chemistry of the surface and the shape must be maintained to preserve a superhydrophobic effect and waxes are good at both, being able to reorder when warm and dry to hide hydrophilic groups and regenerate surface structure. It has only been possible to copy this in a limited manner so far. Some other plants and many animals use hair-like structures to generate the roughness required to repel water. This strategy is very effective against condensing water as the water nucleates among the hairs so does not reach the surface. The well known plant Lady’s mantle, Alchemilla mollis, is thought to use this to collect dew drops on its leaves and allow them to flow to its base. Diving beetles and spiders use a combination of large hairs (setae) and smaller ones (microtrichia). The smaller ones are arranged in a denser array should the larger ones fail for any reason. The hair-like structures are particularly suited to generating persistent air layers underwater, as required by these animals to breathe. The structures have a relatively large area at the interface and a lower area at any other point, making the

18

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 1.9 Superhydrophobic structures on various organisms: (a) wax on cabbage, (b) wax on Lotus leaves, (c) bent “hairs” on water boatman Notonecta glauca, (d) Alchemilla Mollis (Ladies Mantle). Source: (a) Reproduced with permission from [Ditsche-Kuru et al. Beilstein J Nanotech. 2011, 2, 137–144] Copyright (2011) Beilstein, (b) Reproduced with permission from [ ] Copyright (2009) American Chemical Society, (c) Reproduced with permission from [Ditsche-Kuru et al., Beilstein J Nanotech. 2011, 2, 137–144] Copyright (2011) Beilstein, open access, (d) Reproduced with permission from [Shirtcliffe et al., Langmuir 2009, 25(24), 14121–14128] Copyright (2009) American Chemical Society.

surface energy minimum deep and therefore stable. Most aquatic insects achieve this by having structures that come out from their body and then all turn to parallel at the interface, as shown in Figure 1.9c. The second advantage of this system is that there is a degree of elasticity in the structures themselves, so when a pressure wave passes they can move with it and the liquid interface remains attached without slipping deeper into the structure or ejecting gas bubbles [44]. So far these structures have been used less for self-cleaning purposes. Drops become attached to crossing hairs, because the crossing point represents a minimum in energy, and then cannot move very easily. On water insects the structures do not cross because they are all aligned in the same direction so rapid drop movement in one direction and self-cleaning is possible, although if particles penetrated between the structures they would

Superhydrophobicity and Self-Cleaning

19

become difficult to remove as water does not penetrate under normal conditions, unlike on the heirarchical Lotus leaf. Moth’s eyes are structured in a semi-ordered fashion, their primary functions being optical, light gathering and anti-reflection. For these purposes the moth’s eyes are structured with conical projections in a hexagonal array. In order for them not to scatter light or to act as lenses they are smaller than the wavelengths of visible light. This generates a gradient of effective density from the eye to the air and results in very low reflection. There are two huge advantages for the moth, predators do not see the reflection and more light is gathered, helping them to see at night. This type of structure has raised a great deal of interest and ideally would be combined with a self-cleaning function. The obvious use of such a surface would be as a coating on solar cells to reduce scattering and enable self-cleaning. Some artificial surfaces of this type have been designed and show promise although, as with the moth’s eyes, the self-cleaning cannot easily be optimised at the same time as the anti-reflectance [45, 46]. Cicadas and other insects use similar structures to repel water from their wings, but they are not as effective as multiscale structures and are unlikely to be self-cleaning without a supporting very small-scale pattern to prevent impinging drops from becoming stuck in the Wenzel state. On the other hand the spiny structures on the feet of water striders for example, Gerris remingis, contain nanogrooves and remain dry even after long periods in contact with water. The water strider uses them to glide over the interfacial tension of water bodies, allowing it to collect insects without superhydrophobic surfaces that become captured by it [47]. Butterfly wings also repel water effectively; their wing scales are structured and canted to provide a saw-tooth surface. Experiments show that these structures have a preferred direction for water to move off them [48]. This effect is of potential use in a variety of applications because the pattern can be continued over large areas whereas a pattern of decreasing contact angles could not be. The saw-toothed structure probably generates an uneven contact angle, causing the drop to stick in one direction but roll in the other [49]. The butterfly wing pattern could be used over large areas and drop motion can be driven by gravity or vibration. Geckos also use soft projections on their feet, but these are spread out at the ends so that they can attach to walls. They function as an extremely compliant surface, enabling close contact with smooth and rough surfaces to maximise surface interactions. The link to superhydrophobicity and self-cleaning is that they must remain clean to function and have been shown to have some self-cleaning properties. Despite adhering strongly to surfaces their surface energy is low and particles are easily transferred from them to most surfaces, in other words they can wipe or walk off particles [50, 51]. 1.2.4

Superhydrophilic Self-Cleaning Surfaces

The opposite of superhydrophobicity is superhydrophilicity. In this case the surface energy of the base material is high, causing water to spread out over it, in other words to have a relatively low contact angle. In this case Wenzel’s equation can be used to estimate the contact angle that will be observed on the rough surface. Unlike the superhydrophobic case there is no alternative state that takes over at high roughness, so the predicted contact angle soon reaches 0◦ . Roughness can easily be increased past this point and the question

20

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

then is what will happen. The surface wets fully but the rate of wetting inside the roughness of the surface can be more rapid than on a flat surface. Aqil et al. [52] showed that the rate of spread of a drop is greater on a rough surface and slows down less as it approaches 0◦ . In fact the base of the drop can spread so fast that it breaks away from the top, generating a fried-egg shape. Such surfaces can clean themselves of hydrophobic particles when wet because water enters below the particles and lifts them off as they become attached to the air/water interface. This is also extremely effective against biological contamination but relies on the presence of lots of water [53]. Hydrophilic surfaces are usually strongly adhesive because they have a high surface energy and become coated with material that can bind chemically to them. This can make them very difficult to clean. Materials with a surface energy the same as that of water can remain very clean when wet because it is energetically favourable for water molecules to sit at the surface. An example of this is polyethylene oxide (PEO) which is used to prevent biofouling in many applications for exactly this reason. The surfaces do not self-clean, but do not become fouled in the first place. If surfaces of this type dry out they can become contaminated and often become damaged because they shrink when dry. Some self-cleaning products use a combination of superhydrophilic surface properties along with a photocatalyst. The high roughness and hydrophilicity generate a water film when wet and the photocatalyst oxidises organic species to charged species, making surfactants or ultimately completely oxidising them. This combined system allows the surface energy to be higher than that of water with the photocatalysis removing the film of organic matter that spreads over the surface in dry conditions. When wet the high roughness and surface energy causes a water film to lift off the contaminants.

1.2.5

Functional Properties of Superhydrophobic Surfaces

Various routes exist to fabricate superhydrophobic surfaces, allowing a plethora of different material types to be used. Properties inherent in the chosen material can be utilised, for example electrical conductivity, although many other characteristics can be built into materials through the careful design of structure and chemistry. Over the past years attention has moved towards developing advanced materials with properties suitable for a range of applications; either for research or industrial requirements. A number of commercial products making use of superhydrophobicity are already available on the market, with others currently under development. In addition, many patents have been granted for various possible applications of self-cleaning surfaces [54]. Many of these products are used primarily for their superhydrophobic self-cleaning abilities, such as materials used in construction for windows or roofing tiles. Material coatings used in this industry are also widely available, with paints used to form a barrier against bio-fouling or graffiti-resistant layers. The use of non-wettable textiles is also progressing rapidly, with research geared towards identifying methods to modify conventional fabrics and to generate new materials possessing useful functions. These market-led products are mostly driven by novelty, with some products offering better value over existing counterparts in that they have increased functionality [55].

Superhydrophobicity and Self-Cleaning

1.2.5.1

21

Evaporation and Condensation Resistance

The evaporation of liquid drops from solid surfaces has been thoroughly investigated over past decades, although only a few have looked at the interesting phenomena of evaporation from structured superhydrophobic surfaces. On flat solid surfaces two evaporation models were proposed by Picknett et al. in 1977: a droplet having diminishing surface contact area with a constant contact angle, or having a constant contact area with a reducing contact angle [56]. For volatile liquid droplets resting on low surface energy substrates the former model was found to be followed when the initial contact angle was 90◦ the latter model takes precedence, giving non-linear evaporation rates [57]. Xu et al. showed how the three-phase contact line of a sessile droplet remains pinned on some superhydrophobic surfaces through the whole evaporation process [58], such that the constant area model is followed. Water resting on the surface of a lotus leaf may penetrate into the roughness, interacting strongly with the natural wax and hindering droplet movement. Most studies examining the effects of liquid evaporation focus on large droplets a few mm in diameter, although condensation onto a surface occurs through the interaction of much smaller droplets. Cheng et al. showed experimentally the dramatic difference in wetting characteristics on superhydrophobic lotus leaves, comparing macro-scale water drops to condensation of water from the vapour phase [59]. When wetted with condensing water the high contact angles associated with superhydrophobicity are no longer observed. A more in-depth investigation carried out by McCarthy using topographically patterned silicon pillars as model surfaces gave similar results [60]. During steady condensation, wherein the droplet radius increases proportionally with time, the drop interactions are ultimately important to the dynamics, with droplets coalescing to increase the average droplet size. Baysens et al. studied the growth dynamics of condensing water droplets on model surfaces [61] with the initial nucleating droplets being much smaller than feature sizes and forming in between as if they were on planar surfaces. As the drops grow, their initial areas often remain in the fully wetted Wenzel state but new areas covered by coalesced droplets mostly rest in the Cassie–Baxter regime. In their particular case after a very short time (∼1 s) the water was found to move from the slightly wetted features to the more stable Wenzel state, Figure 1.10. Surfaces that go into the Cassie–Baxter state have been proposed for use in condensers because a layer of condensed water at the surface of the cooling pipes is often the limiting factor in condenser design. In the other direction, surfaces that are used for boiling often generate a layer of water vapour at their surface and superhydrophilic surfaces have been suggested for use here to rewet as soon as possible after a bubble forms. 1.2.5.2

Frost/Ice Resistance

Frosting of surfaces can be a serious problem due to changes in surface properties or accumulation of ice or snow. Situations involving reduced friction due to frosting are particularly dangerous, as well as icing of overhead power cables or aeroplane wings, which can result in the failure of electrical insulation or mechanical properties. Bridges are often closed due to the danger of ice falling from supports. Some superhydrophobic surfaces have, however, been shown to reduce freezing temperatures, reduce ice bond strength and even prevent ice formation at their surface [62–64]. Frost formation occurs via two main processes: nucleation and crystal growth, being dependent upon the water

22

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 1.10 Condensation within structured surfaces [61]. Source: Reproduced with permission from [Narhe, Langmuir 2007, 23, 6486] Copyright (2007) American Chemical Society.

vapour overcoming a Gibbs free-energy barrier [65]. Freezing of water has been shown to be significantly delayed when depositing droplets onto cooled superhydrophobic surfaces [66]. Microtextures were found to delay freezing with drops rolling off without leaving behind a film of ice or freezing. Control samples using flat copper showed that slower moving droplets leave a film as they move, which freezes immediately. The low contact area and air gap is thought to reduce the transfer of heat from the liquid to the surface, preventing

Superhydrophobicity and Self-Cleaning

23

freezing. In attempts to mimic natural ice build-up researchers have prepared glazed ice surfaces by spraying super-cooled microdroplets of water onto surfaces held below 0 ◦ C. Kulinich et al. investigated the effects of flat versus rough hydrophobic surfaces on the adhesion strength of such ice coatings [67, 68]. Comparison of surfaces based on their contact angle hysteresis showed a direct relationship with adhesion strength. The size of features presented by superhydrophobic surfaces has also been shown to play a role in anti-icing. Using particles with diameters in the range 20 nm to 20 μm. Gao et al. showed that larger features gave rise to an increase in the probability of ice formation [69]. Clearly, control over frosting processes is achievable through the use of structured hydrophobic materials. In some cases the superhydrophobic surfaces became damaged by freezing cycles, but a great deal of research is being invested in this area as the potential for ice resistant structures and vehicles is very attractive. 1.2.5.3

Anti-Fouling

Biofouling is the accumulation of biological material on surfaces. Biofouling is a vast area, spanning from problems relating to underwater structures, such as oil-rigs and ship hulls which suffer damage and increased operational and maintenance costs due to accumulation of biomatter, to the medical industry wherein avoiding even small amounts of contamination is paramount. The use of superhydrophobic materials to reduce fouling has been demonstrated through a number of investigations. The main advantage of this approach is that a persistent effect would be expected without releasing chemicals into the environment. A reduction in the surface area in contact with the liquid carrying the contaminants is often accredited to a reduction in the level of fouling [70], although the shear forces involved in liquid flow and the chemistry presented at the surface have also been shown to impact on protein adsorption/desorption [71]. Koc et al. showed that a reduction in the size of features presented at a surface, from micro- to nano-scale, drastically reduced the amount of protein adsorbed, particularly under flow conditions. An initial increase in protein adsorption was observed compared to flat surfaces of the same chemistry, possibly due to slight ingress of the surfactant protein solution into the pores. Others have shown that antifouling can be achieved during short-term exposure, but long term exposure, as would be experienced in a marine environment, gave rise to a gradual deterioration of superhydrophobic surfaces due to fouling [72]. In some cases a similar effect is used where the scale of roughness is set so that the attachment points of a target organism cannot fit the surface and the surface is usually hydrophilic so that water squeezes into the interface and excludes other molecules. This approach is used by some marine organisms [73]. 1.2.5.4

Anti-Corrosion

As much as antifouling is important for materials in contact with water, superhydrophobic coatings also present opportunities for use as anti-corrosion treatment. Metals in particular are vulnerable to oxidation, leading to degradation of their mechanical properties. By application of water-repellent materials to their surfaces such breakdown of the metal can

24

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

be slowed down or even prevented altogether by preventing the formation of electrochemical cells [74]. Ishizaki et al. demonstrated the use of nanostructured cerium oxide films functionalised by fluoroalkylsilanes to protect a magnesium alloy from a corrosive salt solution [75]. The superhydrophobic materials were found to be extremely corrosion resistant, with the adhesion of the overlayer also being shown to be good. Some coatings, however, are not as well adhered to the underlying substrate. Yuan et al. demonstrated the use of vinyl modification of copper substrates, from which a fluoropolymer was grafted to form a highly non-wettable and anti-corrosive layer [76]. The copper was first etched to roughen it, with a fluorosilane then being used to form a conformal hydrophobic coating. Electrochemical methods are often used, either to initially roughen surfaces or to deposit coatings. Xu et al. used a facile electrochemical machining process to hydrophobise a magnesium alloy, again using a fluorosilane as the chemical modifier [77]. Electrochemical deposition of polypyrrole has been shown to be a rapid and effective means to fabricate relatively good non-wettable surfaces [78]. Within 3 s the zinc could be modified with a 2 μm thick hydrophobic coating having a contact angle of 125◦ . Super water-repellent coatings are used in this capacity to exclude liquid water from the metal interface, they have shown positive effect in simple tests, but are seldom suggested as the final external coat due to fragility problems. 1.2.5.5

Transparent and Anti-Reflective Properties

Anti-reflective or optically transparent properties, such as those described for the moth eyes, are sought after in many material and device applications, such as electronic devices, sight correction and sun glasses, lenses and mirrors. Although not a requirement, for many of these possible applications, the ability to self-clean is also a useful capability. One example of particular relevance in the current environmentally-friendly era is the use of transparent, self-cleaning coatings on solar panels. These are not always accessible after installation with maintenance being kept to a minimum by correct selection of surface properties to remain free of dust and to transfer the maximum amount of light to the photocell. The roughness of the surface is key in terms of the anti-reflection properties as scattering is enhanced by roughness, therefore reducing transparency. For this reason the features required for anti-reflection must be kept smaller than the wavelength of the light that is to be transmitted. When considering visible light this means features should be less than 380–750 nm. Many routes to investigate these properties make use of silica/silicon nanomaterials, with silica nanoparticles being used to impart controlled film thickness and roughness, whilst also allowing ease of chemical modification via silane functionalisation [79, 80]. Silica nanoparticles can be applied as coatings by using charged polymers to attract particle layers. This has been demonstrated by Bravo et al. forming transparent films of 20–50 nm silica particles using layer-by-layer deposition of poly(sodium 4styrenesulfonate) (PSS) and poly(allylamine hydrochloride) (PAH) [81]. By controlling aggregation and layering of the particle layers almost complete visible transparency was achieved, with further chemical modification rendering the surfaces superhydrophobic, Figure 1.11. In most cases anti-reflection is the main function of the layer with water repellency and then self-cleaning being fringe benefits.

Superhydrophobicity and Self-Cleaning

25

Figure 1.11 Transparent silica nanoparticle superhydrophobic surface [81]. Source: Reproduced with permission from [Bravo, Langmuir 2007, 23, 7293–7298] Copyright (2007) American Chemical Society.

These silica nanoparticle surfaces have been shown to demonstrate excellent water repellency as well as antireflective characteristics, and have experimentally shown selfcleaning for 2000 h of outdoor exposure [82]. Polymer coatings have distinct benefits over their inorganic counterparts, being largely flexible and having controllable toughness and stiffness. Researchers have reported an inexpensive method to produce superhydrophobic polymer coatings through in situ polymerisation of various monomers [83]. Phase separation of the polymerising solution gave rise to a network of pores in the hydrophobic polymer, giving rise to its non-wettable nature. Altering the composition of the monomer mixture gave control over surface morphology from nano- to micro-roughness, whilst photografting was used to adjust the surface chemistry. Transparent films were afforded by a reduction in the architecture to nanoscale features.

1.3

Materials and Fabrication

As discussed above there are many examples of superhydrophobicity in nature, having wide-ranging applications from self-cleaning to liquid harvesting. Synthetic methods to mimic such surfaces have been developed making use of many different types of materials, allowing a range of properties of the final surface. The required characteristics are that the surface must be rough, of the order of nano- to microscale and present a hydrophobic chemistry. Dual-scale roughness is advantageous to form surfaces having both high water contact angles and low sliding angles [84]. The material itself could be inherently hydrophobic, with this being roughened to form a simple superhydrophobic surface – such as roughening a poly(tetrafluoroethylene) (PTFE) plate [85]. If the material is not chemically hydrophobic, then a chemical coating is normally added after the surface structure is formed, as in the case of fluoro/alkylsilane treatment often used to form many reported superhydrophobic materials. Other methods produce a coating that presents both the structure and the chemistry required.

26

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Since the early 1940s researchers have been interested in the wetting properties of surfaces, with more and more focus on the applications of such knowledge being brought about by technology and industrial applications of the 1990s. The number of methods reported in the literature to produce materials/ surfaces/ coatings presenting superhydrophobic properties has increased substantially over the past decade, with endless strategies demonstrating potential for scale-up of materials processing. The careful control over etching processes can lead to degradation of polymers such that they are morphologically altered. High-energy oxygen species present in gas-phase plasmas have been used to etch fluorinated polymers to create surfaces with water contact angles of ∼170◦ [86]. Similar methods have been used to roughen many polymers, such as polypropylene [87]. Ellinas et al. used colloidal lithography of polystyrene to mask off a polymethylmethacrylate (PMMA) sheet before etching with an oxygen-plasma, as shown in Figure 1.12 [88]. Polymer microbeads were spin-coated onto the substrate to form highly ordered structures to pattern the surface. A similar approach was also used to selectively etch silicon substrates to form features with very high aspect-ratios.

Figure 1.12 PMMA pillar arrays produced via plasma etching: (a) −60 V bias, 1 min etch; (b) –80 V bias, 1 min etch; (c) −60 V bias, 2 min etch; (d) −80 V bias, 1 min etch. Source: Adapted with permission from [Ellinas, Microelectronic Eng. 2011, 88, 2547–2551] Copyright (2011) Elsevier.

Superhydrophobicity and Self-Cleaning

27

Langmuir–Blodgett deposition of micro- and nano-scale silica particles is also commonly used to present ordered features, with the structures usually being rendered hydrophobic by post modification with a hydrophobic silane [89]. Phase-separation of polymers can lead to the micro-/nano-scale features required for superhydrophobicity. This method of modification is carried out in-situ with a surface coating being formed. Hydrophobic phase-separation of sol–gel based solutions has afforded superhydrophobic materials which display very high water contact-angles [90]. Polymers are more often used, in part due to their flexibility leading to greater durability. A copolymer equimolar in fluorinated acrylate and methyl methacrylate monomers has been shown to produce hexagonally-packed nano-scale pores during drying under strictly controlled humidity conditions [91]. Through fine-tuning of both the size of these pores and the thickness of the deposited film the coating can present superhydrophobicity and be optically transparent. Electrodeposition under diffusion limited conditions is also commonly used for the fabrication of rough structures. Various TiO2 nanostructures can be formed on titanium surfaces. Lai et al. prepared nano-pore, nano-tube and nano-vesuvianite structures, which could be used to alter water contact modes at the nanometre scale [92]. Anodization of aluminum gives rise to porous membranes, which can be further chemically treated to form superhydrophobic surfaces [93]. Many other methods are available making use of a range of materials. Although early work looked mainly at textiles or wax-based materials, more recent work has been driven to examine potential modification strategies for metals. Through a two-pronged approach researchers are looking to develop robust methods to produce durable superhydrophobic properties via cost-effective routes, either by direct roughening of inherently hydrophobic materials or, more widely, by adding an overlayer to alter the outermost layer of the materials. A number of recent reviews cover the fabrication of superhydrophobic materials/ coatings in depth [94–96].

1.4

Future Perspectives

Through our inspiration from nature we now look to mimic the properties of surfaces that surround us in order to improve our innovations and technologies. Over the past decades we have advanced our understanding of the interface between liquids and solids such that we can now design and fabricate surfaces to exhibit specific properties, alone or in combination, such as water repellency and anti-fouling/ anti-corrosion. These have either been driven by fundamental research or the need for new coatings, such as advances in photocells as well as advances in micro-structuring technology. Applications for superhydrophobic materials are widespread, only really limited by the cost and fragility of typical superhydrophobic surfaces. Outdoor weather-proof paints, easy-clean textiles and self-cleaning glass used for windows of apartment or office block skyscrapers are all examples of how superhydrophobicity is used in our surroundings. Superhydrophilicity, the flip side of the coin, is used in self-cleaning windows and roof tiles and also in mist-free mirrors, causing the condensed water to spread to a film and also remove some contamination. Technologies within laboratory settings are also being improved. Lab-on-a-chip devices of interest to a multi-disciplinary audience for the miniaturisation

28

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

and development of high-throughput processes are possibly the next application to be advanced by the incorporation of superhydrophobic surfaces. Clinical diagnostic devices making use of such technologies may benefit from an enhancement in device longevity and sensitivity of analyte detection through the anti-fouling capabilities of superhydrophobic and superhydrophilic surfaces. Our ability to prepare complicated surfaces structured on a micro- and nano-scale has improved massively in the last few years and has enabled various technologies to advance from the laboratory into the real world as their cost has fallen to a level where application is possible. The development and production of both random roughness and more advanced sculptured surfaces has resulted in both simple and advanced self-cleaning surfaces becoming available. The combination of superhydrophobicity and other properties, such as anti-reflection or ice protection, is possible, although the optimum topography and chemistry of the surface will often be specific for the application.

References R 1. Lotus Effect was trademarked in 1998 by W. Barthlott and his colleagues. 2. Roach, P., Shirtcliffe, N. and Newton, M. (2008) Progress in superhydrophobic surface development. Soft Matter, 4, 224–240. 3. Extrand, C. (2006) Encyclopedia of Surface and Colloid Science, 2nd edn (ed. P. Somasundaran), Taylor and Francis, New York, pp. 5854–5868. 4. Bhusan, B. and Jung, Y. (2011) Superhydrophobicity, self-cleaning, low adhesion, and drag reduction. Prog. Mater. Sci., 56, 1–108. 5. Guo, Z., Liu, W. and Su, B. (2011) Superhydrophobic surfaces : From natural to biomimetic to functional. J. Colloid Interface Sci., 353, 335–355. 6. Genzer, J. and Marmur, A. (2008) Biological and synthetic self- cleaning surfaces. MRS Bull., 33, 742–746. 7. Ma, M., Hill, R. and Rutledge, G. (2008) A review of recent results on superhydrophobic materials based on micro- and nanofibers. J. Adhes. Sci. Technol., 22, 1799–1817. 8. Extrand, C. (2006) Encyclopedia of Surface and Colloid Science (ed. P. Somasundaran), Taylor and Francis, New York, pp. 5854–5868. 9. Young, T. (1805) An essay on the cohesion of fluids. Philos. Trans. R. Soc. London, 95, 65. 10. de Gennes, P. (1895) Wetting: Statics and dynamics. Rev. Mod. Phys., 57 (3): 827–863. 11. Fabretto, M., Sedev, R. and Ralston, J. (2003) 3rd International Symposium on Contact Angle, Wettability and Adhesion, vol. 3 (ed. K.L. Mittal, VSP International Science Publishers, pp. 161–173. 12. Olsen, D.A., Joyner, P.A. and Olson, M.D. (1962) Sliding of water drops on microstructured hydrophobic surfaces. J. Phys. Chem., 66 (5), 883–886. 13. Furmidge, C.G.L. (1962) Studies at phase interfaces. I. The sliding of liquid drops on solid surfaces and a theory for spray retention. J. Colloid Sci., 17, 309. 14. Shanahan, M.E.R. and DeGennes, P.G. (1986) The ridge produced by a liquid near the triple line solid liquid fluid. Compt. Rend. Ser. II, 302, 517. 15. Gao, L. and McCarthy, T. (2008) Teflon is hydrophilic. Comments on definitions of hydrophobic, shear versus tensile hydrophobicity, and wettability characterization. Langmuir, 24, 9183–9188.

Superhydrophobicity and Self-Cleaning

29

16. De Souza, E., Gao, L., McCarthy, T. et al. (2008) Effect of contact angle hysteresis on the measurement of capillary forces. Langmuir, 24, 1391–1396. 17. Wenzel, R.N. (1936) Resistance of solid surfaces to wetting by water. Ind. Eng. Chem., 28, 988–994. 18. Cassie, A.B.D. and Baxter, S. (1944) Wettability of porous surfaces. Trans. Faraday Soc., 40, 546. 19. De Gennes, P., Brochard-Wyart, F. and Quere, D. (2003) Capillarity and Wetting Phenomena: Drops, Bubbles, Pearls, Waves, Springer. 20. Quere, D., Lafuma, A. and Bico, J. (2003) Slippy and sticky microtextured solids. J. Nanotech., 14, 10. 21. Gao, L. and McCarthy, T. (2007) How Wenzel and Cassie were wrong. Langmuir, 23 (7), 3762–3765. 22. McHale, G. (2007) Cassie and Wenzel: Were they really so wrong? Langmuir, 23 (15), 8200–8205. 23. Panchagnula, M. and Vedantam, S. (2007) Comment on how Wenzel and Cassie were wrong by Gao and McCarthy. Langmuir, 23 (26), 13242. 24. Gao, L. and McCarthy, T. (2009) An attempt to correct the faulty intuition perpetuated by the Wenzel and Cassie “Laws”. Langmuir, 25 (13), 7249–7255. 25. Quere, D. (2002) Surface chemistry: Fakir droplets. Nature Mater., 1, 14. 26. Patankar, N. (2004) Transition between superhydrophobic states on rough surfaces. Langmuir, 20, 7097. 27. Reyssat, M., Yeomans, J. and Quere, D. (2008) Impalement of fakir drops. Europhys. Lett., 81, art. 26006. 28. Dorrer, C. and Ruhe, J. (2007) Condensation and wetting transitions on microstructured ultrahydrophobic surfaces. Langmuir, 23 (7), 3820–3824. 29. Barthlott, W. and Neinhuis, C. (1997) Purity of the sacred lotus, or escape from contamination in biological surfaces. Planta, 202 (1), 1–8. 30. Krumpfer, J., Bian, P., Zheng, P. et al. (2011) Contact angle hysteresis on superhydrophobic surfaces: An ionic liquid probe fluid offers mechanistic insight. Langmuir, 27 (6), 2166–2169. 31. Krumpfer, J. and McCarthy, T. (2011) Dip-coating crystallization on a superhydrophobic surface: A million mounted crystals in a 1 cm2 array. J. Am. Chem. Soc., 133 (15), 5764–5766. 32. Bico, J., Marzolin, C. and Quere, D. (1999) Pearl drops. Europhys. Lett., 47 (2), 220–226. 33. Mockenhaupt, B., Ensikat, H., Spaeth, M. and Barthlott, W. (2008) Superhydrophobicity of Biological and Technical Surfaces under Moisture Condensation: Stability in Relation to Surface Structure. Langmuir, 24 (23), 13591–13597. 34. Parker, A.R. and Lawrence, C.R. (2001) Water capture by a desert beetle. Nature, 414 (6859), 33–34. 35. Milne, A.J.B., Elliott, J.A.W., Zabeti, P. et al. (2011) Model and experimental studies for contact angles of surfactant solutions on rough and smooth hydrophobic surfaces. Phys. Chem. Chem. Phys., 13, 16208–16219. 36. Ferrari, M., Ravera, F., Rao, S. and Liggieri, L. (2006) Surfactant adsorption at superhydrophobic surfaces. Appl. Phys. Lett., 89, 053104. 37. Shirtcliffe, N., McHale, G., Newton, M. et al. (2004) Dual-scale roughness produces unusually water-repellent surfaces. Adv. Mater., 16 (21), 1929–1932.

30

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

38. Herminghaus, S. (2000) Roughness-induced non-wetting. Europhys. Lett., 52, 165. 39. Tuteja, A., Choi, W., Ma, M. et al. (2007) Designing superoleophobic surfaces. Science, 318, 1618–1622. 40. Cheng, Y., Rodak, D., Wong, C. and Hayden, C. (2006) Effects of micro- and nanostructures on the self-cleaning behaviour of lotus leaves. Nanotechnology, 17 (5), 1359–1362. 41. Nosonovsky, M. (2007) Multiscale roughness and stability of superhydrophobic biomimetic interfaces. Langmuir, 23 (6), 3157–3161. 42. Koch, K., Bohn, F. and Barthlott, W. (2009) Hierarchically sculptured plant surfaces and superhydrophobicity. Langmuir, 25 (24), 14116–14120. 43. Koch, K., Dommisse, A., Niemietz, A. et al. (2009) Nanostructure of epicuticular plant waxes: Self-assembly of wax tubules. Surf. Sci., 603, 1961–1968. 44. Flynn, M.R. and Bush, W.M. (2008) Underwater breathing: The mechanics of plastron respiration. J. Fluid Mech., 608, 275–296. 45. Xiu, Y., Zhang, S., Yelundur, V. et al. (2008) Superhydrophobic and low light reflectivity silicon surfaces fabricated by hierarchical etching. Langmuir, 24, 10421–10426. 46. Park, Y., Im, H., Im, M. and Choi, Y. (2011) Self-cleaning effect of highly waterrepellent microshell structures for solar cell applications. J. Mater. Chem., 21, 633– 636. 47. Gau, X. and Jiang, L. (2004) Biophysics: Water-repellent legs of water striders. Nature, 432, 36. 48. Zheng, Y., Gao, X. and Jiang, L. (2007) Directional adhesion of superhydrophobic butterfly wings. Soft Matter, 3, 178–182. 49. Kusumaatmaja, H. and Yeomans, J. (2009) Anisotropic hysteresis on ratcheted superhydrophobic surfaces. Soft Matter, 5, 2704–2707. 50. Hansen, W.R. (2005) Evidence for self-cleaning in gecko setae. PNAS, 102(2), 385– 389. 51. Autumn, K. and Hansen, W. (2006) Ultrahydrophobicity indicates a non-adhesive default state in gecko setae. J. Comp. Physiol., 192, 1205–1212. 52. McHale, G., Shirtcliffe, N., Aqil, S. et al. (2004) Topography driven spreading. Phys Rev Lett., 93(3),036102. 53. Genzer, J. and Efimienko, K. (2006) Recent developments in superhydrophobic surfaces and their relevance to marine fouling: A review. Biofouling, 22 (5), 339–360. 54. Nosonovsky, M. and Bhushan, B. (2008) Multiscale Dissipative Mechanisms and Hierarchical Surfaces-Friction, Superhydrophobicity, and Biomimetics, Springer, New York. 55. Blossey, R. (2003) Self-cleaning surfaces – virtual realities. Nat. Mater., 2, 301–306. 56. Picknett, R.G. and Bexon, R. (1977) Evaporation of sessile or pendant drops in still air. J. Colloid Interface Sci., 61, 336–350. 57. Birdi, K.S., Vu, D.T. and Winter, A. (1989) A study of the evaporation rates of small water drops placed on a solid-surface. J. Phys. Chem., 93, 3702–3703. 58. Zhang, X.Y., Tan, S.X., Zhao, N. et al. (2006) Evaporation of sessile water droplets on superhydrophobic natural lotus and biomimetic polymer surfaces. ChemPhysChem., 7, 2067–2070. 59. Cheng, Y.T. and Rodak, D.E. (2005) Is the lotus leaf superhydrophobic? Appl. Phys. Lett., 86, 144101.

Superhydrophobicity and Self-Cleaning

31

60. Wier, K.A. and McCarthy, T.J. (2006) Condensation on ultrahydrophobic surfaces and its effect on droplet mobility: Ultrahydrophobic surfaces are not always water repellant. Langmuir, 22, 2433–2436. 61. Narhe, R.D. and Beysens, D.A. (2007) Growth dynamics of water drops on a squarepattern rough hydrophobic surface. Langmuir, 23, 6486–6489. 62. Nakajima, A., Hashimoto, K. and Watanabe, T. (2001) Recent studies on superhydrophobic films. Monatsh. Chem., 132, 31–41. 63. Sarkar, D.K. and Farzaneh, M. (2009) Superhydrophobic coatings with reduced ice adhesion. J. Adhesion Sci. Technol., 23, 1215–1237. 64. Menini, R. and Farzaneh, M. (2009) Elaboration of Al(2)O(3)/PTFE icephobic coatings for protecting aluminum surfaces. Surf. Coat. Technol., 203, 1941–1946. 65. Fletcher, N.H. (2009) The Chemical Physics of Ice, Cambridge University Press, Cambridge. 66. Tourkine, P., Le Merrer, M. and Qu´er´e, D. (2009) Delayed freezing on water repellent materials. Langmuir, 25, 7214–7216. 67. Kulinich, S.A. and Farzaneh, M. (2004) Alkylsilane self-assembled monolayers: Modeling their wetting characteristics. Appl. Surf. Sci., 230, 232–240. 68. Kulinich, S.A. and Farzaneh, M. (2009) Ice adhesion on super-hydrophobic surfaces. Appl. Surf. Sci., 255, 8153–8157. 69. Cao, L.L., Jones, A.K., Sikka, V.K. et al. (2009) Anti-Icing Superhydrophobic Coatings. Langmuir, 25, 12444–12448. 70. Marmur, A. (2006) Super-hydrophobicity fundamentals: Implications to biofouling prevention. Biofouling, 22, 107–115. 71. Koc, Y., de Mello, A.J., McHale, G. et al. (2008) Nanoscale superhydrohobicity: Suppression of protein adsorption and promotion of flow induced detachment. Lab Chip, 8, 582–586. 72. Zhang, X., Shi, F., Niu, J. et al. (2008) Superhydrophobic surfaces: From structural control to functional application. J. Mater. Chem., 18, 621–633. 73. Callow, J. and Callow, M. (2011) Trends in the development of environmentally friendly fouling-resistant marine coatings. Nature Commun., 2. doi: 10.1038/ncomms 1251 74. Barkhudarov, P.M., Shah, P.B., Watkins, E.B. et al. (2008) Corrosion inhibition using superhydrophobic films. Corros. Sci., 50, 897–902. 75. Ishizaki, T., Masuda, Y. and Sakamoto, M. (2011) Corrosion resistance and durability of superhydrophobic surface formed on magnesium alloy coated with nanostructured cerium oxide film and fluoroalkylsilane molecules in corrosive NaCl aqueous solution. Langmuir, 27(8), 4780–4788. 76. Yuan, S.J., Pehkonen, S.O., Liang, B. et al. (2011) Superhydrophobic fluoropolymermodified copper surface via surface graft polymerisation for corrosion protection. Corros. Sci., 53 (9), 2738–2747. 77. Xu, W., Song, J., Sun, J. et al. (2011) Rapid fabrication of large-area, corrosion-resistant superhydrophobic Mg alloy surfaces. ACS Appl. Mater. Interfaces, 3 (11), 4404– 4414. 78. Hermelin, E., Petitjean, J., Lacroix, J.C. et al. (2008) Ultrafast electrosynthesis of high hydrophobic polypyrrole coatings on a zinc electrode: Applications to the protection against corrosion. Chem. Mater., 20, 4447–4456.

32

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

79. Li, Y., Liu, F. and Sun, J.Q. (2009) A facile layer-by-layer deposition process for the fabrication of highly transparent superhydrophobic coatings. Chem. Commun., 2730– 2732. 80. Deng, X., Mammen, L., Zhao, Y. et al. (2011) Transparent, thermally stable and mechanically robust superhydrophobic surfaces made from porous silica capsules. Adv. Mater., 23 (26), 2962–2965. 81. Bravo, J., Zhai, L., Wu, Z. et al. (2007) Transparent superhydrophobic films based on silica nanoparticles. Langmuir, 23, 7293–7298. 82. Manca, M., Cannavale, A., De Marco, L. et al. (2009) Durable superhydrophobic and antireflective surfaces by trimethylsilanized silica nanoparticles-based sol-gel processing. Langmuir, 25, 6357–6362. 83. Levkin, P.A., Svec, F. and Frechet, J.M.J. (2009) Porous polymer coatings: A versatile approach to superhydrophobic surfaces. Adv. Funct. Mater., 19, 1993–1998. 84. Gao, L. and McCarthy, T.J. (2006) The “lotus effect” explained: Two reasons why two length scales of topography are important. Langmuir, 22, 2966–2967. 85. Zhang, J.L., Li, J.A. and Han, Y.C. (2004) Superhydrophobic PTFE surfaces by extension. Macromol. Rapid. Commun., 25, 1105–1108. 86. Shiu, J.-Y., Kuo, C.-W. and Chen, P. (2005) Fabrication of tunable superhydrophobic surfaces. Proc. SPIE, 5648, 325–332. 87. Youngblood, J.P. and McCarthy, T.J. (1999) Ultrahydrophobic polymer surfaces prepared by simultaneous ablation of polypropylene and sputtering of poly(tetrafluoroethylene) using radio frequency plasma. Macromolecules, 32, 6800– 6806. 88. Ellinas, K., Smyrnakis, A., Malainou, A. et al. (2011) “Mesh-assisted” colloidal lithography and plasma etching: A route to large-area, uniform, ordered nano-pillar and nanopost fabrication on versatile substrates. Microelectron. Eng., 88, 2547–2551. 89. Szu, P.S., Yang, Y.M. and Lee, Y.L. (2007) Hierarchically structured superhydrophobic coatings fabricated by successive Langmuir-Blodgett deposition of micro-/nano-sized particles and surface silanization. Nanotechnology, 18, 465604. 90. Shirtcliffe, N.J., Mchale, G., Newton, M.I. et al. (2005) Porous materials show superhydrophobic to superhydrophilic switching. Chem. Commun., 25, 3135–3137. 91. Yabu, H. and Shimomura, M. (2005) Single-step fabrication of transparent superhydrophobic porous polymer films. Chem. Mater., 17, 5231–5234. 92. Lai, Y.K., Gao, X.F., Zhuang, H.F. et al. (2009) Designing superhydrophobic porous nanostructures with tunable water adhesion. Adv. Mater., 21, 3799–3803. 93. Shibuichi, S., Onda, T., Satoh, N. and Tsujii, K. (1996) Super water-repellent surfaces resulting from fractal structure. J. Phys. Chem., 100, 19512–19517. 94. Yan, Y.Y., Gao, N. and Barthlott, W. (2011) Mimicking natural superhydrophobic surfaces and grasping the wetting process: A review on recent progress in preparing superhydrophobic surfaces. Adv. Colloid Interface Sci., 169, 80–105. 95. Liu, K.S. and Jiang, L. (2011) Bio-inspired design of multiscale structures for function integration. Nano Today, 6 (2), 155–175. 96. Guo, Z.G., Liu, W.M. and Su, B.L. (2011) Superhydrophobic surfaces: From natural to biomimetic to functional. J. Colloid Interface Sci., 353 (2), 335–355.

Part II Applications of Self-Cleaning Surfaces

2 Recent Development on Self-Cleaning Cementitious Coatings Daniele Enea Department of Architecture, University of Palermo, Italy

2.1

Introduction

The increased interest in the sustainability of the whole building process, ruled and controlled by a regulatory and legal system, is increasingly changing management strategies in construction, favouring a preventive policy instead of widespread intervention when breakdown occurs. This new approach requires knowledge of the durability of materials and building components and their ability to maintain acceptable performance characteristics over time. Reducing maintenance costs, focusing attention on the external building envelope, requires surfaces to be more durable. To address this issue, the scientific community, together with the construction industry, has, in the last two decades, carried out studies and research to improve the performance of building surfaces. Nanotechnologies have introduced significant innovations in several fields, particularly in the building construction one, where the application of a new technology based on photocatalysis has aroused increasing interest in the scientific community. The discovery of the different behaviour of matter at the macroscale and nanoscale is at the base of nanotechnology; different laws rule at different scales, leading to changes in properties with size. Nanotechnology deals with the electronic properties and the electronic effects of nanomaterials. Matter reduced in at least one dimension to a size less that 100 nm can be

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach, First Edition. Edited by Walid A. Daoud. © 2013 John Wiley & Sons, Ltd. Published 2013 by John Wiley & Sons, Ltd.

36

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

considered as a nanomaterial.The reduction in size means nanomaterials have an increased surface area, a main aspect influencing and determining specific properties. Nanotechnology is thus applied to modify the properties of matter, forming new materials with new properties, useful for environmental applications: from purification of industrial and vehicular exhaust emissions, to the clean and renewable solar energy. Photocatalysis is important in this regard and is applied to reduce environmental pollution. This is a clean technology based on the use of solar energy, activating the photocatalytic principles dispersed on surfaces and able to contribute to the reduction of environmental pollution and energy consumption. These innovative materials allow the acceleration of the reactions of chemical decomposition of atmospheric organic and inorganic pollutants and their availability in the construction building market is increasing. These materials – solid semiconductors – work by accelerating the oxidative process that leads to the complete mineralization of air pollutants which thus become harmless substances. The most commonly used solid semiconductor is titanium dioxide (TiO2 ), due to its widespread availability, its most efficient photoactivity, highest stability and lowest cost. Seminal works in this area started in 1972, when the Fujishima–Honda effect was developed, consisting in the production of oxygen gas bubbles at an electrode of TiO2 placed in electrical contact with a piece of platinum metal, both immersed in water and exposed to light; at the platinum electrode was observed the production of hydrogen [1]. The study showed that TiO2 subjected to a light source has a strong oxidant power, because it can oxidize water to oxygen gas. Other research, carried out especially in Japan, showed interesting results regarding the capability of TiO2 to oxidize almost all types of organic compounds. The electronic and optical properties of TiO2 have several applications in gas sensors, antireflection coatings for solar cells, antibacterial filters, nano-films and, most noteworthy, in the superhydrophilic and self-cleaning glass and building surfaces. Due to their extent and the limited costs to be covered by innovative photocatalytic materials, these surfaces are the most suitable.

2.2

Atmospheric Pollution: Substances and Laws

Among atmospheric pollutants, the most dangerous for human health are nitrogen dioxide (NO2 ), volatile organic compounds (VOCs) and particulate matter (PM). 2.2.1

Nitrogen Oxides

Nitrogen dioxide is a toxic gas, absorbing visible solar radiation and contributing to impaired atmospheric visibility. It absorbs visible radiation and has a potentially direct role in global climate change and plays a critical role in determining ozone concentrations in the troposphere. It is produced by the oxidation of nitric oxide in the atmosphere; this reaction commonly occurs in most combustion processes. Moreover, large-scale production of nitrogen oxides (NOx ) is due to internal combustion engines, particularly concentrated in congested urban areas, and stationary sources (heating and power generation).

Recent Development on Self-Cleaning Cementitious Coatings

37

The World Health Organization (WHO) has investigated NO2 , finding adverse health effects even when the annual average nitrogen dioxide concentration complied with the WHO annual guideline [2] value of 40 μg m–3 . 2.2.2

Particulate Matter

Airborne particulate matter is a complex mixture of components with different chemical and physical characteristics. Due to this heterogeneity, it is impossible to unambiguously establish health risks, without a correlation with particle size, source and chemical composition. Particulate matter is classified by aerodynamic diameter, as size is a critical determinant of the likelihood and site of deposition within the respiratory tract, so it is often designated as PM10 and PM2.5 , with particle diameters 10 and 2.5 μm, respectively. The two indicators differ for thoracic coarse mass particulate matter and PM10 includes these bigger particles and PM2.5 . PM2.5 is more dangerous than PM10 , as there is a high probability of deposition in the smaller conducting airways and alveoli. In urban environments, it is possible to find even ultrafine particles, smaller than 0.1 μm (100 nm), different in particle mass, origin, physical characteristics and chemical composition. The largest particles of PM10 are mechanically produced by the break-up of solid particles and often adhere to biological matter, thus containing dust from roads and industrial activities, and biological matter such as pollen grains and bacterial fragments. Coarse particles may also be formed from the incomplete combustion of vehicular engines and electric plants and are known as fly ash. PM2.5 is derived from gases, but combustion processes may also generate primary particles in this size range. Typically, these particles originate as ultrafine particles produced by nucleation–condensation of low-vapour-pressure substances formed by high-temperature vaporization or by chemical reactions in the atmosphere. The smaller particles ( E g . In this case, a hole (h + ) remains in the valence band, so these electron–hole pairs diffuse on the surface of the photocatalytic particle, participating in chemical reactions with the adsorbed molecules. Among others, titanium dioxide is the most used in photochemical processes due to its chemical stability, harmless nature and, compared to other semiconductor metal oxides, relative cheapness.

Recent Development on Self-Cleaning Cementitious Coatings

39

red1 ox1 Ec

–

TiO2

Eg

hv + red2

Ev ox2

Figure 2.1 The mechanism of photoreaction of a semiconductor particle.

The band gap of semiconductors is smaller than that of insulating materials, so radiation with energy greater than 3.2 eV for TiO2 in the anatase form and 3.0 eV in the rutile form is able to promote electrons to the conduction band. This energy is often supplied by a non-electric source, such as heat or light. Particularly, a quantum of light with wavelength lower than 380 nm, in the ultraviolet range, produces the electron–hole pairs, according to the equation: TiO2 + hν → h+ + e− The electron–hole pairs are able to react and decompose oxygen and water, present in the atmosphere, generating OH• , hydroxyl radicals, and O− 2 , superoxide ions, according to the following equations: H2 O + h+ → OH• + H+ O2 + e− → O− 2 These two powerful oxidizing agents will then disintegrate and rearrange the structure of some atmospheric pollutants and convert them, through redox reactions occurring on the surface of the catalyst, into limestone, nitrates and CO2 , which are easily washed away by rain [7, 8].

2.4

Self-Cleaning Surfaces

Nowadays, there is no American (ASTM) or International (ISO) standard reporting a description of self-cleaning surfaces, but the term is commonly referred to the superhydrophilic and hydrophobic properties. These two characteristics of a surface are strictly related to the contact angle between liquid drops coming into contact with the surface itself, as shown in Figure 2.2. The contact angle, ϕ, gives a quantitative measure of the wettability of a solid surface by a liquid and depends on the superficial tensions at the interface. Low values of the contact angle mean the liquid spreads on the surface until the complete wetting of the surface when ϕ = 0 and the surface is thus defined as hydrophilic.

40

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 2.2 The contact angle ϕ is formed between a liquid drop and a solid surface.

The higher the contact angle, the poorer the wetting of the surface and the surface is then known as hydrophobic. There is a range of values of contact angle between a solid surface and a liquid drop and it depends on the recent history of the interaction. The Loto-effect is characteristic of the lotus leaves (Figure 2.3), a genus of aquatic plants of the Nelumbonaceae family, having a superhydrophobic property such that water drops in contact have a contact angle greater than 130◦ . This effect ensures the self-cleaning property. Pollutant particulate, in fact, adheres to these water drops, and is thus removed. When titanium dioxide in the mineral form of anatase is irradiated by UV light, a superhydrophilic property develops. The contact angle formed between the surfaces treated with TiO2 and water drops is close to 1◦ . This phenomenon causes the water to spread and wet the surface, creating a nanometer-sized film, which is durable for two days after irradiation of the surface with UV light. This irradiation leads to the formation of oxidative agents able to decompose organic and inorganic materials, increasing the self-cleaning

Figure 2.3 Computer graphic of lotus leaf surface. Source: Image courtesy of W. Thielicke, http://wthielicke.gmxhome.de/ last accessed 15/01/2013.

Recent Development on Self-Cleaning Cementitious Coatings

41

property. This sheeting action not only helps rinse away loosened dirt and other organic material but also minimizes spots and streaks by helping the surface dry more quickly and uniformly. Main applications of this property are the development of nanometric TiO2 films for glass substrates (e.g. driving mirrors, highway sound barriers, solar cell cover glass, etc.) Focusing our interest on the construction industry, many experimental studies demonstrate the efficiency of photocatalytic cementitious products made with titanium dioxide and cement [9], depending upon the formulation used, in terms of: • Self-cleaning property: the capability to favour the removal of most organic and some inorganic pollutants that deposit on cementitious surfaces, causing stains and discoloration; • Pollution reduction: the capability to remove significant amounts of environmental pollutants deemed harmful to human health. 2.4.1

Mechanisms of Photo-Reduction of Air Pollutants

When air pollutants (e.g. NOx , SOx , VOC and particulate matter) come into contact with a photocatalytic surface containing titanium dioxide, they undergo a process of absorption, decomposition and final transformation into limestone and mineral salts which can be easily washed away by rain. Organic pollutants are thus degraded with the formation of CO2 . These photocatalytic reactions take place on a titanium dioxide surface and have a multi-phasic character, both at TiO2 /water and TiO2 /air interfaces. Photochemical degradation of air pollutants is a consequence of the decomposition of oxygen and water on surfaces treated with TiO2 and the formation of hydroxyl radicals and highly reactive superoxide ions. As a result of the photocatalytic oxidation, all elements presenting in a molecule of air pollutant are mineralized to inorganic species: carbon to CO2 , hydrogen to H2 O, halogens to halide ions, sulfur to sulfates, and phosphorus to phosphates, respectively. Several researches document the breaking up of the dioxin benzene ring by OH radicals, produced by photo-oxidation of water [10]. The decomposition of nitrogen oxide in air takes place through oxidation under ultraviolet light to nitric acid, HNO3 , and partially to nitrogen dioxide, NO2 . When NO2 is formed, part of the gas may escape from the photocatalytic surface, but in the presence of a cementitious matrix the gas may be effectively trapped together with the nitric acid formed, and can be easily removed from the surface by atmospheric water [11–13]. The system obtained by mixing TiO2 with cement well creates the conditions for environmental photocatalysis. Most of the photo-oxidizing compounds, including NO2 and SO2 , are acidic. The basic nature of the cement matrix fixes both the polluting reagent and the photo-oxidation products on its surface. 2.4.2

Some Experimental Evidences

In recent years many research programs have started in this field, motivated by the growing interest and multiple applications of photocatalytic materials; one of these was the “Photocatalytic Innovative Coverings Applications for Depollution Assessment” (PICADA) project.

42

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 2.4 The canyon street pilot (www.picada-projet.com). Source: Reproduced by permission of C.T.G. S.p.A.

This research program started in 2002 and ended in 2005, with the participation of the European Community, four industrial partners and some European research centres, for a total investment of €3.4 million. The aim of the project was to develop a range of photocatalytic covering materials and to evaluate their effect on a large scale, typically in street canyons, to better understand the photocatalytic reaction mechanisms and their effect on cleaning and de-pollution. The project was divided into eight work-packages focusing on the improvement of material properties, pre-development of applications and drafting guidelines. A wide range of mineral and organic coatings was selected for testing, in different thickness and with added nano-sized TiO2 . Several laboratory procedures were elaborated and some of these were codified by the Italian Standardisation Organisation (UNI) and are described in the next section. Further experience of a canyon street pilot (Figure 2.4), simulating an urban environment, was realized monitoring NOx and O3 with chemiluminescence analyzers. The difference between the NOx levels in the canyons indicated the significant capability of photocatalytic coatings to remove NOx from the air. NOx recorded concentrations in treated surfaces were 40 to 80% lower than those observed in the untreated reference canyon. Cassar evaluated the behaviour of different nano-sized TiO2, showing no direct correlation between increasing the specific surface area of the photocatalyst and the photocatalytic activity of the TiO2 /cement system. Laboratory tests showed a better efficiency of 150 nm TiO2 particles with respect to 15–20 nm ones, in the degradation of two different dyes, Rhodamine B and Bromocresol Green; thus on decreasing the specific surface area, the photocatalytic activity increased by about 10%. Moreover, mixing different particle sizes of photocatalyst gives a synergic effect, providing the best efficiency with a mixture of different particle sizes of TiO2 [14]. Cassar et al. tested the efficiency towards photo-degradation of unsaturated hydrocarbons and polycondensate aromatic compounds from tobacco cigarette ash by cementitious mortars with added TiO2 (1–2%) [15]. Diamanti et al. evaluated the photocatalytic activity of fiber-reinforced mortars based on white cement and titanium dioxide, in terms of the photo-degradation of 2-propanol, a

Recent Development on Self-Cleaning Cementitious Coatings

43

model reactant of VOC present in the atmosphere, to acetone and water by UV irradiation. The best results were achieved by spreading 2% TiO2 aqueous suspension on the surface samples so as to form a nanometric layer, improving the wettability of the envelope. On the other hand, samples with 3% TiO2 added and mixed with cement showed the best results in terms of maintenance of colour after exposure to outdoor urban environments [16]. Because of the activation of TiO2 dispersed in a cementitious matrix, strictly related to the UV light, and because of the wide use of colour in the building envelope by designers and architects, the influence of the addition of mineral pigments on the photocatalytic and depolluting performances of finishing products was evaluated. Mineral pigments were added in amounts from 0.1 to 0.5% to provide a significant coloration of surfaces and revealed a moderate decrease in the photocatalytic activity, in terms of NOx abatement, evaluated through the procedure of the UNI 11247:2010, described in the following section. The analysis was focused on three different building products: plasters, finishing coatings and paints, setting the samples in different ageing conditions (inside laboratory and outdoor conditions). The tested pigments were yellow, brown, red, blue and green, dry added to the binder mixtures. Preliminary tests were made on them using absorption spectrophotometry to evaluate the absorption spectrum. Even the lighter colours, such as yellow and red, absorbing the light radiation from nearultraviolet wavelengths, decreased the overall absorption of light radiation of wavelength λ < 410 nm in the mortars, thus lowering the photocatalytic activity (Figure 2.5). External ageing reduced the degradation of nitrogen dioxides more than under laboratory conditions (Figure 2.6). The least reduction was found with the brown pigment and among others, paint was the most effective.

1.20 plaster 1.1

cementitious coat NHL coat

1.0

paint

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.00 1910

250

300

350

400

450

500

550

600

650

700

750

800

850

900

950

1000 1050 11000

Figure 2.5 Absorption spectra of outdoor aged samples with 0.1% yellow pigment.

44

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

(a)

NOx reduction for base products added with 0.1% of pigment 100 90 80 70 60

plaster

50

cementitious coat

40

NHL coat

30

paint

20 10 0

blue

brown

yellow

red

Laboratory ageing conditions

NOx reduction for base products added with 0.1% of pigment

(b) 100 90 80 70 60

plaster

50

cementitious coat

40 NHL coat

30 paint

20 10 0

blue

brown

yellow

red

Outdoor ageing conditions

Figure 2.6 A comparison between reference products and those containing 0.1% of pigment, under laboratory (a) and outdoor (b) ageing conditions [17].

The different behaviour of the two finishing coatings is probably due to the natural hydraulic lime and its interaction with TiO2 . It would be necessary to allow longer curing, up to two months, to obtain acceptable performances in terms of NOx reduction [17].

2.5

Main Applications

The use of these materials is highly suitable in polluted urban areas, such as areas congested by vehicular traffic, tunnels and historical centres of the city, where narrow streets cannot

Recent Development on Self-Cleaning Cementitious Coatings

45

Figure 2.7 The three shells of the Dives in the Misericordia church in Rome [13]. Source: Reproduced by permission of C.T.G. S.p.A.

be easily closed to traffic. Moreover, using these materials increases the durability and facilitates the maintenance of external surfaces. The building market provides a wide range of photocatalytic building materials: coatings, such as plasters, paints, tiles and asphalt pavements, and also structural materials, such as pavement blocks and concrete. The first application of these materials on a large scale was the “Dives in Misericordia” church in Rome, whose three shells were made of concrete blocks based on the photocatalytic principle (Figure 2.7). This very prestigious and symbolic structure, designed by architect Richard Meier for the 2000 Jubilee, required the use of a particular concrete type. This concrete, in addition to being highly resistant and durable, ensures time-enduring white to the built elements due to the self-cleaning properties of the final surfaces [18]. This experience is considered unique, due to the complexity of the design and the impossibility of building the structures in a different way, ensuring the necessary durability for a symbolic architecture. The particular shape of the three shells obliged the designer to use cast concrete blocks, each with a different geometry and placing, instead of surface coatings. The activation of photocatalytic material occurs only in the presence of ultraviolet radiation, so the reaction takes place on the external surfaces and involves only a nanometric layer. Other applications of photocatalytic concrete structures are limited to buildings of particular importance and complex geometry, thus the building market tends towards the use of coatings, where the low thickness ensures the same efficiency of the concrete surfaces whilst reducing the use of titanium dioxide. Some recent examples of photocatalytic applications for the renovation of historical buildings were also completed in Italy, with the application of finish coatings and paints after the repair phase. Figure 2.8 shows the renovation of the facade of an Italian church, where photocatalytic paint was applied. The worthiness of the monument, made of historical

46

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 2.8 The facade of the Matrice church in Cittanova (RC) (www.nonsolocittanova.it). Source: Image courtesy of A. Mesiti, Cittanova (RC), www.nonsolocittanova.it.

masonry, and the necessity to preserve the original plaster led to the choice of a coating product based on natural hydraulic lime with a low content of cement as being more suitable for this intervention.

2.6

Test Methods

As mentioned before, there is no worldwide standard to evaluate the self-cleaning effect, but there are many standards to evaluate other parameters strictly related to this property. One of these parameters is colour, the durability of which is influenced by the capability of surfaces to remain clean. The Italian Standardisation Organisation (UNI) in recent years has published three standards dealing with the photocatalytic activity of cement-based materials to evaluate the capability of these materials to degrade air pollutants (NOx test, UNI 11247:2009; BTEX test, UNI 11238–1:2007 and Rhodamine B test, UNI 11259:2008). The above mentioned standards refer to direct procedures to measure the abatement of air pollutants. Nonetheless, there are other indirect approaches, not codified by standards, to evaluate the photocatalytic activity. Among others, X-ray photoelectron spectroscopy (XPS) and Raman spectroscopy are suitable to investigate the external nanolayers of the building envelope where photocatalytic reactions take place. These techniques are able to evaluate and quantify the products of the air pollutants degradation, thus providing information on the efficiency of the photocatalyst on the surface. 2.6.1

Colour

Monitoring the colour of a surface helps to evaluate the efficiency in terms of maintaining surfaces clean, as a result of the presence of a photocatalyst.

Recent Development on Self-Cleaning Cementitious Coatings

47

The UNI EN ISO 3668:2002 deals with the visual comparison of paints and varnishes commonly used in the paint and related coatings industry. The tolerances and differences are expressed in terms of approximately uniform visual colour perception in the CIE 1976 (L∗ a∗ b∗ ) colour space, well known as CIELAB colour space, and other colour spaces. CIELAB is an approximately uniform colour space based on nonlinear expansion of the tristimulus values and takes differences to produce three axes that approximate the perception of lightness–darkness (L∗ axis), redness–greenness (a∗ axis) and yellowness– blueness (b∗ axis). CIELAB colour space was designed by the International Commission on Illumination ´ (CIE means Commission Internationale de l’Eclairage) to approximate human vision. ∗ The total colour difference, E ab , between two colours, is a parameter calculated by the measurement of L∗ , a∗ , b∗ and is given by the formula:  ∗ E ab = (L ∗ )2 + (a ∗ )2 + (b∗ )2 ∗ The magnitude, E ab , gives no indication of the character of the difference since it does not indicate the relative quantity and direction of hue, chroma, and lightness differences. The direction of the colour difference is described by the magnitude and algebraic signs of the components L∗ (positive means lighter, negative means darker), a∗ (positive = redder, negative = greener) and b∗ (positive = yellower, negative = bluer). ∗ starting from zero, The classification proposed is a scale of the colour difference E ab that means no perceptible difference, to five representing very significant difference. This classification does not account for the different colour perception by the human eye that is more sensitive to some colours than others. Therefore, within the visible spectrum, ∗ , can be perceptible or insignificant to the human eye, depending the colour difference, E ab on the spectrum area considered. The concept of Just Noticeable Difference (JND) was defined for this reason, in which the acronym means the hardly noticeable colour difference to the human eye. In the CIE 1931 colour space, MacAdam, in the 1940s, defined some ellipses in the xy chromaticity diagram, being the brightness constant, the contours of which (Figure 2.9) vary in size and tolerance and represent the JND of chromaticity [19]. An assessment of JND, although quite approximate, was conducted by Sharma [20], ∗ of about 2.3 as the limit for the colour difference perceptible to defining the value of E ab the human eye. A similar treatment was carried out for the Cit`e de la Musique and Beaux Arts building in Chambery (Figure 2.10), in the south of France, close to the Italian border, between 1999 and 2001, of pre-cast concrete structural elements, based on the photocatalytic principle. The monitoring of the colour of its grey elements was carried out after three years from the completion and the results were a medium colour difference, in the four facades that were differently oriented, lower than 1 point [21].

2.6.2

Photocatalytic Degradation of Nitrogen Oxides

The UNI 11247:2010 standard deals with the determination of the catalytic degradation of nitrogen oxides in air by photocatalytic inorganic materials (cementitious and ceramic materials). The tests are carried out on samples set inside a reaction chamber, coming into contact with a continuous gas flow of 3 l min−1 , with fixed NOX concentration (equal to

48

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach 0.9 520 0.8

540

0.7 560 0.6 500 580

0.5 y 0.4

600 620

0.3

490

700

0.2 480 0.1

0.0 0.0

470 460 0.1

380 0.2

0.3

0.4 x

05

0.6

0.7

0.8

Figure 2.9 Standard deviations of chromaticity from indicated standards on CIE 1931 standard chromaticity diagram (Creative Commons Attribution-Share Alike 3.0 Unported licence).

Figure 2.10 The Cite` de la Musique and Beaux Arts building in Chambery [22]. Source: Reproduced by permission of C.T.G. S.p.A.

Recent Development on Self-Cleaning Cementitious Coatings

49

0.55 ppm, of which 0.15 ppm is NO2 and 0.4 ppm NO) in N2 , corresponding to a possible atmospheric pollution. The results can be expressed as a percentage of NOX decomposition by a photocatalytic sample under UV radiation, supplied by a UV lamp, 300 W power and irradiance at 365 nm, providing a light intensity of 20 ± 1 W m−2 . The concentration of NOx inside the reaction chamber is measured with a chemiluminescence NOx meter in two different phases: the first one is in the dark and measurements are taken after 30 and 60 min with the UV lamp switched off. In the second phase, with the UV lamp on, measurements are taken after 30 and 60 min, when the concentrations have reached equilibrium. The photocatalytic activity (AF ) in terms of nitrogen oxides degradation is calculated by the formula: (CB − CL )FI AF = CB S Where: CB CL F S I

is the concentration of NOX in dark conditions, at equilibrium (μg m−3 ) is the concentration of NOX in light conditions, at equilibrium (μg m−3 ) is the gas flow (m3 h−1 ) is the geometrical surface area of the sample (m2 ) is the adimensional intensity of the luminous flow, obtained by the ratio of the experimentally measured intensity I’ (W m−2 ) and 1000 W m−2 , corresponding to about 100.000 Lux, the average value of sunlight at noon on an average July day in Italy.

The photocatalytic activity of nitrogen oxides degradation can also be calculated as the percentage reduction of the nitrogen oxides – NOx – equal to the difference between the measured concentrations of NO, NOx and NO2 in dark, CB , and light conditions, CL . NOx = 100

(CB − CL ) % CB

The test set-up is represented in Figure 2.11. 2.6.3

Photocatalytic Degradation of Micro-Pollutants in Air

The UNI 11238-1:2007 standard is referred to photocatalytic cementitious materials and deals with the determination of the photocatalytic degradation of organic micro-pollutants present in the atmosphere particularly all those included in the acronym BTEX (benzene, toluene, ethylbenzene and xylenes). These organic pollutants are some of the volatile organic compounds (VOCs) present in petroleum derivates such as gasoline and diesel fuel. Using this standard, it is possible to measure the degradation of several compounds of a standardized mixture of BTEX promoted by photocatalytic cementitious materials, under an irradiance of 1000 μW cm–2 in the UV-A spectral band. The test procedure is based on the measurement of the equilibrium concentration of every component of the BTEX mixture inside a photo-chemical reactor where the sample is set, with a constant gaseous stream of a mixture of air and BTEX. These equilibrium concentrations are measured with the UV-A light source switched on and off. The

50

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach mass flow meters

control box lamp

V1

reactor flowmeter

air NOx

V2

mixing chamber

V3

pump

NOx analyser

Figure 2.11 Operational scheme of the test of photocatalytic activity (UNI 11247:2010). Source: Reproduced by permission of C.T.G. S.p.A.

photocatalytic activity of the sample is then determined as a function of the concentration of the compound, the mixture flow, the irradiance in the UV-A spectral band and the surface area of the sample. The result is expressed for every component of the BTEX in terms of photocatalytic activity measured in (μg m−2 h−1 )/(μg m−3 ) or in m h−1 . The standard defines two different BTEX mixtures, at low and high concentration, as in Table 2.1, and the pollutant concentrations in the gaseous stream and the irradiation levels are comparable to those found in real ambient conditions. The temperature has to be constant at 23 ◦ C during the entire test, the gaseous stream has to ensure 3 ± 0.5 air changes per hour and a gas chromatography analyser has to analyse and measure the concentrations of the organic compounds, before and after contact with photocatalytic surface of the sample set inside the photochemical reactor. The dimensions of the sample are related to the volume of the photochemical reactor and the ratio is 3 ± 1 m2 /m3 , the thickness has to be between 3 and 20 mm.

Table 2.1

Pollutants concentrations of the two mixtures used in the test.

Component Nitrogen Oxygen Relative humidity Benzene Toluene Ethylbenzene Xylenes

BTEX low concentration mixture 79 21 10.3 100 100 100 100

± ± ± ± ± ± ±

1% 1% 0.5 g m−3 10 ppb 10 ppb 10 ppb 10 ppb

BTEX high concentration mixture 79 21 10.3 300 300 300 300

± ± ± ± ± ± ±

1% 1% 0.5 g m−3 30 ppb 30 ppb 30 ppb 30 ppb

Recent Development on Self-Cleaning Cementitious Coatings

51

The test has two different stages; in the first, the concentrations of organic compounds are measured in dark conditions, after 24 h of direct contact of the gaseous mixture flow with the surface of the sample, until equilibrium is reached. Then, the circuit is closed and the sample irradiated with the UV-A lamp for 24 h at least. The concentrations of all BTEX are measured at equilibrium, confirmed after a measurement after 12 h. The photocatalytic activity (ACAT ) of the sample is calculated through the following formula: ACAT =

1000(CALIM − CIRRAD )F l ACIRRAD

m h−1

Where: CALIM is the initial concentration of each organic compound of BTEX at equilibrium (μg m−3 ) CIRRAD is the concentration of each organic compound of BTEX at equilibrium during the irradiation phase (μg m−3 ) F is the gas flow (m3 h−1 ) A is the geometrical surface area of the sample (m2 ) I is the irradiance on the surface of the sample at the UV-A spectral band (μW cm−2 ) The UNI 11238–2:2007 standard is referred to photocatalytic ceramic materials and is very similar to the Part 1 of the standard, dealing with the determination of the photocatalytic degradation of BTEX. 2.6.4

Photocatalytic Degradation of Rhodamine B

Rhodamine B is a red fluorescent dye, which is used in the UNI 11259:2008 test to evaluate the photocatalytic activity of hydraulic binders. This is a colorimetric method and measurements are referred to CIELAB colour space, particularly to the a∗ coordinate, which represents the colorimetric axis with red and green in opposition. The concentration of the Rhodamine B in distilled water has to be 0.05 ± 0.005 g l−1 and the minimum value of hydraulic binder a∗ is 12. The test lasts for 26 h and is based on the measure of the evolution of the a∗ coordinate of a cementitious sample treated with Rhodamine B, under UV light, ensuring an irradiance equal to 3.75 ± 0.25 W m−2 . The sample has a standard composition and each mixture contains 450 ± 2 g of hydraulic binder, 1350 ± 5 g of normalised sand and 225 ± 1 g of water so as to keep the water/cement ratio at 0.50. Mixing has to be at two different rates, a low rate of 140 ± 5 revolutions per min−1 for 60 s after adding water to the binder and 30 s at a high rate of 285 ± 10 revolutions per min−1 . After a 90 s stop, the revolutions have to restart for 60 s at the high rate. Sample dimensions are 16 × 14 × 4 cm3 with 0.5 cm tolerance and each test deals with four samples with the same concentration. Sample aging takes place in water at 20 ± 1 ◦ C for 7 days and in atmospheric air for a further 7 days.

52

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 2.12 The application of Rhodamine B to the samples.

All the surface of the samples, except the centre part, has to be covered with a hydrophobic material, silicone preferably, so as to mark off a central area of 22 ± 2 cm2 . Then the untreated area has to be wet with the Rhodamine B 0.5 ml solution (Figure 2.12) and set in dark conditions for 24 h at 20 ± 1 ◦ C and 60 ± 10% RH. The first measurement, a∗ (0), as an average of three measurements on the circular area, has to be taken with a colorimeter or a spectrophotometer, the second after 4 h, a∗ (4), and the final one after 26 h, a∗ (26). The result is the average of the measurements taken on three samples; the values obtained should be within 10% of each other, otherwise the sample has to be discarded and the fourth sample is considered. The sample can be considered photocatalytic with respect to the Rhodamine B, if the results satisfy the following: R4 > 20% R26 > 50% where: R4 = 100

a ∗ (0) − a ∗ (4) % a ∗ (0)

R26 = 100

a ∗ (0) − a ∗ (26) % a ∗ (0)

Recent studies showed a possible interaction with other organic admixtures in the products impacting the efficiency of this method to evaluate the photocatalytic activity of hydraulic binders.

Recent Development on Self-Cleaning Cementitious Coatings

2.6.5

53

Spectroscopic Techniques

X-ray photoelectron spectroscopy can provide information on the electronic structure and atomic quantitative analysis in the top 5 nm of the surface. It is based on X-rays, of energy hυ, exciting electrons from the valence band of surface atoms which are then ejected. The kinetic energy of these photoelectrons is measured by an electron energy analyser and the binding energy of the photoelectron can be calculated by this measured energy and the intrinsic characteristic of the spectrometer. The calculated peaks are easily identified using tabulated binding energy values from XPS handbooks, yielding information on chemical composition and bonding environments. Raman spectroscopy provides information on both the crystal structure and bonding characteristics of matter. It is based on inelastic scattering of monochromatic light, usually from a laser in the visible, near-infrared or near-ultraviolet range. The energy difference between the incident and reflected beam corresponds to a change in molecular vibration. These characteristic bond vibrations allow chemical and crystal state identification of matter. Particularly, Dalton et al. used these two techniques to evaluate the photocatalytic oxidation of NOx gases promoted by a TiO2 -treated surface, under UV irradiation. The XPS technique was applied to differentiate and quantitatively measure the adsorbed nitrogen species, such as NH3 , NO, NO2 and NO3 − . Raman spectroscopy was used to characterize the chemical bonds of the adsorbed atoms and molecules. They showed the mechanism of NOx removal based on its degradation into harmless nitrates adsorbed on the TiO2 surface [23]. A similar technique is molecular absorption spectroscopy in the ultraviolet (UV) and visible (VIS) which measures the absorption of radiation in its passage through a gas, a liquid or a solid. The wavelength region generally used is from 190 to about 1000 nm and the result of the absorption of photons of energy in this range of wavelengths is an absorption spectrum. Through this technique it is possible to provide two different analyses: qualitative and quantitative. The qualitative method leads to the identification of an analyte by comparing the absorption spectrum of the unknown substance with spectra of known substances. The quantitative method is based on the relation between absorbed radiation intensity and concentration. A known analyte can be determined by measuring the absorbance at one or more wavelengths and using the Beer–Lambert law and the molar absorption coefficient to calculate its amount concentration. Comparelli et al. monitored the photodegradation of Methyl Red, an organic dye, promoted by TiO2 -based photocatalysts. Due to its nature as a pH indicator, being red for pH less than 4.2 and yellowing with pH greater than 6.2, Methyl Red has an absorption spectrum depending on pH, so UV/VIS absorption spectroscopy was used [24]. Rashed et al. used the same method to evaluate and monitor the photocatalytic degradation of Methyl Orange in a solution of TiO2 [25].

2.7

Future Developments

The application of photocatalytic materials is continuously increasing in building construction. Several technologies are being developed to improve the efficiency of TiO2 and other photocatalysts as depolluting agents for superficial applications.

54

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Japanese researchers are studying the possibility to realize an energy saving cooling method by covering the external walls of a building with titanium dioxide film [26]. The improvement in photocatalytic technologies is focusing on responding to society’s needs: environmental protection, enhanced aesthetic value and durability of buildings. In terms of air pollution, researchers aim at increasing the decomposition properties of photocatalytic surfaces towards nitrogen and sulfur oxides. Great interest is arising in photocatalytic road surfaces and technologies to reduce the impact and the negative effect of particulate and powder, together with the development of systems able to measure in situ the reduction in air pollutants [27]. Hunger et al. developed a laboratory set-up to assess the decomposition of NOx promoted by concrete paving stones and derived a reaction model to predict quantitatively the efficiency in air-purifying of these concrete materials [28]. The efficiency of bituminous pavements was also studied by Da Rios et al. through laboratory tests on cementitious photocatalytic mortar applied on an open graded bituminous layer. Photocatalytic activity was evaluated in terms of nitrogen dioxides abatement with the equipment described in the UNI 11247:2010 and results confirm efficiency up to 40%. The elaboration of a mathematical model to simulate the NOx degradation mechanisms was conducted [29]. The Italian research group at Palermo is beginning a research program with the C.T.G. S.r.l., the research and development branch of the Italcementi group, and Hydratite, the local industrial partner, entitled “Assessment of the durability of photocatalytic cementitious materials aimed at the maintenance scheduling and planning” and coordinated by Prof. G. Alaimo. This program, following the methodology of the ISO 15686 and the UNI 11156:2006, aims to assess the durability of plasters, coatings and paints, in terms of photocatalytic activity, expressed by NOx reduction, maintenance of colour, resistance to saline aggressive atmospheres and resistance to abrasion of surfaces. The methodology foresees the exposure of samples to natural environment and artificial conditions produced by a climatic chamber (temperature range of 0 to 80 ◦ C, relative humidity range 0 to 100%, rain and UV irradiation) and the comparison of results to make a hypothesis on the rescaling of laboratory and outdoor parameters decay.

References 1. Fujishima, A. and Honda, K. (1972) Electrochemical photolysis of water at a semiconductor electrode. Nature, 238, 37–38. 2. Forastiere, F., Peters, A., Kelly, F.J. and Holgate, S.T. (2005) Nitrogen dioxide, in Air Quality Guidelines: Global Update 2005, Particulate Matter, Ozone, Nitrogen Dioxide and Sulphur Dioxide, World Health Organization Europe. 3. Samet, J.M., Brauer, M. and Schlesinger, R. (2005) Particulate matter, in Air Quality Guidelines: Global Update 2005, Particulate Matter, Ozone, Nitrogen Dioxide and Sulphur Dioxide, World Health Organization Europe. 4. Harrison, R.M. (2005) Source of air pollution, in Air Quality Guidelines: Global Update 2005, Particulate Matter, Ozone, Nitrogen Dioxide and Sulphur Dioxide, World Health Organization Europe.

Recent Development on Self-Cleaning Cementitious Coatings

55

5. (2003) Chemical Sampling Information: Benzene, Occupational Safety and Health Administration, United States Department of Labor. 6. Rothenberg, G. (2008) Catalysis: Concepts and Green Applications, Wiley-VCH, Weinheim. 7. Linsebigler, A.L., Lu, G. and Yates, J.T. (1995) Photocatalysis on TiO2 surfaces: principles, mechanisms, and selected results. Chem. Rev., 95, 735–758. 8. Hashimoto, K., Irie, H. and Fujishima, A. (2007) TiO2 photocatalysis: a historical overview and future prospects. AAPPS Bull., 17 (6), 12–28. 9. Cassar, L., Beeldens, A., Pimpinelli, N. and Guerrini, G.L. (2007) Photocatalysis of cementitious materials, RILEM Int. Symposium on Photocatalysis, Environment and Construction Materials, Florence, RILEM PRO 55, pp. 131–145. 10. Choi, W. (2006) Pure and modified TiO2 photocatalysts and their environmental applications. Catal. Surv. Asia, 10(1), 16–28. 11. Yumoto, H., Matsudoa, S. and Akashib, K. (2002) Photocatalytic decomposition of NO2 on TiO2 films prepared by arc ion plating. Vacuum, 65 (3–4), 509–514. 12. Ibusuki, T. and Takeuchi, K. (1994) Removal of low concentration nitrogen oxides through photoassisted heterogeneous catalysis. J. Mol. Catal., 88 (1), 93–102. 13. Cassar, L. (May 2004) Photocatalysis of cementitious materials: clean buildings and clean air. Mater. Res. Soc. Bull., 29, 328–331. 14. Cassar, L. (2005) Nanotechnology and photocatalysis in cementitious materials. Proceedings of NICOM’2, Bilbao, November 2005. 15. Cassar, L., Pepe, C., Pimpinelli, N. et al. (1997) Materiali Cementizi e Fotocatalisi, FAST, Milan. 16. Diamanti, M.V., Ormellese, M. and Pedeferri, M.P. (2008) Characterization of photocatalytic and superhydrophilic properties of mortars containing titanium dioxide. Cement Concrete Res., 38, 1349–1353. 17. Enea, D. and Guerrini, G.L. (2010) Photocatalytic properties of cement-based plasters and paints containing mineral pigments. J. Transport. Res. Board, 2141, 52–60. 18. Cassar, L., Pepe, C., Tognon, G. et al. (2003) White cement for architectural concrete, possessing photocatalytic properties. Proceedings of the 11th Int. Congress on the Chemistry of Cement, Durban, South Africa, Vol. 4, pp. 2012. 19. MacAdam, D.L. (1942) Visual sensitivities to color differences in daylight. J. Opt. Soc. Am., 32 (5), 247–274. 20. Sharma, G. (2003) Digital Color Imaging Handbook, CRC Press. 21. Guerrini, G.L. and Guillot, L. (2006) Realizzazione di edifici con utilizzo di cementi fotocatalitici. Proceedings of the 16th congress C.T.E., Parma, Vol. 2, pp. 941– 950. 22. Chiesa, G., Elias, G., Franchi, A. and Migliacci, A. (2009) Vademecum della progettazione consapevole, in Costruire per la qualit`a della vita: Expo 2015 un’occasione concreta, FAST, Milan. 23. Dalton, J.S., Janes, P.A., Jones, N.G. et al. (2002) Photocatalytic oxidation of NOx gases using TiO2 : a surface spectroscopic approach. Environ. Pollut., 120, 415– 422. 24. Comparelli, R., Cozzoli, P.D., Curri, M.L. et al. (2004) Photocatalytic degradation of methyl-red by immobilized nanoparticles of TiO2 and ZnO. Water Sci. Technol., 49, 183–188.

56

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

25. Rashed, M.N. and El-Amin, A.A. (2007) Photocatalytic degradation of methyl orange in aqueous TiO2 under different solar irradiation sources. Int. J. Phys. Sci., 2 (3), 073–081. 26. Irie, H., Sunada, K. and Hashimoto, K. (2006) Present situations and future prospects in TiO2 photocatalytic technologies. Optronics, 98, 102–111. 27. Venturini, L. and Bacchi, M. (2009) Research, design and development of a photocatalytic asphalt pavement. Proceedings of the II International conference Environmentally Friendly Roads, Enviroad, Warsaw, October 2009. 28. Hunger, M. and Brouwers, H.J.H. (2009) Self-cleaning surfaces as an innovative potential for sustainable concrete. in Excellence in Concrete Construction through Innovation, Proceedings of the International Conference on Concrete Construction, eds M.C. Limbachiya and H.Y. Kew, Taylor & Francis Group, London. 29. Da Rios, G., Lambrugo, S. and Bacchi, M. (September 2008) Analisi sperimentale per pavimentazioni urbane fotocatalitiche. Proceedings of the 17th S.I.I.V. congress, Enna.

3 Recent Progress on Self-Cleaning Glasses and Integration with Other Functions Baoshun Liu1,2 , Qingnan Zhao1 and Xiujian Zhao1 1

State Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, PR China 2 School of Material Science and Engineering,Wuhan University of Technology, PR China

3.1

Introduction

Self-cleaning surfaces have drawn much attention around the world in recent years, both in scientific research and commercial applications. They can prevent ice build-up on the surface as well as exposure to environmental pollution. The self-cleaning surfaces can be applied to a variety of applications including automotives, buildings, optical and household [1]. Self-cleaning glasses can be fabricated by coating with self-cleaning films, which give good optical transparency on the glass substrates. The area of selfcleaning glasses has become of great scientific and commercial importance as the future prospects of this application are huge. When compared with traditional glasses, the selfcleaning glasses utilize their self-cleaning properties to protect themselves from environmental pollution, leading to reduction in maintenance costs as well as improved quality of life. Further to the commercial applications, the self-cleaning glasses enrich and extend the scientific scope of traditional glasses. The coatings used in self-cleaning glasses include mainly (photoinduced) superhydrophilic coatings, superhydrophobic coatings and photocatalytic self-cleaning coatings. In comparison with other self-cleaning surfaces, the

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach, First Edition. Edited by Walid A. Daoud. © 2013 John Wiley & Sons, Ltd. Published 2013 by John Wiley & Sons, Ltd.

58

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

self-cleaning glasses require high optical transparency in the visible light region. The super hydrophilic and hydrophobic coatings can create an anti-fog effect while the transparency is retained at high levels on both cold and rainy days. The photocatalytic properties of the glasses enhance the decomposition of the organic substances adsorbed on the selfcleaning surface, thus keeping them clean. This chapter presents and discusses the theoretical fundamentals and recent research progress for a range of self-cleaning glasses, such as photocatalytic self-cleaning glasses, inorganic hydrophobic self-cleaning glasses, and self-cleaning glasses modified by organic molecules, while the function integration of self-cleaning glasses is also discussed.

3.2

Theoretical Fundamentals for Self-Cleaning Glasses

The theoretical fundamentals related to the self-cleaning glasses including the theories of wettability, photoinduced hydrophilicity and heterogeneous photocatalysis are discussed in this section. 3.2.1 3.2.1.1

Wettability Young’s Equation

The ability for wettability is evaluated by the contact angle of a droplet on a solid surface, which is determined by the force balance among three interfacial tensions [2], as shown in Eq. (3.1). γ cos θ = γSG − γSL

(3.1)

where γ is the surface tension of the liquid, that is, the interfacial energy between the liquid and the gas, γ SG is the interfacial energy between the solid, and the gas and γ SL is the interfacial energy between the solid and the liquid. In general, if the contact angle of water is less than 90◦ , the solid surface is considered as hydrophilic, while a surface with a contact angle greater than 90◦ is defined as hydrophobic. Similarly, a surface having a water contact angle smaller than 5◦ is considered as superhydrophilic, while a surface with a contact angle greater than 150◦ is usually considered as superhydrophobic. Both superhydrophilic surfaces and superhydrophobic surfaces can be considered as self-cleaning surfaces with the function of anti-fogging [3]. 3.2.1.2

The Composite Flat Surface

The nature of the surface used in the self-cleaning glasses is heterogeneous, which leads to further analysis of the theory for composite heterogeneous surfaces. There are three important parameters that determine the surface wettability – the surface roughness, the relative area fraction of the chemically different portions of the films and their dimensions, and the surface energy of the portions of the film. The wettability of two-component and flat surfaces is given by the Cassie equation [4]. cos θ = f 1 cos θ1 + f 2 cos θ2

(3.2)

Recent Progress on Self-Cleaning Glasses and Integration with Other Functions

59

where f1 and f2 are the area fractions of materials 1 and 2 (f1 + f2 = 1), and θ 1 and θ 2 are the contact angles of the pure materials 1 and 2 on the flat surface. Equation (3.2) can be used to analyze the wetting of rough surfaces in the case where the surface pores are filled with air or water, as applied to porous coatings. For the pores filled with air, the Cassie equation is reduced to the Cassie–Baxter equation [5]: cos θ = f 1 cos θ1 − f 2

(3.3)

This equation is used to describe the nature of superhydrophobic films. In the case where the pores are filled with water, the superhydrophilic behavior is described by the modified Cassie equation. cos θ = f 1 cos θ1 − f 2

(3.4)

where f2 represents the area of a droplet in contact with filled pores. 3.2.1.3

Unflat Composite Surface

The unflat composite surfaces have chemical heterogeneities of atomic or molecular dimensions. In this case, a different mechanism for the wetting process is suggested [6]. (1 + cos θ )2 = f 1 (1 + cos θ1 )2 + f 2 (1 + cos θ2 )2

(3.5)

If the solid/liquid interface is not flat, the f values are modified to explain the real contact area between the droplet and the surface, as described by the Wenzel model [7], where the rough component is introduced. cos θ = γ cos θY

(3.6)

where γ is the ratio between the real area of the rough surface and the projected area. For a two-component surface, the roughness can be taken into account by introducing the parameters γ 1 and γ 2 and describing the respective ratios between the real surface area of the rough surface divided by the projected areas. Introducing the two parameters (γ 1 and γ 2 ) into the Cassie equation, Eq. (3.6) becomes cos θ =

f 2 γ2 f 1 γ1 cos θ1 + cos θ2 = F1 cos θ1 + F2 cos θ2 f 1 γ1 + f 2 γ2 f 1 γ1 + f 2 γ2

(3.7)

According to the above theories, the manipulation of surface wettability through a combination of chemical and structural modifications has recently been of great interest in many applications. In general, increase in surface roughness can increase the hydrophobicity of the hydrophobic surface while also increasing the hydrophilicity of the hydrophilic surface. 3.2.2

Photoinduced Hydrophilicity

When a TiO2 film is illuminated with UV light, the water contact angle approaches 0◦ . Then, a pattern of two different types of area with hydrophobic and hydrophilic properties is formed. This co-existence of hydrophobicity and hydrophilicity results in the TiO2 surface being amphiphilic [8,9], as shown in Figure 3.1. In addition, the hydroxy groups anchoring on the TiO2 surface also play a significant role in this photoinduced hydrophilicity. An infrared spectrum showed a reversible presence and a peak decay assigned to the formation of hydroxy groups, which is assumed to be the dissociative adsorption of water molecules at

60

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 3.1 (a) Friction force microscopic (FFM) image (5 × 5 mm) for a rutile (110) surface before UV illumination; (b) FFM image (5 × 5 mm) of the same surface after illumination; (c) a medium-scale FFM image (1000 nm × 1000 nm) of the framed area in (b); (d) a highermagnification topographic image (240 × 240 nm) of the framed area in (c). For (d), the sample C 1999, John Wiley & Sons. stage was rotated 45◦ from the position in (c) [9]. Source: Copyright 

oxygen vacancy sites on the TiO2 surface. It is considered that the introduction of hydroxy groups onto the hydrophilic surface is crucial because the water molecule is favored to be physically adsorbed on the TiO2 surface. Besides the hydroxy groups, bridged oxygen molecules can be removed from the TiO2 surface under UV light illumination. Both surfaces (110) and (100) present similar photoinduced hydrophilicity [10,11], while surface (001) of the TiO2 is much slower. Figure 3.2 shows the atom arrangement in (110), (100), and (001) surfaces of TiO2 (rutile). Bridging oxygens, which are higher in position and energetically more reactive than their surrounding atoms, exist on both (110) and (100) planes. However, the (001) surface differs from the (110) and (100) surfaces in that all the surface Ti cations are fourfold coordinated with two oxygen atoms within the surface plane and two in the plane below. Hence the (001) surface lacks bridged oxygen atoms and, therefore, it is inert to UV illumination. XPS measurements showed that there are a few Ti3 + signals in the Ti2p spectrum on surfaces (110) and (100) due to the removal of bridged oxygen. However, none of the Ti3 + signals was observed from surface (001) as no bridged oxygen attached onto

Recent Progress on Self-Cleaning Glasses and Integration with Other Functions

61

(110) plane [110] –

[110] –

[001] (100) plane [100] [010] –

[001] (001) plane [001] [010] –

[100] Ti cations O anions (white atoms are bridging site O)

Figure 3.2 Schematic illustration of the atomic alignments on ideal TiO2 (110), (100), and C 1999, American Chemical Society. (001) single-crystal faces [10]. Source: Copyright 

the (001) surface. Therefore, the UV illumination could induce some oxygen vacancies on the TiO2 surface, which may be one of the reasons for the photoinduced hydrophilicity. In addition, the TiO2 film becomes hydrophobic in the absence of light after several days, so the wettability of TiO2 films can be tuned by UV light. Because droplets can completely spread on the TiO2 surface under UV illumination, TiO2 films/coatings have become an essential feature for applications in anti-fogging and self-cleaning glasses. The observation of friction force microscopy (FFM) showed the co-existence of hydrophobic and hydrophilic areas on the TiO2 surface. Seki presents a theoretical method to analyze photoinduced hydrophilic conversion processes of TiO2 surfaces. The simplest possible relation between the interfacial energy, γ SG − γ SL and the hydrophilic region surface fraction c, is given by the following linear function: γSG − γSL = γ1 c + γ2

(3.8)

Then, Young’s equation can be deduced f ≡ cos θ =

γSG − γSL = f1 c + f2 γ

(3.9)

When c = 0, cos θ has the value f2 , whereas at c = 1 cos θ has the value f1 + f2 . In the dark, the value of c decreases and, therefore, the hydrophilic regions become hydrophobic [12]. According to Young’s theory, the cosine of the contact angle is a function of the interfacial energy between the solid and the liquid, which in turn changes with the surface fraction of hydrophilic regions. As the UV light illumination proceeds, the surface fraction of the hydrophilic regions increases, which lowers the interfacial energy between solid and

62

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

liquid, resulting in the Cassie–Baxter lowering of the contact angle of water. Therefore, the cosine of the contact angle is more appropriate to extract from the rates of both hydrophilic conversion processes and back-processes.

3.2.3

Heterogeneous Photocatalysis

A photocatalytic effect was discovered, which can effectively remove the pollutants in air and water. If a photocatalytic material is coated onto glass, self-cleaning can also occur as the pollutants adsorbed can be degraded by photocatalysis. TiO2 is a material with dual functions of photocatalytic and photoinduced hydrophilic effects, so coating TiO2 onto a glass surface can improve the self-cleaning function. It is generally considered that photocatalysis originates from the generation of photoinduced electrons and holes, thus transport, trapping and transfer of electrons and holes in the semiconductor bulk and on the surface are very important. Many materials have a photocatalytic effect, and TiO2 is the most studied because of its high chemical stability and durability. The band gap of the TiO2 is 3.2 eV for the anatase phase and 3.0 eV for the rutile phase. [13,14] When the TiO2 is illuminated by UV light with photon energy greaterer than its band gap, an electron and a hole will be produced in the conduction and valence band, respectively. For nanosized TiO2 , the electrons and holes transport at a fast rate to the surface before recombination occurs. [15–17] The electron is trapped on the surface by oxygen molecules, resulting in the formation of O2 − ions, while the hole is captured by hydroxy groups, leading to the formation of OH• . The O2 − and OH• free groups are found to be capable of efficiently removing organic pollutants. This is accepted as the common origin of the photocatalysis. The generation, transport, recombination and trapping of photoinduced electrons and holes are shown in Figure 3.3. The function of photocatalysis is believed to be important in self-cleaning glasses.

hv +o (a)

(a)

∗

o

+ D+

(c)

(b) A

o

o

(d) +o

D

A–

Figure 3.3 Processes occurring on a bare TiO2 particle after UV excitation [13]. Source: C 2008, Elsevier. Copyright 

Recent Progress on Self-Cleaning Glasses and Integration with Other Functions Light

Light

63

Water

CO2 CO2 TiO2

TiO2

Figure 3.4 Schematic diagram of the decontamination process occurring on the superhyC 2008, Elsevier. drophilic self-cleaning surface [13]. Source: Copyright 

3.3 Self-Cleaning Glasses Based on Photocatalysis and Photoinduced Hydrophilicity Since Fujishima and Honda discovered the method of water photo-splitting using a bare TiO2 electrode [18] in 1972, heterogeneous photocatalysis based on TiO2 has been extensively studied. Since the 1970s, photocatalysis has been proved to be a key technique for removing different organic pollutants, such as volatile organic compounds (VOCs) in a sustainable fashion [19]. The glass used in daily life is often exposed to many kinds of pollutants, such as dust, bacteria, and other organic pollutants (e.g. formaldehyde, acetone, benzene and ethylene). The bacteria can be eliminated by photocatalysis, while in the same process, due to the oxidation of photoinduced holes, the organic substances are decomposed as carbon dioxide and water. Importantly, because photocatalysis can work at extremely low light intensity, it can be used in many places. Using the properties of a photoinduced superhydrophilic surface, the small dust that adheres on the TiO2 coating can be easily removed by rainwater, as shown in Figure 3.4. Therefore, the photocatalytic and photoinduced superhydrophilic effect can be easily integrated together to realize the excellent self-cleaning function of glass with a TiO2 coating. The results of the outdoor exposure test, carried out by Fujishima’s group (Kanagawa Academy of Science and Technology), for a PVC tent material with and without TiO2 coating is shown in Figure 3.5. It can be seen that the PVC tent with TiO2 coating shows better properties of self-cleaning [13, 19] as the TiO2 coatings made use of solar light to clean the TiO2 surface. Besides PVC, the self-cleaning properties of TiO2 coatings have also been studied on ceramic tiles, highway tunnel lamps and glasses [20]. 3.3.1

Self-Cleaning Glasses with Pores

Nanoporous TiO2 thin films with controlled microstructure have good photocatalytic activity as well as photoinduced hydrophilicity, thus enhancing the self-cleaning glass applications. The introduction of nanopores not only increases the surface area of the TiO2 films but also the surface roughness. Nanoporous TiO2 thin films on soda-glass substrates were prepared by the sol–gel method by adding PEG in the precursor. The results showed that the hydroxy group content on the nanoporous TiO2 films increased with the weight of

64

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 3.5 Outdoor exposure test for a PVC tent material (The left-hand side of the tent material was coated with TiO2 ). (a) Picture taken July 22, 2004. (b) Picture taken April 23, C 2008, Elsevier. 2007, [13]. Source: Copyright 

PEG. The photocatalytic activity and photoinduced hydrophilicity first increased and then decreased with the addition of excess PEG [21, 22], as shown in Figure 3.6. Introducing nanopores into TiO2 films also increases the surface roughness, which in turn enhances the photoinduced hydrophilicity. Apart from introducing organic substances followed by calcination to etch the TiO2 surface, another method involves plasma etching which roughens the smooth polycrystalline TiO2 film and produces a nanoporous surface morphology, as shown in Figure 3.7. [23] The plasma etching enhances the hydrophilicity of the surface. Furthermore, it is believed that the photocatalytic activity can increase due to the high surface area of the TiO2 .

(a)

(b)

100

100

Region 1

Region 2

40

60

40

2000PEG

20

0

0.5

1

1.5

Weight of PEG / g

30 25

b

20

c

15

d

10 e

5

4000PEG 0

a

35 Contict rate (degree)

Degradation rate %

80

0 2

-5

0

10

30

60

UV Cumination Time (mins)

120

1

3

5

7

sterage time in dark room (days)

Figure 3.6 (a) Relationship between the amount of PEG and DDVP degradation rate for TiO2 thin films (b) Changes in the water contact angle when illuminated by UV light for 120 min and subsequently stored in a dark room for 7 days ((a) 0 g, (b) 0.25 g, (c) 0.5 g, (d) 1.0 g and C 2007, American Chemical Society. (e) 2.0 g PEG) [21]. Source: Copyright 

Recent Progress on Self-Cleaning Glasses and Integration with Other Functions

65

Figure 3.7 SEM images of TiO2 sol–gel films taken after plasma treatment and annealing in C air. The films were treated with CF4 plasma for (a) 0 s and (b) 45 s [23]. Source: Copyright  2007, American Chemical Society.

3.3.2

Doping to Realize Visible-Light-Induced Self-Cleaning Glasses

Pure TiO2 films show excellent photocatalytic activity and photoinduced hydrophilicity under UV light illumination but there is difficulty in making full use of the visible light for the applications of self-cleaning glasses. One possible way is by doping the TiO2 film with impurity elements such as Ni, N, Fe, Nb, Cr, Sn, Ce and V [24–32] in. Asahi et al. used magnetron sputtering to coat nitrogen-doped TiO2 film onto a glass surface, and they found good photocatalytic activity of the nitrogen-doped TiO2 [24]. We implemented the same method to fabricate nitrogen-doped TiO2 films on soda-lime-silica glass substrates. The as-prepared N-doped TiO2 film showed better optical transparency and was photocatalytically active under visible light [26], as shown in Figure 3.8. It is worth

(b)

(a) 100

UV

NTiO-3 80 60

NTiO-1

40 20 0 200

400

600

800

Wavelength/nm

1000

k/min-1×10-6

NTiO-2

18 16 14 12 10 8 6 4 2 0

UV

Visible

Visible Visible UV

NTiO-1 NTiO-2

NTiO-3

Sample No.

Figure 3.8 (a)Transmittance spectra of N – doped TiO2 films; (b) the photocatalytic activity C 2010, of nitrogen-doped TiO2 films under UV and visible lights [26]. Source: Copyright  Elsevier.

66

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

mentioning that the doped TiO2 thin films could possibly influence the fabrication of self-cleaning glasses that respond to visible light. The first principle calculations based on density functional theory (DFT) indicated that the N2p and O2p combine to form hybridized orbitals. N2p orbitals broaden the valence band of TiO2 and introduce gap states in the TiO2 . In this way the TiO2 bandgap is reduced, resulting in the absorbance of visible light. Metal ion doped TiO2 filmcan also be used in the self-cleaning coatings. V-doped TiO2 porous films were prepared using the micro arc oxidation method [28]. The V-doped layers revealed an enhanced hydrophilicity compared to the pure TiO2 films. V-doped TiO2 films were also prepared by the sol–gel dip-coating method, and the Ti0.85 V0.15 O2 films showed excellent properties of photocatalysis and hydrophilicity under visible light illumination. These films could potentially be used as self-cleaning coatings under daily conditions [29]. Different from the N-doped TiO2 , V ions replace Ti ions. V3d orbitals combine together with Ti3d orbitals to form a conduction band, resulting in the narrowing of the bandgap of TiO2 and visible light absorption. Transparent C–N–F-codoped TiO2 films with enhanced visible light photocatalytic activity and non-light activated super-wettability were prepared by a simple layer-by-layer dipcoating method using TiO2 sol and NH4 F methanol solution as precursors [33]. The contact angles of the C–N–F-codoped TiO2 films were 2.3∼3.1◦ in the absence of any illumination and these increased slowly in the dark ( 7) compared to the sample A coated

Table 4.4

XPS data of the samples. S2p

Sample A B

C1s

Ti2p

Atomic ratio (%)

BE (eV)

Atomic ratio (%)

BE (eV)

Atomic ratio (%)

Ti2p3/2 BE (eV)

Ti2p1/2 BE (eV)

2.1 7.2

103.30 103.54

5.9 5.5

283.7 283.7

49.3 49.3

458.1 458.1

464.1 464.1

Self-Cleaning Surface of Clay Roofing Tiles

111

Figure 4.20 High-resolution XPS spectra of Ti2p, coated tiles A and B.

with a thick TiO2 layer. The lower acidity of the surface contributes to a higher number of hydroxyl ions which were adsorbed and detected [54]. The presence of hydroxyl ions on the sample surface may also suggest the presence of Ti3 + on the surface [52, 55] but, considering that the OH− ions could come from H2 O and O–H species that are not directly bonded with Ti, evidently this area of investigation is a very complex issue. 4.2.1.2.2 Nanocomposite Titania Coatings Based on Anionic Clays. Layered double hydroxides (LDH), known as anionic clays could also be defined as hydrotalcite-like materials due to their structural similarities. These materials represent a very important group considering their wide range of application, such as catalysts, catalyst supports, anion exchangers and adsorbents [56]. Their introduction in the field of photocatalysis

Figure 4.21 High-resolution XPS spectra of O1s, coated tiles A and B.

112

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

seems to be a novel trend in the science world. The performance of TiO2 could be increased through the interaction with layered double hydroxides [57]. Our preliminary investigations [58] were dedicated to the design of photocatalytic active inorganic nanocomposite coatings based on TiO2 and ZnAl anionic clays. The idea was to implement novel nanocomposite coatings based on anionic clays that should provide, besides enhanced photocatalytic activity, also better compatibility with mineral substrates – clay roofing tiles. The aim of this research was to increase TiO2 performance through the interaction with ZnAl-LDH (layered double hydroxides – LDH) having the formula [Zn1-x Alx (OH)2 ](CO3 )x/2 ·m H2 O, where x = Al/[Zn + Al]. Layered double hydroxides ZnAl-LDH in our preliminary investigations [58] were prepared by the low supersaturation coprecipitation method with the addition of Zn(NO3 )2 ·6H2 O and Al(NO3 )2 ·6H2 O solution at constant pH (9.0–9.5) [59]. The temperature during the synthesis was constant (40 ◦ C). After ageing (24 h), drying (100 ◦ C) and washing of the obtained samples (pH = 7), the ZnAl-LDH powders were designed. Wet impregnation was used for TiO2 loading onto calcined ZnAl-LDH (500 ◦ C/5 h) to remove the excess of water. The impregnated samples were used for suspension preparation by the addition of H2 O2 and polyethylene glycol (PEG 4000). The prepared suspension was sprayed (p = 6.5 bar) onto clay roofing tiles in 5 cycles, dried (100 ◦ C) for 5 h and calcined at 500 ◦ C. The morpholology of the coated tile profile revealed a distinct border between the tile and the coating. The loading of the coating inside the pores of the clay roofing tile was visible, suggesting a satisfactory compatibility. The presence of layered double hydroxides on the surface of the roofing tile was proved by the presence of plate-like particles in the sand-rose formation, Figure 4.22 [56]. By measuring the contact angle on the coated tile (water as the measuring liquid), it was observed that the surface was superhydrophilic after 3.5 h of UV irradiation (light intensity of 0.912 mW cm–2 ) which enhanced the self-cleaning effect of the coated tiles. When these results were compared with the coatings based on commercial TiO2 suspensions [60], this ZnAl-Ti coating was significantly better when it came to the hydrophilic effect. Namely, the contact angle decreased to 24◦ , Figure 4.23, whereas in

(a)

(b)

Figure 4.22 (a) SEM micrographs of the coated clay roofing tile, (x1000); (b) coating on the surface of the clay roofing tile, ( × 25000).

Self-Cleaning Surface of Clay Roofing Tiles

reference tile

92.55°

113

coated tile

33.3°

coated tile

24.27°

UV irradiation time, 0h

reference tile

94.83° UV irradiation time, 3.5h

Figure 4.23 Contact angle measurements of the reference tile and the coated tile.

the investigation with a commercial TiO2 layer the contact angle (CA) was 46◦ after UV irradiation for the same time. The novel photocatalytic materials based on LDHs could be valuable self-cleaning materials. Synthesis, research and application of these “smart” materials which possess good compatibility with mineral substrates (due to their structural similarity) would enhance the overall quality of clay roofing tiles. However, serious attention must be paid to the selection of the coating precursors considering the sensitivity of the tiles to traces of NO3 − , SO4 2− and Cl− . This shows the great importance of a proper washing procedure. 4.2.2 4.2.2.1

Monitoring the Characteristics of Coated Clay Roofing Tiles Photocatalytic Activity Evaluation Based on the Degradation of Pollutants

A broad range of pollutants, both organic and inorganic, has been used to assess the photocatalytic efficiency of coatings. They can be classified into three categories: dyestuffs [56, 57], organic compounds [47] and inorganic gases [58]. Dyes are degraded by TiO2 under the influence of UV or solar light. Decomposition is assessed by discoloration measurements (color removal ratio), as well as chromatographic

114

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

investigations. Commonly used dyestuffs with different water solubility are: Methylene Blue (MB) and thionine (TH) as typical heteropolyaromatic compounds; Methyl Red (MR) and Methyl Orange (MO) as azo-compounds and Rhodamine B (RB) as a fluorescent pigment. The photocatalytic decomposition of MB has been reported to be an oxidative process [61]. The reason for the widespread application of MB is the fact that it is mainly nontoxic and convenient for the use as a dye. MB exhibits a strong absorption in the visible region (λmax = 664 nm; ε = 7402,8 mM−1 ) but not in the UVA region. It presents the perceived effectiveness of the MB test, which is considered to be a standard test for photocatalytic surfaces by the International Organization of Standardization (ISO 10678:2010). However, the assessment of the dye decomposition processes by discoloration measurements is still a subject of discussion.

4.2.2.2

Photocatalytic Decomposition of Dyes in Aqueous Solution

The functionality of the photocatalytic coatings tested by MB and RB in the case of porous substrates such as clay roofing tiles should be renewed by a pre-absorption process. Major attention must be paid to the selection of a suitable sample preparation. Namely, clay roofing tiles (without photocatalytic coating) as porous materials also absorb a certain quantity of the dye solution, reducing its concentration. In order to assess the contribution of the photocatalytic activity of the prepared mesoporous titania coating on the tile surface, the existing photocatalytic tests should be modified compared to the tests where a non-porous substrate is used. During our investigation [49, 50] the pre-absorption of the chosen dye was established until the absorption process was complete. Then the decomposition rate of the chosen dyes under UVA/VIS light irradiation (photocatalytic activity) was determined by recording the absorption spectrum. The photocatalytic activity measurement of the TiO2 coatings by discoloration of the MB/RB, in the procedure adjusted to porous substrate, was conducted as follows. First, the degree of absorption of the dissolved MB/RB molecules was measured by a pre-absorption test. A glass cylinder (test cell) with an inner diameter of 3 cm and a height of 6 cm was attached to the substrate using silicon glue. Both the test cell and the substrate were marked as test samples. The concentrations of MB/RB solutions for the pre-absorption test and for the photocatalytic test were 20 and 10 μmol L−1 , respectively. 12 ml of the MB/RB solution was poured into the test cell for the preabsorption test. A part of the tile was submerged in the MB/RB solution. The absorption of the MB/RB (20 μmol L−1 ) by the tile sample was allowed to proceed in the dark for 12 h. The procedure was continued in the dark with the test solutions of MB/RB (10 μmol L−1 ) for 24 to 36 h (until the absorption of the dye was complete). The absorption was considered complete if the differences in the concentration of MB measured after 30, 60, 120 and 150 min were less than 5%. When the absorption of the dye was complete, the test samples were irradiated for 1.5, 2.5, 3.5 and 24 h, Figures 4.24a and b [59, 60], (Osram Eversun lamp; UVA for MB/I = 0.67 mW cm−2 and VIS light for RB/I = 0.50 mW cm−2 , distance between the lamps and the reactor 18 cm). The photocatalytic activity of the materials was monitored with a UV/VIS spectrophotometer (Evolution 600, Thermoscientific, Britain, water as the reference sample) by measuring the absorption spectra of MB/RB as a function of the irradiation time [26, 49],

Self-Cleaning Surface of Clay Roofing Tiles

Methylene blue/ rhodamine B

Glass tube Glass tank

TiO2 coated fired roofing tile

(a)

115

(b)

Figure 4.24 Measurement of the photocatalytic activity of TiO2 coating: (a) in the protective box, (b) schematic view of the measurements [49].

while the calculation of the photocatalytic activity of the TiO2 coatings was done using the relation (4.1): TiO2 activity =

C0 − C C1 · · 100% C0 C0

(4.1)

where C0 is the concentration of the test solution of the dye before irradiation; C is the concentration of the dye after UV irradiation; C1 is the concentration of the dye after the pre-absorption test. The concentrations C and C1 were determined from the calibration curves showing the dependence of the absorbance at λmax (664 nm for MB and 554 nm for RB) of dye solutions as a function of the concentrations of MB/RB solutions, Figure 4.25. The photocatalytic activity of the examined samples was higher with decreased absorbance as a consequence of reducing the dye solutions concentration, Figure 4.26. The existence of an active sample, based on the obtained spectra during the irradiation time, was proved. 4.2.2.3

Degradation of a Model Organic Component in a Gas Medium

The photocatalytic activity of a coated tile could be evaluated by monitoring the degradation of model organic compounds in a gas medium by FT-IR. In our experiment, the degradation of isopropanol, as a model organic compound, in a gas medium (in the presence of a TiO2 coated tile) was carried out in a cylindrical reactor (1.4 L in volume) covered by a quartz glass [50,61]. The main parts of the equipment, beside the reactor made of stainless steel (1) were a 300 W Xenon lamp (2), humidity control part with molecular sieves (3) diaphragm pump and (4) FT-IR spectrometer (Perkin Elmer Spectrum BX, US) (5). All parts were connected by Teflon tubes. The valves were built at appropriate places. The coated tile sample was placed in the reactor in the position where the light intensity of the Xe lamp was 30 W m–2 . The system was hermetically sealed and the diaphragm pump was used to ensure steady air flow through the system. Before the start of the experiment the relative humidity at 23 ◦ C in the system was kept in the range 25–30% by drying the air through molecular sieves. This could prevent the influence of high relative humidity on the rate of

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Absorbance

116

2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 400

300

500

600

700

800

700

800

Wavelength (nm)

Absorbance

(a) 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 300

400

500 600 Wavelength (nm) (b)

Figure 4.25 Absorption spectrum for: (a) MB solution, (b) RB solution [62].

0.5 Absorbance

0h

0h 1.5 h 2.5 h 3.5 h 24 h

0.6

0.4 24 h

0.3 0.2 0.1 0.0 500

520

540

560

580

600

620

640

660

680

700

Wavelength (nm)

Figure 4.26 Dependence of the absorbance spectra of MB solution (10 μmol L–1 ) on irradiation time [49].

Self-Cleaning Surface of Clay Roofing Tiles OH H3C

O

O CH3

H3C

isopropanol

O

117

CO2 + H2O

CH3

acetone

Figure 4.27 Reaction of decomposition of isopropanol into acetone and further to CO2 and H2 O.

ADS

the degradation of the pollutant (isopropanol). The relative humidity and temperature were controlled by a thermometer and a hygrometer installed in the reactor. Each experiment was performed injecting 3 μL of isopropanol (∼800 ppm in the gas phase for the described reactor system) through a septum in the reacting system. Acetone was then oxidized to CO2 and water, Figure 4.27. The UV lamp (i.e. the light source) was turned on approximately 1–6 h after isopropanol injection, when constant readings of isopropanol were achieved. This state occurred when the adsorption of isopropanol onto the coated samples was accomplished. The infrared part of the spectrum was blocked by an adequate filter. The monitoring of the oxidation of isopropanol to acetone in the presence of the coated tile exposed to the light irradiation and the subsequent degradation of acetone was followed by calculation of the peak’s area characteristics for isopropanol and acetone, at 951 and 1207 cm−1 , respectively (Figure 4.28), measured by a FT-IR spectrometer and analysed by software. The photocatalytic activity of the samples was evaluated from the rate of the initial acetone formation because the photocatalytic oxidation of isopropanol to acetone at room temperature was rapid, and in a moment the slope gave an evaluation of the formation rate as linear, Figure 4.29. The conversion of isopropanol to acetone is shown in Figure 4.29. The value Cinitial is the threshold concentration of isopropanol in the gas phase, Co is the constant concentration of the isopropanol in the system after the required absorption time (tabsorption ).

0.040 0.035 0.030 0.025 0.020 0.015 0.010 0.005 0.000 –0.005

Acetone Acetone Isopropanol Isopropanol

2000

1800

1600

1400

1200

1000

800

Wavenumber (cm–1)

Figure 4.28 The infrared spectra of reaction products as a function of the coated tile irradiation.

118

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach 1000 concentration isopropanol [ppm]

900

concentration acetone [ppm]

concentration (ppm)

800 Cinitial

700 600

start of UV exposure at t = 0 500 400

linear approximation of rate of acetone formation

300 tabsorption

200

Co

100 0 0

2

4

6

8

10 time (h)

12

14

16

18

20

Figure 4.29 The course of the photocatalytic degradation of isopropanol to acetone.

Tile samples coated with modified Degusa TiO2 suspension, (M3 and M5 layers) fired at 150 ◦ C, are chosen as model systems. The greatest value of the isopropanol decomposition rate is obtained in the case of the tile samples coated with 3 layers (M150 3), see Table 4.5. However, a big difference between this value and that for the tile coated with a commercial TiO2 suspension (3 layers), A3 coating, was identified. 4.2.2.4

Water Contact Angle Values and the Topography of the Coated Tile Samples

The investigation of hydrophilicity of photocatalytic active coatings is of great importance since this characteristic is mandatory for a self-cleaning material. The designing of active coatings which would not reduce the high porosities of roofing tiles, as well as their mechanical properties and aesthetic appearance, is especially challenging. When a selfcleaning material absorbs water, water molecules come into contact directly with the coated surface [63]. As far as clay roofing tiles are concerned, the • OH and O2 • forms, which are created by the tile surface are very important for the photocatalytic processes. They need to be on the surface of the coating tiles in order to be active as microorganism inactivators.

Table 4.5

Rate of acetone formation. Tile sample

Conc. acetone (ppm h–1 )

A3

M150 3

M150 5

112

181.58

119.19

Self-Cleaning Surface of Clay Roofing Tiles

119

Table 4.6 Values of the contact angle of the coated tilesa (A) and the tile without coating (R). Sample R A1 (1 layer) A3 (3 layers) A5 (5 layers) a

Contact angle, (θ c ) before/after irradiation (◦ ) 63/64 88/74 85/72 96/73

Tile surface state hydrophilic hydrophilic hydrophilic hydrophobic/hydrophilic

Coated tiles fired at 150 ◦ C.

The photocatalytic values of a self-cleaning material could be correlated with the water contact angle (CA) values. The CA values of the tiles coated with a commercial photocatalytic TiO2 suspension (samples A) and with a suspension formulated in the laboratory (samples M), compared with the values of the reference tiles (R), without coatings, are presented in Tables 4.6 and 4.7. The measurements of the water contact angle were done before and after the light irradiation with an Osram Vitalux lamp [26]. After the light irradiation procedure, a reduction in water contact angle was noticed for all samples, Tables 4.6 and 4.7. This phenomenon is caused by a reasonable photon energy quantity absorption by the TiO2 -coated surface. The formed quantity of Ti3 + [63–68] is responsible for the photocatalytic activity and superhydrophilicity of the coated tiles. The results in Table 4.6 indicate that the samples with A coatings have slightly hydrophilic surfaces, which is confirmed by the CA values. The highest hydrophilicity for the A group of samples was noticed in the case of the A3 sample before, as well as after the irradiation procedure (Table 4.6, Figure 4.30). The CA values for the samples of the M series (Table 4.7, Figure 4.31) are much lower than those in the A series. Obviously, the hydrophilicity of these tiles is higher in comparison to those coated with the commercial suspension A, enriching their self-cleaning properties in this case.

Table 4.7 Values of the contact angle of the coated tiles (M) in comparison with the tile without coating (R). Sample R tile M120 1a (1 layer) M120 3a (3 layers) M120 5a (5 layers) M150 1b (1 layer) M150 3b (3 layers) M150 5b (5 layers) a b

Coated tiles fired at 120 ◦ C. Coated tiles fired at 150 ◦ C.

Contact angle,(θ c ), before/after irradiation (◦ )

Tile surface state

63/64 28/360 nm. The inactivation of E. coli mediated by the CuO suspensions was investigated as a function of the suspenson parameters: specific surface area of the Cu oxides (40–77 m2 g–1 ) as reported by Bandara et al., 2005, amount of CuO, light intensity and fate of the Cu1 + -ion within the inactivation process. The specific surface area of the CuO in suspension or deposited on cotton as investigated at a later stage was observed to be important during the E. coli inactivation kinetics. Figure 7.15a shows the results of the E. coli viability with CuO particulate in the dark. It is seen that within 17 h the E. coli decreases from a concentration of 106 CFU ml–1 to the CFU detection limit of 2.0 CFU ml–1 . The CuO dispersions in the dark in contact with E. coli lead to the abatement of E. coli. Figure 7.15b shows that, in the absence of CuO or in the presence

Self-Cleaning Textiles Modified by TiO2 and Bactericide Textiles Modified by Ag and Cu

219

109 DARK

E. coli (CFU ml−1)

107 Control, no CuO 105

CuO 77 m2 g–1

103

101

00

05

10

(a)

15 Time (h)

20

25

109 LIGHT

E. coli (CFU ml−1)

107

105

Control, no CuO Commercial 6.8 m2 g–1 40 m2 g–1 50 m2 g–1 77 m2 g–1

103

101 0 (b)

1

2

3

4

Time (h)

Figure 7.15 (a) Inactivation of E. coli as a function of time in the dark and in the presence of CuO (77 m2 g–1 ) added to the solution in a concentration of 1 g l−1 . (b) Inactivation of E. coli in solution in the presence of CuO (1 g l −1 ) powders with different specific surface areas at pH 7.1. Filter cutoff ≤ 360 nm.

of commercial CuO (0.3–0.5 m2 g–1 ), a negligible inactivation of E. coli takes place under light irradiation in the visible region. Figure 7.15b shows that the inactivation of E. coli is more efficient as the specific surface area of the CuO increases. No E. coli were observed in the CuO suspensions after 4 h illumination for CuO (40–77 m2 g–1 ) under light (light > 360 nm). The irradiance of the lamp with light above 360 nm was 29.2 mW cm–2 . The light induced inactivation of E coli in CuO suspensions (1 g l–1 ) was complete within 4 h.

220

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

The bacterial inactivation monitoring of E. coli (Escherichia coli CCT-1457) was carried out in a LB culture medium with glucose and a minimum in salt concentration (GMM). The mixture was: 20 g glucose (D-glucose monohydrated), 3 g K2 HPO4 , 7 g KH2 PO4 , 1 g (NH4 )2 SO4 , 0.5 g sodium citrate and 0.1 g MgSO4 •7H2 O to 1 l of distilled water. The pH of the medium was adjusted to 7.1 using 1 M NaOH. All experiments were conducted using a water bath at 37 ◦ C. Aliquots of the overnight culture medium were inoculated into a fresh medium and incubated aerobically at 37 ◦ C until the absorbance value reached ∼0.05 at 550 nm, which corresponds to 106 –108 CFU ml–1 , on the McFarland scale (Murray et al., 1995). CuO was then added to the samples at concentrations from 0.1 to 2 g l–1 . An aliquot of 400 μl of the samples was deposited on an agar PCA plate after preselected reaction times. The cytotoxicity of E. coli when using CuO (77 m2 g–1 ) was found for CuO concentrations as low as 0.2 g l–1 . A reaction mechanism is suggested for the Fenton-like reactions due to the Cu-ions/CuO action and the reactive oxygen species generated in solution (Bandara et al., 2005). By XPS it was observed that CuO in contact with the bacterial suspension changes its surface oxidation state from Cu2 + to Cu1 + . A redox process takes place in the dark that is accelerated under light, as shown in Figure 7.15. The surface composition of CuO remained stable during 4 h which is the time needed to inactivate the E. coli suspension.

7.9 Inactivation of E. coli by Pretreated Cotton Textiles Modified with Cu/CuO at the Solid/Air Interface The innovative high surface area CuO (65 m2 g–1 ) on cotton effective in bacterial inactivation in the dark and under sunlight irradiation was reported recently (Torres et al., 2010). The bacterial inactivation by cotton/CuO textiles of E. coli pretreated by RF-plasma was observed to be more favorable than CuO coating non-pretreated cotton by 40–50% in terms of the time required for the bacterial inactivation. The same Cu surface concentration was used in both cases, as determined by X-ray fluorescence (XRF). Figure 7.16a shows a negligible decrease in E. coli on the cotton alone in the dark, but with the coated cotton E. coli inactivation proceeds in the dark within 6 h, as shown by trace c. The bacterial inactivation was accelerated at low levels of visible light, as shown by trace d. This occurs due to the highly oxidative radicals generated by the CuO semiconductor. Figure 7.16d and show a shorter bacterial inactivation time of 3–5 h when the light intensity was increased, revealing the truly photocatalytic nature of bacterial inactivation on Cu/CuO cotton textile.

7.10 Direct Current Magnetron Sputtering (DC and DCP) of Nanoparticulate Continuous Cu-Coatings on Cotton Textile Inducing Bacterial Inactivation in the Dark and Under Light Irradiation A recent study reported features for Cu direct current magnetron sputtering (DC) on cotton mediating inactivation of E. coli (Castro et al., 2010). Sputtering for 40 s deposited 4 × 1016 atoms Cu cm–2 (or a thickness of 3 nm, equivalent to 15 atomic layers), the threshold amount of Cu required for complete bacterial inactivation. Cu-ionic species play

Bacterial Survival (CFU ml−1)

Self-Cleaning Textiles Modified by TiO2 and Bactericide Textiles Modified by Ag and Cu

221

1.0E+08 a b

1.0E+06

1.0E+04 d

e

c

1.0E+02

1.0E+00 0

1

2

3 Time (h)

4

5

6

Figure 7.16 E. coli inactivation mediated by: (a) cotton alone in the dark, (b) cotton under visible light irradiation from an Osram Lumilux T8-L18W source with 5 lamps (4.7 mW cm–2 ), (c) RF-plasma cotton/CuO (65 m2 g–1 ) with a loading of 0.71% wt/wt in the dark (), (d) same as (c) but under light irradiation of 1 lamp (1.2 mW cm–2 ) (), (e) same as (d) under light irradiation of 5 lamps (4.7 mW cm–2 ) (◦).

1.0E+06

1.0E+04

1.0E+02

629

–2

Irradiance (mW /cm )

Bacterial concentration (CFU ml−1)

a key role during the E. coli inactivation. Figure 7.17 shows the inactivation of E. coli on Cu-sputtered samples. Cotton by itself did not inactivate the E. coli (not shown to avoid overcrowding in the figure). The 40 s Cu DC sputtered samples had a loading of 0.060% Cu wt/wt polyester and inactivated bacteria within 120 min in the dark and within 30 min under visible light. The 180 s sputtered DC samples had a loading of 0.294% Cu wt/wt

400

500

600

400nm

λ (nm)

1.0E+00 0

30

60

90

120

Time (min)

Figure 7.17 E. coli inactivation for cotton: light (♦), dark () and Cu DC-magnetron sputtered cotton samples at times: 20 s dark (•), light (◦); 40 s dark (), light ();180 s dark (), light ().The spectral distribution of the visible light source used of 1.2 mW cm–2 is shown in the insert.

222

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

polyester and inactivated bacteria within 120 min in the dark and within 30 min when irradiated with 1% of the intensity of solar light. Therefore, it is not the Cu/CuO deposited on the cotton that leads to bacterial inactivation, but the Cu-ionic species able to diffuse through the bacteria cell-wall porins, that lead to bacterial inactivation. The sputtering time of 40 s led to the most favorable structure–reactivity for the Cucotton leading to the shortest bacterial E. coli inactivation. The sputtering for 40 s produces the optimal balance of film thickness, crystallite size and roughness, therefore exposing maximum surface area to the bacterial sample. Employing longer Cu sputtering times of 180 s Cu did not lead to faster E. coli inactivation kinetics (Figure 7.16). The mechanism of bacterial inactivation under light is suggested to follow the CuO semiconductor mechanism in the dark and under visible light (Bandara et al., 2005; Paschoalino et al., 2008). Cu forms CuO in contact with air (O2 ) and the bacterial inactivation of E. coli under light with respect to dark inactivation can be understood by the semiconductor behavior of CuO under light. Under irradiation CuO leads to surface oxidative radical formation + CuO + hν(< 660 nm) → CuO(e− cb , hvb )

(7.2)

•− The e− cb in Eq. (7.2) leads in Eqs. (7.3) and(7.4) to the formation of O2 radicals  − CuO ecb + O2 → CuO + O•− 2  +  − CuO ecb → CuO Cu    + CuO Cu + O2 → CuO Cu2+ + O•− 2 +

O•− 2

(7.3) (7.4) (7.5)

•

leads to the formation of the HO2 radical. The The equilibrium between H and HO2 • in Eq. (7.6) generates H2 O2 (Eq. (7.7)). •− • H+ + O− 2 ⇔ HO2 (partial O2 ), pK a = 4.8

(7.6)

with k7 = 0.6–2.3 M−1 s−1 . − 2+ CuO(Cu+ ) + H+ + HO− 2 (partial O2 ) → CuO(Cu ) + H2 O2

(7.7)

This study also identified the Cu-ions on the cotton intervening in the E. coli inactivation by XPS, reported the topography of Cu cotton fabrics compared to cotton alone, by confocal microscopy, and the increase in hydrophobicity of the Cu-cotton as a function of Cu-loading together with the amount of Cu on the cotton as a function of sputtering time, by XRF. A second recently published study used bipolar asymmetric DC-pulse magnetron sputtering (DCP) to deposit Cu-particles on cotton with a higher energy than DC-magnetron sputtering (DC) (Osorio et al., 2011). In Figure 7.18 the fastest E. coli inactivation was observed within 10 min when Cu was sputtered on cotton for 60 s and led to a bacterial inactivation time lower than the 30 min needed with DC (Figure 7.17). Practically no E. coli inactivation was observed in the dark in Figure 7.18 for cotton samples. The Cu-loaded samples sputtered for 20, 40 and 60 s showed an increased bactericide activity in the dark. The bactericide activity was not increased upon light irradiation with the same source of visible light used in Figure 7.17. This suggests that the Cu is mainly present as Cu0 /Cu1 + /Cu2 + and not in the form of CuO. In the latter case, the CuO semiconductor would enhance the bacterial inactivation kinetics under visible light compared to inactivation in the dark. This is shown in Figure 7.17. The Cu-film prepared

Self-Cleaning Textiles Modified by TiO2 and Bactericide Textiles Modified by Ag and Cu

223

300 mA DC-Pulse Magnetron Sputtering

E. coli (CFU ml−1)

1.E+08

1.E+06

1.E+04

1.E+02

1.E+00 0

30

60

90

120

Time (min)

Figure 7.18 E. coli inactivation for Cu DC-pulse magnetron sputtered cotton samples. Cotton alone dark ( . . . x), light (- - + ); 4 s dark (), light (); 20 s dark (), light (); 40 s dark (), light (♦); 60 s dark (•), light (◦).

by DCP shown in Figure 7.18 attained a surface density of 1.7 x 1017 atoms Cu cm−2 . A nominal thickness was measured on silica wafers of ∼30 nm or ∼150 Cu-layers. Applying a current of 300 mA DCP for 4 s, the threshold loading of Cu necessary to induce bactericide inactivation was 0.048% wt Cu/wt cotton. This Cu-film in Figure 7.17 contains 1016 atoms/Cu cm–2 having a thickness of 1.0–1.2 nm (5–6 Cu-layers). The fact that the bacterial inactivation time proceeds within different times with DCP compared to DC may be due to the different Cu-nanoparticle microstructure on the cotton. The Cu-microstructure seems to be a function of (i) the Cu-ion energies, that is, about 5–20 eV for the Cu-ions in DC compared to energies exceeding 100 eV in DCP (Lin et al., 2010); (ii) the density of the ions in the plasma in the sputtering chamber; and (iii) the applied current density per cm2 at the Cu-target.

7.11

Future Trends

Figure 7.19 shows the importance of antimicrobial textiles in the columns: clothing and sporting goods. The sub-categories are associated with the use of nanoparticles of all types or their combinations, by themselves or deposited on supports like textiles, for anti-odorant, health and disinfection purposes. The inventory of products related to cosmetics (137), clothing (155), personal care (193), sporting goods (93), sunscreen (33), and filtration (43) shows that the cosmetics, clothing and personal care sub-categories remain the largest. The textile group is one of the largest groups (Source: Woodrow Wilson, 2007). The biggest problems ahead could be: • There are not enough regulatory standards available for nanoparticles and textiles modified by nanoparticles for diverse uses like: disinfection, self-cleaning, deodorant textiles. Rapid standardization of the products is required at this point in time.

224

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Health and Fitness Subcategory 200 Mar 8, 2006 Aug 25, 2009

193

Number of Products

150

155 137

100 93

50 43 33

0

Cosmetics

Clothing

Personal Care

Sporting Goods

Sunscreen

Filtration

Figure 7.19 Number of products per sub-category within the category Health and Fitness.

• About 7–10% of patients and people visiting hospitals as well as hospital workers acquire infections while staying in these hospitals. Therefore, the full development of textiles with bactericide, antiviral, and antifungal properties is a necessity in economic terms. Low cost metal/oxides and combinations thereof with effective bacterial inactivation performance should be developed on textiles used in health care related places. • Many new toxic bacterial strains have developed that are resistant to antibiotics. Therefore, wide spectrum innovative bactericide tissues to inactivate resistant microbes should be developed. Moreover, the loading, stability and operational lifetime have to be set and warranted to offer commercial products to attract the confidence of the market place.

References Abidi, D.N., Hequet, E., Tarimala, S. and Dai, L. (2007) Cotton fabric surface modification for UV-protection using sol-gel methods. Appl. Polym. Sci., 10, 111–11. Amberg, M., Greder, K., Barbadoro, M. et al. (2008) Electrochemical behavior of nanoscale Ag-coatings on PET fibers. Plasma Proc. Polym., 5, 874–880. Bandara, J., Bowen, P., Soare, L. et al. (2005) Photocatalytic storing of O2 as H2 O2 mediated by high surface area CuO. Langmuir, 21, 8554–8559.

Self-Cleaning Textiles Modified by TiO2 and Bactericide Textiles Modified by Ag and Cu

225

Borkow, J. and Gabbay, J. (2008) Biocidal textiles can help fight nosocomial infections. Med. Hypothesis, 70, 990–994. Bozzi, A., Yuranova, T. and Kiwi, J. (2005) Self-cleaning of wool-polyamide and polyester textiles due to surface-TiO2 rutile modification under daylight irradiation. J. Photochem. Photobiol. A, 172, 27–34. Bozzi, A., Yuranova, T., Guasaquillo, I. et al. (2005) Self-cleaning of cotton textiles modified with TiO2 at low temperatures under daylight irradiation. J. Photochem. Photobiol. A., 174, 156–164. Castro, C., Sanjines, R., Pulgarin, C. et al. (2010) Structure-reactivity relations of the Cucotton sputtered layers during E. coli inactivation in the dark and under light. Photochem. Photobiol. A, 216, 295–302. Daoud, W. and Xin, H. (2004) Nucleation and growth of anatase crystallites on cotton fabrics a low temperature. J. Amer. Ceram. Soc, 87, 953–955. Daoud, W. and Xin, H. (2005) Synthesis of single-phase nanocrystallites at room temperature. Chem. Comm., 2110–2112. Daoud, W., Leung, K., Tung, S. et al. (2008) Self-cleaning keratins. Chem. Mater., 20, 1242–1254. Dastjerdi, R., Montazer, M. and Shahsavan, S. (2009) A new method to stabilize nanoparticles on textile surfaces. Colloid Surf., A, Physicochem. Eng. Aspects, 345, 202–210. Dhananjeyan, M., Kiwi, J. and Thampi, R. (2000) Performance of titania and iron oxide immobilized on derivatized polymer films for the mineralization of pollutants. Chem. Comm., 1443–1444. Geranio, L., Heuberger, M. and Nowack, L. (2009) The behavior of silver nano-textiles during washing. Environ. Sci. Technol., 4, 8113–8118. Hegemann, D., Amberg, M., Ritter, A. and Heuberger, M. (2009) Recent developments in Ag-metallized textile plasma sputtering. Mater. Technol., 24, 41–45. Hegemann, D., Hossain, M. and Balazs, M. (2007) Nanostructured plasma coatings to obtain multifunctional textile surfaces. Prog. Org. Coatings, 58, 237–240. Jin, M., Zhang, M., Nishimoto, S. et al. (2007) Light-stimulated composition conversion in TiO2 based nanofibers. J. Phys. Chem. C, 111, 658–665. Karin, D. and Gulneth, Y. (1987) Binding and activation of molecular oxygen by copper complexes. in Progress in Inorganic Chemistry (ed. S. Lippard), vol. 35, WileyBlackwell, pp. 220–237. Kasanen, J., Suvanto, M. and Pakkanen, Z. (2009) Self-cleaning TiO2 multilayer coatings fabricated by polymer glass. J. Appl. Polym. Sci., 111, 2597–2602. Kreutler, B. and Bard, A. (1978) Heterogeneous photocatalytic synthesis of methane from acetic acid-new photo-Kolbe reaction pathway. J. Am. Chem. Soc. 100, 2239–2240. Lin, J., Moore, J., Sproul, W. et al. (2010) The structure and properties of CrN coatings deposited using DC, pulsed-DC and modulated pulse power magnetron sputtering. Surf. Coat. Technol. 204, 2230–2239. Malnick, S., Bardenstein, L., Huszar, M. et al. (2008) Pyjamas and sheets as potential source of nosocomial pathogens. J. Hosp. Infect., 1–3. Mej´ıa, I., Mar´ın, M., Restrepo, G. et al. (2009a) Self-cleaning TiO2 cotton pretreated by RF-plasma and UV-C-light (185 nm) in vacuum and also under atmospheric pressure. Appl. Catal. B, 91, 481–488.

226

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Mej´ıa, I., Mar´ın, M., Restrepo, G. et al. (2009b) Self-cleaning of pretreated Nylon by UVClight (185 nm) and plasma RF under atmospheric pressure. ACS Appl. Mater. Interfaces, 1, 2190–2198. Mej´ıa, M., Mar´ın, M., Restrepo, G. et al. (2010) Magnetron-sputtered ag-modified cotton textiles active in the inactivation of airborne bacteria. ACS Appl. Mater. Interfaces, 2, 230–235. Mihailovic, D., Saponjik, Z., Radetic, M. et al. (2010) The functionalization of polyester fabrics with alginates and TiO2 nanoparticle. Carbohydrate Polym., 79, 526–532. Montazer, M. and Afgeh, M. (2007) X-linking and antimicrobial finishing of cotton fabric. J. Appl. Polymer Sci., 103, 178–185. Murray, P., Baron, E., Pfaller, A. et al. (1995) Manual of Clinical Microbiology, 6th edn, Am. Soc. of Microb., Washington, D.C. Osorio, P., Sanjines, R., Ruales, R. et al. (2011) Antimicrobial Cu-functionalized surfaces prepared by bipolar asymmetric DC-pulsed magnetron sputtering (PMS). J. Photochem. Photobiol. A., 220, 70–76. Paschoalino, M., Guedes, N., Jardim, W. et al. (2008) Inactivation of E coli mediated by high surface area CuO accelerated by light irradiation >360 nm. J. Photochem. Photobiol. A., 199, 105–111. Radetic, M., Ilic, V., Vodnik, V. et al. (2008) Antibacterial effect of silver nanoparticles deposited on corona-treated polyester and polyamide fibers. Polym. Adv. Technol., 19, 1816–1821. Rincon, A.-G. and Pulgarin, C. (2005) Use of a coaxial photocatlytic reactor (CAPHORE) on the photo-assisted treatment of mixed E coli and Bacillus sp and bacterial community present on waste waters. Catal. Today, 101, 331–344. Ritter, A. and Reifler, A. (2009) Quick screening method for the photocatlytic activity of nanoparticles powder materials. Appl. Catal. A, 35, 271–276. Shirley, A. (1979) A Correction and reference factors for XPS elements. Phys. Rev., B5, 4709–4717. Tolman, W. (1997) Making and breaking the O–O bond: New insights in copper complexes. Acc. Chem. Res., 30, 227–240. Torres, A., Ruales, C., Pulgarin, C. et al. (2010) Enhanced inactivation of E coli by RF-plasma pretreated cotton/CuO (65 m2 /g) under visible light. Appl. Catal. B, 2, 2547–2552. Tung, S. and Daoud, W. (2009a) Photocatalytic self-cleaning keratins. A feasibility study. Acta Biomater., 5, 53–56 (2005). Tung, W., and Daoud, W. (2009b) Effect of wettability and silicon surface modification on self-cleaning functionalization of wool. J. Appl. Polym. Sci., 12, 235–243. Wagner, C., Riggs, L., Davis, J. et al. (eds) (1989) Handbook of X-ray Photoelectron Spectroscopy, Perkin Elmer Corp, Eden–Prairie, MN 55344, USA. Wang, H., Xin, H., Yang, H. et al. (2004) The characteristics and photocatalytic activities of silver doped ZnO nanocrystalls. Appl. Surf. Sci., 227, 312–317. Woodrow Wilson (2007) Project on Emerging Nanotechnologies, www. nanotechproject.org, Pira Nanotextiles. Yuranova, T., Mosteo, R., Bandara, J. et al. (2006a) Cotton textiles modified by SiO2 /TiO2 with self-cleaning properties. J. Mol. Catal. A, 244, 160–167.

Self-Cleaning Textiles Modified by TiO2 and Bactericide Textiles Modified by Ag and Cu

227

Yuranova, T., Rincon, A.-G., Bozzi, A. et al. (2003) Antibacterial textiles prepared by RF-plasma and vacuum-UV mediated deposition of silver. J. Photochem. Photobiol. A. Chem., 161, 27–34. Yuranova, T., Rincon, A.-G., Pulgarin, C. et al. (2006b) Bactericide cotton textiles active in E coli abatement prepared under mild preparation conditions. J. Photochem. Photobiol. A., 181, 363–369.

8 Liquid Flame Spray as a Means to Achieve Nanoscale Coatings with Easy-to-Clean Properties Mikko Aromaa1 , Joe A. Pimenoff 2 and Jyrki M. M¨akel¨a1 1

8.1

Aerosol Physics Laboratory, Department of Physics, Tampere University of Technology, Finland 2 Beneq Oy, Vantaa, Finland

Gas-Phase Synthesis of Nanoparticles

Nanotechnology, as a generic name for a vast array of industrial sub-groups and academic disciplines, is applied all over the world. The market for nanotechnology is likewise huge. Also, much research is done in this field at the moment. The National Nanotechnology Initiative is a US Government program for coordinating research and development of nanotechnology in the US. The budget for this programme alone is US$ 1.6 billion for 2010 (The National Nanotechnology Initiative, 2009). The investments in nano-related activities have skyrocketed during the beginning of the twentyfirst century. Titanium dioxide (TiO2 ) and silicon dioxide (SiO2 ) are among the most produced materials, together with carbon black. Currently, most of the nanoparticles in use are produced via gas-phase synthesis, which is the most efficient way to produce nanomaterials in large quantities (Ulrich, 1984; Stark and Pratsinis, 2002; Pratsinis, 1998). There are several gas-phase methods, for example, flame methods, hot wall reactors, plasma, laser, and so on, but the fundamental principle in all is to bring material into the aerosol phase as vapour and subsequently synthesize nanoparticles via gas-to-particle conversion (Kodas and HampdenSmith, 1999; Samy El-Shall and Edelstein, 1996).

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach, First Edition. Edited by Walid A. Daoud. © 2013 John Wiley & Sons, Ltd. Published 2013 by John Wiley & Sons, Ltd.

230

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

There are several comprehensive textbooks available about aerosol formation in gas phase. Further information to supplement this brief review can be found in some of those (e.g. Friedlander, 2000; Samy El-Shall and Edelstein, 1996; Hinds, 1999, 2005; Kodas and Hampden-Smith, 1999). In the gas phase, the atoms and molecules transform into particles as long as the circumstances are favourable. The transformation takes place as the vapour is supersaturated, meaning that the partial vapour pressure of the material is higher than its critical vapour pressure. The supersaturation may occur by lowering the temperature or via chemical reaction. In chemical reactions, molecules react in order to form other materials. If the critical vapour pressure for this newly formed material is higher than its partial vapour pressure, molecules can nucleate into small clusters, that is, nuclei. Nucleation is the formation of new nuclei. In the aerosol phase, it means that molecules gather into clusters and form liquid or solid particles. The movement of molecules in the aerosol phase is a random process. Molecules collide randomly but molecules forming clusters need favourable circumstances. There are always several types of molecules in the aerosol. Here, we consider (i) vapour molecules, the compound which condenses and creates the nanoparticles in (ii) a carrier gas such as air, nitrogen or, for example, a mixture of nitrogen, oxygen, carbon dioxide, water vapour, carbon monoxide, NOx and so on. The vapour molecules form clusters when two of them collide in a random fashion. The molecules can also detach from the cluster randomly. In a gaseous system, both of these processes are occurring simultaneously. Molecules are attached to each other by van der Waals forces in the cluster. The clusters are thermodynamically unstable. The Kelvin equation below presents the diameter of what would be a stable nucleus: dp∗ =

4γ M ρ RT ln S

where γ is the surface tension, M the molar mass, ρ the density, R the universal gas constant, T the temperature and S the saturation ratio. The saturation ratio for material x is defined as the ratio of the partial pressure px of x in the gas phase and the critical vapour pressure pS (T) of x. S=

px ps (T )

If S is smaller than 1 the vapour is undersaturated, when S equals 1 the vapour is saturated and for S larger than 1 the vapour is supersaturated. When the cluster size exceeds the Kelvin size, the cluster will become a particle. Nucleation of new particles can either be heterogeneous or homogenous. In heterogeneous nucleation, a nucleus of a foreign substance is needed for molecules to attach to. Homogenous nucleation, however, does not need any foreign material, but instead the molecules start to clusterize and as they grow bigger than the critical size, they stay as clusters. For example, the nucleation of TiO2 molecules is homogenous because of the particle nature of even one single TiO2 molecule. Both heterogeneous and homogenous nucleation can also be divided into heteromolecular and homomolecular nucleation. In heteromolecular nucleation two or more types of molecules form the nuclei and in homomolecular nucleation only similar molecules form the nucleus.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

231

Formation of new particles takes place by either evaporation–condensation-limited nucleation or collision-limited nucleation mechanisms. In the evaporation–condensation mechanism, the formation of stable clusters is a limiting process for nucleation. This usually occurs for materials which have their vapour pressure within a couple orders of magnitude of the vapour pressure of water. Clusters larger than the Kelvin diameter are stable while those with a smaller diameter evaporate. Therefore, evaporation and condensation affect the particle size distribution for these particles. In collision-limited nucleation, dp∗ is smaller than one molecule. When supersaturation reaches really high levels, for example, through rapid chemical reactions, one can consider that even one molecule is larger than the Kelvin diameter and thus, thermodynamically, it may be considered as a particle. Therefore, since each collision, produces a particle, that is, even a collision of only two molecules, the collision of particles limits the formation of larger particles in the system. The formation rate of particles larger than some given size may then be treated as the nucleation rate. When a precursor in the gas phase reacts to form product species with a relatively low critical vapour pressure, a high supersaturation is reached. This leads to the possibility of collision-limited nucleation. Here, the chemical reaction taking place is actually limiting the nucleation rate. For example, thermal decomposition of titanium (IV) isopropoxide vapour in the gas phase into TiO2 is limited by the collisions while the growth of the particles happens due to collisions and coagulation. After nucleation, the particles start to grow. Growth needs a sufficient amount of vapour around the nuclei. Condensation of material on top of the particle is directly proportional to the vapour pressure of the material. Thus, we obtain an equation for the condensation growth rate for dp < λ: ddp 2M( p − pd ) = β1 √ dt ρp NA 2π mkb T and for dp > λ: 4DV M( p − ps ) dd p = β2 dt RTρp dp where M is the molar mass, p the vapour pressure, pd the vapour pressure on the particle surface, ρ p the density of the particle, NA Avogadro’s constant, m the mass of one vapour molecule, kB the Boltzmann constant, Dv the diffusion rate, ps the critical vapour pressure, R the universal gas constant, λ the mean free path in the gas and β 1 and β 2 are the Fuchs correction terms. The Fuchs correction can be expressed in the following way: 1.33K n(1 + K n) 1 + 1.7K n + 1.33K n 2 (1 + K n) β2 = 1 + 1.7K n + 1.33K n 2 β1 =

where Kn is the Knudsen number, which is dimensionless and defined as K n = 2λ , where λ dp is the mean free path in the gas and dp is the particle size. For a continuum regime, K n  1. For a free-molecule regime, K n  1. It is called a transition regime when K n ≈ 1. The role of the Fuchs correction is to match the two regimes, ballistic collision and continuous flow on a particle surface. The number of particles in the aerosol phase is changing all the time. Particles collide physically and stick onto each other forming aggregates. Aggregation is also one of the

232

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

key factors enabling particle growth. The change in total number concentration of the aggregates can be presented as a function of total number concentration of aggregates, N∞ , and the frequency of the collisions between aggregates: dN∞ 1 2 = − β N∞ dt 2 where β is the frequency function of collisions between the aggregates. This also applies for individual molecules in a collision-limited case. The coagulation coefficient for two single particles of sizes di and dj can be presented:   di + d j 8(Di + D j ) −1 βi j = 2π (Di + D j )(di + d j ) + di + d j + 2gi j ci j (di + d j ) where Di is a diffusion coefficient for a particle di , gij is the transition parameter and cij is a function of temperature and colliding particle masses. During growth, sintering will also occur. When two particles collide they can stick onto each other and form a new, bigger particle. Growth by coagulation can be divided into two types: kinematical coagulation and thermal coagulation. If collision occurs by electric force, gravity or another external force it is kinematical coagulation. However, if collision is due to Brownian motion the term “thermal coagulation” is used. The process is called sintering if the two colliding particles deform at collision and fuse together completely or partially. It is also possible that the particles do not fuse at all. In this case, the process is called agglomeration (Figure 8.1a). It has been suggested that partial fusing such as neck formation should be called aggregation (Figure 8.1b) when weak van der Waals forces are not the primary forces keeping the particles together. Aggregates are sometimes called

Figure 8.1 (a) agglomeration (b) aggregation (c) sintering.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

233

“hard” agglomerates as opposed to “soft” agglomerates which do not have a solid-state neck (Tsantilis and Pratsinis, 2004; Grass, Tsantilis and Pratsinis, 2006). Sintering (Figure 8.1c) is a process that depends remarkably on temperature, particle size and material properties. The driving force of sintering is the surface excess of an agglomerate compared to a spherical particle of the same mass and material. By nature, surface tension is trying to decrease the surface excess of the agglomerate. The decrease in radius is also driving towards sintering. There are several different methods for sintering: surface diffusion, lattice diffusion, grain boundary diffusion, evaporation–condensation reaction and viscous flow. Surface diffusion is believed to be the dominant mechanism at the nanoscale, regardless of the material, because nanoparticles have a great amount of surface atoms that have good mobility because of the curvature of the surface. The characteristic sintering time τ f and characteristic coagulation time τ c together define how sintering happens. Coagulation increases the surface area of a single aggregate whereas sintering will tend to decrease it. Thus, the surface excess is driving the sintering of particles. The change in the surface area of a particle, Figure 8.1a, can be expressed as: 1 dN∞ 1 da = − (a − as ) dt N∞ dt τf where aS is the surface area if the aggregate originates from a completely spherical particle.

8.2

Aerosol Reactors

There are several different methods available to create nanoparticles using aerosol synthesis. Several good reviews have been written about gas-phase synthesis methods. Kodas and Hampden-Smith (1999) have divided aerosol methods, in which particles are formed via gas-to-particle conversion, into two categories: physical and chemical routes. Physical routes include laser ablation, nozzles, supercritical spraying and free-convective plumes. Chemical routes include hot-wall techniques, laser as photothermal or photochemical reactors, plasma reactors and flame techniques (Kodas and Hampden-Smith, 1999; Swihart, 2003). Next, we will take a brief look into the chemical nanoparticle synthesis routes. Spray pyrolysis is also introduced as one technique, although it is not a gas-to-particle synthesis route, it is widely used and can also be utilised in flame spray synthesis. 8.2.1

Hot Wall Reactors

In hot wall reactors, a gas stream is heated by the thermal supply on the reactor walls (e.g. Backman, Tapper and Jokiniemi, 2004). The thermal energy drives the precursor species to react, thus forming nanoparticles. A typical hot wall reactor is a tube furnace. Carrier gases and the precursor are directed through the tube. In the furnace, aerosol processes take place by the influence of heat. Precursor materials react and start to nucleate, forming nanoparticles. Condensation, coagulation and agglomeration take place after the nuclei have formed. A schematic of the processes is presented in Figure 8.2. Typically, precursor materials need to be introduced in the furnace as gases or small droplets. Droplets can be produced, for example, by using a bubbler or evaporator. Usually,

234

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 8.2 A schematic of a hot wall reactor.

an inert carrier gas needs to be used. The inert gas is also used in the bubbler. The size and composition of the nanoparticles can be adjusted by tuning the gas flow rate and precursor feed rate. The temperature ramp along the axis in the furnace is also a parameter for particle size modification. Rapid quenching after the reactor can be used in order to stop the particle growth and agglomeration. The nanoparticles are usually collected using a filter after the reactor (e.g. Backman, Tapper and Jokiniemi, 2004). Titanium dioxide synthesis in hot wall reactors is performed by feeding gaseous titanium tetrachloride (TiCl4 ) and oxygen together with inert gas into the reactor. TiCl4 subsequently reacts with oxygen and forms TiO2 particles and chlorine gas. 8.2.2

Laser Reactors

Laser reactors are divided into photothermal and photochemical reactors, depending on the mechanism of the chemical reactions for the precursor material. Photothermal laser reactors are slightly similar to hot wall reactors. The difference is that the precursor stream is led through a laser beam. The laser beam heats up the precursor that runs through it. The acquired thermal energy drives the chemical reactions in the precursor materials, thus forming nanoparticles. Also in this case, coagulation and agglomeration take place after nucleation and condensation. A schematic of the process is presented in Figure 8.3. In photothermal laser reactors, precursor materials are typically in gaseous form (e.g. Cannon et al., 1982a, 1982b). Photochemical laser reactors operate using the same principle as presented for photothermal laser reactors. However, now UV light drives the chemical reactions instead of heat. Chemical dissociation in precursor materials happens solely due to the illumination by the laser. Therefore, the temperature is not high in the reactor, which is a clear benefit of the method (Morita et al., 1999). 8.2.3

Plasma Reactors

Plasma reactors are also thermal reactors. There are two main types of high temperature plasma reactors: direct current (DC) arc and high frequency induction plasma. Together with the main types of plasma, some other types of plasma reactors are also used. A

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

235

Figure 8.3 A schematic of a laser reactor.

schematic of a radio frequency (RF) plasma is shown in Figure 8.4. The operation temperatures of plasma reactors reach several thousand Kelvin. The plasma gas is ionized and is thus conductive. In the reactor, plasma is created from a gas that flows through the DC arc or the induction zone. The precursor is fed into the plasma either co-axially, as illustrated in Figure 8.4, or from the side. In highly turbulent plasma flames, all the materials are vaporized. However, some residual mode precursor materials will exist because of turbulence. The reactor needs to be cooled in order to prevent it from melting at high temperatures (e.g. Pfender, 1999). 8.2.4

Flame Reactors

Flame reactors (Figure 8.5) consist of a flame and precursor material feed. The flame can be produced from either a pre-mixed gas stream or it can be a diffusion flame in which combustion gases mix by diffusion after exiting the burner nozzle. The precursor materials can be introduced into the flame in either gaseous or liquid form. Solid precursor materials

Figure 8.4 A schematic of a RF plasma torch in a reactor. For example, inductively coupled plasma operates like this. For other plasma reactor types, the RF unit in the schematic can be replaced by, for example, DC arc wires or induction electrodes depending of the type of plasma reactor.

236

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 8.5 A schematic of a flame reactor.

are also a possibility in some cases. Combustion gases are typically methane or natural gas and oxygen or air (Pratsinis, 1998). Hydrogen and oxygen are alsopossible choices (M¨akel¨a et al., 2004). However, no fuel is necessarily needed to be supplied in gaseous form, but the precursor solution alone can act as fuel and combust together with oxygen (e.g. Bickmore et al., 1998). Flame methods are easily scalable and can operate in open air as opposed to other methods presented earlier in this chapter (Wegner and Pratsinis, 2003, 2005). The nature of combustion drives the precursor materials easily into oxide end products. However, the flame reactor can also be enclosed in a glove box in order to produce purely metallic nanoparticles and even metal salts (Grass and Stark, 2005, 2006a, 2006b). Liquid flame spray (Tikkanen et al., 1997a) is one special design of the flame reactors and it will be presented more thoroughly in the next section. 8.2.5

Spray Pyrolysis

Spray pyrolysis is widely used in thin film coating synthesis. It differs from other methods presented here in the way in which the coating is formed. In all the other methods presented, if they are used in the synthesis of a coating, it is made by depositing nanoparticles onto the surface. However, in spray pyrolysis, the precursor solution is atomized and sprayed directly onto a surface. Atomization can be done by, for example, a spray nozzle or a bubbler. The surface must be heated for chemical reactions to happen on the surface. Chemical reactions transform the precursor into a thin film. This film is not necessarily in nanoscale, but can achieve a thickness of several microns (e.g. Mooney and Radding, 1982). A schematic of spray pyrolysis deposition is shown in Figure 8.6.

Figure 8.6 A schematic of spray pyrolysis using an atomizer as the spray source.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

8.3

237

Liquid Flame Spray

Liquid flame spray (LFS) is one special case of flame aerosol synthesis methods. LFS is patented for material synthesis (e.g. Tikkanen et al., 1997b). Originally it was used for glass colouring. Commercial names nHALO (hot aerosol layering operation) (Pimenoff, Hovinen and Rajala, 2009) and DND (direct nanoparticle deposition) (Rajala, Janka and Kykk¨anen, 2003) are also used. nHALO is in use for float glass colouring (Gross et al., 1999) and ceramic tile coating and DND is used in manufacturing amplifying optical fibre. 8.3.1

Synthesis of Nanoparticles via LFS

The fundamental operating principle of LFS is a turbulent, high-temperature (Pitk¨anen et al., 2005), high-velocity (Keskinen et al., 2008) hydrogen–oxygen flame. The combustion gases are directed into the flame coaxially and the precursor solution is injected through a thin needle situated in the middle of the burner. In the burner H2 is used as the atomising gas for a liquid precursor and also as one of the combustion gases. The hydrogen flow speed can be approximately up to the speed of sound in the burner nozzle. By means of conventional two-fluid atomisation the hydrogen gas flow atomises the liquid capillary into micron-sized droplets. The hydrogen flow is also the main fuel supply in the combustion. The precursor liquids can also contain alcohols or organic solvents, which act as fuel in the flame, but this is not essential and water-based precursor liquids can therefore also be used. The burning LFS torch is shown in Figure 8.7 with indicated approximate temperature regions (Tikkanen et al., 1997a; Pitk¨anen et al., 2005). Liquid precursor droplets evaporate in the flame and evaporated compounds form nanoparticles in the gas phase as described before. On the left in Figure 8.8, the specific process stages are shown. First, the atomized precursor droplets evaporate. The precursor material evaporates together with the solvent. If the solvent is organic, it burns forming CO2 . The evaporated precursor then reacts chemically or decomposes into the vapour of the product material. Nucleation takes place at this point. After nucleation molecules and particles collide and grow via coagulation and sintering. The temperature of the flame decreases and soon sintering turns into aggregation and finally agglomeration. There are several parameters present in LFS for altering the particle size. The amount of combustion gases and material feed rates are the key factors in tuning the particle size (Aromaa, Keskinen and M¨akel¨a, 2007). Typically, the primary particle size for LFS-produced nanoparticles can be set from 2 nm up to 50–100 nm. Nanoparticles have a certain size distribution in the gas phase as they are produced. Airborne particles usually have a log-normal size distribution due to the formation and growth of the particles. The nature of a flame synthesis tends to drive the geometric standard deviation (GSD) of the log-normal distribution towards values of 1.35–1.40, which can be considered quite narrow. Liquid methods for nanoparticle synthesis always have broader distributions. There is also another route available for nanoparticle synthesis using LFS. In Figure 8.8 it is presented on the right-hand side. One can deliberately synthesize nanoparticles using liquid-to-solid reactions in the flame, for example, by choosing a precursor material that has a low volatility. Thus almost all the particles go through liquid-to-solid transformation and only a small amount of precursor material evaporates and forms particles via the gas phase. That is, when the liquid droplets do not have sufficient time for the evaporation process

238

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 8.7 LFS burner and temperatures inside the flame. The length of the flame and temperature zones depend on the combustion gas flows and liquid feed rate.

before the temperature reaches a point where thermal decomposition of the precursor material takes place. This results in somewhat larger particles, from 50 nm up to several hundreds of nanometres in diameter. The time needed for evaporation is dependent on the volatility of the solvent and the precursor material. Decomposition reactions of the precursor material also have an effect on the formation of residual particles. Residual particles cannot be avoided when particles are synthesised from liquid precursors. Nearly all the material evaporates from the droplet, but some is still left in the liquid form and the residual amount goes through liquid-to-solid transformation usually via chemical reactions. By choosing the precursor material so that it is very volatile, together with the solvent, and also so that decomposition is slow, residual particles can be avoided. However, there are still some residual particles but the particles are few in number or they have the same size as the particles that are produced via gas-to-particle transformation. The residence time of precursor droplets in the flame cannot be easily controlled. The gas feed rates, of course, have an effect on the droplet velocities but gas feed is not the limiting factor in residual particle formation. Burner geometry also affects the residence time, but the physical and chemical properties of the precursor are still predominant. A schematic of these two particle synthesis routes is shown in Figure 8.8. 8.3.2

Multicomponent Nanoparticles

One characteristic advantage of the LFS process is the use of multiple materials in one step (Keskinen, 2007a, 2007b, 2004). One can synthesize multi-component particles, either by adding a small amount of dopant materials within the precursor solution or simply by

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

239

Figure 8.8 Two different nanoparticle synthesis routes for LFS. Gas-to-particle route on the left and liquid-to-particle route on the right. The typical distribution consists of two sizeses 10 and 150 nm, formed via the gas-to-particle route and liquid-to-solid route (residual particles), respectively.

240

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 8.9 Different possible particle morphologies.

mixing two precursors together. The advantage of mixing precursors or introducing dopants is homogenous mixing of the resulting particles. Particles can obtain different morphologies depending on their vapour pressures and solubilities. Different particle morphologies are presented in Figure 8.9. In option 1 of Figure 8.9, there are particles of materials A (light grey) and B (dark grey) both approximately the same size and homogenously mixed. In option 2, there are dopants of material B on top of material A. The difference between the two options is only in the feeding rates of materials. Introducing a second material in the synthesis can decrease the particle size of the primary particles compared to the synthesis of individual components as different vapours disturb the growth. In option 3, material A nucleates first and material B condenses as a shell on top of the nanoparticles of material A. In option 4, two materials appear in the same particle, either in separate phases (4a) or as an alloy (4b). 8.3.3

Synthesis and Deposition of Nanoparticle Coatings

Nanoparticles can be deposited in situ on a surface to prepare a coating on the surface (M¨akel¨a et al., 2006). The substrate to be coated is swept through the flame or it resides in the flame to collect the nanoparticle, as illustrated in Figure 8.10. Nanoparticles attach onto the surface and stay there by van der Waals forces (Hinds, 1999). As the layer of particles grows, they continuously stick onto each other through van der Waals forces. If the top of the substrate is in the liquid phase the particles can also be attached mechanically as they sink a little onto the surface. Since the deposition of the particles occurs gently, as described, the porosity of the coating is very high, reaching even 95%. The level of porosity can be decreased by sintering or annealing the coating after deposition. The final amount of nanoparticles on the substrate, or the thickness of the coating is determined by the residence time of the substrate within the flame. However, there are several limitations for certain substrate materials. Some materials may be damaged due to the thermal energy of the flame. If the residence time is too long, the thermal energy of the flame has time to dissipate into the substrate. However, even paper and other easily flammable materials can be coated as long as the residence time is kept sufficiently short.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

241

Figure 8.10 Synthesis of nanoparticle coating. The substrate is moving through the flame from left to right and the nanoparticles are deposited onto the surface in a vertical motion.

The nanoparticles which are formed in the flame (Pitk¨anen et al., 2005) are carried to the close vicinity of the substrate by forced convection and turbulent diffusion. The eventual mechanism for nanoparticle deposition onto the surface, however, is mainly thermophoresis. For particles larger than, for example, 100 nm, impaction may occur, but for nanoparticles only thermophoresis and diffusion are relevant processes to carry the particles onto the substrate. In most cases, and especially in the case of a moving substrate, the substrate is not as hot as the flame. Also, due to efficient particle attachment on the substrate by van der Waals forces, the particle concentration is assumed to be close to zero on the surface. Thus, there is a thermal gradient from the flame towards the substrate, as well as a concentration gradient, which also slightly affects the particle flux. With a turbulent flame, the boundary layer formed above the surface is fairly distinct, being of the order of hundreds of micrometres or so (M¨adler et al., 2006). The particles need to penetrate through this layer before they deposit on the surface. For nanoparticles in the size range of the free molecular regime (dp < λ), that is, with size around 100 nm or less, we may use the following equations to estimate the particle flux towards the surface, in units of particles/m2∗ s: J = Jdiffusion + JThermophoresis,freemolec 0.55N ηg dN ∇ ln T − dx ρg   0.55ηg T ) Nδ D+ ≈− δ ρg T     0.55ηg T =Tav (Tδ − T0 ) Nδ  ≈− D T =Tav + δ ρg T =Tav Tav = Nδ vdep = −D

242

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

where the diffusion coefficient of the particle is given by D =

kB T C C 3πηdp

using the Cunningham

slip correction CC = 1 + dλp [2.34 + (Allen and Raabe, 1985). In the final form given above, the flux has been written as a product of particle concentration Nδ at the distance of boundary layer height from the substrate, and deposition velocity vdep . Using a temperature difference of, for example, 1000 ◦ C across a distance of 100 μm, one obtains the thermophoretic deposition velocity of 0.5–1 m s–1 . This may be considered fairly high in the scale of the flame of an ordinary liquid flame spray, being usually around 10 cm. Thus, in many cases for particle deposition from the highly concentrated flame onto the substrate at a significantly lower temperature than the flame itself, particle deposition due to thermophoretic flux is very efficient. In addition to the most significant deposition route (Figure 8.11, route (d)) described above, there are a few more routes of nanoscale coating formation. The precursor droplets might splash directly onto the surface without evaporation or reaction (Figure 8.11, route (a)). Then, the coating process is similar to spray pyrolysis, introduced in Section 8.2.5, as the chemical reactions for the precursor happen on the surface and are driven by the heat from the flame. The amount of material sprayed on the surface is dependent on the concentration of the precursor solution. Another possibility is to deliberately form residual particles as described before and deposit these on the surface (Figure 8.11, route (b)). There is also a possibility of chemical vapour deposition (CVD) on the surface (Figure 8.11, route (c)) (Choy, 2003). For some materials, the temperature of the flame might exceed the nucleation temperature at the coating distance. In this case, gaseous compounds deposit and nucleate on the surface. Island growth occurs as the substrate stays longer in the coating area. Moreover, the vapour forms a solid coating on the substrate (Figure 8.11, route (c). d 1.05 exp(−0.39 λp )]

Figure 8.11 Four different nanoscale coating deposition routes. In route (a) the initial droplet splashes onto the surface. In route (b) reacted droplets deposit on to the surface. In route (c) vapour deposits directly on to the surface. In route (d) nanoparticles deposit on to the surface. Source: According to Choy, 2003.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

243

Occurrence of CVD can be tuned by using a longer coating distance so there is enough time for nucleation in the flame.

8.4 Liquid Flame Spray in Synthesis of Easy-to-Clean Antimicrobial Coatings Liquid flame spray has been used successfully in the synthesis of antimicrobial coatings (Keskinen et al., 2006). We summarize the synthesis and deposition of photoactive titanium dioxide (Fujishima and Honda, 1972) nanoparticles. We also take a look at doping titanium dioxide with antibacterial silver in order to boost the easy-to-clean effect. By using antibacterial silver, the amount of titania can be decreased as bacteria do not attach to the coating as easily as to pure titania coatings. 8.4.1

Synthesis of Titanium Dioxide

Titanium dioxide particles are widely produced using aerosol methods. Titania has also been produced using LFS. The high temperature and rapid cooling in the flame are in favour of anatase formation. Therefore, LFS has been used in the synthesis of easy-to-clean and antimicrobial coatings. Titania is synthesized from a metal alkoxide precursor. Titanium tetraisopropoxide Ti(O − iC3 H7 )4 , also referred to as TTIP) is dissolved in 2-propanol and used as a precursor solution. The titanium concentration in the liquid can be quite low. Even the order of grams of atomic titanium in a litre of solvent is sufficient. In the flame, TTIP decomposes at a temperature as low as 250 ◦ C (Chen et al., 1993) according to the following overall reaction (Courtecuisse et al., 1996): TTIP → TiO2 + 4C3 H6 + 2H2 O Propene combusts into carbon dioxide and titania molecules grow via coagulation. Another possible reaction is hydrolysis of the alkoxide groups, which happens according to the following overall reaction (Courtecuisse et al., 1996): TTIP + 2H2 O → TiO2 + 4C3 H7 OH Anatase and rutile are commercially the most important crystal structures of titania. Thermodynamically, anatase is the most stable form for a crystal size below 11 nm. From crystal size 11 to 35 nm brookite is the most stable and above 35 nm rutile is the most favourable (Zhang and Banfield, 2000; Gribb and Banfield, 1997). The anatase-to-rutile ratio can be tuned by adjusting the process parameters. Most of the particles formed via gasto-particle conversion are found as anatase (Figure 8.12a) as the crystalline size is smaller than the limiting crystalline size for anatase. However, it is also possible to synthesize rutile, which has a small crystalline size, which is seen in the HR-TEM image in Figure 8.12b. The amount of residual particles on the surface is also an easy way to tune the anatase to rutile ratio. It seems that almost all the residual particles forming via liquid-to-solid reaction are rutile. They are also much larger in size, which is quite obvious. The large size favours rutile formation (Shannon and Pask, 1965; Gribb and Banfield, 1997; Ahonen et al., 2002). Figure 8.13 shows a TEM image of a residual rutile particle among many smaller anatase particles formed via the gas-to-particle route.

244

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 8.12 HR-TEM images of (a) anatase [100] particle (b) rutile [100] particle. The scale bars are 5 nm each. Source: Reprinted with permission from Dr Jakob B. Wagner, Lund University, Sweden Copyright (2013) Jakob B. Wagner.

8.4.2

Deposition of the Titania Coatings

Particle deposition can be done in a chamber. Beneq, a company using LFS for commercial applications, has designed special coating lines for coating ceramic tiles and float glass in an industrial scale. The coating lines usually comprise two sections for pre- and post-heating and a coating zone in between. A simple schematic of a line is shown in Figure 8.14.

Figure 8.13 A TEM image of a large residual particle (bottom left) and many smaller anatase particles. The scale bar is 200 nm.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

Pre-heating zone

245

After cooling zone

Figure 8.14 A simple design of a LFS coating line for float glass and ceramic tiles. The tiles move on a conveyor belt from left to right and go through the pre-heating, coating and after cooling zones.

In the pre-heating zone, the object is transported on a conveyor belt through an oven to run up a heating ramp. The ramp is important in order to avoid cracks and total breakdown of the substrate because of thermal shock. After the pre-heating, the conveyor belt transports the tile to the coating zone. Here, LFS generated particles are deposited on the substrate. A curtain of flames or a sweeping spot-like burner can be utilized. From here, the substrate is taken into a cooling zone, where it is cooled to room temperature. Precise cooling of the object is important so that there are no residual stresses left in the glass or the glazing of the tile. The LFS unit can also be merged into a manufacturing line for float glass or ceramic tiles. Doing the coating as described above, and for specific substrates and coating materials, results in a uniform cauliflower-shaped coating of titanium dioxide. There is always excess coating but it can be wiped away. Figure 8.15a presents a SEM image of the

Figure 8.15 (a) SEM image of the cauliflower shaped titanium dioxide nanoparticle coating on top of a ceramic tile. The scale bar is 100 nm. Source: SEM image kindly provided by M.Sc. Minna Piispanen Copyright (2013) Minna Piispanen, Abo ˚ Akademi University, Turku, Finland. (b) ceramic tiles undergoing the coating process.

246

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

coating. The SEM image is taken after all the loose excess of the coating has been wiped away from the coated tile (Figure 8.15b).

8.4.3

Doping of the Coatings

Silver can be added as a dopant on the coatings. The role of silver ions is to keep bacteria from attaching on the surface and, in addition, to kill the bacteria. As silver is already keeping the bacteria from adhering on the surface, photocatalytic activity itself can be achieved using a smaller amount of titanium dioxide as there are less bacteria to degrade compared to a pure titania surface. By combining these two properties a surface can be well protected from bacteria. Even a small amount of silver does the trick. The coating can be prepared in two steps or the particles can also be synthesised in one single step. If the coating is done in two different steps, the titania particles are first deposited on the surface and the silver particles are deposited on top of the titania layer. The synthesis can also be combined by mixing the precursor materials together and doing the coating in one step. In Figure 8.16, the difference between these two methods is visualised. Silver particles are synthesized from silver nitrate, which dissolves poorly in alcohols. However, the studied doping percentages of silver were 1–8 wt% of synthesized titania (Keskinen et al., 2007a). As the concentration of the precursor for titania synthesis is relatively low, silver nitrate dissolved well for the low concentrations. Figure 8.17 a shows the morphology of preparing the coating in two separate steps. First, the titania layer is deposited and after that, silver particles are deposited on top. The amount of silver visible on the surface is larger because all the silver lies on top. However, if the precursors are mixed, silver particles cover the titania particles and are evenly distributed throughout the coating. The formation of titania particles as a function of the amount of silver dopant has been studied by Keskinen et al. It was found that titania particles do not grow as big if there is silver present, when compared to pure titania synthesis. It is proposed that silver vapour disturbs the grain growth of titania and therefore the coagulation process is slower. A HR-TEM image of an individual silver particle next to a sintered titania cluster is shown in Figure 8.16. A high angle annular dark field TEM image is also shown in Figure 8.18. In this image, the existence of silver is clearly seen. The bright spots in the image are silver particles.

Figure 8.16 Particle distribution over coatings in (a) two-step and (b) one-step process. Here, large spheres are titania and small ones are silver.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

247

Figure 8.17 A HR-TEM image of an individual silver dopant particle next to a titania cluster. The scale bar is 5 nm. Source: Reprinted with permission from Dr Jakob B. Wagner, Lund University, Sweden Copyright (2013) Jakob B. Wagner.

8.4.4

Performance of the Antimicrobial Easy-to-Clean Coatings

The functionality of the surface has been characterized by Keskinen et al. (2006). The sample coatings were made on stainless steel and characterised by growing a biofilm on top and then illuminating the samples with UV-light. The model bacterium used in this study was Deinococcus Geothermalis, which is typically found in hot springs and is a primary bacterium to adhere on a surface. Other bacteria will easily attach to it and start to form

Figure 8.18 A high angle annular dark field TEM image of silver particles (bright spots). The scale bar is 10 nm. Source: Reprinted with permission from Dr Jakob B. Wagner, Lund University, Sweden Copyright (2013) Jakob B. Wagner.

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Cells cm–2

248

Dark

360 nm

Steel (uncoated); 78.8 %

Dark

360 nm

Titania; 97.3 %

Dark

360 nm

One-step; 94.1 %

Figure 8.19 Biofilm growth tests of LFS coated stainless steel. The percentage denotes the amount of destroyed bacteria cells after UV-illumination (λ = 360 nm, 24 h, 0.1 mW m–2 ) (column on the right) compared to the amount of original living bacteria before illumination on the same surface (column on the left). Source: Reprinted with permission from [Keskinen et al., 2006] Copyright (2006) Springer Science + Business Media.

a dense biofilm. Silver ions were introduced in order to prevent D. Geothermalis from attaching to the surface and thus boost the easy-to-clean effect. Even a small amount of silver dopant is sufficient to prevent some bacteria from attaching onto the surface. Here, a silver dopant of 1 wt% of titania was used to enhance the antibacterial functionality. Figure 8.19 shows the results from bacteria growth tests on LFS-coated stainless steel surfaces. The figure is divided into three parts. First, there is a reference sample without coating, then a titania-coated sample and finallya titania silver composite coating (onestep). The three cases are divided into two. The uncoated control sample has been kept in the dark and bacteria have been growing on it for 24 h. The bar on the right is a sample that has been illuminated for 24 h with UV light that has a wavelength of 360 nm and intensity of 0.1 mW m–2 , much less than solar UV-radiation outside on a sunny day. The percentage denotes the amount of bacteria cells that are destroyed after UV-illumination. There are two things to compare in Figure 8.19. First, we can compare the growth of bacteria on the surface by looking at the dark control samples. From these, it is obvious that less bacteria grow on the titania-coated easy-to-clean surface compared to the uncoated sample. Interestingly, a remarkably smaller amount of bacteria grows on the titania-silver composite coating. Silver, being widely used as an antibacterial agent, prevents bacteria cells from attaching onto the surface. Secondly, we can compare the photocatalytic efficiency of the samples. UV-light deteriorates bacteria, so we show the effect of UV-light only for the uncoated reference sample. However, looking at the LFS-coated easy-to-clean surface, one can see that 97% of the original bacteria cells are destroyed. Moreover, switching to the

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

249

titania-silver composite coating, the photocatalytic effect stays the same while 94% of the cells are destroyed after illumination for 24 h.

8.5

Summary

In this chapter, we have presented aerosol methods for the synthesis of nanoparticles. We demonstrated the liquid flame spray method as an aerosol-based coating method to synthesize and manufacture nanoscale coatings on surfaces. In the LFS process, the aerosol is formed in the gas phase, just prior to deposition. Subsequently, after formation, the particles are deposited into a nanostructured coating by gas phase diffusion and thermophoresis. Thus, the deposition process being rather gentle for the nanoparticles, the morphology of the film coating is often fairly porous. For the lowermost layer of particles, and in the case of very thin coatings, that is, submonolayer, the attachment of the nanoparticles onto the surface is due to van der Waals forces. If the surface is locally in the liquid phase, such as the case may be for a flame-treated float glass or a vitreous glaze, submerging of the particle and dissolution of the particulate compound into the matrix of the substrate is possible. We also demonstrated the use of liquid flame spray as a tool for synthesis of multi-component nanoparticles and coatings. By adding silver as a dopant for nanoscale titanium dioxide particles one can improve the antimicrobial properties of an easy-to-clean surface.

References Ahonen, P.P., Moisala, A., Tapper, U. et al. (2002) Gas-phase crystallization of titanium dioxide nanoparticles. J. Nanopart. Res., 4, 43–52. Allen, M.D. and Raabe, O.G. (1985) Slip correction measurements of spherical solid aerosol particles in an improved Millikan apparatus. Aerosol Sci. Technol., 4, 269–286. Aromaa, M., Keskinen, H. and M¨akel¨a, J.M. (2007) The effect of process parameters on the Liquid Flame Spray generated titania nanoparticles. Biomol. Eng., 24, 543–548. Backman, U., Tapper, U. and Jokiniemi, J.K. (2004) An aerosol method to synthesize supported metal catalyst nanoparticles. Synth. Met., 142, 169–176. Bickmore, C.R., Waldner, K.F., Baranwal, R. et al. (1998) Ultrafine titania by flame spray pyrolysis of a titanatrane complex. J. Eur. Ceram. Soc., 18, 287–297. Cannon, W.R., Danforth, S.C., Flint, J.H. et al. (1982a) Sinterable ceramic powders from laser-driven reactions: I, process description and modeling. J. Am. Ceram. Soc., 65, 324–330. Cannon, W.R., Danforth, S.C., Haggerty, J.S. and Marra, R.A. (1982b) Sinterable ceramic powders from laser-driven reactions: II, powder characteristics and process variables. J. Am. Ceram. Soc., 65, 330–335. Chen, S., Mason, M.G., Gysling, H.J. et al. (1993) Ultrahigh vacuum metalorganic chemical vapor deposition growth and in-situ characterization of epitaxial TiO2 films. J. Vac. Sci. Technol. A, 11 2419–2429. Choy, K.L. (2003) Chemical vapour deposition of coatings. Prog. Mater. Sci., 48, 57–170. Courtecuisse, V.G., Chhor, K., Bocquet, J.-F. and Pommier, C. (1996) Kinetics of the titanium isopropoxide decomposition in supercritical isopropyl alcohol. Ind. Eng. Chem. Res., 35, 2539–2545.

250

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Friedlander, S.K. (2000) Smoke, Dust and Haze: Fundamentals of Aerosol Dynamics, 2nd edn, Oxford University Press, USA, pp. 432. Fujishima, A. and Honda, K. (1972) Electrochemical photolysis of water at a semiconductor electrode. Nature, 238, 37–38. Grass, R.N. and Stark, W.J. (2005) Flame synthesis of calcium-, strontium-, barium fluoride nanoparticles and sodium chloride. Chem. Commun., 13, 1767–1769. Grass, R.N. and Stark, W.J. (2006a) Gas phase synthesis of fcc-cobalt nanoparticles. J. Mater. Chem., 16, 1825–1830. Grass, R.N. and Stark, W.J. (2006b) Flame spray synthesis under a non-oxidizing atmosphere: Preparation of metallic bismuth nanoparticles and nanocrystalline bulk bismuth metal. J. Nanopart. Res., 8, 729–736. Grass, R.N., Tsantilis, S. and Pratsinis, S.E. (2006) Design of high-temperature, gas-phase synthesis of hard or soft TiO2 agglomerates. AIChE J., 52, 1318–1325. Gribb, A.A. and Banfield, J.F. (1997) Particle size effects on transformation kinetics and phase stability in nanocrystalline TiO2 . Am. Mineral., 82, 717–728. Gross, K.A., Tikkanen, J., Keskinen, J. et al. (1999) Liquid flame spraying for glass coloring. J. Therm. Spray Technol., 8, 583–589. Hinds, W.C. (1999) Aerosol Technology: Properties, Behavior, and Measurement of Airborne Particles, 2nd edn, Wiley-Interscience, 504 pp. Hinds, W.C. (2005) Physical and chemical changes in the particulate phase, in Aerosol Measurement: Principles, Techniques and Applications, 2nd edn (eds P.A. Baron and K. Willeke), Wiley-Interscience, pp. 83–98. Keskinen, H., Aromaa, M., Heine, M.C. and M¨akel¨a, J.M. (2008) Size and velocity measurements in sprays and particle-producing flame sprays. Atomization Sprays, 18, 619– 644. Keskinen, H., M¨akel¨a, J.M., Aromaa, M. et al. (2006) Titania and titania-silver nanoparticle deposits made by Liquid Flame Spray and their functionality as photocatalyst for organicand biofilm removal. Catal. Lett., 111, 127–132. Keskinen, H., M¨akel¨a, J.M., Aromaa, M. et al. (2007a) Effect of silver addition on the formation and deposition of titania nanoparticles produced by liquid flame Spray. J. Nanopart. Res., 9, 569–588. Keskinen, H., M¨akel¨a, J.M., Heikkinen, R. et al. (2007b) Synthesis of Pd-alumina and Pd-lanthana suspension for catalytic applications by one-step liquid flame spray. Catal. Lett., 119, 172–178. Keskinen, H., M¨akel¨a, J.M., Vippola, M. et al. (2004) Generation of silver/palladium nanoparticles by liquid flame spray. J. Mater. Res., 19, 1544–1550. Kodas, T.T. and Hampden-Smith, M.J. (1999) Aerosol Processing of Materials, Wiley-VCH Verlag GmbH, 680 pp. Mooney, J.B. and Radding, S.B. (1982) Spray pyrolysis processing. Annu. Rev. Mater. Sci., 12, 81–101. Morita, H., Kihou, Y., Semba, K. and Taibu, T. (1999) Laser synthesis of organic materials in the gas phase: A noble method to prepare nanoparticles. J. Photopolym. Sci. Technol., 12, 95–100. M¨adler, L., Roessler, A., Pratsinis, S.E. et al. (2006) Direct formation of highly porous gassensing films by in situ thermophoretic deposition of flame-made Pt/SnO2 nanoparticles. Sensor. Actuat. B: Chem., 114, 283–295.

Liquid Flame Spray as a Means to Achieve Nanoscale Coatings

251

M¨akel¨a, J.M., Hellst´en, S., Silvonen, J. et al. (2006) Collection of liquid flame spray generated TiO2 nanoparticles on stainless steel surface. Mater. Lett., 60, 530–534. M¨akel¨a, J.M., Keskinen, H., Forsblom, J. and Keskinen, J. (2004) Generation of metal and metal oxide nanoparticles by liquid flame spray process. J. Mater. Sci., 39, 2783–2788. NNI, The National Nanotechnology Initiative (2009) Research and Development Leading to a Revolution in Technology and Industry. Supplement to the President’s FY 2010 Budget, National Nanotechnology Coordination Office, Arlington, VA, U.S.A. Pfender, E. (1999) Thermal plasma technology: Where do we stand and where are we going? Plasma Chem. Plasma Process., 19, 1–31. Pimenoff, J.A., Hovinen, A.K. and Rajala, M.J. (2009) Nanostructured coatings by liquid flame spraying. Thin Solid Films, 517, 3057–3060. Pitk¨anen, A., M¨akel¨a, J.M., Nurminen, M. et al. (2005) Numerical study of silica particle formation in turbulent H2/02 flame. IFRF Combustion J., article no. 200509, 29 pp. Pratsinis, S.E. (1998) Flame aerosol synthesis of ceramic powders. Prog. Energy Combust. Sci., 24, 197–219. Rajala, M., Janka, K. and Kykk¨anen, P. (2003) An industrial method for nanoparticle synthesis with a wide range of compositions. Rev. Adv. Mater. Sci., 5, 493–497. Samy El-Shall, M. and Edelstein, A.S. (1996) Formation of clusters and nanoparticles from a supersaturated vapor and selected properties, in Nanomaterials: Synthesis, Properties and Applications (eds A.S. Edelstein and R.C. Cammarata), Institute of Physics Publishing, U.K., pp. 11–54. Shannon, R.D. and Pask, J.A. (1965) Kinetics of the anatase-rutile transformation. J. Am. Ceram. Soc., 48, 391–398. Stark, W.J. and Pratsinis, S.E. (2002) Aerosol flame reactors for manufacture of nanoparticles. Powder Technol., 126, 103–108. Swihart, M.T. (2003) Vapor-phase synthesis of nanoparticles. Curr. Opin. Colloid Interface Sci., 8, 127–133. Tikkanen, J., Eerola, M., Pitk¨anen, V. and Rajala, M. (1997b) Menetelm¨a ja laite materiaalin ruiskuttamiseksi (Method and apparatus for spraying materials). Pat. FIN 98832 Appl. 954370 15.9.1995, in Finnish. Tikkanen, J., Gross, K.A., Berndt, C.C. et al. (1997a) Characteristics of the liquid flame spray process. Surf. Coat. Technol., 90, 210–216. Tsantilis, S. and Pratsinis, S.E. (2004) Soft- and hard-agglomerate aerosols made at high temperatures. Langmuir, 20 5933–5939. Ulrich, G.D. (1984) Flame synthesis of fine particles. Chem. Eng. News, 62 (32), 22–29. Wegner, K. and Pratsinis, S.E. (2003) Scale-up of nanoparticle synthesis in diffusion flame reactors. Chem. Eng. Sci., 58, 4581–4589. Wegner, K. and Pratsinis, S.E. (2005) Gas-phase synthesis of nanoparticles: Scale-up and design of flame reactors. Powder Technol., 150, 117–122. Zhang, H. and Banfield, J.F. (2000) Phase transformation of nanocrystalline anatase-torutile via combined interface and surface nucleation. J. Mater. Res., 15, 437–448.

9 Pulsed Laser Deposition of Surfaces with Tunable Wettability Evie L. Papadopoulou1,2 1

9.1

Institute of Electronic Structures and Lasers, Foundation for Research and Technology-Hellas, Greece 2 Current address: Istituto Italiano di Tecnologia, Genova, Italy

Introduction

In nature, many surfaces are highly hydrophobic and have self-cleaning capability. These surfaces include plants, the leaves of which are covered by waxy coatings, the chemical compositions of which render them hydrophobic [1], as well as insects [2]. Very often, these hydrophobic coatings provoke water repellency and prevent contamination by particles, resulting in self-cleaning effects of the surfaces. Technology often imitates nature, so over the last years the fabrication and study of surfaces that present special wetting properties has gained a lot of attention, opening the way to a variety of applications. The wetting properties of surfaces depend on both the free energy and the morphological structure of the surface in question. By changing one of these two factors one can tailor the surface wettability [3]. Therefore, by chemically modifying the free energy of a surface with a given geometrical structure, one can modify the wettability of the surface. Correspondingly, for a surface with specific free energy, one can tailor its wetting properties by tailoring its geometrical features. Furthermore, the switching between two wetting states of a surface, that is, between hydrophobicity and hydrophilicity, can be achieved by applying an external stimulus to the surface. Such responsive surfaces, where the initial hydrophobic state is rendered hydrophilic upon application of the external stimulus, add functionality to potential applications. The key issue here is the reversible control of the

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach, First Edition. Edited by Walid A. Daoud. © 2013 John Wiley & Sons, Ltd. Published 2013 by John Wiley & Sons, Ltd.

254

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

switching, that is, the restoration of the initial hydrophobic state upon removal of the external stimulus. For many applications it would be advantageous to dynamically control the wetting properties of a surface, such as the contact angle, the droplet mobility or the effective area of the liquid/olid interface [4–7]. Such control leads to functional surfaces, the wettability of which can be reversibly and controllably switched from hydrophobic to hydrophilic. Potential applications of such surfaces include self-cleaning surfaces, lab-on-chip devices, smart windows, microfluidic devices, adjustable lenses and so on [8–12]. The development of such responsive surfaces requires the tailoring of their wettability. To this end, it is evident that the more hydrophobic a surface is initially, the more pronounced the switching to the hydrophilic state will be. The initial hydrophobicity can be enhanced by covering the surface with an appropriate coating with low surface energy. Nevertheless, in order to attain superhydrophobicity, it is essential to combine the appropriate hydrophobic coating with a morphology that presents hierarchical roughness on micro- and nano-length scales. In nature this is a common strategy known as the “lotus effect” [1]. The sacred lotus (Nelumbo nucifera) leaves owe their self-cleaning capability to the superhydrophobic nature of their surface. This surface is composed of a hierarchical structure, on the microand nano- length scales, covered by a hydrophobic wax layer. The high roughness leads to a reduced contact area between the surface and the liquid drop, with droplets residing only on the surface tips, whereas the hydrophobic coating lowers the free surface energy. Hence, as the droplet rolls off the leaf, it collects any dirt particles residing on it. During the last years, due to the technological advancement, many ways to fabricate hydrophobic and superhydrophobic surfaces have been developed [5, 13]. One of the simplest methods for growing thin film or nanostructured coatings, with reproducible and controlled results, is pulsed laser deposition (PLD). However, despite its facile and widespread use, it has not been used to its full potential in growing hydrophobic coatings. There are only a few reports connecting the produced nanostructures and coatings with wetting properties. From the abundance of materials with interesting wetting properties, we will focus on metal oxides. In particular ZnO and TiO2 have drawn considerable attention, due to their outstanding electrical and optical properties. In addition, their photocatalytic properties, in conjunction with their special wetting properties, make these materials very promising for applications. In particular, the hydrophilicity that can be induced by applying external stimuli, can be used to realize self-cleaning surfaces [14]. In summary, in this chapter, we concentrate on nanostructured, metal oxide surfaces grown by pulsed laser deposition (PLD). In addition, a combined two-step approach is presented for the fabrication of surfaces with hierarchical roughness on the micro- and nano-length scales.

9.2 9.2.1

Basic Theory of Wetting Properties of Surfaces Planar Surfaces

When a small liquid droplet is placed on a flat, solid surface, the droplet will either spread on the surface, causing the wetting of the surface, or remain as a droplet [15]. The shape

Pulsed Laser Deposition of Surfaces with Tunable Wettability

255

γlg γls

CA

γag

Figure 9.1 Droplet residing at planar surface.

of the droplet is defined by the equilibrium of the surface tensions (γ ) acting at the threephase contact line formed along the solid/liquid (SL), liquid/gas (LG) and solid/gas (SG) interfaces, as drawn in Figure 9.1, leading to the well-known Young’s equation [16, 17]: cos θ0 =

γSV − γSL γLG

(9.1)

Here, θ 0 is the equilibrium angle that characterizes the contact. When θ 0 > 90◦ the surface is called hydrophobic, and when θ 0 < 90◦ the surface is called hydrophilic. 9.2.2

Rough Surfaces

The effect of surface roughness has been theoretically addressed and described by two models. In the Wenzel model [18], the liquid is assumed to completely penetrate within the rough surface (Figure 9.2a) and the apparent contact angle, θ W , is given by: cos θW = rW cos θ0

(9.2)

where rw is the ratio of the actual over the projected surface area of the substrate, and θ 0 is the intrinsic contact angle on a flat surface of the same nature as the rough one. Since rW is always greater than unity, it results in the contact angle of an initially hydrophilic, flat surface (i.e. θ 0 < 90◦ ) decreases when its surface is roughened (i.e. θ W < θ 0 ). In contrast, roughening an initially hydrophobic, flat surface (θ > 90◦ ) always increases its hydrophobicity (θ W > θ 0 ). On the other hand, Cassie and Baxter [19] assumed that the liquid does not completely penetrate the roughened solid, residing instead on top of the protrusions (Figure 9.2b). As a result, a liquid/gas interface is created underneath the droplet. In this case, the contact

(a)

(b)

Figure 9.2 Droplet residing at a rough surface, in (a) the Wenzel state and (b) the Cassie– Baxter state.

256

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

angle, θ CB , is the average between the value on air (i.e. 180◦ ) and the value on the flat solid (i.e. θ 0 ), and is given by: cos θCB = −1 + f (1 + cos θ0 )

(9.3)

where f defines the fraction of the projected solid surface wetted by the liquid. Since f is always lower than unity, this model predicts the enhancement of hydrophobicity, independent of the value of the initial contact angle θ 0 . The smaller the solid–liquid contact area (i.e. the smaller f), the larger the increase in the resulting contact angle. Recently, the problem of the effect of binary roughness on the wetting properties of a surface was addressed [20–22]. Assuming that the two types of roughness occupy the same fraction of surface, f, that is, they are homothetic [23], the Cassie–Baxter equation becomes:  = −1 + f 2 (1 + cos θ0 ) cos θCB

(9.4)

Also in this case f < 1 and by comparing Eqs. (9.3) and (9.4), it comes out that the contact angle of a surface presenting binary roughness is always higher than in the single roughness case.

9.3

Roughening a Flat Surface

In general, the fabrication of hydrophobic or superhydrophobic surfaces involves the growth of a surface with high roughness and its subsequent coating with a material of low surface energy. However, as we have learnt from nature, the low surface energy coating is not a prerequisite for achieving superhydrophobicity. What is of equal importance is the ability to control the morphological structure underneath the coating, on the micro- and nanolength scales, that is, the hierarchical roughness of the surface. For example, the leaves of the lotus plant have contact angle (CA) of about 160◦ . The large value of the CA cannot be accounted for by the paraffin wax, with which the leaves are covered, that mostly contains –CH2 – groups, and not the lower energy –CH3 groups or fluorocarbons [5]. In contrast, it is the synergistic effect of the hierarchical roughness and the low surface energy coating that explains the superhydrophobicity of the lotus leaves. Indeed, the surface of the lotus leaves are comprised of micrometre-scale papillae, each of which is covered by branch-like nanostructures, resulting in hierarchical roughness on the micro- and nanolength scales [24]. Experimentally, Zhang et al. [25] fabricated ZnO surfaces, exhibiting micro-roughness, nano-roughness and hierarchical roughness on the micro- /nano- length scales, using wet chemistry. They found that the contact angle on the first surface had the smallest value, while CA attained the largest value on the latter surface. This fact alone gives today’s technology immense opportunities for the fabrication of surfaces with hierarchical roughness and potential superhydrophobicity. Artificial hydrophobic or superhydrophobic surfaces have thus been prepared by several methods in recent years [5, 13]. In the following, we will focus on the fabrication of nanostructured surfaces with special wetting properties grown by pulsed laser deposition (PLD).

Pulsed Laser Deposition of Surfaces with Tunable Wettability

257

Rotatable substrate holder Focusing lens

Ablation Plume

Pulsed laser beam Target

To vaccum pumps

Figure 9.3 Schematic diagram of PLD. Source: Reproduced from M. Ashfold et al., copyright

 C 2003 with permission of The Royal Society of Chemistry.

9.3.1

PLD Technique Overview

One of the simplest methods of growing films, as well as nanostructures, is pulsed laser deposition (PLD) [26]. Since the mid-1960s, when the technique was developed, PLD has become increasingly popular for growing thin films and nanostructures due to the fact that it can be used for the deposition of a very wide range of materials, both inorganic [27] and organic [28]. Figure 9.3 shows a schematic diagram of the PLD technique: a pulsed laser beam is focused on the target. The target absorbs the laser energy and vaporizes locally, creating a plume that contains neutral atoms, electrons, ions and molecules. The plume is ejected forward and propagates towards the substrate, which is placed parallel to the target. The energetic species comprising the plume, exhibit high reactivity and high surface mobility when they reach the substrate, and thus the growth of the material takes place. During the deposition, the target is rotating in order to avoid ablation of the same, which will cause deflection of the plume [29]. The growth usually takes place either in vacuum or in an ambient gas, usually O2 or Ar. Normally, ablation takes place when the energy of the laser beam exceeds a threshold value that depends on the material. The main advantage of the PLD technique is the transfer of the stoichiometry of the target material, even for chemically complex materials. This fact makes PLD an attractive deposition technique. The parameters that must be tuned in order to control the deposition include: (i) laser beam wavelength and duration, (ii) laser beam energy, (iii) substrate temperature, (iv) presence and nature of gas, (v) partial pressure of gas. By tuning these parameters one can tailor the structural characteristics [30] (e.g. crystallinity, surface morphology, phase) as well as other physical properties, such as electrical properties [31, 32] (e.g. conductivity, carrier concentration, carrier mobility) or magnetic properties [33]. 9.3.2

Nanostructures Grown by PLD

In this section is given a short review of different metal oxide nanostructures fabricated by PLD and their corresponding wetting states.

258

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 9.4 SEM images of KrF-deposited films in 1 Pa of oxygen at a substrate temperature of (a) 500◦ C, (b) 600◦ C and (c) 700◦ C. Source: With kind permission from Springer C 2008. Science + Business Media: D. Valerini, 

Nanostructured ZnO thin films have been produced by nanosecond KrF (248 nm) laser ablation. The dependence of the depositions on the various parameters, as described above, has been investigated in depth. Depositions at constant oxygen pressure (10−2 mbar) but with increasing substrate temperature (500–700 ◦ C) show that the roughness of the surfaces increases, and that hexagonal, pyramid-like structures emerge from the surface when the substrate temperature reaches 700 ◦ C [34]. The field emission scanning electron microscopy (FE-SEM) images are shown in Figure 9.4. At such high substrate temperatures, the deposited species have enough kinetic energy to form clusters that will grow in a hexagonal pattern, due to the wurtzite structure of ZnO. In a similar fashion, depositions at a constant substrate temperature (650 ◦ C) and fluence (1.5 J cm–2 ), describe the important role of the oxygen pressure [35]. In Figure 9.5 the FE-SEM images depict three nanostructured films grown in different oxidizing environments. As the pressure increased from 5 × 10−4 to 5 × 10−2 mbar, the surface attained a well-defined morphology, resulting in the formation of almost hexagonal particles for the highest pressure implemented. The increase in oxygen pressure also resulted in a simultaneous increase in the particle diameter. At the higher O2 content, the structures exhibit columnar characteristics, as seen in the crosssection image of Figure 9.5, suggesting that the growth has a preferred orientation, that is, perpendicular to the substrate (along the c–axis). Mean roughness, measured by atomic force microscopy (AFM), was also increased by increasing oxygen pressure, by almost an order of magnitude. At such high pressures, the ablated species condense in the plume and generate nanoparticles, which subsequently create the source of growth of nanorods after they reach the substrate [36, 37]. Furthermore, Valerini et al. [34] have grown hexagonal nanostructures at lower temperatures (∼550 ◦ C) using a nanosecond ArF (193 nm) laser for ablation. In this case, the

Pulsed Laser Deposition of Surfaces with Tunable Wettability

259

Figure 9.5 Top view FE-SEM images of the samples prepared at (a) 5 × 10-4 mbar, (b) 5 × 10-3 mbar and (c) 5 × 10-2 mbar of partial oxygen pressure. The insets depict the cross-section images of the respective sample. All scale bars are 100 nm. Source: Reprinted from E.L. C 2009, with permission Papadopoulou et al. Thin Solid Films 518, 1267 (2009), copyright  from Elsevier.

nanostructures are decorated by a much narrower tip, exhibiting hierarchical structure. The ArF (193 nm) laser beam has higher photon energy than the KrF (248 nm) one, which means that the target material is ablated by higher energy photons, resulting in more energetic species in the plume. Hence, the ablated species acquire the required kinetic energy before they reach the substrate, forming nanostructures even at lower temperatures. Especially, ZnO is known for the ease with which it can be grown into one-dimensional (1D) structures, that is, nanorods, nanowires, nanotubes and so on. Sun et al. [38] have investigated the growth of ZnO nanorods by PLD with a nanosecond ArF (193 nm) laser on Si substrates. The dependence on the fluence was investigated and it was found that nanorods were grown at low fluences, while increasing fluence led to a progressive evolution from nanorod growth to film growth. The dependence of the growth on the substrate type has also been explored [39]. On a Si substrate the nanorods are well aligned and exhibit needle-like morphology, 200–800 nm long and 20–60 nm wide. However, when a thin ZnO seed layer was used, the nanorods were found to grow longer (1–1.2 μm) and thinner (6–20 nm). In the latter case, the ZnO seed layer provides the nucleation sites for the growth of thinner rods. Okada et al. [40, 41] have used nanosecond KrF (248 nm) ablation to grow ZnO nanorods. The nanorods were approximately 6 μm long and ∼200–500 nm wide. The growth took place in relatively high oxygen partial pressure (∼1 mbar) and high substrate temperature

260

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Figure 9.6 SEM images of ZnO nanowalls grown on sapphire substrate with PLD under 268 torr of pressure at 1000 ◦ C: (a) top view and (b) tilted view. Tubular ZnO are marked with dashed circles. Source: Reprinted with permission from Cao et al. J. Phys. Chem. C 113, 10975 C 2009, American Chemical Society. (2009). Copyright 

(∼700 ◦ C). As said before, at these high pressures the nanoparticles that are formed in the plume are the major species transported on the substrate and initiate the nanorod growth. By controlling thus the size distribution and density of these nanoparticles, one can control the growth of the nanorods. Furthermore, Tien et al. [42] studied the effect of temperature on the nanorod growth. Growth at elevated temperatures (∼800 ◦ C) provided the deposited species with sufficient kinetic energy, thus increasing their surface mobility so as to reach low energy nucleation sites. The resulted nanorods had narrow ends, of about 50–90 nm and height ∼6 μm. Two-dimensional (2D) nanostructures have also been fabricated by PLD [43, 44]. ZnO nanowalls, with thickness of tens of nanometres, and height of a few micrometres, were grown in a high pressure PLD procedure, using a nanosecond KrF (248 nm) laser, and either Ar or O2 as carrier gas, at high partial pressure. The 2D nanowalls can be seen in Figure 9.6. Femtosecond lasers have also been used for nanostructure growth with PLD. The main advantage of femtosecond ablation is the absence of interaction between the laser beam and the expanding plume, and the non-thermal melting [45]. ZnO nanowires have been fabricated by femtosecond PLD, with a Ti:Sapphire (800 nm) [46] and a laser with wavelength 1.05 μm [47]. In both works, seed layers were acting as catalysts for the nanorod growth. In the latter case, the ZnO seed layer and the longer ablation wavelength resulted in better alignment of the nanorods. A two-step method has also been used to fabricate hierarchical ZnO nanostructures on Si [48]. The first step involves microstructuring of the flat silicon (Si) surface by irradiating the Si surface by a regenerative amplified Ti:Sapphire (800 nm) delivering 150 fs pulses at a repetition rate of 1 kHz. During the irradiation process, a constant pressure of an etching gas (SF6 ) was maintained in a vacuum chamber. The sample was mounted normal to the laser beam. The irradiation results in conical structures in the Si surface, of micrometre scale. The experimental details, as well as how the morphological characteristics of the microstructures (in particular height and density) depend on the irradiation fluence, are described in detail in Ref. [49]. In the second step, the patterned surfaces are coated with ZnO, using PLD, in a flowing oxygen environment. The surface

Pulsed Laser Deposition of Surfaces with Tunable Wettability

261

Figure 9.7 (a) Side FE-SEM image of Si microstructure, irradiated at a laser fluence of 0.17 J cm–2 , (b) Sample A: the same microstructured Si, covered with ZnO, after PLD, (c) Sample B: Si microstructure, irradiated at 2.1 J cm–2 , covered with ZnO, after PLD. The insets show the ZnO nano roughness, in one micro-cone (scale bar is 100 nm). Source: Reprinted with permission from E.L. Papadopoulou et al. J. Phys. Chem. C 113, 2891 (2009). Copyright  C 2009, American Chemical Society.

morphology, prior to and after the PLD process, is depicted in Figure 9.7. In Figure 9.7a the microstructured Si surface, irradiated by the smallest fluence is shown. The surface morphology comprises a highly uniform and densely packed array of micrometre-sized conical structures. At higher fluences, the aspect ratio of the conical structures increases, leading to a significant enhancement of the overall roughness. Figure 9.7b and c depict FE-SEM images of the morphology of Si, microstructured at the lowest (Sample A) and highest (Sample B) fluence, respectively, acquired after the deposition of the nano-grained ZnO film. As clearly seen, a significant enhancement of the nanoscale roughness is attained. The micrometre-scale conical structures have been decorated by nano-sized protrusions, resulting in a hierarchically rough surface. The nanoscale features are more pronounced in Sample A than in Sample B. In both cases, the ZnO coating is deposited under identical conditions to those used for the nanostructured thin film shown in Figure 9.5c. The initial wetting state of ZnO is found to depend to some extent on the growth parameters. For example, when the deposition takes place at low oxygen content (5 × 10−4 mbar), resulting in the surface depicted in Figure 9.5a, it is slightly hydrophilic, exhibiting CA ∼ 75◦ . At higher oxygen pressures (Figure 5b and c), the surfaces attain hydrophobicity, with CA ∼ 95◦ . However, when ZnO is deposited on the microstuctured Si, resulting in a hierarchical rough surface, there is a significant increase in the contact

262

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

angles which, for the samples depicted in Figure 9.7, attain the values ∼120◦ for Sample A and ∼140◦ for Sample B. PLD has also been used for the growth of other metal oxide materials, apart from ZnO. For example, Co3 O4 nanorods have been fabricated by PLD using the third harmonic of a nanosecond Nd:YAG laser (355 nm) for the ablation, at room temperature and with subsequent annealing [50]. The surfaces exhibit hierarchical roughness on the micro-length scale. The micrometre roughness was induced by polystyrene spheres immobilized on the substrate and used as templates. The secondary roughness on the nano-length scale was induced by depositing Co3 O4 nanorods by PLD, and was improved by increasing the oxygen pressure during deposition. The surfaces were initially superhydrophilic, but became superhydrophobic with CA as high as 152.6◦ upon chemical modification of the surface with fluorosilane. The same group has also grown columnar arrays of hematite (Fe2 O3 ) exhibiting hierarchical roughness using the same PLD protocol [51]. Ngom et al. [52] have improved the initial hydrophilicity of the as-deposited ZnO nanorods by W doping. The initially hydrophilic ZnO nanorod surface (∼70◦ ) became hydrophobic (∼110◦ ) upon doping with 2% W. Another metal oxide that has attracted considerable attention is TiO2 . TiO2 exists in three crystallographic phases: rutile, anatase and brookite. In this case, PLD gives us the opportunity to control easily not only the nanostructured morphology of the surface, but also the crystallographic phase of the nanostructure [30]. This is very important, since it has been shown by various researchers that the photocatalytic behaviour of TiO2 depends to some extent on the crystallographic phase [53,54]. For example, Walczak et al. [30] showed that by increasing the laser fluence, the anatase phase is diminished, and by decreasing the oxygen pressure during the deposition one can obtain the pure rutile phase. In contrast, Gyorgy et al. [55, 56] fabricated TiO2 films with pure anatase phase, at lower substrate temperatures and lower fluence. Li et al. [57] performed PLD under high oxygen pressure, using a template of polystyrene (PS) spheres, as described in the previous section, in order to induce hierarchical structure. The procedure resulted in an ordered nanocolumn array with a hexagonal closed packed (HCP) arrangement, where the TiO2 column grew on the PS nanospheres. The nanocolumns exhibit a very rough surface, comprised of many nanobranches growing vertically on the surface of the PS spheres. Despite the rutile target, the grown surfaces were amorphous. In addition, the hierarchical TiO2 surfaces were superhydrophilic, with CA ∼ 0◦ . Thin films, without the PS template, that were grown under identical conditions, exhibited CA ∼ 15◦ . Hence, the initially hydrophilic surfaces became superhydrophilic upon roughening. It is very interesting that PLD has also been used for growing different materials on less expected substrates, like fabrics. Daoud et al. [58] have used PLD to deposit Teflon on cellulosic fibres. The surface of the textile was converted from superhydrophilic (CA ∼ 0◦ ) to superhydrophobic (CA = 151◦ ) after the deposition took place in vacuum. Popescu et al. [59] recently used a similar PLD method to control the wetting properties of ZnO grown on hydrophilic textiles. They also used a nanosecond KrF laser (248 nm), to deposit thin films or nanoparticles on cotton/polyester fabrics. They found that when the deposition took place in an oxygen ambient, the resulting surfaces were hydrophilic. However, when deposition took place in vacuum, the textile surfaces had a contact angle as high as 157◦ . Despite our comments on the wetting properties for a few of the nanostructures developed by PLD, most of the aforementioned nanostructures have not been tested for their wetting

Pulsed Laser Deposition of Surfaces with Tunable Wettability

263

properties. Hence, the scope for potential investigations and the room for improvement in the existing knowledge on wetting properties is evident.

9.4

Switchable Wettability

Here, we will focus on responsive surfaces, grown by PLD, where changes in the surface wettability can be triggered by external stimuli. The resulting change is normally measured by monitoring changes in the CA between the droplet of water, or any aqueous solution, and the surface. Such responsive surfaces offer new opportunities for manipulating liquids. As discussed in the Section 9.1, a great variety of applications can be developed by utilizing responsive surfaces. A variety of external stimuli has been used [7] to induce hydrophilicity in hydrophobic surfaces, such as electric field [60], UV light [66], temperature [61], mechanical bending [62] and so on. To the best of our knowledge, only UV light and the electric field have been used to trigger wettability changes in samples grown by PLD. 9.4.1

Photoinduced Wettability on PLD Structures

Hydrophilicity can be induced in some hydrophobic surfaces upon irradiation with UV light. The irradiated, hydrophilic surface, returns to its initial, hydrophobic state after storage in the dark for a few days. It is known that many metal oxides can be photo-switched between the two wetting states, by alternating UV irradiation and dark storage [63–65]. The ZnO nanostructured films presented in Figure 9.5 were hydrophilic immediately after deposition, having CA ∼ 45◦ . However, after a few days storage in the dark, the films become hydrophobic, that is, the surface energy of the as-deposited film changes after deposition. It seems that immediately after growth, the as-deposited material is in an energetically metastable phase and the equilibrium wetting state is reached after the interaction of the surface with the ambient. Figure 9.8a depicts the contact angle evolution with UV irradiation time for the aforementioned samples. The sample grown at the lower oxygen pressure (5 × 10−4 mbar) is slightly hydrophilic, having an initial contact angle of 73◦ . 100 5 × 10–4mbar 5 × 10–3mbar 5 × 10–2mbar

80 70 60

5 × 10–4mbar 5 × 10–3mbar 5 × 10–2mbar

1 CA/(CA)L=0

CA (degrees)

90

0.8

0.6

50 40 (a)

0.4 0

40

120 80 UV exposure time (min)

160 (b)

0

40

80

120

160

UV exposure time (min)

Figure 9.8 (a) Water contact angle dependence on the UV irradiation time, for the nanostructured films presented in Figure 9.5; (b) the normalized contact angle dependence. Source: C 2009 Reprinted from E.L. Papadopoulou et al. Thin Solid Films 518, 1267 (2009), copyright  with permission from Elsevier.

264

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

After dark storage

As deposited

UV Irradiation Dark storage or Heating

Figure 9.9 Various stages of the water contact angle. The transition between the two wetting states is evident. Source: Reprinted from E.L. Papadopoulou et al. Thin Solid Films 518, 1267 C 2009 with permission from Elsevier. (2009), copyright 

The samples grown at partial oxygen pressure 1 and 2 orders of magnitude higher have a contact angle of 95◦ . Under UV irradiation for 2 h, the slightly hydrophilic surface shows a rather weak response, the CA reaching about 45◦ (i.e. ∼38% change). The hydrophobic surfaces also become hydrophilic, acquiring a CA< 50◦ (i.e. 52% change). By dividing each CA value of Figure 9.8a by the value at t = 0, we obtain the normalised contact angle. In Figure 9.8b the normalized CA evolution is shown. Evidently, the CA reduction rate, which is a measure of the efficiency of the light-induced process, is higher for the roughest samples. The aforementioned changes in the wetting states are reversible, and storage in the dark for about a week restores the initial wetting state of the surface. Interestingly, heating the samples to 200 ◦ C for 1 h, also leads to the restoration of the initial wetting state. In Figure 9.9, the images of the droplets at various stages of the process are shown. It has been shown that hierarchical roughness amplifies the photo-switching effect [48, 66, 67]. In the previous section, we described a method for growing hierarchically rough surfaces, on the micro-/nano-length scales, by first structuring flat Si and subsequently depositing ZnO by PLD. Hence, the micrometre-scale conical structures are decorated by nano-sized protrusions, resulting in a hierarchically rough surface. The nanoscale features are more pronounced in Sample A, than in Sample B. The corresponding contact angles were measured to be ∼120◦ for Sample A and ∼140◦ for Sample B. The normalized contact angle evolution with UV irradiation time for the different ZnO structures is depicted in Figure 9.10. Both structured samples exhibit a significant photo-induced transition to superhydrophilicity, reaching almost 0◦ (complete wetting) in a short time. In contrast, the nanostructured ZnO thin film shows a weak response to UV irradiation, as the wetting angle change in this case is much smaller for the same irradiation time. The hydrophobicity for Samples A and B was restored within 24 h of storage in the dark, whereas the corresponding flat sample required several days of dark storage in order to return to its initial wetting state. Alternatively, as in the case of the nanostructured films, thermal heating at 200 ◦ C for 1 h can switch all surfaces to their original hydrophobic state (Figure 9.11a). All samples were subjected to numerous switching cycles without observing any deterioration of either the irradiation efficiency or the reversibility behaviour (Figure 9.11b). Since the initial contact angle of the nanostructured film is lower than 90◦ while that for the samples presenting hierarchical roughness is higher than 90◦ , it follows that the structured surfaces are consistent with the Cassie–Baxter model, since this is the only model that predicts a rise in the contact angle upon structuring of an initially hydrophilic surface. Hence, Eq. (9.4) was used to calculate the f parameter for the two samples, resulting in the values f ∼ 0.7 for Sample A, and f ∼ 0.5 for Sample B. Therefore, Sample B exhibits higher total roughness and, being in a Cassie–Baxter state, is less wetted by the water droplet. The evolution of contact angle in both samples is gradual, at short illumination

Pulsed Laser Deposition of Surfaces with Tunable Wettability

265

1.0

CA / (CA)t=0

0.8 0.6 0.4 Sample A Sample B flat ZnO/Si

0.2 0.0

0.1

1 Time (min)

10

Figure 9.10 The normalized water contact angle dependence on the UV irradiation time for ZnO deposited on microstructure Si (Sample A and Sample B). Flat ZnO/Si refers to the nanostructure film, depicted in Figure 9.5(c). Source: Reprinted with permission from C 2009, American E.L. Papadopoulou et al. J. Phys. Chem. C 113, 2891 (2009). Copyright  Chemical Society. (a) UV Irradiation

Dark Storage or Heating

Contact angel (deg)

150

100

50 UV

0 (b)

0

Dark UV Heat 1

Dark UV Dark UV Heat Heat 2 3 Cycle number

Dark UV Dark UV Heat Heat 4 5

Figure 9.11 (a) Images depicting the droplet of water on Sample B before (left panel) and after (right panel) UV irradiation, (b) restoration of the initial hydrophobicity occurs after dark storage or heating. The switching can take place several times, and no deterioration of the effect is observed. Source: Reprinted with permission from E.L. Papadopoulou et al. J. Phys. C 2009, American Chemical Society. Chem. C 113, 2891 (2009). Copyright 

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Composite roughness factor, f

266

Sample A Sample B

1.0

Wenzel State 0.8

Cassie-Baxter State

0.6

0.4

0

5

10

15

Time (min)

Figure 9.12 The dependence of the roughness factor, f, on the UV irradiation time for the ZnO samples exhibiting hierarchical roughness. Source: Reprinted with permission from E.L. C 2009, American ChemPapadopoulou et al. J. Phys. Chem. C 113, 2891 (2009). Copyright  ical Society.

times, followed by an abrupt change taking place at wetting angles lower than ∼90◦ . This sharp transition towards superhydrophilicity suggests that the wetting state of the drop switches from the Cassie–Baxter to the Wenzel one, as the latter is the model predicting the possibility of superhydrophilicity for very rough surfaces. The factor f can be calculated by substituting the respective contact angles for the different exposure times in Eq. (9.4). The resulting diagram, in Figure 9.12, depicts the corresponding evolution of the composite factor f up to the transition to the Wenzel state. The transition takes place when f takes values higher than unity, meaning that the Cassie–Baxter model is not valid. It is interesting, perhaps, to notice that while nanostructured films do not become superhydrophilic, even at long UV illumination times, hierarchically structured samples do. Furthermore, it seems that Sample B, which has higher micro-scale roughness, reaches superhydrophilicity much more abruptly, while Sample A, which has more pronounced nanoscale roughness, reaches superhydrophilicity at shorter times. This might be a suggestion that the nanoscale roughness plays an important role in the efficiency of the switching process [48]. The mechanism of the photo-induced hydrophilicity of surfaces in the case of wide band gap metal oxides, such as ZnO and TiO2 , was explained by Sun et al. [63] in 2001 in the frame of creation of oxygen vacancies. It is known that when the surface of these oxides is illuminated by UV light, electron–hole pairs are generated in their lattice. Some of the holes react with the lattice oxygen, forming oxygen vacancies at the surface, described in Eqs. (9.5)–(9.7): ZnO + 2hν → 2h+ + 2e− O2− + h+ → O− 1 1 − + O1 + h → O2 + ♦ 2 where ♦ is the oxygen vacancy.

(9.5) (9.6) (9.7)

Pulsed Laser Deposition of Surfaces with Tunable Wettability

267

At the same time, some electrons react with the metal ion (Zn2 + ) in the lattice, forming Zn + defective sites at the surface (surface trapped electrons): − Zn2+ + e− → Zn+ s + O2

(9.8)

These defective sites tend to react with oxygen molecules adsorbed on the surface when in ambient: − 2+ Zn+ s + O2 → Zns + O2

(9.9)

In the case of TiO2 , the mechanism of defective sites at the surface is similar [63, 68]. Meanwhile, water and oxygen may compete to dissociatively adsorb on these vacancies. The Zn + defective sites are kinetically more favorable for hydroxyl group adsorption than oxygen adsorption. Consequently, the hydrophilicity of the surface is increased, and thus the contact angle is reduced. It has also been demonstrated that after the hydroxy group adsorption, the surface becomes energetically unstable. Since the oxygen adsorption is thermodynamically favoured, it is more strongly bonded on the defect sites than the hydroxy group. Therefore, the hydroxy groups adsorbed on the defective sites can be replaced gradually by oxygen atoms when the UV-irradiated samples are placed in the dark. Subsequently, the surface regains its original wetting state (before any UV irradiation). Heat treatment accelerates the removal of surface hydroxy groups and, as a result, the hydrophilic surface converts faster to the hydrophobic one. It is shown, hence, how surface chemistry and surface roughness act in conjunction in the case of photoinduced wettability changes: the surface chemistry provides a photosensitive surface, which can be switched between hydrophilicity and hydrophobicity, while the surface roughness enhances these properties. 9.4.2

Electrowetting on PLD Structures

Another way to trigger wettability response on hydrophobic surfaces is to apply an electric field between the surface in question and a droplet residing on it. This phenomenon is called electrowetting. Electrowetting was first applied to metal surfaces [69], where electrolysis of the liquid droplets occurs after the application of a few hundred millivolts. To overcome this difficulty, a dielectric layer was used in order to coat the conducting surface, resulting in larger changes in the wettability [70]. The phenomenon was called electrowetting-ondielectric (EWOD). To date, electrowetting experiments have been focused on Si [60, 71–73], carbon nanostructures [74–76] and polymeric materials [77–79]. Only very recently have there been reports on EWOD where metal oxides are used as lower electrodes [80–83]. Even though PLD is a widespread technique for growing nanostructures, it has not been exploited regarding EWOD. The basic electrowetting set-up is shown in Figure 9.13. An external electric field is applied between the hydrophobic surface (lower electrode) and the droplet resting on it (upper electrode), via a metallic wire (e.g. Au or Pt) immersed in the droplet. Upon application of the electric field between the two electrodes, electric charge is accumulated at the solid/liquid interface, lowering the surface energy, and consequently rendering the initially hydrophobic surface more hydrophilic [84]. When a dielectric layer intervenes between the lower and the upper electrodes, the charge accumulation at the interface is

268

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach Platinum upper electrode Droplet

Hydrophobic layer

γLG Dielectric layer

γSL

θ0

V

γSG

Lower electrode Substrate

Figure 9.13 A schematic representation of EWOD. The solid line of the droplet shows the shape of the droplet before the application of the electric field, while the dashed line shows the shape of the droplet after the application of the electric field. Source: Reprinted with permission C 2008, American Chemical from J.L. Campbell et al. Langmuir 24, 5091 (2008). Copyright  Society.

much higher, resulting in greater changes in the contact angle, and thus the electrowetting effect is more pronounced. The dependence of the surface tension (γ ) and the applied potential (V) is described by Lippmann’s equation: 1 γ = γ0 − cV 2 2

(9.10)

where γ 0 is the solid–liquid interfacial tension at zero potential, c is the capacitance per unit area and V is the applied potential. The capacitance equals c = εr ε0 /d where εr is the dielectric constant of the intermediate layer of thickness d and ε0 is the dielectric constant of air. The shape of a liquid drop in contact with a flat surface is described through the precise equilibrium that results from the balance between the surface tensions (γ ) at the three-phase contact line formed along the solid/liquid (indicated by γ SL ), liquid/gas (γ LG ) and solid/gas (γ LG ) interfaces. The force balance leads to the well-known Young’s equation: γSL = γSG − γLG cos(θ0 )

(9.11)

From Eqs. (9.10) and (9.11), one obtains the so called Lippmann–Young’s equation [85, 86]: cos(θ ) = cos(θ0 ) +

1 1 1 1 ε0 εr 2 V cV 2 = cos(θ0 ) + 2 γLG 2 γLG d

(9.12)

where θ 0 is the CA at zero applied potential and θ is the CA at applied potential V. Evidently, from Eq. (9.12), as the applied voltage, V, increases, the contact angle of the droplet decreases, rendering the surface more hydrophilic. As shown before, in the case of roughened surfaces, the drop does not wet the whole area under its radius, but only a part of it, as described by the theories of Wenzel and

Pulsed Laser Deposition of Surfaces with Tunable Wettability

269

Cassie–Baxter (Eq. (9.2)–(9.4)). The roughness factor, f, must, therefore, be included in the above expression, which becomes: cos(θ ) = cos(θCB ) +

1 1 ε0 εr 2 f V 2 γLG d

(9.13)

Equation (9.13) describes EWOD in the case of superhydrophobic surfaces, where the initial wettability is described by the Cassie–Baxter model [87, 88]. From Eqs. (9.12) and (9.13) it is obvious that for high enough applied voltages, total wetting is possible. However, in reality, the contact angle saturates after a threshold voltage value. Moreover, reversibility of the droplet to its initial state should also be possible upon removal of the applied voltage. This is not the case, however, and reversibility is still a challenge. Reversibility of EWOD on structured surfaces has been associated with the transition from the Cassie–Baxter to the Wenzel state. When, at zero applied bias, the Cassie–Baxter state is of lower energy, the transition of the droplet back to the hydrophobic state requires the dissipation of a considerable amount of energy. In addition to this energy barrier that must be overcome, there are various dissipative forces acting on both smooth and rough surfaces, which oppose the motion of the liquid. EWOD has been performed on metal oxide nanostructures grown by PLD [83]. In particular ZnO and TiO2 nanostructures, grown on flat Si or microstructured Si, have been used in EWOD experiments as lower electrodes. A 200 nm thick layer of SiNx was evaporated on the structures, to act as an insulator. SiNx is hydrophilic, so a few monolayers of DMDCS ((CH3 )2 SiCl2 , silane group) are deposited on the dielectric layer in order to render the surface hydrophobic. The system can then be seen as two capacitors in series, the total capacitance of which is determined by the smaller capacitance (since the thickness of the silane layer is ∼100 times smaller than that of the SiNx one). As a result, the EWOD effect occurs principally due to the SiNx layer. In Figure 9.14 different TiO2 nanostructures are depicted. All three samples were grown using a nanosecond KrF laser source (248 nm) at the same temperature (650 ◦ C) and laser fluence (3 J cm–2 ), but different oxygen contents were used during deposition (0.05 mbar for TO1 and TO3, and 0.5 mbar for TO2). TO1 and TO2 were grown on crystalline Si, while TO3 was grown on microstructured Si, at fluence 2.1 J cm–2 . As discussed in a previous section, the oxygen pressure during PLD plays an important role in the stoichiometry of the grown sample. Despite the morphological similarities observed in samples TO1 and TO2, the different oxygen content used during deposition is expected to have resulted in differences in the oxygen stoichiometry of the samples. Indeed, it is known that a large amount of surface traps exists in TiO2 , created during deposition [89, 90]. So, it is expected that the lower oxygen pressure at which sample TO1 was grown will have induced more oxygen vacancies (traps) on the surface (and in the bulk) of the sample. The effect of the different amount of oxygen traps on the surfaces is clearly seen in Figure 9.15: the actuation of the droplet starts immediately after the application of the electric field for TO2, while for TO1 the actuation of the droplet starts after the applied voltage has exceeded a threshold value, V0 ≈ 13 V. This was explained as the result of the larger amount of surface traps on TO1, which causes the formation of a thicker electrical double layer at the TiO2 /SiNx interface [83, 88]. Furthermore, the reduction of the CA is about 15% for

270

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach (a)

(c)

TO1 (b)

TO2

(d)

cross section of TO1

TO3

Figure 9.14 FE-SEM images of the nanostructure TiO2 films (a) TO1, (b) TO2, (c) the crosssection of TO1, where the different layers can be distinguished, Source: Reprinted with permisC 2010, sion from E.L. Papadopoulou et al. J. Phys. Chem. C 114, 10249 (2010). Copyright  American Chemical Society. (d) TO3, where the TiO2 has been deposited on microstuctured Si, structured by the highest fluence (2.1 J cm–2 ). Source: E.L. Papadopoulou et al. J. Adhesion Sci. Technol. 26, 2143 (2012), reprinted by permission of the publisher (Taylor & Francis Ltd, http://www.tandf.co.uk/journals).

TO2 and double that (30%) for TO1. The insets in Figure 9.15, show the theoretical fit to Eq. (9.12). Clearly, theoretical and experimental points follow the same behaviour until saturation of CA occurs. We should note that, at a given applied voltage, the contact angle saturates in both samples. Upon the decrement of the applied voltage, the wetting state recovers its initial hydrophobic state. EWOD has also been performed in nanostructured ZnO films, however, electrolysis took place before saturation of the contact angle. In TO3, the TiO2 was deposited on microstructured Si, such as that described in a previous section. Hence, the initial hydrophobicity is enhanced (132◦ ) due to the microstructured substrate. However, the effect on the electrowetting behaviour is not spectacular, as one would expect for highly hydrophobic surfaces. Nevertheless, the application voltage for droplet actuation is much lower in comparison with the corresponding bare Si microstructures, which might be an advantage for applications.

9.5

Concluding Remarks

Pulsed laser deposition is shown to be an effective and flexible technique for growing nanostructures of metal oxides, mainly ZnO and TiO2 , with hydrophobic surfaces. These

Pulsed Laser Deposition of Surfaces with Tunable Wettability (a)

271

(b)

105

0.4

105 cra (CA)

100 95

0.2

100

0.0

85

0

200

400

600

CA

CA

90 95

0

90

180 0.1

90

0.0

75

-0.1

85

70 0

5

10

15 20 25 Applied Bias (V)

30

35

40

0

2

4

6 8 10 Applied Bias (V)

12

14

16

(d)

100

98

95

96

90

94 0.4

cra (CA)

80 75

400

200

40

30

240 0.10

92 90 88

20 Applied Bias (V)

0

0

0.00 -0.15 -0.10

86 10

80

160

0.05

0.2

0.0

70

0

cra(CA)

85

600

CA

CA

-0.2

–2

(c)

cra (CA)

80

16

14

12

10

8

6

4

2

0

Voltas (V) (Down)

Figure 9.15 Electrowetting behaviour of samples TO1 (left column) and TO2 (right column). The change of the CA as a function of increasing bias for (a) TO1 and (b) TO2, and decreasing bias for (c) TO1 and (d) TO2. In (a) and (c) we should note the threshold voltage (V0 ) for electrowetting, at ∼13.5 V. In the insets, the corresponding CA is depicted against the applied voltage. In the case of TO1, V0 has been subtracted. The solid lines are the fits to Lippmann’s equation. Source: Reprinted with permission from E.L. Papadopoulou et al. J. Phys. Chem. C C 2010, American Chemical Society. 114, 10249 (2010). Copyright 

materials have special wetting properties, and their wetting state can be altered by the application of different external stimuli. Nevertheless, PLD nanostructures have not been studied extensively for their wetting properties. We believe that there might be ample scope for studying wetting properties of solid surfaces, in synergy with the plentiful nanostructure designs that can be grown by PLD.

References 1. Barthlott, W. and Neinhuis, C. (1997) Purity of the sacred lotus, or escape from contamination in biological surfaces. Planta, 202, 1–8. 2. Wagner, T., Neinhuis, C. and Barthlott, W. (1996) Wettability and contaminability of insect wings of insect wings as a function of their surface sculptures. Acta Zool., 77, 213–225.

272

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

3. Sun, T.L., Feng, L., Gao, X.F. and Jiang, L. (2005) Bioinspired surfaces with special wettability. Acc. Chem. Res., 38, 644–652. 4. Ichimura, K., Oh, S-K and Nakagawa, M. (2000) Light-driven motion of liquids on a photoresponsive surface. Science, 288, 1624–1626. 5. Ma, M. and Hill, R.M. (2000) Superhydrophobic surfaces. Curr. Opin. Colloid Interface Sci., 11, 193–202. 6. Liu, Y., Mu, L., Liu, B. and Kong, J. (2005) Controled switchable surfaces. Chem.-A Eur. J., 11, 2622–2631. 7. Gras, S.L., Mahmud, T., Rosengarten, G. et al. (2007) Intelligent control of surface hydrophobicity. Chem. Phys. Chem., 8, 2036–2050. 8. Ganesh, V. A., Raut, H.K., Nair, A.S. and Ramakrishna, S. (2011) A review on selfcleaning coating. J. Mater. Chem., 21, 16304–16322. 9. Bhushan, B. and Jung, Y.C. (2011) Natural and biomimetic artificial surfaces for superhydrophobicity, self-cleaning, low adhesion and drag reduction. Prog. Mater. Sci., 56, 1–108. 10. Li, Y., Zhang, J. and Yang, B. (2010) Antireflective surfaces based on biomimetic nanopillared arrays. Nano Today, 5, 117–127. 11. Nosonovsky, M. and Bhushan, B. (2009) Superhydrophobic surfaces and emerging applications: non-adhesion, energy, green engineering. Curr. Opin. Colloid Interface Sci., 14, 270–280. 12. Cohen Stuart, M.A., Huck, W.T.S., Genzer, J. et al. (2010) Emerging applications of stimuli-responsive polymer materials. Nat. Mater., 9, 101–113. 13. Feng, X. and Jiang, L. (2006) Design and creation of superwetting/antiwetting surfaces. Adv. Mater., 18, 3063–3078. 14. Watanabe, T., Nakajima, A., Wang, R. et al. (1999) Photocatalytic activity and photoinduced hydrophilicity of titanium dioxide coated glass. Thin Solid Films, 351, 260– 263. 15. Quere, D. and Reyssat, M. (2008) Non-adhesive lotus and other hydrophobic materials. Philos. Trans. R. Soc. Ser. A, 366, 1539–1556. 16. Young, T. (1805) An essay on the cohesion of fluids. Philos. Trans. R. Soc., 95, 65–87. 17. De Gennes, P.G. (1985) Wetting: Statics and dynamics. Rev. Mod. Phys., 57, 827– 863. 18. Wenzel, R.N. (1936) Resistance of solid surfaces to wetting by water. Ind. Eng. Chem., 28, 988–994. 19. Cassie, A.B.D. and Baxter, S. (1944) Wettability of porous surfaces. Trans. Faraday Soc., 40, 546–551. 20. Nosonovsky, M. and Bhushan, B. (2007) Hierarchical roughness optimization for biomimetic superhydrophobic surfaces. Ultramicroscopy, 107, 969–979. 21. Patankar, N.A. (2004) Mimicking the lotus effect: mimicking of double roughness structures and slender pillars. Langmuir, 20, 8209–8213. 22. Lafuma, A. and Quere, D. (2003) Superhydrophobic states. Nat. Mater., 2, 457–460. 23. Verplnack, N., Conffinier, Y., Thomy, V. and Boukherroub, R. (2007) Wettability switching techniques on superhydrophic surfaces. Nanoscale Res. Lett., 2, 577– 596. 24. Feng, L., Li, S., Li, Y. et al. (2002) Superhydrophobic surfaces: from nature to applications. Adv. Mater., 14, 1857–1860.

Pulsed Laser Deposition of Surfaces with Tunable Wettability

273

25. Zhang, J., Huang, W. and Han, Y. (2006) Wettability of zinc oxide surfaces with controllable structures. Langmuir, 22, 2946–2950. 26. Chrisey, D.B. and Hubler, G.K. (1994) Pulsed Laser Deposition of Thin Films, John Wiley and Sons Inc., New York. 27. Ashfold, M.N.R., Clayessens, F., Fuge, G.M. and Henley, S.J. (2004) Pulsed laser ablation and deposition of thin films. Chem. Soc. Rev., 33, 23–31. 28. Chrisey, D.B., Pique, A., McGill, R.A. et al. (2003) Laser deposition of polymer and biomaterial films. Chem. Rev., 103, 553–576. 29. Cultrera, L., Zeifman, M.I. and Perrone, A. (2006) Investigation of liquid droplets, plume deflection and a columnar structure in laser ablation of silicon. Phys. Rev. B, 73, 075304. 30. Walczak, M., Papadopoulou, E.L., Sanz, M. et al. (2009) Structural and morphological characterization of TiO2 nanostructured films grown by pulsed laser deposition. Appl. Surf. Sci., 255, 5267–5270. 31. Papadopoulou, E.L., Varda, M., Kouroupis-Agalou, K. et al. (2008) Undoped and Aldoped ZnO films with tuned properties grown by pulsed laser despotion. Thin Solid Films, 516, 8141–8145. 32. Yan, F., Zhu, T.J., Lai, M.O. and Lu, L. (2010) Infuence of oxygen pressure on the ferroelectric properties of BiFeO3 thin films LaNiO3 /Si substrates via laser ablation. Appl. Phys. A, 101, 651–654. 33. Magaraggi, R., Hambe, M., Kostylev, M. et al. (2011) Probing La0.7 Sr0.3 MnO3 multilayers via spin wave resonances. Phys. Rev. B, 84, 104441. 34. Valerini, D., Caricato, A.P., Lomascolo, M. et al. (2008) Zinc oxide nanostructures grown by pulsed laser deposition. Appl. Phys. A, 93, 729–733. 35. Papadopoulou, E.L., Zorba, V., Pagkozidis, A. et al. (2009) Reversible wettability of ZnO nanostructured thin films prepared by pulsed laser deposition. Thin Solid Films, 518, 1267–1270. 36. Kawakami, M., Hartano, A.B., Nakata, Y. and Okada, T. (2003) Synthesis of ZnO nanorods by nanoparticle assisted pulsed laser deposition. Jpn. J. Appl. Phys., 42, 33–35. 37. Okada, T., Kawashima, K. and Ueda, M. (2005) Ultraviolet lasing and field emission characteristics of ZnO nanorods synthesized by by nanoparticle assisted pulsed laser deposition. Appl. Phys. A, 81, 907. 38. Sun, Y., Doherty, R.P., Warren, J.L. and Ashfold, M.N.R. (2007) Effect of incident fluence on the growth of ZnO nanorods by pulsed excimer laserdeposition. Chem. Phys. Lett., 447, 257–262. 39. Sun, Y., Fuge, G.M. and Ashfold, M.N.R. (2006) Growth mechanisms of ZnO nanorods formed by pylsed laser deposition. Super. Micro., 39, 33–40. 40. Okada, T., Agung, B.H. and Nakata, Y. (2004) ZnO nanorods synthesized by nanoparticle assisted pulsed laser deposition. Appl. Phys. A, 79, 1417–1419. 41. Hartanto, A.B., Ning, X., Nakata, Y. and Okada, T. (2004) Growth mechanism of ZnO nanorods from nanoparticles formed in a laser ablation plume. Appl. Phys. A, 78, 299–301. 42. Tien, L.C., Pearton, S.J., Norton, D.P. and Ren, F. (2008) Synthesis and microstructure of vertically aligned ZnO nanowires grown by high pressure assisted pulsed laser deposition. J. Mater. Sci., 43, 6925–6932.

274

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

43. Cao, B.Q., Matsumoto, T., Matsumoto, M. et al. (2009) ZnO nanowalls grown with high pressure PLD and their applications as field emitters and UV detectors. J. Phys. Chem. C, 113, 10975. 44. Premkumar, T., Zhou, Y.S., Lu, Y.F. and Baskar, K. (2010) Optical and field emission properties of ZnO nanostructures deposited using high pressure pulsed laser deposition. Appl. Mater. Inter., 2, 2863–2869. 45. Klini, A., Manousaki, A., Anglos, D. and Fotakis, C. (2005) Growth of ZnO thin films by ultra violet pulsed laser ablation: study of plume dynamics. J. Appl. Phys., 98, 123301. 46. Zhang, Y., Russo, R.E. and Mao, S.S. (2005) Femtosecond laser assisted growth of ZnO nanowires. Appl. Phys. Lett., 87, 133115. 47. Liu, B., Hu, Z., Che, Y. et al. (2008) Growth of ZnO nanoparticles and nanorods with ultrafast pulsed laser deposition. Appl. Phys. A, 93, 813. 48. Papadopoulou, E.L., Barberoglou, M., Zorba, V. et al. (2009) Reversing photoinduced wettabiility transition of hierarchical ZnO structures. J. Phys. Chem. C, 113, 2891– 2895. 49. Zorba, V., Stratakis, E., Barberoglou, M. et al. (2008) Tailoring the wetting response of silicon surfaces via fs laser structuring. Appl. Phys. A, 93, 819–825. 50. Li, L., Li, Y., Gao, S. and Koshizaki, N. (2009) Ordered Co3 O4 hierarchical nanorod arrays: tunable superhydrophobicity without UV irradiation and transition to superhydrophobicity. J. Mater. Chem., 19, 8366–8371. 51. Li, L. and Koshizaki, N. (2010) Vertically aligned and ordered hematite hierarchical columnar arrays in applications in field emission, superhydrophilicity and photocatalysis. J. Mater. Chem., 20, 2972–2978. 52. Ngom, B.D., Sakho, O., Ndiaye, S. et al. (2011) Photon induced tunable and reversible wettability of pulsed laser deposited W-doped ZnO samples. Eur. Phys. J. Appl. Phys., 55, 20501. 53. Fotou, G.P. and Pratsinis, S. (1996) Photocatalysis of phenol and salicylic acid by aerosol-made titania powders. Chem. Eng. Commun., 151, 251–269. 54. Yamaki, T., Sumita, T., Yamamoto, S. and Miyashita, A. (2002) Preparation of epitaxial TiO2 films by PLD for photocatalyst applications. J. Cryst. Growth, 237–239, 574. 55. Gy¨orgy, E., Socol, G., Axente, E. et al. (2005) Anatase phase TiO2 thin films obtained by pulsed laser deposition for gas sensing applications. Appl. Surf. Sci., 247, 429– 433. 56. Gy¨orgy, E., P´erez del Pin, A., Sauthier, G. et al. (2007) Structural, morphological and local electrical properties of TiO2 thin films grown by pulsed laser deposition. J. Phys. D: Appl. Phys., 40, 5246–5251. 57. Li, Y., Sasaki, T., Shimizu, Y. and Koshizaki, N. (2008) Hexagonal closed-packed hierarchical, amorphous TiO2 nanocolumn arrays: transferability, enhanced photocatalytic activity, and superamphiphilicity without UV irradiation. J. Am. Chem. Soc., 130, 14755–14762. 58. Daoud, W.A., Xin, J.H., Zhang, Y.H. and Mak, C.L. (2006) Pulsed laser deposition of superhydrophobic thin Teflon films on cellulosic fibres. Thin Solid Films, 515, 835–837. 59. Popescu, A.C., Duta, L., Dorcioman, G. et al. (2011) Radical modification of the wetting behavior of textiles coated with ZnO thin films and nanoparticles when changing the ambient pressure in the pulsed laser deposition process. J. Appl. Phys., 110, 064321.

Pulsed Laser Deposition of Surfaces with Tunable Wettability

275

60. Krupenkin, T.N., Taylor, J.A., Schneider, T.M. and Yang, S. (2004) From rolling ball to complete wetting: the dynamic tuning of lights on nanostructured surfaces. Langmuir, 20, 3824–3827. 61. Jones, D.M., Smith, J.R., Huck, W.T. and Alexander, C. (2002) Variable adhesion of micropatterned thermoresponsive polymer brushes: AFM Investigations of Poly(N-isopropylacrylamide) brushes prepared by surface-initiated polymerizations. Adv. Mater., 14, 1130–1134. 62. Zhang, J.L., Lu, X.Y., Huang, W.H. and Han, Y.C. (2005) Reversible superhydrophobicity to superhydrophilicity transition by extending and unloading an elastinc polyamidefilm. Macromol. Rapid Commun., 26, 477–480. 63. Sun, R., Nakajima, A., Fujishima, A. et al. (2001) Photoinduced surface wettability conversion of ZnO and TiO2 thin films. J. Phys. Chem. B, 105, 1984. 64. Feng, X., Zhai, J. and Jiang, L. (2005) The fabrication and switchable superhydrophobicity of TiO2 nanorod films. Angew. Chem. Int. Ed., 44, 5115–5118. 65. Miyauchi, M., Nakajima, A., Watanabe, T. and Hashimoto, K. (2002) Photocatalysis and photoinduced hydrophilicity of various metal oxide thin films. Chem. Mater., 14, 2812–2816. 66. Wang, S., Song, Y. and Jiang, L. (2007) Photoresponsive surfaces with controllable wettbility. J. Photochem. Photobiol. C, 8, 18. 67. Rosario, R., Gust, D., Garcia, A.A. et al. (2004) Lotus effect amplifies light induced contact angle switching. J. Phys. Chem. B, 108, 12640–12642. 68. Miyauchi, M., Nakajima, A., Fujishima, A. et al. (2000) Photoinduced surface reactions TiO2 and SrTiO3 films: photocatalytic oxidation and photoinduced hydrophilicity. Chem. Mater., 12, 3–5. 69. Froumkine, A. (1936) Actual. Sci., 373, 1. 70. Berge, B. (1993) Electrocapillarite et mouillage de films isolant par l’eau. C. R. Acad. Sci. III, 317, 157–163. 71. Krupenkin, T.N., Taylor, J.A., Wang, E.N. et al. (2007) Reversible wetting-dewetting transitions, on electrically tunable superhydrophobic nanostructured surfaces. Langmuir, 23, 9128–9133. 72. Verplanck, N., Galopin, E., Camart, J.-C. et al. (2007) Reversible electrowetting on superhydrophobic silicon nanowires. Nanolett., 7, 813–817. 73. Barberoglou, M., Zorba, V., Pagkozidis, A. et al. (2010) Electrowetting properties of micro/nano structured black silicon. Langmuir, 26, 13007–13014. 74. Dhindsa, M.S., Smith, N.R., Heikenfeld, J. et al. (2006) Reversible electrowetting on vertically aligned superhydrophobic carbon nanofibers. Langmuir, 22, 9030– 9034. 75. Zhu, L., Xu, J., Xiu, Y. et al. (2006) Electrowetting of aligned carbon nanotubes. J. Phys. Chem. B, 110, 15945–15950. 76. Han, Z., Tay, B., Tan, C. et al. (2009) Electrowetting control of Cassie-to-Wenzel transitions in superhydrophobic carbon nanotube-based nanocomposites. ACS Nano, 3, 3031–3036. 77. Herbertson, D.L., Evans, C.R., Shirtcliffe, N.J. et al. (2006) Electrowetting on superhydrophobic SU-8 patterned surfaces. Sens. Actuators A, 130–131, 189–193. 78. Seyrat, E. and Hayes, R.A. (2001) Amorphous fluoropolymers as insulators for reversible low-voltage electrowetting. J. Appl. Phys., 90, 1383.

276

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

79. Bayiati, P., Tserepi, A., Petrou, P.S. et al. (2007) Electrowetting on plasma deposited fluorocarbon hydrophobic films for biofluid transport in microfluidics. J. Appl. Phys., 101, 103306. 80. Campbell, J.L., Breedon, M., Latham, K. and Kalantar-zadeh, K. (2008) Electrowetting of superhydrophobic ZnO nanorods. Langmuir, 24, 5091–5098. 81. Wu, J., Xia, J., Lei, W. and Wang, B.P. (2010) Electrowetting on ZnO nanowires. Appl. Phys. A, 99, 931–934. 82. He, S., Zheng, M., Yao, L. et al. (2010) Preparation and properties of ZnO nanostructures by electrochemical anodization method. Appl. Surf. Sci., 56, 2557–2562. 83. Papadopoulou, E.L., Pagkozidis, A., Barberoglou, M. et al. (2010) Electrowetting properties of ZnO and TiO2 nanostructured thin films. J. Phys. Chem. C, 114, 10249– 10253. 84. Moon, H., Cho, S.-K., Garrell, R.L. and Kim, C.J. (2002) Low voltage electrowetting on dielectric. J. Appl. Phys., 92, 4080–4087. 85. Welters, W.J.J. and Fokkink, L.G.J. (1998) Fast electrically switchable capillary effects. Langmuir, 14, 1535–1538. 86. Ivosevic, N. and Zutic, V. (1998) Langmuir, 14, 231. 87. Tokkerli, A. (2003) Droplet Microfluidics on a Planar Surface. PhD Thesis. Technical Research Centre of Finland. 88. Papadopoulou, E.L., Zorba, V., Stratakis, E. and Fotakis, C. (2012) Properties of silicon and metal oxide electrowetting systems. J. Adhesion. Sci. Technol., 26, 2143–2163. 89. Zhao, H., Zhang, Q. and Weng, Y.-X. (2007) Deep surface trap filling by photoinduced carriers and interparticle electron transport TiO2 nanocrystalline film with time-resolved visible and mid-IR transient spectroscopies. J. Phys. Chem. C, 111, 3762–3769. 90. Boschloo, G. and Fitzmaurice, D. (1999) Spectroelectrochemical investigation of surface states in nanostructured TiO2 electrodes. J. Phys. Chem. B, 103, 2228–2231.

10 Fabrication of Antireflective Self-Cleaning Surfaces Using Layer-by-Layer Assembly Techniques Yu-Min Yang Department of Chemical Engineering, National Cheng Kung University, Taiwan

10.1

Introduction

To cope with the need for the wide range of possible applications, new multifunctionality of surfaces is always desirable. For self-cleaning surfaces, physical and chemical properties, such as wettability, photochemical activity, and bacteria-resistance are most important. Considerable success in the realization of self-cleaning surfaces has actually been achieved by coating methods. The coatings, however, usually enhance the reflection of the transparent substrates. This is a great disadvantage to various applications such as glazing, display devices, optical materials, and solar cells. Therefore, it is most desirable that antireflective properties can still be exhibited by the functional coatings. In this chapter, key topics in the field of multifunctional self-cleaning surfaces, concentrating on the materials and recent development of fabrication techniques using layer-by-layer (LbL) assembly are summarized. The principles of antireflection and methods of solutionbased LbL assembly techniques are demonstrated. Mechanisms of both hydrophilic and hydrophobic self-cleaning are briefly described and related to the material properties of the assemblies. Characterization of the as-fabricated surfaces by various methods is also provided. This chapter also highlights challenges still to be met, together with recent innovations that may overcome them.

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach, First Edition. Edited by Walid A. Daoud. © 2013 John Wiley & Sons, Ltd. Published 2013 by John Wiley & Sons, Ltd.

278

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

10.2

Antireflective Coatings

Optical reflection is a fundamental phenomenon occurring when light propagates across a boundary between two media which have different refractive indices. For normal incidence in air environment, the reflection of light at the air/substrate interface is given by the simplified Fresnel’s equation [1]: R = [(n s − 1) / (n s + 1)]2

(10.1)

in which ns and 1 are the refractive indices of the solid substrate and air, respectively. According to Eq. (10.1), the reflection at air/SiO2 (refractive index: 1.46), air/glass (refractive index: 1.52), air/anatase TiO2 (refractive index: 2.52), and air/rutile TiO2 (refractive index: 2.76) interfaces is 3.5%, 4.3%, 18.6%, and 21.9%, correspondingly. For the fabrication of self-cleaning surfaces by coating methods, therefore, reflection issues will be encountered unavoidably. There are two major approaches to achieve low reflection: 1. Interference multiple layers 2. Inhomogeneous layer with gradient refractive index. 10.2.1

Interference Multiple Layers

As shown in Figure 10.1, the principle of antireflection (AR) is interference of the reflected light from the medium/coating and coating/substrate interfaces. When these two waves are out of phase, destructive interference occurs and, consequently, both beams cancel each other partially or totally before they exit the surface. For an ideal homogeneous single-layer coating, the minimum reflection is given by Rm = (n 2c − n a · n s )2 /(n 2c + n a · n s )2

(10.2)

where nc , na , and ns are the refractive indices of the coating, medium and substrate, respectively [1–3]. The requirements for the minimum reflection to be zero (Rm = 0) in air

Air,na

+ t nf Thin film

θ

ns Substrate

=

Thickness, t Optical thickness, nf × t

Figure 10.1 Schematic diagram of a single antireflective thin film on a substrate.

Fabrication of Antireflective Self-Cleaning Surfaces

279

medium (na = 1) are that the relationship between the indices of the coating and substrate must be n c = (n s )1/2

(10.3)

and the thickness of the coating should also meet the quarterwave optical thickness requirement t = λm /4 · n c

(10.4)

where λm is the wavelength of minimum reflectivity [1]. For an AR coating on a glass (n = 1.52), it must have an index of 1.23 from Eq. (10.3). This value is lower than that of any homogeneous dielectric material [3]. Fabrication/application of single-layer AR coating is limited by the availability of suitable material. To obtain low surface reflection, however, one may make the refractive index of the coating, such as TiO2 film, lower than that of the substrate materials through an increase in the porosity or a coupling of material with low refractive index, such as SiO2 [1, 4–13]. Fortunately, precise control over the porosity, thickness, and composition of the coating film can be achieved by a simple method now known as layer-by-layer (LbL) assembly [14, 15]. Solution-based LbL assembly techniques will be mentioned in Section 10.3. Besides single-layer AR coating, delicate double-layer, three-layer, and multilayer coatings have been designed to achieve a better AR effect [2, 3, 16]. 10.2.2

Inhomogeneous Layer with Gradient Refractive Index

Refractive index (n)

Since the application of conventional single-layer AR coatings is limited to a single wavelength and at normal incidence of light, the gradient refractive index AR coating becomes a desirable choice. Inhomogeneous layers have long been used for optical coatings [3, 9, 16, 17]. As shown in Figure 10.2, inhomogeneous layers differ from normal homogeneous layers in that a smooth variation of refractive index prevails throughout the thickness in the former rather than the step change in refractive index in the latter. Other than a linear profile for the gradient- refractive index layer, different profiles such as parabolic, cubic, and exponential have also been extensively investigated. Gradient refractive index antireflection films are usually produced by chemical etch/leach processes.

ns

ns nc (x) nair

nair

0

d

0

d

Distance from the original substrate surface (x) (a)

(b)

Figure 10.2 Representation of a coating exhibiting linear inhomogeneity (b) as compared to that exhibiting a step change (a) in refractive index.

280

10.3

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

Solution-Based Layer-by-Layer (LbL) Assembly Techniques

Many studies and potential applications involving transparent self-cleaning surfaces depend on the availability of simple and reliable thin film deposition techniques. The ability to assemble organic or inorganic films a monolayer at a time allows for LbL control of thickness, composition, and physical properties. Such control provides an important path for the creation of transparent self-cleaning surfaces. Several solution-based approaches are currently available for controlling growth with monolayer precision [18, 19]. 10.3.1

Electrostatic Assembly

The concept of depositing particles onto solid substrates in an electrostatic LbL (ELbL) manner can be traced back to the work of Iler in the mid-1960s [20]. This approach relies on electrostatic interaction between oppositely charged surfaces and can be readily generalized to include organic, inorganic, and organic–inorganic hybrid systems. It was extended by Decher and coworkers in the early 1990s to a combination of linear polycations and polyanions [14]. The scope of the ELbL technique was later expanded to include inorganic nanoparticles, biomolecules, clays, and dyes in polyelectrolyte multilayer assemblies [15]. The method of sequential build-up of multilayers was even applied to colloidal particles, thus permitting the formation of composite core–shell particles [21]. Figure 10.3 shows

Substrate

Cationic

Anionic Nanoparticle: Polyelectrolyte: repeat

Polyelectrolyte / Polyelectrolyte

Nanoparticle / Nanoparticle

Polyelectrolyte / Nanoparticle

Figure 10.3 Schematic diagram showing electrostatic layer-by-layer assembly of oppositely charged polyelectrolytes, polyelectrolyte/nanoparticle, and nanoparticle/nanoparticle, respectively, on a substrate. Repetition of steps allows the preparation of multilayers.

Fabrication of Antireflective Self-Cleaning Surfaces

281

schematically multilayer assembly by consecutive adsorption of anionic and cationic polyelectrolytes, nanoparticles, and polyelectrolyte/ nanoparticle, respectively. 10.3.2

Langmuir–Blodgett (LB) Assembly

One of the earliest methods of LbL assembly is the classic, Langmuir–Blodgett (LB) technique [22–24], which was developed over 75 years ago. By modifying the Langmuir balance, one can deposit monolayers on a solid substrate to create mono- and multilayer films. Modifications include a dipping device to lower or raise the substrate through the monolayer, an automated movable barrier, which moves during the deposition process so as to maintain a control value of surface pressure, and a surface pressure sensor, which controls the movable barrier. Molecules, usually amphiphiles, are first spread on the gas/liquid interface. After the evaporation of solvent, the molecules are compressed to a close-packed monolayer with the desired target surface pressure using a movable barrier. They are then transferred as a monolayer assembly to a solid substrate by passing through the interface. In the traditional vertical dipping mode, the substrate is immersed (lowered) and emersed(raised) through the interface. Successive dipping of a substrate covered by a monolayer in and out of a liquid builds up multilayers. The most common interface at which to create the Langmuir monolayer is the air/water interface. Figure 10.4 shows the typical procedure of the LB assembly. The LB technique is also frequently used to make monolayers of particles at air/water or air/oil interfaces [25–27]. Recently, the possibility of preparing two-dimensional colloidal crystals from silica spheres by the LB method was studied and was successful in depositing monolayers consisting of hexagonally close-packed (hcp) arrays of silica on various substrates [28–32]. The results suggest that a successful synthesis of ordered

Substrate Gas-phase

Gas-phase Amphiphiles

Liquid-phase

Liquid-phase

Gas-phase

Gas-phase Barrier

Liquid-phase

Liquid-phase

Figure 10.4 Schematic representation of a close-packed Langmuir monolayer preparation and the Langmuir–Blodgett (LB) deposition of the monolayer onto a substrate.

282

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

:Adsorbed surfactant

SiO2

SiO2

Figure 10.5 Schematic representation of silica particles, surface modified by adsorption of surfactant molecules, floating on the air/water interface. Source: Reprinted with permission from [32], Copyright (2006) Elsevier Ltd.

monolayers of monodisperse silica with the LB technique depends critically on the hydrophilic/hydrophobic balance. This balance was investigated by modification of the surface of the silica particles through the adsorption of surfactants [29–32]. As shown in Figure 10.5, the surface-modified solid particle then behaves like an amphiphile at the air/water interface due to the hydrophilic–hydrophobic balance, and the fabrication of particulate LB films is therefore made possible. 10.3.3

Self-Assembly

While electrostatic assembly and the LB technique require the formation of oppositely charged surfaces and a preformed monolayer at the surface of a liquid, respectively, selfassembled monolayers form spontaneously upon immersion of the substrate into a uniform solution [33]. In this technique, the resulting monolayer is chemically anchored to the surface, while electrostatic assembly and LB monolayers are physically linked to it. Thus, these self-assembled monolayers (SAMs) are supposed to be stronger and more resistant than their electrostatic assembly and LB counterparts. However, the more narrow range of substrates available for SAM and the difficulty of constructing multilayers still allow a great deal of interest in electrostatic assembly and LB technique layers. SAMs have been successfully constructed on gold through a thiol group and on silica through a trichlorosilane group. The latter category is especially attractive because of the large number of silica-like substrates, such as glass and metallic oxides, and the low roughness that can be achieved on some microelectronic silicon wafers covered with native silica. The commonly assumed mechanism in silanization has three distinct phases, as shown in Figure 10.6 [34, 35]. First, the aliphatic chains are strongly attracted to the clean silica surfaces through the trichlorosilane group, which acts like the polar head of an amphiphilic molecule. Like any oxide, silica surface is hydrated; surface groups are silanols and silica substrates are covered with a water film (one or several layers thick). The trichlorosilane groups are then hydrolyzed when they get close enough to the surface. Chains are then bonded through hydrogen bonds to the surface silanols and to their close neighbors. This situation is followed by water elimination leading to a network in which each chain is linked to the surface and to the other chains. This exceptional property gives very stable and well-oriented layers.

Fabrication of Antireflective Self-Cleaning Surfaces –HCl

Si Cl Cl

Cl

OH

Si

H O

Cl

OH

H

O

O

O

OH

Si O

O

O

O

Si

H

O H

OH

Si

Si O

–H2O

Cl Si Cl

H2O OH

283

H O

Si

O H

O H

H O

O

Si O

O

Si O

O

Si

O

O

Si O

Figure 10.6 Schematic representation of the commonly accepted mechanism of the silanization reaction. The physical adsorption of the trichlorosilane molecules is followed by their hydrolysis and, inally, by water elimination, leading to a chemically anchored monolayer. Source: Adapted with permission from [35] Copyright (1991) American Chemical Society.

Although self-assembled monolayers and multilayers of organic molecules have been extensively studied, multilayered hybrid films have also been created by a sequential adsorption and chemical activation process [18, 36]. Another modification of the self-assembly technique involves the sequential adsorption of the components of thermodynamically stable and insoluble layered metal phosphonates [18, 37].

10.4 10.4.1

Mechanisms of Self-Cleaning Hydrophilic Surfaces

A surface is usually called “superhydrophilic” when the advancing contact angle of water on the surface is less than 5◦ [12, 38–40] or the time for complete wetting by small droplets of water has been observed to be less than 0.5 s [8, 41]. By utilizing the photocatalytic and superhydrophilic properties of TiO2 , considerable success in the realization of self-cleaning surfaces has been achieved [38, 40–48]. In these cases, photocatalysis causes the coating to chemically break down organic dirt adsorbed on the surface, while hydrophilicity causes water to form sheets rather than droplets, and dirt is washed away. A thorough discussion of the theory of photocatalysis and superhydrophilicity is beyond the scope of this chapter, therefore only a brief summary follows. Greater detail can be found in one of several review papers on the subject [49–52]. The basic working principles of these self-cleaning surfaces, however, can be illustrated in Figure 10.7. As a semiconductor under normal conditions, TiO2 absorbs light with energy equal to or greater than its band gap energy, resulting in excited charged carriers: an electron, e− , and a hole, h + . Although the fate of most of these charge carriers is rapid recombination, some migrate to the surface. There, holes cause the oxidization of adsorbed organic molecules while electrons eventually combine with atmospheric oxygen to give the superoxide radicals, which quickly attack nearby organic molecules. The result is a cleaning of the surface by “cold combustion”, the conversion of organic molecules to carbon dioxide and water at ambient temperatures [47].

284

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

–

Photo excited electron

+

Photo excited hole

hv

(a)

–

–

VB

+

hv

+

–

+

(a) electron-hole recombination on the surface

+

+ –

(b)

+

+

–

+

TiO2

(d)

(c) D A

(a)

D+

+

–

A–

CB

(b) electron-hole recombination in the bulk (c) e– + O2 → O2– O2– + H2O → HO2 · + OH– HO2 · + H2O → H2O2 + OH · H2O2 → 2OH · 2OH

· + organic → CO2 + H2O

2H2O → H2O2 + 2H+ (d) h H2O2 → 2OH · ++

2OH

· + organic → CO2 + H2O

Figure 10.7 The semiconductor (TiO2 ) undergoes photo-excitation upon illumination by UV light. The electron and the hole that result can follow one of several pathways. Source: Adapted with permission from [47] Copyright (2005) RSC.

10.4.2

Hydrophobic Surfaces

It is commonly accepted in the literature that a “superhydrophobic” surface should exhibit a water advancing contact angle larger than 150◦ and a very small contact angle hysteresis (the difference between advancing and receding contact angles) [53,54]. Plant leaf surfaces that exhibit unusual wetting characteristics of superhydrophobicity have been investigated and documented [55, 56]. In these leaves even dew and fog, but especially rain, lead to the complete removal of particulate contamination. This is called the Lotus effect, because it can be demonstrated beautifully in the great peltate leaves of the sacred lotus, which is a symbol for purity in eastern Asian religions. Efforts to mimic this biological cleaning mechanism resulted in a 1998 patent issued to biologists Neinhaus and Barthlott describing the “Lotus Effect” [57]. (1998). It is well-known now that the wettability of naturally occurring solid surfaces with liquids is governed by both the chemical properties and the microstructure of the surface. Once again, a thorough discussion of the fundamental theories of superhydrophobicity and self-cleaning is beyond the scope of this chapter, therefore only a brief summary follows. Greater detail can be found in one of several review papers on the subject [40, 47, 58]. The different results obtained from smooth and rough surfaces with respect to wettability and particle removal are summarized in Figure 10.8. As shown in Figure 10.8a, particles

Fabrication of Antireflective Self-Cleaning Surfaces

(a)

285

(b)

Figure 10.8 Schematic diagram showing the connection between roughness and selfcleaning. While on smooth surfaces the particles are mainly redistributed by water (a), they adhere to the droplets surfaces on rough surfaces and are removed from the surfaces when the droplets roll off (b). Source: With kind permission from Springer Science + Business Media: C 1997. [56], 

on smooth surfaces are mainly displaced to the sides of the droplet and re-deposited behind the droplet, but not removed. On the other hand, the adhesion between particle and rough surface is minimized due to the small contact area. Therefore, the particles may adhere to the surface of the droplet and are removed from the rough surface when the droplet rolls off, as shown in Figure 10.8b [56].

10.5 Fabrication of Antireflective Self-Cleaning Surfaces Using Electrostatic Layer-by-Layer (ELbL) Assembly of Nanoparticles 10.5.1

Superhydrophilic Self-Cleaning Surfaces with Antireflective Properties

Zhang et al. [12] reported a self-cleaning particle coating with antireflection properties by using the ELbL assembly technique. A sub-monolayer of SiO2 particles was covered with TiO2 nanoparticles with the help of oppositely charged polyelectrolytes, to generate a lowrefractive-index film and maintain the self-cleaning properties at the same time. The same research group [13] examined further the possibility of creating the dual functions of selfcleaning and antireflection in double-layered TiO2 –SiO2 films that consisted of dense top layer of TiO2 and porous bottom layer of SiO2 . The films were prepared by ELbL assembly of SiO2 nanoparticles and titanate nanosheets with polycations. Kim et al. [6, 7] reported that TiO2 thin films assembled with TiO2 nanoparticles and oppositely charged polyelectrolytes or titanium (IV) bis (ammonium lactato) dihydroxide (TALH) were fabricated via the ELbL assembly method. The coating procedures for the fabrication of (TiO2 /TALH) multilayer thin film were repeated 30 times. This film may show a high transmittance in the visible light range for optical applications as well as good photocatalytic performance to decompose the Methyl Orange molecules with UV light irradiation. On the other hand, a double-layer AR thin film assembled with 21 bilayers of poly(diallyldimethylammonium

286

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

chloride)/TALH and 20 bilayers of poly(allylamine hydrochloride)/poly(acrylic acid) was also reported. [6, 7] While polyelectrolytes were frequently used in the ELbL assembly processes of multilayer nanoparticulate thin films, as mentioned above, multilayer thin films containing oppositely charged nanoparticles without any polyelectrolytes have seldom been reported. Yang and Tsai [11] demonstrated the feasibility of ELbL assembly of all TiO2 nanoparticles for preparing self-cleaning surfaces on glass substrates without using any polyelectrolytes. The isoelectric point (PI) of TiO2 nanoparticles was determined and the charge was controlled by adjusting the pH of the nanoparticle suspension below or above the PI = 6.5. Mean values of the zeta-potential of 130 and −85 mV at pH values of 1.95 and 11.87 were measured for positively charged TiO2 nanoparticles and negatively charged TiO2 nanoparticles, respectively. The quartz crystal microbalance was also used to measure the amounts of TiO2 nanoparticles adsorbed in each sequential deposition and confirmed the stepwise growth of the films. Self-cleaning nanoparticulate thin films up to 6 bilayers, for which the average transmittance for light in the region of 400–800 nm is about the same as that of plain glass, were fabricated. The linear build-up of film, however, cannot be realized by this process. This may be due to the drastic change in pH values in the alternative depositions of positively charged TiO2 and negatively charged TiO2 nanoparticles. Lee et al. [8] showed that TiO2 /SiO2 all-nanoparticle coatings prepared by using ELbL assembly of oppositely charged nanoparticles exhibit antireflection, antifogging, and self-cleaning functionalities. The amphoteric properties of TiO2 and SiO2 nanoparticles are shown in Figure 10.9 [59]. It is noteworthy that the pH of each nanoparticle suspension was adjusted to the same value of 3.0. Coatings of up to 30 TiO2 /SiO2 bilayers were reported. The multilayers comprising 7 nm TiO2 and 22 nm SiO2 nanoparticles showed linear growth behavior with an average bilayer thickness of 19.6 nm. Furthermore, their results showed that the refractive index does not change as a function of the number of deposited bilayers and the multilayers

45 22 nm Sio2

Zeta-potential (mV)

30

7 nm Tio2 15 0 –15 –30 –45 –60

1

3

5

7

9

11

pH

Figure 10.9 Zeta-potential of suspended nanoparticles determined by dynamic light scattering as a function of nanoparticle suspension pH. Source: Reproduced from [59] by permission of Michael F. Rubner, ACS Publications.

Fabrication of Antireflective Self-Cleaning Surfaces

287

ELbL assembly with periodic calcination

MeOH/HCI Water solution wash treatment and drying

Glass Substrate

Water wash and drying

Water wash and drying

Anionic SiO2 suspension

Cationic TiO2 suspension

Calcination 550°C 3hr

Self-cleaning surface

repeat Figure 10.10 Fabrication of transparent and superhydrophilic self-cleaning surfaces by TiO2 /SiO2 all-nanoparticle coatings with periodic calcinations every 20 or 30 bilayers.

have an average refractive index at 633 nm of 1.28. However, antireflection, antifogging (superhydrophilicity), and self-cleaning properties were reported only for all-nanoparticle thin-film coatings up to 6 TiO2 /SiO2 bilayers. Lin et al. [60] extended the work of Lee et al. [8] to the design and fabrication of transparent and superhydrophilic self-cleaning surfaces of TiO2 /SiO2 all-nanoparticle coatings up to 120 bilayers. This is the thickest transparent and superhydrophilic self-cleaning surface that has ever been prepared by this method, as reported in the open literature. It is noteworthy that in preparing a thin film beyond 40 TiO2 /SiO2 bilayers, however, the occurrence of microcracks in the thin film after calcination was encountered. This might result in a drastic decrease in transmittance. Fortunately, the problem was overcome by using an appropriate periodic calcination every 20 bilayers or even every 30 bilayers, as shown in Figure 10.10. The typical materials used by Lin et al. [60] for fabricating superhydrophilic self-cleaning surfaces with antireflective properties are summarized in Table 10.1. The experimental transmittance spectra of some of these as-fabricated thin films in the visible light region (400–800 nm) are shown in Figure 10.11. All of these experimental results showed the antireflective properties of the thin films of all-nanoparticle coatings

Table 10.1

Typical materials used in the ELbL assembly process.

Materials

Physical Properties

Nanoparticle

SiO2 TiO2 (Anatase)

RI = 1.46, Dia. = 7, 22 RI = 2.52, Dia = 7

Polyelectrolyte

PAA PAH

Mw = 100 000 Mw = 70 000

Substrate Silane

Glass Dodecyltrichlorosilane

RI = 1.52 Mw = 303.78

PAA: poly(acrylic acid); PAH: poly(allylamine hydrochloride); RI: refractive index; Dia.(nm): diameter; Mw (g mol–1 ): molecular weight.

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach

98

X–4 96 X–3 94

X–2

92

X–1 plala glass

90 400 (c) 100 Transmittance (%)

X–6 X–5

500 600 700 Wavelength (nm)

X–30

98 96

X–20

94 X–10

92 90 400

800

500 600 700 Wavelength (nm)

800

(d) 100

X–60

90 80 70 60 50 400

(b) 100

Transmittance (%)

Transmittance (%)

(a) 100

Transmittance (%)

288

X–90

98 96 94 92

500 600 700 Wavelength (nm)

Transmittance (%)

(e)

800

90 400

500 600 700 Wavelength (nm)

800

100 98

X–120

96 94 92 90 400

500 600 700 Wavelength (nm)

800

Figure 10.11 Experimental transmittance spectra in visible light region of (TiO2 /SiO2 )X nanoparticulate thin films on both sides of glass substrates. (a) x = 1–6, (b) x = 10, 20, 30, (c) x = 60, (d) x = 90, (e) x = 120. Source: Adapted with permission from [60] Copyright (2011) Taiwan Institute of Chemical Engineers.

on a glass substrate. Furthermore, average transmittance values higher than 93.5% were exhibited by all the nanoparticulate thin films and the experimental results were only, at the very most, 2% lower than those predicted by the theory, as shown in Figure 10.12. Figure 10.13a shows the photocatalytic decomposition of Methylene Blue (MB) molecules by immersing (TiO2 /SiO2 )x nanoparticulate thin film in 10 ppm aqueous MB solution under UV light irradiation. The decrease in the ratio A/Ao , in which Ao is the initial

Fabrication of Antireflective Self-Cleaning Surfaces

Average transmittance (%)

100

0

200

400

600

289

Thickness (nm) 800 1000 1200 1400 1600 1800 2000 2200 2400

95 90 85 Theoretical ELBL assembly with one calcinations ELBL assembly with periodic calcinations every 30 bilayers ELBL assembly with periodic calcinations every 20 bilayers Plain glass

80 75 70 65

0

10

20

30

40

50 60 70 80 Number of bilayers (x)

90

100

110

120

Figure 10.12 Variations of experimental average transmittance with number of bilayers of TiO2 /SiO2 nanoparticulate thin films. Refractive index = 1.30, bilayer thickness = 20.18 nm. Source: Adapted with permission from [60] Copyright (2011) Taiwan Institute of Chemical Engineers.)

absorption intensity at wavelength 664 nm of the MBsolution, as a function of irradiation time gives evidence for photocatalytic decomposition of MB molecules with (TiO2 /SiO2 )x nanoparticulate thin film. Furthermore, the rates of A/Ao decrease were found to be dependent on the bilayer number. The higher the bilayer number, the faster the decrease of the A/Ao ratio. This can be explained by the ability of the TiO2 layers to absorb UV light, which is determined by the bilayer number and absorption onset. [13] Under the conditions of Lin et al. [60], MB molecules can be completely decomposed by (TiO2 /SiO2 ) 120 in 150 min. The self-cleaning properties of (TiO2 /SiO2 )x nanoparticulate thin films were also revealed by the photocatalytic decomposition of MB molecules, which were adsorbed on the surfaces of thin films by immersing in 50 ppm aqueous MB solution for 1 h, under UV irradiation, as shown in Figure 10.13b. The rates of photocatalytic decomposition were again found to be dependent on the bilayer number. The higher the bilayer number, the faster the rate of photocatalytic decomposition of MB molecules. Under the conditions of Lin et al. [60], MB molecules on the surfaces can be completely decomposed by (TiO2 /SiO2 )120 in 120 min. It is noteworthy that the as-fabricated nanoparticulate thin films (TiO2 /SiO2 )x are superhydrophic. The water contact angles measured on (TiO2 /SiO2 )x with x = 1, 3, 5, 10, 30, 60 and 120 showed that they were always lower than 5◦ (not shown). As shown in Figure 10.14, the water contact angles on the dodecyltrichlorosilane-modified (TiO2 /SiO2 )x nanoparticulate thin films were measured to be almost the same at 125◦ . This angle is higher than that measured on the dodecyltrichlorosilane-modified plain glass at 105◦ (not shown). The self-cleaning properties were evaluated by monitoring the contact angle at the surface after UV irradiation as a function of time. Figure 10.14 shows significant decrease in water contact angle during irradiation. It was found that the rates of water contact angle decrease were dependent on the bilayer numbers of the nanoparticulate thin films. Under

290

Self-Cleaning Materials and Surfaces: A Nanotechnology Approach 0-bilayer 10-bilayer 30-bilayer

1.0

60-bilayer 90-bilayer 120-bilayer

30-bilayer 60-bilayer

1.0

0.6

0.6 A/A0

0.8

A/A0

0.8

90-bilayer 120-bilayer

0.4

0.4

0.2

0.2 MB conc : 10 ppm

0.0

0

30

MB conc : 50 ppm

60 90 Time (min)

120

150

0.0

0

30

60 Time (min)

(a)

90

120

(b)

Figure 10.13 Photocatalytic decomposition of Methylene Blue molecules (a) in 10 ppm aqueous Methylene Blue solution by (TiO2 /SiO2 )x nanoparticulate thin films, (b) which were adsorbed on the surfaces of (TiO2 /SiO2 )x nanoparticulate thin films by immersing in 50 ppm aqueous Methylene Blue solution for 1 h, under UV irradiation. Source: Adapted with permission from [60] Copyright (2011) Taiwan Institute of Chemical Engineers.

Average static contact angle (degree)

140

10-bilayer 30-bilayer 60-bilayer 120-bilayer

1-bilayer 3-bilayer 5-bilayer

120 100 80 60 40 20 0

0

10

20

30 Time (min)

40

50

60

Figure 10.14 Variations of water contact angle on silanized (TiO2 /SiO2 )x nanoparticulate thin films with UV irradiation time. Source: Adapted with permission from [60] Copyright (2011) Taiwan Institute of Chemical Engineers.

Fabrication of Antireflective Self-Cleaning Surfaces

291

the conditions of Lin et al. [60], the water contact angle on (TiO2 /SiO2 )120 decreased from 125◦ to about 0◦ in 60 min.

10.5.2

Superhydrophobic Self-Cleaning Surfaces with Antireflective Properties

The Lotus effect [55, 56], a synonym of superhydrophobic self-cleaning, of lotus leaves is caused by both the hierarchical roughness of the leaf surface from micrometer-sized papillae having nanometer-sized branch-like protrusions and the intrinsic material hydrophobicity of a surface layer of epicuticular wax covering these papillae [61,62]. Actually, the natural surfaces have inspired researchers to consider biomimetic approaches for generating functional artificial surfaces. A recent review of the design and creation of surfaces with special wettability, such as superhydrophilicity, superhydrophobicity, superoleophilicity, superoleophobicity, superamphiphilicity, superamphiphobicity, superhydrophobicity/superoleophilicity, and reversible switching between superhydrophobicity and superhydrophilicity is available [41]. A comprehensive overview of the characterization and technological processes to produce man-made surfaces with similar properties to the biological ones can be found in a recently published book [63]. While surface roughness is necessary for surface wettability, surface roughness must be minimized to reduce the light scattering so that light transparency can be achieved. Establishment of the appropriate surface structure length scale, therefore, is crucial to the fabrication of thin films that exhibit both properties. Xiu et al. [64] demonstrated that a sol–gel process using a eutectic liquid can be invoked to form superhydrophobic, optically transparent films on glass slides. Previous investigations of the optical transparency of superhydrophobic films have also been reviewed by Xiu et al. [64]. Among them, an ELbL processing scheme that can be utilized to create transparent superhydrophobic films from SiO2 nanoparticles of various sizes has been demonstrated by Bravo et al. [65]. Their films consisted of three main parts: adhesion, body, and top layers. Bilayers of cationic and anionic polyelectrolytes were deposited to create the adhesion-promoting multilayer. For polyelectrolyte-silica body layers, two different sized silica nanoparticles were used. Furthermore, top layers containing only one-size of silica nanoparticles were also added to enhance the two-scale roughness necessary for superhydrophobic films. The final assembly was rendered superhydrophobic with sintering and silane treatment. As shown in Figure 10.15, optical transmission levels above 90% throughout most of the visible region of the spectrum were realized in optimized coatings. As shown in Figure 10.16, advancing water droplet contact angles as high as 160◦ with low contact angle hysteresis (160◦ and a roll-off angle variously taken to be