A-PDF Merger DEMO : Purchase from www.A-PDF.com to remove the watermark
Volume 3
7390 tp.indd 1
11/20/09 11:49:33 AM
ANNUAL REVIEW OF NANO RESEARCH Series Editors: Guozhong Cao (University of Washington, USA) C Jeffrey Brinker (University of New Mexico and Sandia National Laboratories, USA)
Vol. 1: ISBN-13 ISBN-10 ISBN-13 ISBN-10
978-981-270-564-8 981-270-564-3 978-981-270-600-3 (pbk) 981-270-600-3 (pbk)
Vol. 2: ISBN-13 978-981-279-022-4 ISBN-10 981-279-022-5 ISBN-13 978-981-279-023-1 (pbk) ISBN-10 981-279-023-3 (pbk) Vol. 3: ISBN-13 978-981-4280-51-8 ISBN-10 981-4280-51-8
YHwa - Annual Rev of Nano Res Vol. 3.pmd
2
11/25/2009, 6:24 PM
Volume 3
Editors
Guozhong Cao Qifeng Zhang University of Washington, USA
C. Jeffrey Brinker University of New Mexico and Sandia National Laboratories, USA
World Scientific NEW JERSEY
7390 tp.indd 2
•
LONDON
•
SINGAPORE
•
BEIJING
•
SHANGHAI
•
HONG KONG
•
TA I P E I
•
CHENNAI
11/20/09 11:49:33 AM
Published by World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
ANNUAL REVIEW OF NANO RESEARCH, Vol. 3 Copyright © 2010 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.
ISBN-13 978-981-4280-51-8 ISBN-10 981-4280-51-8
Printed in Singapore.
YHwa - Annual Rev of Nano Res Vol. 3.pmd
1
11/25/2009, 6:24 PM
TABLE OF CONTENTS
Preface Contributing Authors
xv xvii
Chapter 1. Nanoscale Biosensors and Biochips Wayne R. Leifert, Richard V. Glatz, Kelly Bailey, Tamara Cooper, Marta Bally, Brigitte Maria Stadler, Erik Reimhult and Joseph G. Shapter 1. General Introduction 2. Biological Detectors Used in Biosensing and Biochips 2.1. G-Protein Coupled Receptor Biosensors (GPCRs) 2.2. Pore-Forming Proteins 2.3. Cell- and Viral-Based Sensing 3. Lipid Supports for Biosensor and Biochip Fabrication 3.1. Why Functionalize Biosensors with Lipid Membranes? 3.2. Methods to Assemble Supported Lipid Membranes 3.3. Supported Lipid Membrane Platforms 3.4. Advanced Sensors Functionalized with Lipid Membranes 3.5. Future Perspectives 4. Nanopatterning for Biosensing and Biochip Fabrication 4.1. Parallel Nanopatterning Methods 4.2. Serial Nanopatterning Methods 5. Sensing Substrates: A Closer Look at Nanotubes 5.1. Carbon Nanotube Electrodes for Communicating with Redox Proteins 5.2. Aligned Carbon Nanotube Electrodes for Direct Electron Transfer to Enzymes 6. Reporter Technologies: Nano-Sized Labels for Biosensing Applications 6.1. Biosensors Utilizing Optical Reporting 6.2. Biosensors Utilising Electrochemical Reporting
v
1
1 3 3 13 16 25 25 27 29 32 33 34 34 38 40 40 43 45 46 50
vi
Annual Review of Nano Research
7. Biosensing Applications 7.1. Medical 7.2. Food and Wine 7.3. Explosives and Biowarfare 7.4. Environmental 8. Conclusion References
53 53 55 56 57 59 60
Chapter 2. Surface Modifications and Applications of Magnetic and Selective Nonmagnetic Nanoparticles Rui Shen and Hong Yang
83
1. Introduction 2. General Approaches to Surface Modification of Nanostructures 2.1. Adsorption and Self-Assembly 2.2. Surface Modification Based on Organic Reactions 2.3. Surface Modification Based on Polymerization 2.4. Surface Modification with Inorganic Layers Based on Sol-Gel Approaches 2.5. Surface Modification with Multiple or Composite Layers 2.6. Experimental Designs 2.7. Surface Modification in the Synthesis of Hollow Spheres 3. Surface Modification of Magnetic Nanostructures 3.1. Oxides 3.2. Metals 3.3. Metal Alloys 4. Surface Modification in the Synthesis of Higher-Ordered and Complex Nanostructures 4.1. Hollow and Yolk-Shell Nanostructures 4.2. Anisotropic and Onion-Like Nanostructures 4.3. Other Higher Ordered Nanostructures 5. Applications of Surface-Modified Magnetic Nanoparticles 5.1. Surface Modifications in Nonbiological Applications 5.2. Surface Modifications in Biological Applications 6. Conclusion
83 86 87 90 92 94 99 100 102 103 104 108 111 114 115 120 122 127 127 129 137
Table of Contents
vii
Acknowledgments References
137 137
Chapter 3. Progress in Bionanocomposite Materials Eduardo Ruiz-Hitzky, Margarita Darder and Pilar Aranda
149
1. Introduction 2. Bionanocomposites for Bioplastics 3. Bionanocomposites for Biomedical Applications 4. Bionanocomposites for Sensor Devices and Other Applications 5. Concluding Remarks Acknowledgments References
149 152 162 171 180 181 181
Chapter 4. Mesoporous Silica Nanoparticles: Synthesis and Applications Juan L. Vivero-Escoto, Brian G. Trewyn and Victor S.-Y. Lin 1. Introduction 2. Synthesis of Mesoporous Silica Nanoparticles 2.1. Control of Morphology 2.2. Control of Surface Functionalization 3. Catalysis 3.1. Cooperative Catalysis (Acid/Base) 3.2. Gatekeeping Effect 3.3. Other Applications in Catalysis 4. Biotechnological and Biomedical Applications 4.1. Uptake and Intracellular Performance of MSNs 4.2. Controlled Delivery Systems 4.3. Biosensors 4.4. Multimodal Cell Imaging 5. Conclusions and Outlook Acknowledgments References
191
191 193 194 199 202 202 204 206 208 209 212 219 222 225 226 226
viii
Annual Review of Nano Research
Chapter 5. Nanostructured Mesoporous Materials as Drug Delivery Systems Isabel Izquierdo-Barba, Daniel Arcos and Maria Vallet-Regí
233
1. Introduction 2. Cytotoxicity, Biocompatibility and Bioactivity of Silica Mesoporous Materials 3. Tailoring Mesoporous Drug Delivery Systems-Textural Properties Considerations 3.1. Pore Diameter 3.2. Surface Area 3.3. Pore Volume 3.4. Increasing the Surface Area - The Hybrid Route 4. Surface Functionalization of Mesoporous Drug Delivery Systems 5. Dosage in Mesoporous Materials 6. Mesoporous Materials for Intracellular Targeting 6.1. Cell Mechanism for Particles Internalization 6.2. Microstructural Considerations for SiO2 Nanoparticles Intracellular Targeting 7. Stimuli-Responsive Mesoporous Materials 7.1. Drug Release Mediated by Chemical Stimuli 7.2. Drug Release Mediated by Thermal Stimuli 7.3. Drug Release Mediated by Photo-Chemical Stimuli 7.4. Drug Release Mediated by Magnetic Stimuli 8. Conclusions and Outlook Acknowledgments References
234
247 250 254 254
Chapter 6. Chemical Synthesis, Self-Assembly and Applications of Magnetic Nanoparticles Sheng Peng, Jaemin Kim and Shouheng Sun
275
1. Introduction 1.1. General Background 1.2. Chemical Syntheses of Nanoparticles
275 275 277
236 239 239 244 244 245
256 259 260 263 263 264 267 269 269
Table of Contents
2. Ferrite Nanoparticles: MFe2O4 (M = Fe, Mn, Co) 2.1. Chemical Syntheses of Spherical Ferrite Nanoparticles 2.2. Shape-Controlled Synthesis and Self-Assembly 2.3. Surface Modification for Biological Applications 3. Metallic Iron, Cobalt and Iron-Cobalt Alloy Nanoparticles 3.1. Synthesis and Stabilization of Metallic Fe, Co, and FeCo Particles 3.2. Self-Assembly, Shape-Controlled Synthesis of Fe and Co 4. Tetragonal (L10-Phase) Hard Magnetic FePt Nanoparticles and Their Applications 4.1. General Chemical Syntheses of fcc-FePt Nanoparticles and the Phase Change via Thermal Treatment 4.2. Shape Controlled FePt Nanoparticles and Their Self-Assembly 4.3. Synthesis of Dispersible fct-FePt Nanoparticles 5. Rare-Earth Hard Magnets: Going Into Nanoscale 6. Summary and Outlook Acknowledgments References Chapter 7. Recent Development and Applications of Nanoimprint Technology Xing Cheng and L. Jay Guo 1. Introduction 2. Material Flow Behavior and the Associated Polymer Chain Alignment in NIL 2.1. Polymer Chain Alignment in Nanoimprinted Polymer Micro- and Nanostructures 2.2. Improving the Performance of Polymer Electronics by Nanoimprint-Induced Chain Orientation 3. Reversal Nanoimprint Lithography 3.1. Principles of Reversal Nanoimprint 3.2. Residual Layer Removal in Reversal Nanoimprint 3.3. Building 3D Polymer Nanostructures 3.4. Process Yield of Reversal Nanoimprint
ix
279 280 283 286 288 289 295 298 299 299 301 303 307 307 307
317 317 320 320 323 325 325 326 328 332
x
Annual Review of Nano Research
4. Recent Applications of NIL 4.1. Organic Solar Cells with Imprinted Nanoscale Morphology 4.2. Nanoimprinting Nafion® Film for Micro Fuel Cell Applications 5. Roll-To-Roll Nanoimprint Lithography (R2RNIL) 6. Conclusion Acknowledgments References Chapter 8. Three-Dimensional Nanostructure Fabrication by Focused-Ion-Beam Chemical-Vapor-Deposition Shinji Matsui 1. Introduction 2. Three-Dimensional Nanostructure Fabrication 2.1. Fabrication Process 2.2. Three-Dimensional Pattern Generating System 3. Nanoeletromechanics 3.1. Young’s Modulus Measurement 3.2. Free-Space-Nanowiring 3.3. Nanoelectrostatic Actuator 4. Nanooptics: Brilliant Blue Observation from a Morpho-Butterfly-Scale Quasi-Structure 5. Nanobiology 5.1. Nanoinjector 5.2. Nanomanipulator 6. Summary References
335 335 339 341 346 348 348
351 351 352 353 356 359 359 364 371 373 376 376 379 382 382
Chapter 9. Dye-Sensitized Solar Cells Based on Nano-Structured Zinc Oxide Qifeng Zhang and Guozhong Cao
385
1. Introduction 2. Nanostructures Offering Large Specific Surface Area
385 390
Table of Contents
2.1. ZnO Nanoparticulate Films 2.2. Nanoporous Structured ZnO Films 2.3. Other Nano-Structured ZnO Films 3. Nanostructures with Direct Pathway for Electron Transport 3.1. ZnO Nanowires 3.2. ZnO Nanotubes 3.3. ZnO Nanotips 3.4. ZnO Nanoflowers 3.5. Dendritic ZnO Nanowires 4. Core-shell Structures with ZnO Shell for Reduced Recombination Rate 4.1. Fabrication of Core-Shell Structures and Influence of Shell Thickness 4.2. The Role of ZnO Shell 5. Light Scattering Enhancement Effect 5.1. ZnO Aggregates 5.2. One-Dimensional ZnO Nanostructures for Light Scattering 6. Limitation on ZnO-Based DSSCs 6.1. Instability of ZnO in Acidic Dyes 6.2. Low Electron Injection Efficiency 6.3. New Types of Photosensitizers for ZnO 7. Conclusion and Outlook 7.1. Surface Modification of ZnO Aggregates - An Indirect Method for TiO2 Aggregates 7.2. Hydrothermal Growth of TiO2 Nanoparticle Aggregates 7.3. Emulsion-Assisted Synthesis of TiO2 Nanostructure Aggregation 7.4. Electrostatic Spray Deposition Fabrication of TiO2 Aggregates 7.5. Synthesis of Porous-Structured TiO2 Spheres Acknowledgments References
xi
390 394 397 399 400 403 403 404 405 406 407 408 410 412 415 416 416 420 422 423 426 427 428 429 429 430 430
xii
Annual Review of Nano Research
Chapter 10. Nanocomposites as High Efficiency Thermoelectric Materials Suraj Joottu Thiagarajan, Wei Wang and Ronggui Yang 1. Introduction to Thermoelectricity 2. Nanocomposites as Highly Efficient Thermoelectric Materials 2.1. Modeling of Phonon Transport 2.2. Modeling of Electron Transport 3. Synthesis of Thermoelectric Nanocomposites 3.1. Preparation of Nanocomposites by Compaction Techniques 3.2. Synthesis of Nanocomposites by Phase Separation 4. Recent Achievements in Thermoelectric Nanocomposites 4.1. Bi2Te3-Based Nanocomposites for Low Temperature Applications 4.2. Medium Temperature Materials 4.3. High Temperature Materials 5. Summary Acknowledgments References
441
442 450 452 456 459 460 467 469 470 473 477 479 480 481
Chapter 11. Nanostructured Materials for Hydrogen Storage Saghar Sepehri and Guozhong Cao
487
1. Introduction 2. Hydrogen Storage by Physisorption 2.1. Nanostructured Carbon 2.2. Zeolites 2.3. Metal – Organic Frameworks 2.4. Clathrates 2.5. Polymers with Intrinsic Microporosity 3. Hydrogen Storage by Chemisorption 3.1. Metal and Complex Hydrides 3.2. Chemical Hydrides 3.3. Nanocomposites
487 490 491 493 494 495 497 498 499 502 504
Table of Contents
xiii
4. Summary Acknowledgments References
511 511 512
Chapter 12. Recent Advances in the Characterization of Mesoporous Materials by Physical Adsorption Matthias Thommes
515
1. Introduction 2. General Aspects of Surface and Pore Size Analysis by Physisorption 3. Pore Condensation and Hysteresis in Mesoporous Materials 3.1. Pore Condensation 3.2. Interpretation of Adsorption Hysteresis 4. Comments to Mesopore Size Analysis 4.1. Classical Methods 4.2. Pore Size Analysis by Non Local Density Functional Theory (NLDFT) 4.3. Hysteresis and Pore Size Analysis 5. Summary and Conclusion Acknowledgment References
516 521 524 524 526 542 542 543 546 548 550 550
This page intentionally left blank
PREFACE
Annual Review of Nano Research publishes excellent review articles in selected topic areas authored by those who are authorities in their own subfields of nanotechnology with two vital aims: (1) to present a comprehensive and coherent distilling of the state-of-the-art experimental results and understanding of theories detailed from the otherwise segmented and scattered literature, and (2) to offer critical opinions regarding the challenges, promises, and possible future directions of nano research. The third volume of Annual Review of Nano Research includes 11 articles offering a concise review detailing recent advancements in a few selected subfields in nanotechnology. The first topic to be focused upon in this volume is the bio-applications of and bio-inspired nanostructured materials. The second featured subfield is the recent advancement in the synthesis and fabrication of nanomaterials or nanostructures. Energy related applications of nanostructures and nanomaterials are the third focal topic in this volume. We want to thank all the contributing authors for their time and efforts devoted to the excellent review articles published in this volume. Mr. Yeow-Hwa Quek from World Scientific Publishing was responsible for much of the coordination necessary to make the publication of this volume possible.
Guozhong Cao Seattle, WA Qifeng Zhang Seattle, WA C. Jeffrey Brinker Albuquerque, NM
xv
This page intentionally left blank
CONTRIBUTING AUTHORS
Aranda, Pilar * Instituto de Ciencia de Materiales de Madrid, Spain Arcos, Daniel * Universidad Complutense de Madrid, Madrid, Spain * Networking Research Center on Bioengineering, CIBER-BBN, Spain Bailey, Kelly * Commonwealth Scientific and Industrial Research Organization (CSIRO), Australia * The University of Adelaide, Australia Bally, Marta * Institute for Biomedical Engineering, Switzerland Cao, Guozhong * University of Washington, USA Cheng, Xing * Texas A&M University, USA Cooper, Tamara * Commonwealth Scientific and Industrial Research Organization (CSIRO), Australia * The University of Adelaide, Australia Darder, Margarita * Instituto de Ciencia de Materiales de Madrid, Spain * Instituto Madrileño de Estudios Avanzados en Materiales (IMDEAMateriales), Spain Glatz, Richard V. * South Australian Research and Development Institute (SARDI), Australia xvii
xviii
Annual Review of Nano Research
Guo, L. Jay * The University of Michigan, USA Izquierdo-Barba, Isabel * Universidad Complutense de Madrid; Madrid, Spain * Networking Research Center on Bioengineering, CIBER-BBN, Spain Kim, Jaemin * Brown University, USA Leifert, Wayne R. * Commonwealth Scientific and Industrial Research Organization (CSIRO), Australia Lin, Victor S.-Y. * U.S. Department of Energy Ames Laboratory, USA * Iowa State University, USA Matsui, Shinji * University of Hyogo 3-1-2 Koto, Japan Peng, Sheng * Brown University, USA Reimhult, Erik * Laboratory for Surface Science and Technology, Switzerland Ruiz-Hitzky, Eduardo * Instituto de Ciencia de Materiales de Madrid, Spain Sepehri, Saghar * University of Washington, USA Shapter, Joseph G. * Flinders University, Australia Shen, Rui * University of Rochester, USA
Contributing Authors
xix
Stadler, Brigitte Maria * The University of Melbourne, Australia Sun, Shouheng * Brown University, USA Thiagarajan, Suraj Joottu * University of Colorado, USA Thommes, Matthias * Quantachrome Instruments, USA Trewyn, Brian G. * U.S. Department of Energy Ames Laboratory, USA * Iowa State University, USA Wang, Wei * University of Colorado, USA Yang, Hong * University of Rochester, USA Yang, Ronggui * University of Colorado, USA Vallet-Regí, Maria * Universidad Complutense de Madrid; Madrid, Spain * Networking Research Center on Bioengineering, CIBER-BBN, Spain Vivero-Escoto, Juan L. * U.S. Department of Energy Ames Laboratory, USA * Iowa State University, USA Zhang, Qifeng * University of Washington, USA
CHAPTER 1 NANOSCALE BIOSENSORS AND BIOCHIPS
Wayne R. Leifert1,*, Richard V. Glatz2, Kelly Bailey3,4, Tamara Cooper3,4, Marta Bally5, Brigitte Maria Stadler6, Erik Reimhult7 and Joseph G. Shapter8 1
Commonwealth Scientific and Industrial Research Organization (CSIRO), Division of Human Nutrition, Adelaide, Australia; 2South Australian Research and Development Institute (SARDI), Department of Entomology, Adelaide, Australia; 3Commonwealth Scientific and Industrial Research Organization (CSIRO), Division of Molecular and Health Technologies, Adelaide, Australia; 4 School of Molecular and Biomedical Science, The University of Adelaide, Adelaide, Australia; 5Laboratory of Biosensors and Bioelectronics, Institute for Biomedical Engineering, ETH Zurich, Zurich, Switzerland; 6Centre for Nanoscience and Nanotechnology, Department of Chemical and Biomolecular Engineering, The University of Melbourne, Melbourne, Australia; 7Laboratory for Surface Science and Technology, Department of Materials, ETH Zurich, Zurich, Switzerland; 8School of Chemistry, Physics and Earth Sciences, Flinders University, Adelaide, Australia *Corresponding author, Email:
[email protected]
1. General Introduction Recent advances in molecular biology, surface chemistry, protein purification, signal transduction/amplification, lipid chemistry and nanofabrication technologies have converged in a relatively new field of science, that of molecular biosensing. The growing level of interest in biosensing research has seen the formation of a range of specific journals reporting on advances in the field. Biosensors and biochips have many potential (and some current) applications including provision of point-ofcare diagnostic tools, high-throughput drug discovery tools and in-field sensing tools for a variety of compounds including toxins and/or contaminants. Biosensors are generally accepted as being analytical
1
2
Leifert et al.
devices based on a biologically active or biologically mimetic compound (a detector) coupled to a physical signal transduction mechanism (a transducer/reporter). The interaction of analyte with its biomolecular detector is thus exploited to produce an effect that can be measured through the transducer. Biochips generally refer to an array of individual biosensors, and can also be referred to as nucleotide or protein microarrays. This review explores examples of the various approaches to biosensor and biochip fabrication, including examination of components important to each system. We discuss advantages and disadvantages of biosensors that exploit whole cells as detectors and those which make use of one or several specific molecules, the most widely utilized being membrane-associated proteins, such as G-protein coupled receptors (GPCRs) and ion-channels, due to their diversity and current importance for sensing and screening technologies. There are two major challenges to establishing functionally active biological components within a sensor or chip design, these being the controlled capture and positioning of the biological detector onto a surface and maintenance of the detector’s functional/structural integrity. In the case of membrane proteins, it is vital that an appropriate hydrophobic environment is available to maintain protein structure and function. The use of lipid supports for this purpose, is being widely studied, and is discussed in this review. In combination with appropriate surface compositions, various techniques of nanopatterning technologies are important in the fabrication of biosensors in order to control the location, distribution, amount and orientation of the biomolecules on the surface. There are also a range of physical and biological substrates on which sensor and chip platforms are built and we pay particular attention to nanotubes as physical substrates due to their promising application in electrochemical biosensing. In order to detect changes occurring at the sensor or chip surface, a variety of transduction techniques exist and we discuss a range of nanosized reporter labels for their potential for biosensing applications. Finally, we touch on some of the applications and specific analytes commonly monitored using biosensing today.
Nanoscale Biosensors and Biochips
3
2. Biological Detectors Used in Biosensing and Biochips Since the first description of the biosensor concept by Clark Jr. and Lyons in 1962 involving glucose oxidase and the oxygen electrode [1], a variety of front-end biological detectors have been investigated for use in biosensing. These bio-recognition elements include enzymes, membrane proteins such as cell-receptors and ion-channels, antibodies, nucleic acids, virus particles and intact cells. We begin the discussion with an investigation of membrane proteins, which as diverse biomolecules crucial in cellular signaling and communication as well as regulation of transport into and out of the cell are of significant interest in the fabrication of biosensors and biochips. Because of their diversity, biological importance and level of characterization, the two key classes of membrane proteins utilized are the GPCR family and ion-channels. 2.1. G-Protein Coupled Receptor Biosensors (GPCRs) 2.1.1. Importance of GPCRs Many disease processes involve aberrant or altered GPCR signaling dynamics and GPCRs represent the most significant target class for medicinal pharmaceuticals (≈50% of marketed drugs, see Table 1) [2]. GPCRs are associated with almost every major therapeutic category or disease class, including pain, asthma, inflammation, obesity, cancer, as well as cardiovascular, metabolic, gastrointestinal and central nervous system diseases [3]. It is this vitally important function of these cellsurface receptors combined with the huge diversity of specific ligands they bind, which makes GPCRs so physiologically significant and attractive for biosensing applications. Therefore, there is a need to develop sophisticated and appropriate GPCR biosensors for the detection of a variety of ligands (Figure 1). This section focuses on some of the available cell-free GPCR assay nanotechnologies [4] and describes some of the more sophisticated functional GPCR biosensors. Cell-free biosensors have the potential advantage of being applicable to a range of assay environments in which cells may be damaged.
4
Leifert et al.
Table 1. Examples of some pharmaceuticals which target GPCRs for the indicated condition or disease state.
Brand Name (generic)
G-protein coupled receptor(s)
Disease/Indication
Zyprexa (Olanzapine)
Serotonin 5-HT2 and Dopamine
Schizophrenia, Antipsychotic
Risperdal (Risperidone)
Serotonin 5-HT2
Schizophrenia
Claritin (Loratidine)
Histamine H1
Rhinitis, Allergies
Imigran (Sumatriptan)
Serotonin 5-HT1B/1D
Migraine
Cardura (Doxazosin)
α-adrenoceptor
Prostate hypertrophy
Tenormin (Atenolol)
β1-adrenoceptor
Coronary heart disease
Serevent (Salmeterol)
β2-adrenoceptor
Asthma
Duragesic (Fentanyl)
Opioid
Pain
Imodium (Loperamide)
Opioid
Diarrhea
Cozaar (Losartan)
Angiotensin II
Hypertension
Zantac (Ranitidine)
Histamine H2
Peptic ulcer
Cytotec (Misoprostol)
Prostaglandin PGE1
Ulcer
Zoladex (Goserelin)
Gonadotrophin-releasing factor
Prostate cancer
Requip (Ropinirole)
Dopamine
Parkinson’s disease
Atrovent (Ipratropium)
Muscarinic
Chronic obstructive pulmonary disease (COPD)
GPCR activation can be initiated by a wide variety of stimuli such as light, odorants, neurotransmitters and hormones (Figure 1). In cells, the extracellular ligand is specifically and sensitively detected by a cell surface GPCR. Once binding/recognition takes place, the GPCR triggers the activation of a cellular heterotrimeric G-protein (guanine nuceleotidebinding protein) complex consisting of Gα, Gβ and Gγ subunits (Figure 1). Finally, the “signal transduction” cascade (in whole cells at
Nanoscale Biosensors and Biochips
5
least) involves the activated G-proteins altering the activity of downstream “effector” protein(s) to yield a response that can be used to detect the binding event. This is the basis of many existing screening assays and some biosensor designs have exploited this approach, combined with different surface-attachment and transduction technologies. In cell-free GPCR assays discussed in this section, host cells which have been transfected with DNA encoding a particular GPCR of interest allow the cellular expression of the GPCR. Subsequently, the cells are treated in such a way as to allow a partial purification of the GPCRs (in their cell membranes) to obtain an ongoing supply of GPCRs. The purification process can result in small (nanometer scale) crude membrane fragments containing the GPCRs, which are suitable as detector molecules for biosensor applications [5]. 2.1.2. Surface Capture of GPCRs The arraying of membrane GPCRs has required appropriate surface chemistry for the immobilization of the lipid phase containing the GPCR of interest [6-8]. Surface modification with γ-aminopropylsilane (an amine presenting surface) provided the best combination of properties to allow surface capture of the GPCR-G-protein complex from crude membrane preparations, resulting in microspots of approximately 100 µm diameter. Atomic force microscopy (AFM) demonstrated that the height of the supported lipid bilayer was approximately 5 nm, corresponding to GPCRs confined in a single, supported lipid layer scaffold [9]. Using these chemically-derivatized surfaces, it was possible to demonstrate capture of fluorescently labeled β1, β2, and α2A subtypes of the adrenergic receptor, as well as neurotensin-1 receptors and D1dopamine receptors. Dose-response curves using the fluorescentlylabeled ligands gave IC50 values in the nM range suggesting that the GPCR-G-protein complex was largely preserved and biologically intact in the microspot. Furthermore, good long-term stability was achieved. Waller et al. [10] conjugated dextran beads with dihydroalprenolol, an antagonist of β 2 -adrenergic receptors (β-AR). This allowed the capture of solubilized β-AR to this immobilized surface ligand. The βAR was expressed as a fusion protein with green fluorescent protein
6
Leifert et al. GPCR ligands Acetylcholine
Ghrelin
Opioids
Adenosine
Glucagon
Orexin
Adrenaline
Glutamate
Oxytocin
Adrenocorticotropic hormone
Gonadotropin-releasing hormone
Parathyroid hormone
Angiotensin II
Growth hormone-releasing factor
Photons (light)
Bradykinin
Growth-hormone secretagogue
Platelet activating factor
Calcitonin
Histamine
Prolactin releasing peptide
Chemokines
Luteinising hormone
Prostaglandins
Cholecystokinin
Lymphotactin
Secretin
Corticotropin releasing factor
Lysophospholipids
Serotonin
Dopamine
Melanocortin
Somatostatin
Endorphins
Melanocyte-stimulating hormone
Substances P, K
Endothelin
Melatonin
Thrombin
Enkephalins
Neuromedin-K
Thromboxanes
Fatty acids
Neuromedin-U
Thyrotropin
Follitropin
Neuropeptide-FF
Thyrotropin releasing hormone
GABA
Neuropeptide-Y
Tyramine
Galanin
Neurotensin
Urotensin
Gastric inhibitory peptide
Noradrenaline
Vasoactive intestinal peptide
Gastrin
Odorants
Vasopressin
ligand
GPCR
lipid bilayer membrane
γ
α β
γ
Ni2+ SUPPORT
MATRIX
Figure 1. A list of some of the known endogenous and exogenous GPCR ligands and a schematic depicting the transmembrane topology of a typical “serpentine” G-protein coupled receptor (GPCR) with its associated heterotrimeric G-protein complex. The membrane patch containing the GPCR with associated G-proteins is schematically shown attached to a theoretical solid support matrix. The receptor polypeptide chain traverses the plane of the membrane phospholipid bilayer seven times. The hydrophobic transmembrane segments of the GPCR are indicated by spirals. The ligand can bind to the receptor from the “extracellular” (outer) surface or depending on the receptor type, to a site deep within the receptor, surrounded by the transmembrane regions of the receptor protein. In this way, the receptor can act as a detector of its ligands. The G-proteins (Gα and Gβγ) are shown to interact with the “cytoplasmic” side of the receptor.
Nanoscale Biosensors and Biochips
7
(GFP), thus allowing fluorescent measurement of bound receptor possible. Thus, it was possible to screen for ligands of the β-AR through the reduction in fluorescence as receptor was removed from the beads due to competition of free (added) ligand. Another successful bead-based approach used paramagnetic beads to capture CCR5 receptors from a cell lysate held within a lipid bilayer [11]. More recently, site directed immobilization of membrane extracts containing either the M2muscarinic receptor or H1-histamine receptor, using complementary oligonucleotides has been investigated (unpublished data). Sequencedirected immobilization of oligo-tagged vesicles carrying GPCRs could potentially lead to the development of a self-sorting array platform for a large number of different receptor sub-types, through the inherent selectivity of complementary strands of oligonucleotides. 2.1.3. Ligand-Binding at GPCRs Ligand binding to a GPCR attached to a surface has been reported for the chemokine CCR5 receptor using surface plasmon resonance (SPR) [12]. For such GPCR surface display, purification of the GPCR has not always been necessary and crude membrane preparations have either been fused with an alkylthiol monolayer (approximately 3 nm thickness) formed on a gold-coated glass surface, or onto a carboxymethyl modified dextran sensor surface [13]. One problem of surface based assays is the difficulty in obtaining the correct orientation of the receptor once attached to the surface. This problem was overcome by using conformationally-dependent antibodies [14]. In this biosensor application, SPR has a distinct advantage as a screening tool since this technique can detect the cognate ligand without requiring fluorescent or radio-labeling. This allows SPR to be used in complex fluids of natural origin thus simplifying the development of assay technologies. Martinez et al. [15] used total internal reflection fluorescence (TIRF) to demonstrate ligand binding to the neurokinin-1 GPCR by surface immobilization of membrane fragments containing this receptor. The GPCR was expressed as a biotinylated protein using mammalian cells and could be selectively immobilized on a quartz sensor surface coated
8
Leifert et al.
with streptavidin (streptavidin binds biotin with extremely high affinity). Using this approach, it was not necessary to detergent-solubilize and reconstitute the neurokinin-1 receptors, thus avoiding the deleterious effect(s) associated with such processes. The preparation of the biotinylated receptors allowed for a high-affinity interaction between biotin and streptavidin and thus a template-directed and uniform orientation of the neurokinin-1 receptor on the support matrix. TIRF measurements were made using a fluorescent-labeled agonist (i.e., the cognate agonist substance-P labeled with fluorescein). The highly sensitive TIRF fluorescence detection methodology was able to resolve the binding of fluorescently-tagged ligand (agonist) to as little as one attomol of receptor molecules [15]. This sensitivity far exceeds that of current physical approaches to biosensing (e.g. gas chromatography, mass spectrometry) and is a major reason why biosensing has become an important research area. 2.1.4. Detecting GPCR Conformational Changes The detection of intrinsic conformational changes in the GPCR following ligand (agonist) activation generally involves the use of fluorescence-based techniques and has been limited to date. One study demonstrated the immobilization of β2-adrenergic receptors onto glass and gold surfaces [16]. The receptors were site-specifically labeled with the fluorophore tetramethyl-rhodamine-maleimide at Cysteine 265 (Cys265) and the agonist-induced signal was large enough to detect using a simple intensified charge-coupled device (ICCD) camera image. Therefore, it was suggested that the technique may be useful for drug screening with GPCR arrays. In a recent study, ligand binding to the β2-adrenergic receptor has been demonstrated using plasmon waveguide resonance (PWR) [17]. Using this technique, changes in the refractive index upon ligand binding to surface-immobilized receptor results in a shift in the PWR spectra. Previously, PWR technology was used for detection of conformational changes in a proteolipid membrane containing the human δ-opioid receptor following binding of several types of ligands [18]. Although the
Nanoscale Biosensors and Biochips
9
ligands in that study [18] were of similar molecular weight, there were distinctly different refractive index changes induced by ligand binding and these were too large to be accounted for by differences in the mass alone. The inference from this finding was that a ligand-specific conformation change in the receptor protein may have been detected, a phenomenon that further increases the specificity of ligand-mediated PWR spectral alterations. 2.1.5. GTP Binding at G-Protein Subunits GPCR biosensors can also utilize the use of non-hydrolyzable GTP-analogs such as radiolabeled [35S]GTPγS or fluorescent-tagged Europium-GTP, which bind to the receptor-activated form of the Gα subunit targeting the site of guanine nucleotide exchange (GDP for GTP on the Gα subunit of the Gαβγ heterotrimer). Guanine nucleotide exchange is a very early, generic event in the signal transduction process of GPCR activation which can be measured without the need for intact cells and is, therefore, an attractive event to monitor. The radiolabeled [35S]GTPγS or fluorescent Europium-GTP binding assays measure the accumulative level of G-protein activation following agonist activation of a GPCR by determining the binding of these non-hydrolyzable analogs of GTP to the Gα subunit. Therefore, they are defined as “functional” assays of GPCR activation because GTP-binding indicates that a cellular response will occur, not just that a binding event was detected at the receptor (which may or may not lead to a cellular response). This is important for screening applications to detect novel compounds which activate or block GPCRs. Ligand regulation of the binding of [35S]GTPγS is one of the most widely used assay methods to measure receptor activation of heterotrimeric G-proteins, as discussed elsewhere in detail [19,20]. The move toward a fluorescent based Europium-GTP assay partly overcomes some of the limitations of radioactive-based assays and has already been successfully used with the following GPCRs, motilin, neurotensin, M1-muscarinic and α2Aadrenergic receptors [21,22].
10
Leifert et al.
2.1.6. G-Protein Dissociation GPCR biosensors can involve the detection of the final stage of activation of the G-protein heterotrimeric complex, that being the putative dissociation or rearrangement of the subunits following GPCRinduced G-protein activation [23,24]. This level of GPCR activation has currently not been investigated in great detail and may prove to be extremely valuable in future functional biosensor applications as Gprotein signalling always involves alterations to the heterotrimer structure. An advantage of cell-free GPCR assays involving G-proteins is that GDP can be used to “reset” all GPCRs to the inactive state, thereby allowing for greater resolution of receptor activation by effectively removing background signalling. Bieri et al. [25] used carbohydrate-specific biotinylation chemistry to achieve appropriate orientation and functional immobilization of the solubilized bovine rhodopsin receptor with high contrast micropatterns of the receptor being used to spatially separate protein regions. This reconstituted GPCR:G-protein system provided relatively stable results (over hours) with the added advantage of obtaining repeated activation/deactivation cycles of the GPCR:G-protein system, as occurs in vivo. Measurements were made using SPR detection of G-protein dissociation from the receptor surface following the positioning of the biotinylated form of the rhodopsin receptor onto a self-assembled monolayer (SAM) containing streptavidin. Although SPR is useful for the study of G-protein interactions, it may not be well suited to detect binding of small ligand molecules directly due to its reliance on changes in mass concentration. An advantage of repeated activation/deactivation cycles of GPCRs is that different compounds may be tested serially with the same receptor preparation, allowing for discernment of differential activation of the same receptors by different ligands. The above approach appears promising for future applications of chip-based technologies in the area of GPCR biosensor applications. The well known and highly utilized interaction between Ni2+ and histidine residues (most often used for purification oligohistidixnetagged proteins) may be a useful means of attachment for GPCRs and/or
Nanoscale Biosensors and Biochips
11
G-proteins. To this end, different surface chemistries are being investigated to optimize the affinity interaction [26,27]. Modifying the surface of epoxy-activated dextran beads by forming a Ni2+-NTA conjugate was shown to produce beads with a surface capable of binding hexahistidine (his)-tagged β1γ2 subunits [28]. Tethered β 1γ2 subunits were then used to capture Gαs subunits which in turn were capable of binding membrane preparations containing a β2-adrenergic receptor-GFP fusion protein. Alternatively, a fluorescent labeled ligand binding to the tethered β2-adrenergic receptor could be detected; the whole complex being measured using flow cytometry. Flow cytometry’s greatest advantage is its ability to be multiplexed, where different molecular assemblies can be made in one sample and then be discriminated by their unique spectral characteristics [10,28-30]. Indeed, particle-based nanotechnologies, e.g. quantum-dots [5,31,32], constitute another emerging enabling technology for GPCR biosensor applications. 2.1.7. GPCRs as Biological Detectors of Volatiles Many organisms, from nematodes to mammals, use GPCRs to sense volatile compounds and the neural signal transduction due to GPCRvolatile interactions is the basis of smell [33-36]. Vertebrates are known to utilize olfactory receptors (ORs) that are similar to other metabolic GPCRs [35]. Invertebrate ORs were discovered relatively recently and it appears that these ORs are atypical GPCRs in that they have reversed membrane topology and that they apparently each form dimers with the same highly conserved OR-like receptor which can function as an ion channel, independent of ligand-mediated activation of the OR with which it has dimerized [37-39]. Due to the inherent OR-volatile specificity and high affinity of the OR-volatile interaction, ORs are obvious candidates as biomolecules that could be adapted to detect specific volatiles. Biosensors of volatiles would have a multitude of potential applications, particularly for sensing hidden entities (e.g. explosive screening) and in the agrifood industry (e.g. quality control, fermentation monitoring). The current attempts at volatile biosensing are not only limited by generic issues such as instability of membrane proteins but also by the poor level
12
Leifert et al.
of functional characterization of the different ORs [40]. Therefore, current attempts to produce a sensor for volatile compounds are limited to the most highly characterized receptors, these being human OR17-40 [41], I7 from rat [42], and Odr10 from the nematode Caenorhabditis elegans [43,44]. Hou et al. [45] reported an attempt to produce a biosensor that utilized the membrane fraction of recombinant yeast expressing rat I7 receptor, which is known to be activated by heptanal and octanal. This design used a gold electrode that was functionalized with biotinylated I7specific antibody, to which the membrane fractions were applied. Specific odorants were applied to the I7-presenting surface and interactions were monitored through variation in polarization resistance. Ligand binding was discernable although a specific response was difficult to resolve at ligand concentrations below 10-12 M, which is nevertheless more sensitive than current detection technologies. A common difficulty of presenting such ligands to the biosensor surface is the need for an organic solvent to carry the ligand. In the case discussed, dimethyl sulphoxide was used for this purpose and was found to only alter polarization resistance by 10%, even at the highest concentration tested (0.1 mM). Whole yeast cells have also been utilized to produce a I7-based biosensor [42]. Yeast cells were engineered to express I7 and a mammalian G-protein capable of linking ligand-mediated receptor stimulation to activate a MAP kinase pathway that induced synthesis of luciferase. Thus, in the presence of the luciferin substrate, ligand binding was detected as a dose-dependent fluorescent response. Importantly, sensitivity of the response was altered by the type of G-protein used to couple the OR to downstream elements. Another design utilized the human OR1740 protein expressed in yeast and receptor-containing nanosomes were produced by sonication of recombinant yeast membranes [46]. The nanosomes were then captured on a SAM functionalized with biotinyl groups that were used for attachment of neutravidin and then a biotinylated monoclonal antibody specific to the receptor. Interestingly, the myc-tagged receptors were functional when immobilized via a C-terminal tag but not when attached by the N-terminus.
Nanoscale Biosensors and Biochips
13
A crude Odr10-based olfactory biosensor was produced by expressing the OR in bacteria, obtaining a membrane fraction and coating a quartz crystal microbalance (QCM) chip with the membranes [47]. The authors of this study reported that ligand application could be detected as a change in QCM frequency. Another attempt utilized mammalian (HEK-293) cells to express Odr10, and SPR was then used to detect binding of the applied ligand (diacetyl) [48]. 2.1.8. The Future of GPCR Biosensors With growing interest and commercial investment in GPCRs in areas such as drug targets, orphan receptors, high throughput screening of drugs etc., greater attention will focus on biosensor development to allow for miniaturization, ultra-high throughput, and, eventually, microarray/biochip assay formats that will require nanotechnology-based approaches. The production of stable, robust, cell-free signaling assemblie’s comprising receptor and appropriate molecular switching components will form the basis of future GPCR/G-protein platforms which should be adaptable for laboratory- and field-based applications as microarrays and biosensors. 2.2. Pore-Forming Proteins Ion-channels are transmembrane proteins that regulate the transport of ions and/or small molecules across the lipid membrane. They can exist in the open or closed state, which can be regulated by a range of stimuli, including a change in membrane potential, mechanical stress or the binding of a ligand [49]. Some of the most commonly investigated poreforming peptides, are bacterial porins (e.g. OmpF) [50], α-hemolysin from the human pathogen Staphylococcus [51] and Gramicidin, a polypeptide antibiotic [52]. These have all been studied for their adaptability to a sensor platform [53-57]. Lipid supports, which are essential for the functional capture of the channels, are discussed in section 3.
14
Leifert et al.
a) (i)
(ii)
b) (i)
(ii)
Figure 2. Pore-based Sensing. (a) Schematic of an ion-channel switch (sandwich) developed by Cornell et al. [59] (i) The lipid bilayer is composed of archaebacterial membrane spanning lipids (MSL) and half membrane spanning tethered lipids (DLP). MAAD are spacer molecules attached to the gold surface via a sulfur-gold bond. The mobile lipids (DPEPC/GDPE) and ion channels (Gα) attached to antibodies (Fab) via a streptavidin linker (SA), can move throughout the bilayer, unlike the immobilized ion channels (GT). (ii) Mobile channels (Gα) become cross-linked to tethered antibodies (Fab) on the membrane spanning lipids (MSLα) in the presence of analyte (A), preventing formation of complete channels and therefore decreasing measured current. Reproduced with permission [59]. (b) Braha et al. [57] demonstrated simultaneous stochastic sensing for zinc, cobalt and cadmium ions with an engineered protein pore. (i) Schematic representation of the pore with the single metal binding site (metals are represented by different sized, or filled, balls) in the lumen of the channel. Each time metal ions bind to the pore, the current is modulated, as illustrated in the trace which reflects the currents flowing through the pores that were recorded during the application of a +40 mV membrane potential. Arrows indicate the current through the fully opened pore (ii). Reproduced with permission [57].
Nanoscale Biosensors and Biochips
15
Native or engineered membrane-bound ion-channels are a promising class of receptors for biosensing applications as they allow the sensitive detection of analytes and produce an output (electrical current) which is inherently suitable for digitization [58]. Ion-channel switches, first reported by Cornell et al. [59] (Figure 2a) and single-channel stochastic sensors [58] (Figure 2b), both demonstrate mechanisms by which analyte detection and quantitation can be determined by measured changes in current due to ion-channel activity. An important thing to consider when utilizing ion-channels is that the current, produced by the movement of ions through a pore in a lipid membrane, is dependent on the accessibility of a clear passage through the pore and therefore, can fluctuate when ion-channels do not traverse the membrane or when the pores are partially or fully blocked. 2.2.1. Ion-Channel Switch The ion-channel switch described by Cornell et al. [59] comprises a gold electrode, to which a lipid membrane carrying gramicidin ionchannels bound to antibodies, are tethered. A current is produced (turned on) when ions flow through the channel (in the presence of an applied potential). It is subsequently switched off when mobile channels diffusing in the outer half of the membrane, become cross-linked to antibodies immobilized at the membrane surface [59] (Figure 2). The ion-channel switch has since demonstrated specific signals derived from interactions with a range of analytes including bacteria, DNA, proteins and drugs [60]. Recently, Oh et al. [61] reported the detection of influenza A virus in clinical samples using the ion-channel switch biosensor. 2.2.2. Stochastic Sensing Stochastic sensing involves monitoring current that flows through a single pore and the alterations of this flow in response to analyte binding events (Figure 2). Each time an analyte binds to a binding site within the pore, the current flow is altered and during the on/off equilibrium of the binding analyte, a characteristic flow pattern is produced and monitored.
16
Leifert et al.
The frequency of the current fluctuations is related to concentration of the analyte, whereas the current signature (revealed by fluctuation duration and amplitude) is related to the analyte’s identity (reviewed by Bayley and Cremer [58] as well as Schmidt [62]). Some reported desirable attributes of stochastic sensing include fast detection, as well as the sensitivity and reversibility of sensor elements [57]. Current fluctuations monitored through a single pore, have demonstrated the capability for a variety of analytes, including metals [57], organic molecules [63] and oligonucleotides [64]. Shim and Gu [65] and Kang et al. [66] recently demonstrated increased stability of a single pore chip by encapsulation of the associated lipid bilayer within an agarose gel. This resulted in a more robust chip that can be stored for longer periods than bilayers alone. Pore-based transduction systems such as these have great potential in the field of biosensing. 2.3. Cell- and Viral-Based Sensing The biological component of a biosensor is currently most often an enzyme, antibody or other sub cellular component (such as the receptors and ion-channels mentioned in previous sections). The purification of these proteins can be labor intensive, expensive and the resulting product incompletely purified or unstable. Whole-cell sensors preserve the localization and temporal control of protein function and can utilize reporting processes that may involve multiple enzymes and signaling cascades. These types of reporting system are advantageous when biological/metabolic relevance of an analyte is important rather than simply its detection. Thus, unlike purified enzymes and antibodies, cell biosensors can report on bioavailability, metabolic regulation, toxicity, genotoxicity (DNA damage) etc. The key challenge for cell-based biosensors is the maintenance of cell viability under assay conditions. This requirement may lessen the utility of cell-based systems in field applications or environments that are deleterious to cell viability. Cellular responses can be specific to a substance (e.g. GPCR-ligand interactions) or a general response to adverse environmental conditions (e.g. regulated apoptosis) and each of these could be monitored depending on the specific biosensor design. Often cells are genetically
Nanoscale Biosensors and Biochips
17
modified with a reporter protein such as luciferase, GFP or βgalactosidase. The genes for these reporters are placed under the control of a promoter that responds to the analyte of interest resulting in the expression of the reporter protein, which can be detected. This approach has been exploited in identifying many environmental pollutants that induce particular promoters and for measuring stress responses. With advances in nanotechnology, cell-based nanobiosensors are now emerging with increasing sophistication, sensitivity and detection methods that are available, although most are still in the proof-of-concept stage. Different cell types can be more applicable to certain applications and/or measurements. In the following sections we discuss nanobiosensors utilizing bacterial, yeast, fungal, algal and mammalian cells. Several examples of cellular GPCR-based (olfactory) biosensor designs are discussed earlier (see 2.1.7) 2.3.1. Bacterial Biosensors Bacteria are particularly exploited in biosensors since they contain many defence mechanisms against analytes of interest such as environmental pollutants, including mercury and arsenic [67]. Bacteria can be easily produced at low cost and genetically manipulated and as mentioned previously. Reporter genes are often fused to DNA elements that respond to the presence of these analytes to produce a signal [67]. For example, bacteria can be used to detect genotoxic agents since increased DNA damage (due to the presence of a genotoxin) results in degradation of the endogenous repressor of SOS genes and subsequent expression of genes associated with DNA repair [68]. Other stress responses such as oxidation, nutrient starvation, membrane damage, heat shock and apoptotic responses can also be measured [67,68]. In a “lab-on-a-chip” format, E. coli were genetically engineered such that the activation of the fabA, dnaK or grpE promoters produced the enzyme β-galactosidase [69]. These bacteria were applied to electrochemical cells (100 nL capacities) on a silicon chip as a broth or immobilized within agar. The cells contained embedded electrodes for electrochemical measurements. Upon exposure to the representative toxicant phenol, and in the presence of the substrate
18
Leifert et al.
p-aminophenyl-β-D-galactopyranoside, β-galactosidase was expressed because of activation of the relevant promoters. β-galactosidase cleaved p-aminophenyl-β-D-galactopyranoside into p-aminophenol and β-Dgalactopyranoside. p-aminophenol is electrochemically active and the application of a 220 mV potential produces oxidation of p-aminophenol molecules, which is converted to a current that can be monitored to detect phenol concentrations as low as 1.6 ppm. The fabA promoter gave the largest response to phenol exposure which corresponded to the increased sensitivity of this promoter towards membrane damage inflicted by phenol. It is hoped that advantages such as the small sampling requirement, high signal to noise ratio, potential for highthroughput and high degree of robustness will combine to make this platform suitable for field applications. Carbon nanotubes are also being exploited in bacterial nanobiosensors. Pseudomonas putida have been coated onto an osmium redox polymer on a carbon nanotube-modified electrode and covered by a dialysis membrane to form an amperometric biosensor with increased electron transfer efficiency [70]. The respiratory activity of the cells was correlated with the oxidation of glucose, measured via the osmium redox polymer that acted as an electron acceptor. The system was then modified to measure phenol, which could be detected in an artificial wastewater sample, using phenol adapted P. putida. 2.3.2. Fungal and Algae Cell Biosensors The use of fungal cells (such as yeasts) in biosensors can provide the advantages of using bacteria but being eukaryotic cells they may provide information that is more relevant to higher eukaryotic organisms. This is a particularly important attribute for toxicity and drug screening. Fungal cells remain relatively easy to culture and genetically manipulate, and can be more robust with regard to pH, ionic strength and temperature than mammalian cells [71] due to their resistant cell walls. Wild-type cells can be used as biological oxygen demand sensors or to detect catabolic substrates. Oxygen consumption can be correlated to many physiologically relevant processes such as cell viability, protein synthesis and mitochondria function. Saccharomyces cerevisiae cells have been
Nanoscale Biosensors and Biochips
19
immobilized onto amine functionalized polystyrene nanobeads that were loaded with the oxygen-sensitive fluorescent ruthenium (II) [72]. This allowed optical detection of oxygen consumption since in the presence of molecular oxygen, the fluorescence of ruthenium (II) was quenched. In the presence of high concentrations of glucose, there was increased fluorescence indicating that cellular respiration and oxygen consumption was increased compared to when glucose was absent. Genetically modified cells can report on gene regulation in response to environmental factors or can be engineered to express and monitor the activation of other receptors such as those involved with olfaction or disease processes. An olfactory receptor (a GPCR) that responds to 2,4dinitrotoluene (DNT), a mimic for the explosive trinitrotoluene (TNT), has been identified and yeast were engineered to express the receptor and its associated signaling components including the G-proteins, adenylyl cyclase and a cAMP responsive DNA element that promoted the expression of GFP upon stimulation of the receptor [73] (for more discussion of the application of explosives detection see section 7.3). S. cerevisiae has also been engineered to express the human olfactory receptor, OR17-40. These cells could be immobilized on interdigitated gold microelectrodes coated with poly-L-lysine and the conductance on the surface of the electrode was shown to be modified by receptor-ligand interactions [74]. Algae could potentially be exploited for environmental biosensing since the inhibition of algal photosynthesis can be correlated to toxic effects of pollutants such as herbicides. (for further discussion on environmental monitoring with biosensors, see section 7.4). Inhibition of photosynthesis by photosystem II (PSII)-inhibiting herbicides (e.g. atrazine) can be measured as a change in chlorophyll fluorescence, caused by these compounds blocking the PSII quinone-binding site thereby inhibiting photosynthetic electron flow. This approach was demonstrated by using Chlorell vulgaris entrapped on quartz microfiber filters, and using a fibre-optic bundle to monitor chlorophyll fluorescence, which increased in the presence of atrazine [75]. Changes in chlorophyll fluorescence in response to exposure to formaldehyde and methanol vapor have also been monitored in a biochip platform that
20
Leifert et al.
could be used to simultaneously test many different algal strains with different sensitivities to toxicants [76]. 2.3.3. Mammalian Cell Biosensors Mammalian cell biosensors are finding particular usefulness within the pharmaceutical industry since using whole cells can maximize the information content of the assay and allow examination of a compound’s action on cells close to the intended target, within the context of all cellular machinery. This is particularly useful in the pharmaceutical industry that requires technologies that can provide reliable predictive information on lead compounds early in their development to reduce unwarranted development costs [77]. Impedance-based technologies (reviewed by McGuiness in 2007 [78]) can be used to detect cell death or proliferation, as well as smaller changes caused by receptor signaling and resulting in cytoskeletal rearrangements or changes in cell-cell interactions and adherence. Sensor chip-based impedance spectroscopy has been applied to measure the activation of GPCRs binding with neuropeptide Y (these receptors are implicated in human breast carcinoma [79]). Adenylyl cyclase activity in MCF-7 mamma carcinoma cells adhered to a microelectrode array, was stimulated with forskolin resulting in reduced impedance [79]. This effect could be blocked by pre-treatment with neuropeptide Y, which is known to inhibit adenylyl cyclase activity. Ligand-binding to various GPCRs in adherent cells has also been characterized using mass redistribution cell assay technologies (MRCAT) and resonant waveguide grating (RWG) [80]. Another example of a mammalian cell biosensor used malignant cells taken from a specific patient and exposed to chemotherapeutics to predict if that patient’s response will be favorable. Live metastatic human mammary cancer cells have been adhered to the gold surface of a QCM to sense for disruption of microtubules within the tumor cells in response to anti-tumor agents taxol and nocodazole [81]. To detect the excretion of interleukin-2 from mouse T-cells, silica nanoparticles were used to form a nanoparticle layer between two layers of gold, and antibodies specific to interleukin-2 were immobilized onto the sensor
Nanoscale Biosensors and Biochips
21
surface. Concanavalin-A was used to trigger secretion of interleukin-2 from the cells which was detected by the antibodies with a limit of 10 pg/ml using localized SPR [82]. The authors suggest that this technique has potential to be applied to high-throughput cell analysis systems for reporting on various cell activities and functions. These examples do not require cell engineering and are label-free, making them less invasive and therefore possibly more physiologically relevant. Other label-free technologies for whole cells have recently been more thoroughly reviewed [77]. However, most mammalian cell biosensors do rely on engineering cells to produce signals such as fluorescence, or over expression of a target that can produce an observable change in cellular physiology. These genetically encodable fluorescent biosensors have recently been reviewed [83] and will not be covered in detail here. Nanotechnologies are also being developed to sense changes within whole cells. Plasmonic biosensors or gold nanoparticles (20 nm) have been functionalized with an anti-actin antibody and a TAT-HA2 peptide, which mediated the endocytotic uptake of the nanoparticles into the cell, and their subsequent release from endosomes into the cytoplasm [84]. Binding of the nanosensors to actin could be measured since this brought the probes into such close proximity that the plasmon resonance became red-shifted, which could be detected by darkfield reflectance imaging or confocal microscopy, with detection limited to between 623-643 nm. It is hoped that further advances will enable monitoring of cytoskeletal rearrangements made as a biological response. Other optical sensing components can be entrapped within inert nanosized polymer particles termed PEBBLEs (probes encapsulated by biologically localized embedding) [85]. Possible probes include calcium sensitive dyes, pH sensitive dyes or enzymes such as horseradish peroxidase which can be used to detect reactive oxygen species [86] leading to applications such as analysis of effects of drugs, toxins or environment, on cell physiology. The inert polymer is permeable allowing the encapsulated probe to interact with analytes, and protects both the dye from interference by biological conditions, and the cell from any dye-associated toxicity. PEBBLEs are introduced into cells through surface modification with peptides that mediate cellular uptake, transfection using lipid reagents, picoinjection or gene gun
22
Leifert et al.
bombardment. PEBBLEs containing the BME-44 ionophore, ETH 5350 chromophore, and ionic exchanger KTFPB, have been used to measure increases in potassium in rat C6 Glioma cells treated with kainic acid to open potassium channels [87]. Calcium-sensing PEBBLEs have also been use in SY5Y neuroblastoma cells to report on calcium released from mitochondria in response to exposure to the neurotoxin m-dinitrobenzene [88]. Currently, cell-based approaches usually utilize a large population of cells within which different responses are occurring, and the responses are averaged. However, the response of individual cells can be different in an environment that is free of influences from neighboring cells. Optical nanosensors are being developed to allow intracellular measurement of biological processes within single live cells. Tapered optical fibres with nanosized tips (30-50 nm) have been applied to a wide range of applications (reviewed in Leung et al. [89]) and can be used to probe conditions inside a cell. The tip of the nanofibre is approximately 10-fold smaller than the wavelength of excitation light transmitted along the fibre. Photons travel as far along the fibre as possible but cannot escape from the tip although an evanescent field continues to travel a short distance through the remainder of tip providing excitation light for molecules that are in close proximity (<100 nm) to the nanoprobe [90]. Optical fibres have been derivatized with a substrate of caspase-9, a biomarker of apoptosis that once cleaved, yields a fluorescent product. Manipulation of the fibre into MCF-7 cells exposed to an inducer of apoptosis (ALA-PDT) allowed the monitoring of caspase-9 activity detected by illumination of the fibre optic probe. The level of activity was indicative of the amount of fluorescent product formed in the presence of active caspase-9 [90]. 2.3.4. Cell Immobilization and Arrays Cells can also be immobilized to form arrays for advanced chipbased applications in medical diagnostics, or the detection of environmental pollutants in the field. Cell immobilization can facilitate the display of a variety of targets or enable measurement of physiological responses. Amphibian tumor cells (FT cells) have been cultured on a
Nanoscale Biosensors and Biochips
23
Cr/Au microelectrode array chip [91]. Oxidation of norepinephrine (released by the cells upon stimulation with ATP via an endogenous receptor) and an increase in intracellular calcium were measured. In this example ATP was detected using an endogenous receptor, an approach that may produce results with increased physiological relevance. However, cells could also be engineered to express recombinant receptors. Recombinant-cell microarrays have also been produced using suitable surfaces such as glass, silicon or tissue culture polystyrene that can undergo surface modifications (reviewed by Hook et al. [92]). Nucleic acid microarrays have also been produced and generally utilize DNA for expression of a desired protein or interfering RNA (RNAi). The relevant nucleic acid is firstly spotted and DNA-gelatin mixtures are often used to ensure spatial confinement of the nucleic acid molecules. Cells are then applied and allowed to attach to the surface. The nucleic acid detaches (or desorbs) from the surface by reversal of the hydrophobic (or electrostatic) interactions between the nucleic acid and the surface. A transfection reagent generally facilitates the nucleic acid being taken up by cells where it is then expressed or used to silence genes within the cell. Often the coding sequence of a reporter protein such as GFP is fused to the nucleic acid sequence of interest, allowing determination of transfection and/or expression levels. Transfected-cell microarrays have been applied to express GPCRs in a format where greater than 3000 receptor:ligand interactions could be measured in a single 96-well plate. The cells were loaded with a Fluo-4 calcium indicator dye and agonist binding was detected as an increase in intracellular calcium measured using fluorescence microscopy [93]. This functional screen could be applied to the “deorphanization” of GPCRs with no known ligand or for drug screening for characterized GPCRs. In recent times, several polymers have been investigated as substrates upon which cells can be cultured, with the aim of culturing cells directly on precoated surfaces which could then be used as biomolecular detectors. Lakard et al. [94] tested three polymer substrates for their ability to allow rat neuronal cells to adhere and proliferate, namely polyethyleneimine (PEI), polypropyleneimine (PPI) and polypyrrole. Data indicated that PEI and PPI were the best candidates
24
Leifert et al.
and did not affect cell morphology. In addition these polymers had advantages such as strongly binding to electrode surfaces and were insoluble in most organic solvents. Subsequently, the same researchers were able to culture neuronal cells expressing ORs in defined positions on a silicon wafer with the aim of producing an olfactory biosensor [95]. Single cells can also be trapped within minute depressions on a CDlike chip containing microfluidic channels along with passive valves and chambers allowing sample loading and waste storage [96]. Centrifugal force was used to load HEK293 or Jurkat cells into the depressions and assays performed using paraformaldehyde or UV irradiation to detect cytotoxicity or apoptosis, respectively. Compared to testing a larger population of cells, this single cell assay showed a higher rate of survival when cytotoxicity was measured presumably due to the lack of proteases and other toxicants released from neighboring lysed cells. 2.3.5. Virus-Containing Biosensors While virus particles (virions) are not cellular, they are generally composed of a relatively complex mixture of peptides, exceeding that of the receptor based system, and may often be membrane-enveloped. Therefore we discuss virion-based biosensors at this point in the review, after the more complex cellular biosensors. Being proteinacious and genetically encoded, virions can be simply produced and can display peptides in a biologically functional form, either through genetic engineering or chemical conjugation of peptides or small molecules. In addition, virions can also be conjugated to fluorescent moieties such as a quantum dots or fluorophores for optical monitoring [97]. An important goal in the development of sensitive imaging sensors is the ability to specifically target cells and tissues of interest to allow sensitive imaaging or delivery of therapeutics. The abnormal characteristics of tumor cells produce cell-surface or extracellular matrix proteins that can be used as markers to distinguish tumor cells from normal tissue. Often the higher metabolic activity of tumors gives rise to over expression of a number of receptors such as the folic acid or transferrin receptors [97,98]. Virions have been utilized from Cowpea mosaic virus (CPMV), bacteriophages and other viruses that are not typically human pathogens and are
Nanoscale Biosensors and Biochips
25
therefore unlikely to cause infection in humans, reducing costs associated with minimization of biohazards (e.g. quarantine laboratories and protective equipment) [97]. High yields of CPMV can be obtained from infected plants and virions are simple and inexpensive to purify. The virions are relatively thermally stable and can tolerate a wide pH range and a variety of solvents, making them useful candidates for field applications. By modifying virions for surface display of folate and blocking the remainder of the virion surface with PEG, the viruses could be targeted to tumors that over express folic acid receptors [98]. Philamentous bacteriophages also tolerate a wide range of conditions particularly high salt concentrations, low pH, presence of chaotropic agents and prolonged storage. Similarly to eukaryotic viruses, bacteriophages can be engineered for surface display of a desired peptide that could detect a specific cellular target or specifically modify the behaviour of a target cell, in a measurable way. Networks of phages and gold nanoparticles (44 nm in diameter) were observed to form spontaneously and phage particles displaying the peptide CDCRGDCFC, have been use to recognize αv integrins present in high levels on the surface of melanoma cells, where binding resulted in receptor-mediated phage internalization [99]. This was detected using dark field microscopy utilizing the large degree of scattering resulting from gold nanoparticles incorporated into cells (after washing), which are ideal contrast agents. Additionally, surface enhanced Raman scattering (SERS) spectra also correlated with the level of cell binding and internalization and could therefore also be used to monitor these processes. 3. Lipid Supports for Biosensor and Biochip Fabrication 3.1. Why Functionalize Biosensors with Lipid Membranes? Lipid membranes are versatile and convenient alternatives to study the properties of natural cell membranes (see sections 2.1 and 2.2). Due to a combination of factors such as ease of formation, control over complexity, stability and the applicability of a large range of analytical techniques, artificial lipid membranes now have a central role in
26
Leifert et al.
membrane research. Research on membrane sensor platforms has, in particular, been stimulated by the possibility of studying membrane proteins in a near-native environment, but it is also emerging as a nanoscale surface functionalization platform in its own right for controlled bioresponse. With more than 50% of all drug targets being membrane proteins, which require a lipid membrane to remain functional [100], target applications for commercial biosensors are dominated by the quest for high-throughput membrane protein drug screening assays and more predictive in vitro admetox platforms. Additionally, the development of artificial tongues and noses built on biological principles is also being sought. Traditionally, lipid membranes have been regarded as having two important biological functions: (i) acting as an electrochemical barrier between cells and the environment and between different cellular compartments; and (ii) as scaffolds for membrane proteins. In recent years, it has been increasingly realized that the dynamically rearranging lipid membrane, containing a multitude of different molecules, could on its own perform important messenger and switching functions [101,102]. Its role in controlling protein and cell function might be much greater than previously thought [102]. This has led to an interest in more sophisticated membrane sensor configurations which offer the possibility of obtaining richer information about the biophysical state of the membrane, in real time. Desired features include ability to measure/characterize changes in density, thickness and ordering of the lipids, sub-micron domain chemical composition, membrane asymmetry, surface charge, mechanical properties and morphology. Obtaining this kind of information with the necessary level of detail requires sensing at the nanolevel and in many of the approaches now being developed, nanostructured sensors play a key role. This section primarily describes lipid membrane architectures, which can be used to study membrane properties at, or in close proximity to, a solid interface. Interest in such systems is high due to their stability and large range of highly sensitive analytical label-free methods available to characterize the membranes and their associated interactions or alterations. In particular, the integration of membranes with electrochemical and optical readout schemes allows for simultaneous
Nanoscale Biosensors and Biochips
27
measurements of binding, conformational changes and functional response of membrane and incorporated membrane proteins. Several different surface-based membrane platforms have been developed over the years. They can be divided into hybrid lipid bilayers (Figure 3a), solid-supported lipid bilayers (SLB, Figure 3b), tethered lipid bilayers (tSLB, Figure 3c), polymer-cushioned lipid bilayers (pSLB, Figure 3d), pore-spanning lipid membranes (nano- or microBLM, Figure 3e) and tethered liposomes (Figure 3f).
Figure 3. Classification of different membrane functionalized sensor platforms. (a) Hybrid lipid bilayer with a lipid monolayer formed onto an alkane self-assembled monolayer (SAM) functionalized sensor substrate; (b) supported lipid bilayer (SLB) selfassembled on hydrophilic support; (c) tethered lipid bilayer (tSLB) self-assembled on covalently attached hydrophobic molecules, often derived from lipids, with a hydrophilic spacer layer attached to the substrate; (d) polymer-supported lipid bilayer (pSLB) assembled either (i) directly on a polymer cushion with adjusted wetting behavior or (ii) using hydrophobic molecules as anchors that are incorporated into the polymer matrix; (e) pore-spanning lipid membrane assembled across a nano- or micro-sized aperture in a support.; (f) tethered liposomes.
3.2. Methods to Assemble Supported Lipid Membranes With the exception of tethered liposomes (hollow phospholipid bilayer vesicles self-assembled in water from amphiphilic molecules [103]), which only require adequate substrate attachment, the formation of a lipid membrane on a sensor surface requires detailed control of the
28
Leifert et al.
interaction of the physisorbed lipids (and incorporated proteins) with the substrate. Several methods have been developed to self-assemble sensorsupported lipid membranes, and are schematically represented in Figure 4: self-assembly from vesicles in solution (Figure 4a) [104,105], Langmuir-Blodgett type deposition (Figure 4b) [106] and self-assembly from lipid dispersion directly on the surface by painting and detergent dialysis (Figure 4c) [107]. Additional methods such as stamping of a membrane (including native cell membranes) onto a surface have also been developed but do not yield high and homogenous coverage beyond that which allows qualitative binding studies [108].
Figure 4. Methods used to assemble lipid membranes on substrates. (a) Self-assembly from vesicles on (i) hydrophilic substrates, (ii) hydrophobic substrates and (iii) hydrophobic tethers with hydrophilic spacers (tSLB); (b) assembly from Langmuir films to (i) first monolayer deposited on hydrophilic substrate, (ii) bilayer completed by Langmuir-Blodgett deposition, or (iii) bilayer completed by Langmuir-Schäfer deposition; and (ci) detergent dialysis, (cii) painting and solvent extraction.
The original and possibly still most used methods are based on spreading the membrane from a solvent onto a surface (or across an aperture in a support). However, these methods are becoming increasingly replaced by those relying on assembly from pre-formed small (20-200 nm in diameter) unilamellar liposomes, which are fused into an SLB (for example) [109,110]. Also, Langmuir-Blodgett deposition techniques are less used, because they require more complex setups and controlled environments and thus typically have lower
Nanoscale Biosensors and Biochips
29
reproducibility despite the advantage of control over membrane tension during deposition. The main reason for the emergence of self-assembly from vesicles (compared to solvent spreading methods) is that it allows for the formed membrane to remain free of solvents, which render many transmembrane proteins non-functional. Additionally, liposome composition and other properties can be easily controlled, including by the use of vesicles harvested directly from cells, and thus provides a simpler and more versatile way of controlling the composition and lateral distribution of lipid and protein material on the sensor surface. 3.3. Supported Lipid Membrane Platforms The literature on sensing using supported lipid membranes is vast and only the wider context and a few selected examples are discussed here. For an in-depth review of the many suggested membrane sensor platforms, we refer the reader to a recent review by Janshoff and Steinem [111] or to topical reviews on specific configurations [112,113]. The first developed and most widespread in vitro method for studying ion-channel function is the bilayer lipid membrane (BLM), spanning across apertures between two aqueous compartments [114]. While a successful approach under laboratory conditions, this platform has a few major drawbacks for general application: (i) instability manifested in the collapse of the membrane within a few hours of preparation; (ii) preparation requiring solvents which stay in the membrane, making it incompatible with many proteins which alter conformation and function in the presence of solvent; and (iii) limited to only electrochemical and fluorescence based sensing techniques. As a result, there has been an increasing emphasis on solid-supported membrane platforms in recent years. Solid-supported platforms offer inherent stability thanks to the underlying support. In their original form, these solid-supported bilayers [115-118] (hybrid bilayer Figure 3a; SLB Figure 3b) had a severe drawback for studying membrane protein function in that they offer very little space between the lipid membrane and the solid support to accommodate hydrophilic domains of the integrated protein and ions transported across the membrane [111,119], although their assembly on electrodes [115,120-122] and use for characterization of inserted
30
Leifert et al.
membrane proteins [123-128] have been demonstrated. However, due to their simple geometry and proximity to the surface they offer an excellent model system for studying the self-assembly of lipid membranes and to probe membrane physicochemical properties by a wide range of techniques [105,109,111,113,129-131]. Further, decoupling of the membrane from the surface by forming a tethered supported lipid bilayer (Figure 4c), first demonstrated by Vogel and coworkers [132], has been achieved primarily by forming a supported lipid bilayer onto a hydrophobic anchor layer separated from the surface by a short hydrophilic polymer spacer [59,111,133-144], typically oligo(ethylene glycol) with a lipid anchor [111,132,134-137,145,146]. The additional space of a few nanometers allows for integration of small transmembrane proteins without undesired surface interaction, but the reservoir is still too restricted to monitor continuous ion transport and the preparation can be difficult for solvent-free membranes [145]. Knoll and coworkers have shown how to reproducibly achieve high-insulating tSLBs by creating membranes on ultra-flat substrates and improving stability by using tethers derived from phytanoyl thiolipids inspired by those found in members of the Archaea [145,146]. A competing platform using similar tethers based on cholesteryl anchors has been shown [136,147]. Tethering membranes directly to pre-immobilized transmembrane proteins, which has the advantage of yielding 100% orientation of proteins, has also been shown [133,148]. Recently, the fusion of native membrane fragments to form the upper leaflet of tSLBs has also been reported with high sealing resistance [149]. This method avoids the needs for reconstitution of sensitive proteins. Membranes can also be formed on top of hydrogels and other thicker (∼10 nm) polymer cushions (Figure 3d) [113,150-153]. However their application to biosensing is so far limited. Tanaka et al. [113,154,155] have demonstrated the formation of lipid membranes on hydrogels (mainly cellulose derivatives), but cushions such as PEI, PEG and pHresponsive polymers have also been used [153,156-159]. These membranes, also formed directly from erythrocyte ghosts, have been shown to retain the mobility, density, function and orientation of transmembrane proteins containing large hydrophilic domains, like celladhesion receptors [155,160]. Thick, fully decoupling, hydrophilic
Nanoscale Biosensors and Biochips
31
spacer layers at the sensor substrate can also be obtained by selfassembling the polymer with the SLB, without covalent binding to the substrate. Methods based on self-assembly from PEG-lipsomes and stabilization with trehalose to create robust membranes, have been demonstrated by Albertorio et al. [161,162]. Due to advances in sensor nanofabrication technology and in lipid self-assembly, free-spanning membranes are currently experiencing a renaissance as nano-BLMs (Figure 3e). Fabricating apertures in the submicron range, allows for greatly enhanced membrane temporal stability [112,163,164]. Furthermore, decreased size and advances in nanoscale control of the surface properties make it possible to self-assemble nanoBLMs from small unilamellar vesicles. Several publications report the development of platforms based on nanoporous solid substrates [152,165,166]. The best results have been achieved by forming the membrane after rendering the top surface of anodized alumina foils with dense sub-100 nm orifices made hydrophobic by an alkanethiol SAM. The system thus comprises spanning membranes sealed and stabilized to the solid support by forming a hybrid bilayer. Nano-BLMs are stable for tens of hours and activity of single ion-channels inserted into the membrane can be measured [164,166-168]. A recently proposed method to completely decouple the membrane from the substrate but still keep it in proximity of the surface is to tether intact liposomes to the sensor substrate through several hydrophilic linkers (Figure 3f) [169-172]. While early demonstrations utilized tight binding by for example lipid-biotin-avidin linking, the major linking strategy has become the use of complementary DNA-anchors that allow for tagging and addressing libraries of different vesicles [169,173,174]. Importantly, the size of tethers and vesicles is suitable to capture the membrane system within the evanescent sensing zone of optical and acoustic sensors [175,176]. When tethered to an SLB on the sensor, liposomes retain high lateral fluidity allowing liposome-liposome interactions (and therefore potentially those between inserted protein) to be studied [177,178] while immobilization onto micro- or nanopatterned surfaces creates stable arrays. Stamou and others demonstrated addressing of single liposomes per spot onto large areas with applications in, for example, affinity studies for GPCRs [170,179] (also see section
32
Leifert et al.
2.1.2), and recently similar addressing plus sorting of functionality was also shown [174,176]. The main drawback of the tethered vesicle system is that it does not allow for detailed measurements of membrane ordering or transport of material across the membrane. Micropatterning of planar lipid membranes on solid and polymer supports has also been demonstrated, which although not yet developed with the same ability to label and target a certain functionality to an array spot (as allowed by tethered vesicles), nevertheless shows promise as a potential strategy to selectively assemble planar membranes and confine them by materials contrast [180], microcontact printing of membrane or diffusion barriers [154,181-184], photopolymerization [185,186], polymer wettability contrast [154], “nanoshaving” by AFM [187,188], polymer lift-off [189], microspotting onto tethers [147] or microfluidics [189-191]. At least the two latter methods allowed for arraying of membrane function and subsequent array bioaffinity sensing. 3.4. Advanced Sensors Functionalized with Lipid Membranes The discussion of membrane platforms in earlier sections concentrated on the study of transmembrane proteins and charge translocation across the membrane. Despite the many advantages for functional biosensing conferred by electrochemical sensing methods, not least of which is the ability to characterize ion-channel and iontransporter function [192] (see also section 2.2), the bulk of work on lipid bilayers has been on characterizing the assembly and properties of assembled membranes using optical and acoustic sensing methods. Typical methods to characterize the formation of planar lipid membranes include fluorescence recovery after photobleaching (FRAP) [193], QCM with dissipation monitoring (QCM-D) [194] and electrochemical impedance spectroscopy (EIS) [140]. These methods facilitate probing for completeness (FRAP, EIS), lateral fluidity (FRAP) or formation kinetics and mass (QCM-D). Sub-micron miniaturization of the membrane combined with the realization of the importance of lipid organization (especially for complex biological mixtures) makes for example, optical and acoustic evanescent probing of lipid distribution and membrane conformation increasingly interesting. Recent data from
Nanoscale Biosensors and Biochips
33
self-assembly of supported lipid bilayers demonstrate that combining different biosensing techniques also provides the potential to perform detail characterization of morphological transitions from vesicular layers to planar bilayers [105] and of morphological changes in a supported lipid bilayer itself [130]. Furthermore, advances in waveguide spectroscopy have made it possible to probe transitions in supported membrane structures in terms of thickness, density and lipid alignment [129,131,195,196]. This information can be used for biosensor applications involving influence of ligand binding, drugs and ions on membrane properties (see GPCRs and ion channels; section 2). As mentioned earlier, this methodology could differentiate between different ligands binding to GPCRs reconstituted into waveguide-supported lipid bilayers by the specific induced conformational changes to the GPCR and surrounding membrane [197]. With the advent of nanooptical (in particular nanoplasmonic) sensors, it has been demonstrated that single liposomes and supported lipid bilayer islands can be confined to a single nanoplasmonic sensing element [198,199]. The high local sensitivity of such sensors combined with approaches such as the liposome arraying technology described earlier makes it possible to envision ultra-dense membrane affinity or more advanced array sensor platforms in the future. 3.5. Future Perspectives The increasing number of publications in recent years demonstrates the interest for lipid membrane based biosensors. A recent trend in the field is the utilization of nanoscale optical and electrochemical sensor architectures [168,198] as well as microfluidics [200,201]. Future developments will address the miniaturization of sensor elements and production of membrane arrays. For integration into commercial biosensors, especially for field applications, an important goal is to improve the robustness of membrane functionalized sensors by increasing their stability [66,147,202,203]. An additional area of further development is the creation of more sophisticated sensor integrated membranes which mimic a greater range of biological membranes and their properties. Such functionalization could greatly enhance our understanding of the molecular basis of membrane function in biology.
34
Leifert et al.
Combined with sensor platforms that probe not only the affinity of ligands but also the biophysical properties and organization within the membrane, direct probing of spatial and temporal dynamics of membrane signaling at the nanolevel could be another area of investigation [80]. 4. Nanopatterning for Biosensing and Biochip Fabrication A number of nanopatterning technologies are being developed in order to control the location, distribution, amount or conformation and orientation of biomolecules in the nanorange [204]. The sensitivity and selectivity of a biosensor relies on the specific interaction between biomolecules, hence uncontrolled, non-specific interactions have to be suppressed in order to avoid false responses [205-216]. Parallel nanofabrication approaches enable fast production of a large number of samples that can be applied over a large surface area. Alternatively, serial nanofabrication methods, although slower, usually offer better control over size and composition. In addition, the combination of topdown approaches with self-assembly (bottom-up) concepts is further increasing the flexibility to position biomolecules onto a pre-defined area. The most prominent nanofabrication methods will be addressed in this section. 4.1. Parallel Nanopatterning Methods The ability to produce high-quality, large scale nanopatterns quickly and cheaply is challenging. A large number of different nanopatterning methods are being considered for novel biosensing platforms including the next generation of photolithography (see section 4.1.1 e.g., Extreme Ultraviolet Interference Lithography [217]), soft lithography (see section 4.1.2 e.g., replica molding and microcontact printing) [218,219], nanoimprint lithography (see section 4.1.3) [220,221] and nanosphere lithography (NSL, see 4.1.4 e.g., colloid lithography [222] or colloidal block-copolymer micelle lithography [223]).
Nanoscale Biosensors and Biochips
35
4.1.1. Photolithography Photolithography is a well-established patterning method enabling fast and highly reproducible creation of micron- and sub-100 nm structures by illuminating a photosensitive polymer. In particular, while illumination through a mask is successful in the micron-range, maskless approaches (so called interferometric lithography) were found to be valuable for creation of periodic large-scale nanofeatures over a large surface area [217,224-227]. Subsequently, biological contrasts (i.e. active nanopatches) embedded in a non-interacting PEGylated background could subsequently be incorporated into such a pre-pattern [228]. Alternatively, photolithography can be used to create nanowire[229,230] or carbon nanotube- [231] field effect transistor biosensors (see section 5 for more detailed discussion of nanotubes). 4.1.2. Soft Lithography Microcontact printing is a popular parallel micron- and nanopatterning approach, which was first introduced in 1993 by Whitesides and coworkers who created alkanethiol patterns on gold [232]. This three-step patterning method consists of: (1) production of the re-usable master, (2) formation of the elastomeric stamp from the master and (3) the “inking” of the stamp and printing of its features onto a substrate [219,233]) which enables patterning of a large variety of biomolecules (proteins [234-238], DNA [239-241], supported lipid bilayers [182,183,242] or liposomes [171]) crucial for creating biosensing platforms in the micron- and nanorange. Recently, an approach to automate the microcontact printing process has been reported, making soft lithography even more competitive (inexpensive and higher throughput) when compared with photolithography [243]. Supramolecular nanostamping is an alternative printing approach introduced by Yu et al. [244], which enables high-resolution DNA nanopatterning [245-247].
36
Leifert et al.
4.1.3. Nanoimprint Lithography Nanoimprint lithography [248-250] and closely related techniques [251-253] facilitate imprinting of a rigid nanostructured mold into a thin layer of spin-coated resist. The excess resist is then removed in the deformed areas via reactive ion etching (pattern transfer). There are different ways that nanostructures created via imprint lithography can be used for biosensing platforms. They have been utilized in a “lab on a chip” device providing a nanoscaled channel network. The imprinted (polymer) contrast was used as an etching mask [254-256] or a step in a lift-off process [257,258] in order to provide access to a bio-active nanostructured surface for biosensing applications. Alternatively, PEGbased UV-curable polymers have been directly imprinted and further decorated with proteins [259-262]. 4.1.4. Nanosphere Lithography NSL is a bottom-up approach enabling cheap, fast and large-scale production of nanopatterns via colloidal self-assembly [222]. The concept was introduced in the early 1980s and is simple and straightforward and utilizes a particle monolayer as a mask for contact imaging [263], etching or material deposition [264-267]. In recent years, impressive advances have been made to facilitate defect-free particle monolayers and subsequently, high fidelity bio-active nanopatterns [268275]. NSL is not only interesting for the creation of high-density nanoarrays of biomolecules, but also for nanopatterned metal colloids that were used as read-out systems for biosensing applications based on localized surface plasmon resonance (LSPR) [276-278], as demonstrated by Van Duyne and coworkers [279-282] or Frederix et al. [283,284]. By monitoring changes in the UV/visual absorption band of the nanoparticles adsorption of chemical or biological species could be detected. For instance, amyloid-β derived diffusible ligands (ADDLs) could be detected in the cerebrospinal fluid of an Alzheimer’s disease patient using a nanoscale optical biosensor [279] (Figure 5). The “sandwich assay” was used as a biosensor where the signal is generated by a tagged reporter molecule binding to a biological detector that is
Nanoscale Biosensors and Biochips
37
Figure 5. Biomarker detection using a nanoscale optical biosensor. (a) Illustration of the biosensing design for the detection of amyloid-β-derived diffusible ligands (ADDLs) using a sandwich assay (schematic in the top right corner). The changes of the optical properties (LSPR) of silver nano-triangles created via NSL (atomic force microscopy image on the bottom right) upon the adsorption of the biomolecules were monitored. (b) A sandwich assay and a LSPR nanosensor (i) were used to analyze human cerebrospinal fluid (CSF) from an aging person (ii) and an Alzheimer’s disease patient (iii). The LSPR spectra of each adsorption step were monitored and the presence of ADDL was detected only in the sample of the Alzheimer’s disease patient (shift in the LSPR spectra in (iii). Reprinted with permission [279].
38
Leifert et al.
attached to a surface with linker molecules (often oligonucletoides or proteins). Alternatively, holes in a conductive gold film created via NSL were utilized in LSPR and were shown to be a sensitive detection option for several bio-recognition events [199,285]. 4.2. Serial Nanopatterning Methods Direct-writing techniques such as e-beam lithography [286], focused ion-beam lithography [287] or dip pen nanolithography (DPN) [288,289] can virtually produce any type of nanostructure with resolution between 5-50 nm. Although e-beam lithography and focused ion-beam lithography are able to create a large variety of nanopatterns [290-293], the cost and time to write them are disadvantageous when compared to the parallel nanopatterning methods described in section 4.1. These important disadvantages have seen the use of these two nanopatterning techniques limited. DPN, on the other hand, has become a very popular and versatile nanopatterning approach since its invention in 1999 [294]. DPN is a direct-writing, scanning probe based technique where an inked AFM tip is used to transfer biomolecules to a pre-defined location on a surface. Using this approach, nanoarrays of DNA [295,296], peptides [297,298], proteins [299,300], lipids [301], viruses [302-304] and bacteria [305] have been created. In addition, enzymes patterned via DPN have been used to perform localized reactions on a surface [306308]. In order to produce nanoarrays for parallel high-throughput screening, the inherent slowness of the process when using a single AFM tip is being overcome by either using an array of individually actuated tips [309-311] or of passive tips. While the former concept enables the creation of complex chemical patterns, the latter is much simpler and was recently used to create 450,000 sub-100 nm features in less then 30 minutes (Figures 6a and 6b) [312]. Figure 6c shows an example whereby massive parallel DPN was used to create a large-scale nanoarray of supported phospholipid bilayers. The recent advances in parallelization of the patterning approaches together with the fact that virtually any biomolecules can be arranged into any nanoshape, offers unique opportunities for designing multi-component biosensing platforms.
Nanoscale Biosensors and Biochips
39
Figure 6. Dip Pen Nanolithography: (a) Schematic illustration of massive parallel dip pen nanolithography using a 2D cantilever array. (b) A micrograph and a SEM (inset) image of such a cantilever array. Reprinted with permission [312]. (c) Fluorescent images of supported phospholipid bilayer patterns; (i) lower magnification image, (ii) close-up in the array, (iii) a two-component bilayer nanoarray using two different inks (phospholipids doped with two different fluorescent dyes) is shown. Scale bars are 5 µm. Reprinted with permission [301].
All of the parallel and serial nanopatterning methods discussed in section 4 have the potential (in combination with an appropriate surface chemistry) to be implemented as a biochip especially when integrated with other bio-analytical components into a small portable device, a “lab on a chip” [313-317]. Such nanoarrays are expected to significantly impact future biosensing applications. However, the reliable and largescale production of a heterogeneous nanoarray still needs to be achieved. Without this, a meaningful nanoarray based biosensing platform cannot be established since commercial drug discovery or diagnostic screening
40
Leifert et al.
applications require sensitive and selective parallel analysis of a large number of biomolecules and their interactions in a short time. 5. Sensing Substrates: A Closer Look at Nanotubes The term nanotube is now used to describe a variety of hollow structures made from a wide range of materials including carbon [318,319], boron nitride [320], titanium dioxide [321], silica [322] and even “soft” matter such as peptides [323]. Many of these structures have unique properties that have been exploited for biosensing [322-326]. This discussion will concentrate on carbon nanotubes as they are easily the most utilized structure for development of biosensors, and many carbon nanotube-based approaches have been applied to detect a range of analytes [319]. We focus on electrochemical approaches as this represents the most active (and best developed) area of biosensing with carbon nanotubes. The great promise of nanotubes as biosensing elements is the potential to develop systems where direct electron transfer between enzymes and electrodes is possible. This innovation is key to the development of mediatorless (third-generation) enzyme biosensors, where no co-substrate is required in the recycling of the enzyme back to its active form. The mediatorless enzyme biosensor using nanotubes is most obviously applicable to the oxidoreductase enzymes where redox reactions cause electron flow and the extremely high conductivity of the nanotubes is used to detect this flow. 5.1. Carbon Nanotube Electrodes for Communicating with Redox Proteins Carbon nanotubes consist of graphene sheets wrapped into a hollow cylinder with the ends capped or open [318,324]. In the case of multiwalled carbon nanotubes (MWNTs) the concentric graphite tubules are in the range of 2 to 25 nm in diameter with 0.34 nm between tubule sheets. With single-walled carbon nanotubes (SWNTs) a single graphene sheet is rolled seamlessly into individual cylinders (typically of 1-2 nm) with capped ends containing carbon atoms which are all sp2. SWNTs
Nanoscale Biosensors and Biochips
41
can be metallic conductors, semiconductors or small-band gap semiconductors depending on their diameter and chirality [318]. Closed nanotubes can be opened in oxidizing environments such as nitric acid [327]. Open-ended nanotubes have been shown to have excellent electron transfer properties [324] compared with closed nanotubes. The open ends of the carbon nanotubes typically contain carboxylate and quinone functionalities in common with edge-planes of pyrolytic graphite, allowing linking of functionalized nanotubes with edge-planes of pyrolytic graphite, while the nanotube walls have similar electron transfer properties to the basal planes of pyrolytic graphite. Electrodes have been made using either MWNTs or SWNTs. In many nanotube electrodes thus far presented in the literature, the electrode is prepared by forming a paste with a filler compound and packing this into an electrode body, or simply by dispersing the tubes in a solvent and drop-coating onto the electrode to leave a bed of nanotubes on the electrode surface [328,329]. The first example of achieving electron transfer to proteins using carbon nanotube-modified electrodes was by Davis et al. [330] where an electrode of MWNTs was first opened in nitric acid and then mixed with nujol, bromoform, mineral oil or water. Cytochrome c and azurin were subsequently adsorbed onto and/or within the tubes with retained activity. The electrodes were shown to have an excellent ability to probe the redox sites of these proteins which was superior to that provided by edge-plane pyrolytic graphite. Similar results have been obtained by others who probed redox proteins with their active sites close to the protein surface, such as cytochrome c [331,332] and horseradish peroxidase (HRP) [333,334]. A more recent example [328] of this approach initially coated SWNTs in the biocompatible polymer chitosan, which has the dual effect of making the nanotubes more dispersible in water in addition to removing the hydrophobic character of the outer nanotube surface allowing adsorption of a biological molecule, in this case glucose oxidase. Using this approach, the detection limit of glucose was 0.01 mM with a response time of 10-15 seconds. The electron transfer rate of this electrode type was also shown to be superior to similar MWNTs or nanoparticle electrodes highlighting, the importance of the reduced size
42
Leifert et al.
of the SWNTs which likely allows a closer approach to the redox active centre of the probed enzyme. This is advantageous as most redox active biological molecules have their redox centers embedded deep within the protein’s quaternary structure [335]. For example, in the case of glucose oxidase, the smallest distance between the protein exterior and its redox active center, Flavin adenine dinucleotide (FAD), is 13 Å [336]. Consequently, electrons cannot be efficiently transferred between the enzyme and the electrode and hence, mediators or redox relays are required. The use of nanotubes overcomes this obstacle by essentially “plugging into” the enzyme and getting close to the active centre which facilitates efficient electron transfer. Boron-doped nanotubes [337] have been used to detect glucose with high sensitivity and selectivity, and importantly, in blood plasma with little sample preparation. The low potential at which the glucose is observed allows its detection with only minor disruption from common interferents such as ascorbic acid, acetaminophen and uric acid, giving improved resolution of the electrochemical signal. The approach to attachment of biological species to carbon nanotubes is generic and should work for most biomolecules, which means the range of potential applications is enormous. Recent work utilized organophosphorus hydrolase (OPH) non-specifically bound to horizontally-aligned SWNTs on a SiO2 substrate [338]. This electrode hydrolyses organophosphates (OPs) many of which are used as insecticides and other pesticides [339]. Changes to OPH upon exposure to OPs cause changes to nanotube conductance, which is monitored to determine a response. This type of electrode gives a real-time response and has been used for multiple analyses. Currently, a key challenge is to produce multiplexed electrodes that allow monitoring of two or more active species (for example, by coadsorbing enzymes [340]) and facilitating simultaneous detection of a number of analytes. Ultimately, it could be possible to co-adsorb enzymes and their mediators, or perhaps enzymes and cofactors, to build smart electrodes which only are activated in certain situations that cause release of the activating element. Studies involving direct electron transfer to enzymes, illustrate the potential advantages of carbon nanotubes modified electrodes. However,
Nanoscale Biosensors and Biochips
43
these studies employed randomly entangled nanotubes which give a poorly defined electrode surface and poorly defined protein immobilization. Aligned nanotubes electrodes will provide a more controlled surface upon which to immobilize thus improving communication with redox proteins [324,341]. Additionally, covalent attachment of the bioentity to the nanotubes promises to significantly increase efficiency of electron transfer, especially for the cases where distances between the redox centre and the probe are relatively large, for example, when biomolecules are attached to the apex of nanotubes. One approach is to vertically-align nanotubes using chemical vapor deposition (CVD) to produce vast arrays of long nanotubes [342]. Biosensors based on this design have been demonstrated to for example, detect glucose [343]. The main drawback of this approach is that high temperatures are required to make the nanotubes and adhesion of the nanotubes to the substrates is often weak, meaning that their use in longterm or field applications is questionable. Chemically producing the vertically-aligned nanotube arrays has the advantage of using a strong covalent bond for nanotube attachment, making further modification of the nanotubes straightforward. 5.2. Aligned Carbon Nanotube Electrodes for Direct Electron Transfer to Enzymes One the earliest examples of the fabrication of aligned carbon nanotubes electrodes used a short thiol to make a amine terminated SAM on a gold surface [341] which could be further reacted to attach acidfunctionalized SWNTs [327] to the substrate. The free end of the nanotubes, which still had available active groups, was then covalently bound to the enzyme microperoxidase MP-11 [341]. Attaching MP-11 to the aligned SWNT modified gold electrodes and subsequent electrochemical interrogation showed peaks characteristic of the heme redox active center of MP-11. Further to this, Willner et al. [344] attached the active centre of glucose oxidase (FAD) to the end of a vertically-aligned nanotube and then reconstituted the bound enzyme by wrapping apo-glucose oxidase around the FAD. This construct was able to detect glucose. Other early work detected hydrogen peroxide with
44
Leifert et al.
myoglobin- or horseradish peroxidase-modified nanotubes attached to pyrolytic graphite electrodes [345]. A disadvantage of using thiol-gold based SAMs is that despite the strong interaction previously observed between a thiol and gold [346], studies of alkanethiols on gold have revealed that they are susceptible to thermal instability [347,348], UV photoxidation [349] and adsorbatesolution interchange leading to poor long-term stability [350]. Therefore, the ability to use a different substrate for thiol attachment would be desirable. In view of the importance of silicon as the primary semiconductor material in modern microelectronic devices, efforts to control its electronic properties and tailor the chemical and physical characteristics of its surface are of major importance. Early work in this area reported the preparation of well-aligned carbon nanotube arrays on silicon (100) surfaces by reaction of hydride-terminated silicon (100) with ethyl undecylenate, producing SAMs that were linked by stable silicon–carbon covalent bonds [351]. However, the presence of a SAM of organic material hinders electron transport between carbon nanotubes and the underlying silicon substrate.
Figure 7. Nanotube sensor substrates. Schematic of general approach to attachment of a biomolecule to aligned carbon nanotubes anchored to a silicon substrate. Initial attachment of the nanotube to the silicon is done via a condensation reaction between surface –OH groups and the carboxylic acid groups of the oxided carbon nanotubes. Subsequently, unreacted acid groups on the nanotube are available for further modification to directly attach the biomolecule or as shown in the figure condensation reactions allow attachment of an intermediate which has a high affinity for the bioentity to be attached.
A new approach to covalently attach carbon nanotubes to silicon (without the use of intermediate molecules) has been developed using hydroxyl terminated silicon as the substrate [350]. This approach yielded
Nanoscale Biosensors and Biochips
45
vertically-aligned, shortened carbon nanotube architectures on a silicon (100) substrate. Compared to older techniques, the new approach has several advantages including the lower temperatures involved in preparation and the possibility for subsequent modifications. Electrochemical analysis of this interface demonstrated excellent conductivity to the substrate, a factor that is likely to see this approach adopted for numerous potential applications [352,353]. The attachment of SWCNTs directly to the silicon surface provides a simple and novel avenue for the fabrication and development of silicon-based electrochemical and bio-electrochemical sensors. As outlined in Figure 7 and earlier in this section, the approaches to attach biomolecules to functionalized nanotubes are well established [324,341] and future research will be aimed at further developing this biosensor platform. 6. Reporter Technologies: Nano-Sized Labels for Biosensing Applications Various transduction techniques including electrochemical, optical, piezoelectric, and thermometric have been applied to development of biosensing platforms. Many of these have been mentioned throughout previous sections in this review. Reviews on these different groups of biosensor transduction systems, including optical [354] (e.g. SPR [355]), piezoelectric [356], electrochemical [192,357,358], and thermal [359] can be found in the scientific literature. The demand for unsophisticated, low cost, portable yet sensitive biosensing devices, has directed biosensor research towards development of novel biosensing platforms with improved signal generation and transduction mechanisms. To this end, nanoparticle labels conjugated to reporter molecules show great potential for generation of strong optical or electrical signals upon biomolecular recognition. They have therefore been explored as promising alternatives to conventional labels that commonly include enzymes, fluorescent dyes or radioactive conjugates. There are a range of general reviews available on the use of nanoparticles as reporters [360-368]. In this section, we highlight properties of nanometer-sized labels that provide unique means for signal amplification, while overcoming limitations of traditional labels. We
46
Leifert et al.
focus on optical and electrical biosensors, currently the two most common transduction approaches. A selection of biosensors using nanosized signal generation elements, such as metallic and semiconductor nanoparticles or vesicles formed from amphiphilic molecules, will be discussed. 6.1. Biosensors Utilizing Optical Reporting Currently, biosensors relying on transduction of an optical signal are by far the most widespread [369]. Generally, optical read-out relies on generation of a colorimentric, fluorescent or chemiluminescent signal which can be visualised by eye, using e.g. CCD cameras or reflectometers as well as confocal and flatbed scanners. A large range of fluorescent (or fluorescence quenching) molecules are available for tagging of reporter elements. These include GFP (and variants such as yellow- and cyan-fluorescent proteins; YFP and CFP) [370,371], anthozoan GFP-like proteins [372-377], anthozoan non-fluorescent (quenching) chromoproteins [372], biarsenical ligands [378-380] and lanthanide probes [381,382], amongst others. Fluorescent components can be used as stand alone reporters (i.e. to report on the presence or absence of the fluorescent compound) or can be used with spectrally matched fluorescent/bioluminescent (or quenching) partners to produce resonance energy transfer (RET) assays that produce a fluorescent signal that is indicative of an interaction between labeled partners. RET assays may therefore be designed to potentially provide information about biological interactions or biosensor construction. Alternatively, label-free biosensors based on changes in optical properties (e.g. refractive index) of a thin film upon adsorption of biomolecules, have also been reported [383]. 6.1.1. Metallic Nanoparticle Labels Metallic nanoparticles exhibit unique optical and catalytic properties (see reviews [363,384]) that have seen them utilized for a variety of biological applications. A main advantage of gold colloid labels is that staining protocols involving the wet chemical deposition of metal on the
Nanoscale Biosensors and Biochips
47
Figure 8. Optical biosensors using nanometer-sized reporter labels. (a) Scanometric detection of a DNA array using silver amplified gold colloids; (i) working principle: in the presence of tagged DNA, a sandwich complex is formed using an oligonucleotide reporter tagged with a gold colloid followed by silver amplification; (ii) visualization of the tagged nanoparticles using a commercial flatbed scanner. Reproduced with permission [388]. (b) Multiplexing of a scanometric assay using silver amplified gold colloids and Raman active dyes. (i) working principle: the spectrum of a Raman dye is used as a “barcode” to individualize different oligonucleotide reporters (i.e. multiplexing); (ii) scanometric detection of a protein microarray based on silver amplified gold. The spectrum of three different Raman dyes is used for multiplexing. Top array image: the three protein targets are present. Bottom array image: only the protein target associated to cy5 is present. Reproduced with permission [400]. (c) Test strip immunoassay for botulism toxin. (i) working principle: dye-coated liposomes carrying a receptor (GT1b) for the toxin is mixed with the sample. After migrating along the strip, liposomes carrying the toxin are immobilized in the detection zone coated with anti-BT antibodies, resulting in a colorimetric signal; (ii) dose curve response obtained from scanned images of the strips. Reproduced with permission [419].
48
Leifert et al.
nanoparticles have long been available and can be used to increase the size of the colloid (and therefore the sensitivity of the assay) after biorecognition. Silver staining is most commonly used (Figure 8a). This feature makes it possible to detect low particle concentrations (to picomolar levels [385]) using the naked eye alone [385,386]. Quantification of the signal can be achieved using conventional flatbed scanners [387-389] or a CCD camera [390]. When conjugated to oligonucleotides, gold nanoparticles were also shown to confer very sharp melting profiles upon hybridised nucleotide strands [387]. This provides increased accuracy in discriminating single base mismatches, compared to standard assays performed using oligonucleotides labeled with fluorophores. Taton and coworkers [387] took advantage of this property and reported (after a silver amplification step) a femtomolar (50 fM) DNA detection limit on a microarray imaged with a conventional flatbed scanner. A commercially available device (Verigene® System by Nanosphere Inc., http: // www. nanosphere. Us / VerigeneSystem _ 4411. aspx ) uses side illumination to detect as little as 0.0025 probes/µm2 [388] (Figure 8a). This technology has been applied to a variety of biological assays [389,391-393]. Metallic nanoparticles also have the property of scattering light of a specific wavelength upon illumination with white light. Several groups have taken advantage of this phenomenon, commonly referred to as resonance light scattering (RLS), to develop new biosensing platforms. With this approach, a very intense signal is produced and no bleaching or quenching effects are observed (as for conventional fluorophores) [394]. Typical setups for the detection of RLS from nanoparticles in a microarray format are based on side illumination combined with TIRF [395-397] or on dark-field illumination [398]. With an RLS scanner, sensitivities were at least 50 times better than confocal scanners, with limits of detection in the femtomolar range (5-10 fM) being reported [396]. Multiplexing can potentially be achieved using metallic nanoparticles since the scattered wavelength depends on properties of the particles such as composition, size and shape [394,399] each of which could be varied within a single biosensor platform. Gold and silver nanoparticles were also shown to enhance the Raman scattering signal of
Nanoscale Biosensors and Biochips
49
organic molecules; this effect is commonly referred to as Surface Enhanced Raman Scattering (SERS). Raman spectra are generated upon illumination of a sample with laser light that produces excitation of vibrational and rotational states of chemical bonds which translate as wavelength shifts. Several DNA [400] and protein [401-403] biosensor setups have taken advantage of SERS and have used specific Raman spectra of (fluorescent and non-fluorescent) Raman dyes in the vicinity of metal nanoparticles as individual barcodes for multiplexing and identification of various reporter molecules (Figure 8b). The level of specificity of resultant Raman spectra allowed parallel use of up to six distinguishable labels and a femtomolar limit of detection [401]. These studies illustrate the potential for multiplexed microarray platforms relying on visualization of metallic nanoparticles. 6.1.2. Quantum Dot Labels Now that difficulties associated with their water solubility and functionalization are being overcome, quantum dots appear to be a promising alternative to conventional fluorescent labeling (with organic dyes) for a variety of biological applications. These semiconductor nanocrystals (with sizes usually ranging from 2 to 10 nm) exhibit unique luminescent characteristics; they were shown to be brighter and more stable against photobleaching than conventional fluorophores. Moreover, they exhibit a very broad adsorption spectrum and an emission spectrum that is narrow and size dependent (i.e. tunable) making them ideal candidates for multiplexed assays. Several groups have therefore started using quantum dot labels usually coupled to fluorescent microarray readout systems [404-408]. Up to four different quantum dots could be detected simultaneously using one excitation light source [409]. We refer the reader to a range of general reviews on biological applications and properties of quantum dots [362,410-412]. 6.1.3. Liposomes as Optical Labels Liposomes have been investigated as labels for a variety of biosensing applications because signaling molecules (such as enzymes,
50
Leifert et al.
fluorescent dyes and importantly, membrane proteins (see section 2)), can be encapsulated in their interior, bound to their surface or inserted within the bilayer (see reviews [368,413]). Signal amplification relies on the fact that a large number of signaling molecules can be associated to one binding event. Several portable immuno- or DNA sensors based on lateral flow assays, have been developed for field applications such as rapid detection of food and waterborne cellular pathogens [414-417], toxins [418,419] or pathogenic spores [420-422]. Flow assays have been performed on membrane strips (dip-stick sensor), in microcapillaries or in microfluidic channels. For example, in a colorimetric dip-stick assay [414-416,418422], reagents (sample and dye-loaded liposomes tagged with a reporter molecule) migrate on a membrane via capillary action until they reach the “capture zone” where a sandwich complex is formed. This results in a colorimetric signal that can be detected visually and quantified with a scanner [414,418,419,423] or hand-held reflectometer [416,420-422,424] (Figure 8c). Using the latter, nanomolar detection limits are commonly reported [424]. Fluorescence-based assays have utilized microcapillary columns [417,425-427] or microfluidic channels [415,428,429] coated with reporter molecules for target immobilization and formation of a sandwich complex with liposomes. Alternatively, complexes have been formed on the surface of a magnetic microparticle prior to immobilization in the channel using a magnet [430-432]. The signal was detected either directly using microscopy, or upon vesicle lysis and transport of the released dye to the detector situated at the end of the channel. This approach provided picomolar limits of detection (e.g. 5.5 pM for a cholera toxin assay [429]) after lysis of sensor liposomes. 6.2. Biosensors Utilising Electrochemical Reporting In an electrochemical biosensor, a biological signal is translated into an electric signal that is used as the means of detection. Advances in micro- and nanotechnology, as well as in the semiconductor industry, have opened the way for development of novel electrochemical biosensing platforms [192]. Such platforms have been presented as a low
Nanoscale Biosensors and Biochips
51
cost alternative to optical transduction systems for the production of small, hand-held (i.e. portable) sensors with simple read-out systems [365,433]. 6.2.1. Metallic and Semiconductor Nanoparticles as Electrochemical Reporters Metallic and semiconductor nanoparticles have also attracted much interest in biosensor research due to their unique electrical properties (see reviews [360,365,384]). The first electrical biosensing devices used metallic particles to measure changes in electrical conductivity between two microelectrodes upon biorecognition. In the assays presented by Park et al. [434] and Velev et al., [435] the biological sensor molecules were immobilized between two microelectrodes. Upon binding of a target molecule, a sandwich complex was formed with a reporter molecule tagged with a gold colloid. The change in conductivity after a silver enhancement step could be related to the amount of target material present in solution and limits of detection down to 0.2 pM were reported [435]. Similar biosensors making use of capacitance changes have also been described [436]. Alternatively, nanoparticles can be detected by monitoring label dissolution after biorecognition and surface immobilization. This can be achieved both directly by voltammetry [433,437,438] or potentiometry [439] measurements upon particle oxidation, or indirectly by stripping voltammetry [440-449] or stripping potentiometry [450-452]. In the latter method, chemical dissolution of the label after biorecognition is followed by a concentration step involving electrodeposition of the metal ions on the electrode. This is followed by electrochemical dissolution that results in an electrical signal, which can be used for quantification (Figure 9a). A great variety of metallic and semiconductor labels have been utilized including gold and silver colloids [440,442,448,449], indium microrods [451] and nanoparticles of CdS [443,444,446] or PbS [445,447]. Using PbS nanoparticles, a detection limit of ~0.2 pM was reported. An interesting feature of semiconductor nanoparticles is that multiplexing can be achieved using the defined stripping profiles as a barcode for the identification of different labels [453-455] (Figure 9b), with one study measuring up to five targets simultaneously [453].
52
Leifert et al.
Figure 9. Electrochemical biosensors using nanometer- sized reporter labels. (a) Detection of surface bound PbS nanoparticles using stripping voltammetry. Reproduced with permission [447]. (b) Assay multiplexing by use of three different semiconductor nanoparticle labels; (i) assay scheme; (ii) anodic stripping voltammetry analysis in the presence of the three lables (left) or only one label (right). Reproduced with permission [455]. (c) Amplified DNA detection using negatively charged liposomes and faradaic impedance spectroscopy. Reproduced with permission [459].
6.2.2. Liposomes as Electrochemical Reporters Electrochemical biosensors based on the release of encapsulated redox markers from phospholipid vesicles after biorecognition have also been described. Quantification is usually achieved by amperometry [456,457] or voltammetry [458]. Liposome-based electrochemical detection was also applied in lateral flow biosensor formats [456,457]. Alternative electrochemical approaches were proposed by Willner and coworkers [366] who described a transduction mechanism whereby the biorecognition event induced an increase in electron-transfer resistance at the electrode-solution interface which could be detected
Nanoscale Biosensors and Biochips
53
using faradaic impedance spectroscopy. In this assay format, negatively charged liposomes are used as labels for detection of a DNA target. Upon hybridization of the tagged reporter DNA, the interface between the sensor and the solution becomes highly negative and electrostatically repels positively-charged redox probes (which are present in the electrolyte) thereby increasing the electron transfer resistance (Figure 9c). Further amplification could be achieved by producing vesicle networks using biotin-streptavidin as a linker [459,460]. 7. Biosensing Applications The purpose of this section is to recognize current applications, discuss analytes of interest for currently utilized biomolecules and to give context to some of the immobilization transduction strategies discussed throughout this review. This is by no means a comprehensive review but more-so an overview of the biosensing field with some recent examples. Biosensors offer enormous potential to detect a wide range of analytes in a range of disciplines and we highlight applications in health care, the food industry, environmental monitoring, and defense/security. The primary goals when developing commercial biosensors for these applied uses are increased speed, heightened sensitivity, improved accuracy, portability and minimization of sample preparation. To facilitate adoption of next generation biosensors, these assay variables must be clearly superior to current standards (e.g. gas chromatography, mass spectrometry and high performance liquid chromatography). 7.1. Medical As biomarker research continues to identify new protein and chemical markers associated with a given pathology, commercial interest grows in producing pre-symptomatic diagnostic tools that can detect subclinical concentrations of known markers. Medical diagnostics requires fast, accurate, inexpensive devices. The most common analytes targeted in medical biosensing include glucose, lactate, urea, creatinine, cholesterol, uric acid and DNA.
54
Leifert et al.
Glucose is a common target analyte due to the prevalence of diseases characterized by altered sugar metabolism, such as types of diabetes (reviewed by Wang [461]). Glucose oxidase is commonly used as the biomolecular sensor element in glucose biosensors due to its affinity for glucose and the electrochemical signal produced by its interaction with glucose. Recently, work has become more focused on micro-type biosensors for applications in embedded glucose monitoring systems within the human body. Jia et al. recently reported on carbon nanotubebased needle-type glucose biosensors with this application in mind [462] (further discussion on nanotube based glucose sensing can be found in section 5). An alternative to the electrochemical approach for glucose biosensing is the use of optical transduction techniques (as reviewed by Pickup et al. [463]). One such approach was described by Wang et al. who reported the use of a fluorescent reporter system whereby oxygen consumed as a result of glucose oxidation caused a detectable change in fluorescence quenching [464] with the level of change being proportional to the glucose concentration of human serum samples being monitored. Urea concentration estimation is important for monitoring kidney function and any related disorders. The enzyme urease, which catalyses the reaction of urea into ammonia and carbon dioxide is commonly used as the biorecognition element for urea sensor systems. Recent studies include those by Jha et al. who entrapped urease in polyvinyl alcohol and a polyacrilamide polymer membrane, and monitored levels of NH4+ which is indicative of urease activity in the presence of urea [465]. Uric acid is used as an indicator of a wide range of conditions such as leukemia, pneumonia, kidney injury, hypertension and ischemia. The detection scheme for uric acid utilizes the enzymatic activity of uricase which produces a decrease in oxygen that is proportional to the concentration of uric acid. This system was recently used to determine the uric acid concentrations in human serum and urine [466]. Creatinine is another key medical analyte that is monitored to determine renal, muscular and thyroid dysfunction. Recent studies demonstrated an amperometric sensor based on creatinine amidinohydrolase and sarcosine oxidase [467] which are both used to detect the presence of this creatinine. Other analytes that have been
Nanoscale Biosensors and Biochips
55
measured by biosensors include cholesterol [468], insulin [469], nitric oxide (e.g. detected by myoglobin [470] or peroxidase activity [471]) and cytochrome c [472] all of which have diagnostic uses. In addition to the diagnostic applications mentioned here, biochips have potential in the drug discovery arena, a lucrative area of research that has utilized many of the existing cell-based biosensors. Biochips carrying biomolecules that are involved in certain diseases may provide a high-throughput method for drug screening of these targets (as discussed in more detail in section 2.1 of this review). 7.2. Food and Wine Food monitoring is important for the detection/quantification of microbial content, freshness/quality, and toxic ingredients (including pesticides and allergens). Additionally, testing manufacturing processes, such as fermentation for wine and beer production are also important to ensure quality and consistency of the end product. The ability to monitor multiple analytes with increased sensitivity will lead to improved control of production processes and improved profitability. The commercial demand for fast and sensitive food analysis methods has paved the way for the applicability of biosensors in this field. A review of enzymebased biosensors in food analysis has been published by Prodromidis and Karayannis [473] who discussed detection of glucose, fructose, sucrose, lactulose, lactose, lactic acid, malic acid, citric acid, glutamic acid, ascorbic acid (vitamin c), ethanol, and lysine, as well as the freshness indicators such as inosine which is used to monitor the progress of fermentation in wine-making since it is a fermentation byproduct that effects wine quality [474]. Additionally, inosine is an indicator of microorganism presence in food products [475]. Mycotoxins, which are toxic secondary metabolites produced by filamentous fungi, are dangerous contaminants which can be found in various food-stuffs, particularly grains. Mycotoxin-targeting biosenors based on enzyme and DNA bio-elements, have previously been reviewed by Prieto-Simon et al. [476]. More recently the same authors described an electrochemical immunosensor capable of detecting an example of
56
Leifert et al.
one such toxin, Ochratoxin A, in wine [477]. Some mycotoxins (e.g. aflatoxins) have been designated as biowarfare agents (discussed in section 7.3) due to their potential carcinogenicity [476]. 7.3. Explosives and Biowarfare Due to the nature of analytes targeted for explosives detection, a method known as “stand-off” detection is advantageous. This involves the detection of chemical vapors without necessarily seeing the explosives (e.g. if they are purposely concealed). Examples of biosensors developed for explosives detection have recently been reviewed by Smith et al. [478] TATP (triacetonetriperoxide) and TNT (2,4,6 – trinitrotoluene) are two explosives which have been investigated using immunosensors, enzymatic sensors, biologically inspired/biomimetic sensors and whole-cell biosensors (cell-based biosensing is discussed in more detail in section 2.2.2) [478]. Guan et al. also investigated the use of stochastic sensing (see section 2.2.2) to detect TNT using a genetically engineered pore-forming protein [479]. Interestingly, several mammalian ORs (see 2.1.7) have been isolated (Olfr226 from rat; Olfr2 and MOR226-1 from mouse) that are responsive to 2,4-dinitrotoluene [73] and therefore, could potentially be applied as olfactory sensors for certain explosives. “Biologically inspired” (biomimetic) sensors utilize a more robust substitute to mimic the biological element within the system. This may be important in some field applications where conditions to maintain an active biomolecule may not be easily adapted to a hand-held device. Additionally, in an application such as landmine detection, the outcome of false negatives due to sensor protein denaturation (e.g. receptor/antibody) is not trivial [480]. Some of these biological mimics adapted to a variety of biosensing/biochip applications (not only security), include aptamers (single stranded DNA acting as receptor/protein mimics) [481] and carbon nanotubes (as ion-channel mimics) [482,483]. There are a number of biological agents that are tagged as a potential security threats. These include botulism toxin, Smallpox virus,
Nanoscale Biosensors and Biochips
57
Hemorrhagic fever viruses (including Ebola and Yellow fever viruses), Fracisella tularensis (causing tularemia), Yesinis pestis (causing plague), and Bacillus anthracis (anthrax) [484]. Other toxins that pose a threat include ricin and diphtheria toxins. In the area of bio-defence, sensitive, early and accurate detection of agents such as these is of high importance. Huelseweh et al. have developed a simple and rapid modification to the ELISA technique that allows simultaneous detection of a number of different biowarfare agents on a protein chip using biotinylated antibody recognition and streptavidin-HRP signal amplification [485]. Investigations into detection of smallpox virions [486] and antibodies to the Ebola virus [487] were conducted using optical immunosensors as a faster, portable alternative to the ELISA assays commonly employed. Halverson et al. [488] investigated the three proteins that comprise the active anthrax toxin, protective antigen (PA), lethal factor (LF) and edema factor (EF), and the effect of the pore made by a fragment of PA, in the presence and absence of LF and EF. The change in current through the pore was monitored as an indication of the presence of the anthrax proteins. Another study described construction of a hand-held SPR-based device capable of detecting ricin at 200 ng/ml in 10 minutes [489]. While this device is still in its prototype stage and researches admit to certain limitations of the system as it stands, it is still a step towards a generic hand-held sensing device. Additionally, the SPR technique can potentially be adapted to monitor a wide range of interactions and therefore is likely to find use in many applications. 7.4. Environmental Detection of environmental pollutants in air, water and soil, is important for the health and well-being of humans and other biodiversity that rely on these resources. A recent review by Wanekaya et al. [490] covers many of the current biosensing developments in the environmental field. Other reviews in this area have focused on particular classes of biological detectors, such as enzymes [491] or whole cells [67], while others have focused on particular analytes such as heavy
58
Leifert et al.
metals [492-495], dioxins [496], pesticides [497], endocrine-disrupting compounds [498], or water-born pathogens [499]. These sensor technologies are moving from proof-of-concept to real-world applications where river water, wastewater, groundwater and soil have been tested for a number of the analytes mentioned above (for a review of literature involved see Rodriguez-Mozaz et al. [500]). Some recent studies into development of biosensors targeting organophosphorous (OP) pesticides provide an example of some approaches to environmental sensing. OP pesticides have been widely utilized to control agricultural and household pests, however, they are also toxic to humans and other mammals. They constitute a range of chemical structures and exhibit a range of physicochemical properties, with their primary toxicological action arising from inhibition of the enzyme, acetylcholinesterase (AChE), which is important to nerve impulse responses [501]. It is this enzyme which is predominantly used as the sensor biomolecule for current OR biosensor designs. Recently, Istamboulie et al. [502] captured recombinant histidine-tagged AchE proteins on magnetic beads using Ni2+-histidine affinity. These beads can be attached to the surface of a working electrode for amperometric transduction, by the application of a magnetic field. An advantage of this system was the re-usability of the electrode. Another report recently described the use of screen-printed electrodes for production of an amperometric biosensor using immobilized AchE [503,504]. Vamvakaki et al. [505] presented a pH-sensitive fluorescent indicator to monitor the activity of AchE encapsulated within liposomes. The internal environment of the liposome has been reported to improve enzyme stability (for more discussion of liposomes see section 3), and liposomes can be adaptable to sol-gel matrices for biochip/array pesticide-screening applications. Microcantilevers were also recently explored for their use in detecting OPs by Karnati et al. [506]. In addition, this group investigated an alternative enzyme (organophosphorous hydrolase, OPH) for its ability to hydrolyse OP. The detection was based on the deflection of the cantilever due to changes in enzyme conformation induced by the hydrolysis of OP by cantilever immobilized OPH. Nanotubes have also been used in conjunction with this OPH as discussed in section 5.1.
Nanoscale Biosensors and Biochips
59
8. Conclusion It is evident from the increasing number of biosensor-related publications and the array of associated approaches, that biosensing is seen as an important area of research worldwide. This interest is set to increase as more biologically active proteins are characterized and the first commercial biosensors are developed. Currently, molecular biosensing is still a relatively new field of research with few commercial examples. In this review, we have discussed some of the key classes of biological sensors being studied (membrane proteins, enzymes, whole cells and virions) and their current assay technologies, the use of lipid supports, nanopatterning approaches, use of carbon nanotubes and current biosensing applications and their target analytes. Currently, the main factors limiting the wide-spread use of biosensors include (1) the need to overcome issues of functional integrity of sensor proteins/cells under harsh purification or assay conditions, (2) limited ability to correctly and consistently orientate sensor proteins, (3) limitations to use of lipid supports, particularly with regard to targeted insertion of membrane proteins and even coverage of biochips, (4) need for metabolic relevancy of detected events for medical applications, (5) lack of portability and (6) expense. Many of these limiting factors currently drive aspects of biosensor design, often restricting the approach that can be used for a given application, however, advances in biosensor-related technologies should see design flexibility improve. There are some clear trends emerging and the next decade of biosensor research is likely to be characterized, for example, by increasing use of cell-free approaches. Cell-based approaches, which are in a sense traditional, are often utilized and may lend metabolic relevancy to a sensing event. To this end, it is likely that culturing surfaces will be developed which can be applied to a potential biosensor surface (e.g. chips and electrodes) and facilitate cell growth, proliferation and differentiation directly on targeted areas of the surface. However, cellular approaches are inherently limited by the need to maintain cells and their fragility under assay conditions, particularly in the field. In addition, cell-based sensors are largely confined to medical and other screening applications.
60
Leifert et al.
Other trends are likely to include miniaturization of sensing elements, increased use of biochips containing sensor arrays, multiplexing of sensing and transduction systems, advances in the use of lipid supports and membrane protein sensors, increased use of optical biosensors that utilize a range of novel fluorescent reporters, increased use of microfluidics and biosensing of volatiles. It is also likely that good progress will be made in attempts to increase efficiency of electron transfer for electrochemical approaches. In combination, these advances will see commercially produced biosensors become increasingly commonplace and the range of biosensing applications expand rapidly. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
L. C. Clark Jr and C. Lyons, Ann. N. Y. Acad. Sci. 102, 29 (1962) A. Wise, K. Gearing and S. Rees, Drug Discov. Today. 7, 235 (2002) K. L. Pierce, R. T. Premont and R. J. Lefkowitz, Nat. Rev. Mol. Cell Biol. 3, 639 (2002) W. R. Leifert, A. L. Aloia, O. Bucco, R. V. Glatz and E. J. McMurchie, J. Biomol. Screen. 10, 765 (2005) J. L. Swift, M. C. Burger, D. Massotte, T. E. Dahms and D. T. Cramb, Anal. Chem. 79, 6783 (2007) Y. Fang, A. G. Frutos and J. Lahiri, J. Am. Chem. Soc. 124, 2394 (2002) Y. Fang, A. G. Frutos, B. Webb, Y. Hong, A. Ferrie, F. Lai and J. Lahiri, BioTechniques. Suppl, 62 (2002) Y. Fang, A. G. Frutos and J. Lahiri, Chembiochem. 3, 987 (2002) Y. Fang, J. Peng, A. M. Ferrie and R. S. Burkhalter, Anal. Chem. 78, 149 (2006) A. Waller, P. Simons, E. R. Prossnitz, B. S. Edwards and L. A. Sklar, Comb. Chem. High Throughput Screen. 6, 389 (2003) T. Mirzabekov, H. Kontos, M. Farzan, W. Marasco and J. Sodroski, Nat. Biotechnol. 18, 649 (2000) N. M. Rao, V. Silin, K. D. Ridge, J. T. Woodward and A. L. Plant, Anal. Biochem. 307, 117 (2002) O. Karlsson and L. Stefan, Analytical Biochemistry. 300, 132 (2002) P. Stenlund, G. J. Babcock, J. Sodroski and D. G. Myszka, Anal. Biochem. 316, 243 (2003) K. L. Martinez, B. H. Meyer, R. Hovius, K. Lundstrom and H. Vogel, Langmuir. 19, 10925 (2003)
Nanoscale Biosensors and Biochips 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
28.
29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
61
L. Neumann, T. Wohland, R. J. Whelan, R. N. Zare and B. K. Kobilka, Chembiochem. 3, 993 (2002) S. Devanathan, Z. Yao, Z. Salamon, B. Kobilka and G. Tollin, Biochemistry. 43, 3280 (2004) I. D. Alves, S. M. Cowell, Z. Salamon, S. Devanathan, G. Tollin and V. J. Hruby, Mol. Pharmacol. 65, 1248 (2004) G. Milligan, Trends Pharmacol. Sci. 24, 87 (2003) C. Harrison and J. R. Traynor, Life Sci. 74, 489 (2003) H. Frang, V. M. Mukkala, R. Syysto, P. Ollikka, P. Hurskainen, M. Scheinin and I. Hemmila, Assay Drug Dev. Technol. 1, 275 (2003) M. Engstrom, A. Narvanen, J. M. Savola and S. Wurster, Peptides. 25, 2099 (2004) W. R. Leifert, A. L. Aloia, O. Bucco and E. J. McMurchie, Mol. Membr. Biol. 22, 507 (2005) M. Bunemann, M. Frank and M. J. Lohse, Proc. Natl. Acad. Sci. U. S. A. 100, 16077 (2003) C. Bieri, O. P. Ernst, S. Heyse, K. P. Hofmann and H. Vogel, Nat. Biotechnol. 17, 1105 (1999) D. L. Johnson and L. L. Martin, J. Am. Chem. Soc. 127, 2018 (2005) C. S. Barton, R. V. Glatz, A. S. Martin, L. Waniganayake, E. J. McMurchie, W. R. Leifert, G. Wei, D. J. Bray, B. Raguse and L. Wieczorek, Journal of Bionanoscience. 1, 22 (June 2007) P. C. Simons, M. Shi, T. Foutz, D. F. Cimino, J. Lewis, T. Buranda, W. K. Lim, R. R. Neubig, W. E. McIntire, J. Garrison, E. Prossnitz and L. A. Sklar, Mol. Pharmacol. 64, 1227 (2003) A. Waller, P. C. Simons, S. M. Biggs, B. S. Edwards, E. R. Prossnitz and L. A. Sklar, Trends Pharmacol. Sci. 25, 663 (2004) L. A. Sklar, J. Vilven, E. Lynam, D. Neldon, T. A. Bennett and E. Prossnitz, BioTechniques. 28, 976 (2000) X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li, G. Sundaresan, A. M. Wu, S. S. Gambhir and S. Weiss, Science. 307, 538 (2005) R. F. Heuff, J. L. Swift and D. T. Cramb, Phys. Chem. Chem. Phys. 9, 1870 (2007) L. Buck and R. Axel, Cell. 65, 175 (1991) Q. Gao and A. Chess, Genomics. 60, 31 (1999) H. Breer, Anal. Bioanal Chem. 377, 427 (2003) E. A. Hallem, A. Dahanukar and J. R. Carlson, Annu. Rev. Entomol. 51, 113 (2006) A. Chesler and S. Firestein, Nature. 452, 944 (2008) D. Wicher, R. Schafer, R. Bauernfeind, M. C. Stensmyr, R. Heller, S. H. Heinemann and B. S. Hansson, Nature. 452, 1007 (2008)
62
Leifert et al.
39.
K. Sato, M. Pellegrino, T. Nakagawa, T. Nakagawa, L. B. Vosshall and K. Touhara, Nature. 452, 1002 (2008) G. Gomila, I. Casuso, A. Errachid, A. Ruiz, E. Pajot, J. Minic, T. Gorojankina, M. A. Persuy, J. Aioun, R. Salesse, J. Bausells, G. Villanueva, G. Rius, Y. Hou, N. Jaffrezic, C. Pennetta, E. Alfinito, V. Akimov, L. Reggiani, G. Ferrari, L. Fumagalli, M. Sampietro and J. Samitier, Sens. Actuator B-Chem. 116, 66 (2006) V. Jacquier, H. Pick and H. Vogel, J. Neurochem. 97, 537 (2006) H. Zhao, L. Ivic, J. M. Otaki, M. Hashimoto, K. Mikoshiba and S. Firestein, Science. 279, 237 (1998) P. Sengupta, J. H. Chou and C. I. Bargmann, Cell. 84, 899 (1996) Y. Zhang, J. H. Chou, J. Bradley, C. I. Bargmann and K. Zinn, Proc. Natl. Acad. Sci. U. S. A. 94, 12162 (1997) Y. Hou, N. Jaffrezic-Renault, C. Martelet, A. Zhang, J. Minic-Vidic, T. Gorojankina, M. A. Persuy, E. Pajot-Augy, R. Salesse, V. Akimov, L. Reggiani, C. Pennetta, E. Alfinito, O. Ruiz, G. Gomila, J. Samitier and A. Errachid, Biosens. Bioelectron. 22, 1550 (2007) J. Vidic, M. Pla-Roca, J. Grosclaude, M. A. Persuy, R. Monnerie, D. Caballero, A. Errachid, Y. Hou, N. Jaffrezic-Renault, R. Salesse, E. Pajot-Augy and J. Samitier, Anal. Chem. 79, 3280 (2007) J. H. Sung, H. J. Ko and T. H. Park, Biosens. Bioelectron. 21, 1981 (2006) J. Y. Lee, H. J. Ko, S. H. Lee and T. H. Park, Enzyme Microb. Technol. 39, 375 (2006) B. Alberts, A. Johnson, J. Lewis, M. Raff, K. Roberts and P. Walter (2002) Molecular Biology of the Cell. Garland Science (New York) S. W. Cowan, T. Schirmer, G. Rummel, M. Steiert, R. Ghosh, R. A. Pauptit, J. N. Jansonius and J. P. Rosenbusch, Nature. 358, 727 (1992) L. Z. Song, M. R. Hobaugh, C. Shustak, S. Cheley, H. Bayley and J. E. Gouaux, Science. 274, 1859 (1996) B. A. Wallace, Annu. Rev. Biophys. Biophys. Chem. 19, 127 (1990) D. P. Nikolelis, C. G. Siontorou, U. J. Krull and P. L. Katrivanos, Anal. Chem. 68, 1735 (1996) S. Blake, T. Mayer, M. Mayer and J. Yang, Chembiochem. 7, 433 (2006) R. Capone, S. Blake, M. R. Restrepo, J. Yang and M. Mayer, J. Am. Chem. Soc. 129, 9737 (2007) T. Stora, J. H. Lakey and H. Vogel, Angewandte Chemie-International Edition. 38, 389 (1999) O. Braha, L. Q. Gu, L. Zhou, X. F. Lu, S. Cheley and H. Bayley, Nat. Biotechnol. 18, 1005 (2000)
40.
41. 42. 43. 44. 45.
46.
47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57.
Nanoscale Biosensors and Biochips 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.
63
H. Bayley and P. S. Cremer, Nature. 413, 226 (2001) B. A. Cornell, V. L. Braach-Maksvytis, L. G. King, P. D. Osman, B. Raguse, L. Wieczorek and R. J. Pace, Nature. 387, 580 (1997) G. Woodhouse, L. King, L. Wieczorek, P. Osman and B. Cornell, Journal of Molecular Recognition. 12, 328 (1999) S. Y. Oh, B. Cornell, D. Smith, G. Higgins, C. J. Buffell and T. W. Kok, Biosens. Bioelectron. 23, 1161 (2008) J. Schmidt, Journal of Materials Chemistry. 15, 831 (2005) L. Q. Gu, O. Braha, S. Conlan, S. Cheley and H. Bayley, Nature. 398, 686 (1999) S. Howorka, L. Movileanu, O. Braha and H. Bayley, Proc. Natl. Acad. Sci. U. S. A. 98, 12996 (2001) J. W. Shim and L. Q. Gu, Anal. Chem. 79, 2207 (2007) X. F. Kang, S. Cheley, A. C. Rice-Ficht and H. Bayley, J. Am. Chem. Soc. 129, 4701 (2007) K. Yagi, Appl. Microbiol. Biotechnol. 73, 1251 (2007) S. J. Sorensen, M. Burmolle and L. H. Hansen, Curr. Opin. Biotechnol. 17, 11 (2006) R. Popovtzer, T. Neufeld, E. Z. Ron, J. Rishpon and Y. Shacham-Diamand, Sensors Actuators B: Chem. 119, 664 (2006) S. Timur, U. Anik, D. Odaci and L. Gorton, Electrochemistry Communications. 9, 1810 (2007) K. H. Baronian, Biosens. Bioelectron. 19, 953 (2004) D. R. W. Yina Kuang, Biotechnol. Bioeng. 96, 318 (2007) V. Radhika, T. Proikas-Cezanne, M. Jayaraman, D. Onesime, J. H. Ha and D. N. Dhanasekaran, Nat. Chem. Biol. 3, 325 (2007) M. Marrakchi, J. Vidic, N. Jaffrezic-Renault, C. Martelet and E. Pajot-Augy, Eur. Biophys. J. 36, 1015 (2007) C. Vedrine, J. C. Leclerc, C. Durrieu and C. Tran-Minh, Biosens. Bioelectron. 18, 457 (2003) B. Podola, E. C. Nowack and M. Melkonian, Biosens. Bioelectron. 19, 1253 (2004) B. Xi, N. Yu, X. Wang, X. Xu and Y. A. Abassi, Biotechnol. J. 3, 484 (2008) R. McGuinness, Curr. Opin. Pharmacol. 7, 535 (2007) P. Wolf, A. Rothermel, A. G. Beck-Sickinger and A. A. Robitzki, Biosens. Bioelectron. (2008) Y. Fang, G. Li and A. M. Ferrie, J. Pharmacol. Toxicol. Methods. 55, 314 (2007) K. A. Marx, T. Zhou, A. Montrone, D. McIntosh and S. J. Braunhut, Anal. Biochem. 361, 77 (2007) T. Endo, S. Yamamura, K. Kerman and E. Tamiya, Anal. Chim. Acta. 614, 182 (2008)
64
Leifert et al.
83. 84.
S. B. VanEngelenburg and A. E. Palmer, Curr. Opin. Chem. Biol. 12, 60 (2008) S. Kumar, N. Harrison, R. Richards-Kortum and K. Sokolov, Nano Lett. 7, 1338 (2007) S. Buck, H. Xu, M. Brasuel, M. Philbert and R. Kopelman, Talanta. 63, 41 (2004) A. K. Poulsen, A. M. Scharff-Poulsen and L. F. Olsen, Anal. Biochem. 366, 29 (2007) M. Brasuel, R. Kopelman, T. J. Miller, R. Tjalkens and M. A. Philbert, Anal. Chem. 73, 2221 (2001) H. A. Clark, R. Kopelman, R. Tjalkens and M. A. Philbert, Anal. Chem. 71, 4837 (1999) A. Leung, P. M. Shankar and R. Mutharasan, Sensors and Actuators B: Chemical. 125, 688 (2007) T. Vo-Dinh, P. Kasili and M. Wabuyele, Nanomedicine. 2, 22 (2006) X. Feng, J. Castracane, N. Tokranova, A. Gracias, G. Lnenicka and B. G. Szaro, Biosens. Bioelectron. 22, 3230 (2007) A. L. Hook, H. Thissen and N. H. Voelcker, Trends Biotechnol. 24, 471 (2006) B. W. Matthews, J. Mol. Biol. 33, 491 (1968) S. Lakard, G. Herlem, N. Valles-Villareal, G. Michel, A. Propper, T. Gharbi and B. Fahys, Biosens. Bioelectron. 20, 1946 (2005) S. Lakard, N. Morraznd-Villeneuve, E. Lesniewska, B. Lakard, G. Michel, G. Herlem, T. Gharbi and B. Fahys, Electrochim. Acta. 1114 (2007) S. Lee, J. Y. Kang, I. Lee, S. Ryu, S. Kwak, K. Shin, C. Kim, H. Jung and T. Kim, Sensors and Actuators A: Physical. 143, 64 (2008) M. Manchester and P. Singh, Adv. Drug Deliv. Rev. 58, 1505 (2006) G. Destito, R. Yeh, C. S. Rae, M. G. Finn and M. Manchester, Chem. Biol. 14, 1152 (2007) G. R. Souza, D. R. Christianson, F. I. Staquicini, M. G. Ozawa, E. Y. Snyder, R. L. Sidman, J. H. Miller, W. Arap and R. Pasqualini, Proc. Natl. Acad. Sci. U. S. A. 103, 1215 (2006) J. Drews, Science. 287, 1960 (2000) G. Di Paolo and P. De Camilli, Nature. 443, 651 (2006) D. M. Engelman, Nature. 438, 578 (2005) D. D. Lasic (1993) Liposomes: From Physics to Applications. Elsevier L. K. Tamm and H. M. McConnell, Biophys. J. 47, 105 (1985) E. Reimhult, M. Zach, F. Hook and B. Kasemo, Langmuir. 22, 3313 (2006) M. Stelzle, G. Weissmuller and E. Sackmann, J. Phys. Chem. 97, 2974 (1993) H. T. Tien and Z. Salamon, Bioelectrochem. Bioenerg. 22, 211 (1989) C. Danelon, J. B. Perez, C. Santschi, J. Brugger and H. Vogel, Langmuir. 22, 22 (2006)
85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99.
100. 101. 102. 103. 104. 105. 106. 107. 108.
Nanoscale Biosensors and Biochips
65
109. R. P. Richter, R. Berat and A. R. Brisson, Langmuir. 22, 3497 (2006) 110. V. I. Silin, H. Wieder, J. T. Woodward, G. Valincius, A. Offenhausser and A. L. Plant, J. Am. Chem. Soc. 124, 14676 (2002) 111. A. Janshoff and C. Steinem, Anal. Bioanal Chem. 385, 433 (2006) 112. E. Reimhult and K. Kumar, Trends Biotechnol. 26, 82 (2008) 113. M. Tanaka and E. Sackmann, Nature. 437, 656 (2005) 114. S. D. Ogier, R. J. Bushby, Y. Cheng, S. D. Evans, S. W. Evans, A. T. A. Jenkins, P. F. Knowles and R. E. Miles, Langmuir. 16, 5696 (2000) 115. O. Purrucker, H. Hillebrandt, K. Adlkofer and M. Tanaka, Electrochimica Acta. 47, 791 (2001) 116. A. A. Brian and H. M. McConnell, Proc. Natl. Acad. Sci. 81, 6159 (1984) 117. M. Grandbois, H. Clausen-Schaumann and H. Gaub, Biophysical Journal. 74, 2398 (1998) 118. C. Steinem, H. J. Galla and A. Janshoff, Physical Chemistry Chemical Physics. 2, 4580 (2000) 119. L. H. He, J. W. F. Robertson, J. Li, I. Karcher, S. M. Schiller, W. Knoll and R. Naumann, Langmuir. 21, 11666 (2005) 120. X. M. Bin, I. Zawisza, J. D. Goddard and J. Lipkowski, Langmuir. 21, 330 (2005) 121. J. Ekeroth, P. Konradsson and F. Hook, Langmuir. 18, 7923 (2002) 122. C. Steinem, A. Janshoff, W. Ulrich, M. Sieber and H. Galla, Biochim. Biophys. Acta. 1279, 169 (1996) 123. A. Devadoss, M. S. Palencsar, D. C. Jiang, M. L. Honkonen and J. D. Burgess, Analytical Chemistry. 77, 7393 (2005) 124. A. Devadoss and J. D. Burgess, Journal of the American Chemical Society. 126, 10214 (2004) 125. J. Salafsky, J. T. Groves and S. G. Boxer, Biochemistry. 35, 14773 (1996) 126. H. T. Tien and A. L. Ottova, Electrochimica Acta. 43, 3587 (1998) 127. J. D. Burgess, M. C. Rhoten and F. M. Hawkridge, Langmuir. 14, 2467 (1998) 128. Z. C. Bi, H. T. Tien and A. L. Ottova, Progress in Biochemistry and Biophysics. 28, 447 (2001) 129. A. Mashaghi, B. Swann, J. Popplewell, M. Textor and E. Reimhult, Anal. Chem. 80, 3666–3676 (2008) 130. C. Merz, W. Knoll, M. Textor and E. Reimhult, Biointerphases. 3, FA41 (2008) 131. Z. Salamon and G. Tollin, Biophysical Journal. 80, 1557 (2001) 132. H. Lang, C. Duschl and H. Vogel, Langmuir. 10, 197 (1994) 133. F. Giess, M. G. Friedrich, J. Heberle, R. L. Naumann and W. Knoll, Biophys. J. 87, 3213 (2004) 134. M. L. Wagner and L. K. Tamm, Biophys. J. 79, 1400 (2000)
66
Leifert et al.
135. A. T. A. Jenkins, R. J. Bushby, N. Boden, S. D. Evans, P. F. Knowles, Q. Y. Liu, R. E. Miles and S. D. Ogier, Langmuir. 14, 4675 (1998) 136. L. J. C. Jeuken, S. D. Connell, P. J. F. Henderson, R. B. Gennis, S. D. Evans and R. J. Bushby, Journal of the American Chemical Society. 128, 1711 (2006) 137. S. Terrettaz, M. Mayer and H. Vogel, Langmuir. 19, 5567 (2003) 138. O. Purrucker, A. Fortig, R. Jordan and M. Tanaka, Chemphyschem. 5, 327 (2004) 139. N. Bunjes, E. K. Schmidt, A. Jonczyk, F. Rippmann, D. Beyer, H. Ringsdorf, P. Graber, W. Knoll and R. Naumann, Langmuir. 13, 6188 (1997) 140. R. Naumann, E. K. Schmidt, A. Jonczyk, K. Fendler, B. Kadenbach, T. Liebermann, A. Offenhausser and W. Knoll, Biosensors and Bioelectronics. 14, 651 (1999) 141. E. K. Schmidt, T. Liebermann, M. Kreiter, A. Jonczyk, R. Naumann, A. Offenhausser, E. Neumann, A. Kukol, A. Maelicke and W. Knoll, Biosensors and Bioelectronics. 13, 585 (1998) 142. C. Steinem, A. Janshoff, K. von dem Bruch, K. Reihs, J. Goossens and H. J. Galla, Bioelectrochemistry and Bioenergetics. 45, 17 (1998) 143. C. Daniel, K. E. Sohn, T. E. Mates, E. J. Kramer, J. O. Radler, E. Sackmann, B. Nickel and L. Andruzzi, Biointerphases. 2, 109 (2007) 144. H. M. Keizer, B. R. Dorvel, M. Andersson, D. Fine, R. B. Price, J. R. Long, A. Dodabalapur, I. Koper, W. Knoll, P. A. V. Anderson and R. S. Duran, ChemBioChem. 8, 1246 (2007) 145. R. Naumann, S. M. Schiller, F. Giess, B. Grohe, K. B. Hartman, I. Karcher, I. Koper, J. Lubben, K. Vasilev and W. Knoll, Langmuir. 19, 5435 (2003) 146. S. M. Schiller, R. Naumann, K. Lovejoy, H. Kunz and W. Knoll, Angewandte Chemie-International Edition. 42, 208 (2003) 147. Y. Deng, Y. Wang, B. Holtz, J. Y. Li, N. Traaseth, G. Veglia, B. J. Stottrup, R. Elde, D. Q. Pei, A. Guo and X. Y. Zhu, J. Am. Chem. Soc. 130, 6267 (2008) 148. K. Ataka, F. Giess, W. Knoll, R. Naumann, S. Haber-Pohlmeier, B. Richter and J. Heberle, Journal of the American Chemical Society. 126, 16199 (2004) 149. S. R. Jadhav, D. X. Sui, R. M. Garavito and R. M. Worden, J. Colloid Interface Sci. 322, 465 (2008) 150. O. Purrucker, S. Gonnenwein, A. Fortig, R. Jordan, M. Rusp, M. Barmann, L. Moroder, E. Sackmann and M. Tanaka, Soft Matter. 3, 333 (2007) 151. H. Hillebrandt, G. Wiegand, M. Tanaka and E. Sackmann, Langmuir. 15, 8451 (1999) 152. C. Hennesthal, J. Drexler and C. Steinem, Chemphyschem. 3, 885 (2002) 153. J. Y. Wong, J. Majewski, M. Seitz, C. K. Park, J. N. Israelachvili and G. S. Smith, Biophysical Journal. 77, 1445 (1999)
Nanoscale Biosensors and Biochips
67
154. M. Tanaka, A. P. Wong, F. Rehfeldt, M. Tutus and S. Kaufmann, Journal of the American Chemical Society. 126, 3257 (2004) 155. M. Tanaka, S. Kaufmann, J. Nissen and M. Hochrein, Physical Chemistry Chemical Physics. 3, 4091 (2001) 156. L. Renner, T. Osaki, S. Chiantia, P. Schwille, T. Pompe and C. Werner, J. Phys. Chem. B. 112, 6373 (2008) 157. B. Deme and D. Marchal, European Biophysics Journal with Biophysics Letters. 34, 170 (2005) 158. J. C. Munro and C. W. Frank, Langmuir. 20, 10567 (2004) 159. C. Rossi, J. Homand, C. Bauche, H. Hamdi, D. Ladant and J. Chopineau, Biochemistry. 42, 15273 (2003) 160. S. Goennenwein, M. Tanaka, B. Hu, L. Moroder and E. Sackmann, Biophysical Journal. 85, 646 (2003) 161. F. Albertorio, V. A. Chapa, X. Chen, A. J. Diaz and P. S. Cremer, J. Am. Chem. Soc. 129, 10567 (2007) 162. F. Albertorio, A. J. Diaz, T. L. Yang, V. A. Chapa, S. Kataoka, E. T. Castellana and P. S. Cremer, Langmuir. 21, 7476 (2005) 163. A. Simon, A. Girard-Egrot, F. Sauter, C. Pudda, N. P. D'Hahan, L. Blum, F. Chatelain and A. Fuchs, Journal of Colloid and Interface Science. 308, 337 (2007) 164. X. Han, A. Studer, H. Sehr, I. Geissbühler, M. Di Berardino, F. K. Winkler and L. X. Tiefenauer, Advanced Materials (Accepted). (2007) 165. J. Drexler and C. Steinem, Journal of Physical Chemistry B. 107, 11245 (2003) 166. W. Romer and C. Steinem, Biophysical Journal. 86, 955 (2004) 167. E. K. Schmitt, M. Nurnabi, R. J. Bushby and C. Steinem, Soft Matter. 4, 250 (2008) 168. E. K. Schmitt, M. Vrouenraets and C. Steinem, Biophys. J. 91, 2163 (2006) 169. S. Svedhem, I. Pfeiffer, C. Larsson, C. Wingren, C. Borrebaeck and F. Hook, Chembiochem. 4, 339 (2003) 170. S. M. Christensen and D. Stamou, Soft Matter. 3, 828 (2007) 171. D. Stamou, C. Duschl, E. Delamarche and H. Vogel, Angew. Chem. Int. Ed Engl. 42, 5580 (2003) 172. C. Yoshina-Ishii and S. G. Boxer, J. Am. Chem. Soc. 125, 3696 (2003) 173. B. Stadler, D. Falconnet, I. Pfeiffer, F. Hook and J. Voros, Langmuir. 20, 11348 (2004) 174. B. Stadler, M. Bally, D. Grieshaber, J. Voros, A. Brisson and H. M. Grandin, Biointerphases. 1, 142 (2006) 175. A. Graneli, M. Edvardsson and F. Hook, ChemPhysChem. 5, 729 (2004) 176. G. Klenkar, B. Brian, T. Ederth, G. Stengel, F. Höök, J. Piehler and B. Liedberg, Biointerphases. 3, 29 (2008)
68
Leifert et al.
177. J. J. Benkoski and F. Hook, J. Phys. Chem. B. 109, 9773 (2005) 178. Y. H. M. Chan, P. Lenz and S. G. Boxer, Proc. Natl. Acad. Sci. U. S. A. 104, 18913 (2007) 179. R. Michel, I. Reviakine, D. Sutherland, C. Fokas, G. Csucs, G. Danuser, N. D. Spencer and M. Textor, Langmuir. 18, 8580 (2002) 180. J. T. Groves, N. Ulman and S. G. Boxer, Science. 275, 651 (1997) 181. J. S. Hovis and S. G. Boxer, Langmuir. 16, 894 (2000) 182. J. S. Hovis and S. G. Boxer, Langmuir. 17, 3400 (2001) 183. L. A. Kung, L. Kam, J. S. Hovis and S. G. Boxer, Langmuir. 16, 6773 (2000) 184. A. R. Sapuri-Butti, Q. J. Li, J. T. Groves and A. N. Parikh, Langmuir. 22, 5374 (2006) 185. K. Morigaki, Biointerphases. 3, FA85 (2008) 186. Y. Ueda, K. Morigaki, Y. Tatsu, N. Yumoto and H. Imaishi, Biochem. Biophys. Res. Commun. 355, 926 (2007) 187. B. L. Jackson and J. T. Groves, J. Am. Chem. Soc. 126, 13878 (2004) 188. J. J. Shi, J. X. Chen and P. S. Cremer, Journal of the American Chemical Society. 130, 2718 (2008) 189. J. M. Moran-Mirabal, J. B. Edel, G. D. Meyer, D. Throckmorton, A. K. Singh and H. G. Craighead, Biophysical Journal. 89, 296 (2005) 190. L. Kam and S. G. Boxer, J. Am. Chem. Soc. 122, 12901 (2000) 191. S. Kunneke and A. Janshoff, Angew. Chem. -Int. Edit. 41, 314 (2001) 192. D. Grieshaber, R. MacKenzie, J. Voros and E. Reimhult, Sensors. 8, 1400 (2008) 193. E. Kalb, S. Frey and L. K. Tamm, Biochimica Et Biophysica Acta. 1103, 307 (1992) 194. C. A. Keller and B. Kasemo, Biophys. J. 75, 1397 (1998) 195. T. R. Khan, H. M. Grandin, A. Mashaghi, M. Textor, E. Reimhult and I. Reviakine, Biointerphases. 3, (2008) 196. R. Horvath, G. Fricsovszky and E. Papp, Biosens. Bioelectron. 18, 415 (2003) 197. V. J. Hruby and G. Tollin, Curr. Opin. Pharmacol. 7, 507 (2007) 198. A. B. Dahlin, M. P. Jonsson and F. Hook, Adv. Mater. 20, 1436 (2008) 199. A. Dahlin, M. Zach, T. Rindzevicius, M. Kall, D. S. Sutherland and F. Hook, J. Am. Chem. Soc. 127, 5043 (2005) 200. K. A. Burridge, M. A. Figa and J. Y. Wong, Langmuir. 20, 10252 (2004) 201. J. M. Moran-Mirabal, J. B. Edel, G. D. Meyer, D. Throckmorton, A. K. Singh and H. G. Craighead, Biophys. J. 89, 296 (2005) 202. S. Daniel, F. Albertorio and P. S. Cremer, MRS Bull. 31, 536 (2006) 203. D. K. Shenoy, W. R. Barger, A. Singh, R. G. Panchal, M. Misakian, V. M. Stanford and J. J. Kasianowicz, Nano Lett. 5, 1181 (2005)
Nanoscale Biosensors and Biochips
69
204. T. Blattler, C. Huwiler, M. Ochsner, B. Stadler, H. Solak, J. Voros and H. M. Grandin, J. Nanosci Nanotechnol. 6, 2237 (2006) 205. K. Morigaki, K. Kiyosue and T. Taguchi, Langmuir. 20, 7729 (2004) 206. F. F. Rossetti, M. Bally, R. Michel, M. Textor and I. Reviakine, Langmuir. 21, 6443 (2005) 207. N. B. Holland, Y. X. Qiu, M. Ruegsegger and R. E. Marchant, Nature. 392, 799 (1998) 208. E. Osterberg, K. Bergstrom, K. Holmberg, T. P. Schuman, J. A. Riggs, N. L. Burns, J. M. Vanalstine and J. M. Harris, J. Biomed. Mater. Res. 29, 741 (1995) 209. H. J. Griesser, P. G. Hartley, S. L. McArthur, K. M. McLean, L. Meagher and H. Thissen, Smart Mater. Struct. 11, 652 (2002) 210. M. Amiji and K. Park, Biomaterials. 13, 682 (1992) 211. J. L. Dalsin, L. J. Lin, S. Tosatti, J. Voros, M. Textor and P. B. Messersmith, Langmuir. 21, 640 (2005) 212. L. M. Feller, S. Cerritelli, M. Textor, J. A. Hubbell and S. G. P. Tosatti, Macromolecules. 38, 10503 (2005) 213. P. Kingshott and H. J. Griesser, Curr. Opin. Solid State Mat. Sci. 4, 403 (1999) 214. E. Ostuni, R. G. Chapman, R. E. Holmlin, S. Takayama and G. M. Whitesides, Langmuir. 17, 5605 (2001) 215. S. Pasche, S. M. De Paul, J. Voros, N. D. Spencer and M. Textor, Langmuir Langmuir. 19, 9216 (2003) 216. V. Zoulalian, S. Monge, S. Zurcher, M. Textor, J. J. Robin and S. Tosatti, Journal of Physical Chemistry B. 110, 25603 (2006) 217. H. H. Solak, J. Phys. D: Appl. Phys. 39, R171 (2006) 218. B. Michel, A. Bernard, A. Bietsch, E. Delamarche, M. Geissler, D. Juncker, H. Kind, J. P. Renault, H. Rothuizen, H. Schmid, P. Schmidt-Winkel, R. Stutz and H. Wolf, IBM Journal of Research and Development. 45, 697 (2001) 219. S. A. Ruiz and C. S. Chen, Soft Matter. 3, 168 (2007) 220. L. J. Guo, J. Phys. D Appl. Phys. 37, R123 (2004) 221. H. Schift, J. Vac. Sci. Technol. B. 26, 458 (2008) 222. C. L. Haynes and R. P. Van Duyne, J. Phys. Chem. B. 105, 5599 (2001) 223. R. Glass, M. Moller and J. P. Spatz, Nanotechnology. 14, 1153 (2003) 224. M. G. Capeluto, G. Vaschenko, M. Grisham, M. C. Marconi, S. Luduena, L. Pietrasanta, Y. F. Lu, B. Parkinson, C. S. Menoni and J. J. Rocca, IEEE Trans. Nanotechnol. 5, 3 (2006) 225. P. W. Wachulak, M. G. Capeluto, M. C. Marconi, D. Patel, C. S. Menoni and J. J. Rocca, Journal of Vacuum Science and Technology B. 25, 2094 (2007)
70
Leifert et al.
226. M. G. Capeluto, P. Wachulak, M. C. Marconi, D. Patel, C. S. Menoni, J. J. Rocca, C. Iemmi, E. H. Anderson, W. Chao and D. T. Attwood, Microelectronic Engineering. 84, 721 (2007) 227. H. H. Solak, C. David, J. Gobrecht, L. Wang and F. Cerrina, Microelectron Eng. 61-2, 77 (2002) 228. B. Stadler, H. H. Solak, S. Frerker, K. Bonroy, F. Frederix, J. Voeroes and M. Grandin, Nanotechnology. 18, (2007) 229. F. Patolsky, B. P. Timko, G. F. Zheng and C. M. Lieber, MRS Bull. 32, 142 (2007) 230. F. Patolsky, G. F. Zheng and C. M. Lieber, Anal. Chem. 78, 4260 (2006) 231. J. V. Veetil and K. M. Ye, Biotechnol. Prog. 23, 517 (2007) 232. A. Kumar and G. M. Whitesides, Applied Physics Letters. 63, 2002 (1993) 233. Y. N. Xia and G. M. Whitesides, Annual Review of Materials Science. 28, 153 (1998) 234. W. Lee, S. S. Lim, B. K. Choi and J. W. Choi, J. Microbiol. Biotechnol. 16, 1216 (2006) 235. J. P. Renault, A. Bernard, A. Bietsch, B. Michel, H. R. Bosshard, E. Delamarche, M. Kreiter, B. Hecht and U. P. Wild, J. Phys. Chem. B. 107, 703 (2003) 236. D. Lehnert, B. Wehrle-Haller, C. David, U. Weiland, C. Ballestrem, B. A. Imhof and M. Bastmeyer, J. Cell Sci. 117, 41 (2004) 237. G. Csucs, T. Kunzler, K. Feldman, F. Robin and N. D. Spencer, Langmuir. 19, 6104 (2003) 238. H. Schmid and B. Michel, Macromolecules. 33, 3042 (2000) 239. D. I. Rozkiewicz, J. Gierlich, G. A. Burley, K. Gutsmiedl, T. Carell, B. J. Ravoo and D. N. Reinhoudt, ChemBioChem. 8, 1997 (2007) 240. H. Tan, S. Huang and K. L. Yang, Langmuir. 23, 8607 (2007) 241. S. A. Lange, V. Benes, D. P. Kern, J. K. H. Horber and A. Bernard, Anal. Chem. 76, 1641 (2004) 242. P. Kim, S. E. Lee, H. S. Jung, H. Y. Lee, T. Kawai and K. Y. Suh, Lab on a Chip. 6, 54 (2006) 243. E. B. Chakra, B. Hannes, G. Dilosquer, C. D. Mansfield and M. Cabrera, Rev. Sci. Instrum. 79, 9 (2008) 244. A. A. Yu, T. A. Savas, G. S. Taylor, A. Guiseppe-Elie, H. I. Smith and F. Stellacci, Nano Lett. 5, 1061 (2005) 245. O. Akbulut, J. M. Jung, R. D. Bennett, Y. Hu, H. T. Jung, R. E. Cohen, A. M. Mayes and F. Stellacci, Nano Letters. 7, 3493 (2007) 246. A. A. Yu, T. Savas, H. I. Smith and F. Stellacci, Abstracts of Papers of the American Chemical Society. 231, (2006) 247. A. A. Yu and F. Stellacci, Journal of Materials Chemistry. 16, 2868 (2006)
Nanoscale Biosensors and Biochips
71
248. S. Y. Chou, P. R. Krauss and P. J. Renstrom, Appl. Phys. Lett. 67, 3114 (1995) 249. S. Y. Chou, P. R. Krauss and P. J. Renstrom, Science. 272, 85 (1996) 250. S. Y. Chou and P. R. Krauss, Microelectron Eng. 35, 237 (1997) 251. E. Kim, Y. N. Xia, X. M. Zhao and G. M. Whitesides, Adv. Mater. 9, 651 (1997) 252. M. Colburn, T. Bailey, B. J. Choi, J. G. Ekerdt, S. V. Sreenivasan and C. G. Willson, Solid State Technol. 44, 67 (2001) 253. M. Colburn, S. Johnson, M. Stewart, S. Damle, B. Bailey, B. Choi, M. Wedlake, T. Michaelson, S. V. Sreenivasan, J. G. Ekerdt and C. G. Willson, Proc. SPIE-Int. Soc. Opt. Eng. 3676, 379 (1999) 254. J. D. Hoff, L. J. Cheng, E. Meyhofer, L. J. Guo and A. J. Hunt, Nano Lett. 4, 853 (2004) 255. T. Ohtake, K. Nakamatsu, S. Matsui, H. Tabata and T. Kawai, J. Vac. Sci. Technol. B. 22, 3275 (2004) 256. H. E. Gaubert and W. Frey, Nanotechnology. 18, 7 (2007) 257. D. Falconnet, D. Pasqui, S. Park, R. Eckert, H. Schift, J. Gobrecht, R. Barbucci and M. Textor, Nano Lett. 4, 1909 (2004) 258. P. Maury, M. Escalante, M. Peter, D. N. Reinhoudt, V. Subramaniam and J. Huskens, Small. 3, 1584 (2007) 259. A. P. Zhu, M. B. Chan-Park and J. X. Gao, J. Biomed. Mater. Res. Part B. 76B, 76 (2006) 260. M. B. Chan-Park, Y. H. Yan, W. K. Neo, W. X. Zhou, J. Zhang and C. Y. Yue, Langmuir. 19, 4371 (2003) 261. M. C. Lensen, P. Mela, A. Mourran, J. Groll, J. Heuts, H. T. Rong and M. Moller, Langmuir. 23, 7841 (2007) 262. B. K. Lee, H. Y. Lee, P. Kim, K. Y. Suh, J. H. Seo, H. J. Cha and T. Kawai, Small. 4, 342 (2008) 263. U. C. Fischer and H. P. Zingsheim, Journal of Vacuum Science and Technology. 19, 881 (1981) 264. H. W. Deckman and J. H. Dunsmuir, Applied Physics Letters. 41, 377 (1982) 265. J. C. Garno, N. A. Amro, K. Wadu-Mesthrige and G. Y. Liu, Langmuir. 18, 8186 (2002) 266. J. C. Hulteen, D. A. Treichel, M. T. Smith, M. L. Duval, T. R. Jensen and R. P. Van Duyne, J. Phys. Chem. B. 103, 3854 (1999) 267. J. C. Hulteen and R. P. Van Duyne, Journal of Vacuum Science and Technology A: Vacuum, Surfaces, and Films. 13, 1553 (1995)
72
Leifert et al.
268. N. D. Denkov, O. D. Velev, P. A. Kralchevsky, I. B. Ivanov, H. Yoshimura and K. Nagayama, Nature. 361, 26 (1993) 269. N. D. Denkov, O. D. Velev, P. A. Kralchevsky, I. B. Ivanov, H. Yoshimura and K. Nagayama, Langmuir. 8, 3183 (1992) 270. F. Burmeister, W. Badowsky, T. Braun, S. Wieprich, J. Boneberg and P. Leiderer, Appl. Surf. Sci. 145, 461 (1999) 271. F. Burmeister, C. Schafle, T. Matthes, M. Bohmisch, J. Boneberg and P. Leiderer, Langmuir. 13, 2983 (1997) 272. A. D. Ormonde, D. Erin, M. Hicks, J. Castillo and R. P. van Dyne, Langmuir. 20, 6927 (2004) 273. X. D. Wang, C. J. Summers and Z. L. Wang, Nano Lett. 4, 423 (2004) 274. Y. J. Sun and G. C. Walker, J. Phys. Chem. B. 106, 2217 (2002) 275. T. M. Blattler, A. Binkert, M. Zimmermann, M. Textor, J. Voros and E. Reimhult, Nanotechnology. 19, 10 (2008) 276. N. Nath and A. Chilkoti, Analytical Chemistry. 74, 504 (2002) 277. L. M. Liz-Marzan, Langmuir. 22, 32 (2006) 278. A. D. McFarland and R. P. Van Duyne, Nano Letters. 3, 1057 (2003) 279. A. J. Haes, L. Chang, W. L. Klein and R. P. Van Duyne, J. Am. Chem. Soc. 127, 2264 (2005) 280. A. J. Haes, W. P. Hall, L. Chang, W. L. Klein and R. P. Van Duyne, Nano Lett. 4, 1029 (2004) 281. A. J. Haes and R. P. Van Duyne, J. Am. Chem. Soc. 124, 10596 (2002) 282. A. J. Haes, S. Zou, G. C. Schatz and R. P. Van Duyne, J. Phys. Chem. B. 108, 6961 (2004) 283. F. Frederix, K. Bonroy, W. Laureyn, G. Reekmans, A. Campitelli, W. Dehaen and G. Maes, Langmuir. 19, 4351 (2003) 284. F. Frederix, J. M. Friedt, K. H. Choi, W. Laureyn, A. Campitelli, D. Mondelaers, G. Maes and G. Borghs, Anal. Chem. 75, 6894 (2003) 285. M. P. Jonsson, P. Jonsson, A. B. Dahlin and F. Hook, Nano Letters. 7, 3462 (2007) 286. T. R. Groves, D. Pickard, B. Rafferty, N. Crosland, D. Adam and G. Schubert, Microelectronic Engineering. 61-2, 285 (2002) 287. A. A. Tseng, Small. 1, 924 (2005) 288. D. S. Ginger, H. Zhang and C. A. Mirkin, Angewandte Chemie-International Edition. 43, 30 (2004) 289. K. Salaita, Y. H. Wang and C. A. Mirkin, Nat. Nanotechnol. 2, 145 (2007) 290. J. Lussi, C. Tang, P. Kuenzi, U. Staufer, G. Csucs, J. Voeroes, G. Danuser, J. A. Hubbel and M. Textor, Nanotechnology. 16, 1781 (2005) 291. G. Zhang, T. T., T. Zako, H. T., M. T., Y. Kanari, T. Funatsu and I. Ohdomari, Small. 1, 833 (2005)
Nanoscale Biosensors and Biochips 292. 293. 294. 295. 296. 297. 298. 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. 310. 311. 312. 313. 314. 315. 316. 317. 318.
73
J. Jiang, X. M. Li, W. C. Mak and D. Trau, Advanced Materials. 20, 1636 (2008) Z. Y. Suo, R. Avci, X. H. Yang and D. W. Pascual, Langmuir. 24, 4161 (2008) R. D. Piner, J. Zhu, F. Xu, S. H. Hong and C. A. Mirkin, Science. 283, 661 (1999) L. M. Demers, D. S. Ginger, S. J. Park, Z. Li, S. W. Chung and C. A. Mirkin, Science. 296, 1836 (2002) S. W. Chung, D. S. Ginger, M. W. Morales, Z. F. Zhang, V. Chandrasekhar, M. A. Ratner and C. A. Mirkin, Small. 1, 64 (2005) Y. Cho and A. Ivanisevic, J. Phys. Chem. B. 108, 15223 (2004) H. Z. Jiang and S. I. Stupp, Langmuir. 21, 5242 (2005) M. Lee, D. K. Kang, H. K. Yang, K. H. Park, S. Y. Choe, C. Kang, S. I. Chang, M. H. Han and I. C. Kang, Proteomics. 6, 1094 (2006) K. B. Lee, E. Y. Kim, C. A. Mirkin and S. M. Wolinsky, Nano Lett. 4, 1869 (2004) S. Lenhert, P. Sun, Y. H. Wang, H. Fuchs and C. A. Mirkin, Small. 3, 71 (2007) J. C. Smith, K. B. Lee, Q. Wang, M. G. Finn, J. E. Johnson, M. Mrksich and C. A. Mirkin, Nano Lett. 3, 883 (2003) C. L. Cheung, J. A. Camarero, B. W. Woods, T. W. Lin, J. E. Johnson and J. J. De Yoreo, J. Am. Chem. Soc. 125, 6848 (2003) R. A. Vega, C. K. F. Shen, D. Maspoch, J. G. Robach, R. A. Lamb and C. A. Mirkin, Small. 3, 1482 (2007) S. Rozhok, C. K. F. Shen, P. L. H. Littler, Z. F. Fan, C. Liu, C. A. Mirkin and R. C. Holz, Small. 1, 445 (2005) P. Xu and D. L. Kaplan, Adv. Mater. 16, 628 (2004) P. Xu, H. Uyama, J. E. Whitten, S. Kobayashi and D. L. Kaplan, J. Am. Chem. Soc. 127, 11745 (2005) J. Hyun, J. Kim, S. L. Craig and A. Chilkoti, J. Am. Chem. Soc. 126, 4770 (2004) D. Bullen, S. W. Chung, X. F. Wang, J. Zou, C. A. Mirkin and C. Liu, Appl. Phys. Lett. 84, 789 (2004) D. Bullen and C. Liu, Sens. Actuator A-Phys. 125, 504 (2006) X. F. Wang, D. A. Bullen, J. Zou, C. Liu and C. A. Mirkin, J. Vac. Sci. Technol. B. 22, 2563 (2004) K. Salaita, Y. H. Wang, J. Fragala, R. A. Vega, C. Liu and C. A. Mirkin, Angew. Chem. -Int. Edit. 45, 7220 (2006) L. F. Kang, B. G. Chung, R. Langer and A. Khademhosseini, Drug Discovery Today. 13, 1 (2008) P. S. Dittrich and A. Manz, Nature Reviews Drug Discovery. 5, 210 (2006) E. Delamarche, D. Juncker and H. Schmid, Advanced Materials. 17, 2911 (2005) J. Pihl, M. Karlsson and D. T. Chiu, Drug Discovery Today. 10, 1377 (2005) J. Khandurina and A. Guttman, Journal of Chromatography A. 943, 159 (2002) S. Bellucci, Phys. Stat. Sol. (c). 2, 34–47 (2005)
74
Leifert et al.
319. K. S. Coleman, Annu. Rep. Prog. Chem., Sect. A: Inorg. Chem. 104, 379 (2008) 320. M. Terrones, J. M. Romo-Herrera, E. Cruz-Silva, F. López-Urías, E. MuñozSandoval, J. J. Velázquez-Salazar, H. Terrones, Y. Bando and D. Golberg, Materials Today. 10, 30 (2007) 321. D. V. Bavykin, J. M. Friedrich and F. C. Walsh, Adv. Mater. 18, 2807 (2006) 322. B. He, S. J. Son and S. B. Lee, Langmuir. 22, 8263 (2006) 323. M. Yemini, M. Reches, J. Rishpon and E. Gazit, Nano Lett. 5, 183 (2005) 324. J. Wang and Y. Lin, TrAC Trends in Analytical Chemistry. 27, 619 (2008) 325. W. C. Poh, K. P. Loh, W. De Zhang, S. Triparthy, J. S. Ye and F. S. Sheu, Langmuir. 20, 5484 (2004) 326. S. Liu and A. Chen, Langmuir. 21, 8409 (2005) 327. J. Liu, A. G. Rinzler, H. Dai, J. H. Hafner, R. K. Bradley, P. J. Boul, A. Lu, T. Iverson, K. Shelimov, C. B. Huffman, F. Rodriguez-Macias, Y. Shon, T. R. Lee, D. T. Colbert and R. E. Smalley, Science. 280, 1253 (1998) 328. Y. Zhou, H. Yang and H. Y. Chen, Talanta. 76, 419 (2008) 329. V. B. Kandimalla, V. S. Tripathi and H. Ju, Biomaterials. 27, 1167 (2006) 330. J. J. Davis, R. J. Coles, H. Allen and O. Hill, Journal of Electroanalytical Chemistry. 440, 279 (1997) 331. J. Wang, M. Li, Z. Shi, N. Li and Z. Gu, Anal. Chem. 74, 1993 (2002) 332. G. Wang, J. Xu and H. Chen, Electrochemistry Communications. 4, 506 (2002) 333. K. Yamamoto, G. Shi, T. Zhou, F. Xu, J. Xu, T. Kato, J. Jin and L. Jin, Analyst. 128, 249 (2003) 334. Y. Zhao, W. Zhang, H. Chen and Q. Luo, Analytical Sciences. 18, 939 (2002) 335. A. Heller, Acc. Chem. Res. 23, 128 (1990) 336. H. J. Hecht, D. Schomburg, H. Kalisz and R. D. Schmid, Biosensors and Bioelectronics. 8, 197 (1993) 337. X. Chen, J. Chen, C. Deng, C. Xiao, Y. Yang, Z. Nie and S. Yao, Talanta. 76, 763 (2008) 338. N. Liu, X. Cai, Y. Lei, Q. Zhang, M. B. Chan-Park, C. Li, W. Chen and A. Mulchandani, Electroanalysis. 19, 616 (2007) 339. V. B. Kandimalla and H. Ju, Chemistry. 12, 1074 (2006) 340. D. R. Jeykumari and S. S. Narayanan, Biosens. Bioelectron. 23, 1686 (2008) 341. J. J. Gooding, R. Wibowo, J. Q. Liu, W. Yang, D. Losic, S. Orbons, F. J. Mearns, J. G. Shapter and D. B. Hibbert, J. Am. Chem. Soc. 125, 9006 (2003) 342. M. L. Terranova, V. Sessa and M. Rossi, Chemical Vapour Deposition. 12, 315 (2006) 343. C. K. Tan, K. P. Loh and T. T. John, Analyst. 133, 448 (2008)
Nanoscale Biosensors and Biochips
75
344. F. Patolsky, Y. Weizmann and I. Willner, Angew. Chem. Int. Ed Engl. 43, 2113 (2004) 345. X. Yu, D. Chattopadhyay, I. Galeska, F. Papadimitrakopoulos and J. F. Rusling, Electrochemistry Communications. 5, 408 (2003) 346. D. Zanchet, H. Tolentino, M. C. Martins Alves, O. L. Alves and D. Ugarte, Chemical Physics Letters. 323, 167 (2000) 347. E. Delamarche, B. Michel, H. Kang and C. Gerber, Langmuir. 10, 4103 (1994) 348. A. B. Horn, D. A. Russell, L. J. Shorthouse and T. R. E. Simpson, J. Chem. Soc., Faraday Trans.,. 92, 4759 (1996) 349. M. H. Schoenfisch and J. E. Pemberton, J. Am. Chem. Soc. 120, 4502 (1998) 350. J. Yu, J. G. Shapter, J. S. Quinton, M. R. Johnston and D. A. Beattie, Phys. Chem. Chem. Phys. 9, 510 (2007) 351. J. Yu, D. Losic, M. Marshall, T. Böcking, J. J. Gooding and J. G. Shapter, Soft Matter. 2, 1081 (2006) 352. J. Yu, S. Mathew, B. S. Flavel, M. R. Johnston and J. G. Shapter, J. Am. Chem. Soc. 130, 8788 (2008) 353. J. Yu, J. G. Shapter, M. R. Johnston, J. S. Quinton and J. J. Gooding, Electrochimica Acta. 52, 6206 (2007) 354. M. D. Marazuela and M. C. Moreno-Bondi, Analytical and Bioanalytical Chemistry. 372, 664 (2002) 355. J. Homola, S. S. Yee and G. Gauglitz, Sensors and Actuators B-Chemical. 54, 3 (1999) 356. R. L. Bunde, E. J. Jarvi and J. J. Rosentreter, Talanta. 46, 1223 (1998) 357. D. R. Thevenot, K. Toth, R. A. Durst and G. S. Wilson, Biosens. Bioelectron. 16, 121 (2001) 358. N. Jaffrezic-Renault and S. V. Dzyadevych, Sensors. 8, 2569 (2008) 359. K. Ramanathan and B. Danielsson, Biosens. Bioelectron. 16, 417 (2001) 360. E. Katz and I. Willner, Angewandte Chemie-International Edition. 43, 6042 (2004) 361. C. M. Niemeyer, Angewandte Chemie-International Edition. 40, 4128 (2001) 362. S. G. Penn, L. He and M. J. Natan, Current Opinion in Chemical Biology. 7, 609 (2003) 363. M. Seydack, Biosensors and Bioelectronics. 20, 2454 (2005) 364. C. S. Thaxton, D. G. Georganopoulou and C. A. Mirkin, Clinica Chimica Acta. 363, 120 (2006) 365. J. Wang, Analytica Chimica Acta. 500, 247 (2003) 366. I. Willner, B. Willner and E. Katz, Bioelectrochemistry. 70, 2 (2007) 367. P. Alivisatos, Nature Biotechnology. 22, 47 (2004) 368. K. A. Edwards and A. J. Baeumner, Talanta. 68, 1421 (2006)
76
Leifert et al.
369. M. Bally, M. Halter, J. Voros and H. M. Grandin, Surface and Interface Analysis. 38, 1442 (2006) 370. R. Y. Tsien, Annu. Rev. Biochem. 67, 509 (1998) 371. Y. A. Labas, N. G. Gurskaya, Y. G. Yanushevich, A. F. Fradkov, K. A. Lukyanov, S. A. Lukyanov and M. V. Matz, Proc. Natl. Acad. Sci. U. S. A. 99, 4256 (2002) 372. V. V. Verkhusha and K. A. Lukyanov, Nat. Biotechnol. 22, 289 (2004) 373. N. G. Gurskaya, A. F. Fradkov, A. Terskikh, M. V. Matz, Y. A. Labas, V. I. Martynov, Y. G. Yanushevich, K. A. Lukyanov and S. A. Lukyanov, FEBS Lett. 507, 16 (2001) 374. R. Ando, H. Hama, M. Yamamoto-Hino, H. Mizuno and A. Miyawaki, Proc. Natl. Acad. Sci. U. S. A. 99, 12651 (2002) 375. J. S. Marchant, G. E. Stutzmann, M. A. Leissring, F. M. LaFerla and I. Parker, Nat. Biotechnol. 19, 645 (2001) 376. D. M. Chudakov, V. V. Belousov, A. G. Zaraisky, V. V. Novoselov, D. B. Staroverov, D. B. Zorov, S. Lukyanov and K. A. Lukyanov, Nat. Biotechnol. 21, 191 (2003) 377. D. M. Chudakov, A. V. Feofanov, N. N. Mudrik, S. Lukyanov and K. A. Lukyanov, J. Biol. Chem. 278, 7215 (2003) 378. S. R. Adams, R. E. Campbell, L. A. Gross, B. R. Martin, G. K. Walkup, Y. Yao, J. Llopis and R. Y. Tsien, J. Am. Chem. Soc. 124, 6063 (2002) 379. B. N. Giepmans, S. R. Adams, M. H. Ellisman and R. Y. Tsien, Science. 312, 217 (2006) 380. C. Hoffmann, G. Gaietta, M. Bunemann, S. R. Adams, S. Oberdorff-Maass, B. Behr, J. P. Vilardaga, R. Y. Tsien, M. H. Ellisman and M. J. Lohse, Nat. Methods. 2, 171 (2005) 381. P. R. Selvin, Annu. Rev. Biophys. Biomol. Struct. 31, 275 (2002) 382. K. J. Franz, M. Nitz and B. Imperiali, Chembiochem. 4, 265 (2003) 383. X. Fan, I. M. White, S. I. Shopova, H. Zhu, J. D. Suter and Y. Sun, Analytica Chimica Acta. 620, 8 (2008) 384. M. C. Daniel and D. Astruc, Chemical Reviews. 104, 293 (2004) 385. Z. F. Ma and S. F. Sui, Angewandte Chemie-International Edition. 41, 2176 (2002) 386. K. Glynou, P. C. Ioannou, T. K. Christopoulos and V. Syriopoulou, Analytical Chemistry. 75, 4155 (2003) 387. T. A. Taton, C. A. Mirkin and R. L. Letsinger, Science. 289, 1757 (2000) 388. J. J. Storhoff, S. S. Marla, P. Bao, S. Hagenow, H. Mehta, A. Lucas, V. Garimella, T. Patno, W. Buckingham, W. Cork and U. R. Müller, Biosensors and Bioelectronics. 19, 875 (2004) 389. J. M. Nam, C. S. Thaxton and C. A. Mirkin, Science. 301, 1884 (2003)
Nanoscale Biosensors and Biochips
77
390. I. Alexandre, S. Hamels, S. Dufour, J. Collet, N. Zammatteo, F. De Longueville, J. L. Gala and J. Remacle, Analytical Biochemistry. 295, 1 (2001) 391. Y. P. Bao, M. Huber, T. F. Wei, S. S. Marla, J. J. Storhoff and U. R. Muller, Nucleic Acids Research. 33, (2005) 392. M. Huber, T. F. Wei, U. R. Muller, P. A. Lefebvre, S. S. Marla and Y. P. Bao, Nucleic Acids Research. 32, (2004) 393. D. G. Georganopoulou, L. Chang, J. M. Nam, C. S. Thaxton, E. J. Mufson, W. L. Klein and C. A. Mirkin, Proceedings of the National Academy of Sciences of the United States of America. 102, 2273 (2005) 394. J. Yguerabide and E. E. Yguerabide, Journal of Cellular Biochemistry. 71 (2001) 395. D. I. Stimpson, J. V. Hoijer, W. T. Hsieh, C. Jou, J. Gordon, T. Theriault, R. Gamble and J. D. Baldeschwieler, Proceedings of the National Academy of Sciences of the United States of America. 92, 6379 (1995) 396. P. Bao, A. G. Frutos, C. Greef, J. Lahiri, U. Muller, T. C. Peterson, L. Warden and X. Xie, Anal. Chem. 74, 1792 (2002) 397. P. Francois, M. Bento, P. Vaudaux and J. Schrenzel, Journal of Microbiological Methods. 55, 755 (2003) 398. S. J. Oldenburg, C. C. Genick, K. A. Clark and D. A. Schultz, Analytical Biochemistry. 309, 109 (2002) 399. T. A. Taton, G. Lu and C. A. Mirkin, Journal of the American Chemical Society. 123, 5164 (2001) 400. Y. C. Cao, R. C. Jin, J. M. Nam, C. S. Thaxton and C. A. Mirkin, Journal of the American Chemical Society. 125, 14676 (2003) 401. Y. W. C. Cao, R. C. Jin and C. A. Mirkin, Science. 297, 1536 (2002) 402. J. Ni, R. J. Lipert, G. B. Dawson and M. D. Porter, Analytical Chemistry. 71, 4903 (1999) 403. D. S. Grubisha, R. J. Lipert, H. Y. Park, J. Driskell and M. D. Porter, Analytical Chemistry. 75, 5936 (2003) 404. A. Zajac, D. S. Song, W. Qian and T. Zhukov, Colloids and Surfaces BBiointerfaces. 58, 309 (2007) 405. D. Gerion, F. Q. Chen, B. Kannan, A. H. Fu, W. J. Parak, D. J. Chen, A. Majumdar and A. P. Alivisatos, Analytical Chemistry. 75, 4766 (2003) 406. M. Nichkova, D. Dosev, A. E. Davies, S. J. Gee, I. M. Kennedy and B. D. Hammock, Analytical Letters. 40, 1423 (2007) 407. E. R. Goldman, T. J. O'Shaughnessy, C. M. Soto, C. H. Patterson, C. R. Taitt, M. S. Spector and P. T. Charles, Analytical and Bioanalytical Chemistry. 380, 880 (2004) 408. M. Shingyoji, D. Gerion, D. Pinkel, J. W. Gray and F. Q. Chen, Talanta. 67, 472 (2005)
78
Leifert et al.
409. D. Gerion, W. J. Parak, S. C. Williams, D. Zanchet, C. M. Micheel and A. P. Alivisatos, Journal of the American Chemical Society. 124, 7070 (2002) 410. W. C. W. Chan, D. J. Maxwell, X. H. Gao, R. E. Bailey, M. Y. Han and S. M. Nie, Current Opinion in Biotechnology. 13, 40 (2002) 411. J. Riegler and T. Nann, Analytical and Bioanalytical Chemistry. 379, 913 (2004) 412. W. J. Parak, D. Gerion, T. Pellegrino, D. Zanchet, C. Micheel, S. C. Williams, R. Boudreau, M. A. Le Gros, C. A. Larabell and A. P. Alivisatos, Nanotechnology. 14, R15 (2003) 413. H. A. H. Rongen, A. Bult and W. P. vanBennekom, Journal of Immunological Methods. 204, 105 (1997) 414. S. Park and R. A. Durst, Analytical Biochemistry. 280, 151 (2000) 415. M. B. Esch, A. J. Baeumner and R. A. Durst, Analytical Chemistry. 73, 3162 (2001) 416. A. J. Baeumner, R. N. Cohen, V. Miksic and J. H. Min, Biosensors and Bioelectronics. 18, 405 (2003) 417. J. A. Ho, H. Hsu and M. Huang, Analytical Biochemistry. 330, 342 (2004) 418. J. A. A. Ho and R. D. Wauchope, Analytical Chemistry. 74, 1493 (2002) 419. S. Ahn-Yoon, T. R. DeCory and R. A. Durst, Analytical and Bioanalytical Chemistry. 378, 68 (2004) 420. H. A. Hartley and A. J. Baeumner, Analytical and Bioanalytical Chemistry. 376, 319 (2003) 421. A. J. Baeumner, B. Leonard, J. McElwee and R. A. Montagna, Analytical and Bioanalytical Chemistry. 380, 15 (2004) 422. J. T. Connelly, S. R. Nugen, W. Borejsza-Wysocki, R. A. Durst, R. A. Montagna and A. J. Baeumner, Analytical and Bioanalytical Chemistry. 391, 487 (2008) 423. H. W. Wen, W. Borejsza-Wysocki, T. DeCory and R. Durst, Analytical and Bioanalytical Chemistry. 382, 1217 (2005) 424. A. J. Baeumner, C. Jones, C. Y. Wong and A. Price, Analytical and Bioanalytical Chemistry. 378, 1587 (2004) 425. J. A. A. Ho and R. A. Durst, Analytica Chimica Acta. 414, 61 (2000) 426. J. A. Ho, S. Zeng, M. Huang and H. Kuo, Analytica Chimica Acta. 556, 127 (2006) 427. M. Lee, R. A. Durst and R. B. Wong, Talanta. 46, 851 (1998) 428. L. E. Locascio, J. S. Hong and M. Gaitan, Electrophoresis. 23, 799 (2002) 429. J. A. A. Ho, L. C. Wu, M. R. Huang, Y. J. Lin, A. J. Baeumner and R. A. Durst, Analytical Chemistry. 79, 246 (2007) 430. S. Kwakye and A. Baeumner, Analytical and Bioanalytical Chemistry. 376, 1062 (2003) 431. N. V. Zaytseva, V. N. Goral, R. A. Montagna and A. J. Baeumner, Lab on a Chip. 5, 805 (2005)
Nanoscale Biosensors and Biochips
79
432. N. V. Zaytseva, R. A. Montagna and A. J. Baeumner, Analytical Chemistry. 77, 7520 (2005) 433. K. Kerman, M. Kobayashi and E. Tamiya, Measurement Science and Technology. 15, R1 (2004) 434. S. J. Park, T. A. Taton and C. A. Mirkin, Science. 295, 1503 (2002) 435. O. D. Velev and E. W. Kaler, Langmuir. 15, 3693 (1999) 436. L. Moreno-Hagelsieb, P. E. Lobert, R. Pampin, D. Bourgeois, J. Remacle and D. Flandre, Sensors and Actuators B-Chemical. 98, 269 (2004) 437. M. Ozsoz, A. Erdem, K. Kerman, D. Ozkan, B. Tugrul, N. Topcuoglu, H. Ekren and M. Taylan, Analytical Chemistry. 75, 2181 (2003) 438. T. M. H. Lee, L. L. Li and I. M. Hsing, Langmuir. 19, 4338 (2003) 439. J. Wang, D. K. Xu and R. Polsky, Journal of the American Chemical Society. 124, 4208 (2002) 440. L. Authier, C. Grossiord, P. Brossier and B. Limoges, Analytical Chemistry. 73, 4450 (2001) 441. M. Dequaire, C. Degrand and B. Limoges, Analytical Chemistry. 72, 5521 (2000) 442. H. Cai, Y. Xu, N. N. Zhu, P. G. He and Y. Z. Fang, Analyst. 127, 803 (2002) 443. J. Wang, G. D. Liu, M. R. Jan and Q. Y. Zhu, Electrochemistry Communications. 5, 1000 (2003) 444. W. Sun, J. Zhong, B. Zhang and K. Jiao, Analytical and Bioanalytical Chemistry. 389, 2179 (2007) 445. W. Sun, J. H. Zhong, P. Qin and K. Jiao, Analytical Biochemistry. 377, 115 (2008) 446. N. N. Zhu, A. P. Zhang, P. G. He and Y. Z. Fang, Analyst. 128, 260 (2003) 447. N. N. Zhu, A. P. Zhang, Q. J. Wang, P. G. He and Y. Z. Fang, Electroanalysis. 16, 577 (2004) 448. M. Rochelet-Dequaire, B. Limoges and P. Brossier, Analyst. 131, 923 (2006) 449. Q. Pan, R. Y. Zhang, Y. F. Bai, N. Y. He and Z. H. Lu, Analytical Biochemistry. 375, 179 (2008) 450. P. Rijiravanich, M. Somasundrum and W. Surareungchai, Anal. Chem. 80, 3904 (2008) 451. J. Wang, G. D. Liu and Q. Y. Zhu, Analytical Chemistry. 75, 6218 (2003) 452. A. N. Kawde and J. Wang, Electroanalysis. 16, 101 (2004) 453. J. Wang, G. Liu and A. Merkoci, J. Am. Chem. Soc. 125, 3214 (2003) 454. G. D. Liu, T. M. H. Lee and J. S. Wang, Journal of the American Chemical Society. 127, 38 (2005) 455. J. A. Hansen, R. Mukhopadhyay, J. O. Hansen and K. V. Gothelf, Journal of the American Chemical Society. 128, 3860 (2006) 456. V. N. Goral, N. V. Zaytseva and A. J. Baeumner, Lab on a Chip. 6, 414 (2006)
80
Leifert et al.
457. S. Kwakye, V. N. Goral and A. J. Baeumner, Biosensors and Bioelectronics. 21, 2217 (2006) 458. S. Viswanathan, L. C. Wu, M. R. Huang and J. A. A. Ho, Analytical Chemistry. 78, 1115 (2006) 459. F. Patolsky, A. Lichtenstein and I. Willner, Angewandte Chemie-International Edition. 39, 940 (2000) 460. F. Patolsky, A. Lichtenstein and I. Willner, Journal of the American Chemical Society. 123, 5194 (2001) 461. J. Wang, Electroanalysis. 13, 983 (2001) 462. J. Y. Jia, W. J. Guan, M. H. Sim, Y. Q. Li and H. Li, Sensors. 8, 1712 (2008) 463. J. C. Pickup, F. Hussain, N. D. Evans, O. J. Rolinski and D. J. S. Birch, Biosens. Bioelectron. 20, 2555 (2005) 464. X. D. Wang, T. Y. Zhou, X. Chen, K. Y. Wong and X. R. Wang, Sensors and Actuators B-Chemical. 129, 866 (2008) 465. S. K. Jha, A. Topkar and S. F. D'Souza, J. Biochem. Biophys. Methods. 70, 1145 (2008) 466. Y. Zhang, G. M. Wen, Y. H. Zhou, S. M. Shuang, C. Dong and M. M. F. Choi, Biosens. Bioelectron. 22, 1791 (2007) 467. A. Ramanavicius, Analytical and Bioanalytical Chemistry. 387, 1899 (2007) 468. S. K. Arya, M. Datta and B. D. Malhotra, Biosens. Bioelectron. 23, 1083 (2008) 469. M. G. Zhang, C. Mullens and W. Gorski, Anal. Chem. 77, 6396 (2005) 470. L. Zhang, G. C. Zhao, X. W. Wei and Z. S. Yang, Electroanalysis. 17, 630 (2005) 471. E. Casero, M. Darder, F. Pariente and E. Lorenzo, Anal. Chim. Acta. 403, 1 (2000) 472. D. Ashe, T. Alleyne and E. Iwuoha, Biotechnol. Appl. Biochem. 46, 185 (2007) 473. M. I. Prodromidis and M. I. Karayannis, Electroanalysis. 14, 241 (2002) 474. M. Gamella, S. Campuzano, A. J. Reviejo and J. M. Pingarron, Anal. Chim. Acta. 609, 201 (2008) 475. C. Cordella, J. F. Antinelli and D. Cabrol-Bass, TrAC, Trends Anal. Chem. 22, 115 (2003) 476. B. Prieto-Simon, T. Noguer and M. Campas, TrAC, Trends Anal. Chem. 26, 689 (2007) 477. B. Prieto-Simon, M. Campas, J. L. Marty and T. Noguer, Biosens. Bioelectron. 23, 995 (2008) 478. R. G. Smith, N. D'Souza and S. Nicklin, Analyst. 133, 571 (2008) 479. X. Y. Guan, L. Q. Gu, S. Cheley, O. Braha and H. Bayley, Chembiochem. 6, 1875 (2005) 480. A. W. Czarnik, Nature. 394, 417 (1998) 481. S. P. Song, L. H. Wang, J. Li, J. L. Zhao and C. H. Fan, TrAC, Trends Anal. Chem. 27, 108 (2008)
Nanoscale Biosensors and Biochips
81
482. P. Nednoor, V. G. Gavalas, N. Chopra, B. J. Hinds and L. G. Bachas, Journal of Materials Chemistry. 17, 1755 (2007) 483. M. Majumder, X. Zhan, R. Andrews and B. J. Hinds, Langmuir. 23, 8624 (2007) 484. A. J. Russell, J. A. Berberich, G. E. Drevon and R. R. Koepsel, Annu. Rev. Biomed. Eng. 5, 1 (2003) 485. B. Huelseweh, R. Ehricht and H. J. Marschall, Proteomics. 6, 2972 (2006) 486. K. A. Donaldson, M. F. Kramer and D. V. Lim, Biosens. Bioelectron. 20, 322 (2004) 487. A. Petrosova, T. Konry, S. Cosnier, I. Trakht, J. Lutwama, E. Rwaguma, A. Chepurnov, E. Muhlberger, L. Lobel and R. S. Marks, Sensors and Actuators B-Chemical. 122, 578 (2007) 488. K. M. Halverson, R. G. Panchal, T. L. Nguyen, R. Gussio, S. F. Little, M. Misakian, S. Bavari and J. J. Kasianowicz, J. Biol. Chem. 280, 34056 (2005) 489. B. N. Feltis, B. A. Sexton, F. L. Glenn, M. J. Best, M. Wilkins and T. J. Davis, Biosens. Bioelectron. 23, 1131 (2008) 490. A. K. Wanekaya, W. Chen and A. Mulchandani, Journal of Environmental Monitoring. 10, 703 (2008) 491. A. Amine, H. Mohammadi, I. Bourais and G. Palleschi, Biosens. Bioelectron. 21, 1405 (2006) 492. S. Ramanathan, M. Ensor and S. Daunert, Trends Biotechnol. 15, 500 (1997) 493. N. Verma and M. Singh, Biometals. 18, 121 (2005) 494. J. J. Gooding, E. Chow and R. Finlayson, Aust. J. Chem. 56, 159 (2003) 495. D. A. Blake, R. M. Jones, R. C. Blake, A. R. Pavlov, I. A. Darwish and H. N. Yu, Biosens. Bioelectron. 16, 799 (2001) 496. P. A. Behnisch, K. Hosoe and S. Sakai, Environ. Int. 27, 413 (2001) 497. M. Trojanowicz, Electroanalysis. 14, 1311 (2002) 498. S. Rodriguez-Mozaz, M. P. Marco, M. J. L. de Alda and D. Barcelo, Analytical and Bioanalytical Chemistry. 378, 588 (2004) 499. A. Rasooly and K. E. Herold, J. AOAC Int. 89, 873 (2006) 500. S. Rodriguez-Mozaz, M. J. L. de Alda, M. P. Marco and D. Barcelo, Talanta. 65, 291 (2005) 501. C. N. Pope, Journal of Toxicology and Environmental Health-Part B-Critical Reviews. 2, 161 (1999) 502. G. Istamboulie, S. Andreescu, J. L. Marty and T. Noguer, Biosens. Bioelectron. 23, 506 (2007) 503. A. Hildebrandt, J. Ribas, R. Bragos, J. L. Marty, M. Tresanchez and S. Lacorte, Talanta. 75, 1208 (2008) 504. A. Hildebrandt, R. Bragós, S. Lacorte and J. L. Marty, Sensors and Actuators B: Chemical. 133, 195 (2008)
82
Leifert et al.
505. V. Vamvakaki and N. A. Chaniotakis, Biosens. Bioelectron. 22, 2848 (2007) 506. C. Karnati, H. W. Du, H. F. Ji, X. H. Xu, Y. Lvov, A. Mulchandani, P. Mulchandani and W. Chen, Biosens. Bioelectron. 22, 2636 (2007)
CHAPTER 2 SURFACE MODIFICATIONS AND APPLICATIONS OF MAGNETIC AND SELECTIVE NONMAGNETIC NANOPARTICLES
Rui Shen and Hong Yang* Department of Chemical Engineering, University of Rochester, Gavett Hall 206, Rochester, New York 14627-0166, USA *Corresponding author, Email:
[email protected]
1. Introduction Magnetic nanoparticles (NPs) are an important class of nanomaterials because of their applications in various areas 1. They have been investigated for biological labeling, magnetic resonance imaging (MRI), targeted drug carriers for cancer therapy, sensing, magnetic separation, and ferrofluids in heat transfer, dampers and actuators 2-5. The response of magnetic nanoparticles to an external field is an appealing feature for their uses in drug delivery and biological separation including cell sorting. Magnetic particles exposed in alternating magnetic fields can produce thermal energy and enables the effective hyperthermia therapy 5. The spin carried in the magnetic particles can response to the microenvironments around and be exploited in MRI. The ease to control the property of magnetic nanoparticles is important features for their widely spread uses. Common bulk magnetic materials can be classified into diamagnetism, paramagnetism, ferromagnetism, ferrimagnetism and antiferromagnetism based on the arrangement, orientation, and strength of magnetic dipole moments. The most useful magnetic nanoparticles are made of ferro- or ferri-magnetic materials and often have the superparamagnetic properties with blocking temperatures typically below 83
84
Shen and Yang
zero degree Celsius 1. Thus, solution phase synthetic methods can typically be applied to the magnetic materials without the concern of extensive uncontrollable aggregations among the particles because of the interaction of permanent magnetic dipoles. These wet chemistry approaches include microemulsion, hydrothermal, solvothermal, and nonhydrolytic methods 1. The classes of magnetic materials that can be produced in nanostructured forms range from metals, to metal alloys, to oxides and ferrites. The size, shape and size distribution of a few magnetic nanoparticle can be controlled precisely at the nanometer scale. To explore the potentials of high quality magnetic nanoparticles, an important requirement is the fine and sometimes precise control of surface chemistry. Surface modification can be used to prevent aggregation, improve stability in suspension, and enhance compatibility of nanoparticles with solid matrices or biological environments. It also provides the means for functional groups used for further grafting or conjugation of additional functional groups or molecules. For clinical applications magnetic nanoparticles should be able to disperse in water and other hydrophilic media, biocompatible and selective to specific targets 1, 6. Magnetic nanoparticles with toxic elements such as cobalt and nickel need to be shielded from the biological media with biocompatible layers. As particles have higher surface energy than bulk materials, magnetic nanoparticles of metals or metal alloys can be oxidized easily even under mild temperatures. Silica and other coatings can alter the oxidation profiles by preventing or slowing down the diffusion of oxygen to magnetic cores 7, 8. Furthermore, various monodisperse magnetic nanoparticles developed in nonhydrolytic solvent systems are capped with surfactant and highly hydrophobic 6. These particles cannot be readily applicable in biological systems which are aqueous media. Stabilization of magnetic particles in water and other polar solvents is a critical issue in various applications of ferrofluids as well. For magnetic nanoparticles such as the widelystudied FePt alloys, inert coating can prevent the aggregation during the post-synthetic heat treatment process so that small size with high coercivity and preferred crystal phase may be obtained 9. A range of surface modification methods have been developed in order to improve the surface characteristics of magnetic nanoparticles
Surface Modifications of Nanoparticles
85
and to achieve targeted specific applications. When ten-nanometer-sized immunoglobulin M (IgM) molecules are used in cell sorting, the chosen nanoparticles should be in the similar size range in order to have maximum valence for the specific binding and small enough to evade biological cleaning process (Figure 1) 10. Besides size, nanoparticles also have to possess an hydrophobic layer for high stability, a hydrophilic layer for biocompatibility and an outmost functional layer for cell recognition, as shown in Figure 1c 11.
Figure 1. Schematic illustrations of (a) an 8-10 nm finite-sized (IgM)5 molecule 10, (b) the height, h of ligands and the core diameter, D of a bi-functional magnetic nanoparticle (BMNP) 10 (Reprinted with permission from Reference 10, © 2006, Royal Society of Chemistry) and (c) the concept of creating multifunctional nanoparticles through surface modifications 11. (Reprinted with permission from Reference 11, © 2007, Academic Press Inc, Elsevier Science.)
Surface composition and structure are critically important in the design and synthesis of surface coating layers. A variety of approaches have been developed for the surface modifications of magnetic nanoparticles. They include covalent attachment and adsorption of either small molecules or polymers, resulting in core-shell structures based on polymerization of organic monomers or polycondensation of inorganic precursors through Stöber and other sol-gel methods 12-15. In this review, we first survey the methods for surface modifications developed for both magnetic and nonmagnetic nanoparticles. As surface structure is one of the most important factors in determining the
86
Shen and Yang
feasibility of a coating approach, methods developed for nonmagnetic nanoparticles of metals, alloys and oxides can often be applicable to magnetic counterparts that share similar surface chemistry, or vice versa. Effects of solvent, surface capping agent and shape on the strategy of surface modifications are covered. After the discussion on general approaches, we discuss the surface modifications of various classes of magnetic nanoparticles of metals, alloys and metal oxides. The methods developed for magnetic nanostructures with different shapes are presented to highlight the broad applications of some popular synthetic methods. Surface modifications in creating anisotropic and higherordered nanostructures are followed. The applications of surface modification including its use as a tool for creating new nanostructural types and in biotechnology are given towards the end of the review.
2. General Approaches to Surface Modification of Nanostructures The strategy for modification or functionalization of nanoparticles depends on the specific atomic structures of the surfaces and their interactions with ligands. Nanoparticles with chemical functional groups on their surface can be modified with organic, inorganic molecules, regardless whether they are synthesized in aqueous or nonhydrolytic solutions. Hydroxyl group is one of the common functional groups and it can react with carboxyl group via oxygen atom or with various silane groups through -O-Si 16. For nanoparticles synthesized from nonhydrolytic solutions, they have to be modified first in some cases to introduce functional group such as hydroxyl and mercapto groups 6. Amine and oxysilane are two other popular functional groups that are used for surface modification of nanoparticles 17. Nanoparticles of semiconducting quantum dots (QDs), Pt, Ag and Au can be modified with mercapto groups using the well developed thiolate chemistry in the design 18. The selective surface species can often prevent the nanoparticles from aggregation. To fulfill specific applications, the primer molecules may be necessary to graft and activate the surface for further functionalization by other chemical or
Surface Modifications of Nanoparticles
87
biological molecules. Usually, these primers have dual functional groups, one for surface binding and the other for initiating the designed chemical reactions. Silane coupling agent is such a compound which can change the surface of nanoparticles to vitreophilic and make the easy deposition of coating layers 19. Electrostatic interaction 20, 21 and other types of van der Waals interaction 22 are sufficient for the surface modification of nanoparticles in some cases. Since these types of interactions, which are physical adsorption in nature, tend to be weak, chemical reaction is needed if strong binding is necessary for the surface modifications. The surface structures often determine the types of reactions. For example, Au and Ag nanoparticles usually require sulfur as the bridging element due to the strong affinity of these metals with thiol group 23. Other suitable chemical reactions for the formation of surface coating include silanization and polymerization 17, 24. Emulsion and self assembly are two commonly used strategies in the solution phase surface modification of nanoparticles 25-28. Some other approaches include ozone treatment 29, 30, microwave assisted modification 31, coprecipitation 32 and dry-mechanical coating technique 33. For surface modification with oxides, sol-gel method is rather useful and the Stöber method is among the most widely used strategy that was originally developed for making silica layers 19, 34, 35. 2.1. Adsorption and Self-Assembly 2.1.1. Modification through Adsorption of Organic Molecules Physical adsorption is among the methods that are conceptually straight forward for modifying the surface to become hydrophilic or more stable in suspensions 36-39. The conditions used are generally not complicated, since small molecules or even polymers can attach to the surface of nanoparticles through exchange with the original ligands or adsorption, and such procedures can be conducted at ambient room temperatures.
88
Shen and Yang
2.1.2. Modification through Self Assembly and Layer-by-Layer Deposition Self-assembly is a method capable for making one, two or even three-dimensional structures of nanomaterials 40, 41. While self-assembly on planar structure has been widely studied, it is not a straight forward technique when nanometer-sized colloidal particles are involved. The major driving force for self-assembly include electrostatic interaction, surface tension, capillary force, hydrophobic interaction and bio-specific recognition 42. Host-guest interaction is typically seen in biological systems, but also regularly used in the assembly of nonbiological molecules through these weak interactions. Long carbon chain surfactants are routinely used as stabilizing agents and serve as a good platform for the host-guest interaction. Figure 2 shows a schematic illustration of using cyclodextrin (CD) to modify oleic acid-capped iron oxide nanoparticles 43. The interaction between the hydrophobic chains of oleic acid and cavity of CD molecules is the key in this process. After the modification, the surface of these nanoparticles change from hydrophobic to hydrophilic, and the particles can be readily dispersed in aqueous solution.
Figure 2. Schematic illustration showing the surfaces of surfactant capped nanoparticles modified by cyclodextrin through host-guest chemistry 43. Reprinted with permission from Reference 43, © 2003, American Chemical Society.
Surface Modifications of Nanoparticles
89
Electrostatic interaction is dominant in the layer-by-layer (LbL) self assembly of polyelectrolyte shells 26-28. This method provides a route for the deposition of various chemicals onto carbon nanocubes 44. Caruso group has focused on the modification of nanostructures using self assembly methods for some time 28, 42, 45, 46. The thickness of the shells can be adjusted by sequentially depositing oppositely charged polyelectrolytes or even nanoparticles on the surface through predominantly such electrostatic interaction. Up to eight layers of polyelectrolytes on the surface of 35 nm Au nanoparticles have been demonstrated based on this method 47. Layer-by-layer assembly can be formed with or without templates 48. Nanostructures with different compositions and shapes can be used as templates as well. Besides molecular and polymeric species, nanoparticles can also be incorporated into the shell layers 49. Such surface self-assembly can be used for the formation of hollow spheres or capsules from the core-shell structures. The coating of nanoparticles with polyelectrolyte layers and nanoparticles has been achieved by polymerization and silanization methods. By removing the templates through calcination or etching, hollow capsules, feasible candidates for drug delivery carriers that are composed of polymer and/or nanoparticles can be formed 27, 50. The most commonly used cation-anion polyelectrolyte pairs include poly(styrene sulfonate) (PSS) and poly(diallyldimethylammonium chloride) (PDADMAC) 45, 47, PSS and poly(allylamine hydrochloride) (PAH) 28, 49, 50, PSS and poly(diallydimethyl-ammonium) (PDDA) or poly(pyrrole) and poly(N-methylpyrrole) 44, 46. Sequential assembly of nanoparticles and polyelectrolytes can apply to coat the surface of cores 27, 51. Functional moiety of small molecules or polymers can be incorporated into the multiple component shell structures. While the electrostatic interaction can be used to produce multilayers, the cation and anion polyelectrolytes are often intertwined and no clear boundary exists between them, especially for thick films 48. This self-assembly approach has been adopted for surface modification of particles with polymers 28, 47, 50. TEM image revealed that a single gold nanoparticle could be coated uniformly with multiple layers of polymer. The coating was consisted of one monolayer of
90
Shen and Yang
10-mercaptodecanesulfonate followed by eight layers of oppositely charged PSS and PDADMAC 47. Modification of nanoparticles based LbL approaches is not as straight though as in the cases of micron-sized spheres. Achieving uniformity among large quantity of small nanoparticles with satisfied yield can be challenging. 2.2. Surface Modification Based on Organic Reactions Since physisorption is a relatively weak interaction, the attached molecules or polymers are very sensitive to the environments such as pH value and temperature, and easy to be dissolved. Direct grafting through chemical adsorptions has been applied for the surface modification of nanoparticles having strong covalent bonds with functional molecules 52. For instance ether bonds generated with dextran nanoparticles with aliphatic and aromatic groups are used to make the surface hydrophobic 53. Since hydroxyl group (-OH) is commonly used on the surface of nanoparticles, many of the chemical reactions target the interactions with the molecules that have such functional moiety 54, 55. Silanization is one such method where silanol group-containing silane coupling agents (SCAs) react with surface species 17, 19, 34. Hydrolysis and polycondensation of -Si-(OH)3 itself and with surface hydroxyl groups are powerful combination to create new coating layers. These silane groups can have different organic moieties at the one end of the SCAs and a functional group at the other for further conjugation with molecules. As covalent bonding between SCAs and the surface species can ensure strong linkage, silanization can be used to create ultrathin single molecular layer on the surface of nanoparticles. Thus SCAs are regularly used as primers to activate the surface of nanoparticles. (3Aminopropyl)triethoxysilane (APS) and (3-mercaptopropyl) trimethoxysilane (MPS) are the most often used primers for this purpose 56-59. Poly ethylene glycol (PEG) 57, amino 60-62, vinyl 63 and phenyl 64 are some of the popular functional groups at the other end of the ligand. Pre-determined organic reactions can take place on the surface of nanoparticles upon the activation by primers. For example, sulfur bridge can be created on gold nanoparticles to create tetrathiafulvalene-modified
Surface Modifications of Nanoparticles
91
gold nanoparticles (TTF-S-gold), which can be used as coating layer on the Pt electrode to catalyze redox reactions 65. Likewise, TiO2 nanoparticles with surface hydroxyl group can form a surface complex with the active NCO groups on tolylene diisocyanate (TDI) 66. Cycloaddition and Brust-Schiffrin method are two other popular approaches to the functionalization of nanoparticle surfaces 67-69. When the direct chemical reaction with the surface cannot be implemented, covalent attachment can take place after the initial surface treatment 43, 70. For instance, oleic acid-capped γ-Fe2O3 nanoparticles are first ligand exchange with phosphonic acid-azide or carboxylic acidalkyne before the formation of bonds between the functional groups and surface through the “click chemistry” 71. Trichloro-s-triazine (TsT) is a good linker molecule to bridge methoxy poly(ethylene glycol) (mPEG) and dopamine 72. It has three chlorine groups which can react sequentially with other molecules. In general, mPEG is first conjugated with one of these three chlorine groups to form mPEG-TsT, while another chlorine reacts with dopamine to form mPEG-TsT-dopamine. The catecol unit on the dopamine moiety replaces the oleylamine and oleate groups on the surfaces of Fe3O4 nanoparticles. TsT can also be used for the coupling of other molecules or polymers. Oligomeric phosphines with different functional moieties can bind to the surface of semiconducting quantum dots (QDs) and prevent them from aggregation 73. In such case, the outer most layers are designed in such way to be compatible with microenvironments of the solvents. Further conjugation with biomolecules can be achieved through the interaction with the outer functional groups. Generally, attachment through covalent bond takes place directly on the surface of nanoparticles. Figure 3 shows an example of the surface modification of iron oxide via covalent bond 17. Methoxy-poly (ethylene glycol) silane (mPEG-Sil) can have strong interaction with surface and replace the capping agent of oleic acid. The direct consequence is the ligandexchanged particles changes from hydrophobic to hydrophilic. These modified particles can also disperse well in many aqueous biological media that have relatively high salt concentrations which destabilize the dispersion of nanoparticles (Figure 3b).
92
Shen and Yang
Figure 3. (a) Scheme showing surface modification of oleic acid capped iron oxide through covalent bond with mPEG-Sil and (b) photographs of Fe2O3 dispersed in various biologically relevant media (From left to right: 0.9% saline, 25-mM phosphate buffer at pH 7.2, limulus amebocyte lysate (LAL) aqueous solution; and endothelial based medium-2 (EMB-2).17 (Reprinted with permission from Reference 17, © 2008, American Chemical Society.)
2.3. Surface Modification Based on Polymerization Different polymerizations have been applied in the surface modification of nanoparticles 74-77. The suitable types of reactions can be free radical or ionic polymerization. They can also undergo chain transfer, or occur in emulsion if solvents are involved. Solvent, initiator, monomer to initiator ratio, and surface structure can affect the choice of methods used to graft polymers onto the given nanoparticles. Nanostructures of metal, metal alloy, metal oxide and other materials have all been modified with polymers.
Surface Modifications of Nanoparticles
93
Poly(methyl methacrylate) (PMMA) is a popular polymer that is used for surface modification. Functionalization of CdS nanocrystals by PMMA can be achieved via free radical polymerization 74. Assynthesized CdS nanoparticles are first grafted with methacryloxypropyltrimethoxysilane (MPS). The dangling double-bond moiety of MPS on the nanoparticles can polymerize upon initiation by 2, 2-azobic isobutyronitrile (AIBN). Radical polymerization is employed to deposit methyl methacrylate on titania nanoparticles 75. In this case, 6-palmitate ascorbic acid (6-PAA) is used first in treating TiO2 surface to allow phase transfer of nanoparticles from water into toluene where radical polymerization occurs. After the coating with PMMA, TiO2 nanoparticles become thermal stability 76. By using reversible additionfragmentation chain transfer (RAFT) polymerization, poly(2(dimethylamino) ethyl methacrylate), poly (acrylic acid) and polystyrene (PS) can all be grafted onto the surface of gold nanorods synthesized in aqueous solutions using trisodium citrate as stabilizing agent 77. Biocompatible polymers are designed as the shell layers in core-shell or core-corona-shell nanoparticles, largely depending on the specific biological applications 78. Among the variety of available biocompatible polymers 25, 74, PEG and its derivatives are heavily studied because they are hydrophilic and capable to prevent unspecific protein adsorption which is important for intravenously delivered drugs to have extended half-life in blood circulation 79. Further conjugation with functional moieties is often required in such applications. The selection of polymers however often depends on the targeted applications. If hydrophobic surface and chemical stability of the coating are required, PS can be a better candidate than many others 24. Dendrimer is another special class of polymers that can have many structural varieties and local morphologies which can be appealing for the design of new drug carriers with excellent release profile 78, 80. Figure 4 shows the structure of polystyrene-based dendrimer which has different nods and pockets with each new generations and an outer active layer of hydroxyl functional groups 81, 82.
94
Shen and Yang
Figure 4. Structure of dendrimer-like polystyrene 81. (Reprinted with permission from Reference 81, © 2008, American Chemical Society.)
2.4. Surface Modification with Inorganic Layers Based on Sol-Gel Approaches Inorganic materials are usually chosen to improve the stability and to introduce new electronic, photonic, magnetic, mechanical, and surface chemical properties of the nanoparticles 83-86. The common choices of inorganic layers include silica, titania, zirconia and other metal oxides that are readily obtained by solution phase approach such as sol-gel method, although other classes of materials including even pure metals which can also be used as shell layers. Among the inorganic materials, silica has been widely used since the invention of Stöber method originally designed for the preparation of silica nanoparticles with well controlled spherical shape and size using alcoholic solvents and silicon alkoxide and other precursors 19, 87. The hydrolysis and polycondensation of silica precursors can be catalyzed by either base or acid. The reasons for choosing other oxides such as zirconia, titania or alumina in surface modification are more diverse than silica and mostly related to the
Surface Modifications of Nanoparticles
95
specific requirements of optical, electrical and mechanical properties 88, 89. Ternary oxide ceramic coatings have been developed. For instance, yttria-stabilized zirconia (YSZ), a solid oxygen ion conductor, can bind to silica particles as thin as ten nanometers using ultrasound-assisted deposition 88. In the following subsections, we will review the sol-gel approach for the commonly-seen binary oxide surfaces and the Stöber method in the context of surface modification. 2.4.1. Sol-Gel Methods Sols are small colloidal nanoparticles in solution and form interconnected network of metal oxides, which are called gels, upon further polycondensation in the presence of acid or base catalysts 90. The common metal oxides such as silica, zirconia and titania are frequently made from the precursors of metal alkoxides based on the sol-gel methods. The widely used silica precursors are short carbon chain alkoxyl silanes, particularly tetraethyl orthosilicate (TEOS) 89, 91, 92. Ammonia and sodium hydroxide (NaOH) are the typical base catalysts, while hydrochloric acid (HCl) and nitric acid (HNO3) can be used as acid catalysts for the hydrolysis and condensation of TEOS in water and alcohol mixtures. NH4OH Si(OC2 H 5 ) 4 + H 2O → Si(OH )4 + C2 H 5OH
Si(OC2 H 5 ) 4 + C2 H 5OH + H 2O → [ Si(OH ) 4 ]n • xC2 H 5OH • xH 2O
[ Si(OH ) 4 ]n • xC2 H 5OH • xH 2O → SiO2 + C2 H 5OH + H 2O
In the case of TEOS as the precursor, the hydrolysis leads to the formation of silanol and ethanol in the solvent mixture during the first step. These silanol cross links to form oligomers which further condense catalytically through -OH interactions on the surfaces of nanoparticles and form silica. The second and third steps are the polycondensation, during which period dense silica are formed. Other silane and halide precursors have also been studied for the formation of silica sol-gel for surface modification of nanoparticles. 3-aminopropyltrimethoxysilane
96
Shen and Yang
(APTMS) and methacryloxypropyltri-methoxysilane (MPTMS) are among the most widely used precursors 93, 94. Similarly, there are different precursors that can be used in the formation of zirconia coating 88, 95-99. The alkoxide form is still the popular choice, and zirconium propoxide and butoxide are among the commonly used 95, 100. The formation of zirconia sol-gel also undergoes the hydrolysis and polycondensation steps during the coating process: NaOH Zr (OC4 H 9 ) 4 + H 2 O → Zr (OH ) 4 + C 4 H 9 OH
Zr (OH ) 4 + C4 H 9OH + H 2O → [Zr (OH ) 4 ]n • xC4 H 9OH • xH 2O [ Zr (OH ) 4 ]n • xC4 H 9OH • xH 2 O → ZrO2 + C4 H 9OH + H 2O
Yttrium is sometime introduced to stabilize the certain types of zirconia structures at around room temperature range 88. Reaction temperature is generally higher for the formation of zirconia sol-gel and the deposition of zirconia crystals on the surface of nanoparticles than that with silica coating. Monoclinic is usually the stable crystal phase made at low temperatures, and tetragonal and cubic phases form upon being annealed at high temperatures 98, 99, 101, 102. Surface of zirconia coating on nanoparticles of metals or metal oxides is typically faceted and not as smooth as that of silica 97. However, using ultrasound-assisted deposition YSZ can be uniformly deposited on the surface of silica particles 88. The general mechanism for the formation of other metal oxides such as titania is similar to those for zirconia, and includes hydrolysis and polycondensation steps using either acid or base catalyst. 2.4.2. Stöber Method This method is mainly referred to the surface modification of nanoparticles with silica shells. Since its invention, it has been modified and improved for use on nanoparticles with different compositions, sizes, shapes and surface chemistry 19, 87. There are several advantages of the Stöber method: First, the synthesis can take place in solvents with a range of hydrophilicity or hydrophobicity. A primer can be introduced if necessary to active the surface of nanoparticles for highly hydrophobic
Surface Modifications of Nanoparticles
97
solvents. Second, the formation of silica shells can not only prevent the nanoparticles from coalescence but also generate functional surfaces for further modification. The silica coated nanoparticles are often dispersed readily in aqueous solutions. Third, silica coating can improve the biocompatibility of nanoparticles due to its low toxicity 19. Figure 5 shows silica coated gold nanoparticles using the Stöber method 19. Silica precursors-(3-aminopropyl)-trimethoxysilane can easily react at the surface region of the nanoparticles in aqueous phase. The control of uniformity in thickness can reach low nanometer regime. For nanoparticles synthesized from hydrophobic solvents, Stöber method cannot always be applied directly and usually a primer is required to activate the surface 17. Mostly these primers used for Stöber method are silane-based agents such as APTMS or MPTMS, which can change the surface of nanoparticles hydrophilic. Stöber method can be used to coat silica on nanoparticles of different surface chemistry from aqueous or nonhydrolytic solvents 103. Metal 17, 19, 104, metal oxide 17 and alloy nanostructures 105, 106 have been modified with
Figure 5. TEM image of SiO2 coated nanoparticles 19. (Reprinted with permission from Reference 19, © 1996, American Chemical Society.)
98
Shen and Yang
silica shells this way. The capping reagents used for the synthesis of these nanoparticles can affect the efficiency of the surface modification. The commonly used capping agents in nonhydrolytic systems such as acryl amine, oleic acid and long carbon chain thiol groups need to be replaced at least in part before the proper deposition of silanol groups. While Stöber method is widely used in modifying spherical nanoparticles, achieving uniformity in coated layers for nanowires, nanocubes, nanorods, and nanobars is quite feasible. Figure 6 shows the silica coating on quantum dots of CdSe/ZnS, nanobars and rods of Ln(BDC)1.5(H2O)2, where Ln=Eu 3+, Gd 3+, or Tb 3+ and BDC= 1,4-benzenedicarboxylate, and silver nanowires using the Stöber method 26, 104, 107-110. These silica layers can be as thin as a few nanometers, and still maintain their uniformity indicating an excellent level of control of deposition.
Figure 6. TEM images showing different types and shapes of at (a-d) as-made and (e-h) silica coated nanoparticles: (a and e) QD of CdSe/ZnS 9 (Reprinted with permission from Reference 9, © 2006, American Chemical Society); (b and f) nanobars and (c and g) nanorods of Ln(BDC)1.5(H2O)2, where Ln=Eu 3+, Gd 3+, or Tb 3+ and BDC=1,4benzenedicarboxylate 107 (Reprinted with permission from Reference 107, © 2007, American Chemical Society); and (d and h) silver nanowires 104 (Reprinted with permission from Reference 104, © 2002, American Chemical Society).
Surface Modifications of Nanoparticles
99
2.5. Surface Modification with Multiple or Composite Layers Besides single layer of coating, corona-shell double layers can be created by using a combination of different materials and nanoparticles. Figure 7 shows a TEM image of three-layered gold-silica-PS particles made by using both sol-gel and polymerization methods 111. In this procedure, gold nanoparticles are coated with silica by seeded growth technique and the surface is then activated with MPTMS. Styrene monomers are polymerized on the silica surface initiated by potassium persulfate. TEM images showed the distinctive layered structures 111. Nanoparticles can also be sandwiched in two oxide layers made by solgel methods, as being demonstrated with silica-Au-silica core-double shell particles 112.
Figure 7. TEM image of gold-silica-polystyrene nanostructures 111 (Reprinted with permission from Reference 111, © 2004. Academic Press Inc, Elsevier Science).
Layer-by-layer deposition is another very useful method for making multiple-layered nanostructures. Figure 8 shows a schematic illustration of self-assembly of polyelectrolytes and CdTe quantum dots on nanoparticles of magnetite (Fe3O4) 40. The self-assembly starts with
100
Shen and Yang
physical adsorption of PAH cationic polyelectrolyte, followed by the deposition of anion PPS with sulfonate groups. Once multilayered electrolytes form, surfactant capped CdTe nanoparticles can be incorporated into PSS/PAH layers through the electrostatic interaction. This LbL assembly is a useful approach to the fabrication of multifunctional nanostructures.
Figure 8. Scheme showing the formation of Fe3O4/(PE3/CdTe)n based on the LbL selfassembly 113. PE is abbreviated for polyelectrolyte (Reprinted with permission from Reference 113, © 2004, American Chemical Society).
2.6. Experimental Designs As nanomaterials generated in solutions are stabilized by surfactants, ligand exchange with small molecules is common 57, 114-116. Figure 9 shows three different types of nanoparticles that have been used in ligand exchange for surface modification. In this process, surfactants with designed structural properties replace the capping agents through either kinetic stabilization or chemical bonding. Typically ligand exchange leads to good stability and creates new functional groups for applications
Surface Modifications of Nanoparticles
101
that require specific selectivity, surface chemistry and hydrophilicity. Surface composition and structure of the particles are critical for the selection of new ligands. For ferrite nanoparticles, the capping agent with carboxylic group can be readily exchanged with functional siloxide (Figure 9a) 57. Similarly, amine on ZnSe quantum dots can be replaced by mercaptopropionic acid (MPA) because of the formation of S-Mn or S-Zn bonds (Figure 9b) 115. On the surface of FePt nanoparticles, iron atom binds to the carboxylic derivatives of oleic acid, while platinum atom interact preferably with oleylamine 114. Both oleylamine and oleic acid can be exchanged with mercaptoalkanoic acid through the interaction between carboxylate group and iron, and mercapto group with platinum (Figure 9c) 114.
Figure 9. Surface modification of nanoparticles based on ligand exchange: (a) oleic acid exchanged with silane on the surface of CoFe2O4 nanoparticles 57 (Reprinted with permission from Reference 57, © 2007, American Chemical Society); (b) amine exchanged with mercaptopropionic acid (MPA) on Mn doped ZnSe quantum dots 115 (Reprinted with permission from Reference 115, © 2007, American Chemical Society); and (c) oleylamine and oleic acid exchanged by mercaptoalkanoic acid on FePt nanoparticles 114 (Reprinted with permission from Reference 114, © 2006, American Chemical Society).
102
Shen and Yang
Another widely used ligand exchange system is thiolate on gold and silver, which relies on the strong chemical interactions between sulfur and these two metal atoms 18, 117. Poly(allylamine) on gold and silver nanoparticles can be replaced by ω-functionalized alkylthiols to increase their stability in water 118. These ω-functional groups can be acid, alcohol, amine, and biotin. Phosphine, which is used as stabilized capping agent to synthesize gold or silver nanoparticles can be exchanged with ω-functionalized thiols groups as well 119, 120. The ligand exchange happened between triphenylphosphine and octadecanethiol. Through a three-step ligand exchange process. Phosphine ligands are first replaced by thoil groups in the form of AuCl(PPh3). This step is followed by further removal of PPh3 or AuCl(PPh3) groups. Thiol groups then reorganize and form crystallinelike shells on the surface of Au nanoparticles 120. Guest molecules with hydrophobic cavity such as α-cyclodextrin (CD) can be used in modification of nanoparticles through interaction with the aliphatic thiol molecules on the surface 121. Among the various microenvironments for the reactions, emulsions including micelles or reverse micelles are capable to control the liquid mixtures in confined spaces. The nanoparticles modified inside emulsions can have various sizes, shapes and different functional molecules. Silica nanoparticles can be modified with dye molecules using different types of water in oil microemulsions 122. Besides direct ligand exchange and emulsion, microwave-assisted modification 31, dry mechanical coating technique 33, and ozone deposition are some other techniques that have been developed to facilitate the surface treatment 123-128.
2.7. Surface Modification in the Synthesis of Hollow Spheres One active research area in nanomaterials is the synthesis of hollow nanostructures from coated nanoparticles. Such hollow nanostructures have potential applications in biotechnology 129-131, reaction engineering 132, electrochemical engineering 133, 134 and other areas. For particles with size greater than 100 nm, interfacial polymerization, polycondensation, LbL deposition, nozzle reactor spray drying and emulsion are some of
Surface Modifications of Nanoparticles
103
the approaches 40, 135-137. For those smaller than 100 nm, the synthesis of hollow particles are not straightforward in many cases 138. Using organic or inorganic nanostructures as templates, ceramics and polymers alike can be deposited on the surfaces through either chemical or physical interaction 107, 129, 139-150. The core materials can be removed through calcinations 143, 144, 147, 148 or etching 107, 129, 139-142, 145, 146, 150 to create void spaces. Much attention has been focused on the combination of LbL self assembly and templating method, with which the thickness of hollow shells can be controlled by alternately depositing oppositely charged polyelectrolytes on the cores 27, 42, 46, 47. Self-templating is an approach that combines emulsion method with template 151. In this approach, surfactants are used to form micelles in water and as template at the same time. By adjusting pH and controlling polycondensation process, hollow spheres are fabricated. Biological or chemical functional groups can further be grafted onto hollow particles 107. The functionalization of interior surface is nontrivial and needs further study 152-155, although optically and magnetically active metal and metal oxide particles and polymers have successfully been deposited onto the interior surface or in the void regions.
3. Surface Modification of Magnetic Nanostructures The general approaches described in the previous sections have been used for the surface modification of magnetic nanoparticles and selection of a particular method is largely based on the surface chemistry. While surface modification can address a range of the needs for the applications of magnetic nanoparticles in bio-imaging, separation, ferrofluids, it can also change the properties, such as decrease of the saturated magnetization 17. So there need to develop material and application specific approaches. For instance, secondary nanostructures that are composed of many small superparamagnetic nanoparticles have been synthesized in order to maintain the saturated magnetization close to that of the bulk counterpart without losing their unique properties 156, 157. In this section, we review approaches developed on the surface modifications of different classes of magnetic nanoparticles.
104
Shen and Yang
3.1. Oxides Among the magnetic oxides and ferrites that can be produced as nanoparticles, iron oxides are perhaps the most widely studied materials in this category because of the availability of a range of synthetic methods for making high quality products and their broad applications. In recent years, size, shape, composition and crystal phase specific nanostructures of iron oxides can be generated with accuracy reaching sub-nanometer level 1. Among the various forms of iron oxides, maghemite (Fe2O3) and magnetite (Fe3O4) are the two best known materials and the focus of many studies 60, 62. Silica, titania, zirconia and other inorganic materials are commonly used as coating materials, while organic and polymeric materials have also been applied 24, 79. We begin with silica as it is the popular choice of material for coating iron oxide nanoparticles. Iron oxide nanoparticles produced from both aqueous and nonhydrolytic solutions have been successfully modified with silica coatings, although those made in nonhydrolytic solvents may need additional treatment first.
Figure 10. TEM image showing mPEG-exchanged iron oxide nanoparticles that are subsequently coated with SiO2.
Surface Modifications of Nanoparticles
105
When the widely used oleic acid and oleylamine are the capping agents in the synthesis of iron oxide nanoparticles in organic media, ligand exchange with short carbon chain molecules such as APTMS, MPTMS and other siloxane-based compounds is not always effective. PEG-containing compounds, owing to their unique amphiphilic property, are shown to be the linker molecules for surface modification of those nanoparticles synthesized in organic media. The presence of mPEG-sil at the surface of these nanoparticles can serve as reactive sites for the formation of SiO2 shells trough the Stöber method. In such case, mPEGsil replaces the surface capping groups of oleic acid and oleylamine in toluene and subsequently reacts with TEOS to form silica on the surface. Figure 10 shows the TEM image of silica coated iron oxide nanoparticles using this approach 95, 158-161. Other oxides including TiO2, ZrO2 and Al2O3 have also been used for the coating of iron oxide nanoparticles in different applications 95, 159, 160. Liu et al. used Al2O3 as the coating for Fe3O4 particles for the affinity capture of uropathogenic E. coli. 160. Phosphate moiety on the pigeon ovalbumin (POA) was used as a linker for the conjugation of alumina on the surface. The core-shell nanoparticles can be applied for the immobilization of E. coli within 30 s. External magnetic field can be applied for the separation of free E. coli from the coated particles because of the good magnetic response of these magnetic nanoparticles. Chen et al. coated Fe3O4 nanoparticles with TiO2 to analyze phosphopeptides using surface-assisted laser desorption/ionization mass spectrometry, because titania nanoparticles have been used as packing materials in columns for enrichment of these peptides 159. Protein digest products can be concentrated in an external magnetic field on Fe3O4 nanoparticles coated with phosphopeptide-loaded TiO2 layers. These coated magnetic nanoparticles can be introduced directly into a mass spectrometer for the detection of proteins because of the high concentration of phosphopeptides on Fe3O4@TiO2 nanoparticles. Zirconium dioxide has also been coated onto Fe3O4 nanoparticles and used for concentrating phosphopeptides 95. Besides the inorganic coatings, polymers and biomolecules have been used for the formation of core-shell nanostructures of magnetic oxides. The commonly used methods include free radical polymerization 74, 76,
106
Shen and Yang
emulsion polymerization 162-164 and graft polymerization 165, though chain transfer polymerization 77 and atom transfer radical polymerization (ATRP) 166 are also used for the surface coating of magnetic nanoparticles. Wang et al. developed the method for producing polystyrene layer uniformly on Fe2O3 nanoparticles using solvent free ATRP method (Figure 11) 24. The monodisperse nanoparticles stabilized with oleic acid were first ligand-exchanged with 2-bromo-2-methylpropionic acid (BrMPA) which works as initiator for the polymerization of styrene monomers. As the initiators were concentrated on iron oxide, polymerization was confined to the surface region. Gel permeation chromatography (GPC) measurement indicates that the molecular weight is about 6900 and polydispersity is 1.23. Dey also synthesized polystyrene coated magnetite nanoparticles using the ATRP method 166. The core-shell nanoparticles had good monodispersity with a diameter of about 7 nm. Wuang et al. deposited polypyrrole (PPY) on Fe3O4 nanoparticles based on a facile mini-emulsion polymerization method 167, 168. Magnetite nanoparticles were first transferred into an aqueous solution using sodium dodecylbenzene sulfonate (NaDS). Pyrrole was added to this solution to form the magnetite-monomer dispersion. A mixture of FeCl3•6H2O and poly(vinyl alcohol) (PVA), which was used as surfactant, was added into this dispersion in an icecooled bath. The mini-emulsion ensured that the magnetic nanoparticles distributed evenly inside the PPY matrix. The outer surface of PPY could be further modified with herceptin, a drug for cancer therapy. The herceptin-functionalized magnetic nanoparticles could be taken up seven times higher than those of nonfunctionalized particles. PEG-related polymers are often used for the surface coating of magnetic nanoparticles because of their hydrophilicity and biocompatibility 79. Flesch et al. used PEG functionalized with methacrylate groups (PEG-MA) to coat maghemite nanoparticles using inverse emulsion polymerization method 79. The surface of maghemite nanoparticles was activated with methacryloxypropyotrimethoxysilane (MPS) first before PEG-MA was covalent bonded with these MPS molecules using dibenzoyl peroxide (PBO) as the initiator. Compared to the direct copolymerization method, inverse emulsion can increase the
Surface Modifications of Nanoparticles
107
local concentration of PEG-MA to generate thick coating on nanoparticles. Other polymers have also been coated onto the surface of iron oxide nanoparticles using ligand exchange and covalent attachment or grafting technique 60, 61, 169-171. In general, polymer coating is uniform in thickness regardless of the morphology and size of the nanoparticles.
Figure 11. TEM image of polystyrene 24 (Reprinted with permission from Reference 24, © 2003, American Chemical Society).
Besides oxide and polymeric coatings, metal was used to modify nanoparticles of iron oxide as well. Figure 12a shows an example of using gold as coating material for making Fe3O4@Au core-shell nanostructures 172. The reduction of chloroauric acid by sodium citrate in oleylamine and chloroform led to the formation of gold metal, although the direct observation of gold layer using TEM was not obvious. Interestingly, the reverse coating, i.e. Au@Fe3O4 core-shell nanostructures was also possible (Figure 12b) 173. The formation of iron oxide coating was achieved through first decomposition of iron carbonyl followed by the oxidation of resulting iron formed in the early step to magnetite.
108
Shen and Yang
Figure 12. Schematic illustration of the synthesis and TEM images of (a) Fe3O4@Au 173 (Reprinted with permission from Reference 173. © 2007, American Chemical Society) and (b) Au@Fe3O4 core-shell nanostructures 172 (Reprinted with permission from Reference 172, © 2005, American Chemical Society).
3.2. Metals The commonly used magnetic metal nanoparticles are made of the three elements: iron, cobalt and nickel. These metal nanoparticles have higher saturated magnetization and show ferromagnetic properties at smaller size than their oxide counterparts. The metal-based magnetic nanoparticles however can be readily oxidized if exposed to air, water and many other mild oxidizing agents. Thus even though as-synthesized colloidal magnetic nanoparticles of metals are capped with surfactants, they are unstable and form amorphous oxidized shells on the surface if no additional treatment is introduced before their exposure to oxidation environments. Controlled oxidation is used for the formation of crystalline shells on the surface of metal nanoparticles to improve the chemical stability of metal cores 7. Peng et al. synthesized uniform Fe@Fe3O4 core-shell nanoparticles by controlled oxidation method using (CH3)3NO (Figure 13) 7. The outer Fe3O4 layer dramatically increased the stability of iron cores. These Fe@Fe3O4 core-shell nanoparticles were easily transferred into aqueous solutions after ligand exchange.
Surface Modifications of Nanoparticles
109
Nickel and cobalt magnetic nanoparticles can also be prepared with corresponding oxides as protecting outer layers. Lee et al. synthesized NiO coated nickel nanoparticles using Ni(acac)2 as precursor, oleylamine as capping agent and trioctylphosphine (TOP) as solvent 174. Oxidation occurred on the surface of as-synthesized nickel nanoparticles in air. The Ni@NiO core-shell nanoparticles had good affinity for polyhistidine that was used in the separation and purification of proteins and superparamagnetic properties from mainly nanometer-sized nickel cores. The formation of core-shell structure can affect both the surface chemistry and the magnetic properties of the metal cores. Seto et al. demonstrated the size-dependant magnetic properties of Ni-NiO coreshell nanoparticles that ferromagnetism existed even below the typical critical size for superparamagnetism of nickel 175. This observation suggests that there existed spin exchange coupling between superparamagnetic Ni cores and antiferromagnetic NiO shells.
Figure 13. Hysteresis curves and TEM images of Fe@Fe3O4 core-shell nanoparticles (Reprinted with permission from Reference 7, © 2006, American Chemical Society).
7
110
Shen and Yang
Wiedwald et al. investigated the formation of Co@CoO core-shell nanostructures and its surface modification with hydrogen and oxygen plasma 8. They found that average magnetization and coercive field could be tuned by controlling the thickness of CoO layer. The magnetic property of such Co@CoO core-shell nanoparticles however could depends highly on the nanostructures 176. Exchange biasing between the ferromagnetic cobalt core and antiferromagnetic CoO shell could be detected only when the shell reached certain thickness and unusual magnetic behaviors were also observed due to defects in CoO. Hou et al. showed the synthesis of Sm2O3 nanoparticles coated Co shell and their reductive annealing to SmCo5 at high temperatures 177. Authors indicated that the formed SmCo5 nanoparticles were not only stable in air but also with high magnetic coercivity. Besides shells of native and controlled oxides, as-synthesized magnetic metal nanoparticles have been modified with inorganics or organics. Silica coating is one of the most often used inorganic materials in the surface modification of magnetic metal nanoparticles 178. Fernández-Pacheco et al. coated iron nanoparticles with silica using arc-discharge method 178. The magnetic response of coated iron nanoparticles was much stronger than other inorganic encapsulated magnetic nanomaterials. Other inorganic materials including Pt, Ag and Au have also been explored as coating for magnetic metal nanoparticles. Platinum and gold were electrochemically deposited on the surface of Fe, Co or Ni nanostructures 179. In this method, core metals were electrodeposited on glassy carbon substrates and Pt or Au shells formed by immersing the substrates into chloroplatinic or chlorolauric solution, respectively, to partially replace the cores. The Co@Ag core-shell nanostructures were synthesized by reacting Co nanoparticles capped with oleic acid with Ag2SO4 solution 180. For molecular coatings, while thiols are commonly used for noble metal nanoparticles, they can also be used in capping magnetic nanoparticles of non-noble metals. Dodecanethiol could adsorb on the surface of Ni nanoparticles directly during the synthesis 181. The dedecanethiol could not only control the size of Ni nanoparticles but also increase their dispersibility in nonaqueous solvents. Surface modification with such small molecules can affect the properties of
Surface Modifications of Nanoparticles
111
nonmagnetic metal nanoparticles dramatically 182, 183. Gold, silver and copper nanoparticles can all show magnetic properties once their surfaces is modified with ligands, such as thiol groups 184, 185. Strong interaction between sulfur and metal atoms at the interface regions results in the magnetic property. The chemically induced magnetic properties of these metal nanoparticles are largely surface phenomenon and thus weak when compared with traditional metallic ferromagnetic materials.
3.3. Metal Alloys Among the nanostructured magnetic alloys, FePt is the most widely studied one in recent years because of its high magnetoanisotropy and good chemical stability, though other materials have also been examined 1. The face centered tetragonal (fct) FePt is the alloy that has the high coercivity and small domain wall width. The surface modification of fct FePt alloy nanoparticles has been focused on the maintenance of the size while tuning of the magnetic and crystal phase properties through postsynthesis treatments 9, 105, 106. Tsang and co-workers developed a method for uniformly coating fct FePt nanoparticles with silica using Stöber method 106. The FePt nanoparticles were around 3 nm in diameter based on X-ray diffraction (XRD) analysis. In the absence of the silica coating, the diameter increased to 23.8 nm when the particles were annealed at 800°C. With the silica shells, the growth of nanoparticles was prevented and the nanoparticles remained to be separated from each other. Silica shells acted as a barrier layers during the annealing process. These silica coated FePt nanoparticles became highly monodispersed and could form two and three dimensional self-assembly under an external magnetic field, as shown in Figures 14a and 14b 105. Besides silica coating which is magnetically inert but can improve the stability of FePt cores, the coating itself can also be made of magnetic materials. Iron oxide (Fe3O4) was used as soft magnetic coating to generate new magnetic properties in the FePt@Fe3O4 coreshell nanoparticles 186, 187. The iron oxide shell was made with a thickness ranging from 0.5 to 3 nm on 4-nm FePt nanoparticles 187. These uniform core-shell nanostructures could be transformed into hard magnetic materials based on the principle of exchange coupling. Iron
112
Shen and Yang
Figure 14. (a-d) TEM images and (e) hysteresis loop of (a-b) FePt@SiO2 105 (Reprinted with permission from Reference 105, © 2008, American Chemical Society) and (c-e) FePt@FexOy 186 core-shell nanoparticles (Reprinted with permission from Reference 186, © 2005, American Chemical Society).
Surface Modifications of Nanoparticles
113
oxide shells acted as both magnetically soft matrix and the barrier materials for preventing FePt cores from coalescence. Liu and others synthesized FePt@Fe3O4 core-shell nanoparticles by a two-step polyol process186. First 2.6-nm FePt nanoparticles were synthesized by reducing of iron and platinum precursors. These FePt particles were then used as cores for the deposition of iron oxides (Figures 14c and 14d). The iron oxide shells could stabilize the FePt cores when the ensembles of particles were treated at 550°C. The magnetic hysteresis of these multiphase nanomaterials showed relatively high coercivity and magnetic remanence in a smooth hysteresis loop, showing the exchange behaviors between magnetically hard cores and soft shells (Figure 14e). Deposition of semiconducting materials on FePt alloy nanoparticles were achieved using a step-wise deposition 188. Figure 15 shows the sequential deposition of sulfur and CdS on FePt nanoparticles. The deposition of sulfur could be due to the formation of both S-Fe and S-Pt bonds. The reactive sulfur shell could then be converted chemically into CdS semiconducting materials in the subsequent step. The FePt cores seemed to be intact throughout this reaction sequence, judging by the TEM images (Figure 15).
Figure 15. Schematic illustration and TEM image of FePt@CdS core-shell nanostructures made from sequential reactions 188 (Reprinted with permission from Reference 188, © 2004, American Chemical Society).
114
Shen and Yang
Besides iron platinum, there are several other magnetic alloys that have been made into nanoparticles 177, 189. The surface modification is generally required before these particles can be used in various applications. Klem et al. developed the method for using protein cages to encapsulate CoPt nanoparticles 190. The interior surface of the protein was thought to strongly bind to fct phase CoPt nanoparticles. The protein cage provided a confined reaction environment and peptides offered the binding sites for nucleation and growth of CoPt nanoparticles. FeCo is another magnetic alloy that has been studied because of its relatively high saturated magnetization when compared to their oxide nanoparticles 191. As-synthesized FeCo nanoparticles however are easy to be oxidized and their chemical instability has limited the applications 192. Surface modification is necessary for stabilizing FeCo nanoparticles from oxidation. Bai et al. synthesized FeCo@Au/Ag core-shell nanoparticles by combining sputtering technique with nanocluster deposition process 193. Gold and silver coating improved the stability and biocompatibility of the magnetic nanoparticles through further modification. Seo et al. used graphitic sheets to coat FeCo nanoparticles through chemical vapor deposition (CVD) method 194. Single graphene sheet could be deposited on the surface of FeCo nanoparticles which had high saturated magnetization for magnetic resonance imaging (MRI) application. The coating also had high optical absorbance in near infrared (NIR) region. Finally, Teng et al. synthesized SmCo5@Fe2O3 core-shell nanoparticles by thermal deposition of iron pentacarbonyl on SmCo5 cores 189. The iron oxide shell protected SmCo5 cores from rapid oxidation in air even under room temperatures.
4. Surface Modification in the Synthesis of Higher-Ordered and Complex Nanostructures Design of multifunctional nanostructures has emerged as an active reach area in recent years. Among them, hollow and yolk-shell structures have been produced and explored for applications in drug delivery and nanoreactor 27, 51, 78, 143, 195-199. Other complex structures such as dumbbell and onion-like structures have also been developed. The introductions of second or third components in the new structures are
Surface Modifications of Nanoparticles
115
often realized based on the strategy for surface modification. We review some of these new nanostructures and synthetic strategies in this section.
4.1. Hollow and Yolk-Shell Nanostructures The recent study on hollow structured particles aims at their potential applications in biological systems 129-131, reactor design 132 and electrochemistry 133, 134. The single-step approaches include nozzle reactor spray drying and emulsion, interfacial polymerization, LbL deposition and self assembly 40, 135-137, 200. They are largely used to produce particles with diameter larger than 100 nm 138. The challenge is to produce structurally well-defined hollow particles with diameter less than 100 nm. Using organic and inorganic templates, either ceramics or polymers can be deposited on the surface through chemical reaction or physical adsorption 107, 129, 139-150. Nanosized hollow structures are generally prepared by removing templates through calcination 143, 144, 147, 148 and etching 107, 129, 139-142, 145, 146, 149. LbL self assembly on templates can be used to adjust the thickness of hollow shells by alternately depositing oppositely charged polyelectrolytes on templates 27, 42, 46, 47. It works very well for micron and sub-micron templates, but is also used on nanometer-sized particles. Self-templating is a technique that combines the emulsion and template 151. In general, surfactants form micelles in aqueous solutions at the proper concentrations first. These micelles can then template the formation of hollow spheres with different composites, sizes and morphologies by adjusting pH and by controlling the polycondensation of inorganic precursors. Further modification with biological or chemical functional groups can be achieved through grafting onto the surface of hollow particles 107. The functionalization of interior surface is an important because many applications of hollow particles aim at taking the advantages of the interior void spaces 152-155. Magnetic particles and polymers have been deposited on the interior surface or in the void space. Jang et al. synthesized nanosized polystyrene hollow spheres using microemulsion polymerization method (Figures 16a and 16b) 140. First, poly(methyl methacrylate) (PMMA) cross linked on the surface of
116
Shen and Yang
micelles, followed by surface deposition of polystyrene to form coreshell spheres in oil-in-water micoremulsion. By etching away the core, PS hollow spheres formed. Khanal and others used poly(styrene-b-2vinyl pyridine-b-ethylene oxide) (PS-PVP-PEO) triblock copolymer as template to make hollow silica nanospheres (Figures 16c and 16d). Silica formed selectively in the PVP block which catalyzed the hydrolysis and polycondensation of tetramethoxysilane (TMOS). The weak interaction of PEO block with silanol groups also facilitated the localized deposition of silica.
Figure 16. Schemes and TEM images showing the formation of hollow nanostructures using micelle and block-copolymer as templates: (a and b) micelle as template 140 (Reprinted with permission from Reference 140, © 2002, American Chemical Society) and (c-d) tri-block copolymer as template 144 (Reprinted with permission from Reference 144, © 2007, American Chemical Society).
Surface Modifications of Nanoparticles
117
Hollow spheres can be produced without the use of emulsion. Figure 17 shows the TEM images of hollow spheres of poly(amic acid) ester (PAE) cross-linked with poly (4-vinylpyridine) (PVPy), PS and palladium hollow spheres made using methods other than microemulsions. Kuang et al. used self-assembly of poly(amic acid) ester (PAE) oligomer and poly(4-vinylpyridine) (PVPy) to synthesize polymeric hollow nanospheres (Figure 17a) 137. Fu et al. show that polystyrene hollow nanospheres could be synthesized with surfaceinitiated atom transfer radical polymerizations on silica templates (Figure 17b) 141. Kim et al. produced palladium hollow spheres though adsorption of palladium precursor-palladium acetylacetonate on the surface of silica templates, followed by the thermal treatment at 250°C and the removal of silica template with 10 M HF (Figure 17c) 133.
Figure 17. TEM images of hollow spheres of (a) PAE-PVPy made through self-assembly 137 (Reprinted with permission from Reference 137, © 2003, Royal Chemical Society); (b) PS made by using silica as templates 141 (Reprinted with permission from Reference 141, © 2005, American Chemical Society); and (c) Pd hollow nanostructures assembled on silica templates 133 (Reprinted with permission from Reference 133, © 2002, American Chemical Society).
When a nanoparticle exists or is introduced in the void region of the hollow sphere, yolk-shell structure forms. Templating 14, controlled etching 136 and growth 201 are the common approaches to the preparation of yolk-shell nanostructures (Figure 18). If the removal of cores can be done in a controllable fashion, both hollow and yolk-shell nanostructures can be made from core-shell nanoparticles. Typically, if the yolk-shell nanoparticles are made from coated nanoparticles, the average size of the
118
Shen and Yang
Figure 18. Schematic illustrations showing the preparation of yolk-shell nanostructures using (a) controlled etching 136 (Reprinted with permission from Reference 136, © 2002, American Chemical Society) and (b) confined growth methods 152 (Reprinted with permission from Reference 152, © 2005, American Chemical Society).
yolks is determined by reaction condition 14, 136. The compositions of the yolks are most likely the same as those of the cores. Synthesis of Ausilica yolk-shell nanostructures could be made through controlled etching of gold cores 14. A more elaborated template-etching approach is developed for the synthesis of Au-mesoporous polymer or carbon yolkshell nanostructures (Figure 18a). The polymers infiltrate into the porous channels of mesoporous silica and can be converted into carbon through pyrolysis if necessary. Finally, silica template is removed, leaving behind the yolk-shell nanostructures. In this approach, both the composition and size of the cores (or yolks) are predetermined. Alternatively, yolk-shell nanostructures can also be synthesized using hollow spheres as nanoreactors. By controlling the nucleation and growth of nanoparticles inside the hollow spheres, yolk can be synthesized with the tunable size and composition 132, 152. Figure 18b illustrates a method for making Ag-PPy-CS (polypyrrole-chitosan) yolkshell structures using this method. By controlling the pH of solutions, Ag+ can be confined insides the PPy-CS hollow spheres and forms silver nanoparticles upon photo-reduction. Emulsion method is another
Surface Modifications of Nanoparticles
119
feasible approach for the synthesis of yolk-shell structures 202. Noticeably, most of shell materials used in these approaches are silica and various polymers 14, 199. Figure 19 shows TEM images of FePt@CoS2 201, Au@SiO2 136, SiO2@SiO2 203, Ag@SiO2 152, Sn@SiO2 134 and Fe2O3@SiO2 yolk-shell nanostructures. Again silica is the widely-used material for constructing the shells. Noticeably, both sub-micron and nanometer-sized yolk-shell particles are made. The synthesis of FePt@CoS2 yolk-shell nanoparticles were developed for biological applications 201. The FePt cores could be released from the CoS2 shells after cellular uptake and showed very high activity and toxicity towards cancer cells.
Figure 19. TEM images of (a) FePt@CoS2 201 (Reprinted with permission from Reference 201, © 2007, American Chemical Society), (b) Au@SiO2 136 (Reprinted with permission from Reference 136, © 2002, American Chemical Society), (c) SiO2@SiO2 203 (Reprinted with permission from Reference 203, © 2008, Wiley-VCH), (d) Ag@SiO2 152 (Reprinted with permission from Reference 152, © 2005, American Chemical Society), (e) Sn@SiO2 134 (Reprinted with permission from Reference 134, © 2003, American Chemical Society), and (f) Fe2O3@SiO2 yolk-shell nanoparticles.
120
Shen and Yang
4.2. Anisotropic and Onion-Like Nanostructures While sphere is the commonly observed morphology for nanoparticles after the surface modification because of the minimization of surface energy, several unsymmetrical shapes have been produced in recent years. The anisotropy in the nanostructures can attribute to the subtle difference in surface structures. Dumbbell-like nanostructures refer to those with two or more components connected through single point and they are developed for their multifunctionality. Shi et al. synthesized PbSe nanocrystals on Au-Fe3O4 dumbbell hybrid nanoparticles using an anisotropic growth method (Figure 20) 204. Peanut-like Au-Fe3O4 nanoparticles were synthesized at 300°C by mixing 3-nm Au nanoparticles with Fe(CO)5 in 1-octadecene using oleic acid and oleylamine as capping agents. Such Au-Fe3O4 binary systems could serve as the seeds for the growth of PbSe nanorods. The morphology of these PbSe nanorods changed with the ratio between PbSe precursors and Au-Fe3O4 particle seeds. Lin et al. used Stöber
Figure 20. TEM image of PbSe nanorods growth on Au-Fe3O4 nanoparticles (Reprinted with permission from Reference 204, © 2006, Wiley-VCH).
204
Surface Modifications of Nanoparticles
121
method to make anisotropic Mag-Dye@MSN nanostructures that combined magnetic (Mag) and photonic (Dye) properties with mesoporous silica nanoparticles (MSN) 205. Briefly, magnetic nanoparticles of iron oxide were first coated with SiO2 layer using Stöber method. These coated nanoparticles further incorporate organic dyes of pre-conjugated N-1-(3-trimethoxy-silylpropyl)-N-fluoresceyl thiourea (FITC-APT-MS), TEOS and cationic surfactant of CTAB in an ammonia solution to form dye-encapsulated mesoporous silica coating. Such MagDye@MSN nanostructure is designed for cell tracking and drug delivery. Multilayered coating is a facile approach to introduce functionality with each layer having different composition and distinctive chemical and physical properties. Such onion-like nanostructures have been made in either one-pot or sequential synthesis. Chen et al. coated the yolk egg-like silica magnetic nanostructures with Ag metal using electroless plating to get the sandwich-type nanostructure 86. The yolk egg-like silica magnetic cores were generated with a one-step sol-gel method. Amino groups were introduced on the particle surface using APTMS followed by the deposition of Au colloidal seeding layer for the growth of Ag shells. This onion-like nanostructure possesses both magnetic and optical properties and can be a potential magnetic-field guided photothermal therapeutic agent. Ji et al. synthesized a similar onion-like nanostructure with gold shell and iron oxide-silica core 206. Gold shell was formed by modifying the iron oxide@SiO2 surface with APTMS and by using seeded growth on the surface. Gu et al. prepared Au-SiO2-PS onion-like nanostructures with Stöber method and seeded polymerization 111. MPTMS was used to first modify the surface of AuSiO2 particles followed by the polymerization of styrene monomer using potassium persulfate as the initiator. Shen et al. synthesized silica-coated Pt@Fe2O3 core-double shell nanostructures using sol-gel approach after surface being activated (Figure 21) 86, 206, 207. The three-component structure was clearly visible in the TEM images. Such Pt@Fe2O3@SiO2 may facilitate the chemical conversion of Pt@Fe2O3 nanoparticles to other important materials, such FePt nanoparticles without the coalescence among the core particles during the heat and other treatments.
122
Shen and Yang
Figure 21. TEM images of Pt@Fe2O3@SiO2 core-double shell nanostructures at (a) low and (b) high magnifications 17 (Reprinted with permission from Reference 17, © 2008, American Chemical Society).
4.3. Other Higher Ordered Nanostructures Higher order of complexity such as assembly of multicomponent nanoparticles can be built into nanostructures through surface modification. The driving force for the assembly among the particles can be either specific interactions through capping ligands or nonspecific interactions through sol-gel and polymeric emulsification in confined spaces. Figure 22 shows TEM images of nanorods on spheres, cat paw, and particles-on-particle satellite and dendrite structures. Gold nanorods were adsorbed on the surface of negatively charged poly(Nisopropylacrylamide) (PNIPAM) microgels 208. The adsorption of gold nanorods was obtained by modifying the Ag nanorods using poly(styrene sulfonate) (PSS) and PAH. The binding efficiency heavily depended on the two oppositely charged species on the particles (Figures 22a-22b). The coverage was affected by the ratio between the micron-sized PNIPAM particles and Au nanorods. The thermal response of the PNIPAM microgel was monitored based on the optical properties of Au nanorods. Other nanostructures such as cat-paw, satellite and dendrimerlike Au nanocomposites were fabricated by modifying 13- and 30-nm Au
Surface Modifications of Nanoparticles
123
nanoparticles with complementary oligonucleotides and then mixing them (Figures 22c-22f) 209.
Figure 22. Schematic illustrations (left side panels) and TEM images (right side panels) showing selective types of higher ordered nanostructures: (a and b) PNIPAM microgels decorated with gold nanorods modified with PSS and PAH 208 (Reprinted with permission from Reference 208, © 2007, Wiley-VCH); (c and d) asymmetrically functionalized Au nanoparticles into cat-paw structures; (e and f) satellite structure; and (g and h) dendrite-like structure 209 (Reprinted with permission from Reference 209, © 2006, American Chemical Society).
124
Shen and Yang
Figure 23. EM micrographs of different types of multifunctional magnetic nanostructures: (a) MNP-QD@SiO2 nanocomposites 210 (Reprinted with permission from Reference 210, © 2005, American Chemical Society); (b) PLGA-encapsulated 15-nm Fe3O4 nanocrystals (MNP/DOXO) 25 (Reprinted with permission from Reference 25, © 2008, Wiley-VCH); (c) Fe3O4@SiO2/p-NIPAM-SiO2-Au 12 (Reprinted with permission from Reference 12, © 2008, American Chemical Society); and (d) 30-nm Au on MMP 209 nanostructures (Reprinted with permission from Reference 209, © 2006, American Chemical Society).
Technique based on the nonspecific interactions in sol-gel process and polymerization can be more broadly applied to various materials with different surface chemistry than those based on the specific molecular binding or via charge interaction such as LbL deposition. The indiscriminative nature of sol-gel process and polymerization can be used within emulsions in colloidal media and thus capable for encapsulation of a broad range of nanomaterials. Figure 23 shows
Surface Modifications of Nanoparticles
125
several representative examples involving magnetic nanoparticles. The magnetic particles and quantum dots can be encapsulated simultaneously in silica spheres (Figure 23a) 210. Poly(D, L-lactic-co-glycolic acid) (PLGA) encapsulated Fe3O4 and doxorubicin nanostructures were synthesized using emulsion method (Figure 23b) 25. More complicated hierarchical assemblies were formed by linking different components with polymer network (Figure 23c), in which silica coated magnetite nanoparticles were surrounded with many small silica spheres with poly(N-isopropylacrylamide) as a cross linker to form the support for Au nanocatalyst. 12 As shown in Figure 23d, Au nanoparticles functionalized with 3’-thiolated and 5’-phosphorylated 15-mer oligonucleotide were self-assembled on the surface of 3’-thiol-terminated 30-mer oligonucleotides functionalized magnetic microparticles (MNP) 209. These multifunctional nanostructures can serve as the platforms for targeted biological imaging and drug delivery 12, 25, 210, and other applications depending on the types of primary nanoparticles. Self-assembly of monodisperse nanoparticles is a recently developed method for making secondary magnetic particles, namely supracrystals. The particles of ordered nanoparticle assembly are mostly in colloidal solutions with various additives to confine and stabilize the supracrystals. Ge et al. synthesized 30 to 189 nm colloidal supraparticles from randomlyarranged monodisperse 10-nm Fe3O4 nanoparticles (Figure 24a) 211. These uniform colloidal nanocrystal clusters (CNC) were synthesized from the hydrolysis of FeCl3 with NaOH at 220°C in diethylene glycol solution containing polyacrylic acid (PAA). The CNCs were stabilized through polyacrylate and showed broad magnetic stop bands which covered the entire visible spectrum. They were proposed for uses in microelectromechanical systems or sensors because of their ability to response reversibly to optical cues 211. Zhuang et al. further developed a method for the formation of crystalline supracrystals of Fe3O4 nanoparticles in different diameters (Figure 24b) 13, 156, 212. The primary Fe3O4 nanoparticles were uniform in size and capped with oleic acid that were replaced by dodecyltrimethylammonium bromide (DTAB) to form nanoparticle-micelle aqueous solutions which were added into ethylene glycol (EG) through injection. DTAB molecules left the surface of Fe3O4 nanoparticles because of their good solubility in EG, and
126
Shen and Yang
Figure 24. Representative TEM images of uniform (a) amorphous 211 (Reprinted with permission from Reference 211, © 2007, Wiley-VCH) and (b) crystalline212 supracrystals of magnetic nanoparticles. (Reprinted with permission from Reference 212, © 2008, Wiley-VCH)
aggregation of iron oxide magnetic particles formed as a consequence. By annealing the disordered aggregates of nanoparticles at 80°C, crystalline supralattices formed. Ge et al. used silica to coat iron oxide clusters through Stöber method and emulsion polymerization with PS subsequent as a means for further modification of magnetic nanoclusters 211. The PS was coated either concentrically or eccentrically depending on if cross-linking agent divinylbenzene (DVB) was applied or not. Without DVB, Fe3O4@SiO2 core-shell nanoparticles were localized eccentrically in nanostructures because of the interfacial tension between hydrophilic Fe3O4@SiO2 coreshell nanoparticles and hydrophobic styrene monomers. DVB was introduced to limit the contraction of polystyrene shell and concentric form of Fe3O4@SiO2 core-shell nanoparticles was dominant in this case. Depending on the amount of styrene monomer, ellipsoid or doublet structures formed. Zigzag chains of the eccentric Fe3O4@SiO2@PS nanostructures formed by applying an external magnetic field. This behavior is very different from the alignment of concentric nanoparticles.
Surface Modifications of Nanoparticles
127
5. Applications of Surface-Modified Magnetic Nanoparticles Development of new surface modification approaches to the control of size, shape, composition and surface chemistry of magnetic nanoparticles is driven in large part by the potential applications. In this context, modification is typically intended for tuning the stability, hydrophilicity and biocompatibility of the surface for various applications including imaging, sensing, separation, and actuation. Each application often requires surface to be treated differently to have the specificity. Ligand exchange usually can improve the stability and hydrophilicity to some extent. Self-assembly of monodispersed magnetic nanoparticles driven by the Van der Waals interactions through surface capping agents can be a useful approach to the bottom-up fabrication of different components for device fabrications. For further modification, nanoparticles need to be covalent bonded with specific functional groups in small molecules or polymers to increase the selectivity, biocompatibility and stability. When single component magnetic nanoparticles cannot fulfill the requirements, multifunctional nanostructures are constructed by introducing other functional moiety.
5.1. Surface Modifications in Nonbiological Applications 5.1.1. Controlling Magnetic Properties through Surface Modifications Coating can serve as an active component in the magnetic nanostructures, as being demonstrated in the FePt or Co@CoO systems for exchange-coupling. For instance, wile several intermetallic forms are possible for FexPty nanoparticles, only the fct phase FePt has the high magnetic anisotropy and is magnetically hard 213. The as-made colloidal low-temperature products of FexPty nanoparticles are typically superparamagnetic and possess fcc phase, even if the alloy is Fe50Pt50, the composition for fct phase. The temperature for transition from fcc to fct phases is generally above 400°C. Core-shell nanostructures of Pt@Fe2O3 and FePt@Fe3O4 were prepared as precursors for making exchange-coupled magnetic materials using the bottom-up approach 186, 187, 213. The advantage of core-shell nanoparticles lies in the
128
Shen and Yang
ability to control multiple phase and composition in the same particles and the potential to conduct chemistry in sub 10-nm-sized regime. Pt@Fe2O3 core-shell nanoparticles could transfer into multiphase nanocomposites containing Fe3Pt, FePt, and FePt3 ordered phases by controlling the diameter of Pt cores and thickness of Fe2O3 shells 186, 187. These FePt nanophases have different coercivity (Figure 25). Similarly, FePt@Fe3O4 core-shell nanoparticles were developed for producing magnetically exchange-coupled nanocomposites 213. The magnetically soft Fe3O4 shell can serve as the matrix for dispersing hard FePt cores to maximize the energy product. Thus, surface modification can be useful as a synthetic approach to structurally complex multifunctional nanocomposites.
Figure 25. Hysteresis loops of different FePt phases obtained from various Pt@Fe2O3 core-shell nanoparticles 213 (Reprinted with permission from Reference 213, © 2005, Institute of Physics).
Surface Modifications of Nanoparticles
129
5.1.2. Stabilization of Magnetic Nanostructures and Ferrofluids The reductive annealing of silica-coated disordered fcc FePt and Pt@Fe2O3 nanoparticles is a good example on using shell as passivitive materials in controlling structural property of the magnetic cores. In this case, the transformations of fcc FePt or Pt@Fe2O3 nanoparticles at high temperatures do not result in the sintering due to the material barriers of silica 214. Silica coating can sometimes affect the magnetic hysteresis of the magnetic nanoparticles such as iron oxide 57, 215-220. Ferrofluid represents the traditional applications of magnetic particles in many different areas that include sensing and actuation. An important reason on conducting the surface modification with the magnetic particle in ferrofluids is to improve their chemical and thermal stability in selective liquid media. The stabilization of ferrofluids can be achieved by replacing original capping agents with functional groups on the surface 221. The formation of core-shell nanostructures changes the surface hydrophilicity and allows for the good dispersibility to meet the requirement of long-term stability in liquid suspensions 221. In those chemically intensive processes such as the magnetorheology finishing (MRF), surface modification can not only stabilize the dispersion of magnetic particles but also protect them from being oxidized or corroded under high temperature and harsh pH conditions.
5.2. Surface Modifications in Biological Applications Magnetic nanoparticles have increasingly been developed for the biomedical applications, particularly in the areas of bioanalysis 56, 61, 113, 217, 222, 223 , bioseparation 113, 223-229, biosensing and imaging 62, 230-234, targeted drug delivery 167, 168, 170, 225, 235-240 and cancer therapy 241. The requirements of surface property for these magnetic nanoparticles are often more rigid than those for nonbiological applications. Depending the end use, nanoparticles may need to be hydrophilic and charge-neutral on the surface to escape the cleaning process by the mononuclear phagocyte system (MPS) in vivo 242, 243. Magnetic nanoparticles have to be small enough to achieve high surface-to-volume ratios while maintain high binding rates with good colloidal stability, which means they need
130
Shen and Yang
to be big enough for ligand binding and multivalent interaction. Size distribution of magnetic nanoparticles after the modifications needs to be narrow enough to have the predictable chemical and physical properties. Currently, magnetic nanoparticles can be synthesized with single nanometer accuracy in diameter ranging from 2 to 20 nm to meet different size requirements. These monodisperse nanoparticles are typically produced from hydrophobic solvents 11. Thus the as-synthesized magnetic nanoparticles from nonhydrolytic solvents tend to have hydrophobic surfaces with low biocompatibility and stability in biological environments and often cannot be used for biological applications directly. Surface modification is essential to develop biocompatible material systems. Better selectivity, targeting and other capability of magnetic nanoparticles are often achieved through the grafting functional molecules on the surfaces 4. Antibody-antigen, other bio-macromolecules and small functional molecules are some of the popular classes of ligands that can be
Figure 26. Schematic illustration of the structural requirement and functionality of magnetic nanoparticles for biological applications 11 (Modified with permission from Reference 11, © 2007, Academic Press Inc, Elsevier Science).
Surface Modifications of Nanoparticles
131
covalently bound on the surface of magnetic nanoparticles. These molecules facilitate further binding to the specific receptors such as antigens and targeted proteins, nucleic acids and so on. Figure 26 illustrates some of the necessary surface modifications of nanoparticles for bioanalysis and therapeutic uses. In this example, PEGylated ligands with thiol groups are covalently attached to the particle surface. The PEG functional groups are introduced to improve the resistance of unspecific protein binding, while carboxylic acid groups can further interact with amine-based functional group through electrostatic attraction. Such multifunctional coating results in the proper binding of target biological entity, such as α-chymotrypsin (ChT) 11. 5.2.1. Biological Imaging and Sensing Biological tagging and labeling for MRI is one of the most heavily researched areas among the various biological applications of magnetic nanoparticles, because they offer the advantage of near microscopic resolution to differentiate healthy and pathological tissues 5, 6, 206, 226, 244. The contrast of MRI rises from the difference in signal intensity of tissues in responding to applied radiofrequency fields. Two parameters, i.e. proton density and magnetic relaxation times are important in the response. New magnetic agents are developed to improve imaging contrast and accuracy. Iron oxide nanoparticles are the most often used contrast agents because of their good magnetic property, chemical stability, availability and well-developed protocols for the surface modifications 194. Other magnetic materials such as metals and FeCo alloys cannot be directly applied in MRI imaging because they are not stable and prompted to surface oxidation even at low temperatures. Surface modification can protect the nanoparticles from being oxidized in order to preserve the excellent magnetic property suitable for MRI applications. Dai et al. modified FeCo nanoparticles with graphitic shell 245. The coated FeCo nanoparticles exhibit high saturated magnetization along with long r1 and r2 relaxation times. In vivo MRI experiments indicated long-lasting positive contrast enhancement in rabbits by using these surface-modified contrast agents. Additionally, the graphite-coated
132
Shen and Yang
FeCo nanoparticles showed an absorbance at 808 nm in the near-infrared region, which is useful for controlled drug release and cancer therapy. Other core-shell nanostructures have also been developed for MRI and fluorescence imaging 15. Fe3O4 nanoparticles, the magnetically active component, were coated with mesoporous silica shells using sol-gel method. In vivo imaging showed that these nanoparticles tended to accumulated at tumor sites due to their enhanced permeability and retention effect, which can be detected with T2-weighted MRI technique 15. These mesoporous silica (mSiO2)-coated Fe3O4 nanoparticles were also good candidate for fluorescence imaging by covalently incorporating fluorescein isothiocyanate (FITC) and rhodamine B isothiocyanate (RITC) into silica walls. Figure 27 shows the fluorescence imaging applications of these multifunctional nanoparticles which were injected into the nude mice bearing tumors.
Figure 27. Fluorescent micrograph of Fe3O4@mSiO2 nanostructures in sectioned tumors: (a) immuno-staining of vasculature (brown) with anti-CD31 antibody and counterstaining of nucleus with hematocylin (blue), (b) distribution of R-rhodamine B, and (c) image by merging those of (a) and (b) (Modified with permission from Reference 15, © 2008, Wiley-VCH).
Magnetic nanoparticles can often combined with quantum dots or other luminescent labels to serve as dual functional probes in bioanalysis 246. Etgar et al. synthesized hierarchical nanostructure of γ-Fe2O3 and PbSe nanoparticles for potential biological detection 247. Choi et al. synthesized bifunctional nanocomposites consisting of silicacoated magnetic cores and surface-anchored luminescent lanthanide irons 5, 217, 231, 233, 239. Photoluminescence was dramatically enhance
Surface Modifications of Nanoparticles
133
because of a ligand-to-metal energy transfer between 2,2’-bipyridine4,4’-dicarboxylic acid (BDA) and Ln metal ion. Such bi-functional nanostructures can be imaged based on the luminescent markers and controlled through an external magnetic field. 5.2.2. Controlled Drug Targeting and Releasing Carriers Oral and injecting administrations are major traditional approaches for drug delivery, although most drugs formulated for oral delivery or injection routes are low efficiency and lack of the ability in targeting the specific areas of body. New delivery techniques and drug formations are developed to address the lack of specificity and reduce side effects. Magnetic nanoparticle is a possible candidate as active controlling component for targeted drug delivery and has been studied in drug formation since 1970s 5, 6. Drug molecules can be loaded on exterior or interior of various magnetic nanostructures 10. Ideally, magnetic carriers need to deliver drugs specifically to the target organs with long circulation time driven by an externally magnetic field. The release profiles need be controlled with sufficient accuracy. Another common requirement is the size of magnetic nanoparticles in drug delivery should be smaller than about tens of nanometers to keep good superparamagnetic property and free of uncontrolled aggregation in various dispersion media and biological environments 10. Those magnetic nanoparticles less than 10 nm are easy to be removed by extravasations and renal systems, thus the optimal final size of the particle has to be above 10 nm 62. Surface modification can circumvent some of the size requirements by generating superparamagnetic nanoclusters with diameter larger than 100 nm and through functionalizing the surface with molecules for long circulation time and good specificity 156, 157, 212. As in other cases, ligand exchange and surface conjugation can sometimes be sufficient to adjust the surface chemistry of magnetic nanoparticles to meet the requirements for hydrophilicity, biocompatibility and selectivity in formulating the drugs. A range of pharmaceutical or biological compounds can be cross-linked to magnetic nanoparticles through silane functional groups 62. Among the coupling
134
Shen and Yang
agents, amino conjugated silane is popular in the surface modification of nanoparticles for biological applications as the amino group has strong affinity to various biomolecules 62. In vivo tests show that low or no toxicity exists for the amino-silane functionalized magnetic nanoparticles 233. Charge on the surface of magnetic nanoparticles is important in the control of drug delivery. Takeda et al. developed a drug delivery system composed of protamine sulfate-modified magnetic nanoparticles which were also associated with plasid DNA 230. In vitro cell culture study showed that cationic charge generated from protamine sulfate greatly enhanced the transfection efficiency. Gupta et al. used PEG to modify superparamagnetic iron oxide nanoparticles (SPION) for site specific delivery of drugs 78, 79. PEG is widely used in surface modification of nanoparticles in drug delivery because of its good biocompatibility and ability to prevent unspecific binding. Magnetic nanoparticles coated with PEG can have long circulation time due to the resistance of the protein adsorption 230. Studies shows that PEG modified SPION particles affected the adhesion/viability, morphology and cytoskeletal organization of human dermal fibroblasts 248. Like dendrimers where molecules can be incorporated into the structures via either complexation or encapsulation 249, multicomponent magnetic nanoparticles such as dumbbell or tripods can provide the type of platform for multi-functionality in functions of drug delivery. Heterodimers of nanoparticles with different compositions and shapes have also been reported 1, 250. Two different kinds of functional molecules can be covalently bonded to the two different components of the heterodimers in a particle-specific fashion. With the improvement in control of the surface modification and synthesis, more complex magnetic nanostructures can be expected for drug delivery. 5.2.3. Hyperthermia and Caner Therapy The treatment of malignant tumors with hyperthermia is among the actively studied cancer therapeutic methods. The hyperthermia therapy is based on the observations that tumor cells are more sensitive to the change of temperature when compared with normal tissue cells 250. It is
Surface Modifications of Nanoparticles
135
more benign than other treatments such as chemotherapy and radiotherapy, because it causes fewer side effects. Magnetic fluid can produce uniform thermal energy in localized regions under an external AC magnetic field 251. An interesting research direction is the combination of hyperthermia treatment with chemotherapy or gene therapy 251. Ito and others used magnetite cationic liposomes (MCLs) for hyperthermia therapy under an alternating magnetic field. The heat triggered a three-fold increase in the tumor necrosis factor (TNF)-α gene expression which caused cell death 10, 170, 240, 252, 253. Controlling the interaction of magnetic nanoparticles with cells and their intracellular uptake though surface modification plays a great role in the future cancer diagnosis and therapy using drug, protein or nucleic acid loaded magnetic nanoparticles 252. Different polymers, surfactants and biomolecules have been used for targeting cellular delivery of magnetic nanoparticles. Yu et al. synthesized thermally cross-linked (TCL) SPION which could be accumulated in Lewis lung carcinoma (LLC) tumor by enhanced permeability and retention effect 253. Doxorubicin was loaded on the surface in order to evaluating its toxicity and ability in cancer therapy. These systems showed excellent tumor targeting efficiency and were studied as contrast agents for MRI and drug carriers. Shukoor et al. developed nanocomposites targeted for kidney cancer cells, Caki-1 therapy with high selectivity and low toxicity 201. They used multifunctional polymeric ligand containing 3hydroxytyramine (dopamine), piperazinyl-4-chloro-7-nitrobenzofurazane (pipNBD) and a free amine group to functionalize γ-Fe2O3 nanoparticles for immobilization of dsRNA poly (I:C) (polyinosinic-polycytidyl acid) in diagnosis of Caki-1 cells. The polymeric ligand worked as a crosslinker and provided anchor groups for magnetic nanoparticles and dsRNA. The FePt@CoS2 yolk-shell nanostructures were studied as an anticancer drug 201, although the design was not based on the magnetic property of FePt nanoparticles. FePt@CoS2 nanoparticles decorated with cancer-targeting antibodies caused extensive Hela cells death (Figure 28). The high cytotoxicity of FePt@CoS2 yolk-shell nanoparticles against Hela cells was thought to be the result of dissolution of FePt cores under acidic environment when the particles were inside the secondary
136
Shen and Yang
Figure 28. Cytotoxicity analyses of FePt@CoS2 yolk-shell nanoparticles on Hela cells: (a) after incubated for 72 h with 5 µg/mL FePt@CoS2 nanocrystals and (b) negative control after 48 h with only the growth medium 201. (Reprinted with permission from Reference 201, © 2007, American Chemical Society.)
lysosomes. Platinum ions were responsible for changing the DNA double-helices of Hela cells. 5.2.4. Other Biological Applications Magnetic nanoparticles have also applied to purification, immunoassay, immobilization and amplification. Surface functionalization of magnetic nanoparticles provides an approach for specificity in rapid removal of targeted biomaterials from blood cells or other biological media by applying external magnetic fields 61. Mikhaylova et al. immobilized bovine serum albumin (BSA) on APTMS-modified SPION 56, 231, 254. Based on thermogravimetric and chemical analyses, APTMS-modified SPION showed a higher immobilization capacity than those using the co-precipitated SPION and BSA. In order to increase the efficiency and selectivity in protein or DNA separation, the surface of magnetic nanoparticles is usually modified with complementary sequences of oligonucletide. For example, 5’-amine modified dC6dT25 oligonucleotide was grafted on the activated surface by reacting with the aminopropyltriethoxysilane (APTS) modified nanoparticles which have been treated with glutaraldehyde to capture complementary fluorescencelabeled oligonucleotide through hybridization 254.
Surface Modifications of Nanoparticles
137
6. Conclusion Surface modification is an effective way to not only change the chemical and physical properties of nanoparticles but also introduce new functional components to be complimentary with the magnetic cores. Strategies developed for magnetic and non-magnetic materials can often be used inter-changeably. Recently major advances are made in the areas of developing multifunctionality and supra-structures through surface modifications. While the industrial applications of surface modification of magnetic nanoparticles are focused traditionally on chemical and thermal stability and use relatively simple approaches, the recent advances in biological and other advanced applications call for much stringent and specific requirements. Solvent dispersity and structural diversity in core-shell nanostructures are just a few of the new emerging topics. The novel applications of magnetic nanoparticles in both biological and other areas will rely on the fine controls of not only the core materials themselves but also the surface chemistry.
Acknowledgments We thank U.S. National Science Foundation (DMR-0449849), Environmental Protection Agency (RD-83172201-0) and National Institute of Health (R01 CA134218) for supports.
References 1. 2. 3. 4. 5. 6. 7. 8.
U. Jeong, X. W. Teng, Y. Wang, H. Yang and Y. N. Xia, Adv. Mater., 19, 33 (2007). K. C. R. Bahadur, S. Aryal, S. R. Bhattarai, N. Bhattarai, C. H. Kim and H. Y. Kim, J. Biomater. Sci.-Polym. Ed., 17, 579 (2006). Y. Y. Liu, H. Miyoshi and M. Nakamura, Int. J. Cancer, 120, 2527 (2007). C. Lang, D. Schuler and D. Faivre, Macromol. Biosci., 7, 144 (2007). G. F. Goya, V. Grazu and M. R. Ibarra, Curr. Nanosci., 4, 1 (2008). E. Duguet, S. Vasseur, S. Mornet and J. M. Devoisselle, Nanomedicine, 1, 157 (2006). S. Peng, C. Wang, J. Xie and S. Sun, J. Am. Chem. Soc., 128, 10676 (2006). U. Wiedwald, K. Fauth, M. Hessler, H. G. Boyen, F. Weigl, M. Hilgendorff, M. Giersig, G. Schutz, P. Ziemann and M. Farle, ChemPhysChem, 6, 2522 (2005).
138 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
Shen and Yang D. C. Lee, F. V. Mikulec, J. M. Pelaez, B. Koo and B. A. Korgel, J. Phys. Chem. B., 110, 11160 (2006). H. W. Gu, K. M. Xu, C. J. Xu and B. Xu, Chem. Commun., 941 (2006). C. C. You, A. Chompoosor and V. M. Rotello, Nano Today, 2, 34 (2007). J. P. Ge, T. Huynh, Y. P. Hu and Y. D. Yin, Nano Lett., 8, 931 (2008). J. P. Ge, Y. X. Hu, T. R. Zhang and Y. D. Yin, J. Am. Chem. Soc., 129, 8974 (2007). J. Lee, J. C. Park and H. Song, Adv. Mater., 20, 1523 (2008). J. Kim, H. S. Kim, N. Lee, T. Kim, H. Kim, T. Yu, I. C. Song, W. K. Moon and T. Hyeon, Angew. Chem. Int. Edit., 9999, NA (2008). K. Woo and J. W. Hong, IEEE Trans. Magn., 41, 4137 (2005). R. Shen, P. H. C. Camargo, Y. Xia and H. Yang, Langmuir, 24, 11189 (2008). J. M. Bergen, H. A. Von Recum, T. T. Goodman, A. P. Massey and S. H. Pun, Macromol. Biosci., 6, 506 (2006). L. M. LizMarzan, M. Giersig and P. Mulvaney, Langmuir, 12, 4329 (1996). A. M. De Campos, A. Sanchez, R. Gref, P. Calvo and M. J. Alonso, Eur. J. Pharm. Sci., 20, 73 (2003). Z. W. Deng, M. Chen and L. M. Wu, J. Phys. Chem. C, 111, 11692 (2007). J. F. Ding and J. S. Jiang, J. Inorg. Mater., 22, 859 (2007). K. Aslan, C. C. Luhrs and V. H. Perez-Luna, J. Phys. Chem. B, 108, 15631 (2004). Y. Wang, X. W. Teng, J. S. Wang and H. Yang, Nano Lett., 3, 789 (2003). J. Kim, J. E. Lee, S. H. Lee, J. H. Yu, J. H. Lee, T. G. Park and T. Hyeon, Adv. Mater., 20, 478 (2008). F. Caruso, Adv. Mater., 13, 11 (2001). F. Caruso, Chem.-Eur. J., 6, 413 (2000). F. Caruso, E. Donath and H. Mohwald, J. Phys. Chem. B, 102, 2011 (1998). G. A. Diaz-Quijada and D. D. M. Wayner, Langmuir, 20, 9607 (2004). K. P. Lee, A. I. Gopalan, P. Santhosh, K. M. Manesh, J. H. Kim and K. S. Kim, J. Nanosci. Nanotechnol., 6, 1575 (2006). L. Bin, L. Jun and J. H. Li, Chin. J. Chem. Phys., 20, 495 (2007). T. Y. Liu, L. Y. Huang, S. H. Hu, M. C. Yang and S. Y. Chen, J. Biomed. Nanotechnol., 3, 353 (2007). N. Coowanitwong, C. Y. Wu, M. Cai, M. Ruthkosky, J. Rogers, L. Feng, S. Watano and T. Yoshida, J. Nanopart. Res., 5, 247 (2003). M. Aslam, L. Fu, S. Li and V. P. Dravid, J. Colloid Interface Sci., 290, 444 (2005). J. E. Fuller, G. T. Zugates, L. S. Ferreira, H. S. Ow, N. N. Nguyen, U. B. Wiesner and R. S. Langer, Biomaterials, 29, 1526 (2008). T. I. Armstrong, M. C. Davies and L. Illum, J. Drug Target., 4, 389 (1997). A. S. Ethiraj, N. Hebalkar, S. R. Sainkar, J. Urban and S. K. Kulkarni, Surf. Eng., 20, 367 (2004). J. H. Gao, W. Zhang, P. B. Huang, B. Zhang, X. X. Zhang and B. Xu, J. Am. Chem. Soc., 130, 3710 (2008).
Surface Modifications of Nanoparticles 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.
139
D. X. Li, Q. He, H. F. Zhu, C. Tao and J. B. Li, J. Nanosci. Nanotechnol., 7, 3089 (2007). D. E. Bergbreiter, Angew. Chem.-Int. Edit., 38, 2870 (1999). H. Ai, S. A. Jones and Y. M. Lvov, Cell Biochem. Biophys., 39, 23 (2003). F. Caruso, M. Spasova, V. Saigueirino-Maceira and L. M. Liz-Marzan, Adv. Mater., 13, 1090 (2001). Y. Wang, J. F. Wong, X. W. Teng, X. Z. Lin and H. Yang, Nano Lett., 3, 1555 (2003). A. B. Artyukhin, O. Bakajin, P. Stroeve and A. Noy, Langmuir, 20, 1442 (2004). F. Caruso and H. Mohwald, Langmuir, 15, 8276 (1999). T. Cassagneau and F. Caruso, J. Am. Chem. Soc., 124, 8172 (2002). D. I. Gittins and F. Caruso, Adv. Mater., 12, 1947 (2000). G. Decher, Science, 277, 1232 (1997). Y. Wang, A. S. Angelatos and F. Caruso, Chem. Mat., 20, 848 (2008). Y. J. Wang, V. Bansal, A. N. Zelikin and F. Caruso, Nano Lett., 8, 1741 (2008). F. Caruso, R. A. Caruso and H. Mohwald, Science, 282, 1111 (1998). C. Flesch, C. Delaite, P. Dumas, E. Bourgeat-Lami and E. Duguet, J. Polym. Sci. Pol. Chem., 42, 6011 (2004). A. Aumelas, A. Serrero, A. Durand, E. Dellacherie and M. Leonard, Colloid Surf. B-Biointerfaces, 59, 74 (2007). E. J. Tang, G. X. Cheng, X. L. Ma, X. S. Pang and Q. Zhao, Appl. Surf. Sci., 252, 5227 (2006). T. Tsuruoka, K. Akamatsu and H. Nawafune, Langmuir, 20, 11169 (2004). I. J. Bruce and T. Sen, Langmuir, 21, 7029 (2005). R. De Palma, S. Peeters, M. J. Van Bael, H. Van den Rul, K. Bonroy, W. Laureyn, J. Mullens, G. Borghs and G. Maes, Chem. Mat., 19, 1821 (2007). H. Iida, T. Nakanishi and T. Osaka, Electrochim. Acta, 51, 855 (2005). S. Park and M. J. Weaver, J. Phys. Chem. B, 106, 8667 (2002). D. K. Kim, M. Mikhaylova, Y. Zhang and M. Muhammed, Chem. Mat., 15, 1617 (2003). M. Mikhaylova, D. K. Kim, C. C. Berry, A. Zagorodni, M. Toprak, A. S. G. Curtis and M. Muhammed, Chem. Mat., 16, 2344 (2004). S. Moritake, S. Taira, Y. Ichiyanagi, N. Morone, S. Y. Song, T. Hatanaka, S. Yuasa and M. Setou, J. Nanosci. Nanotechnol., 7, 937 (2007). M. Marini, B. Pourabbas, F. Pilati and P. Fabbri, Colloid Surf. A-Physicochem. Eng. Asp., 317, 473 (2008). E. Ukaji, T. Furusawa, M. Sato and N. Suzuki, Appl. Surf. Sci., 254, 563 (2007). J. Dai, L. Guo, Y. Jiang, Q. Y. Zhu, R. A. Gu, D. X. Jia and W. J. Guo, J. Nanosci. Nanotechnol., 5, 474 (2005). D. Jiang, Y. Xu, B. Hou, D. Wu and Y. H. Sun, J. Solid State Chem., 180, 1787 (2007).
140
Shen and Yang
67.
W. H. Binder, L. Petraru, R. Sachenshofer and R. Zirbs, Mon. Chem., 137, 835 (2006). W. H. Binder and H. C. Weinstabl, Mon. Chem., 138, 315 (2007). M. Dasog, A. Kavianpour, M. F. Paige, H. B. Kraatz and R. W. J. Scott, Can. J. Chem.-Rev. Can. Chim., 86, 368 (2008). M. A. White, J. A. Johnson, J. T. Koberstein and N. J. Turro, J. Am. Chem. Soc., 128, 11356 (2006). H. Mader, X. H. Li, S. Saleh, M. Link, P. Kele and O. S. Wolfbeis, Ann. NY. Acad. Sci., 1130, 218 (2008). J. Xie, C. J. Xu, Z. C. Xu, Y. L. Hou, K. L. Young, S. X. Wang, N. Pourmond and S. H. Sun, Chem. Mat., 18, 5401 (2006). S. Kim and M. G. Bawendi, J. Am. Chem. Soc., 125, 14652 (2003). L. Chen, J. Zhu, Q. Li, S. Chen and Y. R. Wang, Eur. Polym. J., 43, 4593 (2007). E. Dzunuzovic, M. Marinovic-Cincovic, K. Jeremic, J. Vukovic and J. Nedeljkovic, Polym. Degrad. Stabil., 93, 77 (2008). E. Dzunuzovic, K. Jeremic and J. M. Nedeljkovic, Eur. Polym. J., 43, 3719 (2007). J. W. Hotchkiss, A. B. Lowe and S. G. Boyes, Chem. Mat., 19, 6 (2007). S. S. Davis, L. Illum and S. Stolnik, Curr. Opin. Colloid Interface Sci., 1, 660 (1996). C. Flesch, Y. Unterfinger, E. Bourgeat-Lami, E. Duguet, C. Delaite and P. Dumas, Macromol. Rapid. Comm., 26, 1494 (2005). C. X. Zhang, S. O'Brien and L. Balogh, J. Phys. Chem. B, 106, 10316 (2002). R. Matmour and Y. Gnanou, J. Am. Chem. Soc., 130, 1350 (2008). C. Gruttner, V. Bohmer, A. Casnati, J. F. Dozol, D. N. Reinhoudt, M. M. ReinosoGarcia, S. Rudershausen, J. Teller, R. Ungaro, W. Verboom and P. S. Wang, J. Magn. Magn. Mater., 293, 559 (2005). V. Salgueirino-Maceira and M. A. Correa-Duarte, Adv. Mater., 19, 4131 (2007). V. Salgueirino-Maceira, F. Caruso and L. M. Liz-Marzan, J. Phys. Chem. B, 107, 10990 (2003). E. Bourgeat-Lami, J. Nanosci. Nanotechnol., 2, 1 (2002). M. Chen, Y. N. Kim, H. M. Lee, C. Li and S. O. Cho, J. Phys. Chem. C, 112, 8870 (2008). W. Stober, A. Fink and E. Bohn, J. Colloid. Interf. Sci., 26, 62 (1968). S. Ramesh, E. Sominska and A. Gedanken, Ultrason. Sonochem., 9, 61 (2002). W. J. Chen, P. J. Tsai and Y. C. Chen, Small, 4, 485 (2008). B. C. Jeffrey, Boston : Academic Press, (1990). G. C. Chen, C. Y. Kuo and S. Y. Lu, J. Am. Ceram. Soc., 88, 277 (2005). M. Giersig, T. Ung, L. M. LizMarzan and P. Mulvaney, Adv. Mater., 9, 570 (1997). S. H. Liu and M. Y. Han, Adv. Funct. Mater., 15, 961 (2005). X. P. Zhou, X. Q. Wang, F. X. Liu, Z. L. Chen and A. Kasuya, Curr. Nanosci., 4, 88 (2008). C. Y. Lo, W. Y. Chen, C. T. Chen and Y. C. Chen, J. Proteome Res., 6, 887 (2007).
68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.
83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95.
Surface Modifications of Nanoparticles 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122.
141
M. Souza, D. A. G. Aranda, C. A. C. Perez and M. Schmal, Physica Status Solidi aApplied Research, 187, 297 (2001). N. I. Baklanova, B. N. Zaitsev and A. T. Titov, J. Eur. Ceram. Soc., 27, 2503 (2007). L. Kumari, W. Li and D. Wang, Nanotechnology, 19, (2008). C. Suciu, L. Gagea, A. C. Hoffmann and M. Mocean, Chem. Eng. Sci., 61, 7831 (2006). B. Tyagi, K. Sidhpuria, B. Shaik and R. V. Jasra, Ind. Eng. Chem. Res., 45, 8643 (2006). C. Viazzi, J. P. Bonino and F. Ansart, Surf. Coat. Tech., 201, 3889 (2006). C. Viazzi, A. Deboni, J. Z. Ferreira, J. P. Bonino and F. Ansart, Solid State Sci., 8, 1023 (2006). R. I. Nooney, T. Dhanasekaran, Y. M. Chen, R. Josephs and A. E. Ostafin, Adv. Mater., 14, 529 (2002). Y. D. Yin, Y. Lu, Y. G. Sun and Y. N. Xia, Nano Lett., 2, 427 (2002). S. C. Tsang, C. H. Yu, H. L. Tang, H. He, V. Castelletto, I. W. Hamley, T. Narayanan, C. C. H. Lo and K. Tam, Chem. Mater., 20, 4554 (2008). C. H. Yu, N. Caiulo, C. C. H. Lo, K. Tam and S. C. Tsang, Adv. Mater., 18, 2312 (2006). W. J. Rieter, K. M. L. Taylor and W. B. Lin, J. Am. Chem. Soc., 129, 9852 (2007). D. C. Lee, F. V. Mikulec, J. M. Pelaez, B. Koo and B. A. Korgel, 110, 11160 (2006). R. Bakalova, Z. Zhelev, I. Aoki, K. Masamoto, M. Mileva, T. Obata, M. Higuchi, V. Gadjeva and I. Kanno, Bioconjugate Chem., 19, 1135 (2008). R. Bakalova, Z. Zhelev, I. Aoki, H. Ohba, Y. Imai and I. Kanno, Anal. Chem., 78, 5925 (2006). S. Gu, J. Onishi, E. Mine, Y. Kobayashi and M. Konno, J. Colloid Interface Sci., 279, 284 (2004). Y.-L. Shi and T. Asefa, Langmuir, 23, 9455 (2007). X. Hong, J. Li, M. J. Wang, J. J. Xu, W. Guo, J. H. Li, Y. B. Bai and T. J. Li, Chem. Mat., 16, 4022 (2004). H. G. Bagaria, E. T. Ada, M. Shamsuzzoha, D. E. Nikles and D. T. Johnson, Langmuir, 22, 7732 (2006). N. Pradhan, D. M. Battaglia, Y. C. Liu and X. G. Peng, Nano Lett., 7, 312 (2007). L. O. Brown and J. E. Hutchison, J. Am. Chem. Soc., 121, 882 (1999). Y. C. Chung, I. H. Chen and C. J. Chen, Biomaterials, 29, 1807 (2008). R. Sardar, J. W. Park and J. S. Shumaker-Parry, Langmuir, 23, 11883 (2007). M. G. Warner, S. M. Reed and J. E. Hutchison, Chem. Mater., 12, 3316 (2000). G. H. Woehrle, L. O. Brown and J. E. Hutchison, J. Am. Chem. Soc., 127, 2172 (2005). J. Liu, R. L. Xu and A. E. Kaifer, Langmuir, 14, 7337 (1998). R. P. Bagwe, C. Yang, L. R. Hilliard and W. Tan, Langmuir, 20, 8336 (2004).
142
Shen and Yang
123. A. Binder, A. Heel and G. Kasper, Chem. Vapor Depos., 13, 48 (2007). 124. W. Choi, J. Lee, S. Kim, S. Hwang, M. C. Lee and T. K. Lee, J. Ind. Eng. Chem., 9, 96 (2003). 125. J. C. Gao, J. Zou, X. W. Tan and Y. Wang, Trans. Nonferrous Met. Soc. China, 16, 1252 (2006). 126. L. F. Hakim, J. H. Blackson and A. W. Weimer, Chem. Eng. Sci., 62, 6199 (2007). 127. R. P. Hodgkins, A. Ahniyaz, K. Parekh, L. M. Belova and L. Bergstrom, Langmuir, 23, 8838 (2007). 128. L. Q. Jing, B. Q. Wang, B. F. Xin, S. D. Li, K. Y. Shi, W. M. Cai and H. G. Fu, J. Solid State Chem., 177, 4221 (2004). 129. M. Darbandi, R. Thomann and T. Nann, Chem. Mat., 19, 1700 (2007). 130. J. F. Ding and G. J. Liu, J. Phys. Chem. B, 102, 6107 (1998). 131. Y. Hu, Y. Ding, D. Ding, M. J. Sun, L. Y. Zhang, X. Q. Jiang and C. Z. Yang, Biomacromolecules, 8, 1069 (2007). 132. S. Cavaliere-Jaricot, M. Darbandi and T. Nann, Chem. Commun., 2031 (2007). 133. S. W. Kim, M. Kim, W. Y. Lee and T. Hyeon, J. Am. Chem. Soc., 124, 7642 (2002). 134. K. T. Lee, Y. S. Jung and S. M. Oh, J. Am. Chem. Soc., 125, 5652 (2003). 135. S. Stewart and G. J. Liu, Chem. Mater., 11, 1048 (1999). 136. M. Kim, K. Sohn, H. Bin Na and T. Hyeon, Nano Lett., 2, 1383 (2002). 137. M. Kuang, H. W. Duan, J. Wang, D. Y. Chen and M. Jiang, Chem. Commun., 496 (2003). 138. H. Y. Huang, E. E. Remsen, T. Kowalewski and K. L. Wooley, J. Am. Chem. Soc., 121, 3805 (1999). 139. K. Zhang, L. L. Zheng, X. H. Zhang, X. Chen and B. Yang, Colloid Surface A, 277, 145 (2006). 140. J. Jang and H. Ha, Langmuir, 18, 5613 (2002). 141. G. D. Fu, Z. H. Shang, L. Hong, E. T. Kang and K. G. Neoh, Macromolecules, 38, 7867 (2005). 142. G. D. Fu, Z. H. Shang, L. Hong, E. T. Kang and K. G. Neoh, Adv. Mater., 17, 2622 (2005). 143. C. E. Fowler, D. Khushalani and S. Mann, J. Mater. Chem., 11, 1968 (2001). 144. A. Khanal, Y. Inoue, M. Yada and K. Nakashima, J. Am. Chem. Soc., 129, 1534 (2007). 145. S. M. Marinakos, J. P. Novak, L. C. Brousseau, A. B. House, E. M. Edeki, J. C. Feldhaus and D. L. Feldheim, J. Am. Chem. Soc., 121, 8518 (1999). 146. K. Kamata, Y. Lu and Y. N. Xia, J. Am. Chem. Soc., 125, 2384 (2003). 147. A. B. Bourlinos, M. A. Karakassides and D. Petridis, Chem. Commun., 1518 (2001). 148. Z. Y. Zhong, Y. D. Yin, B. Gates and Y. N. Xia, Adv. Mater., 12, 206 (2000). 149. T. K. Mandal, M. S. Fleming and D. R. Walt, Chem. Mat., 12, 3481 (2000). 150. G. Y. Liu, H. Zhang, X. L. Yang and Y. M. Wang, Polymer, 48, 5896 (2007).
Surface Modifications of Nanoparticles
143
151. Q. B. Wang, L. Yan and H. Yan, Chem. Commun., 2339 (2007). 152. D. M. Cheng, X. D. Zhou, H. B. Xia and H. S. O. Chan, Chem. Mat., 17, 3578 (2005). 153. A. Shukla, P. Degen and H. Rehage, J. Am. Chem. Soc., 129, 8056 (2007). 154. Y. Yin, Y. Lu, B. Gates and Y. Xia, Chem. Mat., 13, 1146 (2001). 155. L. Dahne, S. Leporatti, E. Donath and H. Mohwald, J. Am. Chem. Soc., 123, 5431 (2001). 156. J. Q. Zhuang, H. M. Wu, Y. A. Yang and Y. C. Cao, J. Am. Chem. Soc., 129, 14166 (2007). 157. J. P. Ge, Y. X. Hu, M. Biasini, W. P. Beyermann and Y. D. Yin, 46, 4342 (2007). 158. R. Hofer, M. Textor and N. D. Spencer, Langmuir, 17, 4014 (2001). 159. C. T. Chen and Y. C. Chen, Anal. Chem., 77, 5912 (2005). 160. J. C. Liu, P. J. Tsai, Y. C. Lee and Y. C. Chen, Anal. Chem., 80, 5425 (2008). 161. Y. Li, J. S. Wu, D. W. Qi, X. Q. Xu, C. H. Deng, P. Y. Yang and X. M. Zhang, Chem. Commun., 564 (2008). 162. Y. K. Guo, M. Y. Wang, H. Q. Zhang, G. D. Liu, L. Q. Zhang and X. W. Qu, J. Appl. Polym. Sci., 107, 2671 (2008). 163. C. Ladaviere, M. C. Averlant-Petit, O. Fabre, A. Durand, E. Dellacherie and E. Marie, Colloid Polym. Sci., 285, 621 (2007). 164. Y. D. Luo, C. A. Dai and W. Y. Chiu, J. Polym. Sci. Pol. Chem., 46, 1014 (2008). 165. P. Liu, W. M. Liu and Q. J. Xue, Des. Monomers Polym., 7, 253 (2004). 166. T. Dey, J. Nanosci. Nanotechnol., 6, 2479 (2006). 167. S. C. Wuang, K. G. Neoh, E. T. Kang, D. W. Pack and D. E. Leckband, Macromol. Rapid Commun., 28, 816 (2007). 168. S. C. Wuang, K. G. Neoh, E. T. Kang, D. W. Pack and D. E. Leckband, J. Mater. Chem., 17, 3354 (2007). 169. F. X. Hu, K. G. Neoh, L. Cen and E. T. Kang, Biomacromolecules, 7, 809 (2006). 170. Y. Zhang and J. Zhang, J. Colloid Interface Sci., 283, 352 (2005). 171. Y. Zhou, S. X. Wang, B. J. Ding and Z. M. Yang, Chem. Eng. J., 138, 578 (2008). 172. H. Yu, M. Chen, P. M. Rice, S. X. Wang, R. L. White and S. H. Sun, Nano Lett., 5, 379 (2005). 173. Z. C. Xu, Y. L. Hou and S. H. Sun, J. Am. Chem. Soc., 129, 8698 (2007). 174. I. S. Lee, N. Lee, J. Park, B. H. Kim, Y. W. Yi, T. Kim, T. K. Kim, I. H. Lee, S. R. Paik and T. Hyeon, J. Am. Chem. Soc., 128, 10658 (2006). 175. T. Seto, H. Akinaga, F. Takano, K. Koga, T. Orii and M. Hirasawa, J. Phys. Chem. B, 109, 13403 (2005). 176. J. B. Tracy and M. G. Bawendi, Phys Rev B, 74, (2006). 177. Y. L. Hou, Z. C. Xu, S. Peng, C. B. Rong, J. P. Liu and S. H. Sun, Adv. Mater., 19, 3349 (2007). 178. R. Fernández-Pacheco, M. Arruebo, C. Marquina, R. Ibarra, J. Arbiol and J. Santamaria, Nanotechnology, 17, 1188 (2006).
144
Shen and Yang
179. S. Papadimitriou, A. Tegou, E. Pavlidou, S. Armyanov, E. Valova, G. Kokkinidis and S. Sotiropoulos, Electrochim. Acta., 53, 6559 (2008). 180. T. Bala, S. K. Arumugam, R. Pasricha, B. L. V. Prasad and M. Sastry, J. Mater. Chem., 14, 1057 (2004). 181. L. Chen, J. M. Chen, H. D. Zhou, D. J. Zhang and H. Q. Wan, Mat. Sci. Eng. AStruct., 452, 262 (2007). 182. T. G. Schaaff and R. L. Whetten, J. Phys. Chem. B, 104, 2630 (2000). 183. M. A. Garcia, J. M. Merino, E. F. Pinel, A. Quesada, J. de la Venta, M. L. R. Gonzalez, G. R. Castro, P. Crespo, J. Llopis, J. M. Gonzalez-Calbet and A. Hernando, Nano Lett., 7, 1489 (2007). 184. J. S. Garitaonandia, M. Insausti, E. Goikolea, M. Suzuki, J. D. Cashion, N. Kawamura, H. Ohsawa, I. G. de Muro, K. Suzuki, F. Plazaola and T. Rojo, Nano Lett., 8, 661 (2008). 185. X. W. Teng, W. Q. Han, W. Ku and M. Hucker, Angew Chem-Int Edit, 47, 2055 (2008). 186. C. Liu, X. W. Wu, T. Klemmer, N. Shukla and D. Weller, Chem. Mater., 17, 620 (2005). 187. H. Zeng, J. Li, Z. L. Wang, J. P. Liu and S. Sun, Nano Lett., 4, 187 (2004). 188. H. W. Gu, R. K. Zheng, X. X. Zhang and B. Xu, J. Am. Chem. Soc., 126, 5664 (2004). 189. X. W. Teng and H. Yang, J. Nanosci. Nanotechnol., 7, 356 (2007). 190. M. T. Klem, D. Willits, D. J. Solis, A. M. Belcher, M. Young and T. Douglas, Adv. Funct. Mater., 15, 1489 (2005). 191. G. S. Chaubey, C. Barcena, N. Poudyal, C. B. Rong, J. M. Gao, S. H. Sun and J. P. Liu, J. Am. Chem. Soc., 129, 7214 (2007). 192. S. Behrens, H. Bonnemann, N. Matoussevitch, A. Gorschinski, E. Dinjus, W. Habicht, J. Bolle, S. Zinoveva, N. Palina, J. Hormes, H. Modrow, S. Bahr and V. Kempter, J. Phys.-Condes. Matter, 18, S2543 (2006). 193. J. M. Bai and J. P. Wang, Appl. Phys. Lett., 87, (2005). 194. W. S. Seo, J. H. Lee, X. M. Sun, Y. Suzuki, D. Mann, Z. Liu, M. Terashima, P. C. Yang, M. V. McConnell, D. G. Nishimura and H. J. Dai, Nat. Mater., 5, 971 (2006). 195. E. Donath, G. B. Sukhorukov, F. Caruso, S. A. Davis and H. Mohwald, Angew. Chem-Int Edit., 37, 2202 (1998). 196. Y. H. Wang, P. L. Chen and M. H. Liu, Nanotechnology, 19, (2008). 197. Y. Q. Zhang, Z. B. Huang, F. Q. Tang and J. Ren, Thin Solid Films, 515, 2555 (2006). 198. H. Y. Wu, Z. L. Liu, X. D. Wang, B. H. Zhao, J. Zhang and C. X. Li, J. Colloid Interface Sci., 302, 142 (2006). 199. J. H. Gao, G. L. Liang, J. S. Cheung, Y. Pan, Y. Kuang, F. Zhao, B. Zhang, X. X. Zhang, E. X. Wu and B. Xu, J. Am. Chem. Soc., 130, 11828 (2008). 200. O. D. Velev, K. Furusawa and K. Nagayama, Langmuir, 12, 2374 (1996).
Surface Modifications of Nanoparticles
145
201. J. Gao, G. Liang, B. Zhang, Y. Kuang, X. Zhang and B. Xu, J. Am. Chem. Soc., 129, 1428 (2007). 202. L. Zhang, S. Z. Qiao, Y. G. Jin, Z. G. Chen, H. C. Gu and G. Q. Lu, Adv. Mater., 20, 805 (2008). 203. T. R. Zhang, J. P. Ge, Y. X. Hu, Q. Zhang, S. Aloni and Y. D. Yin, Angew. Chem., 120, 5890 (2008). 204. W. L. Shi, Y. Sahoo, H. Zeng, Y. Ding, M. T. Swihart and P. N. Prasad, Adv. Mater., 18, 1889 (2006). 205. Y. S. Lin, S. H. Wu, Y. Hung, Y. H. Chou, C. Chang, M. L. Lin, C. P. Tsai and C. Y. Mou, Chem. Mater., 18, 5170 (2006). 206. X. J. Ji, R. P. Shao, A. M. Elliott, R. J. Stafford, E. Esparza-Coss, J. A. Bankson, G. Liang, Z. P. Luo, K. Park, J. T. Markert and C. Li, J. Phys. Chem. C., 111, 6245 (2007). 207. X. Teng and H. Yang, J. Am. Chem. Soc., 125, 14559 (2003). 208. M. Karg, I. Pastoriza-Santos, J. Perez-Juste, T. Hellweg and L. M. Liz-Marzan, Small, 3, 1222 (2007). 209. X. Y. Xu, N. L. Rosi, Y. H. Wang, F. W. Huo and C. A. Mirkin, J. Am. Chem. Soc., 128, 9286 (2006). 210. D. K. Yi, S. T. Selvan, S. S. Lee, G. C. Papaefthymiou, D. Kundaliya and J. Y. Ying, J. Am. Chem. Soc., 127, 4990 (2005). 211. J. P. Ge, Y. X. Hu and Y. D. Yin, Angew. Chem-Int. Edit., 46, 7428 (2007). 212. J. Q. Zhuang, H. M. Wu, Y. G. Yang and Y. C. Cao, Angew. Chem-Int. Edit., 47, 2208 (2008). 213. X. W. Teng and H. Yang, Nanotechnology, 16, S554 (2005). 214. D. Ma, T. Veres, L. Clime, F. Normandin, J. Guan, D. Kingston and B. Simard, J. Phys. Chem. C., 111, 1999 (2007). 215. B. L. Frankamp, N. O. Fischer, R. Hong, S. Srivastava and V. M. Rotello, Chem. Mat., 18, 956 (2006). 216. S. Y. Lee and M. T. Harris, J. Colloid Interface Sci., 293, 401 (2006). 217. Y. Y. Liang, L. M. Zhang, W. Li and R. F. Chen, Colloid Polym. Sci., 285, 1193 (2007). 218. M. Z. Rong, M. Q. Zhang, H. B. Wang and H. M. Zeng, J. Polym. Sci. Pt. B-Polym. Phys., 41, 1070 (2003). 219. Q. X. Wang, B. Z. Song and H. Z. Li, J. Mater. Sci. Technol., 19, 155 (2003). 220. D. N. Williams, K. A. Gold, T. R. P. Holoman, S. H. Ehrman and O. C. Wilson, J. Nanopart. Res., 8, 749 (2006). 221. T. Osaka, T. Matsunaga, T. Nakanishi, A. Arakaki, D. Niwa and H. Iida, Anal. Bioanal. Chem., 384, 593 (2006). 222. M. J. Kim, B. Lim, Y. K. Jeong, Y. W. Cho and Y. H. Choa, J. Ceram. Process. Res., 8, 293 (2007).
146
Shen and Yang
223. R. De Palma, J. Trekker, S. Peeters, M. J. Van Bael, K. Bonroy, R. Wirix-Speetjens, G. Reekmans, W. Laureyn, G. Borghs and G. Maes, J. Nanosci. Nanotechnol., 7, 4626 (2007). 224. S. T. Selvan, P. K. Patra, C. Y. Ang and J. Y. Ying, Angew. Chem-Int. Edit., 46, 2448 (2007). 225. H. Lee, M. K. Yu, S. Park, S. Moon, J. J. Min, Y. Y. Jeong, H. W. Kang and S. Jon, J. Am. Chem. Soc., 129, 12739 (2007). 226. Y. T. Lee, K. Woo and K. S. Choi, IEEE Trans. Nanotechnol., 7, 111 (2008). 227. K. M. K. Selim, J. H. Lee, S. J. Kim, Z. C. Xing, I. K. Kang, Y. Chang and H. Q. Guo, Macromol. Res., 14, 646 (2006). 228. D. E. Sosnovik, M. Nahrendorf and R. Weissleder, Basic Res. Cardiol., 103, 122 (2008). 229. C. Zhang, B. Wangler, B. Morgenstern, H. Zentgraf, M. Eisenhut, H. Untenecker, R. Kruger, R. Huss, C. Seliger, W. Semmler and F. Kiessling, Langmuir, 23, 1427 (2007). 230. A. K. Gupta and A. S. G. Curtis, J. Mater. Sci.-Mater. Med., 15, 493 (2004). 231. N. Morishita, H. Nakagami, R. Morishita, S. Takeda, F. Mishima, B. Terazono, S. Nishijima, Y. Kaneda and N. Tanaka, Biochem. Biophys. Res. Commun., 334, 1121 (2005). 232. P. Phanapavudhikul, S. Shen, W. K. Ng and R. B. H. Tan, Drug Deliv., 15, 177 (2008). 233. S. I. Takeda, B. Terazono, F. Mishima, H. Nakagami, S. Nishijima and Y. Kaneda, J. Nanosci. Nanotechnol., 6, 3269 (2006). 234. M. K. Yoo, I. Y. Kim, E. M. Kim, H. J. Jeong, C. M. Lee, Y. Y. Jeong, T. Akaike and C. S. Cho, J. Biomed. Biotechnol., (2007). 235. S. Liang, Y. X. Wang, J. F. Yu, C. F. Zhang, J. Y. Xia and D. Z. Yin, J. Mater. Sci.-Mater. Med., 18, 2297 (2007). 236. S. Mohapatra, S. K. Mallick, T. K. Maiti, S. K. Ghosh and P. Pramanik, Nanotechnology, 18, (2007). 237. S. Mornet, S. Vasseur, F. Grasset and E. Duguet, J. Mater. Chem., 14, 2161 (2004). 238. M. Uchida, M. L. Flenniken, M. Allen, D. A. Willits, B. E. Crowley, S. Brumfield, A. F. Willis, L. Jackiw, M. Jutila, M. J. Young and T. Douglas, J. Am. Chem. Soc., 128, 16626 (2006). 239. X. M. Wang, R. Y. Zhang, C. H. Wu, Y. Y. Dai, M. Song, S. Gutmann, F. Gao, G. Lv, J. Y. Li, X. M. Li, Z. Q. Guan, D. G. Fu and B. A. Chen, J. Biomed. Mater. Res. Part A, 80A, 852 (2007). 240. Y. Zhang, N. Kohler and M. Q. Zhang, Biomaterials, 23, 1553 (2002). 241. D. Gao, H. Xu, M. A. Philbert and R. Kopelman, Nano Lett., (2008). 242. T. Hyeon, S. S. Lee, J. Park, Y. Chung and H. Bin Na, J. Am. Chem. Soc., 123, 12798 (2001). 243. J. Park, K. J. An, Y. S. Hwang, J. G. Park, H. J. Noh, J. Y. Kim, J. H. Park, N. M. Hwang and T. Hyeon, Nat. Mater., 3, 891 (2004).
Surface Modifications of Nanoparticles
147
244. P. Majewski and B. Thierry, Crit. Rev. Solid State Mat. Sci., 32, 203 (2007). 245. J. Kim, H. S. Kim, N. Lee, T. Kim, H. Kim, T. Yu, I. C. Song, W. K. Moon and T. Hyeon, Angew. Chem-Int. Edit., 47, 8438 (2008). 246. L. Etgar, E. Lifshitz and R. Tannenbaum, J. Phys. Chem. C., 111, 6238 (2007). 247. J. Choi, J. C. Kim, Y. B. Lee, I. S. Kim, Y. K. Park and N. H. Hur, Chem. Commun., 1644 (2007). 248. G. A. Hughes, DM-Dis. Mon., 51, 342 (2005). 249. H. W. Gu, Z. M. Yang, J. H. Gao, C. K. Chang and B. Xu, J. Am. Chem. Soc., 127, 34 (2005). 250. A. K. Gupta and M. Gupta, Biomaterials, 26, 3995 (2005). 251. A. Ito, M. Shinkai, H. Honda and T. Kobayashi, Cancer Gene Ther., 8, 649 (2001). 252. M. K. Yu, Y. Y. Jeong, J. Park, S. Park, J. W. Kim, J. J. Min, K. Kim and S. Jon, Angew. Chem-Int. Edit., 47, 5362 (2008). 253. M. I. Shukoor, F. Natalio, N. Metz, N. Glube, M. N. Tahir, H. A. Therese, V. Ksenofontov, P. Theato, P. Langguth, J. Boissel, H. C. Schröder, W. E. G. Müller and W. Tremel, Angew. Chem-Int. Edit., 47, 4748 (2008). 254. A. del Campo, T. Sen, J. P. Lellouche and I. J. Bruce, J. Magn. Magn. Mater., 293, 33 (2005).
CHAPTER 3 PROGRESS IN BIONANOCOMPOSITE MATERIALS
Eduardo Ruiz-Hitzky1, Margarita Darder1,2 and Pilar Aranda1 1
Instituto de Ciencia de Materiales de Madrid, CSIC. C/Sor Juana Inés de la Cruz 3, Cantoblanco, 28049 Madrid (Spain); 2Instituto Madrileño de Estudios Avanzados en Materiales (IMDEA-Materiales), C/Profesor Aranguren s/n, 28040 Madrid (Spain)
The present review chapter includes an overview on the current state-of-art of bionanocomposites, which are an emerging class of nanostructured biohybrid materials. In the same way than conventional nanocomposites, these biohybrid materials also exhibit both structural and functional properties together with biocompatibility and biodegradability, which can be of great interest for different applications. Three main areas of interest have been identified, those dealing with the development of green nanocomposites, bionanocomposites addressed to biomedical purposes, as well as bionanohybrids for uses in advanced devices.
1. Introduction Nanocomposites are a class of nanostructured organic-inorganic hybrids that are giving rise to advanced materials provided of a broad range of properties with incidence in many applications [1,2]. Among these materials, bionanocomposites are an emerging class of biohybrids made of an inorganic component mixed at the nanometer scale with polymers of natural origin (biopolymers) [3]. Bionanocomposites represent a rapidly growing field of research (Figure 1) due to their current and potential uses as ecological materials that attract scientists and engineers for diversification in enhanced applications [4].
149
150
Ruiz-Hitzky et al.
Figure 1. Publications per year related to biopolymer-based nanocomposites (the patterned area in the last column results from extrapolating the number of publications appeared in 2008 until mid-October). Data collected from the ISI Web of Knowledge [v4.3]-Web of Science.
Most of synthetic bionanocomposites consist in the assembly of biopolymers and silicates belonging to the clay minerals family. These last solids combine singular properties, such as chemical inertness, low or null toxicity and good biocompatibility, with high adsorption ability, cation exchange capacity and elevated surface area [5]. These characteristics are essential to assure strong interactions with biopolymers leading to stable bionano-hybrids through different mechanisms, such as hydrogen bonding, water bridges, electrostatic interactions and van der Waals forces. In this way, biopolymer-clay nanocomposites can be easily prepared by direct adsorption of soluble biopolymers on smectite and microfibrous clay minerals. Smectites are able to intercalate charged biopolymers, as for instance chitosan and gelatin, by ion-exchange reactions giving rise to intercalated compounds or even delaminated composites [4,6]. Sepiolite and palygorskite microfibrous clays, interacts by hydrogen bonding and water bridges with neutral and charged biopolymers, these last being also able to binding the clay surface through ionic bonding [7,8]. Certain modified
Progress in Bionanocomposite Materials
151
polymers (e.g. derivatives of cellulose and starch) can be thermally compounded with clays by adding plasticizers and other additives [9]. Other inorganic particles of different nature and origin (metals and metal oxides, hydroxides, salts, carbonaceous materials, etc.) are also susceptible to assembly natural polymers and conveniently modified biopolymers [4,10,11]. In addition to the synthetic routes above indicated, other procedures such as sol-gel processes, self-assembly, and layer-bylayer adsorption have been also reported [12]. As above indicated, bionanocomposites are generally integrated by components of natural origin showing low or null toxicity. Natural polymers (cellulose, starch, chitin and chitosan, gelatin, etc.) and other biodegradable polymers, such as those containing hydrolysable backbones such as polycaprolactone, are eco-friendly materials able to be degraded in a natural process to simpler compounds, mineralized and redistributed in the environment through elemental C, N and S cycles [13]. Some of them can be processed given the so-called bioplastics or green plastics of great interest for replacing conventional plastics from petroleum industry, in applications including biomedicine, packaging materials, disposable nonwovens, hygiene products, etc. [14-17]. However, they show low mechanical properties and low water resistance limiting, or even avoiding, their practical use. The assembly with inorganic nanoparticles provides bionanocomposites with advantageous properties that derive from the synergistic effect of their both types of components that can determine the enhancement of mechanical properties, the increase of the thermal stability or the improvement of gas-barrier properties of the biopolymers [18]. In addition to these advantages useful for applications as diverse as structural materials, flame retardant and food packaging, bionanocomposites can receive attention for developing sensor devices, pharmaceutical, cosmetic, and even medical uses due to the biocompatibility of the integrating components [18]. All these aspects will be considered in the present contribution, with the aim to review the state of the art in bionanocomposites and to stress new contributions and tendency in Science and Technology of this class of nanomaterials.
152
Ruiz-Hitzky et al.
2. Bionanocomposites for Bioplastics The global strategy addressed to substitute conventional plastics derived from petroleum by polymers of natural origin constitutes a challenge of crucial importance because there are still certain restrictions that limit their extended use. Among other of their applications, food packaging is probably the most interesting one from the point of view of worldwide consumption as well as environmental impact reduction [19]. In this way, much research is currently undertaken using abundant biodegradable polymers, such as starch, cellulose, chitin and chitosan, alginate, polycaprolactone (PCL), polylactic acid (PLA), polyhydroxyalkanoates, etc., as well as their derivatives. [13,16,20-22]. The bioplastics industry’s aim is to use of renewable resources in their manufacture affording products susceptible of biodegradability and compostability, as this cycle (Figure 2) represents the means by which environmentally-damaging CO2 emissions can be reduced and fossil resources conserved for future generations [23].
Figure 2. The life cycle of biodegradable plastics (from Ref. [23]).
Progress in Bionanocomposite Materials
153
Strategies to improve the low mechanical properties and poor water resistance of bioplastics include their blending with other polymers (e.g., starch with hemicellulose and zein, chitosan and whey protein with PCL) [24,25], and incorporation of reinforced additives coming from natural sources (e.g., starch and chitin whiskers, corncobs husks, hemp nanocrystals) [26-28]. From a point of view of biodegradability the use of whiskers, for instance from cellulose or starch, become a more ecological alternative to inorganic nanofillers to improve mechanical properties of bioplastics. Thus, rod-like cellulose whiskers prepared from cotton linter pulp, with an average length of 1.2 µm and a diameter of 90 nm, were employed as reinforcing agents of soy protein isolate (SPI) plastics, resulting materials that show an increase in both the tensile strength and Young’s modulus as well as an improvement in the water resistance [29]. These enhanced mechanical properties have been ascribed to the formation of crosslinked networks caused by intermolecular hydrogen bonds between the nanofillers and the SPI matrix. In a similar way, waxy maize starch nanocrystals (6-8 nm thick, 20-40 nm long, and 15-30 nm wide) have been tested as nanofillers in waxy maize starch plasticized with glycerol [30] as well as in natural rubber [31], leading to an enhancement of mechanical properties in both cases. The incorporation of hemp cellulose nanocrystals to plasticized starch has probe to enhance mechanical properties and also decreases the water sensitivity of the nanocomposites [28]. These performance improvements have been attributed to the chemical similarities between cellulose and starch, the nanometric effect of the bio-filler, and the hydrogen bonding interactions between the two components. Although it will be necessary further studies focused on improving the dispersion of biodegradable whiskers in the biopolymer matrix for the enhancement of the mechanical properties, application of this type of ecomaterials is already a reality. In this way, Fujitsu and NEC have recently commercialized notebook computers and mobile phones based on ecofriendly PLA, either blended with a petroleum derived polymer or reinforced by Kenaf fibers [18]. Even though whiskers are preferable from a point of view of developing fully biodegradable materials inorganic nanofillers constitutes an interesting alternative to prepare improved bioplastic
154
Ruiz-Hitzky et al.
materials. In this way, a recent work compares the reinforcing effect of microcrystalline cellulose (MCC) and a commercial organically modified bentonite (layered silicate) in a PLA matrix [32]. Bentonite offers an improvement in tensile modulus and yield strength as well as a reduction in the oxygen permeability, whereas MCC only shows a better behavior concerning the elongation at break. Comparison of starch-based nanocomposites reinforced with cellulose nanowhiskers and the synthetic layered silicate Laponite B, shows better mechanical properties for the bionanocomposites based on the inorganic nanofiller [33]. Therefore, investigation of new bionanocomposites, especially by using clay minerals, probably constitutes the most employed route addressed to increase mechanical strength, decrease gas permeability and increase water resistance. Bionanocomposites offer the additional advantage that the inorganic nanoparticles can be used as carrier of antimicrobials and other additives. In this way, there is a large number of recent reviews on the topic that not only gather different examples of bionanocomposites but describe the state of the art in the preparation, address technological issues and explore markets forces directed to food packaging applications [9,34-38]. Starch, derived from corn, wheat, rice or potato, and cellulose and its derivatives, are the main neutral polysaccharides used in the preparation of green nanocomposites [39-44]. Most of the bionanocomposites are prepared by incorporation of natural or synthetic clay minerals, with and without organic modifications, as inorganic nanofillers, which results in intercalation or exfoliation compounds. Montmorillonites from different origin, including commercial cloisite, are the usual silicates used as nanocharges to produce a reinforcing effect in the biopolymer matrix, resulting in improved mechanical properties. The use of other fillers, such as tourmaline nanoparticles, has been also explored. The effect of this last type of nanoparticles, which can be considered as isodimensional, in cellulose-tourmaline composite films slightly decreases the tensile strength and the Tg value, but interestingly favors antimicrobial action against Staphylococcus aureus [45]. Native starch is not a true thermoplastic but it can be transformed into a plastic-like material so-called plasticized or thermoplastic starch. In this way, plasticizers such as glycerol, tryethylcitrate or vegetable oils
Progress in Bionanocomposite Materials
155
are added to the biopolymer at a temperature below its decomposition temperature and under shear the polymer melts and flows, allowing its use as an extrusion, injection or blow molding material in a similar way to conventional thermoplastic polymers. The use of plasticizers also contributes to a better dispersion of the nanofiller in the biopolymer matrix, resulting in an enhancement of mechanical properties. However, the resulting bionancomposites may be still quite brittle [42]. The amount of plasticizer also affects and for instance the addition of 5 wt% of glycerol produces mostly exfoliated nanoclay, whereas adding 10-15 wt% produces only intercalation [46]. Although the final properties did not vary significantly with the starch source [47], the sequence in the addition of components seems to affect the characteristics of the bionanocomposites and better mechanical properties are obtained when clay was added to starch and then plasticized [40].
Figure 3. Schematic representation of the process to produce cationic starch/montmorillonite fillers by exfoliation/adsorption technique. (from Ref. [55]).
In addition to montmorillonites, other clays such as kaolinite and hectorite have been also used as nanofillers of plasticized starch as as well as other layered systems including layered double hydroxides (LDHs) and brucite [48-50]. In general, the use of organoclays is less effective than the non-modified silicate to intercalate and delaminate starch, and the degree of intercalation depends on filler concentration as well as on the hydrophobicity of the modified clay [51-54]. The use of cationic starch as a new organomodifier to better match the polarity of the matrix and therefore facilitate the clay exfoliation process have been recently proposed [55]. The modified-starch/montmorillonite organoclay
156
Ruiz-Hitzky et al.
is prepared via exfoliation-adsorption technique (Figure 3) and once dried is blended with starch and glycerol. In contrast to other organoclays, the modified-starch/montmorillonite filler can be easily exfoliated and also shows lower tendency to incorporate glycerol over starch than Namontmorillonite. These characteristics favor the formation of bionanocomposites with highly exfoliated clay, which strongly improves their mechanical properties [55]. Besides improvement of mechanical properties the presence of the nanofiller particles in the biopolymer matrix gives rise to “tortuous” pathways that difficult the gas diffusion through the bionanohybrid. Although enhancement of barrier properties is one of the requisites to favor the use of bionanocomposites in food packaging applications reduction of water absorption remains still a serious drawback. In this way, recent research searches strategies to produce materials with higher water uptake by incorporation of other hydrophobic bioplastics or by substitution of the clay for more hydrophobic nanocharges. Bionanocomposites based on montmorillonite and organomontmorillonites and starch-polycaprolactone blends show a decrease of the water diffusion coefficient with clay incorporation. However elongation and break increase after exposure to humid environments indicating matrix plasticization, being the mechanical properties affected by the nature of the clay modifier [56]. The use of alternative nanofillers, such as multiwalled carbon nanotubes (MWCNTs), has been recently explored [57]. Previous acid treatment of MWCNTs is required to improve dispersibility and adhesion of the nanofiller into the biopolymer matrix. Plasticized starch-MWCNTs bionanocomposites show improved mechanical properties and the presence of the MWCNTs decreases the water uptake at moisture equilibrium [57]. Chitosan is a water soluble polysaccharide obtained by partial deacetylation of chitin, the second most abundant biopolymer in the Nature, which shows interesting properties as film formation ability. Chitosan is positively charged, which makes it an effective binder to negatively charged substrates of different nature (metals, biochemical, macromolecules and cells). Interestingly, it has anti-microbial properties that can be profited when used in the preparation of bionanocomposites
Progress in Bionanocomposite Materials
157
for different applications including food packaging. Although the first reported clay-based bionanocomposite was based on chitosan-smectites [58], few research has been devoted in the exploration of chitosan bionanocomposites for packaging applications. However, improvement of the mechanical properties and preservation of the film formation ability of the biopolymer can be attained in bionanocomposites as reported for chitosan-sepiolite [7]. Actually, chitosan dissolved in water can be intercalated in layered silicates, such as montmorillonite, giving different type of bionanocomposites depending on the amount of incorporated polymer. Controlled ion-exchange reactions produce bionanocomposites with the biopolymer disposed as monolayer between the silicate layers but if the chitosan concentration is high enough the polymer can be incorporated in larger extend accompanied by counter-ions to balance the total charge in the system [58,59]. It has been also reported the possibility to obtain bionanocomposites with a mixed exfoliated-intercalated structure in which the hardness and elastic modulus are gradually enhanced with the content of clay (from 2.5 to 10 wt%) [60]. The mechanical and water vapor barrier properties of chitosan-montmorillonite nanocomposite films are also strongly affected by the clay concentration [61]. As chitosan is a highly hydrophilic biopolymer it is low compatible with organo-clays, such as cloisite 30B, resulting in composite formation [62]. Other authors reported the formation of cloisite 30B-chitosan bionancomposite films with improved tensile, water vapor barrier and water resistance properties as well as superior antimicrobial activity in comparison to montmorillonite-chitosan films [63]. In this case, the preparation method operates through acid treatments of the organo-clay (or the clay) suspension and further assembling with chitosan and glycerin. It can be admitted that this last compound may favor the compatibilization of the biopolymer and the organo-clay. Among other sources of proteins, soybean, gelatin, wheat gluten and whey probably are the most explored ones in view to prepare bioplastics. Soybean protein isolate (SPI) extracted from soybean contains more than 90% of proteins (mainly glycinin and β-conglycinin) and attracts increasing interest because it can be thermoplastically processed by using plasticizer small molecules, such as glycerol. The incorporation of
158
Ruiz-Hitzky et al.
different type of fillers, including polyphosphates [64], different type of clays [65-68], talc powders and zeolites [65], carbon nanotubes [69] or SiO2 nanoparticles [70] have been tested to improve its film properties. The use of clays introduces improvement of the tensile strength decreasing water vapor permeability and water solubility but also elongation at break in the biocomposites [65]. The use of ultrasonic treatments favors delamination of the montmorillonite into the SPI matrix [66]. The amount of incorporate montmorillonite affects to the exfoliation degree therefore to the mechanical strength and thermostability of the bionanocomposites, and the optimal seems to be for 12 wt% clay content [67]. Recently, it has been explored the use of rectorite, a clay mineral consisting of regular interstratification of montmorillonite-type and mica-type layers, as filler, being also observed that the bionanocomposite with a 12 wt% clay loading shows the highest strength [68]. The use of rigid nanoparticles of SiO2 produces a kind of reinforced and toughened bionanocomposite however the amount of nanofiller must be controlled as separation of SPI microphases and SiO2 domains has been observed [70]. More interesting seems to be the use of MWCNTs that with just a 0.25 wt% content results in reinforced and thoughened SPI bionanocomposites provided also of higher water resistance. These effects have been ascribed to interactions at the interface of SPI chain wrapped MWCNT and the interface between penetrating SPI chains and the internal wall of MWCNTs, as well as the associations with the SPI matrix mediated by protruding segments from the channel inside [69]. Gelatin is a protein mainly composed of alpha chains of aminoacids which is obtained by thermal or acid denaturation of collagen [71] and show among other properties excellent swelling and water absorbency, ion-exchange behavior, easy processing as films as well as biocompatibility, non-toxicity and biodegradability. Gelatin has been assembled with a large variety of inorganic solids such as clay minerals, silica particles, hydroxyapatite, calcium triphosphate, zirconium phosphates, vanadium pentoxide, layered perovskites and carbon nanofibers and nanotubes, with the aim to develop materials for different applications, food additives, food packaging, biomedical materials and tissue engineering purposes, as recently reviewed [72]. Gelatin can be
Progress in Bionanocomposite Materials
159
easily intercalated in layered silicates, such as montomorillonite, arranged in mono- or bi-layer disposition [73-75] or even forming more expanded structures [76-78], depending on the initial concentration of gelatin. It is possible to prepare homogeneous films whose transparent aspect decreases with the increase in clay content [73,74]. The partially exfoliated materials show improvements of tensile strength and Young’s modulus with respect to the neat gelatin films [77], varying with montmorillonite content as well as pH of the gelatin matrix [76,79]. It seems that also the clay particle size influences the mechanical properties and it has been found in bionanocomposites based on Cloisite® and Laponite®, fillers with a tenfold aspect ratio, that the filler aspect ratio, dispersion and content were critical for the final properties [77]. The use of fibrous fillers, such as the clay mineral sepiolite, reveals enhanced mechanical properties with a high reinforcement efficiency especially for low loadings (283% Young’s modulus enhancement per unit percent of clay mass in a 0.5 wt% sepiolite-gelatin bionanocomposite) (Figure 4) [8]. Homogeneous and transparent films can be prepared by assembly of gelatin with montmorillonite and sepiolite clay minerals but silicates of large particle size, such as vermiculite, originate inhomogeneous materials ascribed to micro-composites. Layered double hydroxides showing small particle size gives rise to homogeneous films although intercalation does not occur [8]. Interestingly, the wet mechanical strength was also significantly improved in the bionanocomposites, although water uptake still remains one of the main drawbacks of this type of materials for food packaging applications. Poly(lactic acid), PLA, is a biodegradable thermoplastic derived from L-lactic acid produced in the fermentation of cornstarch, that has been widely used in the development of reinforced bioplastics, mainly using organoclays as nanofillers [18,80]. Melt intercalation is the most frequent method employed for the preparation of these bionanocomposites [81-84] but alternatively, nanocomposites can be prepared by in situ polymerization of previously intercalated lactic acid monomers to produce exfoliated systems [85]. PLA-clay bionanocomposites show improved thermomechanical and gas barrier properties compared with the polymer alone and their biodegradability strongly depends on the nature of the layered silicate and the organic
160
Ruiz-Hitzky et al.
modifier, being possible to tailor their biodegradability by choosing the adequate organoclay nanofiller [86]. It has been also reported that a faster hydrolytic degradation takes place for more hydrophilic fillers [87]. It has been also reported that the incorporation of a layered titanate in the PLA matrix results in an enhanced rate of biodegradation under sunshine due to its photocatalytic reactivity, similar to that of TiO2 [88]. Recently has been reported the use of MWCNTs as nanofiller of PLA resulting in improvement of mechanical properties and thermal stability, specially when using MWCNTs submitted to an acid treatment to promote the formation of –COOH groups [89]. An alternatively route to produce a better dispersion of MWCNTs in the PLA matrix implies the direct grafting of PLA to the carbon nanotubes wall to enhance the interfacial interaction between both components [90]. The electrical resistivity of PLA-grafted MWCNTs bionanocomposites is much lower than pure PLA matrix but higher than PLA-MWCNTs. It should be noted that a recent comparative life cycle study of the impact of biodegradable polymers using poly(3-hydroxybutyrate), PHB, combined to sugar cane bagasse and organo-montmorillonites shows that the relatively low Young modulus and high density of the bionanocomposites compared to conventional plastics are a disadvantage for their environmental performance. However, environmental benefits can be obtained thanks to the saving inputs in the bioplastics productions and for certain applications their lower impact values may also favor their use [91]. Other polymers from natural origin, such as soybean and linseed oils as well as natural rubber obtained from certain trees, such as Hevea brasiliensis, have been also combined to inorganic nanocharges to obtain bionanocomposites with improved mechanical properties. In this way, sepiolite [92,93], MWNTs [94], SiC nanoparticles and single-walled carbon nanotubes (SWNTs) [95] have been incorporated in natural rubber leading to an enhancement of mechanical and other physical, and chemical properties of the organic matrix. Bio-based elastomers have been also prepared by combination of acrylated oleic methyl ester [96] and epoxidized soybean oils [97-99] with organically modified clays, observing that the presence of the clay produces bionanocomposites with controlled mechanical and coating properties as well as superior barrier properties towards water vapor.
Progress in Bionanocomposite Materials
161
Figure 4. Young’s modulus (A) and reinforcement efficiency (B) as a function of the sepiolite content in gelatin-sepiolite bionanocomposites (Results from Ref. [8]).
Even though the many examples here introduced, research on green nanocomposites is still in an incipient phase and more research is required to overcome the problems inherent to bioplastics features. Some issues must be directed to explore the use of alternative biopolymers and nanofillers, as well as to develop methodologies of synthesis to attaint an enhanced compatibility with the inorganic moieties. Some of the alternatives should explore the controlled modification of polysaccharides and other polymers of natural origin, as well as the integration of a wide range of “non-pollutant” nanofillers other than silica and silicate, as for instance LDHs, to afford new formulations with not only improved mechanical properties but also water stability and gas permeation barrier in the resulting green nanocomposites. In the field of edible films, although biopolymers have been extensively studied and applied, up to now few works indicate the possibility of incorporation nanoparticles in order to improve the physical properties of these materials. However, the use of nanoparticles may suppose additional advantages for certain applications of bioplastics. In this way, they can be used as carriers of antimicrobials and additives, which is very important for long-term storage of foods of specific desirables
162
Ruiz-Hitzky et al.
characteristics (flavor,...). Finally, as certain authors have noted, better knowledge of the effects of nanoparticles on the human health and environment is necessary for an adequate regulation of their use for food contact applications [35]. 3. Bionanocomposites for Biomedical Applications Biocompatibility of biopolymers and scarce or null toxicity of clay minerals, and other nanoparticles, used to prepare bionanocomposites make these materials as suitable for biomedical applications from tissue engineering to drug delivery systems (DDS). The development of biomaterials for regenerative medicine can be still considered an emerging field [100], being tissue engineering, and essentially bone implants, a fast-growing branch of this research area. Biodegradable synthetic polymers such as polylactic acid (PLA), polyglycolic acid (PGA), polycaprolactones (PCL), polyanhydrides, polyorthoesters, polyfumarates, and polycarbonates are mostly employed for this purpose [101]. Alternatively, natural polymers such as collagen, chitosan and starch are being also applied to develop hybrid materials that offer the required properties to be used as bioresorbable scaffolds: biocompatibility, suitable mechanical properties, cell attachment and proliferation, enough macroporosity with interconnected pores to allow for the transport of nutrients and metabolic wastes, and controlled biodegradability since the rate of biodegradation needs to be on balance with the rate of regeneration of new tissue [100-102]. Collagen, a fibrous protein and a major extracellular matrix component, is one the natural polymers most commonly involved in the development of scaffolds for damaged tissue repair. Many bionanocomposites based on collagen and hydroxyapatite (HAP) [103-105] are tested as implants that try to mimic the composition, nanostructure, porosity, surface roughness and mechanical properties of natural bone [106,107]. In addition to collagen, many other natural polymers are being combined to HAP in order to develop nanocomposites for orthopaedic applications [108]. Some recent examples of these HAP-based scaffolds have been prepared with biodegradable synthetic polymers such as the
Progress in Bionanocomposite Materials
163
thermoplastic polyester PLA [109,110] or polycaprolactone [111], while many other examples involve naturally occurring polymers such as alginate [112], chitosan [113], seroalbumin [114], silk fibroin [115], or cellulose [116]. Copolymers such as poly(hydroxybutyrate-co-valerate) (PHBV) are also good components for the preparation of HAP-based scaffolds [117]. In other cases, blends of chitosan with other polymers including fibroin [118], polygalaturonic acid [119], or poly(hydroxyethylmethacrylate-methylmethacrylate) (p(HEMA-MMA)) [120] are combined to HAP to prepare the scaffolds for tissue repair.
Figure 5. SEM micrographs showing the attachment of osteoblast cells (O) into the macroporous chitosan-CPC composite scaffold, developing long cytoplasmic extensions (E) for anchoring to the pore wall. From Ref. [122].
Other scaffolds are based on calcium phosphates different from hydroxyapatite. Thus, chitosan has been mixed with calcium phosphate cement (CPC) derived from tetracalcium phosphate (TTCP, Ca4(PO4)2O) and dicalcium phosphate anhydrous (DCPA, CaHPO4) to develop
164
Ruiz-Hitzky et al.
bioresorbable implants [121]. A similar nanocomposite based on the same components and using mannitol as porogen was tested as an injectable scaffold [122], with potential applications in dental, craniofacial, and orthopaedic reconstructions due to the excellent cell infiltration into the macropores (Figure 5). In other cases, more soluble calcium phosphates such as β-tricalcium phosphate (β-TCP) and amorphous calcium phosphates (ACP) are preferred for bone tissue repair. Chitosan-ACP materials have been prepared by a urease-assisted procedure [123]. The enzyme produces the hydrolysis of urea, increasing the pH and provoking the chitosan gelation. ACP has been also combined with κ-carrageenan and the resulting macroporous nanocomposites, processed by freeze-drying, showed good compressive mechanical properties that make them suitable for tissue engineering [124]. Besides HAP and calcium phosphates, other inorganic components such as sepiolite, a hydrated magnesium silicate with microfibrous morphology, have been employed. This clay mineral showed high affinity for collagen giving rise to hybrid materials with a high degree of organization, with the sepiolite fibers aligned with the collagen chains [125,126]. The biodegradation rate of this biomaterial can be reduced by treatment with the crosslinker glutaraldehyde that enhances the mechanical properties of the nanocomposite, leading to a longer persistence once implanted in the damaged tissue [127]. Bionanocomposites based on montmorillonite, a layered aluminum silicate, are being also evaluated for application as implants. This silicate has been combined with HAP and chitosan, giving a material with improved nanomechanical properties and higher cell proliferation rate than that observed in chitosan-HAP composites [128]. The chitosanmontmorillonite nanocomposite has been also modified by grafting PDMS chains, which act as a plasticizer to give flexible films, and proposed for application in the biomedical field due to its swelling behavior and water retention properties [129]. Combination of montmorillonite with natural biopolyesters such as PHB and PHBV are also promising materials for biomedical applications [130]. Recently, Al2O3-Zr2O nanoparticles have been also used to reinforce biological matrices such as collagen, enhancing its mechanical and
Progress in Bionanocomposite Materials
165
thermal properties and leading to hybrid materials with potential use in biomedical and bionic applications [131]. Other inorganic components incorporated in bionanocomposites for tissue engineering applications are carbon nanotubes (CNTs). Biocompatible and biodegradable supports for cell growth have been produced from chitosan-MWNTs mixtures by applying the ice segregation induced self-assembly (ISISA) process, that gives rise to a well-defined microchannel porous structure [132]. These porous chitosan-MWNTs materials have been loaded with a bone morphogenetic protein and applied in in vivo experiments as implant for bone formation. A recent work reports the combination of SWNTs with a collagen-fibrin matrix and the processing of the material by the application of mechanical strain, which increases its compactness and causes an alignment of the components in the direction of the applied strain [133]. The SWNTs are useful to provide this nanocomposite with electrical conductivity, which is enhanced by the alignment due to the mechanical strain, giving a material with potential application in cardiac and neural tissue engineering. Hybrid materials applied as bioresorbable scaffolds need to be provided with enough macroporosity to allow for the transport of nutrients and metabolic wastes, and thus to help the growth of new tissue. Several strategies are being employed to generate foam-like bionanocomposites provided with interconnected macropores: solvent casting/particle leaching, freeze-drying, supercritical CO2 drying, fiber bonding and phase separation [100,134]. Unidirectional freezing previous to lyophilization drying [104,132] allows to produce materials with an ordered microchannel structure, while supercritical CO2 foaming generates nanocomposites with a controlled cellular structure, with pore diameters from nano- to micrometric dimensions [135]. The biopolymerHAP mixture can be also processed by electrospinning to obtain fibrous matrixes (Figure 6) that enhance cell attachment and proliferation [136]. Thermally induced phase separation is a novel method that has allowed to process a chitosan/fibroin-HAP nanocomposite with a porosity as high as 94% with interconnected open pore structure and pore size ranging from microns to few hundred microns [118].
166
Ruiz-Hitzky et al.
Figure 6. (A) Schematic of the electrospinning process assisted by a surfactant, 12hydroxysteric acid (HSA), to prepare HA-PLA nanocomposite fibers. (B) FESEM image of the fibers electrospun using 25% HAS. (C) TEM image showing the morphology of the resulting fibers, as well as the homogeneous dispersion of HA nanoparticles within PLA matrix. From Ref. [136].
Other important application of bionanocomposites in the biomedicine field is related to controlled drug release. In some cases, the materials tested as implants can work simultaneously as a drug reservoir for the controlled release of bioactive compounds, for instance morphogenetic proteins (BMPs) that are interesting to promote bone regeneration. This double activity was successfully tested with a BMPloaded HAP-alginate-collagen nanocomposite [137], but other compounds such as vitamins may be also incorporated in the bionanocomposite before implantation [138]. Biodegradable and biocompatible polymers are appropriate drug vehicles that can be released at a constant, predetermined rate and even, targeted to a determinate location in the body [13]. For instance, polylactic and certain polyesters, which are frequently used in DDS, can undergo partial or
Progress in Bionanocomposite Materials
167
total dissolution releasing bioactive compounds in a more or less controlled maner. Inorganic nanoparticles, such as smectites clay minerals and layered double hydroxides (LDHs) have been employed as carriers of drugs with different functionality. In this context antiinflammatory, analgesics and chemotherapeutic agents have been assembled to both smectites and LDHs [5,139,140]. These layered solids can act as hosts with an efficient protection at molecular level of the intercalated drugs controlling their delivery. By including drugs into bionanocomposites it is expected to profit from both components, biopolymers and inorganic hosts, to generate DDS provided of improved stability and controlled release. It can be expected that bionanocomposites formed by intercalation of chitosan in montmorillonite and other clay minerals were able to incorporate different types of molecular drugs. Because of the possibility to control the nature and extent of the electrical charge (positive or negative) of these intercalation compounds, as a function of the chitosan intercalation degree [58,59], either anionic or cationic bioactive molecules can be included to develop new DDS. This approach has been recently confirmed using bionanocomposites based on a derivative of chitosan intercalated in montmorillonite, which is able to act as a DDS using BSA as a protein drug model [141]. Apparently, this bionanocomposite combines the drug adsorption action and the mucosa protection effect of the clay with the mucoadhesive and permeability properties of the biopolymer, being these non-cytotoxic formulations useful as promising biomedical applications [141]. Non-viral gene therapy can be also envisaged using bionanocomposites. In this way, Wang et al. have used a “quaternary chitosan” that is apparently intercalated in rectorite to create a biocompatible substrate for DNA adsorption [142]. Such DNAbionanocomposites have been tested as non viral vectors for gene transfection [142], but some crucial points remain unclear: (i) the DNA adsorption mechanisms to the bionanocomposite, (ii) the transfection pathway, i.e. how the bionanocomposite facilitate the pass through cell membrane and the transference of gene material to the nucleus, (iii) the nature of the bionanocomposite itself, as rectorite is a mica/smectite superstructure (interstratified structure) more complex
168
Ruiz-Hitzky et al.
than montmorillonite and other smectites to give well ordered intercalated compounds. The presence of negative sites in the bionanocomposite, afforded by the positively charged ammonium quaternary groups of the modified chitosan, is responsible for electrostatic interactions with the negatively charged DNA species. This is the same approach based in the intercalation of DNA in a magnesiumaluminium LDH that gives bionanocomposites able to act as nonviral vectors in gene therapy (Figure 7) [143]. It is claimed that the inorganic layered host safely protects the intercalated DNA against harsh condition including strong alkaline, weak acidic environments, and DNase attack, facilitating its transport through the cell membrane. It has been proposed that the acidity media of lysosomes can dissolve the LDH host, releasing the DNA that would be then transferred to the nucleus. This system has been tested in gene therapy for targeted treatment of leukemia and diabetes [144].
Figure 7. Model proposed for the intercalation of DNA in Mg-Al LDH after Ref. [143].
Other bionanocomposites having ability to act as DDS can be prepared using sol-gel routes. In this way, Coradin’s group reports on the preparation of silica-biopolymer nanocomposites capsules of silicacoated agarose and silica-carboxymethylcellulose gels, tested with Rodamine B as a molecule probe used as model for controlled drug delivery making these materials potentially applicable as DDS. [145]. The potentiality of these new nanohybrid capsules is realistic because it can be prepared using other alternative biopolymers, such as
Progress in Bionanocomposite Materials
169
poly(lactide-co-glycolide), which is one of the most popular material used in pharmaceutics drug delivery [145]. For the same purpose, coreshell gelatin-silica nanoparticles have been produced by means of an alternative nanoemulsion route [146]. The degradability of these nanospheres inside the cells has been proved, allowing the release of drug in the intracellular space during this process. Also, silica-alginate bionanocomposites processed as nanospheres by means of spray-drying techniques have been studied as DDS. They are non-cytotoxic bionanocomposites that can enter the intracellular space of fibroblast cells (endocytosis) where they are degraded delivering the concerned drug [147]. DDS based bionanocomposites for in situ delivery have been investigated using porous calcium phosphates-polyhydroxybutyrate (PHB) that are able to entrap the tetracycline antibiotic, releasing it in bone implant and repair [148]. The slow release of this antibiotic can be profited to enhance bone-forming ability via osteoblast cell chemostaxis with other benefices such as the reduction of bone resorption and the promotion of alveolar bone growth [148]. It is well known the use of magnetic nanoparticles for targeted drug release. In this way, magnetite nanoparticles functionalized with biocompatible and biodegradable polymers such as poly(hydroxyethylmethacrylate) (PHEMA) or PLA have been prepared [149]. These magnetic bionanocomposites can be directed and accumulated in a target area with the help of an external magnetic field where the drug is released. It should be taken into account that the direct preparation of magnetic bionanocomposites in presence of the biopolymers (e.g. alginate) following the one-pot synthesis procedure, influences the nature (size and aggregation degree) of the iron oxide magnetic nanoparticles [150]. Also, the effect of magnetite nanoparticles on biopolymers, such as carrageenan, can involve modifications in the rheological properties of the resulting bionanocomposites [151]. Using clay based bionanocomposites, such as sepiolite-xanthan gum polysaccharide, Ruiz-Hitzky et al. [152] have recently developed a new class of bio-hybrids by assembly of viral nanoparticles (Influenza virus) to the composite. The external surface of the microfibrous silicate contains external silanol groups that can easily associate the anionic
170
Ruiz-Hitzky et al.
polysaccharide xanthan gum, giving a bionanocomposite material able to facilitate its binding to the virus particles. As the polysaccharide is negatively charged, it is assumed that electrostatic interactions with the positively viral particles take place forming stable virusbionanocomposite systems (Figura 8). Preliminary experiments carried out in mice demonstrate that virus-bionanocomposites induce the formation of specific antibodies with an efficient protection against Influenza virus [152].
Figure 8. TEM image of viral particles assembled to the xanthan-sepiolite nanocomposites.
On the other hand, as reviewed by Niemeyer [153], biopolymers such as proteins and nucleic acids coated nanoparticles (Au, Ag, ZnS,..) can be used as biolabels incorporating coupling agents such as functional organosilanes. This important topic is receiving great attention in view to applications related to bioanalytical purposes in biomedical diagnostics as well as biosensors and other devices [153]. Porous silicon also exhibits a number of properties that make it an attractive material as DDS [154]. These materials can adsorb proteins and other biological species provided that they were positively charged. It can be expected that the resulting bionanocomposites combined with specific bioactive
Progress in Bionanocomposite Materials
171
species could be used as drug delivery devices able to release the drug under controlled physical stimulus. 4. Bionanocomposites for Sensor Devices and Other Applications Bionanohybrid materials exhibiting suitable functionality are applied as the active part of electrochemical, optical or photoelectrical devices, as well as for many other purposes including application as superabsorbents, pollutants removal or membranes in separation processes. Functionality can be provided by the biopolymer or the inorganic solid constituting the nanocomposite, while in other cases it can be the result of a synergistic effect from both types of components. Typical examples of bio-hybrid materials where the functional properties come from the biological counterpart are the enzyme-based nanocomposites. As the main advantage, these bionanocomposites offer long-time stability due to the protecting effect of the inorganic matrix on the enzyme. The confinement in the inorganic solid prevents the irreversible deformation of the protein structure and preserves its biological activity. As a consequence, this fact will allow the development of devices with an improved long-term stability, from optical and electrochemical biosensors to enzymatic bioreactors. Layered or 2D inorganic solids such as phyllosilicates, phosphates, triphosphochalcogenides, perovskites, and layered double hydroxides (LDHs) have been reported as suitable host matrixes for the entrapment of enzymes. The transport of substrates and products to the immobilized enzyme is guaranteed by the open frame of these inorganic solids. One of the pioneer works, reported by McLaren and Peterson in 1961, described the incorporation of lysozyme, lactoglobulin, pepsin and chymotrypsin into the interlayer space of the montmorillonite layered silicate through a direct intercalation mechanism [155]. The diameter of the intercalated enzymes was determined by XRD measurements from the increase in the basal spacing. More recently, different enzymes and globular proteins have been immobilized in other layered solids such as zirconium phosphate (α-ZrP) [156], layered calcium niobate (HCa2Nb3O10, perovskite) [157], manganese triphosphochalcogenide (MnPS3) [158], and a synthetic magnesium phyllosilicate [159]. In all these examples,
172
Ruiz-Hitzky et al.
the biological macromolecules were incorporated by means of a previous delamination of the layered solid, usually employing quaternary ammonium salts, followed by the restacking of the layers. The activity of the entrapped enzymes was checked in all the resulting bionanocomposite materials, which were proposed as active phases in biosensor devices. LDHs also offer an appropriate open structure for an effective immobilization of enzymes. For instance, urease has been entrapped within Zn-Al hydroxide layers following co-precipitation or delamination-restacking mechanisms [160]. The bio-hybrid materials resulting from both routes were successfully applied as active phases in capacitance biosensors, being the maximum sensitivity achieved for the system prepared by the delamination-restacking procedure. The coimmobilization of chitosan with the LDH gel has been recently reported as a way to increase the biocompatibility, the adhesive ability and the mechanical strength of the immobilization host matrix [161]. The entrapment of polyphenol oxidase (PPO) in this material has allowed the construction of an amperometric phenol biosensor. In both examples, the strong association between the urease and PPO enzymes and the LDH layers achieved with the co-precipitation method makes unnecessary the use of glutaraldehyde, a crosslinker commonly employed to increase the stability in other enzymatic systems. Silica matrices generated from different precursor organoalkoxy- and alkoxy-silanes by mild chemistry procedures are widely applied for the entrapment of proteins, enzymes and antibodies and even yeasts and bacteria [162,163]. The resulting bio-hybrid materials, with the biomolecules encapsulated in the silica network, have been incorporated in a wide number of optical and electrochemical devices, mainly for application as biosensors exhibiting a long-time stability [164-166]. In order to increase the biocompatibility of the silica matrix, certain additives may be incorporated during the sol-gel process. For instance, the more biocompatible environment provided by addition of glycerol allowed to enhance the viability of the entrapped bacteria Serratia marcescens [167], as observed from the metabolic production of the red pigment prodigiosin (Figure 9). Different polysaccharides, including xanthan, locust bean gum and several cellulose derivatives, have been
Progress in Bionanocomposite Materials
173
also mixed with the precursor tetrakis(2-hydroxyethyl) orthosilicate (THEOS) in order to increase the biocompatibility of the silica gel. This feature together with a suitable porosity of the resulting bionanocomposite networks offer a favorable environment for the immobilization of 1→3-β-D-glucanase and α-D-galactosidase, increasing their stability more than a hundred times [168]. In addition to polysaccharides, the proteins bovine serum albumin (BSA), casein and gelatin have been also employed as template in the formation of the silica network from THEOS. The morphology and properties of the resulting porous nanocomposites can be controlled due to the different shape of these proteins and the strong influence of pH and temperature on their structure [169]. It has been also reported the ability of lysozyme to precipitate amorphous silica or titania, in which the enzyme remains entrapped, producing functional biomaterials with antimicrobial activity [170]. Nanocrystalline TiO2 prepared by anodic electrodeposition has been also employed for the entrapment of bacterial photosynthetic reaction center proteins, in order to develop functional bionanocomposite materials that work as active phase in biophotoelectric sensors [171]. Titania nanoparticles have been also grown on a chitosan matrix acting as a directing agent, producing a hybrid porous material by means of supercritical CO2 drying with a homogeneous dispersion of titania nanoparticles. This material has been applied as a bifunctional catalyst for monogliceride synthesis, due to the cooperative effect of basic and acid sites, the NH2 groups from biopolymer and the titanium center, respectively [172]. In many other bionanocomposites the functional properties are mainly provided by the inorganic counterpart. Among them, materials incorporating carbon nanotubes (CNTs), magnetic and metallic nanoparticles are very recent examples. Many biosensor devices have profited from the interesting electrical conductivity properties of CNTs. Their proved biocompatibility allows their integration with a wide variety of biological entities, from nucleic acids to proteins, leading to functional biohybrid materials for bionanoelectronic applications [173,174]. For instance, several glucose biosensors have been developed following different strategies that include layer-by-layer techniques to entrap glucose oxidase (GOD) within the CNTs layer [175,176], one-step
174
Ruiz-Hitzky et al.
electrodeposition method to grow the active phase from a chitosan-CNTGOD solution, acting the biopolymer as a continuous phase in which CNTs and the enzyme are entrapped [177], and the integration of GOD and SWNTs into a bionanocomposite by means of an ionic liquid-like unit [178]. An alternative strategy involves the covalent grafting of the enzyme to chitosan previously combined with CNTs, which is then processed as a film onto the electrode surface. In this example, glutaric dialdehyde was used as cross-linker to immobilize the enzyme glucose dehydrogenase [179]. Precipitation of the enzyme and the CNTs on the electrode surface followed by deposition of a polymer to wrap the active phase is also a useful method to immobilize enzymes, for instance hemoglobin [180] and myoglobin [181] by means of poly(sodium-pstyrene-sulfonate) or Nafion®, respectively. In both cases, the CNTs integrated in the film promote electron transfer between the entrapped protein and the electrode.
Figure 9. The viability of S. marcescens bacteria encapsulated in silica gel is confirmed by production of the red pigment prodigiosin (A), which is maintained over four weeks (B). (Composed from Ref. [167]).
In certain cases, MWNTs are functionalized with the aim to create appropriate groups on their surface. Activation of CNTs with acid produces carboxyl and hydroxyl groups that can be coupled to poly(acrylonitrile-co-acrylic acid), giving nanofibrous membranes with a
Progress in Bionanocomposite Materials
175
high specific surface area and good electrical conductivity where catalase has been successfully immobilized [182]. In a similar way, a sonochemical treatment of MWNTs in acidic solution generated quinone-like functional groups on the tube ends [183], which acted as redox mediators in the electrocatalytic oxidation of dihydronicotinamide adenine dinucleotide (NADH) (Figure 10). These functionalized MWNTs were combined with the enzyme glutamate dehydrogenase and chitosan to construct an amperometric glutamate biosensor [183].
Figure 10. (A) Functional carbon nanotubes (FCNT) with quinone-like groups generated on the tube ends by sonochemical treatment of MWNTs in acidic solution. (B) Electrocatalytic activity of the chitosan-FCNT nanocomposite towards the oxidation of dihydronicotinamide adenine dinucleotide (NADH): (a,b) FCNT-chitosan nanocomposite and (c) as-purchased MWCNTs in the (a) absence and (b,c) presence of 3 mM NADH in 0.1 M PBS, at 5 mV/s. From Ref. [183].
Besides enzymatic biosensors, CNTs are also involved in the design of DNA biosensors. A recent work describes the use of a MWNTschitosan bionanocomposite deposited on a screen printed electrode (SPE) and combined with a double-stranded herring sperm DNA using layerby-layer technique. The resulting device is applied for the detection of deep DNA degradation [184]. SWNts have been also employed in combination with lysozyme and DNA using a layer-by-layer approach, in order to develop multifunctional materials with antimicrobial properties that can be processed as protective coatings [185].
176
Ruiz-Hitzky et al.
Another recent application of CNTs is the development of biofuel cells, concretely microbial fuel cells (MFCs). In such devices, the CNTs constitute the carbon electrodes where a colony of microorganisms is incorporated, and the whole system converts the microbial reducing power to electrical energy. In a recent work, a carbonaceous material derived from MWNTs was used as a biocompatible support for the growth of Staphylococcus aureus, a Gram-positive bacterium, and the resulting bio-hybrid material was proposed for further application in MFCs [186]. In order to increase the specific surface area of the carbon electrode, the MWNTs have been processed as a macroporus scaffold by means of ISISA (ice segregation induced self-assembly) process, with the help of chitosan to homogeneously disperse the CNTs [187]. The biocompatibility and non-cytotoxicity of these porous scaffolds allow the growth and proliferation of E. coli, a hydrogen producing bacteria, resulting in a promising hybrid material for MFCs applications. The electrochemical performance of these systems can be increased by modifying the MWNTs-chitosan scaffolds with Pt nanoparticles, leading to a material with improved catalytic activity towards the methanol oxidation [188]. In addition to the use of CNTs in electrochemical devices, a recent application reports the processing of aqueous suspensions of CNTs dispersed in the biopolymers gellan and xanthan gum for water sensitive inkjet printing. The resulting materials were successfully printed as an ink on transparent poly(ethylene terephthalate) (PET) plastic substrates, giving rise to thin composite films that are not sensitive to solvent vapors but only to water [189]. Bionanocomposites may combine other functional inorganic counterparts that provide new properties to the hybrid system. For instance, magnetic FePt nanoparticles were assembled by a DNAmediated self-assembly route giving rise to magnetic bionanocomposite materials [190]. In a similar way, the assembly of magnetic FePt and nonmagnetic Au nanoparticles was carried out using ferritin, an iron storage protein [191]. Similar materials are commonly prepared by other methods such as dispersion of magnetic nanoparticles in biopolymers, whose biocompatibility will allow the further assembling of bioactive species (enzymes or cells). Recently, Fe3O4 magnetic nanoparticles
Progress in Bionanocomposite Materials
177
(30 nm diameter) were uniformly dispersed in the biopolymer chitosan leading to films with a network-like structure. These bionanocomposites were reported as a favorable environment for the entrapment of tyrosinase, as well as for the preservation of its enzymatic activity, allowing for the construction of amperometric biosensors that show good analytical performance in the detection of phenolic compounds [192]. In other cases, the biopolymer loaded with magnetic nanoparticles is employed for water pollution remediation. For instance, alginate entrapping maghemite and activated carbon was processed as beads and applied to remove methylene blue (MB) and methyl orange (MO) dyes from aqueous solution [193], showing a selective adsorption of MB (Figure 11).
Figure 11. (A) Photograph of magnetic alginate beads attracted by a magnet and (B) adsorption of methylene blue (MB) by the magnetic nanocomposite as a function of time, using starting MB concentrations of (a) 1 and (b) 5 mmol L-1. From Ref. [193].
In a similar way to magnetic nanoparticles, metal nanoparticles have been also dispersed or grown in biopolymers to prepare bioinorganic hybrid materials for several applications. For instance, Pt nanoparticles produced from H2PtCl6·6H2O in ethanolic solution were homogeneously dispersed in κ-carrageenan and the mixture processed as films. Their proton conductivity under high humidity would allow their use as membrane-electrode assemblies in polymer electrolyte fuel cells [194]. Several examples report the growth of gold nanoparticles in the presence of chitosan for different applications. For instance, an amperometric glucose biosensor was constructed by incorporating GOD in the
178
Ruiz-Hitzky et al.
biocompatible hybrid matrix, which was deposited on an electrode previously modified with Au-Prussian Blue (PB) nanoparticles [195]. The Au-chitosan matrix was also employed for immobilization of leukemia cells on electrodes, in order to monitor cell adhesion, proliferation and apoptosis by impedance measurements [196]. A different application involves the dispersion of Au-SiO2 in the chitosan matrix and its use as an effective catalyst of the regioselective hydroamination of alkynes [197]. Another chitosan-based nanocomposite involves antimony oxide bromide (AOB) nanorods synthesized by hydrothermal route [198]. The suitability of this hybrid system to incorporate peroxidase (HRP) allows the construction of a reagentless peroxide biosensor with good sensitivity and long-time stability. In other type of functional bionanocomposites, the new properties arise from the synergistic effect between the biological and the inorganic counterparts. This is the case of bionanocomposites derived from the intercalation of the positively charged polysaccharide chitosan in the layered silicate montmorillonite following an ion-exchange mechanism [58,59], and the assembly of negatively charged polysaccharides (ι-carrageenan, pectin and alginate) in Zn-Al layered double hydroxide through the co-precipitation method [199]. The resulting functional materials have been applied for the first time as active phases in potentiometric sensors for the determination of different anions and cations, taking profit from both their ion-exchange properties and their good stability and robustness. The assembly of chitosan chains by direct adsorption to sepiolite, a microfibrous clay mineral, has been also possible to reverse the starting cation-exchange behavior of the clay and apply the resulting hybrid material in potentiometric sensors for the determination of anions [7]. For a similar application, multifunctional nanocomposites have been prepared from cheap and abundant products such as clay minerals (montmorillonite and sepiolite) and sucrose [200203]. An intermediate caramel-clay material prepared by MW-assisted process was transformed into a carbon-clay nanocomposite, which was then functionalized by reaction with organoalkoxysilanes bearing the desired functional groups. Thus, the functionalized nanocomposites may
Progress in Bionanocomposite Materials
179
act simultaneously as the electronic collector and the sensing phase in potentiometric sensors [202]. Besides electrochemical applications, chitosan-montmorillonite nanocomposites have been employed as biosorbents for the removal of water pollutants and also as superabsorbent materials. The anionexchange property of this nanocomposite, due to the presence of amino groups from chitosan, allows the adsorption of anionic species as tannic acid [204], Congo red [205], and tungsten species [206], and their removal from wastewater. In a similar way, acrylic acid can be adsorbed by the nanocomposite and then polymerized, resulting in an improved and photostable water-superabsorbent material [207]. This material can be prepared by an alternative one-step route that implies an in situ intercalative polymerization process, resulting in a nanocomposite with higher swelling abilities than that obtained by the two-step method [208]. In addition to superabsorbency, the material prepared by the one-step procedure has been also applied for the efficient removal of MB dye from aqueous solutions [209]. Analogous superabsorbent materials are based on polyacrylic acid-grafted collagen and kaolin composites, also prepared in a one-step route, and present high water absorbency values close to 700 g/g [210]. Other explored application of functional bionanocomposites that involve polysaccharides and layered silicates is their use as membranes for pervaporation processes, being the alginate-montmorillonite system a recent example [211]. The hydrophilic character of the clay mineral particles determines a preferential passage of water against organic solvents, such as isopropanol, 1,4-dioxane and THF, being the dehydration activity higher when increase the clay content in the composite membrane. In addition to polysaccharides, as above indicated the fibrous protein gelatin has been also extensively combined with inorganic host solids including zirconium phosphate (α-Zr(HPO4)2·nH2O, α-ZrP) [212], clay minerals and LDHs [8,72], silica [213], and the perovskite CsCa2Nb3O10 [214]. This last solid gives rise to biohybrid materials following a delamination-restacking procedure, and can be processed as selfsupported films that show functionality related to the dielectric behavior of the inorganic moiety, suggesting their application in the microwave
180
Ruiz-Hitzky et al.
industry or in high frequency devices. Gelatin has been also employed as a transparent matrix to disperse montmorillonite previously intercalated with methyl red, a pH indicator [8]. The hybrid material was processed as a self-supported film whose color depends on the acid or basic form of the entrapped dye (Figure 12) and could be applied for pH sensing purposes.
Figure 12. Optical pH sensing using self-supported hybrid films of gelatin incorporating 50% of methyl red intercalated montmorillonite: absorbance spectra acquired from reflectance measurements of a film in basic form (yellow color) kept under an acidic atmosphere at different periods of time (left), and photographs of this film (right) showing the change in color from yellow to pink, related to the transformation of the basic form of methyl red to the acid one.
5. Concluding Remarks Bionanocomposites are nanostructured biohybrid materials resulting from the assembly of species from biological origin combined with particulated inorganic solids. So, this represents a field of research involving a large variety of compounds with different compositions and structures. In this way, it results in a great variety of properties, and therefore in many potential applications derived from these materials.
Progress in Bionanocomposite Materials
181
In many cases the inorganic component is also of natural origin (e.g. clay minerals), giving rise to ecologically acceptable nanocomposites of interest in the development of green nanocomposites. This is particularly important for developing bionanocomposites as food packaging materials provided with enhanced gas and moisture barrier properties, increasing stiffness with lighter weight, strength and thermal stability. Alternatively, the inorganic component is biocompatible (e.g. HAP) and can be useful for applications in biomedicine such as tissue engineering. Applications in bone repair and new DDS are illustrative examples of the interest of bionanocomposites in the biomedical field. In other cases, the inorganic component may afford functional properties allowing the development of new nanostructured biomaterials that can be applied in different type of devices. This is for instance, the case of CNTs-based bionanocomposites where it can be taken advantage from the electronic conductivity inherent to the CNTs to build biosensors. Finally, it can be considered that bionanocomposites are still in an incipient phase of research. Hence a promising future involving the development of advanced bionanohybrid materials from new combinations of inorganic and biological components, novel preparative and processing procedures, as well as the introduction of additives that could confer new properties can be expected. Acknowledgments This work has been supported by the CICYT, Spain (Projects MAT2006-03356), the Comunidad de Madrid, Spain (Project S0505/MAT/000227) and the CSIC (Project PIF08-018-2). M. D. thanks the Spanish MICINN for the award of a Ramón y Cajal contract. References 1. 2. 3.
Functional Hybrid Materials edited P. Gómez-Romero and C. Sánchez, WileyVCH, Weinheim, Germany (2004). P. Gómez-Romero and M. Lira-Cantú in Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, Inc. Hoboken, N.J. (2002) Vol. 13, p. 523. E. Ruiz-Hitzky, M. Darder and P. Aranda in Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, Inc. Hoboken, N.J. (in press).
182 4. 5. 6. 7. 8. 9. 10.
11.
12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.
26. 27. 28. 29.
Ruiz-Hitzky et al. E. Ruiz-Hitzky, M. Darder and P. Aranda, J. Mater. Chem. 15, 3650 (2005). J. H. Choy, S. J. Choi, J. M. Oh and T. Park, Appl. Clay Sci. 36, 122 (2007). S. F. Wang, L. Shen, Y. J. Tong, L. Chen, I. Y. Phang, P. Q. Lim and T. X. Liu, Polym. Degrad. Stab. 90, 123 (2005). M. Darder, M. López-Blanco, P. Aranda, A. J. Aznar, J. Bravo and E. Ruiz-Hitzky, Chem. Mater. 18, 1602 (2006). F. M. Fernandes, A. I. Ruiz, M. Darder, P. Aranda and E. Ruiz-Hitzky, J. Nanosci. Nanotechnol. (in press). J. W. Rhim, Food Sci. Biotechnol. 16, 691 (2007). E. Ruiz-Hitzky, P. Aranda, M. Darder in Bottom-up Nanofabrication: Supramolecules, Self-Assemblies, and Organized Films edited K. Ariga and H. S. Nalwa, American Scientific Publishers, Valencia, USA (2007) Ch. 73. E. Ruiz-Hitzky, M. Darder and P. Aranda in Bio-Inorganic Hybrid Materials edited E. Ruiz-Hitzky, K. Ariga and Y. Lvov, Wiley-VCH, Weinheim, Germany (2007), Ch. 1, p.1. Bio-Inorganic Hybrid Nanomaterials edited E. Ruiz-Hitzky, K. Ariga and Y. Lvov, Wiley-VCH, Weinheim, Germany (2007). R. Chandra and R. Rustgi, Prog. Polym Sci. 23, 1273 (1998). S. Sinha Ray and M. Bousmina, Prog. Mater. Sci. 50, 962 (2005). J. K. Pandey, A. P. Kumar, M. Misra, A. K. Mohanty, L. T. Drzal and R. P. Singh, J. Nanosci. Nanotechnol. 5, 497 (2005). R. A. Gross and B. Kalra, Science 297, 803 (2002). Y. Ikada, H. Tsuji, Macromol. Rapid Comm. 21, 117 (2000). M. Darder, P. Aranda and E. Ruiz-Hitzky, Adv. Mater. 19, 1309 (2007). A. Khare and S. Deshmukh, J. Plast. Film Sheeting 22, 193 (2006). Polymeric Biomaterials edited S. Dumitriu, Marcel Dekker, N.Y. (2002). R. N. Tharanathan, Trends Food Sci. Technol. 14, 71 (2003). V. Siracusa, P. Rocculi, S. Romani, M. Dalla Rosa, Trends Food Sci. Technol. doi:10.1016/j.tifs.2008.07.003. http://www.european-bioplastics.org/index.php?id=182, October 2008. M. Gáspár, Zs. Benkó, G. Dogossy, K. Réczey and T. Czigány, Polym. Degrad. Stab. 90, 563 (2005). I. Olabarrieta (2005). Strategies to improve the aging, barrier and mechanical properties of chitosan, whey and wheat gluten protein films. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-208 (2008-10-18)]. A. Dufresne, Canadian J. Chem.-Rev. Canad. Chimie, 86, 484 (2008). M. Mondragon, K. Arroyo and J. Romero-García, Carbohyd. Polym. 74, 201 (2008). X. Cao, Y. Chen, P. R. Chang, M. Stumborg and M. A. Huneault, J. Appl. Polym. Sci. 109, 3804 (2008). Y. Wang, X. Cao and L. Zhang, Macromol. Biosci. 6, 524 (2006).
Progress in Bionanocomposite Materials 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.
183
H. Angellier, S. Molina-Boisseau, P. Dole and A. Dufresne, Biomacromolecules 7, 531 (2006). H. Angellier, S. Molina-Boisseau and A. Dufresne, Macromol. Symp. 233, 132 (2006). L. Petersson and K. Oksman, Compos. Sci. Technol. 66, 2187 (2006). I. Kvien, J. Sugiyama, M. Votrubec and K. Oksman, J. Mater. Sci. 42, 8163 (2007). P. A. Fowler, J. M. Hughes and R. M. Elias, J. Sci. Food Agric. 86, 1781 (2008). A. Sorrentino, G. Gorrasi and V. Vittoria, Trends Food Sci. Technol. 18, 84 (2007). R. Zhao, P. Torley and P. J. Halley, J. Mater. Sci. 43, 3058 (2008). K. K. Yang, X. L. Wang and Y. Z. Wang, J. Industrial Engineer. Chem. 13, 485 (2007). J. W. Rhim and P. K. W. Ng, Crit. Rev. Food Sci. Nutr. 47, 411 (2007). H. M. Park, M. Misra, L. T. Drzal and A. K. Mohanty, Biomacromolecules 5, 2281 (2004). J. K. Pandey and R. P. Singh, Starch/Stärke 57, 8 (2005). A. C. Wibowo, M. Misra, H.-M. Park, L. T. Drzal, R. Schalek and A. K. Mohanty, Compos. Pt. A-Appl. Sci. Manuf. 37, 1428 (2006). M. Avella, J. J. De Viegler, M. A. Errico, S. Fischer, P. Vacca and M. G. Volpe, Food Chem. 93, 467 (2005). S. Kalambur and S. S. H. Rizvi, J. Appl. Polym. Sci. 96, 1072 (2005). H. M. Park, A. K. Mohanty, L. T. Drzal, E. Lee, D. F. Mielewski and M. Misra, J. Polym. Environ. 14, 27 (2006). D. Ruan, L. Zhang, Z. Zhang and X. Xia, J. Polym. Sci. Polym. Phys. 42, 367 (2003). B.-S. Chiou, D. Wood, E. Yee, S. H. Imam, G. M. Glenn and W. J. Orts, Polym. Engineer. Sci. 47, 1898 (2007). X. Z. Tang, S. Alvi and T. J. Herald, Cereal Chem. 85, 433 (2008). H.-M. Wilhelm, M.-R. Sierakowski, G. P. Souza and F. Wypych, Polym. Int. 52, 1035 (2003). H.-M. Wilhelm, M.-R. Sierakowski, G. P. Souza and F. Wypych, Carbohyd. Polym. 52, 101 (2003). B. Chen and J. R. G. Evans, Carbohydr. Polym. 61, 455 (2005). H. M. Park, X. Li, C. Z. Jin, C.Y. Park, W. J. Cho and C. S. Ha, Macromol. Mater. Eng. 287, 553 (2002). H. M. Park, W. K. Lee, C. Y. Park, W. J. Cho and C. S. Ha, J. Mater. Sci. 38, 909 (2003). Y. Xu, J. Zhou and M. A. Hanna, Cereal Chem. 82, 105 (2005). N. Lilichenko, R. D. Maksimov, J. Zicans, R. Merij Meri and E. Plume, Mechan. Composite Mater. 44, 45 (2008). F. Chivrac, E. Pollet, M. Schmutz and L. Avérous, Biomacromolecules 9, 896 (2008).
184
Ruiz-Hitzky et al.
56.
C. J. Pérez, V. A. Álvarez, I. Mondragón and A. Vázquez, Polym. Int. 57, 247 (2008). X. Cao, Y. Chen, P. R. Chang and M. A. Huneault, J. Appl. Polym. Sci. 106, 1431 (2007). M. Darder, M. Colilla and E. Ruiz-Hitzky, Chem. Mater. 15, 3774 (2003). M. Darder, M. Colilla and E. Ruiz-Hitzky, Appl. Clay Sci., 28, 199 (2005). S. F. Wang, L. Shen, Y. J. Tong, L. Chen, I. Y. Phang, P. Q. Lim, T. X. Liu, Polym. Degrad. Stab. 90, 1234 (2005). J. W. Rhim, Food Sci. Biotechnol. 15, 925 (2006). Y. Xu, X. Ren and M. A. Hanna, J. Appl. Polym. Sci. 99, 1684 (2006). J. W. Rhim, S.-I. Hong, H.-M. Park and P. K. W. Ng, J. Agric. Food Chem. 54, 5814 (2006). J. U. Otaigbe and D.O. Adams, J. Environ. Polym. Degr. 5, 199 (1997). J. W. Rhim, J. H. Lee and H. S. Kwak, Food. Sci. Biotechnol. 14, 112 (2005). K. Dean and L. Yu in Biodegradable Polymers for Industrial Application edited R. Smith, CRC Press, Boca Raton FL, USA (2005) p. 289. P. Chen and L. Zhang, Biomacromolecules, 7, 1700 (2006). J. Yu, G. Cui, M. Wei and J. Huang, J. Appl. Polym. Sci. 104, 3367 (2007). H. Zheng, F. Ai, M. Wei, J. Huang and P. R. Chang, Macromol. Mater. Engin. 292, 780 (2007). F. Ai, H. Zheng, M. Wei and J. Huang, J. Appl. Polym. Sci. 105, 1597 (2007). The Macromolecular Chemistry of Gelatin edited A. Veis, Academic Press, N. Y. (1964). M. Darder, A. I. Ruiz, P. Aranda, H. Van Damme and E. Ruiz-Hitzky, Curr. Nanosci, 2, 231 (2006). O. Talibudeen, Nature 166, 236 (1950). O. Talibudeen, Trans. Faraday Soc. 51, 582 (1955). A. Weiss in Organic Geochemistry edited G. Eglinton and M. T. J. Murphy, Springer Verlag, Berlin, Germany (1969) p. 737. J. P. Zheng, P. Li, Y. L. Ma and K. D. Yao, J. Appl. Polym. Sci., 86, 1189 (2002). Y.Q. Rao, Polymer 48, 5369 (2007). J. F. Martucci, A. Vázquez and R. A. Ruseckaite, J. Thermal Anal. Cal. 89, 117 (2007). S. W. Xu, J. P. Zheng, L. Tong and K. D. Yao, J. Appl. Polym. Sci. 101, 1556 (2006). R. Bhardwaj and A. K. Mohantly, J. Biobased Mater. Bioenergy 1, 191 (2007). S. Sinha Ray and M. Okamoto, Macromol. Rapid Comm. 24, 815 (2003). P. Maiti, K. Yamada, M. Okamoto, K. Ueda and K. Okamoto, Chem. Mater. 14, 4654 (2002). M. Pluta, A. Galeski, A. Alexandre, M.-A. Paul and P. Dubois, J. Appl. Polym. Sci. 86, 1497 (2002). M.-A. Paul, M. Alexandre, P. Degeé, C. Henrist, A. Rulmont and P. Dubois, Polymer 44, 443 (2003).
57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84.
Progress in Bionanocomposite Materials 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102.
103. 104. 105. 106. 107. 108. 109. 110.
185
M.-A. Paul, C. Delcourt, M. Alexandre, P. Degée, F. Monteverde, A. Rulmont and P. Dubois, Macromol. Chem. Phys. 206, 484 (2005). S. Sinha Ray, K. Yamada, M. Okamoto and K. Ueda, Macromol. Mater. Engineer. 288, 203 (2003). M.-A. Paul, C. Delcourt, M. Alexandre, P. Degée, F. Monteverde and P. Dubois, Polym. Degrad. Stab. 87, 535 (2005). R. Hiroi, S. Sinha Ray, M. Okamoto and T. Shiroi, Macromol. Rapid Commun. 25, 1359 (2004). W.-M. Chiu, Y.-A. Chang, H.-Y. Kuo, M.-H. Lin and J.C. Weng, J. Appl. Polym. Sci. 108, 3024 (2008). H.-S. Kim, Y. S. Chae, B. H. Park, J.-S. Yoon, M. Kang and H.-J. Jin, Curr. Appl. Phys. 8, 803 (2008). M. Pietrini, L. Roes, M. K. Patel and E. Chiellini, Biomacromolecules 8, 2210 (2007). L. González, A. Rodríguez and L. Ibarra, J. Appl. Polym. Sci. 43, 1327 (2003). L. Bokobza and J.-P. Chauvin, Polymer 46, 4144 (2005). A. Fakhru’l-Razi, M. A. Atieh, N. Girun, T. G. Chuah, M. El-Sadig and D. R. A. Biak, Compos. Struct. 75, 496 (2006). K. Kueseng and K. I. Jacob, Eur. Polym. J. 42, 220 (2006). L. Zhu and R. P. Wool, Polymer 47, 8106 (2006). H. Uyama, M. Kuwabara, T. Tsujimoto, M. Nakano, A. Usuki and S. Kobayashi, Macromol. Biosci. 4, 354 (2004). Z. S. Liu, S. Z. Erhan and P. D. Calvert, J. Am. Oil Chem. Soc. 81, 605 (2004). Z. E. Liu, Z. Sevim and X. Jingyuan, Polymer 46, 10119 (2005). V. Thomas, D. R. Dean and Y. K. Vohra, Curr. Nanosci. 2, 155 (2006). C. Liu, Z. Xia and J. T. Czernuszka, Chem. Eng. Res. Des. 85, 1051 (2007). M. S. Widmer and A. G. Mikos in Frontiers in tissue engineering edited C. W. Patrick Jr., A. G. Mikos and L. V. McIntire, Elsevier Science Ltd., Oxford (1998) Ch. II.5, p. 107. Y. J. Wang, C. R. Yang, X. F. Chen and N. R. Zhao, Adv. Eng. Mater. 8, 97 (2006). S. Yunoki, T. Ikoma, A. Monkawa, K. Ohta, M. Kikuchi, S. Sotome, K. Shinomiya and J. Tanaka, Mater. Lett. 60, 999 (2006). A. Yokoyama, M. Gelinsky, T. Kawasaki, T. Kohgo, U. König, W. Pompe and F. Watari, J. Biomed. Mater. Res. Part B: Appl. Biomater. 75B, 464 (2005). P. Fratzl, H. S. Gupta, E. P. Paschalis and P. Roschger, J. Mater. Chem. 14, 2115 (2004). P. Fratzl and R. Weinkamer, Prog. Mater. Sci. 52, 1263 (2007). M. C. Rogel, H. Qiu and G. A. Ameer, J. Mater. Chem. 18, 4233 (2008). L. M. Mathieu, T. L. Mueller, P.-E. Bourban, D. P. Pioletti, R. Müller and J.-A. E. Månson, Biomaterials 27, 905 (2006). W. Y. Zhou, S. H. Lee, M. Wang, W. L. Cheung and W. Y. Ip, J. Mater. Sci.-Mater. Med. 19, 2535 (2008).
186
Ruiz-Hitzky et al.
111. H.-W. Kim, J. Biomed. Mater. Res. A 83A, 169 (2007). 112. A. Tampieri, M. Sandri, E. Landi, G. Celotti, N. Roveri, M. Mattioli-Belmonte, L. Virgili, F. Gabbanelli and G. Biagini, Acta Biomater. 1, 343 (2005). 113. V. M. Rusu, C.-H. Ng, M. Wilke, B. Tiersch, P. Fratzl and M. G. Peter, Biomaterials 26, 5414 (2005). 114. A. Sinha, S. Nayar, A. Agrawal, D. Bhattacharyya and P. Ramachandrarao, J. Am. Ceram. Soc. 86, 357 (2003). 115. L. Wang, R. Nemoto and M. Senna, J. Nanoparticle Res. 4, 535 (2002). 116. C. Tsioptsias and C. Panayiotou, Carbohyd. Polym. 74, 99 (2008). 117. N. Sultana and M. Wang, J. Mater. Sci.: Mater. Med. 19, 2555 (2008). 118. H. S. Kim, J. T. Kim, Y. J. Jung, S. C. Ryu, H. J. Son and Y. G. Kim, Macromol. Res. 15, 65 (2007). 119. D. Verma, K. S. Katti and D. R. Katti, Ann. Biomed. Eng. 36, 1024 (2008). 120. K. R. Mohamed, G. T. El-Bassyouni and H. H. Beherei, J. Appl. Polym. Sci. 105, 2553 (2007). 121. H. H. K. Xu, E. F. Burguera and L. E. Carey, Biomaterials 28, 3786 (2007). 122. H. H. K. Xu, M. D. Weir and C. G. Simon, Dent. Mater. 24, 1212 (2008). 123. M. C. Gutiérrez, M. Jobbágy, M. L. Ferrer and F. del Monte, Chem. Mater. 20, 11, (2008). 124. A. L. Daniel-da-Silva, A. B. Lopes, A. M. Gil and R. N. Correia, J. Mater. Sci. 42, 8581 (2007). 125. N. Olmo, M. A. Lizarbe and J. G. Gavilanes, Biomaterials 8, 67 (1987). 126. J. I. Herrera, N. Olmo, J. Turnay, A. Sicilia, A. Bascones, J. G. Gavilanes and M. A. Lizarbe, Biomaterials 16, 625 (1995). 127. N. Olmo, J. Turnay, J. I. Herrera, J. G. Gavilanes and M. A. Lizarbe, J. Biomed. Mater. Res. 30, 77 (1996). 128. K. S. Katti, D. R. Katti and R. Dash, Biomed. Mater. 3, 034122 (2008). 129. D. Depan, B. Kumar and R. P. Singh, J. Biomed. Mater. Res. B-Appl. Biomater. 84B, 184 (2008). 130. P. Bordes, E. Pollet, S. Bourbigot and L. Avérous, Macromol. Chem. Phys. 209, 1473 (2008). 131. Y. Cao, Y. Zhou, Y. Shan, H. Ju and X. Xue, Adv. Mater. 18, 1838 (2006). 132. A. Abarrategi, M. C. Gutierrez, C. Moreno-Vicente, M. J. Hortiguela, V. Ramos, J. L. Lopez-Lacomba, M. L. Ferrer and F. del Monte, Biomaterials 29, 94 (2008). 133. C. M. Voge, M. Kariolis, R. A. MacDonald and J. P. Stegemann, J. Biomed. Mater. Res. Part A 86A, 269 (2008). 134. K. Rezwan, Q. Z. Chen, J. J. Blaker and A. R. Boccaccini, Biomaterials 27, 3413 (2006). 135. Y. Ema, M. Ikeya and M. Okamoto, Polymer 47, 5350 (2006). 136. H. W. Kim, H. H. Lee and J. C. Knowles, J. Biomed. Mater. Res. Part A 79A, 643 (2006).
Progress in Bionanocomposite Materials
187
137. S. Sotome, T. Uemura, M. Kikuchi, J. Chen, S. Itoh, J. Tanaka, T. Tateishi and K. Shinomiya, Mater. Sci. Eng. C 24, 341 (2004). 138. T.-Y. Liu, S.-Y. Chen, J.-H. Li and D.-M. Liu, J. Control. Release 112, 88 (2006). 139. T. Ito, T. Sugafuji, M. Maruyama, Y. Ohwa and T. Takahashi, J. Supramol. Chem. 1, 217 (2001). 140. F. H. Lin, Y. H. Lee, C. H. Jian, J. M. Wong, M. J. Shieh and C. Y. Wang, Biomaterials 23, 1981 (2002). 141. X. Wang, Y. Du and J. Luo, Nanotechnology 19, art. #065707 (2008). 142. X. Wang, X, Pei, Y. Du and Y. Li, Nanotechnology 19, art. #375102 (2008). 143. J. H. Choy, S. Y. Kwak, Y. J. Jeong and J. S. Park, Angew. Chem. Int. Ed. 39, 4042 (2000). 144. S. Y. Kwak, Y. J. Jeong, J. S. Park and J. H. Choy, Solid State Ionics 151, 229 (2002). 145. N. Benmouhoub, N. Simmonet, N. Agoudjil and Th. Coradin, Green Chem. 10, 957 (2008). 146. J. Allouche, M. Boissière, C. Hélary, J. Livage and Th. Coradin, J. Mater. Chem. 16, 3120 (2006). 147. M. Boissière, P. J. Meadows, R. Brayner, C. Hélary, J. Livage and Th. Coradin, J. Mater. Chem. 16, 1178 (2006). 148. L. Medvecky, R. Stulajterova and J. Briancin, Chem. Pap. 61, 477 (2007). 149. S. T. Tan, J. H. Wendorff, C. Pietzonka, Z. H. Jia and G. Q. Wang, Chem. Phys. Chem. 6, 1461 (2005). 150. M. A. Morales, P. V. Finotelli, J. A. H. Coaquira, M. H. M. Rocha-Leao, C. DiazAguila, E. M. Baggio-Saitovitch and A. M. Rossi, Mater. Sci. Eng. C-Biomimetic Supramol. Syst. 28, 253 (2008). 151. A. L. Daniel-Da-Silva, R. Loio, J. A. Lopes-Da-Silva, T. Trindade, B. J. Goodfellow and A. M. Gil, J. Colloid Interface Sci. 324, 205 (2008). 152. E. Ruiz-Hitzky, M. Darder, P. Aranda, M. A. Martín del Burgo and G. del Real (in preparation). 153. C. M. Niemeyer, Angew. Chem. Int. Ed. 40, 4128 (2001). 154. E. J. Anglin, L. Y. Cheng, W. R. Freeman and M. J. Sailor, Adv. Drug Deliv. Rev. 60, 1266 (2008). 155. A. D. McLaren and G. H. Peterson, Nature 192, 960 (1961). 156. C. V. Kumar and A. Chaudhari, J. Am. Chem. Soc. 122, 830 (2000). 157. L. Gao, Q. Gao, Q. Wang, S. Peng and J. Shi, Biomaterials 26, 5267 (2005). 158. T. Coradin, A. Coupé and J. Livage, J. Mater. Chem. 13, 705 (2003). 159. A. J. Patil, E. Muthusamy and S. Mann, J. Mater. Chem. 15, 3838 (2005). 160. C. Forano, S. Vial and C. Mousty, Curr. Nanosci. 2, 283 (2006). 161. E. Han, D. Shan, H. G. Xue and S. Cosnier, Biomacromolecules 8, 971 (2007). 162. J. Livage, T. Coradin and C. Roux in Functional Hybrid Materials edited P. Gómez-Romero and C. Sanchez, Wiley-VCH, Weinheim, Germany (2004) Ch. 11, p. 387.
188
Ruiz-Hitzky et al.
163. 164. 165. 166. 167.
I. Gill and A. Ballesteros, Trends Biotechnol. 18, 282 (2000). R. Gupta and N. K. Chaudhury, Biosen. Bioelectr. 22 2387 (2007). H. Frenkel-Mullerad and D. Avnir, J. Am. Chem. Soc. 127, 8077 (2005). D. Avnir, T. Coradin, O. Lev and J. Livage, J. Mater. Chem. 16, 1013 (2006). N. Nassif, C. Roux, T. Coradin, O. M. M. Bouvet and J. Livage, J. Mater. Chem. 14, 2264 (2004). Y. A. Shchipunov, T. Y. Karpenko, I. Y. Bakunina, Y. V. Burtseva and T. N. Zvyagintseva, J. Biochem. Biophys. Methods 58, 25, (2004). Y. Shchipunov and N. Shipunova, Colloids Surfaces B Biointerf. 63, 7 (2008). H. R. Luckarift, M. B. Dickerson, K. H. Sandhage and J. C. Spain, Small 2, 640 (2006). Y. Lu, Y. Liu, J. Xu, C. Xu, B. Liu and J. Kong, Sensors 5, 258 (2005). A. El Kadib, K. Molvinger, C. Guimon, F. Quignard and D. Brunel, Chem. Mater. 20, 2198 (2008). A. Bianco and M. Prato, Adv. Mater. 15, 1765 (2003). E. Katz and I. Willner, Chem. Phys. Chem. 5, 1085 (2004). Y. D. Wang, P. P. Joshi, K. L. Hobbs, M. B. Johnson and D. W. Schmidtke, Langmuir 22, 9776 (2006). G. D. Liu and Y. H. Lin, J. Nanosci. Nanotech. 6, 948 (2006). X.-L. Luo, J.-J. Xu, J.-L. Wang and H.-Y. Chen, Chem. Comm., 2169 (2005). Y. Zhang, Y. Shen, D. Han, Z. Wang, J. Song, F. Li and L. Niu, Biosens. Bioelectron. 23, 438 (2007). M. Zhang, A. Smith and W. Gorki, Anal. Chem. 76, 5045 (2004). L. Chen and G. Lu, J. Electroanal. Chem. 597, 51 (2006). Y. Li, X. Lin and C. Jiang, Electroanal. 18, 2085 (2006). Z.-G. Wang, Z.-K. Xu, L.-S. Wan, J. Wu, C. Innocent and P. Seta, Macromol. Rapid Comm. 27, 516 (2006). S. Chakraborty and C. R. Raj, J. Electroanal. Chem. 609, 155 (2007). J. Galandova, G. Ziyatdinova and J. Labuda, Anal. Sci. 24, 711 (2008). D. Nepal, S. Balasubramanian, A. L. Simonian and V. A. Davis, Nano Lett. 8, 1896 (2008). A. Morozan, L. Stamatin, F. Nastase, A. Dumitru, S. Vulpe, C. Nastase, I. Stamatin and K. Scott, Phys. Status Solidi A-Appl. Mat. 204, 1797 (2007). M. C. Gutierrez, Z. Y. Garcia-Carvajal, M. J. Hortiguela, L. Yuste, F. Rojo, M. L. Ferrer and F. del Monte, J. Mater. Chem. 17, 2992 (2007). M. C. Gutierrez, M. J. Hortiguela, J. M. Amarilla, R. Jimenez, M. L. Ferrer and F. del Monte, J. Phys. Chem. C 111, 5557 (2007). M. I. H. Panhuis, A. Heurtematte, W. R. Small and V. N. Paunov, Soft Matter. 3, 840 (2007). S. Srivastava, B. Samanta, P. Arumugam, G. Han and V. M. Rotello, J. Mater. Chem. 17, 52 (2007).
168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190.
Progress in Bionanocomposite Materials
189
191. S. Srivastava, B. Samanta, B. J. Jordan, R. Hong, Q. Xiao, M. T. Tuominen and V. M. Rotello, J. Am. Chem. Soc. 129, 11776 (2007). 192. S. F. Wang, Y. M. Tan, D. M. Zhao and G. D. Liu, Biosens. Bioelectron. 23, 1781 (2008). 193. V. Rocher, J. M. Siaugue, V. Cabuil and A. Bee, Water Res. 42, 1290 (2008). 194. M. Fujishima, H. Takatori, K. Yamai, Y. Nagao, H. Kitagawa and K. Uchida, J. Mater. Sci. 43, 3130 (2008). 195. J. Kulys and R. Stupak, Open Nanosci. J. 2, 34 (2008). 196. L. Ding, C. Hao, Y. D. Xue and H. X. Ju, Biomacromolecules 8, 1341 (2007). 197. A. Corma, P. Concepcion, I. Dominguez, V. Fornes and M. J. Sabater, J. Catal. 251, 39 (2007). 198. X. B. Lu, Z. H. Wen and J. H. Li, Biomaterials 27, 5740 (2006). 199. M. Darder, M. López-Blanco, P. Aranda, F. Leroux and E. Ruiz-Hitzky, Chem. Mater. 17, 1969 (2005). 200. M. Darder and E. Ruiz-Hitzky, J. Mater. Chem. 15, 3913 (2005). 201. P. Aranda, M. Darder, R. Fernández-Saavedra, M. López-Blanco and E. RuizHitzky, Thin Solid Films 495, 104 (2006). 202. A. Gómez-Avilés, M. Darder, P. Aranda and E. Ruiz-Hitzky, 46, 923 (2007). 203. R. Fernández-Saavedra, M. Darder, A. Gómez-Avilés, P. Aranda and E. RuizHitzky, J. Nanosci. Nanotechnol. 8, 1741 (2008). 204. J. H. An and S. Dultz, Appl. Clay Sci. 36, 256 (2007). 205. L. Wang and A. Q. Wang, J. Hazard. Mater. 147, 979 (2007). 206. H. Gecol, P. Miakatsindila, E. Ergican, and S. R. Hiibel, Desalination 197, 165 (2006). 207. H. Qiu, J. Yu and J. Zhu, Polym. Polym. Compos. 13, 167 (2005). 208. J. P. Zhang, L. Wang and A. Q. Wang, Ind. Eng. Chem. Res. 46, 2497 (2007). 209. L. Wang, J. P. Zhang and A. Q. Wang, Colloids Surfaces A-Phys. Eng. Aspects 322, 47 (2008). 210. A. Pourjavadi, M. Ayyari and M. S. Amini-Fazl, Eur. Polym. J. 44, 1209 (2008). 211. S. D. Bhat and T. M. Aminabhavi, Sep. Purif. Technol. 51, 85 (2006). 212. Y. Ding, D. J. Jones, P. Maireles-Torres and J. Rozière, Chem. Mater. 7, 562 (1995). 213. T. Coradin, J. Allouche, M. Boissière and J. Livage, Curr. Nanosci. 2, 219 (2006). 214. A. I. Ruiz, M. Darder, P. Aranda, R. Jiménez, H. Van Damme and E. Ruiz-Hitzky, J. Nanosci. Nanotechnol. 6, 1602 (2006).
CHAPTER 4 MESOPOROUS SILICA NANOPARTICLES: SYNTHESIS AND APPLICATIONS
Juan L. Vivero-Escoto, Brian G. Trewyn and Victor S.-Y. Lin* Department of Chemistry and U.S. Department of Energy Ames Laboratory, 0755 Gilman Hall, Iowa State University, Ames, Iowa 50011-3111, USA *Email:
[email protected]
This review chapter outlines the recent breakthroughs in the synthesis and applications of mesoporous silica nanoparticle (MSNs) materials. The current state of knowledge for the surface functionalization of MSNs is also reviewed. We highlight the research developments of MSN-based selective catalysts. Recent advancements in designing MSN nanodevices for applications of biotechnology and biomedicine are discussed, with focus on the areas of controlled release drug/gene delivery, biosensor, and cell imaging. Furthermore, we describe new opportunities that MSN materials could potentially bring to the exciting research field of nanotechnology.
1. Introduction In the last two decades, the world has witnessed an unprecedented pursuit and breakthroughs on the synthesis, characterization, and application of a wide variety of nanoparticles, such as quantum dots, carbon nanotubes, dendrimers, polymers, and inorganic (metal and metal oxide) nanoparticles.1-3 In the case of metal oxides, silica-based nanoparticles has been a particularly exciting research area since the beginning of 1990’s. In contrast to solid silica nanoparticles, structurally ordered mesoporous silica materials, such as MCM-4-6 and SBA-type7
191
192
Vivero-Escoto et al.
mesoporous silicas, offer many advantageous features, such as high surface area, narrow pore size distribution, and high chemical and physical stability. These materials attracted much attention worldwide due to their potential application in catalysis, chromatography, controlled release delivery, sensor design, and new semiconducting nanostructures.8-16 Unfortunately, unlike other monodisperse nanoparticles, most of the mesoporous silica materials are amorphous in particle morphology. Therefore, it is difficult to control and tune the masstransport and surface property of these materials. To fully take advantage of the unique mesoporous structure, it is important to gain the ability of tuning the particle morphology and chemical properties of mesoporous silicas. In this review, we summarize the recent development on the morphology control, such as the synthesis of shape-, size-, and structuretunable mesoporous silica nanoparticle materials (MSNs).17-21 As detailed in the following sections, these synthetic breakthroughs for MSNs not only preserve the outstanding porous structural property of MCM-type materials, but also render much better morphological and chemical properties. This new class of nanoparticle materials opened the doors of promising applications in the fields of catalysis, biotechnology, and biomedicine.16, 22, 23 Moreover, the current pursuit of selectively functionalize the interior/exterior surface of MSNs with a variety of organic moieties has brought the synthesis of organic-inorganic hybrid materials with outstanding functions. For instance, we and others have synthesized MSNs systems for catalysis using the synergy between different organo-functional groups and MSNs. In contrast to many silicabased materials, this MSNs material not only played the role of catalyst support, but also served as an active component of the catalytic mechanism.22, 24 Supramolecular chemistry has also been applied on the surface of MSN. Stoddart, Zink, and co-workers have developed design strategies for organic-inorganic hybrid materials using nanovalves (rotaxanes, and [2]-pseudorotaxanes) and MSNs. These assembles exhibit interesting properties for controlled release applications.12, 25, 26 In addition to the organic-inorganic approach, our research group has investigated a variety of chemical interactions between MSNs and different nanoparticles. We
Mesoporous Silica Nanoparticles
193
have demonstrated the use of inorganic nanoparticles in combination with MSNs for the construction of a new generation of stimuliresponsive controlled release nanodevices. We were able to use monodisperse gold (Au-NP),23 iron oxide (Fe3O4-NP),27 and cadmium sulfide nanoparticles (CdS-NP)10 as removable caps for the regulation of various guest molecules in-and-out of the porous framework of MSN.16 As result of these breakthroughs, novel applications have been developed, such as the internalization of Au-NPs-MSNs in plant cells,23 the efficient transfection of DNA vectors to cancer cells using polyamidoamine dendrimer PAMAM-NPs-MSNs,16 and the controlled release of guest molecules through Fe3O4-NPs-MSNs triggered by the intracellular reducing environment.27 Recently, the applications of MSNs for cell imaging using optical and magnetic resonance (MR) imaging agents have also been demonstrated.28-32 These are a few examples of the current progress of biotechnological applications of MSNs materials that will be highlighted in this review. 2. Synthesis of Mesoporous Silica Nanoparticles The synthesis of structurally ordered (MCM-type) mesoporous silica materials was first reported by scientists of Mobil Company in 1992.4-6 Their method is based on the condensation of silica precursors (i.e, sodium silicate, tetramethylammonium silicate, or tetraethyl orthosilicate) in presence of structure-directing agents (quaternary ammonium cationic surfactants) under hydrothermal and basic conditions. The scope of this micelle-templated approach was extended by a number of variations leading to a wide variety of mesoporous silica materials, such as SBA-,7 MSU-,33 and FSM-types of mesoporous silica.34, 35 In contrast to other crystalline microporous silica materials, such as zeolites, mesoporous silicas, while containing no crystalline domains, exhibit a high degree of pore order, high thermal and hydrolytic stability, narrow pore size distribution, and high surface areas. However, the large size of mesoporous silica (above micron in diameter) and the irregularity in particle shape prevent the applications of these materials in areas such as separation (chromatography) biotechnology, and medicine (Figure 1a).36, 37 Given the fact that particle morphology and surface functionalization
194
Vivero-Escoto et al.
are of primordial importance for controlling the physical, chemical, and biological properties of the mesoporous silicas, new synthetic strategies for controlling the particle morphology and surface functionalization of these materials are needed for the further advancement of this burgeoning research field.
(a)
(b)
Figure 1. SEM micrographs of amorphous silica (a) and MSNs (b).
2.1. Control of Morphology The capability of tuning the morphology of MCM-type materials has a direct outcome in regulating the mass-transport properties of guest molecules (separation and catalysis) and improving the biocompatibility of mesoporous silica particles as demonstrated by our group and others.38, 39 Unger and co-workers,19 were first in reporting the morphology control of MCM-type mesoporous silica particle materials. Their approach was based on the Stöber process for the synthesis of monodisperse silica sphere. They modified this method including a cationic surfactant as structure-directing agent following the procedure previously reported for the synthesis of MCM-41. MSNs with MCM-41 type structure, spherical shape, and submicrometer-size (400-1100 nm) were obtained. Following this approach, several groups have reported on the synthesis of morphologically controlled mesoporous silica particles. For example, Cai et al. tuned the morphology of MSNs simply by selecting different pHadjusting agents.20 This procedure yielded hexagonally ordered nanospheres (roughly 100 nm in diameter) and nanorods (700-1000 nm in length, and 300-500 nm in width). Mou and co-workers,21, 40
Mesoporous Silica Nanoparticles
195
developed a pH-controlled method. The synthesis was carried out through a fast transfer from the original acid solution to basic media modifying the kinetics of silica condensation. This method afforded well order type MCM-41 nanorods and disordered nanospheres. Mann and coworkers41 developed a simple pH quenching procedure of the alkaline synthesis of MCM-41. This method yielded MSNs with hexagonally ordered channels and a wide range of sizes. The common strategy reported in these synthetic procedures was to combine the Stöber process and the micelle-templated approach that is typical for the synthesis of MCM-type materials with modifications in the reaction conditions (pH, temperature, addition time, co-solvents, catalysts, etc). Recently, a novel methodology to synthesize small-size MSNs (20-50 nm) with a wellordered hexagonal arrangement was report by Imai and co-workers.42 The new approach for this synthesis is the use of two surfactants; a cationic surfactant as a structure-directing agent, and a block copolymer as nonionic surfactant to suppress the growth of the silica particles. In addition to this method, Stein and co-workers,43, 44 showed the use of colloidal crystals as templates to synthesize MSNs with cubic or spherical shapes having dimensions from 50-150 nm. While these reports describe useful methods for controlling the particle morphology, the resulting mesoporous silica particles are pure inorganic. The surface functionalization needs to be conducted postsynthetically. For example, the surface functionalization of organic groups could be done with the common “grafting” method with organoalkoxysilanes. As investigated by Stein and others,45 this method would lead to inhomogeneous distribution of organic groups on mesoporous surface because of the diffusional mass-transport limit. To circumvent this problem, our group has developed methods for controlling the morphology of MSNs while incorporating organic functional groups simultaneously by using the co-condensation method.18 This approach relies in the use of different organoalkoxysilanes (OAS) for the synthesis of MSNs. In addition to manipulate the shape of MSNs, this procedure also resulted in multifunctionalized materials. In this method different parameters have to be considered; such as, concentration, molecular size, and hydrophilicity/hydrophobicity of the OAS precursors. By tuning these parameters, we obtained different
196
Vivero-Escoto et al.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Figure 2. SEM micrographs of AP- (a), AAP- (b), AEP- (c), UDP- (d), ICP- (e), CP- (f), Al- (g), and MSNs (h).18
morphologies ranging from spherical, twisted columns, rod-shape, to tubular particles (Figure 2). All of these MSNs materials showed a high surface area (> 600 m2/g), narrow pore size distribution (2.0-2.9 nm), and different mesopore structures as summarized in Table 1. The working
197
Mesoporous Silica Nanoparticles
Table 1. Structural and morphological (size and shape) properties of organofunctionalized MSNs. Samplea
Sb (m2/g)
Vb (cm3/g)
MDb (nm)
OASb (mmol/g)
Shape
PSb
AP-
721.7
0.45
2.37
1.7
Tubular
1-3 µm
AAP-
664.6
0.48
2.59
0.7
Twisted Columns
1-3 µm
AEP-
805.8
0.57
2.60
1.0
Spherical
0.5-2 µm
UDP-
1022.4
0.78
2.86
0.9
Spherical
0.5-2 µm
ICP-
840.1
0.66
2.58
1.5
Spherical
0.1-0.5 µm
CP-
1012.5
0.68
2.35
1.4
Rod
0.5-1 µm
AL-
1080.5
0.65
1.97
1.7
Rod
0.05-0.5 µm
MCM-41
767.1
0.55
2.55
Spherical
0.3-0.6 µm
a
MSNs were functionalized with: 3-aminopropyltrimethoxysilane (AP-); N-(2-aminoethyl)3-amino-propyltrimethoxysilane (AAP-); 3-[3-(2-aminoethylamino)ethylamino]propyltrimethoxysilane (AEP-); ureido-propyltrimethoxysilane (UDP-); 3-isocyantopropyltrimethoxysilane (ICP-); 3-cyano-propytriethoxysilane; allyl-trimethoxysilane (AL-) b BET surface area (S); Mesopore volume (V); Mesopore diameter (MD); Amount of organic group (OAS); and Particle size (PS)
principle is based on the hydrophilic/hydrophobic interaction between the OAS precursors and the CTAB surfactant micelle. To further gaining the morphology control of MSNs by this methodology, we studied the effect of tuning the concentrations of OAS introduced in the synthesis of these materials.46 For instance, in the case of using 3-[2-(2aminoethylamino)ethylamino]propyltrimethoxysilane (AEP-TMS) and 3-cyanopropyl trimethoxysilane (CP-TMS) mixture, particles with different sizes ranging from 3 µm to less than 500 nm were obtained. The different particle shapes obtained from this procedure varied from spherical to rods. Moreover, we demonstrated that the particle morphology of these materials could be tuned and controlled by the hydrophilicity of OAS (AEP-TMS) groups.
198
Vivero-Escoto et al.
Table 2. Structural and morphological (size and shape) properties of MSNs synthesized with RTILs. RTILs
Sb (m2/g)
Vb (cm3/g)
MDb (nm)
Shape/Mesopore Structure
C14MIM-MSN
729
0.664
2.71
Spherical (Hexagonal)
C16MIM-MSN
924
0.950
3.03
Ellipsoids (Hexagonal)
C18MIM-MSN
893
0.995
3.27
Rod (Pseudo-moiré)
C14OCMIM-MSN
639
0.695
2.61
Tubular (Wormhole)
a
MSNs were synthesized with: 1-tetradecyl-3-methylimidazolium bromide (C14MIMBr); 1hexadecyl-3-methylimidazolium bromide (C16MIMBr); 1-octadecyl-3-methylimidazolium bromide (C18MIMBr); and 1-tetradecyl-oxymethyl-3-3methylimidazolium chloride (C14OCMIMCl) b BET surface area (S); Mesopore volume (V); Mesopore diameter (MD); Amount of organic group (OAS); and Particle size (PS)
Another strategy developed by our group to fine-tune the morphology of MSNs was the use of room-temperature ionic liquids (RTILs) as structure-directing agents.47 Interestingly, the MSNs obtained through this method showed various particle shapes such as spheres, ellipsoids, rods, and tubes. We also obtained different mesopores structure, such as hexagonal, rotational moiré type, and wormhole-like (Table 2). It is note worthy to mention that in contrast to Tatsumi’s material,48 the rotational moiré pattern of mesoporous obtained in this report was based on an achiral surfactant. Finally, we used these MSNs for the controlled release of the RTILs as antibacterial agents. In this work we demonstrated the crucial role played by the morphology of MSNs in the release of guest molecules (RTILs) and its effect in their antibacterial behavior. In addition to these achievements, we also published a report on the synthesis of large pores MSNs (3-6 nm) by using a pore expanding-agent (mesitylene).49 The synthesis of this large pore MSNs was carried out to achieve the goal of releasing proteins (i.e. cytochrome c) inside mammalian cells.
Mesoporous Silica Nanoparticles
199
From the exciting results aforementioned we can conclude that we have the ability to synthesize MSNs with the following unique features: • High surface area and large pore volume • Chemical and mechanical stability • Unique mesoporous structure • Narrow and tunable pore size • Well define and monodisperse particle shape • Tunable particle size 2.2. Control of Surface Functionalization The capability of developing methods for controlling the degree of organic functionalization is important in order to fine-tune the chemical properties of MSNs. In this respect, one of the interesting advantages of MSNs is the fact that these materials have two functional surfaces; one external (outside of the particles) and the other one internal (inside of the channels). The ability of selectively and efficiently functionalizing these surfaces is crucial for controlling the chemical and physical performances of MSN. Generally, two well-established methods for the functionalization of MSNs materials have been applied: co-condensation, and grafting.45, 50 The co-condensation method has the advantages of homogeneous surface coverage in one-pot synthesis, better control over the amount of OAS groups incorporated in the MSNs, and the possibility of using a wide variety of organo-functional groups.45 However, this method is not useful to functionalize the external surface of MSNs and there is a limitation in the amount of OAS groups incorporated to the material without compromising the mesostructure. On the other hand, the grafting procedure yields hydrothermally stable MSNs, and can selectively functionalize the exterior surface of MSNs materials.50 The grafting method also presents its own drawbacks, such as heterogeneous surface coverage and the OAS groups tend to be grafted in the opening of the mesoporous blocking the interior of the channels. To overcome these drawbacks some novel approaches have been reported in the literature.51 For instance, our group used the co-condensation method to functionalize the MSNs with a wide
200
Vivero-Escoto et al.
variety of OAS groups (Table 1).18 OAS with different molecular size, hydrophobic/hydrophilic properties, and concentrations were successfully incorporated into the MSNs channels. The co-condensation of two different OAS groups was also carried out,46 which opened the pathway for the synthesis of more sophisticated MSNs for catalysis and biosensors applications, as will be described later. Beside our group, Mann and co-workers have also used the co-condensation strategy to functionalized MSNs through a controlled quenching procedure.52 Tuning the amount of OAS groups incorporated to MSNs by carefully designing the interfacial electrostatic interaction between the surfactant head group and the OAS precursor was investigated by our group.53 We used three anionic groups (thiolate, carboxylate, and sulfonate) to functionalize the interior of MSNs using our previously reported co-condensation method.18 We found that the sulfonate-MSNs had the highest amount of OAS groups (1.56 mmol/g); carboxylateMSNs had 0.97 mmol/g; and finally the thiolate-MSNs incorporated 0.56 mmol/g of OAS groups in the MSNs. Interestingly, these results followed the anionic lyotropic series reported by Larsen and Magid.54 They found that the interactions of anions with the CTAB surfactant micelles followed the order: citrate < CO32- < SO42- < CH3CO2- < F- < OH- < HCO2- < Cl- < NO3- < Br- < CH3C6H4SO3-. We demonstrated that the loading of OAS groups to MSNs could be controlled by the interaction between the anionic group of OAS precursor and the positive charge of the structure-directing agent (CTAB). As was previously mentioned, organic-functionalization of MSNs by the grafting method can selectively decorate either the exterior or interior surface of the material. The synergy yields by this method in combination with the co-condensation method affords materials with very interesting properties. For instance, our group published a report on the internal surface functionalization of MSNs with o-phtalic hemithiacetal (OPTA) group using the co-condensation method, and later the external surface of this MSNs material was functionalized with poly(lactic acid) via the grafting method (Scheme 1).55 This nanodevice was used for the selective sensing of amino-containing neurotransmitters. In this case, the poly-(lactic acid) polymer coated in the exterior of MSNs had the role of gatekeeper to allow the diffusion of specific
Mesoporous Silica Nanoparticles
201
neurotransmitters. The OPTA group was the sensor that detected the neurotransmitters inside the MSN channels. Recently, we used a similar approach to selectively functionalize MSNs with poly(Nisopropylacrylamide) polymer (PNiPAm) in the exterior of the surface. This PNiPAm-coated MSNs exhibited interesting water solubility behavior which is controlled by the critical solution temperature of the polymer.56
Scheme 1. Poly(lactic acid)-coated MSNs with fluorescence probe inside the channels. This system was synthesized based on our capability to functionalize at will both of the surfaces (external/internal) of MSNs.
Different factors can be involved in the efficient performance of the grafting method, such as solvent, temperature, type of OAS precursor, and reaction time. Asefa and co-workers have studied the influence of the polarity and dielectric constant of different solvents in the grafting method.57 Three main properties were affected by the solvent: concentration of grafted groups, degree of site-isolation, and catalytic properties. The researchers found that polar-protic solvents resulted in lower concentrations of grafted groups, higher site-isolated organic groups, and more efficient catalytic properties. On the contrary, dipolaraprotic and nonpolar solvents resulted in larger concentrations of grafted groups, more densely populated organic groups, and poor to efficient catalytic properties.
202
Vivero-Escoto et al.
3. Catalysis The key role that heterogeneous catalysis might play in the development of environmentally-friendly processes (production of chemicals, and petroleum chemistry) has led to a growth in environmentally-related research in the past decade.14 Specifically, silicabased materials such as zeolites have found extensive applications in this field. They have been used as acids, base, and/or redox catalysts.3 However, zeolites present several pore-space limitations regarding to the diffusion of reactants, mainly when large molecules are involved. These mass-transfer limitations have been reduced with the development of larger pore size silica materials.8 Towards this respect, MSN materials present unique possibilities to overcome this limitation. For instance, the capability of tuning the microenvironment of MSNs such as the control of surface functionalization (hydrophobicity/hydrophilicity of the surface), the incorporation of catalytic functions; the control of particle size, and shape afford suitable materials for their application as catalysts.22, 58-60 In this section we will describe some of the strategies that have been developed from our group and others to exploit the aforementioned unique features of MSN materials for their application in catalysis. 3.1. Cooperative Catalysis (Acid/Base) Mimicking the extraordinary catalytic systems found in nature has attracted a lot of attention in the catalysis field. In particular, the development of strategies based on enzymatic principles have been pursued for a long time.61 It is well-established that in the case of carbonyl chemistry, enzymes use a cooperative approach between acid and base residues in their active sites to catalyze the carbonyl activation reaction.62 In order to reproduce this strategy in a synthetic system is clearly necessary to be able to functionalize a solid support with control of spatial distribution and concentration of the acid and base functional groups. We recently published a report on the synthesis and application of multifunctionalized MSN that showed cooperative acid-base
203
Mesoporous Silica Nanoparticles
catalysis.59 The MSNs were functionalized with ureidopropyl (UDP) and 3-[2-(2-aminoethylamino)ethylamino]propyl (AEP) groups. The working principle of this MSN-based catalyst system relies on the ability of the UDP group to activate the carbonyl group,62 and the primary and secondary amines of the AEP group than function as a nucleophile to catalyze aldol, Henry, and cyanosilylation reactions (Scheme 2). Different amounts of organic groups were loaded to react with acetone, 4-nitro-benzaldehyde, and trimethylsilyl cyanide in the reactions aforementioned. We found that indeed there is a synergistic interaction between both functional groups inside the MSNs (Table 3). The reaction rates of the aldol, Henry, and cyanosilylation reactions with the multifunctionalized MSNs were accelerated up to four times. We observed that the most effective ratio is 2/8 AEP/UDP-MSNs for all three model reactions. The higher turnover numbers (TONs) obtained for these model reactions in comparison with different control experiments supported our hypothesis that the improvement in their performance is due to a cooperative effect between the acid (UDP) and nucleophile (AEP) catalyst.
a)
OH
O
O
MSNs catalyst acetone
b)
+ O2N
O2N
O
O2N
O
O MSNs catalyst CH3NO2
O2N
c)
O2N OSi(CH3)3
O
CN
MSNs catalyst (CH3)3SiCN O2N
O2N
Scheme 2. Three model reactions to evaluate the cooperative acid/base concept promoted by MSNs catalyst: aldol reaction (a), Henry reaction (b), and cyanosilylation (c).
204
Vivero-Escoto et al. Table 3. Turnover number (TON) for the MSNs-catalyzed reactions. Reaction
MSNs catlysta
TON
Aldol
2/8 AEP/UDP
22.6
5/5 AEP/UDP
11.9
8/2 AEP/UDP
8.6
AEP
5.4
UDP
0.0
2/8 AEP/UDP
125.0
5/5 AEP/UDP
91.1
8/2 AEP/UDP
65.8
AEP
55.9
UDP
5.8
2/8 AEP/UDP
276.1
5/5 AEP/UDP
170.5
8/2 AEP/UDP
109.4
AEP
111.4
UDP
45.9
Henry
Cyanosilylation
a
MSNs were co-condensated with: 3-[3-(2-aminoethylamino)ethylamino]propyltrimethoxysilane (AEP-); ureido-propyltrimethoxysilane (UDP-)
Asefa and co-workers, followed a similar dual catalysis approach for the Henry reaction of p-hydroxybenzaldehyde.24, 63 However, in this example the silanols groups of the MSNs were used as carbonyl activating sites, and aminopropyl group was grafted as the nucleophilic catalyzing agent. 3.2. Gatekeeping Effect Another approach that enzymes generally use to achieve selectivity is modifying the environment around their active sites.61 By means of physicochemical properties, such as hydrophobicity and polarity, enzymes can selectively catalyze specific substrates. Mimicking this strategy, our group recently controlled the selectivity for competitive
205
Mesoporous Silica Nanoparticles
nitroaldol reaction by using bifunctionalized MSNs.22 Different MSN materials were synthesized combining AEP- as primary group with UDP-, mercaptopropyl (MP-) and allyl (AL-) as secondary groups. These secondary groups were judiciously chosen base on their hydrophilic/hydrophobic properties; UDP- has a hydrophilic character; while MP-, and AL- are more hydrophobic. Thus, we were able to tune the nano-environment around the catalytic group (AEP-). A series of alkoxybenzaldehydes were selected for the nitroaldol reaction (Scheme 3). These alkoxybenzaldehydes showed different hydrophilic/hydrophobic performances depending on the size of the chain. The mono- and bifunctionalized MSNs containing AEP- and AEP/UDP did not show any selectivity regarding to reactants. However, when more hydrophobic secondary groups were used (MP- or AL-) the selectivity of the reaction increased toward the nonpolar, more hydrophobic alkoxybenzaldehyde reactants (Table 4). These results suggested that the hydrophobic secondary groups (MP- and AL-) work as a gatekeepers by selectively allowing more hydrophobic reactants to penetrate into the channels and react with the nucleophilic group (AEP). O NO2
H HO
Product A
HO
MSNs Catalyst O CH3NO2
NO2
H *
n O
*
NH2 n=0 n=3 n=7
Product B
n O n=0 n=3 n=7
HN
NH R = -CH2CH2CH2NHCNNH2 (UDP) -CH2CH2SH (MP) -CH2CH=CH2 (AL) R
MSNs
Scheme 3. Schematic representation of the MSNs-based system developed for testing the competitive nitroaldol reaction.
206
Vivero-Escoto et al.
Table 4. MSNs-catalyzed competitive nitroaldol reactions using the gatekeeping concept. Product Ratio B/Aa
Catalyst Series 1
b
Series 2
Series 3
Series 4
Series 5
AEP
0.97
1.01
0.92
0.79
0.88
AEP/UDP
1.16
1.08
1.10
0.93
0.91
AEP/MP
1.22
1.62
2.21
1.00
1.06
AEP/AL
1.50
1.83
2.58
1.16
1.54
a
TON of product B/TON of product A Product A = 4-(2-nitro-vinyl)-phenol. Series 1: Product B = 1-methoxy-4-(2-nitrovinyl)-benzene; Series 2: Product B = 1-buthoxy-4-(2-nitro-vinyl)-benzene; Series 3: Product B = 1-hepthyloxy-4-(2-nitro-vinyl)-benzene; Series 4: Product B = 1-methyl-4(2-nitro-vinyl)-benzene; Series 5: Product B = 1-nitro-4-(2-nitro-vinyl)-benzene b
Scheme 4. Schematic representation for the oxidative polymerization of 1,4diethynylbenzene into conjugated oligo(phenylene butadynylene) using Cu-MSN and Cu-MAL as catalysts.60
3.3. Other Applications in Catalysis In addition to the cooperative acid/base catalysis and the gatekeeping effect, other applications have been demonstrated based on the ability of tuning the morphology and surface functionalization of MSN materials. For instance, we reported the synthesis of a Cu2+ functionalized MSN
Mesoporous Silica Nanoparticles
207
materials (Cu-MSNs) to catalyzed the formation of a highly conjugated poly(phenylene butadiynylene) polymers (PPB) (Scheme 4).60 Some drawbacks in the synthesis of this type of conducting polymers inside silica mesopores, such as low content and lack of the structural alignment of the polymer within the pores of the material. Our ability to tune the quantity and spatial distribution of catalytic sites, the size and morphology of the pores of MSNs allowed us to control the diffusion rate of monomeric precursors and their local concentrations near their catalytic sites. We demonstrated that through a co-condensation reaction a Cu-MSN material with the suitable amount of catalytic sites and mass transport properties can be obtained. A high degree of polymerization and a linearly aligned PPB conducting polymer was successfully obtained from catalytic Cu-MSNs. The Cu-MSN performance was superior when we compared to a mesoporous alumina material (Cu-MAL) and the bulk polymerization in solution. In addition to the aforementioned report, our group also developed a heterogeneous nucleophilic catalytic system for Baylis-Hillman, acylation, and silylation reactions (Scheme 5).58 This MSNs system is based on the co-condensation of 4-(dimethylamino)pyridine (DMAP) inside of the mesopore channels. The DMAP-MSN material catalyzed the Baylis-Hillman reaction of aryl aldehydes and various α,β unsaturated ketones. High reactivity and selectivity was obtained for the reaction with methyl vinyl ketones. We concluded that the high selectivity of the DMAP-MSNs catalyst is attributed to the “matrix effect”. In the case of the acylation reaction, secondary alcohols reacted rapidly under our reaction conditions. Several control experiments were carried out to demonstrate the efficiency of DMAP-MSNs to catalyze acylation reactions. We used the acylation reaction of 1-(1naphthyl)ethanol to test the recyclability of DMAP-MSN materials. We found that the catalyst was highly recyclable, as it was reused in 10 consecutive cycles. We did not observe a reduction in yield, or chemical/thermal decomposition, or collapse of the structure along this recyclability test. The stability of DMAP-MSNs was also confirmed using cyclohexanol as a substrate. We also demonstrated the successful silylation reaction of several substrates by the DMAP-MSN catalysts.
208
Vivero-Escoto et al.
(a) N NH
N (1) NaH
N
TEOS, CTAB, NaOH(aq)
(2) Cl
N
Si(OEt)3
Si(OEt)3
Co-condensation Reaction
N
DMAP-MSN Si O O O
DMAP-TES
(b) Baylis-Hillman reaction
+ R
OH O
O
O
R
R'
H
OH R'
+
O
R
O R'
+
R'
O R
DMAP-MSN O 1
O R'
R' 2
3
(c) Acylation O
O
O
ROH + O
DMAP-MSN
RO
(d) Silylation
ROH +
Cl Si
RO Si DMAP-MSN
Scheme 5. Synthesis of DMAP-MSNs catalyst (a); and three model reactions to test its performance, Baylis-Hillman reaction (b), acylation reaction (c), and silylation reaction (d).
4. Biotechnological and Biomedical Applications Research and development of nanomaterials in the fields of biotechnology and biomedicine is just a few decades old.1, 2 In the particular case of mesoporous silica-based materials, their potential for biotechnological and biomedical applications started to be exploited at the beginning of this century.11, 64 Previously, the lack of regular shape and size control made silicon-based materials highly toxic.65-70 However, with the development of the aforementioned synthetic methods for controlling the shape-, size-, and functionalization of MSNs; the opportunity of extending their applications in these fields was possible. The applications of MSN materials in the fields of biotechnology and biomedicine have mainly taken three routes: the controlled release of a wide variety of cargos, such as drugs, proteins, and genes; biosensors; and cell imaging. In this section we will describe the most important
209
Mesoporous Silica Nanoparticles
developments and we will highlight the latest advances in these three areas of research. In addition we will summarize some of the basic studies done in the topics of cellular uptake and toxicity of MSNs.
(a)
(b)
Figure 3. TEM images of HeLa cells with endocytosed MSNs.16
4.1. Uptake and Intracellular Performance of MSNs The uptake of MSNs into cells has been widely demonstrated by our group and others (Figure 3).16, 23, 38, 39, 71, 72 The understanding of the internalization mechanism and the final fate of MSNs in the cells is of primordial importance for future development and application. The uptake of extracellular materials into the cell cytoplasm is divided in three different pathways: phagocytosis, micropinocytosis, and endocytosis.73 Phagocytosis and micropinocytosis are pathways to uptake material in the micron-size range. Endocytosis is the mechanism in which fluids, dissolved solutes, and suspended macromolecules are internalized by cells.74 Endocytosis is sub-divided in three different processes; clathrin-dependent, caveolin-dependent, and clathrin/caveolin independent endocytosis.73 The mechanisms in which the MSNs are internalized by cells will determine the pathways that this material will follow intracellularly (i.e. endosomes, lyposomes, etc) before reaching the cytoplasm. The unique features required by MSNs to achieve these goals, such as the release of its cargo, targeting specific sites, sensing intracellular processes or tuning cells metabolism is dependent on the
210
Vivero-Escoto et al.
endocytic pathway. Since the beginning of this exciting research area, our group has been actively involved in its development. We published the first report on the effects of surface functionalization on the uptake of MSNs by cervical cancer cells (HeLa).38 The external surface of MSNs was functionalized with different organo functional groups, fluorescein derivative (FITC-MSN), aminopropyl (AP-MSN), guanidinium derivatives (GP-, and GEGP-MSNs) and folic acid (FAP-MSNs). We found that the external charge of MSNs influences its mechanism of uptake by HeLa cells (Table 5). Moreover, we demonstrated that the endocytosis of MSNs could be tuned by chemically binding specific receptors to its surface (i.e. folic acid). Finally, in this report we were also able to prove that MSNs escaped from endosomes. These findings are crucial for the future progress of MSNs in the area of drug delivery. Mou and co-workers also studied the internalization of MSNs in 3T3-L1 cells.39, 71 Their results confirmed our aforementioned findings. Recently, Zink and co-workers studied the energy-dependence of the cellular uptake of MSNs.75 By incubating the cells with MSNs at different temperatures, they found that the uptake of MSNs is higher at 37°C than at 4°C. This finding demonstrated that the uptake of MSNs is driven by an energy-dependent process. Table 5. Influence of MSNs surface functionalization on its endocytosis pathway in HeLa cells. Material
Surface Potential (mV)
Endocytosis Pathway
FITC-
-34.73 ± 3.50
Clathrin-mediated
AP-
-4.68 ± 1.54
Caveolae-mediated
GP-
-3.25 ± 0.275
Caveolae-mediated
GEGP-
+0.57 ± 0.095
Not determined
FAP-
+12.81 ± 1.60
Clathrin and Receptormediated
In addition to understanding the pathways of internalization of MSNs into cells, another critical issue for the future progress of MSNs in the fields of biotechnology and biomedicine is the in vitro and in vivo toxicity. It has been reported that amorphous silica has a high toxic effect
Mesoporous Silica Nanoparticles
211
for cells.67, 69, 70 In contrast we and others have demonstrated the efficient uptake and biocompatibility of MSNs.38, 71, 76 We have shown that HeLa cells survive and grow for 4-6 days in presence of MSNs in concentrations below 100 µg/mL.16, 38 Recently, Goodisman and coworkers studied the cytotoxicity of MSNs and SBA-15 materials.77 The mitochondrial O2 consumption was reported in HL-60, and Jurkat cells. MSNs material had no effect on the respiration rate according to the authors. In contrast, SBA-15 showed inhibition of cellular O2 consumption in the range of 25-500 µg/mL. This confirmed that the physical properties of MSNs indeed play an important role in toxicity. Our group has also studied the effect of the morphology (shape and size) on the uptake of MSNs by mammalian cells.76 In this report two nanoparticles were investigated, tubes (600 nm in length and 100 nm in width) and spheres (80-100 nm in diameter). The cellular uptake and kinetics were evaluated in Chinese hamster ovarian (CHO) and human fibroblast cells. We found that the uptake of MSNs was both shape and cell-line dependent. CHO cells showed a more efficient uptake capacity than fibroblast cells for both materials. Moreover the rate of uptake was different in the case of fibroblast cells. MSNs with spherical shape reached 100% uptake in 180 min, while tube-shaped MSNs needed 360 min to achieve the same uptake percentage. Vallhov and co-workers also studied the effect of MSN material sizes on human monocytederived dendritic cells (MDDC).78 Two materials with spherical shape, but different size (270 nm and 2.5 µm) were tested. The authors demonstrated that the size and concentration have an effect on the viability, uptake, and immune regulatory markers. They found that smaller MSNs in lower concentrations affected the MDDC viability to a minor degree. Interestingly, their results showed that both materials are localized in different locations inside the MDDC. To further develop the biomedical applications of MSNs in vivo some properties of these materials are necessary to investigate, such as their bio-distribution, uptake efficiency, circulation behavior, long term toxicity, and final fate in living animals. Recently, Mou and co-workers published a report on the cellular uptake efficiency, toxicity, and circulation behavior of MSNs in mice.30 They synthesized a fluorescein labeled MSNs fused to amorphous silica shells of Fe3O4 nanoparticles
212
Vivero-Escoto et al.
(Mag-Dye@MSNs). This system allowed the investigators to locate the materials inside both the cells and animals by either magnetic resonance imaging or fluorescence microscopy. They found low Mag-Dye@MSNs cytotoxicity even in concentrations higher than 200 µg/mL in rat bone marrow stromal cells and NIH 3T3 fibroblast cells. In the case of in vivo experiments, they administered Mag-Dye@MSNs aqueous suspensions through eye vein injection in mice. The authors reported that MagDye@MSNs tended to accumulate in the liver and spleen tissue rather than in kidney. After preliminary results the investigators did not find any abnormal clinical sign in the Mag-Dye@MSNs treated mice during a four-week study period. This report provides some baseline results for future MSNs bio-applications in vivo.
Scheme 6. Schematic representation of the control drug delivery systems (CDS) based on NPs-MSNs materials and disulfide-link trigger system.
4.2. Controlled Delivery Systems The use of MSNs as controlled delivery systems (CDS) has been a major focus in biotechnology applications. Some reasons for this increased interest are the high surface area that facilitates loading a large amount of cargo (drugs, gene, proteins, nutrients or bioimaging agents); the ability of MSNs to be developed into smart site-specific materials
Mesoporous Silica Nanoparticles
213
through independently external and internal functionalization that could control the release of cargo; the efficient uptake of MSNs by mammalian and plant cells; the capping of the pores and achieving zero-premature release and finally, its proven biocompatibility. All these unique features make MSNs a suitable material to achieve the ultimate goal of a controlled-delivery system. This goal includes the release of a pharmaceutical drug, gene or nutrient at the desire time in the specific location without side effects.79, 80 Our group has developed an army of drug delivery systems based on stimuli-responsive caps.10, 16, 23, 27 Our research has focused in the development of nano-caps using inorganic and organic nanoparticles for controlling the release of cargo. For instance, CdS-,10 Fe3O4-,27 and Au-,23 nanoparticles have been used as caps (Scheme 6). These inorganic NPs are attached to the MSNs through a disulfide bond, which is chemicallylabile and can be cleaved by the reducing environment of the cell. One of the outstanding features of these CDS capped with inorganic NPs is that all of these systems showed “zero premature release” before the addition of a reducing agent. The successful performance of these systems was demonstrated by releasing bioimaging agents (i.e. fluorescein, and Texas red), ATP/vancomycin, and β-estradiol in solution and in vitro. In the case of CdS-NPs-MSNs system, a highly efficient loading of vancomycin and ATP was reported, 83.9 and 30.3 mol%, respectively. This CDS was tested using mercaptoethanol (ME) and dithiothreitol (DTT) as gate-opening triggers for releasing the cargo. To investigate its biocompatibility and performance in vitro, ATP-loaded CdS-NPs-MSN materials were introduced into an astrocyte culture. We demonstrated the efficient release of ATP triggered by ME in this cell culture. Following the development of these CDS capped with inorganic NPs we recently reported on the design, synthesis and application of Fe3O4-NPs-MNS materials. In addition to its excellent capping/releasing properties, this CDS could be moved at will in the presence of an external magnet. The Fe3O4-NPs-MSNs system was successfully tested using the bioimaging agent fluorescein as cargo. Less than 1% of fluorescein leached was detected after 132 h proved the excellent performance as “zero premature release” system. After the addition of either DTT or dihydrolipoic acid (DHLA) 40 and 31.4% of the fluorescein loaded was released,
214
Vivero-Escoto et al.
respectively. The biocompatibility and endocytosis of the Fe3O4-NPsMSNs system was successfully demonstrated in HeLa cells. In addition to these breakthroughs, our group also reported, for the first time, the successful application of MSN-based CDS in plant cells. Two systems were developed; the first one contained a fluorescein (FITC)-labeled triethylene glycol (TEG) functionalized MSNs, and the second was capped with Au-NPs. FITC-TEG-MSN materials were used to investigate the influence of MSN functionalization on the endocytosis of plant cells. Tobacco mesophyll protoplasts were used as model systems. We proved the successful internalization of FITC-TEG-MSNs (7±3%) and we found that the MSNs material remained inside of endocytic vesicles for the entirety of the experiment (72 h). On the contrary, FITCMSNs without TEG in the external surface was not endocytosed by the protoplasts. The Au-NPs-MSNs system was used to address the application of MSNs in intact plant tissues using the gene gun system. As a proof-of-principle, we investigated the GFP expression on tobacco cotyledon and maize immature embryos. Indeed, we were able to visualize the GFP-expressing foci cotyledons and maize embryos. One of the unique advantages of this MSN system is its potential of delivering different species in a controlled fashion simultaneously. In this published report we tested this concept using generated transgenic tobacco containing an inducible promoter controlled GFP gene. In this way the expression of GFP in plants can only be observed when the chemical promoter (β-estradiol) is presented. To investigate the performance of our CDS, β-estradiol was loaded into the mesopores of the MSNs and capped with Au-NPs via a disulfide bond, finally the material was coated with GFP vector. Following the release of the promoter (β-estradiol) triggered by DTT, plant cells efficiently expressed the GFP marker gene carried by the Au-NPs-MSNs system. We successfully demonstrated the use of Au-NPs-MSNs as double release system. In addition to these inorganic NP-capped CDS, our group also developed organic NP-capped CDS. Our approach was based on the use of chemically attached polyamidoamine (PAMAM) dendrimers to MSNs.16 A bioimaging molecule (Texas Red), was loaded inside the MSN channels. We observed the efficient uptake of this material by
Mesoporous Silica Nanoparticles
215
HeLa cells through confocal fluorescence microscopy. Finally, pEGFP vector was attached by electrostatic interaction with the positively charged PAMAM G-2 on the surface of MSNs. We demonstrated the successful deliver of pEGFP gene inside of HeLa cells. Interestingly, we found that the transfection efficiency was even better than some commercial transfection reagents. In a different approach, Stoddart, Zink and co-workers developed CDS based on MSNs and supramolecular nanovalves (SNVs).12, 26, 81 These SNVs are based on rotaxanes and [2]-pseudorotaxanes that work as gatekeepers to block the pores of MSN materials (Scheme 7). Different trigger mechanisms have been used, such as light,82 redox,83 pH,84 and competitive binding.85 These SNVs-MSN systems have been successfully evaluated for the release of different types of bioimaging agents. However, because the main driving force for the binding between the stalks and the ring (moving part) is based on electrostatic interaction, the release experiments with these CDS materials have been carried out in organic solvents only. This disadvantage makes the use of these systems prohibitive in biological conditions.
Scheme 7. Graphical representation of CDS based on SNVs-MSNs materials using different trigger startegies (pH, redox, and photo).
216
Vivero-Escoto et al.
Table 6. CDS-MSNs for the release of guest molecules based on different capping and triggering strategies. Cargo
Cap
Triggered by:
Released on:
Ref.
ATP, vancomycin
CdS-NPs
Reducing agents (DTT, ME)
Physiological conditions
10
Fluorescein
Fe3O4-NPs
Reducing agents (DTT, DHLA)
Physiological conditions and in vitro (HeLa cells)
27
β-estradiol, DNA
Au-NPs
Reducing agent (DTT)
In vitro (plant cells)
23
Cholestane, phenanthrene
Coumarin
UV light (λ=310 nm)
Organic solvent (n-hexane)
91
Ru(bipy)32+
Diethylenetriamine
pH, anions
Physiological conditions
89
ATP, DNA
PAMAM
Reducing agents (DTT, TCEP)
Physiological conditions and in vitro (HeLa cells)
16
Ir(ppy)3
[2]-Pseudorotaxane
NaCNBH3
Organic solvent (Toluene:EtOH/ 1:1)
81
Ir(ppy)3, Rhodamine B
[2]-Rotaxane
Fe(ClO4)3, ascorbic acid
Organic solvents (MeCN)
83
Calcein
[2]-Pseudorotaxane PEI-α/γCD
pH
Physiological conditions
87
Rhodamine B
[2]-Pseudorotaxane Cucurbit[6]uril
pH
Physiological conditions
86
Rhodamine B
[2]-Pseudorotaxane αCD
Enzymatic (Porcine liver esterase)
Physiological conditions
88
To overcome this drawback Stoddart, Zink and co-workers have developed a new generation of SNVs using rings with stronger interaction toward the stalks in aqueous solutions. For example, they developed a MSNs system containing bisammonium stalks and cucurbit[6]uril as moving part.86 The release principle in this SNVsMSNs material is based on the variation of pH. This system was successfully tested in aqueous solution for the release of Rhodamine B. Similar approach was reported by Kim and co-workers; however, in this case α/γ-cyclodextrins were used as moving rings.87 Recently, Zink and
Mesoporous Silica Nanoparticles
217
co-workers developed a CDS using α-cyclodextrins (α-CD) as SNVs.88 The α-CD ring is held by a polyethylene glycol thread and kept in place by a large endgroup (adamantyl). This group is chemically bonded to MSNs through an enzymatically cleavable bond. This is the first example of an enzyme-responsive CDS based on MSNs. As proof for the working principle the release of Rhodamine B was triggered by porcine liver esterase in aqueous media. The use of organic/inorganic NPs and SNVs as caps for the controlled release in solution of different guests in MSN systems have been very successful (Table 6). A few lessons have been learned from these initial approaches; such as the possibility of releasing the cargo from MSNs at will using different stimuli-responsive systems. These wide variety of stimuli-responsive systems could be used to take advantage of the different conditions found in cancer and normal cells. Plus, these hybrid systems have the ability of eliminating cargo leaching before reaching the desired target, and most of the systems have been proven to be biocompatible. A different strategy is to chemically bind small organic molecules to the walls of MSNs that could partially block the opening of their pores. For instance, Martinez-Mañez and co-workers attached 3-[2-(2aminoethylamino)ethylamino]propyl-trimethoxysilane to the openings of the MSNs pores (AEP-MSNs).89, 90 The AEP- group by itself or with anions served as gatekeepers for selectively release of [Ru(bipy)3]2+. The working principle for the first approach is based on the hydrogenbonding interaction between amines at neutral pH and their coulombic repulsions in acidic conditions. In the case of the anion-controlled mechanism, the interaction between different anions and the amines at certain pH is responsible for the gatekeeping effect. In addition to this pH-controlled system, Tanaka and others demonstrated using photoresponsive molecules, coumarin and azobenzene derivativesc as caps.91, 92 The advantages of this system was the reversible pore opening and closing process. Moreover, these photo-responsive groups might have another functions such as nano-impellers to enhance the mass transfer flow of guest molecules from the interior of MSNs channels toward the bulk solution. For instance, Zink and co-workers functionalized the
218
Vivero-Escoto et al. Table 7. CDS-MSNs for the release of biogenic agents in vitro. Biogenic agent
Cell culture
Ref.
Cytochrome c
HeLa
49
ATP/Vancomycin
Astrocytes
10
Camptothecin
PANC-1, Capan-1, AsPC-1, SW-480, and MKN-45
94
pEGFP vector
HeLa
16
Propidium Iodide/Camptothecin
PANC-1, and SW-480
72
Paclitaxel
PANC-1
75
interior of MSNs with azobenzene derivatives.93 They used this strategy as photo-controlled nano-impeller for the release of Coumarin 540A. Most of the CDS-MSNs systems described so far have been tested based on the successful release of bioimaging agents (i.e., fluorescein, Texas Red, [Ru(bipy)3]2+, coumarin, and Rhodamine B). However, the ultimate goal of CDS-MSNs is the release of biogenic agents in vitro and in vivo conditions (Table 7). To achieve this goal, our group has shown the efficient release of pEGFP vector, vancomycin and ATP in vitro using NPs as caps (PAMAM or CdS-NPs).10, 16 Recently Zink, Tamanoi, and co-workers demonstrated the efficient release of different hydrophobic anticancer drugs (Table 7).72, 75, 94 For instance, they used an aminopropyl/phosphate modified MSN materials for the release of campthotecin into human cancer cells (PANC-1, AsPC-1, Capan-1, MKN45, and SW480).72 They also demonstrated the photo-driven release of camptothecin and propidium iodide (PI) through nanoimpellers MSN materials in PANC-1 and SW 480 cells.94 In addition to the release of camptothecin and PI, Zink and co-workers have also demonstrated the release of paclitaxel in PANC-1 cells.75 A major target that has been pursued for long time in the drug delivery field is the release of proteins inside human cells and tissues. To achieve this goal it is necessary to have CDSs that are able to protect the protein from the environment until the final target is reached and
Mesoporous Silica Nanoparticles
219
efficiently release the protein without changing its biological activity.95 Our group was able to synthesize extended pore diameter MSNs (LPMSNs) capable of carrying and delivering a membrane impermeable protein (cytochrome c) in HeLa cancer cells.49 We demonstrated that this LP-MSN material efficiently soaked up and released cytochrome c in PBS solution. We found that the release of cytochrome c can be controlled by pH; low pH (5.2) faciliated the release of the protein; meanwhile at neutral pH (7.4) we did not observe a significant release of cytochrome c. The change in electrostatic interactions between LP-MSN material and cytochrome c accounts for this release performance. While LP-MSNs is negatively charged at pH 7.4 (-25.5 mV) a dramatic shift is observed at pH 5.2 (-1.81 mV); in the case of cytochrome c, we did not observe any significant variation in surface charge regardless of pH (from +15 to +17 mV). We confirmed the enzymatic activity of cytochrome c, after release by the oxidation of 2,2’-azino-bis(3ethylbenzthiazoline-6-sulfonate) by hydrogen peroxide. In this report we demonstrated the successful intracellular release of cytochrome c in cancer cells using LP-MSN material as carrier. 4.3. Biosensors The development of biosensor systems based on nanoparticles is a burgeoning research field because of its potential application to a wide range of analytical targets, such as clinical diagnosis, food industry, environmental monitoring and bioassays.96-98 The unique features make MSN material superior immobilization matrices for sensing molecules. In our group we have been pursuing the development of MSN-based biosensors for the detection of neurotransmitters for the understanding of interneuronal chemical communication. We developed a fluorescence sensory system that recognizes amino-containing neurotransmitters, such as dopamine and glucosamine (Scheme 8).64 The system contains an amine-sensitive o-phthalic hemiacetal (OPTA) group. OPTA group has the particular feature that after reacting with molecules containing amine groups a highly fluorescent isoindole is preduced. Besides, the size selectivity inherent to mesoporous materials, we decorated the pores with different secondary functional groups (silanols, propyl, phenyl, and
220
Vivero-Escoto et al.
pentafluorophenyl) to enhance the substrate selectivity by tuning pore hydrophobicity. We indeed observed that the microenvironment played an important role in the reactivity of glucosamine and dopamine with OPTA-MSN materials. For instance, the reaction rate of glucosamine with the silanol containing material is one order of magnitude slower than that of dopamine. The differences in noncovalent interactions of both neurotransmitters with OPTA-MSNs can account for that outcome. Glucosamine might generate strong dipolar interactions with the silanols in the surface of the pore openings slowing the diffusion process within the channels. In the same way, dopamine binding was dependent on the microenvironment of the pore. The reaction rates of dopamine with the pentafluorophenyl-MSN materials were much higher than those of propyl- and phenyl-MSN materials. Hydrophobic π-π interactions can account for this behavior. This report demonstrated that noncovalent interactions inside the channels play an important role for the development of MSN-based biosensors.
Scheme 8. Schematic illustration of MSNs-based biosensor for detecting aminocontaining neurotransmitters.
Mesoporous Silica Nanoparticles
221
Following our interest in the development of biosensors for aminecontaining neurotransmitters, we published a report on the synthesis and application of a poly(lactic-acid)-OPTA-functionalized MSNs (PLAOPTA-MSNs). This system consists of OPTA groups anchored inside the channels of MSNs and the exterior was decorated with poly(lacticacid).55 PLA-OPTA-MSNs were developed to selectively recognize neurotransmitters with similar features, such as dopamine, tyrosine, and glutamic acid. In contrast to our previous work, poly(lactic acid) worked as gatekeeper to regulate the penetration of molecules in and out of the mesopores. To demonstrate the gatekeeping effect the aforementioned neurotransmitters were used as proof-of-principle. Interestingly, we found that dopamine gave the most significant increase in fluorescence intensity. Moreover, the reaction rates for dopamine, tyrosine, and glutamic acid are different by a factor of 4, 10, and 57, respectively. The working principle is based on the electrostatic interactions produced at pH 7.4 between the components of the system. At this pH, PLA-OPTAMSNs, tyrosine, and glutamic acid are negatively charged, while dopamine is positively charged. The positive charged dopamine had a favorable interaction with PLA coating on the MSN and, thus, more readily entered the mesopores. In this report we demonstrated that PLAOPTA-MSNs material is able to selectively recognize between three similar amine-containing neurotransmitters. The last decade has seen an increased intensities in anion recognition.99, 100 Martinez-Mañez and co-workers, reported on the performance of nanoscopic gate-like MSN system for anion recognition.89 The system is based on polyamine-functionalized MSNs materials (PAMSNs). The principle of the system is based on the interaction between the partially protonated polyamines at neutral pH with anions through hydrogen-bonding and electrostatic attractive forces. When the system recognizes the anion, the gate closes stopping the release of the guest molecule ([Ru(bipy)3]2+) from the pores of MSNs. The authors studied different anions, such as fluoride, chloride, bromide, iodide, nitrate, phosphate, sulfate, acetate, and carbonate without notable differences in the release of the bioimaging agent. In contrast, upon the addition of adenosine triphosphate (ATP) and adenosine diphosphate (ADP) the release of [Ru(bipy)3]2+ was stopped as a clear indication of the
222
Vivero-Escoto et al.
recognition of ATP/ADP by this MSN sensor. Interestingly, this effect was not observed upon the addition of guanosine monophosphate, which demonstrates some degree of specificity in this system. The authors proved that PA-MSNs efficiently recognized ATP and ADP against other anions, opening the door to further developments in the area of biosensors using advanced hybrid MSNs systems. 4.4. Multimodal Cell Imaging Multimodal techniques are quickly becoming important tools for developing breakthroughs in the areas of biomedical research, clinical diagnosis, and therapeutics.32 For instance, tracking the distribution of soft tissues in vivo for distinguishing anatomical images and assess disease pathogenesis by biomarkers is crucial for therapeutical treatments.101 The development of methods to determine the fate and distribution of transplanted stem cells is vital for finding future advances in this area.102, 103 The use of inorganic nanoparticles in the biomedical fields has resulted in the development of several techniques for cell imaging. For example, Au nanoparticles, semiconducting quantum dots, and magnetic nanoparticles have found a wide application in detection of biomolecules such as DNA and cancer markers, optical imaging of small organelles and tumors, and cellular magnetotransduction signaling and magnetic resonance imaging (MRI) agents.104-109 However, despite such outstanding progresses these techniques still suffer from some drawbacks such as low target sensitivity, poor spatial resolution, low tissue penetration, and each technique is limited to a single imaging modality.32 To overcome these issues the combination of different imaging methods into a single system has been proposed. Because the unique properties of MSNs; such as biocompatibility, optical transparency, easy modification to embed nanoparticles (i.e. Au, and Fe3O4), and functionalize with optical groups (fluorescein, Rhodamine B); and high surface area have attracted attention as suitable material for multimodal imaging and multifunctional probes (Scheme 9). For instance, Mou and co-workers demonstrated the use of silica coated core-sell superparamagnetic iron oxide nanoparticles attached to fluorescein incorporated MSNs (Mag-Dye@MSNs).28 The authors
Mesoporous Silica Nanoparticles
223
observed the efficient internalization of Mag-Dye@MSNs into human mesenchymal stem cells (hMSCs). The hMSCs were easily imaged with a clinical 1.5-T MRI system. One of the major problems of stem-cell tracking is the low labeling efficiency of magnetic nanoparticles, in this work the authors proved that Mag-Dye@MSNs showed a highly magnetic labeling efficiency. Finally, the authors tested the performance of Mag-Dye@MSNs in vivo implanting with the hMSCs at the olfactory cortex of the brain in mice. Recently, the same group published a report on the performance of Mag-Dye@MSNs in vivo.30 After MagDye@MSNs material was injected to mice, the authors found that the biodistribution of Mag-Dye@MSNs is localized mainly in the spleen and liver, with very little material found in the kidney. After a long term study (three months), the material was still localized in the liver and spleen. This is a strong indication that Mag-Dye@MSNs are resistant to decomposition and not easily excreted from the body. The authors did not find any toxic effect during these experiments.
Scheme 9. Schematic representation of multi-task system for multimodal cell imaging and drug release.
224
Vivero-Escoto et al.
In a different approach, Lin and co-workers grafted MSNs with gadolinium chelates.31 This Gd-MSNs hybrid material was successfully tested in vitro (murine monocytes cell line) and in vivo (mice) for MRI. The authors demonstrated that Gd-MSNs is a highly efficient T1 contrasting agent for intravascular MR imaging and an excellent T2 contrasting agent for MR imaging of soft tissues. Recently some groups have focused on the development of multifunctional MSN materials. In addition to the multimodal cell imaging strategy, these researchers have take advantage of the high surface area of MSNs to release drugs in vitro. For example; Zink, Tamanoi, and co-workers have published a report on the synthesis of cancer cell-specific MSN system containing both optical and magnetic resonance (MR) imaging capability, and able to release hydrophobic cancer drugs.29 This MSN system consists of embedded superparamagnetic iron oxide nanocrystals (20 nm) for MR imaging and magnetic manipulation capabilities; fluorescein isothiocyanate (FITC) were attached to the MSNs surface for optical imaging, folic acid moieties were grafted to MSNs surface for cancer cell-specificity, and this system was loaded with camptothecin and paclitaxel as model chemotherapeutic drugs. The MR imaging capability of this system was tested in solution and in vitro. Pancreatic cancer cells, PANC-1, were used to evaluate the contrasting effect using a clinical MR imaging instrument. The results obtained proved that this MSNs material could be used as MR contrast agent in solution and inside cells. The release of cancer drugs was carried out in cancer cells lines PANC-1 and BxPC3. The researchers found a cell growth inhibition effect after the addition of camptothecin- and paclitaxel-loaded MSN materials to cell cultures. The effect of folic acid functionalization was tested with cancer cells PANC-1 and human foreskin fibroblasts (HFF). Cancer cells PANC-1 over express folate receptors. The authors found that folic acid functionalized MSN systems increased its uptake more than two-fold in PANC-1 against HFF cells. Indeed, the cell growth inhibition effect produced by the release of camptothecin was higher in PANC-1 cells than in HFF cells. The authors demonstrated that multifunctional MSN systems can be successfully designed and applied in vitro.
Mesoporous Silica Nanoparticles
225
5. Conclusions and Outlook
In this review we have described and discussed some of the recent advancements in the design, synthesis and applications of MSNs in the fields of catalysis, biotechnology, and biomedicine. The development of methods for controlling the surface properties and particle morphology of well-defined MSNs materials with high surface area and stable chemical and mechanical properties have significantly impacted their applications. However, despite all these outstanding developments in the area of morphology control and surface functionalization of MSNs there is still a lack of complete understanding of some basic aspects in the synthesis of MSNs. What are the driving forces that account for the synthesis of MSNs with different morphologies and sizes? Can we control these factors to afford MSNs with the desire morphology, pore size/structure, and particle size? What are the quantity limitations of organic groups that can be reacted in both surfaces of the MSNs? Can we further improve the control of the distribution and spatial location in multifunctionalized MSN materials? Is it possible to find a simple and general method to quantify the amount of organic groups functionalized on MSNs? What are the mass-transport properties associated with each morphology and pore structure of MSNs? These are just some of the questions that have to be overcome as the progress continues in this field. We envision that in the near future a better understanding of the factors that affect the synthesis of MSNs will be achieve. Moreover new synthetic methods with better morphology, size and pore structure control will be develop in this burgeoning area of research. Mimicking enzymatic strategies in MSN catalysts has brought outstanding advancements such as the cooperative and gatekeeping catalysis effect. However, new breakthroughs in the synthesis of MSNs materials with better control of the distribution and spatial location of functional groups in the interior and exterior surfaces will result in novel catalytic applications. Moreover, we envision that two areas of MSN catalysis will bring exciting results. The first area is the synthesis or postsynthesis of MSNs in combination with other metal oxides (TiO2, ZrO2, MnO, etc) which can result in new catalysts based on MSN materials. Further, the nanospace formed by the mesoporous and the capability of
226
Vivero-Escoto et al.
functionalize the external surface of MSNs may be excellent tool to exploit this material for asymmetric catalysis. The partial understanding of the cellular uptake mechanisms and cytotoxicity of MSNs has opened the possibility of applying MSNs as CDS, biosensor, and cell imaging systems in vitro with a wide variety of cells. However, some important issues, such as cytotoxicity, cellular uptake and circulation performance of MSNs in living animals have to be completely understood before practical human-level applications can be developed. The application of MSNs to cell tracking has opened the area of multifunctional MSNs. These multi-task materials have several advantages over simple CDS because we can tune the MSNs to work as controlled delivery systems, imaging machines, and biosensors at the same time. In biosensor applications the look for non-invasive systems to measure time-specific properties of cells and organisms has been a longterm goal for the community. Because its biocompatibility and ability to tune its physical, chemical, and biological properties, MSNs will make an important contribution in this field. As progress continues with the design, synthesis, and application of MSNs materials, we envision that further breakthroughs will impact their performance in catalysis, biotechnology, biomedicine and other areas of research. Acknowledgments
The authors thank U.S. National Science Foundation (CHE-0239570; CHE-0809521), U.S. Department of Energy, Office of Basic Energy Sciences (DE-AC02-07CH11358), and Plant Science Institute at Iowa State University for financial support. References 1. 2. 3.
G. A. Ozin and A. C. Arsenault, ‘Nanochemistry: A Chemical Approach to Nanomaterials’, 2005. V. Rotello and Editor, ‘Nanoparticles: Building Blocks for Nanotechnology’, 2004. J. Cejka, H. van Bekkum, and Editors, ‘Zeolites and Ordered Mesoporous Materials: Progress and Prospects. (1st FEZA School held in Prague, Czech Republic 20-21 August 2005.) [In: Stud. Surf. Sci. Catal., 2005; 157]’, 2005.
Mesoporous Silica Nanoparticles 4.
5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
227
J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge, K. D. Schmitt, C. T. W. Chu, D. H. Olson, E. W. Sheppard, and et al., J. Am. Chem. Soc., 1992, 114, 10834. A. Monnier, F. Schuth, Q. Huo, D. Kumar, D. Margolese, R. S. Maxwell, G. D. Stucky, M. Krishnamurty, P. Petroff, and et al., Science, 1993, 261, 1299. C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli, and J. S. Beck, Nature, 1992, 359, 710. D. Zhao, J. Feng, Q. Huo, N. Melosh, G. H. Frederickson, B. F. Chmelka, and G. D. Stucky, Science, 1998, 279, 548. A. Corma, Chem. Rev., 1997, 97, 2373. A. Stein, Adv. Mater., 2003, 15, 763. C.-Y. Lai, B. G. Trewyn, D. M. Jeftinija, K. Jeftinija, S. Xu, S. Jeftinija, and V. S. Y. Lin, J. Am. Chem. Soc., 2003, 125, 4451. M. Vallet-Regi, A. Ramila, R. P. del Real, and J. Perez-Pariente, Chem. Mater., 2001, 13, 308. S. Angelos, E. Johansson, J. F. Stoddart, and J. I. Zink, Adv. Funct. Mater., 2007, 17, 2261. A. B. Descalzo, R. Martinez-Manez, F. Sancenon, K. Hoffmann, and K. Rurack, Angew. Chem., Int. Ed., 2006, 45, 5924. A. Taguchi and F. Schueth, Microporous Mesoporous Mater., 2004, 77, 1. F. Schuth and W. Schmidt, Adv. Mater., 2002, 14, 629. D. R. Radu, C.-Y. Lai, K. Jeftinija, E. W. Rowe, S. Jeftinija, and V. S. Y. Lin, J. Am. Chem. Soc., 2004, 126, 13216. R. I. Nooney, D. Thirunavukkarasu, Y. Chen, R. Josephs, and A. E. Ostafin, Chem. Mater., 2002, 14, 4721. S. Huh, J. W. Wiench, J.-C. Yoo, M. Pruski, and V. S. Y. Lin, Chem. Mater., 2003, 15, 4247. M. Gruen, I. Lauer, and K. K. Unger, Adv. Mater., 1997, 9, 254. Q. Cai, Z.-S. Luo, W.-Q. Pang, Y.-W. Fan, X.-H. Chen, and F.-Z. Cui, Chem. Mater., 2001, 13, 258. H.-P. Lin and C.-Y. Mou, Acc. Chem. Res., 2002, 35, 927. S. Huh, H.-T. Chen, J. W. Wiench, M. Pruski, and V. S. Y. Lin, J. Am. Chem. Soc., 2004, 126, 1010. F. Torney, B. G. Trewyn, V. S. Y. Lin, and K. Wang, Nat. Nanotechnol., 2007, 2, 295. A. Anan, K. K. Sharma, and T. Asefa, J. Mol. Catal. A: Chem., 2008, 288, 1. E. Johansson, E. Choi, S. Angelos, M. Liong, and J. I. Zink, J. Sol-Gel Sci. Technol., 2008, 46, 313. S. Angelos, M. Liong, E. Choi, and J. I. Zink, Chem. Eng. J., 2008, 137, 4. S. Giri, B. G. Trewyn, M. P. Stellmaker, and V. S. Y. Lin, Angew. Chem., Int. Ed., 2005, 44, 5038.
228
Vivero-Escoto et al.
28.
H.-M. Liu, S.-H. Wu, C.-W. Lu, M. Yao, J.-K. Hsiao, Y. Hung, Y.-S. Lin, C.-Y. Mou, C.-S. Yang, D.-M. Huang, and Y.-C. Chen, Small, 2008, 4, 619. M. Liong, J. Lu, M. Kovochich, T. Xia, S. G. Ruehm, A. E. Nel, F. Tamanoi, and J. I. Zink, ACS Nano, 2008, 2, 889. S.-H. Wu, Y.-S. Lin, Y. Hung, Y.-H. Chou, Y.-H. Hsu, C. Chang, and C.-Y. Mou, ChemBioChem, 2008, 9, 53. K. M. L. Taylor, J. S. Kim, W. J. Rieter, H. An, W. Lin, and W. Lin, J. Am. Chem. Soc., 2008, 130, 2154. J. Cheon and J.-H. Lee, Acc. Chem. Res., ACS ASAP. S. A. Bagshaw, E. Prouzet, and T. J. Pinnavaia, Science, 1995, 269, 1242. S. Inagaki, A. Koiwai, N. Suzuki, Y. Fukushima, and K. Kuroda, Bull. Chem. Soc. Jpn., 1996, 69, 1449. S. Inagaki, Y. Fukushima, and K. Kuroda, Chem. Commun., 1993, 680. K. K. Unger, D. Kumar, M. Grun, G. Buchel, S. Ludtke, T. Adam, K. Schumacher, and S. Renker, J. Chromatogr., A, 2000, 892, 47. J. Rejman, V. Oberle, I. S. Zuhorn, and D. Hoekstra, Biochem. J., 2004, 377, 159. I. Slowing, B. G. Trewyn, and V. S. Y. Lin, J. Am. Chem. Soc., 2006, 128, 14792. D.-M. Huang, Y. Hung, B.-S. Ko, S.-C. Hsu, W.-H. Chen, C.-L. Chien, C.-P. Tsai, C.-T. Kuo, J.-C. Kang, C.-S. Yang, C.-Y. Mou, and Y.-C. Chen, FASEB J., 2005, 19, 2014. M.-C. Chao, H.-P. Lin, and C.-Y. Mou, Chem. Lett., 2004, 33, 672. C. E. Fowler, D. Khushalani, B. Lebeau, and S. Mann, Adv. Mater., 2001, 13, 649. K. Suzuki, K. Ikari, and H. Imai, J. Am. Chem. Soc., 2004, 126, 462. F. Li, Z. Wang, and A. Stein, Angew. Chem., Int. Ed., 2007, 46, 1885. Z. Wang and A. Stein, Chem. Mater., 2008, 20, 1029. A. Stein, B. J. Melde, and R. C. Schroden, Adv. Mater., 2000, 12, 1403. S. Huh, J. W. Wiench, B. G. Trewyn, S. Song, M. Pruski, and V. S. Y. Lin, Chem. Commun., 2003, 2364. B. G. Trewyn, C. M. Whitman, and V. S. Y. Lin, Nano Lett., 2004, 4, 2139. S. Che, Z. Liu, T. Ohsuna, K. Sakamoto, O. Terasaki, and T. Tatsumi, Nature, 2004, 429, 281. I. I. Slowing, B. G. Trewyn, and V. S. Y. Lin, J. Am. Chem. Soc., 2007, 129, 8845. J. Liu, Y. Shin, Z. Nie, J. H. Chang, L.-Q. Wang, G. E. Fryxell, W. D. Samuels, and G. J. Exarhos, J. Phys. Chem. A, 2000, 104, 8328. A. Vinu, K. Z. Hossain, and K. Ariga, J. Nanosci. Nanotechnol., 2005, 5, 347. S. Sadasivan, D. Khushalani, and S. Mann, J. Mater. Chem., 2003, 13, 1023. D. R. Radu, C.-Y. Lai, J. Huang, X. Shu, and V. S. Y. Lin, Chem. Commun., 2005, 1264. J. W. Larsen and L. J. Magid, J. Am. Chem. Soc., 1974, 96, 5774. D. R. Radu, C.-Y. Lai, J. W. Wiench, M. Pruski, and V. S. Y. Lin, J. Am. Chem. Soc., 2004, 126, 1640.
29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.
40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.
Mesoporous Silica Nanoparticles 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79.
229
P.-W. Chung, R. Kumar, M. Pruski, and V. S. Y. Lin, Adv. Funct. Mater., 2008, 18, 1390. K. K. Sharma, A. Anan, R. P. Buckley, W. Ouellette, and T. Asefa, J. Am. Chem. Soc., 2008, 130, 218. H.-T. Chen, S. Huh, J. W. Wiench, M. Pruski, and V. S. Y. Lin, J. Am. Chem. Soc., 2005, 127, 13305. S. Huh, H.-T. Chen, J. W. Wiench, M. Pruski, and V. S. Y. Lin, Angew. Chem., Int. Ed., 2005, 44, 1826. V. S. Y. Lin, D. R. Radu, M.-K. Han, W. Deng, S. Kuroki, B. H. Shanks, and M. Pruski, J. Am. Chem. Soc., 2002, 124, 9040. J. M. Notestein and A. Katz, Chem. Eur. J., 2006, 12, 3954. P. M. Pihko, Angew. Chem., Int. Ed., 2004, 43, 2062. K. K. Sharma and T. Asefa, Angew. Chem., Int. Ed., 2007, 46, 2879. V. S. Y. Lin, C.-Y. Lai, J. Huang, S.-A. Song, and S. Xu, J. Am. Chem. Soc., 2001, 123, 11510. J.-S. Chang, B. Chang Ke Liang, D.-F. Hwang, and Z.-L. Kong, Environ. Sci. Technol., 2007, 41, 2064. W. Lin, Y.-W. Huang, X.-D. Zhou, and Y. Ma, Toxicol. Appl. Pharmacol., 2006, 217, 252. S. C. Brown, M. Kamal, N. Nasreen, A. Baumuratov, P. Sharma, V. B. Antony, and B. M. Moudgil, Adv. Powder Technol., 2007, 18, 69. R. Comolli, J. Pathol. Bacteriol., 1967, 93, 241. A. B. Kane, R. P. Stanton, E. G. Raymond, M. E. Dobson, M. E. Knafelc, and J. L. Farber, J. Cell Biol., 1980, 87, 643. Y. Lim, J. H. Kim, K. A. Kim, H. S. Chang, Y. M. Park, B. Y. Ahn, and Y. G. Phee, Toxicol. Lett., 1999, 108, 335. T.-H. Chung, S.-H. Wu, M. Yao, C.-W. Lu, Y.-S. Lin, Y. Hung, C.-Y. Mou, Y.-C. Chen, and D.-M. Huang, Biomaterials, 2007, 28, 2959. J. Lu, M. Liong, J. I. Zink, and F. Tammanoi, Small, 2007, 3, 1341. S. Mayor and R. E. Pagano, Nat. Rev. Mol. Cell Biol., 2007, 8, 603. S. D. Conner and S. L. Schmid, Nature, 2003, 422, 37. J. Lu, M. Liong, S. Sherman, T. Xia, M. Kovochich, A. E. Nel, J. I. Zink, and F. Tamanoi, NanoBiotechnology, 2007, 3, 89. B. G. Trewyn, J. A. Nieweg, Y. Zhao, and V. S. Y. Lin, Chem. Eng. J., 2008, 137, 23. Z. Tao, M. P. Morrow, T. Asefa, K. K. Sharma, C. Duncan, A. Anan, H. S. Penefsky, J. Goodisman, and A.-K. Souid, Nano Lett., 2008, 8, 1517. H. Vallhov, S. Gabrielsson, M. Stromme, A. Scheynius, and A. E. Garcia-Bennett, Nano Lett., 2007, 7, 3576. S. Svenson and Editor, ‘Polymeric Drug Delivery I: Particulate Drug Carriers. (Papers presented as Part of 226th National Meeting of American Chemical Society
230
80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103.
Vivero-Escoto et al. held 7-11 September 2003 in New York, New York.) [In: ACS Symp. Ser.; 2006, 923]’, 2006. R. Haag and F. Kratz, Angew. Chem., Int. Ed., 2006, 45, 1198. R. Hernandez, H.-R. Tseng, J. W. Wong, J. F. Stoddart, and J. I. Zink, J. Am. Chem. Soc., 2004, 126, 3370. T. D. Nguyen, K. C. F. Leung, M. Liong, Y. Liu, F. Stoddart, and J. I. Zink, Adv. Funct. Mater., 2007, 17, 2101. T. D. Nguyen, Y. Liu, S. Saha, K. C. F. Leung, J. F. Stoddart, and J. I. Zink, J. Am. Chem. Soc., 2007, 129, 626. D. Nguyen Thoi, C. F. Leung Ken, M. Liong, D. Pentecost Cari, J. F. Stoddart, and I. Zink Jeffrey, Org. Lett., 2006, 8, 3363. K. C. F. Leung, T. D. Nguyen, J. F. Stoddart, and J. I. Zink, Chem. Mater., 2006, 18, 5919. S. Angelos, Y.-W. Yang, K. Patel, J. F. Stoddart, and J. I. Zink, Angew. Chem., Int. Ed., 2008, 47, 2222. C. Park, K. Oh, S. C. Lee, and C. Kim, Angew. Chem., Int. Ed., 2007, 46, 1455. K. Patel, S. Angelos, W. R. Dichtel, A. Coskun, Y.-W. Yang, J. I. Zink, and J. F. Stoddart, J. Am. Chem. Soc., 2008, 130, 2382. R. Casasus, E. Aznar, M. D. Marcos, R. Martinez-MaZez, F. Sancenon, J. Soto, and P. Amoros, Angew. Chem., Int. Ed., 2006, 45, 6661. R. Casasus, E. Climent, M. D. Marcos, R. Martinez-MaZez, F. Sancenon, J. Soto, P. Amoros, J. Cano, and E. Ruiz, J. Am. Chem. Soc., 2008, 130, 1903. N. K. Mal, M. Fujiwara, and Y. Tanaka, Nature, 2003, 421, 350. T. Tanaka, H. Ogino, and M. Iwamoto, Langmuir, 2007, 23, 11417. S. Angelos, E. Choi, F. Voegtle, L. De Cola, and J. I. Zink, J. Phys. Chem. C, 2007, 111, 6589. J. Lu, E. Choi, F. Tamanoi, and J. I. Zink, Small, 2008, 4, 421. S. P. Hudson, S. White, D. Goradia, H. Essa, B. Liu, L. Qiao, Y. Liu, J. C. Cooney, B. K. Hodnett, and E. Magner, ACS Symp. Ser., 2008, 986, 49. J. Livage, T. Coradin, and C. Roux, J. Phys.: Condens. Matter, 2001, 13, R673. T. Vo-Dinh and Editor, ‘Biomedical Photonics Handbook’, 2003. J. Shi, Y. Zhu, X. Zhang, W. R. G. Baeyens, and A. M. Garcia-Campana, Trends Anal. Chem., 2004, 23, 351. P. D. Beer and P. A. Gale, Angew. Chem., Int. Ed., 2001, 40, 486. W. Verboom and D. N. Reinhoudt, ‘Supramolecular Chemistry of Anions. Edited by A. Bianchi, K. Bowman-James and E. Garcia-Espana’, 1999. R. Weissleder, Science, 2006, 312, 1168. R. Weissleder, H. C. Cheng, A. Bogdanova, and A. Bogdanov, Jr., J. Magn. Reson. Imaging 1997, 7, 258. J. W. M. Bulte, T. Douglas, B. Witwer, S.-C. Zhang, E. Strable, B. K. Lewis, H. Zywicke, B. Miller, P. van Gelderen, B. M. Moskowitz, L. D. Duncan, and J. A. Frank, Nat. Biotechnol., 2001, 19, 1141.
Mesoporous Silica Nanoparticles
231
104. Y. C. Cao, R. Jin, and C. A. Mirkin, Science, 2002, 297, 1536. 105. R. Elghanian, J. J. Storhoff, R. C. Mucic, R. L. Letsinger, and C. A. Mirkin, Science, 1997, 277, 1078. 106. X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li, G. Sundaresan, A. M. Wu, S. S. Gambhir, and S. Weiss, Science, 2005, 307, 538. 107. S. Laurent, D. Forge, M. Port, A. Roch, C. Robic, L. Vander Elst, and R. N. Muller, Chem. Rev., 2008, 108, 2064. 108. I. L. Medintz, H. T. Uyeda, E. R. Goldman, and H. Mattoussi, Nat. Mater., 2005, 4, 435. 109. Y.-W. Jun, J.-W. Seo, and J. Cheon, Acc. Chem. Res., 2008, 41, 179.
CHAPTER 5 NANOSTRUCTURED MESOPOROUS MATERIALS AS DRUG DELIVERY SYSTEMS
Isabel Izquierdo-Barba, Daniel Arcos and Maria Vallet-Regí* Departamento de Química Inorgánica y Bioinorgánica, Facultad de Farmacia, Universidad Complutense de Madrid; Madrid, Spain. Networking Research Center on Bioengineering, Biomaterials and Nanomedicine, CIBER-BBN, Spain *Email:
[email protected]
Recently, research on nanostructured mesoporous materials as drug delivery has experienced an outstanding increase. Since 2001, when MCM-41 was first proposed as drug delivery systems, a wide knowledge, about the relationship between textural and structural properties of these materials and the drug adsorption/release properties has been described. Moreover, the chemical modification by organic functionalization has contributed to precise control in the loading and releasing capacity of such materials. Mesoporous materials are intended for both systemic delivery systems and implantable localdelivery devices. The latter application provides very promising possibilities in the field of bone tissue repair because of the excellent behavior of these materials as bioceramics. This chapter deals with the advances in this field by the control of the textural parameters, surface functionalization, and the synthesis of sophisticated stimuli-response systems.
233
234
Izquierdo-Barba et al.
1. Introduction The development of new active molecules and potential treatments, such as gene therapy, is leading to the evolution of new therapeutic agents but also to the enhancement of the mechanisms to administrate them. In this frame, drug delivery systems (DDS) are called to play a key role for the success of new therapies (1). Basically, a drug delivery system can be described as a formulation that controls the rate and period of drug delivery (i.e. time-release dosage) and target specific areas of the body. Unlike traditional therapies, which show a saw-tooth curve of drug concentration in plasma, DDSs are designed to maintain therapeutic levels during the treatment period (2). Silica mesoporous materials (SMM) have received much attention due to their applications in a wide range of fields, included the biomedical one (3). Since the discovery of the M41S mesoporous materials family by Mobil Corporation scientists in 1992 (4,5), synthesis and applications of mesoporous molecular sieves have been widely developed. Silica-based mesoporous materials are ordered porous structures of SiO2, which exhibit high pore volume, narrow pore size distribution and high surface area. SMMs are synthesized by selfassembly of silica-surfactant composites, in which inorganic species (silica precursors) simultaneously condense giving rise to mesoscopically ordered composites formation (6). After removing the surfactant, a silica based mesostructured solid with the textural properties described above is formed. This last step is commonly carried out through pyrolysis or extraction with the appropriated solvent. Finally the matrices obtained are potential drug carriers featuring: (a) Ordered pore network, which are very homogeneous in size, allowing a high control of the drug load and release kinetics. (b) High pore volume to host the required amount of pharmaceuticals. (c) High surface area, which implies high potential for drug adsorption. (d) Silanol-containing surface that can be functionalized, allowing a better control over the drug loading/release.
235
Nanostructured Mesoporous Materials
Table 1. Biologically active molecules confined into different ordered mesoporus materials by impregnation method, solvent used and maximum loading. Drug
Mesoporous matrix
Solvent
Max. load (%)
Ref
Taxol
FSM
CH2Cl2
6
(7)
Ibuprofen Ibuprofen Ibuprofen Ibuprofen Ibuprofen Ibuprofen Naproxen Amoxicillin Gentamicin Gentamicin Gentamicin Erythromycin Erythromycin Erythromycin Erythromycin Erythromycin Erythromycin Erythromycin Diflunisal Aspirin Aspirin Alendronate Alendronate Alendronate Alendronate BSA BSA BSA BSA L-Trp L-Trp
MCM-4112[a] MCM-4116[b] MCM-4112-NH2[a,c] MCM-4116-NH2[b,c] MCM-48 FDU-5[d] MCM-41-Al[e] SBA-15 SBA-15 PLGA-SiO2[f] PLGA-SiO2[f] MCM-48 FDU-5 [d] FDU-5-C8[d,g] SBA-15 SBA-15-C8[h] SBA-15-C18[i] MCM-41 MCM-41-Al[e] MCM-41 MCM-41-NH2[c] MCM-41 MCM-41-NH2[c] SBA-15 SBA-15-NH2[c] SBA-15 SBA-15-NH2[c] SBA-15-7d[j] SBA-15-7d-NH2[j,c] SBA-15-C3N+Me[k] SBA-15-C3N+Me2C18[l]
Hexane Hexane Hexane Hexane Hexane Hexane Methanol Water Water Water Water Acetonitrile Acetonitrile Acetonitrile Acetonitrile Acetonitrile Acetonitrile Acetonitrile Methanol Water Water Water Water Water Water Water Water Water Water Water Water
23 34 23 33 28.7 20.1 7.3 24 20 22.4 45.6 28 28 12 34 13 18 29 8.7 15 15 14 37 8 22 15.1 10.0 27.0 28.5 4.3 8.2
(8) (8) (9) (9) (10) (10) (11) (12) (13) (14) (14) (10) (10) (10) (15) (15) (15) (15) (11) (16) (16) (17) (17) (17) (17) (18) (18) (18) (18) (19) (19)
[a] C12TAB used as surfactant. [b] C16TAB used as surfactant. [c] Organically modified with amino groups. [d] Large pore 3D bicontinuous cubic Ia-3d mesoporous material. [e] Containing 1.04% (w/w) aluminium. [f] Poly(D,L-lactide-co-glycolide)/mesoporous silica hybrid structure. [g] Organically modified with trimethoxyoctylsilane. [h] Organically modified with trimethoxyoctadecylsilane. [i] Organically modified with carboxylic acid groups. [j] SBA-15 synthesized using seven days of hydrothermal treatment. [k] Organically modified with N-trimethoxysilylpropyl-N,N,N-trimethyl ammonium chloride. [l] Organically modified with octadecyldimethyl (3-trimethoxysilylpropyl) ammonium chloride.
236
Izquierdo-Barba et al.
Table 1 displays some of the most common mesoporous matrixes used to host different drugs. Commonly, the drug is incorporated on the basis of the high capacity of SMM to adsorb and retain molecules inside the pores. The mesoporous material is soaked into the drug solution or suspension for a determined period (commonly between 24-48 hours). The solvent polarity, drug concentration as well as the pH and temperature are carefully chosen. In this way, not only drugs but also numerous proteins and enzymes (cytochrome c, papaine, trypsine, panchreatic lipase, etc.) have been immobilized to be subsequently isolated and purified. Several works have demonstrated that the protein adsorption is conditioned by pH, surface area, pore diameter, temperature and also by the ionic or non-ionic characteristic of the surfactant (20-22). The present chapter reviews the advances of mesoporous materials in the field of biomaterials for DDS applications reached so far. An analysis of the possibilities of these materials is carried out from a critical point of view and future perspectives are also considered. 2. Cytotoxicity, Biocompatibility and Bioactivity of Silica Mesoporous Materials Silica mesoporous materials interact with the physiological environment when performing their functions as drug delivery systems. This is a dynamic two-way process that involves the time-dependent effects of the body on the material and the material on the body. In the case of hard tissue implantation, i.e. when SMM are used as bioceramics, they are also restricted by its long-term biocompatibility and several parameters regarding the site of implantation, the shape and size of material as well as its surface chemistry must be considered. Implantable DDS materials should provide an adequate combination of proper biological response together with the release of necessary drugs against the inflammatory and encapsulation processes. Silica-based systems are among the biomaterials where positive biocompatibility has been observed (23-26) when implanted in bone defects and periodontal sites, mainly those based on SiO2-CaO-NaO-P2O5, SiO2-CaO-P2O5, and
Nanostructured Mesoporous Materials
237
SiO2-CaO systems, obtained without the addition of structure directing agents. However, very few biocompatibility and cytotoxicity studies have been reported on mesoporous SiO2 based materials. Another interesting point is the parameter used to determine the toxicity. Since mesoporous SiO2 is not a water soluble substance, the standard half maximal inhibitory concentration (IC50) is not appropriated. Therefore, quantities as the number of particles required to inhibit normal cell growth by 50% (Q50) could be more appropriated. Di Pasqua et al. (27) have recently reported on the cytotoxicity of MCM-41 with different surface functional groups such as aminopropyl (AP) and mercaptopropyl (MP) groups. Besides, the cytoxicity of nonmesoporous SiO2 nanoparticles was also tested, thus considering the role of the particle size in the cytotoxicity. These authors observed that the cytotoxicity effect decreased in the order MCM-41 > (MP)-MCM-41 > (AP)-MCM-41 ≈ nano SiO2 following the same trend than the magnitude of the surface area in the case of the different MCM-41 derived materials. It is suggested that the cytotoxic effect could be related with the amount of the exposed surface area, so the more efficient functionalization the less toxicity is observed. However, the role of the chemical entities at the surface and the particle size must be also considered. In the case of functionalized and not functionalized MCM-41, the Q50 values obtained were 1.50·1010, 3.60·1010 and 5.76·1010 for MCM-41, (MP)-MCM-41 and (AP)-MCM41, respectively. The cytoxicity of mesoporous silica nanoparticles have been also tested as a function of the surface charge created by means of gradual surface modification with N-trimethoxysilylpropyl-N,N,Ntrimethylammonium chloride (28). The nanoparticles were under 100 nm in diameter and have not shown cytoxicity for concentrations below 100 µg/L, independently of the surface charge. In vivo biocompatibility studies of SiO2 mesoporous materials are also very limited so far. The local effect of SBA-15 has been tested in areas close to brain tissue (temporal lobe) in rats, observing the formation of a well-organized fibrous tissue after 14 days of implantation (29). No pathology, necrosis or acute inflammation was observed after
238
Izquierdo-Barba et al.
this period. Longer term in vivo studies up to four months showed that the biological tissue is self adapted to the material profile, displaying no gliosis or major neuron affection. Recently, Kohane et al. have reported on a systematic biocompatibility study of non-funtionalized SiO2 mesoporous materials, specifically MCM-41, SBA-15 and MCF (30). These authors have tested the effects of both, particles and extracts when injected in subcutaneous, intra-peritoneal and intra-venous locations. The mesoporous silica exhibited a good biocompatibility on histology when implanted subcutaneously. However, intra-peritoneal and intra-venous injections resulted in death for doses over 5 mg in rats. The microscopy analysis of lungs point out that the death could occur due to thrombosis. Anyway, although local tissue reaction seems to be acceptable, they cause severe systemic toxicity. Moreover, the in vitro cytotoxicity was also determined with mesothelial cells, exhibiting a significant degree of toxicity at high concentration. The toxicity appears to result from the particles themselves and not from any contaminants or degradation products and important efforts must be done to modify the SiO2 mesoporous surface to reduce cytotoxicity. These results also suggest that SiO2 mesoporous materials could be better applied in low-irrigated sites, such as subcutaneously or skeletal tissue. However, to best of our knowledge no report has been published on this topic. Positive responses when implanted in bone tissue will provide a clear dimension of the capabilities of these systems for hard tissue applications. In this sense, and concerning permanent implantable DDS intended as bone graft, it should be highlighted that the osteointegration is a fundamental factor to ensure the good performance of this kind of DDSs. During the drug releasing process, both matrix degradation and tissue ingrowth occurs at the implant site. This process goes together with the deposition of calcium and phosphate ions on the material surface to deposit a poorly crystallized apatite phase similar to the inorganic phase of bones (Figure 1) (31-37). This set of reactions modifies the drug release kinetics when compared with other “inert” systems such as PMMA beads. This fact, known as bioactivity, constitutes an added value for a biomaterial respect to the currently used polymer-based devices. All these surface
Nanostructured Mesoporous Materials
239
reactions concerning biocompatibility and bioactivity must be taken into account before their use as DDS for orthopaedic and periodontal applications.
Figure 1. Nanocrystalline apatite (HA) growth onto the surface of a mesoporous matrix under simulated physiological conditions. (A) TEM image obtained from the surface of a ordered mesoporous material after several weeks in simulated body fluid (SBF). (B) Magnification and filtered image of the HA crystal. (C) Fourier pattern showing the hexagonal structure of HA formed.
3. Tailoring Mesoporous Drug Delivery Systems-Textural Properties Considerations The textural properties of mesoporous materials (pore diameter, surface area and pore volume) have been revealed as important factors that govern the drug loading and release, as will be described within the next sections. 3.1. Pore Diameter The adsorption of drugs into mesoporous silica is governed by size selectivity, i.e. the mesopore diameter will determine the size of the guest
240
Izquierdo-Barba et al.
drug. The vast majority of drugs used in clinical practice fall into the nanometer scale and thus they can be easily introduced into the pores of mesoporous matrices (Figure 2). Commonly, pore diameters slightly larger than the drug molecule dimensions (pore diameter/drug size ratio >1) are enough to allow the adsorption of drug inside the pores.
Figure 2. Size of different drugs and biological agents in comparison with mesopore diameter: d ranging between 1.5 and 50 nm size.
One of the most important features of mesoporous matrices is that the mesopore sizes can be tuned from 1.5 nm to several tens of nanometers by changing the chain length of the surfactant, employing polymeric structure-directing agents, or solubilizing auxiliary substances, such as swelling agents, into the micelles (38). For this reason it is expected that mesoporous matrices are able to host from small molecules to macromolecules such as proteins.
Nanostructured Mesoporous Materials
241
Figure 3. (A) Kinetic release constant of ibuprofen as a function of the pore diameter of the mesoporous matrix. (B) Drug load (ibuprofen) capability as a function of surface area of the mesoporous matrix.
Pore diameter has been demonstrated to influence the release rate of molecules as it affects the drug diffusion to the delivery medium. For instance, several MCM-41 mesoporous materials with different pore sizes, ranging from 2.5 nm to 3.6 nm, have been obtained by using cationic surfactants with different length chains (39). Thereafter ibuprofen was confined into the mesopore channels and release profiles into simulated body fluid (SBF) were obtained. The resulting release kinetic parameters, which are summarized in Table 2 and Figure 3A, reveal that the rate of ibuprofen released to the delivery media increases with the pore size in the 2.5–3.6 nm interval. The release profiles from a porous matrix can be predicted as a first approach by the Higuchi model (equation 1) (40): a = kt1/ 2 (1) where a is the amount of drug released after time t and k is the release constant. The relation between the amount of ibuprofen delivered and square root of time resulted linear with regression factors >0.99, pointing to a control of the delivery process by the internal diffusion of ibuprofen thorough the mesopore channels. The values of the slopes k are included in Table 2, showing the increase of this slope with the pore diameter.
242
Izquierdo-Barba et al.
Table 2. Textural properties (pore diameter: Dp and surface area: SBET), ibuprofen loading and release kinetic constant (k) from MCM-41 materials with different pore sizes. Surfactant
Dp (nm)
SBET (m2/g)
IBU loaded (%)
IBU 24h (%)
k (mg/gh1/2)
85%C8TAB-15%C10TAB
2.5
768
11
41
10.0
70%C8TAB-30%C10TAB
2.7
936
19
42
13.7
C12TAB
3.3
1087
23
49
37.0
C16TAB
3.6
1157
34
61
61.0
* C8TAB: octyltrimethylamonium bromide; C10TAB decyltrimethylamonium; C12TAB: dodecyltrimethylamonium bromide; C16TAB: hexadecyltrimethylamonium bromide.
Figure 4. Scheme of the pore size influence on BSA adsorption.
Nanostructured Mesoporous Materials
243
The concept of pore size as key factor in the rate diffusion and release of molecules is not only applicable to 2D-hexagonal structures such as MCM-41, but also to 3D bicontinuous cubic structures such as MCM-48. Thus, the ibuprofen release rate from two mesoporous materials type MCM-48 with different pore diameters (3.6 and 5.7 nm) was studied (10). The obtained results showed a faster ibuprofen rate release for the case of MCM-48 with the largest pore diameter, as it was expected. Pore size is also a limiting factor respect to the amount of drug adsorbed, when the confinement of large molecules, such as proteins and other biologically active molecules, is pursued. This is the case of serum albumins, one of the major components in plasma proteins in humans and the upper mammals. Serum albumins have several physiological functions such as binding, transport, and delivery of fatty acids, porphyrins, bilirubin, steroids, etc. Albumins are also capable of binding a wide variety of drugs that can be later delivered to sites of pharmacological action (41). Therefore, these proteins play an important role on drug delivery, including transport of biologically active molecules to specific places for tissue regeneration technologies. Serum albumin is usually composed of a single-chain of 582 amino acids with an average length of 10 nm and width of 6 nm (42,43). Therefore, the pore size of mesoporous materials seems to be a critical factor for the confinement of this protein and depends on the template and synthesis conditions. Many methods have been reported for controlling the pore size of mesoporous materials (44-46). The most frequently used procedure is the introduction of a swelling agent into the structure directing template. The introduction of these agents has been shown to lead to the swelling of pore diameter, but loss of long-range order of the mesoporous structure is observed. Our research group has recently developed a straightforward way for tailoring the pore size of SBA-15 mesoporous materials and the influence of pore diameter on bovine serum albumin (BSA) loading and release has been evaluated (18). The control over the mesopore size of SBA-15 was performed by increasing the time of the hydrothermal treatment (47,48). Thus, pore diameters ranging from 8.2 nm up to 11.4 nm have been found for SBA-15 materials submitted to hydrothermal treatments for periods of 1 to 7
244
Izquierdo-Barba et al.
days, which seem suitable in size to host BSA. BSA loading experiments revealed that the higher the pore diameter the higher is the amount of BSA loaded in the mesopores, i.e. the protein adsorption is favored when mesopore sizes are enlarged (Figure 4). When loading the protein in a conventional SBA-15 (8.2 nm of diameter size), 15% of BSA is absorbed. The amount of BSA increased to 23%, 25% and 27% for mesoporous matrices exhibiting pore diameters of 9.5 nm, 10.5 nm or 11.4 nm. 3.2. Surface Area The adsorption of drugs into mesoporous materials is a surface phenomenon that is governed by the adsorption properties of mesoporous silica. Therefore, the surface area is expected to be the main factor that determines the amount of drug molecules adsorbed. The synthesis of several MCM-41 mesoporous materials using cationic surfactants with different length chains, as above mentioned, leads to mesoporous matrices with surfaces areas ranging from 768 to 1157 m2/g (Table 2). Figure 3B displays the amount of ibuprofen loaded as a function of the surface area. There is a clear dependence of ibuprofen adsorbed with surface area, i.e. the greater the surface area the higher the amount of drug adsorbed. The influence of surface area on drug adsorption is also demonstrated when evaluating the alendronate adsorption, a potent bisphosphonate able to inhibit bone resorption by osteoclasts, into MCM41 and SBA-15 mesoporous materials (17). Such mesoporous matrices have the same structure (2D-hexagonal and p6mm symmetry) but present different surface area, being 1157 m2/g and 719 m2/g for MCM-41 and SBA-15, respectively. The amount of alendronate adsorbed into MCM-41 and SBA-15 was 14% and 8% respectively, evidencing the dependence of the amount of drug loaded on surface area. 3.3. Pore Volume Drug adsorption is a surface phenomenon governed by attracting interactions between silanol groups in the mesopore walls and the functional groups of the guest molecule, i.e. it is a host-guest or
Nanostructured Mesoporous Materials
245
surface-molecule interaction (49). Therefore it seems reasonable to assume that pore diameter, as the size selectivity limiting factor, together with the surface area are the textural parameters that control drug adsorption. As previously mentioned, drug loading is commonly performed by impregnation methods. Recently, it has been reported that the amount of drug loaded can be increased by performing several successive impregnation methods (50). Following this method intermolecular drug-drug interactions inside the mesopore channels are promoted, as revealed by NMR studies (51). In this case pore volume would determine the total amount of drug adsorbed into the mesopores until complete pore filling. Mesoporous matrices with the greatest pore volumes will lead to the highest amounts of drug adsorbed. 3.4. Increasing the Surface Area - The Hybrid Route The hybrid route, which consists on combining the high and regular porosity of mesoporous materials with the presence of organic groups within the framework, has been also proposed as DDS (52-54). The final compounds are metal-organic frameworks that exhibit outstanding SBET values and that cumulate both high drug loading and a controlled release. These materials are denoted as MIL (Materials Institute Lavoisier). Figure 5 shows the structures of the MIL-100 and MIL-101 compounds. MIL-100 and MIL-101 show surface areas of 3340 and 5510 m2·g-1, respectively, three times and five times larger than SBET values measured for inorganic mesoporous materials. Moreover, MIL-100 is able to adsorb 350 mg·g-1 of IBU and MIL-101 loads 1.4 g·g-1 in the same conditions. These differences were attributed to the pore sizes and structural reasons, particularly the accessible dimensions of the windows of the cages in the solids, which are larger in MIL-101. The behaviour of these materials have been also compared with the silica based MCM-41, with SBET = 1157 m2·g-1. In this sense, MCM-41 and MIL-100 materials showed very similar IBU dosage and kinetics, whereas the drug content of MIL-101 is four times larger than in MCM-41. However, it must be taken into account that these drug adsorption levels are consequence of the high surface area provided by the external micropores, as the IBU molecules are mainly retained outside the zeo-type architecture. This adsorption
246
Izquierdo-Barba et al.
IBU (mg)
mechanism differs from that of silica-based mesoporous materials, which show lower SBET values but possess accessible mesopores for drug storage and release. Comparison of the IBU release profiles are shown in Figure 5(right) for both the MIL materials and MCM-41. Since it is speculated that the larger part of the drug is adsorbed in the outermost micropores of MIL materials and only a small amount is retained in the closed mesopore cavities, the drug release should follow different delivery kinetics from that of MCM-41. As it can be seen, two sections can be observed in the release profiles of MIL materials, corresponding to the release from micropores and mesopores. In MCM-41, as the drug is essentially loaded into the mesopore channels, the release profile only shows one exponential profile. More interesting to observe is that the release from MCM-41 and MIL-100 is very similar although the total surface area is different, whereas for MIL-101, with more open mesopores, is larger. This difference is mainly due to the interaction of IBU molecules with the terephthalic units inside the cavities, which may retain the drug molecules and release them when the micropore windows are clear, i.e. when the drug adsorbed in the outer surface has been already delivered.
Time (days)
Figure 5. Representation of the tetrahedra built up from trimers of chromium octahedra and 1,3,5 BTC (for MIL-100) and 1,4 BDC (for MIL-101) (left). Ibuprofen release as a function of time for MIL-101, MIL-100 and MCM-41 (right).
Nanostructured Mesoporous Materials
247
Although the release profile of ibuprofen from this kind of material could be of interest due to the special nature of MOF matrices, chromium-containing MIL materials are non-biocompatible and new iron-based MOFs within the same structure are currently under development. 4. Surface Functionalization of Mesoporous Drug Delivery Systems A step forward towards the development of mesoporous silicas as DDSs consists in modifying or functionalizing the mesopore silica walls with functional groups. The surface of silica walls is full of silanol groups that can undergo organic modification by covalently linking organic silanes ((RO)3SiR’) (55,56). The drug release can be effectively controlled using different approaches. The first strategy consists in increasing the drug-surface interaction by organically modifying the silica matrix with chemical groups that are able to undergo attracting interactions with the drug molecules through ionic bonds or through ester groups (57). The second strategy for effectively controlling drug release consists in functionalization of mesoporous silica walls with hydrophobic species. The nature of the host-guest chemical interaction can be modified to effectively control drug adsorption and release. Consequently, the choice of the organic modifying group would depend on the targeted drug. The relevance of the chemical nature of the host-guest interaction on the adsorption and release of molecules has been supported by organically modifying MCM-41 mesoporous matrix using several organic groups (chloropropyl, phenyl, benzyl, mercaptopropyl, cyanopropyl and butyl) (58). Ibuprofen was selected as model drug and adsorption and delivery tests were carried out. Different ibuprofen adsorption and delivery were found depending on the organic group modifying the silica walls. Thus, MCM-41 functionalized with polar groups showed greater ibuprofen adsorption than MCM-41 functionalized with non-polar groups. Recently, amino-functionalized MCM-41 and SBA-15 mesoporous materials were proposed as DDSs for alendronate (17). After the loading process the amount of alendronate loaded into amino-modified mesoporous matrices was almost three-fold that of unmodified materials
248
Izquierdo-Barba et al.
(Table 3). This different behavior can be explained by the different chemical interaction between the phosphonate groups in alendronate with silanol groups of unmodified materials and with the amino groups covering the mesoporous silica surface of modified materials. Under the loading conditions (pH 4.8) the three oxygen atoms present in alendronate are deprotonated and they would interact with positively charged amino groups covering the modified silica surfaces. Such interaction is stronger than that of alendronate with the silanol group in the unmodified materials. Therefore, the amount of alendronate adsorbed was 22% and 37% in SBA-15-NH2 and MCM-41-NH2, respectively. These alendronate loads are significantly higher than those of unmodified SBA-15 (8%) and MCM-41 (14%). Table 3. Alendronate loading and release data from MCM-41 and SBA-15 mesoporous matrices before and after functionalization with amino groups. Material
Alendronate loaded (%)
Release after 24h (mg/gSiO2)
Total delivery time (h)
MCM-41 MCM-41-NH2 SBA-15 SBA-15-NH2
13.9 36.6 8.3 22.0
81 103 46 24
72 264* 264 264**
* Incomplete delivery (76%) ** Incomplete delivery (69%)
At physiological pH (pH 7.4), the differences in polarity between the silica surface and alendronate, or between amino-modified surface and alendronate, induce weakening of the adsorbed molecules, which are then slowly delivered to the media. The drug delivery tests indicate that that amino functionalization of mesoporous silica allows a better control on the drug release. In MCM-41-NH2 the amount of alendronate released after 24 hours was 103 mg/g SiO2 (ca. 28 of the total amount loaded) whereas it was 81 mg/g SiO2 (ca. 58% of the total amount loaded) in the unmodified MCM-41 (see Table 3). In the case of SBA-15, the amount of alendronate released after 24 hours was 24 mg/g SiO2 (ca. 11% of the total amount loaded), whereas for unmodified SBA-15, the amount of drug released after 24 hours of assay was 46 mg/g (ca. 55% of the total
249
Nanostructured Mesoporous Materials
amount loaded). In all cases, a noticeable burst effect was observed during the first hours of assay, which could be ascribed to two main factors: the first one involves the release of alendronate adsorbed in the outer surface of the matrix. The second one concerns the gradient produced by the alendronate concentration difference between the delivery medium and the matrix (see Figure 6).
MCM41
1.00
1.00 0.75
0.50
0.50
0.25
0.25
Qt/Qo
0.75
0
0 0
100
200
300
MCM41NH2
1.00
Qt/Qo
SBA15
0
1.00
0.75
0.75
0.50
0.50
0.25
0.25
100
200
300
200
300
SBA15NH2
0
0 0
100
200
Time, h
300
0
100
Time, h
Figure 6. Alendronate release kinetic profiles from MCM-41 and SBA-15 mesoporous materials before and after functionalization with aminopropyl groups.
Moreover, functionalization with amino groups of SBA-15 exhibiting different pore diameters has been demonstrated to result in a higher control of BSA release. The amino groups of SBA-15 modified materials undergo attractive interactions with the carboxylic groups of the protein. It should be noticed that organic functionalization always leads to a decreasing in the mesopore diameter. The BSA molecule is just on the limit of the mesopore dimensions, and thus after aminofunctionalization the amount of BSA loaded decreased compared to unmodified matrices. However, the functionalization with amino group
250
Izquierdo-Barba et al.
has a strong influence on the release kinetic of BSA. As a consequence of the functionalization, the initial burst effect was drastically reduced, and the release of the protein from the mesopores was incomplete from all SBA-15 amino-modified matrices, due to the attracting interaction between the protein and the amino groups grafted to the silica surface (18,59). In other hand, the functionalization of mesoporous matrices using hydrophobic species aims at impeding drug transport out of the matrix because the aqueous delivery solution does not easily penetrate inside the pores. Therefore, SBA-15 mesoporous matrix has been functionalized using hydrophobic octyl (C8) and octadecyl (C18) moieties by treating the mesoporous silica with trimethoxyoctylsilane and trimethoxyoctadecylsilane, respectively (10,15). Adsorption and release tests of erythromycin, a non-polar antibiotic that belongs to the macrolide family, have been carried out. As a result of the organic modification of SBA-15 there was a decrease of the effective pore size and surface area and also there was a decrease of the wettability of the surface by aqueous solutions. The decrease in the surface area leaded to a decrease in the amount of erythromycin loaded. However, this strategy allowed a higher control over the release rate of drug, resulting in a release rate on order of magnitude lower in the C18-SBA-15 sample compared to unmodified SBA-15. Similar results were reported in the literature with mesoporous materials modified by silylation. Ibuprofen was incorporated into the functionalized mesoporous matrix showing a lower drug loading when silylation was carried out (60). 5. Dosage in Mesoporous Materials The dosage, i.e. the actual amount of drugs that can be delivered to the targeted sites, is a very important issue for drug delivery systems. There are several factors that determine the appropriated dosage, including patient characteristics, biodisponibility (in the case of oral delivery), renal elimination, drug-matrix stability, treatment duration, designing of the matrix etc. In the case of oral administration, it is desirable to develop matrixes for once-daily formulations. Under the assumption that all the drug is
Nanostructured Mesoporous Materials
251
released following a zero-order kinetic release, the dosage incorporated should be that indicated for each daily treatment. For example, 40 mg to 500 mg of famotidine can be included into 1.5 g of mesoporous MSU tablets (61). Depending of the dosage per day the therapeutic activity is clearly different, being 40 mg per day of famotidine indicated for the treatment gastric ulcers whereas 500 mg per day is used for the treatment of Zollinger-Ellison syndrome. Therefore, MSU mesoporous materials could be proposed for both treatments, due to its capability for loading and release a wide interval of dosages. Another example is the captopril loading into MCM-41 mesoporous material (62). Captopril is an orally active inhibitor of the angiotensin-converting enzyme and is used for the treatment of hypertension and congestive heart failure. The recommended daily dosage for captopril is from 50 to 100 mg. Taking into account that the loading capability of MCM-41 is of 32% of captopril, then a tablet of only 300 mg of MCM-41 could contain the maximum daily dosage and released after 24 hours in simulated stomach fluid. In the case of bone implants for local drug delivery, the scenario is quite different and new factors take part in the rational design of the DDSs. In these cases, drug biodisponibility is generally much higher compared with oral administration and the drug release must be extended for several days or weeks. Table 4 displays some of the implantable DDS intended as bone grafts. This table collects the actual amounts of drugs (dosages) that can be incorporated into 10 g of mesoporous material. This amount would be appropriated for grafting a bone defect resulted from, for instance, a femur fracture. The dosages per implant are calculated from the maximum load percentages previously reported and displayed in Table 1. Taking into account the daily dosage needed for a 60 kg patient, and the load capacity of mesoporous materials when hosting drugs (dosage/implant), it can be noticed that the reported mesoporous DDSs could release their drug content during several days or even weeks, providing effective doses at the bone tissue. Of course, this assumption only makes sense if the kinetic release is adequate, since the working life of the system not only depends on the drug amount incorporated, but also on the release rate.
252
Izquierdo-Barba et al.
Table 4. Mesoporous matrices, drug dosages and potential delivery time of some mesoporous materials proposed for implantable (bone) drug delivery systems. Mesoporous matrix
Drug
Doses/day
Dosage(2)
SBA-15 SBA15/PLGA SBA-15 SBA-15 SBA-15-NH2 MCM-41-NH2 MCM-41
Gentamicin Gentamicin Erythromycin Amoxicillin Alendronate Alendronate Ibuprofen
150-300 mg 150-300 mg 1.5-3 g(1) 1.5-2 g(1) 5-10 mg(1) 5-10 mg(1) 0.9-1.2 g(1)
2g 4.50 g 3.4 g 2.50 g 2g 2.5 g 7.0 g
(1) Dosages orally administered, taking into account the biodisponibility of the drugs, which are 60, 80, 0.7, and 92% for erythromycin, amoxicillin, alendronate and ibuprofen, respectively. Dosages for gentamicin are those recommended for intra-venous administration. (2) Dosages contained into 10 g of mesoporous silica, which is the approximate amount of silica based material to graft a bone defect in a femur fracture. Smaller periodontal defects usually require between 3-5 g of silica based glass graft.
Figure 7 shows the drug release kinetic profiles for ordered mesoporous materials. Profile “A” is closely related to nonfunctionalized matrices which commonly exhibit a burst effect (kb) followed by a very slow drug release with order-zero kinetic (17). This linear dependence or zero-order kinetic, which is characteristic of pure silica mesoporous matrices exhibiting relatively large mesopore sizes like SBA-15 material, can be described by equation (2):
Qt = k 0t Q0
(2)
where Qt and Q0 are respectively the molecule amount at time t and the initial amount of molecule in the porous matrix and k0 is the zeroorder release constant independent of the molecule concentration in the ordered matrix as well as the solvent accessible area. This kind of profile can be useful in those situations where an immediate high dosage is necessary, for instance acute infections or inflammations. Profile “B” very common in non-functionalized mesoporous matrices such as MCM41 with high surface areas and small mesopores, where the diffusion of
253
Nanostructured Mesoporous Materials
small-drugs to the liquid media can be predicted by first-order kinetic [equation (3)] which contains information about the solvent accessibility and the diffusion coefficient through mesoporous channels.
Qt = 1 − e −k1t Q0
(3)
Profile “C” corresponds to a zero-order kinetic, i.e. to those release processes which are only time-dependent. This kind of profile is highly desirable for long-term drug delivery systems and alendronate/aminofunctionalized SBA-15 system is a clear example of it (17). Finally, profile “D” would correspond to more sophisticated stimuli-responsive system (pH, temperature, magnetic fields, etc.). In these systems the release rate can be controlled by external changes, which opens up an amazing field of new smart DDSs. This kind of matrices will be reviewed in the next section. 1
A
Qt/Q0
Burst effect
0.5
B
Qt = k0t Q0
Qt = kbt Q0
Qt = 1 − e −k1t Q0
Qt/Q0
0
C
Qt = k0t Q0
0.5
D
stimuliresponsive system
0 2
6
10
14
Time (hours)
18
2
6
10
14
18
Time (hours)
Figure 7. Drug release kinetic profiles from ordered mesoporous materials showing the different release kinetics equations.
254
Izquierdo-Barba et al.
6. Mesoporous Materials for Intracellular Targeting 6.1. Cell Mechanism for Particles Internalization Surface functionalized silica nanoparticles can be internalized by animal and vegetal cells (63,64). This fact makes possible the selective delivery of drugs and genes over targeted cell organelles, thus increasing the control release and the therapeutic efficiency. Obviously, mesoporous silica nanoparticles (MSNs) must undergo endocytosis to carry out intracellular drug targeting and three main mechanisms must be considered in the nanoparticles internalization process: - Phagocytosis is probably the most well-known manner in which a cell may import outside materials. This mechanism relays on stretching out pseudopodia and encircling the particles. Once the particle is entrapped, pseudopodia self-fuse resulting into a phagosome that will digest the material within. - Pinocytosis is the mechanism by which a cell is able to ingest droplets of liquid from the extracellular fluid. Pinocytic vesicles tend to be smaller than vesicles produced by other endocytic processes. - Receptor mediated endocytosis is the most specifically-targeted form of the endocytic process. Through receptor mediated endocytosis, active cells are able to take in significant amounts of particular molecules (ligands) that bind to receptor sites extending from the cytoplasmic membrane into the extracellular fluid surrounding the cell. Once freed into the cytoplasm, several small vesicles produced via endocytosis may come together to form a more complex entity named endosome. This last mechanism is the more efficient and allows the internalization of nanoparticles in a selective fashion, by means of the surface functionalization with the appropriated ligands. These ligands can be selectively recognized by a short of cell expressing the corresponding receptors in the cell membrane. Figure 8 represents the endocytosis mechanism undergone by nanoparticles.
255
Nanostructured Mesoporous Materials
Extracellular fluid
Receptor
Endosomal vesicle
Figure 8. Receptor mediated endocytosis of nanoparticles.
Incorporation of functional groups into the mesoporous channels and/or external particle surface of MSNs allows a wide range of manipulation of the surface properties of these materials for controlled release delivery and biosensing applications. If the mechanism mediation is identified, the uptake efficiency can be increased by manipulating the receptor mediated endocytosis. For instance, the uptake of nonfunctionalized MSNs is inhibited by phenylarsine oxide and cytochlasin D in mouse 3T3L1 preadipocytic cells and human mesenchimal stem cells (hMSCs), suggesting that a clathrin- and an actin-dependent endocytosis is involved for these cells (28). Furthermore, a correlation of positive surface charge and the number of fluorescence-labeled cells is observed in both type of cells. Very interesting results are those obtained by Lin and co-workers (65) from HeLa cells. This cancer cells exhibit enhanced uptake efficiency when MSN are functionalized with N-folate-3-aminopropyl
256
Izquierdo-Barba et al.
(FAP). FAP-MSN endocytosis is partially inhibited in presence of folic acid, indicating that the endocytosis of this material is mediated by folic acid receptors on the HeLa cell surface. It is well known that the membranes of human cancer cells are abundant with folate receptors. Therefore, the nanoparticles uptake can be improved by means of surface functionalization with groups for which cancer cells expressing specific receptors or over-expressing non-specific ones. Similarly to 3T3L1 and hMSC cells, HeLa cancer cells better uptake MSNs with positively charged surfaces. The uptake of non-functionalized, negatively charged MSN (-34.73 mV at pH 7.4) by cells has been found to occur through a non-specific adsorptive endocytosis and the resting potentials of cell membranes are normally also negative. By following this strategy, not only nanoparticles uptake can be improved, but also the subsequent release of the nanoparticles from the formed endolysosomal vesicle. 6.2. Microstructural Considerations for SiO2 Nanoparticles Intracellular Targeting Commonly, non-phagocytic eukaryotic cells can internalize particles up to 500 nm in size. Hoekstra et al. (66) reported on a high uptake efficiency of latex particles of 200 nm or smaller. On the contrary, particles larger than 1 µm are hardly internalized. The minimization of particle size to the nanometer range form intracellular delivery is critical in the biological usage of mesoporous silica because most cell uptake occurs in this size range. Several synthetic strategies to control the sizes of mesoporous nanoparticles have been reported (67-69). Small particles with diameters less than 50 nm often have disordered mesostructure. Some works have been success in obtaining well-ordered mesoporous nanoparticles with diameter ranging between 20-50 nm (70), but serious interparticle aggregation can be present; this is a great handicap for their biological application. A suitable strategy to obtain well dispersed mesoporous nanoparticles is separating the nuclei formation and particle growth into two steps in dilute alkaline solution. In this way, non-aggregated and highly ordered nanoparticles of around 110 nm can be obtained. Mesoporous silica nanoparticles with of this size can be internalized into 3T1-L1 fibroblast
Nanostructured Mesoporous Materials
257
cells, accumulated within the cytoplasm and used for fluorescence tag purposes (71). The cellular internalization appears to be generic for various cells with this kind of systems. Both adherent (3T1-L1, MCF7, HeLa, and hMSC) and non-adherent (HSC) cells exhibit nanoparticles internalization.
Figure 9. Different microstructures of SiO2 mesoporous particles. (a) Evaporation induced self-assembly (EISA) method. (b) Aerosol assisted EISA method. (c) Modified Stöber method. (d) Hydrothermal SBA-15 synthesis.
In addition to the nature of the pore system, size, shape and connectivity of mesoporous materials, and depending on application, the morphology of the mesophase may be particularly important. Actually, the synthesis of mesoporous silica-based functional materials for practical applications in biotechnology and biomedicine requires the ability of controlling the particle morphology of these materials. Simple morphologies with short, unhindered path lengths such as small spheres and crystal-like particles as well as short, straight rods are beneficial for applications limited by intra-particle diffusion processes such as catalysis, separation, guest molecule encapsulation and internal surface modification. Thus, not surprisingly, extensive work has been devoted to the morphological control of mesoporous silica (72-76) and organosilicates (77) (Figure 9). Most approaches are based on changes in
258
Izquierdo-Barba et al.
synthesis conditions, including the silica source, the nature of the surfactants, co-surfactants, co-solvents and additives, and the overall composition of the synthesis mixture. Figure 9 shows different particle morphologies obtained through several synthesis routes and conditions. Regular morphologies provide several advantages for drug adsorption and delivery applications. In fact, ordered mesoporous materials with irregular bulk morphology exhibit sustained-release properties, but their drug storage capacity is relatively low and the drug delivery is unbalanced. Moreover, the random aggregation of the amorphous and polydisperse particles of mesoporous silicas in aqueous solutions with high ionic strength complicates their circulation lifetime and cell membrane permeability. Consequently, it is difficult to predict and regulate the biocompatibility of amorphous mesoporous silica materials both in vitro and in vivo. Endocytosis efficiency is also influenced by the particle morphology. The internalization of mesoporous silica particles with spherical and tube-like morphologies have been tested in the presence of Chinese hamster ovarian (CHO) cancer cells and human fibroblast cell lines (78). The rates of endocytosis form both MSNs were similar and rapid for CHO cells. However, the rates of endocytosis for fibroblast cells were different for the MSN with different morphologies. Specifically, the rate of endocytosis for the spherical ones was significantly faster than that of those tubular. This difference in the endocytosis kinetics may be attributed to two variables. One is the particle size, which is similar in width but smaller in length in the case of spherical particles. The other variable is the different aggregation ability between the MSN nanoparticles with different shapes. In this case, tubular particles aggregates into larger particles compared with the spherical ones. Another elegant strategy is to synthesize hollow mesoporous silica (HMS) spheres on the nanoscale with pore channels penetrating from the outside to the inner hollow core (79,80). Such materials can be also synthesized in the presence of organic molecules during the templating stage to produce ordered mesoporous nanospheres with worm-like pores that can be employed as drug-delivery systems (81), although the most important procedure to take into account is the control of the outer sphere morphology. In this sense, in order to direct the hollow structure of the
Nanostructured Mesoporous Materials
259
nanospheres, organic polymers such as polystyrene and polymethyl methacrylate (82) or poly(vinyl pyrrolidone) have been employed (83). Thus-obtained hollow nanospheres generally show larger storage capacity than that observed in irregular bulk particles with similar mesoporous structures. Not only hollow nanospheres have been synthesized but also bulk spherical mesoporous particles can be obtained by fluorocarbon surfactant templating. The weakly acidic conditions needed for these materials promote a slow hydrolysis of silica precursors and the hydrolyzed silica species co-assemble with triblock co-polymer templates to yield well-defined mesophases. The structures and poresizes of such templated mesoporous nanospheres depend on the type of copolymer and the amount of organic additives. Simultaneously, fluorocarbon surfactants surround the silica nanoparticles through S+X−I+ interactions, thereby limiting the growth of mesoporous silica nanospheres (84). Moreover, the control of drug adsorption and release has been tested even with pure silicon mesoporous particles (85) obtained by anodization in HF solutions, although the absence of pore ordering reduces both the molecular selectivity and therefore the drug delivery performance 7. Stimuli-Responsive Mesoporous Materials Together with the development of mesoporous silica as drug delivery systems, the interest in controlling the release of such drug molecules have been growing. In conventional mesoporous systems, as it has been described, the release of adsorbed molecules usually follows a sustained kinetic mechanism that can be expressed in terms of diffusion of adsorbed molecules throughout the mesopore channels in the silica matrix. Release kinetics, therefore, can be interpreted by the Fickian diffusion coefficients dependent on both the molecule and silica matrix characteristics. However, for certain applications in both chemistry and medicine, it is needed to modulate the delivery of adsorbed molecules by responding to environmental stimuli such as pH and temperature changes or even photosensitive modifications. In addition, many important siteselective delivery systems, such as those for highly toxic anti-tumour
260
Izquierdo-Barba et al.
drugs, require “zero-release” before reaching targeted cells or tissues. Actually, one of the main and more specific problems of DDSs at the present is the loss of activity of several drugs before reaching the target tissue, due to the premature active principle degradation. In this sense, stimuli-responsive systems that exhibit “zero premature” release could play a fundamental role in enabling this task. Other specific long-term situations could require to increase or to slow down the drug release, depending on the disease evolution. For this purpose, implantable systems able to respond to external stimuli (magnetic fields, for instance) or internal pH changes would be also very interesting. The current research is focused on the design of ordered mesoporous materials with certain functional groups that respond to environmental changes for modifying the adsorption and release characteristics of the conventional systems. In the latest years, several ordered mesoporous materials have been developed with this stimuli-responsive ability. The following sections deal with these systems attending to the corresponding sort of stimuli. 7.1. Drug Release Mediated by Chemical Stimuli - Disulfide-Reduction Based Gating Groundbreaking systems currently under investigation are those based on capping mesopores with nanoparticles, such as CdS, where the linkage between silica and ceramic particles is verified through disulfide bonds with thiol-functionalized silica (86). Prof. Lin’s research team has developed this gated MSN system by means of an elegant strategy that can be summarized as follows:
(a) Preparation of mecaptopropyl-derivatized mesoporous silica nanosphere material through a co-condensation method (87), and subsequent surfactant removal. (b) Treatment with 2-(pyridyl-disulphanyl) ethylamine to prepare a MSN material containing a chemically labile disulfide bond. (c) Incorporation of the guest molecule into the mesopores by an impregnation method.
Nanostructured Mesoporous Materials
261
(d) Mesopores blocking with mercaptoacetic acid coated CdS nanocrystals (2.0 nm size). These nanocrystals are covalently bonded through the formation of an amide bond as depicted in Figure 10.
Figure 10. Representation of the disulfide-link based CdS MSN controlled release system.
Moreover, pore capping can be performed by using polyamidoamine (PAMAM) dendrimers (88). PAMAM caps can serve as non-viral gene transfection reagents where the plasmid DNA of an enhanced green fluorescence protein (Aequorea victoria) crosses cell membranes through endocytosis to release the plasmid. The released part is sent to the nucleus to produce green fluorescent proteins. Both systems (CdS and PAMAM capping) do not exhibit premature release in phosphate buffer saline solution (pH 7.4) over a period of 12 h. The addition of disulfidereducing agents, such as dithiothretol (DTT) and mercaptoethanol, results in a fast release of the guest molecule in the case of CdS caps, or to a more sustained release in the case of PAMAM capping. In this way,
262
Izquierdo-Barba et al.
the drug is released through a cascade of events comprising the cleavage of bisulfide bond, uncapping of the CdS nanocrystals and drug diffusion. - pH Modification Based Responsive Systems Xiao and co-workers have designed pH-responsive carriers synthesised by oppositely charged ionic interaction between carboxylic acid modified mesoporous silica and a polyelectrolyte solution (89). In these materials, as it is shown in Figure 11, polycations grafted to anionic SBA-15 silica by oppositely charged interaction are acting as closed gates for drug storage into mesopores. When ionized carboxylic groups (COO−) are changed to protonated groups (COOH) due to the surrounding pH, polycations are detached from modified silica surface inducing the drug release from mesopores.
Figure 11. Schematic representation of pH-responsive storage-release drug delivery system. This pH-controlled system is based on the interaction between negative carboxylic acid modified SBA-15 silica rods with polycations (PDDA).
- Redox-Based Gating Excellent work has been also carried out about pore modification with large organic molecules following a supramolecular route employing for instance rotaxane molecules as reported by Prof. Zink’s group (90,91). These systems act as nanovalves operated by a selective redox process that modifies the chemical configuration of the linked rotaxane molecules “opening” and “closing” the mesopore access. Modification of mesopore entrance can even be performed by using other
Nanostructured Mesoporous Materials
263
types of organic functionalities to develop supramolecular mesoporous materials that react to several chemical signals (92). 7.2. Drug Release Mediated by Thermal Stimuli Thermoresponsive mesoporous materials are essentially based on the development of hybrid systems that integrate silica inorganic phase with thermally active polymers such as poly N-isopropylacryl amide (PNIPAm), to produce sponge-like phases (93). Sponge phases are formed by self assembly of amphiphilic templates during the formation of mesoporous inorganic materials, and it consists of a three-dimensional random packing of a multiple connected bilayer of the surfactant and cosurfactant that divides the space in two subspaces filled with solvent, similar to the liposome structure. Pore size and distance between adjacent silicate layers in the porous structure were controlled by changing the hydrophilic domain, for example, by varying the amount of water. Hence, drugs can be loaded in the sponge-like mesoporous domains that serve as reservoirs and the thermal cyclic polymer shrinkage and aperture controls the drug release. 7.3. Drug Release Mediated by Photo-Chemical Stimuli Light of the suitable wavelength can be also employed for triggering the drug release or even molecular recognition systems based in modified mesoporous silica materials. The size-sieving efficiency of the mesoporous silica matrix is combined with the possibility of modifying with the appropriate photo-sensitive organic groups on the pore surface. Therefore, small molecules with certain groups diffuse through mesopores and react with the modified pore walls as described by Lin and co-workers (75). Nevertheless, using this system, molecules are only adsorbed onto pore walls and retained there as a function of their fluorescence characteristics regardless of their release from the matrix. For direct drug release applications, Fujiwara and co-workers have been researching on a photo-controlled release system based on the pore entrance modification with coumarin groups. Such molecules undergo a reversible dimerization by irradiation with UV light at wavelength larger
264
Izquierdo-Barba et al.
than 310 nm, returning to the monomer form by subsequent irradiation at lower wavelengths, about 250 nm. The dimer form of the coumarin, when grafted on the surface of mesoporous silica systems, such as MCM-41, reduces the effective pore size of the matrix and subsequently hinders the adsorption and release of encapsulated molecules into the pore voids. The adequate irradiation of the material virtually “opens” the pore gates and the adsorbed drugs can be released (94,95). 7.4. Drug Release Mediated by Magnetic Stimuli Ordered mesoporous drug-delivery systems that respond to external magnetic fields have been developed. These magnetic-responsive materials can be obtained by direct encapsulation of the magnetic nanoparticles into the mesoporous silica. Such approach yields materials that can be employed as magnetic nanovectors, usually based on magnetite or even iron, covered with mesoporous silica (96). Other very interesting systems currently under investigation are those based on capping mesopores with magnetic nanoparticles that can be alternatively placed and removed. For these systems, mesoporous materials are synthesised by co-condensation method with mercaptopropyl silanes and the so-obtained thiol-functionalized silica is linked through –SH groups with 2-carboxyethyl-2-pyridyl disulfide to yield acid functionalized mesoporous silica. Mesopore entrance of MCM-41 materials have then been closed with magnetic Fe3O4 nanoparticles by placing the acid functionalized silica in a suspension together with magnetite nanoparticles and the incorporated molecule to test drug delivery. Capped materials are then submitted to adequate magnetic fields to remove the magnetite nanoparticles and therefore release the adsorbed drug. The controlled release mechanism of the whole system is due to the reduction of the disulfide linkage between magnetic Fe3O4 nanoparticles and the thiol-functionalized silica mesoporous material by reducing agents such as dihydrolipoic acid or dithiothreitol (97). Mesoporous materials in the form of microspheres with a magnetic component represent a significant advance in the field of drug delivery. The outstanding textural properties of mesoporous materials allow a great load of drug and a controlled release. On the other, the magnetic
265
Nanostructured Mesoporous Materials
properties allow not only the targeting or accumulation of drug in a desired place of the body, but also the possibility of using hyperthermia combined with drugs for the treatment of cancerous diseases. Our research group has developed an aerosol-assisted route to synthesize mesoporous silica microspheres encapsulating magnetic nanoparticles (98). The strategy followed can be summarized in three synthesis steps:
Linker-MSN
Amidation Fe3O4-MSN
Linker cleavage
Fluorescein
Figure 12. Schematic of the stimuli-responsive delivery system (magnet-MSN) based on mesoporous silica nanorods capped with superparamagnetic iron oxide nanoparticles. The controlled-release mechanism of the system is based on reduction of the disulfide linkage between the Fe3O4 nanoparticle caps and the linker-MSN hosts by reducing agents such as DHLA.
(a) Preparation of the ferrofluid. The co-precipitation of Fe(II) and Fe(III) chlorides with ammonium hydroxide at alkaline pH is carried out yielding nanometric magnetite (Fe3O4). The particles are subsequently oxidized to maghemite and finally dispersed in water. The ferrofluid so-obtained is composed of magnetic nanoparticles with an average diameter of 8 nm. (b) Synthesis of the precursor solution. A certain amount of Pluronic P123 is dissolved in a suspension of the ferrofluid in ethanol. The pH of the mixture is adjusted to acid conditions (pH 1.2) before silica precursor was added to the solution.
266
Izquierdo-Barba et al.
(c) Fabrication of the spheres. MMS were synthesized by pyrolisis of an aerosol generated by ultra high frequency spraying of the solution. A piezoelectric ceramic that allows the carrier gas flow to be freely adjusted is located at the bottom of a vessel, which contains the precursor solution of the material. When the piezoelectric transducer is excited near its own resonance frequency, a geyser is formed at the surface of the liquid. This geyser produces ultrafine droplets, which form an aerosol. N2 gas is used as the carrier to convey the aerosol to the pyrolisis zone. The residence time of the particles in the high temperature zone is controlled by the gas flow. The system is designed in such a way that the mesostructure is mainly formed during the droplet drying at the pre-heating site. The temperature at this site is around 100ºC (drying zone) and the furnace is set at 400ºC in order to avoid the particles coalescence on the collection surface. Finally, the powder obtained by the aerosol-assisted method is calcined to remove the surfactant. The final products are silica mesoporous spheres containing magnetic nanoparticles that provide superparamagnetic behaviour to the mesoporous material (Figure 13). The drug incorporation was carried out
Figure 13. SiO2 mesoporous magnetic spheres. (A) TEM image of a γ-Fe2O3 nanoparticle hosted within. (B) TEM images of lamellar structured SiO2 magnetic spheres. (C and D) Functional properties: magnetic response to an external continous field and drug release capability.
Nanostructured Mesoporous Materials
267
by soaking the matrix into a highly concentrated drug solution and subsequent drying. These materials can load a high amount of drug inside the mesoporous network and they can perform controlled delivery because of the confinement of the drug molecules in the ordered mesostructure. Drug release from MMS shows a first faster release, probably due to drug molecules associated to the outer surface. A slower release is observed afterwards. This kind of kinetic is useful for those clinical cases that require a first high dose followed by a more stable dosage. 8. Conclusions and Outlook The research efforts aimed to fabricate more efficient drug delivery systems is one of the most important objectives of biomedical and pharmaceutical sciences. Currently, pharmaceutical companies and device manufactures are actively seeking development opportunities for new drug/device combination products based on their existing drug and device products. However, in the next years we will be witnessing the expansion of biological therapy, which will involve the development of numerous active therapeutic agents (monoclonal antibodies, peptide and protein-based drugs, gene therapy, etc.). Currently, the development of biological therapies based on biotechnology treated proteins has resulted in more than 60 approved drugs by the FDA in the last 20 years, and it is estimated that around 360 candidates to treat more than 200 illnesses are currently under clinical test. General pharmacokinetic/pharmacodynamic principles are just as applicable to biotech agents as they are to traditional small molecule drugs. However, their macromolecular nature and the fact that most biotech drugs are identical or similar to endogenous molecules, lead to new pharmacokinetics-related problems that subsequently result in bioavailability problems (99). These problems are related with the matrix biodegradation and the premature drug release becomes a serious drawback in the case of highly toxic drugs, such as anti-tumorals. During the last years, there has been an increasing number of research groups involved in the synthesis of mesoporous materials as
268
Izquierdo-Barba et al.
drug delivery systems. Nowadays, this field is one of the most evident examples of transversal knowledge among materials science and biomedical applications, and its development offers promising possibilities for better medical treatments. In this chapter we have collected the advances in the field of silica mesoporous materials for drug delivery. Since 2001, our research team gets involved with obtaining a wide knowledge about the relationship between textural properties of SMMs and the drug adsorption/release properties. Many efforts have been aimed to design different pore structures. Nowadays, big macromolecules can be entrapped into cavities suitable for their adsorption and release. The development of cage like structures or plumber’s nightmare pore structures will play an important role in this topic. A second porosity related parameter that should be addressed for drug adsorption and delivery is the pore inlet. Some studies have shown that the immobilisation of enzymes is facilitated in materials with cage like structures and large entrance sites (100). The incorporation of functional chemical groups at the surface of the silica matrix meant a great step toward the “controlled” release of drugs from these systems. A crucial advance can be found with the development of the smart release (stimuli-response) systems. Considering these materials for highly toxic drug delivery systems, mesoporous materials exhibiting “zero premature release” seem to be an excellent alternative to conventional systems. It is clear that to control the drug kinetic release through external stimuli opens a wide field of possibilities in long-term therapies. Finally, the development of mesoporous nanoparticles has allowed the fabrication of drug delivery systems for intracellular targeting, reaching specificity levels not reached before. Biocompatibility is a difficult question to tackle when considering SMMs for drug delivery. Certainly, in vitro citotoxicity studies indicate a positive cell-material interaction, even when mesoporous nanoparticles are internalised into the cytoplasm. However, in vivo biocompatibility tests are very few and some of them indicate that much caution should be paid, especially when peritoneal or intravenous delivery is considered. Due to the chemical and textural properties of silica mesoporous
Nanostructured Mesoporous Materials
269
materials, they seem to be an excellent alternative as drug delivery bioceramics for orthopaedic applications. The bioceramic character of these materials allows repairing bone defects with the added value of releasing adequate drugs helping the bone-repairing process. These delivery systems, therefore, increase the biodisponibility of drugs in the bone tissue compared to oral formulations and come to fill a gap where conventional polymeric systems cannot be used. Acknowledgments The authors would like to thank all members of our research team who have contributed with the results described in this work, and whose names are collected in the reference sections. We also thank the Spanish National Science and Technology Commission (grant MAT-2005-01486) and Autonomous Government of Madrid (grant S-0505/MAT/0324) for financial support. References 1. 2. 3. 4. 5.
6. 7. 8. 9. 10.
R. Langer. Science, 293, 58 (2001). M. Vallet-Regí. Chem. Eur. J., 12, 5934 (2006). M. Vallet-Regí, F. Balas, D. Arcos. Angew. Chem. Int. Ed., 46, 7548 (2007). C.T. Kresge, M.E. Loenowicz, W.J. Roth, J.C. Vartuli, J.S. Beck. Nature, 359, 710 (1992). J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Loenowicz, C.T. Kresge, K.D. Schmitt, C.T.W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker. J. Am. Chem. Soc., 114, 10834 (1992). A. Firouzi, F. Atef, A.G. Oaertly, G.D. Stucky, B.F. Chmelka. J. Am. Chem. Soc., 119, 3596 (1997). H. Hata, S. Saeki, T. Kimura, Y. Sugahara, K. Kuroda. Chem. Mater., 11, 1110 (1999). M. Vallet-Regí, A. Rámila, R.P. del Real, J. Pérez-Pariente. Chem. Mater., 13, 308 (2001). B. Muñoz, A. Rámila, J. Pérez-Pariente, I. Diaz and M. Vallet-Regi. Chem. Mater., 15, 500 (2003). I. Izquierdo-Barba, A. Martínez, A.L. Doadrio, J. Pérez-Pariente, M. Vallet-Regi. Eur. J. Pharm. Sci., 26, 365 (2005).
270
Izquierdo-Barba et al.
11.
G. Cavallaro, P. Pierro, F.S. Palumbo, F. Testa, L. Pasqua, R. Aiello. Drug Delivery, 11, 41 (2004). M. Vallet-Regí, J.C. Doadrio, A.L. Doadrio, I. Izquierdo-Barba, J. Pérez-Pariente. Solid State Ionics, 172, 435 (2004). A.L. Doadrio, E.M.B. Sousa, J.C. Doadrio, J. Pérez-Pariente, I. Izquierdo-Barba, M. Vallet-Regi. J. Control Release, 97, 125 (2004). J.M. Xue, M. Shi. J. Control Release, 98, 209 (2004). J.C. Doadrio, E.M.B. Sousa, I. Izquierdo-Barba, A.L. Doadrio, J. Pérez-Pariente, M. Vallet-Regí. J. Mater. Chem., 16, 462 (2006). W. Zeng, X.F. Quian, Y.B. Zhang, J. Yin, Z.K. Zhu. Mater. Res. Bull. 40, 766 (2005). F. Balas, M. Manzano, P. Horcajada, M. Vallet-Regí. J. Am. Chem. Soc., 128, 8116 (2006). M. Vallet-Regí, F. Balas, M. Colilla, M. Manzano. Prog Solid State Chem 2007: published online 30 October 2007, doi:10.1016/j.progsolidstchem.2007.10.002. F. Balas, M. Manzano, M. Colilla, M. Vallet-Regí. Acta Biomater 2008: published online 15 December 2007, doi:10.1016/j.actbio.2007.11.09. A. Katiyar, L. Ji, P.G. Smirniotis, N.G. Pinto. Micropor. Mesopor. Mat., 80, 311 (2005). J.F. Díaz, JrKJ Balkus. J. Mol. Catal. B: Enz., 2, 115 (1996). H. Takahashi, B. Li, T. Sasaki, C. Miyaaki, T. Kajino, S. Inagaki. Micropor. Mesopor. Mat., 44-45, 755 (2001). P. Kortesuo, M. Ahola, S. Karlsson, I. Kangasniemi, A. Yli-Urpo, J. Kiesvaara. Biomaterials, 21, 193 (2000). M.C. Serrano, M.T. Portolés, R. Pagani, J. Sáez de Guinoa, E. Ruiz-Hernández, D. Arcos and M. Vallet-Regí. Tissue Engineering, 14, 617 (2008). L. Meseguer-Olmo, M.J. Ros-Nicolás, M. Clavel-Sainz, V. Vicente-Ortega, M. Alcaraz-Baños, A. Lax-Pérez, D. Arcos, C.V. Ragel, M. Vallet-Regí. J. Biomed. Mater. Res., 61, 458 (2002). N. Olmo, A.I. Martin, A.J. Salinas, J. Turnay, M. Vallet-Regí, M.A. Lizarbe. Biomaterials, 24, 3383 (2003). A.J. Di Pasqua, K.K. Sharma, Y.L. Shi, B.B. Toms, W. Ouellette, J.C. Dabrowiak, T. Asefa. J. Inorg. Biochem., 102, 1416 (2008). T.H. Chung, S.H. Wu, M. Yao, C.W. Lu, Y.S. Lin, Y. Hung, C.Y. Mou, Y.C. Chen, D.M. Huang. Biomaterials, 28, 2959 (2007). A. Díaz, T. López, J. Manjares, E. Basaldella, J.M. Martínez-Blanes, J.A. Odriozola. Acta Biomaterialia, 2, 173 (2006). S.P. Hudson, R.F. Padera, R. Langer, D.S. Kohane. Biomaterials, 29, 4045 (2008). M. Vallet-Regí, L. Ruiz-González, I. Izquierdo-Barba, J.M. González-Calbet. J. Mater. Chem., 16, 26 (2006). I. Izquierdo-Barba, L. Ruiz-González, J.C. Doadrio, J.M. González-Calbet and M. Vallet-Regí. Solid State Sci., 7, 983 (2005).
12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.
26. 27. 28. 29. 30. 31. 32.
Nanostructured Mesoporous Materials 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.
50. 51. 52. 53. 54. 55. 56.
271
M. Vallet-Regí, I. Izquierdo-Barba, A. Rámila, J. Pérez-Pariente, F. Babonneau and J.M. González-Calbet. Solid State Sci., 7, 233 (2005). P. Horcajada, A. Rámila, K. Boulahya, J. González-Calbet and M. Vallet-Regí. Solid State Sci., 6, 1295 (2004). X. Yan, C. Yu, X. Zhou, J. Tang and D. Zhao. Angew. Chem. Int. Ed., 43, 5980 (2004). A. López-Noriega, D. Arcos, I. Izquierdo-Barba, Y. Sakamoto, O. Terasaki and M. Vallet-Regí. Chem. Mater., 18, 3137 (2006). I. Izquierdo-Barba, D. Arcos, Y. Sakamoto, O. Terasaki, A. Lopez-Noriega and M. Vallet-Regí, Chem. Mater., 20, 3191 (2008). J. Fan, C. Yu, F. Gao, J. Lei, B. Tian, L. Wang, Q. Luo, B. Tu, W. Zhou, D. Zhao. Angew. Chem. Int. Ed., 42, 3146 (2003). P. Horcajada, A. Rámila, J. Pérez-Pariente, M. Vallet-Regí. Micropor. Mesopor. Mater., 68, 105 (2004). T. Higuchi. J. Pharm. Sci., 52, 1145 (1963). D.C. Carter, J.X. Ho. Adv. Protein Chem., 45, 153 (1994). T. Peters. All about Albumin: Biochemistry, Genetics and Medical Applications, Academia Press, San Diego, 1995. S. Sugio, A. Kashima, S. Mochizuki, et al. Protein Eng., 12, 439 (1999). E. Prouzet, T.J. Pinnavaia. Angew. Chem. Int. Ed., 55, 559 (1997). M. Kruk, M. Jaroniec, A. Sayari. Micropor Mesopor Mater., 35, 545 (2000). G.D. Stucky, B.F. Chmelka, D. Zhao, et al. Block polymer processing for mesostructured inorganic oxide materials. WO037705 (1999). M. Kruk, M. Jaroniec, A. Sayari. J. Phys. Chem. B, 103, 4590 (1999). S.A. El-Safty, T. Hanaoka, F. Mizukami. Chem. Mater., 17, 3137 (2005). G. Buntkowsky, H. Breitzke, A. Adamczyk, F. Roelofs, T. Emmler, E. Gedat, B. Grueberg, Y. Xu, H.H. Limbach, J. Shenderovich, A. Vyalikh, G. Findenegg. Phys. Chem. Chem. Phys., 9, 4843 (2007). C. Charnay, S. Begu, L. Tourné-Péteilh, L. Nicole, D.A. Lerner, J.M. Devoisselle. Eur. J. Pharm. Biopharm., 57, 533 (2004). T. Azaïs, C. Tourné-Péteilh, F. Aussenac, N. Baccile, C. Coelho, J.M. Devoisselle, F. Babonneau. Chem. Mater., 18, 6382 (2006). G. Ferey, C. Mellot-Draznieks, C. Serre, F. Millange, J. Dutour, S. Surble, I. Margiolaki. Science, 309, 2040 (2005). C. Serre, F. Millang, C. Thouvenot, M. Nogues, G. Marsolier, D. Louer, G. Ferey. J. Am. Chem. Soc., 124, 13519 (2002). P. Horcajada, C. Serre, M. Vallet-Regí, M. Sebban, F. Taulelle, G. Férey. Angew. Chem. Int. Ed., 45, 5974 (2006). Exxonmobil Research and Engineering Company. Functionalized periodic mesoporous materials, their synthesis and use. US0072674 (2004). F. Hoffmann, M. Cornelius, J. Morell, et al. Angew. Chem. Int. Ed., 45, 3216 (2006).
272
Izquierdo-Barba et al.
57. 58. 59. 60. 61. 62. 63.
C. Tourné-Péteilh, D. Brunel, S. Bégu, et al. New J. Chem., 27, 1415 (2003). P. Horcajada, A. Rámila, G. Férey, M. Vallet-Regí. Solid State Sci., 8, 1243 (2006). S.W. Song, K. Hidajat, S. Kawi. Langmuir, 21, 9568 (2005). Q. Tang, Y. Xu, D. Wu, Y. Sun. Chem. Lett., 35, 474 (2006). Q. Tang, Y. Xu, D. Wu, Y. Sun. J. Solid State Chem., 179, 1513 (2006). F. Qu, G. Zhu, S. Huang, S. Li, S. Qiu. ChemPhysChem., 7, 400 (2006). W. Sun, N. Fang, B.G. Trewyn, M. Tsunoda, I.I. Slowing, V.S. Lin, E.S. Yeung. Anal. Bioanal. Chem., 391, 2119 (2008). F. Torney, B.G. Trewyn, V.S. Lin, K. Wang. Nature Nanotech., 2, 295 (2007). I. Slowing, B.G. Trewyn, V.S.Y. Lin. J. Amer. Chem. Soc., 128, 14792 (2006). J. Rejman, V. Oberle, I.S. Zurrón, D. Hoekstra. Biochem. J., 377, 159 (2004). Y. Lu, H. Fan, A. Stump, T.L. Ward, T. Rieker, C.J. Brinker. Nature, 398, 223 (1999). C.E. Fowler, D. Khushalani, B. Lebeau. S. Mann. Adv. Mater., 13, 649 (2001). Q. Cai, Z.-S Luo, W.-Q Pang, Y.-W Fan, X.-H Chen, F.-Z. Cui. Chem. Mater., 13, 258 (2001). K. Suzuki, K. Ikari, H. Imai. J. Am. Chem. Soc., 126, 462 (2004). Y.S. Lin, C.P. Tsai, H.Y. Huang, C.T. Kuo, Y. Hung, D.M. Huang, Y.C. Chen, C.Y. Mou. Chem. Mater., 17, 4570 (2005). J. Fan, J. Lei, L. Wang, C. Yu, B. Tu, D. Zhao. Chem. Commun., 2140 (2003). D. Zhao, J. Sun, Q. Li, G.D. Stucky. Chem. Mater., 12, 275 (2000). H. Yang, N. Coombs, G.A. Ozin. Nature, 386, 692 (1997). C. Yu, J. Fan, B. Tian, D. Zhao, G.D. Stucky. Adv. Mater., 14, 1742 (2002). A. Sayari, B.-H. Han, Y. Yang. J. Am. Chem. Soc., 126, 14348 (2004). M.P. Kapoor, S. Inagaki. Chem. Mater., 14, 3509 (2002). B.G. Trewyn, J.A. Nieweg, Y. Zhao, V.S.-Y. Lin. Chem. Eng. J., 137, 23 (2008). Y. Zhu, J. Shi, W. Shen, X. Dong, J. Feng, M. Ruan, Y. Li. Angew. Chem. Int. Ed., 44, 5083 (2005). Y. Li, J. Shi, H. Chen, Z. Hua, L. Zhang, M. Ruan, J. Yan, D. Yan. Microporous Mesoporous Mater., 60, 51 (2003). N.E. Botterhuis, Q. Sun, P.C.M.M. Magusin, R.A. van Santen, N.A.J.M. Sommerdijk. Chem. Eur. J., 12, 1448 (2006). J.-F. Chen, H.-M. Ding, J.-X. Wang, L. Shao. Biomaterials, 25, 723 (2004). Y. Zhu, J. Shi, H. Chen, W. Shen, X. Dong. Microporous Mesoporous Mater., 84, 218 (2005). Y. Han, J.Y. Ying. Angew. Chem. Int. Ed., 44, 288 (2005). J. Salonen, L. Laitinen, A.M. Kaukonen, J. Tuura, M. Björkqvist, T. Heikkilä, K. Vähä-Heikkilä, J. Hirvonen, V.-P. Lehto, J. Control. Release, 108, 362 (2005). C.Y. Lai, B.G. Trewyn, D.M. Jeftinija, K. Jeftinija, S. Xu, S. Jeftinija, V.S.-Y Lin. J. Am. Chem. Soc., 125, 4451 (2003). V.S.Y. Lin, C.Y. Lai, J. Huang, S.A. Song, S. Xu. J. Am. Chem. Soc., 123, 11510 (2001).
64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87.
Nanostructured Mesoporous Materials 88.
273
D.R. Radu, C.-Y. Lai, K. Jeftinija, E.W. Rowe, S. Jeftinija, V.S.-Y. Lin. J. Am. Chem. Soc., 126, 13216 (2004). 89. Q. Yang, S. Wang, P. Fan, L. Wang, Y. Di, K. Lin, F.-S. Xiao. Chem. Mater., 17, 5999 (2005). 90. R. Hernández, H.R. Tseng, J.W. Wong, J.F. Stoddart, J.I. Zink. J. Am. Chem. Soc., 126, 3370 (2004). 91. T.D. Nguyen, H.R. Tseng, P.C. Celestre, A.H. Flood, Y. Liu, J.F. Stoddart, J.I. Zink. Proc. Natl. Acad. Sci. USA, 102, 10029 (2005). 92. A.B. Descalzo, R. Martínez-Máñez, F. Sancenón, K. Hoffmann, K. Rurack. Angew. Chem. Int. Ed., 45, 5924 (2006). 93. J.H. Chang, C.H. Shim, B.J. Kim, Y. Shin, G.J. Exarhos, K.J. Kim. Adv. Mater., 17, 634 (2005). 94. N.K. Mal, M. Fujiwara, Y. Tanaka, T. Taguchi, M. Matsukata. Chem. Mater., 15, 3385 (2003). 95. N.K. Mal, M. Fujiwara, Y. Tanaka. Nature, 421, 350 (2003). 96. M. Arruebo, M. Galán, N. Navascués, C. Téllez, C. Marquina, M.R. Ibarra, J. Santamaría. Chem. Mater., 18, 1911 (2006). 97. S. Giri, B.G. Trewyn, M.P. Stellmaker, V.S.-Y. Lin. Angew. Chem. Int. Ed., 44, 5038 (2005). 98. E. Ruiz-Hernández, A. López-Noriega, D. Arcos, I. Izquierdo-Barba, O. Terasaki, M. Vallet-Regí. Chem. Mater., 19, 3455 (2007). 99. L. Tang, A.M. Persky, G. Hochhaus, B. Meibohm. J. Pharm. Sci., 93, 2184 (2004). 100. J. Fan, C. Yu, F. Gao, J. Lei, B. Tian, L. Wang, Q. Luo, B. Tu, W. Zhou and D. Zhao. Angew. Chem. Int. Ed., 42, 3146 (2003).
CHAPTER 6 CHEMICAL SYNTHESIS, SELF-ASSEMBLY AND APPLICATIONS OF MAGNETIC NANOPARTICLES
Sheng Peng, Jaemin Kim and Shouheng Sun* Department of Chemistry, Brown University, Providence, RI 02912, USA *Email:
[email protected]
This chapter reviews recent advances in chemical synthesis, selfassembly, and applications of novel magnetic nanomaterials in the past few years. The chemical syntheses of several important classes of magnetic nanoparticles are covered, including ferrites, metallic iron, cobalt, alloys, and rare-earth hard magnets. Several examples of shapecontrolled synthesis of nanoparticles and shape-induced texture in selfassembled nanoparticle superlattices are also outlined. Potential applications, particularly in data storage, permanent magnet and biomedicine, of these nanoparticles and assemblies are discussed.
1. Introduction 1.1. General Background Materials in nanoscale are often found to exhibit very unique and distinguished properties, comparing to respective bulk forms [1,2]. This is originated from the finite number of atoms within each particle and from the large proportion of surface atoms. Besides such so-called finite size effect, significant structure-property relationships have also been revealed in nanomaterials [3-5]. Thanks to the numerous research efforts over the past decade, great progress has been made on the design and fabrication of nanomaterials with controlled size, shape, crystalline and composition, and with novel physical and chemical properties [6-8]. 275
276
Peng et al.
For magnetic nanomaterials, one of the most technologically important characteristics is the process of magnetization reversal [9]. This property determines greatly the performance of a material and changes drastically with size. According to magnetic domain theory, the formation of a domain wall inside a magnetic particle is not thermodynamically favored when the size decreases to a certain level. In this case, the particle consists of only one magnetic domain with all of the spins aligned in the direction of magnetic easy axis. The magnetization reversal of such a single-domain particle occurs by rotation of its easy axis. The reversal process has a relaxation time τ = τ0eKV/2kT (K: anisotropic constant, V: particle volume, k: Boltzmaann constant, and T: temperature) [10,11]. The KV term represents the magnetic anisotropy (the energy barrier between the two orientations), while kT stands for thermal energy. It is explicit that τ largely depends on the competition between the two energy terms. In reality, when a material that has a large KV value is present in an external field H, its particles tend to align in the direction of the field. An overall magnetic moment M (Fig. 1a) arises from the sum of all particle magnetizations. When all the particles are aligned in the field direction, the overall moment reaches saturation: Ms. As field is slowly released, randomization of the particles leads to a gradual moment drop. When H completely removed, a considerable degree of magnetization is retained with a measurable moment (remanence Mr). M further decreases as H increases in the reverse direction. It becomes zero when the particles are fully randomized, giving no net magnetization; and the field strength is called coercivity Hc (unless otherwise indicated in the paper, Hc is referred to the intrinsic coercivity, not to be confused with the BHc coercivity in an induction curve). Such history-depending H-M curve is called a hysteresis loop and the behavior is named ferromagnetism. However, KV will be comparable to kT when the size of the particle decreases to a certain level. In this case, ambient thermal fluctuation will agitate the magnetization of each particle from one direction to another. In the absence of an external field, the particles are fully randomized giving a net moment of zero. But the magnetization would be readily saturated or reversed once a relatively small field is applied (Fig. 1b).
Magnetic Nanoparticles
277
Figure 1. Schematic illustration of (a) a typical hysteresis loop of a ferromagnetic material and (b) a typical curve for a superparamagnetic material, and (c) the dependence of coercivity Hc on particle size D; when D decreases, Ds is the critical size below which particle is single-domain; Dp is the critical size below which coercivity is zero. (S-D: single-domain, M-D: multi-domain, SPM: superparamagnetic, and FM: ferromagnetic).
Unlike the ferromagnetic H-M curve, it shows no hysteresis. Such a material is said to be superparamagnetic. Such size dependent phenomena can be summarized in a scheme Fig. 1c. The conversion of magnetic properties for most magnetic materials occurs within tens of nanometers. Therefore the most intensive research has been focused on nanoparticles in this size range, because of their great potentials in a wide variety of applications. For example, ferromagnetic nanoparticles with large coercivities can keep their magnetizations for very long time and hence are promising candidates for high performance permanent magnet and ultrahigh density magnetic data storage [12-14]; while superparamagnetic nanoparticles are not subject to strong magnetic interactions in a stable dispersion and could be very useful in biomedical applications [15-17]. 1.2. Chemical Syntheses of Nanoparticles As size does matter in nanoscale, understanding the mechanism of the formation of monodisperse nanoparticles is crucial for the syntheses to achieve desired fine controls. Previously numerous physical methods have been applied to synthesize magnetic nanomaterials, such as melt sputtering, mechanical ball milling, vacuum-deposition and electrodeposition. These methods are able to produce magnetic micro- or nanoparticles with high purity and in large quantities; however, they generally
278
Peng et al.
have little control on particle size, size distribution and structural morphology. A variety of chemical synthesis techniques has been introduced, such as metal salt coprecipitation, solution-phase reduction, thermal decomposition, sol-gel processing, solvothermal syntheses and microemulsion methods [9,18]. Sonication/microwave assisted chemical syntheses and even biogenic syntheses of nanoparticles have also been demonstrated [6,19,20]. In this chapter we will mostly discuss the solution phase based chemical syntheses
(a)
(b)
Figure 2. (a) A scheme of nucleation and growth for the synthesis of monodisperse nanoparticles following La Mer model and (b) a typical hot-injection set-up employed in solution-based chemical synthesis of monodisperse nanoparticles. Reprinted with permission from Ref. 22, C. B. Murray et al., Annu. Rev. Mater. Sci. 30, 545, (2000), Copyright @ Annual Reviews.
Though there has not been a general method/mechanism applicable to the synthesis of all magnetic nanoparticles, one mostly-adopted consensus is that monodisperse particles can be synthesized in solution phase by a process with separated nucleation and growth steps. In a typical La Mer model (Fig. 2a), as the monomer concentration rapidly
Magnetic Nanoparticles
279
reaches above the critical saturation, a burst nucleation event occurs and consumes the precursors preventing further formation of nuclei [21,22]. The as-produced nuclei then grow at the same rate. Under ideal conditions, a dynamic equilibrium can be achieved by controlling the supply rate of precursor species close to the growth rate of the particles at which stage the reaction can be stopped when nanoparticles with desired size have been resulted. Practically, a hot-injection procedure is often employed to achieve the burst nucleation (Fig. 2b), and various surfactant molecules are introduced to capture the particle surface and to prevent particle agglomeration. Alternatively, a heating-up strategy is also widely applied [23], which mixes precursors, surfactants and solvent at low temperature and heats up the mixture to certain temperatures to initiate the particle clustering. This simple procedure provides scale-up potentials for the production of large quantities of nanoparticles. Both hot-injection and heating-up procedures could generate monodispersed nanoparticles (standard deviation in diameter σ < 10%) by tuning the reaction parameters [6,23]. 2. Ferrite Nanoparticles: MFe2O4 (M= Fe, Mn, Co) Ferrites are one of the most important and fascinating classes of magnetic materials. The ferrites discussed in this section adopt cubic spinel structures and have a general formula of MFe2O4, where M can be a wide variety of metal cations [9,24,25]. Usually, oxygen forms a facecentered cubic (fcc) close packing with M2+ cations and Fe3+ cations occupying either tetrahedral or octahedral interstices. Depending on what and where the M2+ cations are in the lattice, ferrites can show a great versatility in structures and properties. Due to their high magnetic permeability and electrical resistivity, ferrites have long been used in high frequency based applications, such as power generation, conditioning and conversion [9,10]. Nano-sized ferrites have attracted great research interests recently due to potential biomedical applications, such as drug delivery, biosensing, magnetic resonance imaging, and ferrofluid hyperthermia [26-30].
280
Peng et al.
2.1. Chemical Syntheses of Spherical Ferrite Nanoparticles Conventionally, ferrites with different M2+ cations and compositions were synthesized in aqueous phase by base-coprecipitation of transition metal and iron salts. Despite its advantage in mass production of ferrite ferrofluids, this method lacks control in particle size and size distribution. Recently, several distinguished synthetic routes have been developed to fabricate monodisperse ferrite nanoparticles with sizes below 30nm. A facile organic phase synthesis utilizing high temperature reaction of metal acetylacetonates and 1,2-alkanediol in the presence long-chain surfactants was developed [31,32]. For example, uniform 6nm singlecrystalline Fe3O4 nanoparticles (Fig. 3a) were synthesized by stepwise heating of a reaction mixture of Fe(acac)3, 1,2-hexadecanediol, oleic acid (OA), oleylamine (OAm) and benzyl ether up to reflux (300°C) for 1hr. The nanoparticle size could be tuned up to 20nm by multiple seedmediated growths using small particles as seeds. This recipe does not require fractionation procedure to achieve the desired size distribution and can be readily extended to the synthesis of other ferrites (MFe2O4, M = Co, Mn, Ni, etc) by adding stoichiometric amount of M(acac)2 together with Fe(acac)3 into the reaction mixture (Fig. 3b). Later the same group further improved this procedure to avoid the multi-step seedmediated processes [33]. By tuning parameters, such as surfactant, alkanediol and solvent amounts, uniform MFe2O4 ferrite nanoparticles with sizes 5-15nm have been synthesized using simple one-pot syntheses.
Figure 3. Transmission electron microscopy (TEM) images of (a) 6nm Fe3O4 particles and (b) 16nm CoFe2O4 particles, and (c) hysteresis loops of 16nm CoFe2O4 particle assembly measured at (A) 10K and (B) 300K. Reproduced with permission from Ref. 32, S. Sun et al., J. Am. Chem. Soc. 126, 273, (2004), Copyright @ American Chemical Society.
Magnetic Nanoparticles
281
Room temperature magnetic measurements of the as-synthesized 16nm Fe3O4 nanoparticles showed superparamagnetic behavior. The nanoparticles reached a saturation magnetization of 83emu/g, close to the value of bulk Fe3O4 magnetite. At lower temperature, the nanoparticles became ferromagnetic with a coercivity of 450Oe at 10K. However, due to their large magnetocrystalline anisotropy, the 16nm CoFe2O4 nanoparticles exhibited a room temperature Hc of 400Oe and a much larger Hc of 20kOe at 10K (Fig. 3c). Note that Fe3O4 nanoparticles are also important in several other connections via phase transformations. Magnetite (Fe3O4) is well-known to undergo a topotactic oxidation to form maghemite (γ-Fe2O3) and further to hematite (α-Fe2O3). These conversions have been confirmed by first annealing the as-synthesized 16nm Fe3O4 particle assembly in O2 at 250°C for 6hr, followed by further treatment of the γ-Fe2O3 assembly in Ar at 500°C. Reductive annealing was also carried out in a reductive atmosphere (Ar + 5% H2) at 400°C; Fe3O4 particle assembly was converted into body-centered cubic (bcc) Fe with a much higher Ms of 186emu/g [32,33]. FeO nanoparticles made from reductive decomposition of Fe(acac)3 in a mixture of pure OA and OAm could be readily converted to Fe3O4 and γ-Fe2O3 by heat treatment in air at 120-200°C [34]. Nonstoichiometric iron oxide nanoparticles were also synthesized and studied [35]. An alternative route to make ferrite nanoparticles is via high temperature decomposition of organometallic precursors (sometimes assisted by an oxidation process using air or organic oxidizers). For
Figure 4. TEM images of (a) 5nm, (b) 12nm, (c) 16nm, and (d) 22nm iron oxide nanoparticles. Reproduced with permission from Ref. 42, J. Park et al., Nat. Mater. 3, 891, (2004), Copyright @ Nature Publishing Group.
282
Peng et al.
example, decomposition of FeCup3 (Cup: C6H5N(NO)O-) with longchain amines as surfactants gave γ-Fe2O3 nanoparticles [36]. Later, a more common precursor was employed; γ-Fe2O3 particles were produced by thermolysis of iron(0) pentacarbonyl Fe(CO)5 in the presence of OA followed by an oxidation process using trimethylamine N-oxide (CH3)3NO [37-39]. This recipe was readily extended to the synthesis of CoFe2O4 particles by replacing Fe(CO)5 with a bimetallic precursor of (η5-C5H5)CoFe2(CO)9 [40]. In the syntheses, metallic nanoparticles were generated from the decomposition of metal-oleate complex (via the reaction between organometallic precursor and OA molecules). Consequent oxidation of the metallic particles resulted in highly crystalline and monodisperse ferrite nanoparticles with sizes from 4nm to 16nm. A one-nanometer-level fine control of the particle size was acquired via similar seed-mediated procedures [41]. Since the metal-oleate intermediate species were the key during the reaction, metal-oleate complexes were prepared directly by reacting metal chlorides with sodium oleate [42]. The synthesis was advantageous in mass production (up to 40g) of uniform iron oxide nanoparticles with sizes 5-22nm. For example, pre-synthesized iron-oleate in 1-octadecene (ODE) was slowly ramped to 320°C, and was aged for 30min to give uniform 12nm nanoparticles. Various solvents with different boiling points (b.p.) were used to control the particle size (Fig. 4), e.g. 1-hexadecene (b.p. 274°C) was used to produce 5nm particles while trioctylamine (b.p. 365°C) led to 22nm ones. Characterizations showed smaller particles were closer to γ-Fe2O3 while the bigger ones were more as Fe3O4. However, the iron oxide prepared by this recipe showed relatively lower magnetic moments (e.g. 16nm Fe3O4 nanoparticles Ms ~ 20emu/g at 5K [42].). By using different metal-oleates, the recipe was also extended readily to the synthesis of MnO nanoparticles, CoO nanorods, and Fe nanocubes. Other methods [43], such as liquid-solid-solution (LSS) phase transfer synthesis [44] and even bacteria/virus assisted bio-production [20] of Fe3O4 nanoparticles, have been reported. Superparamagnetic chalcospinel particles were also synthesized [45].
Magnetic Nanoparticles
283
Figure 5. (a) TEM images of CNCs made of Fe3O4 nanoparticles (5.8nm±0.2nm diameter) viewed along the zone axes of [001] and [011] (scale bars: 20nm). Reproduced with permission from Ref. 63, J. Zhuang et al., Angew. Chem. Int. Ed. 47, 2208, (2008). (b) TEM image and schematic illustration (inset) of polyacrylate-capped Fe3O4 CNCs (scale bar: 100nm); (c) Photographs of the response of CNCs to an external magnetic field; the magnet-sample distance decreases gradually from right to left; e) Dependence of the reflection spectra of the CNCs (avg. diameter 120nm) on the magnet-sample distance. Diffraction peaks show blue-shift (from right to left) as the distance decreases. Reproduced with permission from Ref. 66, J. Ge et al., Angew. Chem. Int. Ed. 46, 7428, (2007), Copyright @ Wiley-VCH.
2.2. Shape-Controlled Synthesis and Self-Assembly Self-assembly here is referred to as spontaneous ordering of particle building blocks via weak interactions to form thermodynamically more stable structures [46,47]. In a sense that nanoparticles in a self-assembled structure resemble the atoms in a crystal lattice, these ordered structures are often called superlattices. Again, like different atoms in the lattice of a compound, multiple types of nanoparticles can be assembled together to form binary or ternary superlattices with various structures [39,48-51]. In order for a self-assembly to take place, the process must lead to a lower Gibbs free energy without the intervention of external forces. This usually requires slow and steady evaporation of the nanoparticle dispersion to give the particles enough time and energy to reorder into superlattices. Binary and ternary assemblies usually require control on more parameters, such as size and ratio of each component; several examples will be discussed later. Other methods include LangmuirBlodgett film, template-assisted assembly DNA-induced assembly and
284
Peng et al.
etc [52-59], will not be discussed. Due to the collective effect, selfassembled superlattices may exhibit different or even new properties comparing to single, unassembled particles, and are thus utilized to build novel nanoscale devices [60,61]. Self-assembly of Fe3O4 nanoparticles into a monolayer has been shown in Fig. 3-4. Stacking layers will lead to bilayer and multi-layer superlattices. In both cases, close-packed structures are frequently observed and will be further discussed in the next section. Another form of self-assembly is the colloidal nanoparticle cluster (CNC). Increasing research interests have been focused on this class of photonic gap materials due to their potential optoelectronic applications. Several methods of fabricating polymer-capped CNCs were reported. Figure 5a illustrates that two CNCs made of monodisperse Fe3O4 nanoparticles show nearly perfect fcc superlattice structures [62,63]. Lattice fringes and zone axes observed were similar to those in an atomic fcc lattice. Another group also reported the fabrication of superparamagnetic Fe3O4 CNCs using polyacrylates (Fig. 5b) [64-67]. They described the photonic CNCs had a significant dependence of the reflection spectra at normal incidence on the sample-magnet distance (Fig. 5c-d). Such fieldresponsive photonic CNCs would be promising for novel optical sensors and color display units. Many research efforts have also been devoted to shape-controlled synthesis [68] and shape-induced texture of ferrite particles. Cube-like and polyhedron-shaped MnFe2O4 nanoparticles were made by controlling surfactant/Fe(acac)3 ratios (Fig. 6a-b) [69]. HRTEM proved the cube-like particles were terminated with {100} planes while the polyhedronshaped ones were truncated dodecahedrons. Controlled evaporation of dilute hexane dispersion led to superlattices with four-fold symmetry for both types of particles (Fig. 6c-d). Although showing similar structures under TEM, XRD illustrated an intensified (400) peak for the assembly of cube-like particles and an enhanced (220) peak for the other (Fig. 6e-f). This revealed the two assemblies had preferred crystal orientation with different planes parallel to the Si substrate ({100} and {110} planes respectively). Not only does the shape of nanoparticles have an effect on self-assembly, but also shows a significant influence on magnetic properties. A much lower coercivity from cubic CoFe2O4 nanoparticles
Magnetic Nanoparticles
285
was reported comparing to spherical ones with equal particle volume. This low coercivity was originated from a much smaller surface anisotropy in the cubic particles due to the less severe surface magnetic disorder and pinning [70].
Figure 6. TEM images of 12nm MnFe2O4 nanoparticles: (a) cubelike, (b) polyhedronshaped, (c) self-assembly of (a), and (d) self-assembly of (b). (e) and (f) relative XRD patterns of (c) and (d). Reproduced with permission from Ref. 69, H. Zeng et al., J. Am. Chem. Soc. 126, 11459, (2004), Copyright @ American Chemical Society.
From thermolysis of Fe(CO)5 in the presence of long-chain amines and air, a mixture of diamond-, sphere-, and triangle-shaped γ-Fe2O3 nanoparticles were synthesized [71]. Detailed study on the synthesis concluded that the diamond-shaped particles were similar to the truncated dodecahedrons, while the triagonal-shaped ones were nanoprisms from highly truncated tetrahedrons. However, a much less literature is found on ferrite nanomaterials with one-dimensional (1-D) structures. Ferrite multi-pods, nanorods and hollow tubes were reported but the materials had poor morphology and wide size distribution [72-76]. Recently, a facile chemical synthesis of hollow Fe3O4 nanoparticles was reported [77]. The synthesis was based on nanoscale Kirkendall effect due to unequal interfacial diffusions between two components during the controlled oxidation of core/shell Fe/Fe3O4 nanoparticles. Different from previous physical methods (e.g. gas phase oxidation and electron beam irradiation [78-80]), which let to insoluble particles, the
286
Peng et al.
monodisperse (σ<7%) hollow Fe3O4 nanoparticles (with tunable outer diameter 10-18nm and inner void 5-10nm) obtained from this method were readily dispersed in hexane. Self-assembled monolayer and multilayer hcp-structures were prepared by slow evaporation of hollow particle dispersion (Fig. 7a-c). Due to the poly-crystalline nature, the asprepared 16nm hollow Fe3O4 particles showed a broader XRD pattern and a more-difficult-to-saturate hysteresis (Fig. 7d) comparing to solid particles with the same size. HRTEM study further suggested possible pores between different Fe3O4 grains in the shell. The crystallinity and pore size of the hollow particles were controlled by reaction temperature. These hollow Fe3O4 nanoshells have been used as the carrier for other nanoparticulate components [81].
Figure 7. TEM images of 16nm hollow Fe3O4 nanoparticle assemblies with (a) monolayer (inset shows HRTEM image of a single particle, bar 10nm), (b) bilayer and trilayer structures. The trilayer assembly illustrates a few defects in the third layer, bars are 20nm. (c) room temperature hysteresis loops of 16nm solid and hollow Fe3O4 particles. Reproduced with permission from Ref. 77, S. Peng et al., Angew. Chem. Int. Ed. 46, 4155, (2007), Copyright @ Wiley-VCH.
2.3. Surface Modification for Biological Applications The monodisperse nanoparticles prepared as described above are generally hydrophobic and only soluble in non-polar or weakly polar solvents, due to the long chain hydrocarbon present at the particle surface [15-17]. Generally, for the nanoparticles to be compatible with biological systems and applicable under physiological conditions, they must be
Magnetic Nanoparticles
287
water-soluble and stable at pHs ranging 5-9, under salt concentrations up to a few hundred mM and at various cell culture temperatures [15,82]. To meet such requirements, the hydrophobic particles should be modified to be hydrophilic and ready for further conjugation with biomnolecules.
Figure 8. Schematic illustration of nanoparticle surface modification strategies.
Two typical approaches are frequently applied in particle surface modification and functionalization: surfactant addition and surfactant exchange [82-84]. Schematic illustration is shown in Fig. 8. The first approach, surfactant addition takes advantage of hydrophobic interaction between hydrocarbon chains and immobilizes the new ligand with a functional group “F” over the original surfactant shell. In this reaction, a double-layer structure forms over the surface of the particles. Depending on the identity of “F”, such modified nanoparticles can be dispersed in various media and further conjugated with other biological species. The second approach, surfactant exchange, utilizes the stronger binding between the new surfactant and the particle surface to replace the original surfactant. The new surfactant usually should be bifunctional ligand with one end capable of anchoring on the particle surface via strong chemical interactions, while with the chain or the other end group “F” having a polar character to make the modified particles water-soluble.
288
Peng et al.
The surfactant addition is represented by a modification process of transferring magnetic iron oxide nanoparticles to aqueous phase using PEGylated-phospholipids (PEG: polyethylene glycol). A commercial phospholipid, DSPE-PEG(2000)Biotin, was applied to intereact with the original OA/OAm surfactant layer of Fe3O4 particles to give a robust double layer structure [33,85,86]. After the modification, the PEG chain in the phospholipid molecules render the Fe3O4 particles water soluble, while the biotin groups make it possible for strepavidin attachment via biotin-avidin interaction. Similarly, amphiphilic polymers, such as dextran, polyvinylpyrrolidone (PVP) and poly(aniline) have also been used to transfer hydrophobic particles into aqueous solution by surface addition [87-91]. Due to the easy availability of the reagents, the method advances in functionalizing magnetic particles with a variety of functional groups, e.g. -COOH, -SH and -NH2, and in facilitating the conjugation of DNA, proteins or antibodies with the particles [32,92,93]. In all these functionalization processes, the development of magnetic particles with thinner coating and smaller overall size is much favored for more efficient magnetic separation, detection and delivery [82]. For example, dopamine is a robust ligand which has high affinity towards an iron-based surface. This dopamine chemistry thus has been widely applied to modify monodispersed Fe3O4, γ-Fe2O3, and even FePt particles [94-97]. Recently, ultrasmall peptide-conjugated Fe3O4 nanoparticles were synthesized using catechol-based surface chemistry [98]. The inorganic Fe3O4 core size was 4.5nm while the overall hydrodynamic size of the conjugated particles was only 8.4nm. 3. Metallic Iron, Cobalt and Iron-Cobalt Alloy Nanoparticles An important class of magnetic materials is the 3d transition metals, including bcc-iron (Fe), fcc- or hcp- cobalt (Co), and fcc-nickel (Ni), which are the only three ferromagnetic elements at room temperature. Several chosen properties are listed in Table 1. As can be seen, Fe has very high saturation moment (218emu/g), more than two times of that of Fe3O4. In fact, iron, as one of the most ubiquitous elements on earth, is the origin of the term “ferromagnetism” [99]. Another important property listed is the Curie temperature Tc. Above Tc, thermal energy kT
289
Magnetic Nanoparticles
overwhelms spin alignment within a magnetic domain and the materials become paramagnetic. It is thus an important practical parameter for high temperature applications. Table 1. Several magnetic properties of the ferromagnetic elements: Fe, Co and Ni (Ref. 99). Element
μH (μB)
σs at 0K (emu/g)
σs at 293K (emu/g)
Tc (°K)
Fe Co Ni
2.22 1.72 0.62
222 162 57
218 161 54
1043 1388 627
Moreover, there are also a great variety of magnetic alloys composed of or containing Fe, Co and Ni, which often exhibit interesting properties. Particularly, the incorporation of Co in Fe gives a further increase in the saturation magnetization. In fact, FeCo alloy (Fe70Co30) has long been known to have the highest magnetization in ambient conditions, near 245emu/g [10,24]. This superiority has made it the key in current hard disk drive write-heads, where a maximum field as high as 2.4T is required during the writing process. Here, we will mostly focus on Fe, Co and FeCo. However, due to the extreme reactivity of these metals, especially Fe, fast oxidation of their fine powders in air produces huge amount of heat leading to spontaneous ignition. As the result, it is traditionally very difficult to fabricate and study such pure metallic nanoparticles, and it is even harder to apply them under ambient condition. Thus an irony has long existed that their nano-forms are strongly favored in the oxides but not the pure metallic species [99]. Recent work has begun to focus not only on the synthesis of metallic Fe and Co nanoparticles, but also on the stabilization of the as-prepared particles for further solution-based applications in air. 3.1. Synthesis and Stabilization of Metallic Fe, Co, and FeCo Particles Conventionally, nanoparticles of metallic Fe and Co have been fabricated via physical methods and have been extensively applied in
290
Peng et al.
ferrofluids and magnetic recording media. More recently, Fe and Co nanoparticles in the size range below 20nm are found to be superparamagnetic at room temperature and their stable dispersions with high magnetic moment are predicted to have important biomedical applications in biomagnetic separation and sensing, therapeutic hyperthermia, drug delivery and MRI contrast enhancement [9,100]. Procedures leading to monodisperse Fe and Co nanoparticles have been summarized [18,23,99]. The most common method is via thermolysis of relative organometallic complexes, in which the metal centers are frequently in zero-valent metastable state and could readily undergo decomposition under mild conditions [101]. One significant class of such complexes is the metal carbonyls, including iron pentacarbonyl Fe(CO)5, dicobalt octacarbonyl Co2(CO)8 and various derivatives. These reactive carbonyls have very high tendency to dissociate the CO’s when heated. Through complicated pathways, multiple metal cluster intermediates are formed and will further catalyze the reaction to give nanoparticles [100]. Polymer-assisted and sonochemical decomposition of Fe(CO)5 were extensively explored as early as in the 1970s [102-105]. Similar to those prepared using physical methods, the Fe particles obtained via such chemical processes often were polydispersed though having high moments provided not exposed in air. Recently, organic-phase themolyses of Fe(CO)5 in the presence of surfactants have been demonstrated to produce monodisperse Fe nanoparticles in dioctyl ether, either with or without Pt nanoparticle as seeds, and using OA/OAm as surfactants [106,107]. Monodisperse Fe particles with various sizes (5-19nm) were yielded by controlling surfactant/iron ratios. Low temperature (10K) magnetic measurements showed high saturation moments (>130emu/g) though the Fe particles had 0.5-2nm surface oxide layers. Alternative precursors, such as iron-oleate and iron(II) bis(trimethylsilyl)amide (Fe[NSi(Me3)2]2), were also introduced to Fe nanoparticle synthesis [42,108]. There has also been reports of Fe nanoparticles via reduction of Fe(acac)3 and other iron salts [109-115]. Although the particle size is well-controlled, the syntheses do reveal that the Fe particles so prepared are extremely reactive and subject to facile oxidation, giving various iron oxide nanoparticles. As a result, the
Magnetic Nanoparticles
291
syntheses have difficulty in producing stable Fe particle dispersions, especially aqueous dispersions. Very recently, a simple one-pot synthesis was developed to make monodisperse Fe nanoparticles (σ < 7%) by Fe(CO)5 thermolysis in ODE solvent and OAm surfactant [116,117]. The as-synthesized Fe particles were readily oxidized during the sample treatment in air, giving core/shell Fe/Fe3O4 with 8nm diameter Fe core and 2.5nm thick Fe3O4 shell (Fig. 9a). The shell thickness was similar to previously reported case [106,118]. HRTEM and XRD proved both components in the core/shell particles were amorphous initially. Follow-up annealing of the particle assembly in Ar crystallized Fe3O4 and Fe at 400 and 500°C respectively. Room temperature magnetic measurements revealed the nanoparticles were superparamagnetic with a saturation moment of 67emu/g particles (equal to 103emu/g of inorganic matters after removal of organic surfactants), which roughly corresponded to the sum of the core- and the shell-magnetizations on a volume basis [117]. Further oxidation occurred readily as the particles were exposed in air, giving a drastically declining moment and a rapid agglomeration of the dispersion.
Figure 9. TEM images of (a) the as-synthesized 8nm/2.5nm Fe/Fe3O4 nanoparticles (inset: HRTEM image); (b) the control-oxidized 5nm/5nm Fe/Fe3O4 particles. (c) The magnetic moment drop of Fe/Fe3O4 particle dispersions of (a) and (b) versus time exposed to air at room temperature. Reproduced with permission from Ref. 117, S. Peng et al., J. Am. Chem. Soc. 128, 10676, (2006), Copyright @ American Chemical Society.
To make Fe/Fe3O4 nanoparticles stable, crystalline Fe3O4 shell was produced by controlled oxidation of the as-synthesized nanoparticles using an oxygen transferring agent trimethylamine N-oxide (CH3)3NO. The shell thickness was tuned by controlling the amount of (CH3)3NO
292
Peng et al.
used. Fig. 9b illustrates typical 5nm/5nm oxidized Fe/Fe3O4 particles. In contrast to the case of the untreated Fe/Fe3O4 particles, the moment of the oxidized ones had a slightly lower initial value of 62emu/g particles, but dropped much slower when exposed in air and was finally stabilized at 56emu/g particles after hours (Fig. 9c). Comparing to Fe, a much larger research literature could be found on Co nanoparticle synthesis probably due to its higher stability. The monodisperse nanocrystals of 3d metals, especially Co, have been fabricated [119]. For example, a fast injection of superhydride into a reaction mixture of cobalt chloride salt, OA, trialkylphosphine, and octyl ether at 200°C gave ε-phase Co nanoparticles, which had a metastable complex cubic structure resembling β-phase Mn and could be converted to hcp- and fcc-phases by thremal annealing [120,121]. The selfassembled array of the ε-Co particles showed interesting spin-dependent electron transport [122]. Similar ε-phase Co nanoparticles were also synthesized using Co2(CO)8 and trialkylphosphine oxides [123-125]. The reaction mechanism and the ligand effect were studied [126]. By switching to acetate precursors and polyol reducing reagents, hcp-Co, Ni and Co/Ni alloy nanoparticles were synthesized [127]. Moreover, fcc-Co nanoparticles were produced using Co2(CO)8 precursor [127]. Such polycrystalline fcc-Co particles contained multi-twinned structures which were frequently observed in noble metal particles [3,128]. Another surfactant, sodium bis-(2-ethylhexyl) sulfosuccinate (NaAOT) was also widely used [129,130]. Co particle synthesis and magnetic characterizations have been well summarized [9,23,127]. Various coatings have been investigated to protect as-synthesized Co nanoparticles, such as Au, Pt, CdSe, and SiO2 [131-136]. However, such processes often yield polydisperse particles with non-uniform coatings and lead to dramatic reduction in magnetic moment due to the large volume proportion of the non-magnetic shell material. Recently a magnetic shell materials (MFe2O4) was explored to coat Co nanoparticles to enhance both chemical and magnetic stabilities while maintaining the overall high moment density [137]. Figure 10a shows monodisperse 11nm Co particles synthesized by decomposition of Co2(CO)8 in tetralin. The as-obtained Co particles served as seeds in a seed-mediated growth, and uniform 11nm/3nm Co/MFe2O4 core/shell composite particles were
Magnetic Nanoparticles
293
made (Fig. 10b). Magnetic measurements were carried out after the syntheses. The uncoated Co particles had Ms ~120emu/gCo-1, close to 75% of the bulk value; while the coated particles exhibited a slightly lower value of 103emu/g, which is reasonable due to the lower moment density of MFe2O4 (Fig. 10c). Stability of the particles were compared both in air as powders and in aqueous buffer as dispersions at elevated temperature (70°C) (Fig. 10d-f). Raw Co particles underwent quick oxidation in both cases while coated Co/MFe2O4 composites showed greatly enhanced stabilities. Particularly in aqueous phase, Co particles were so susceptible to oxidation that the dark-brown color of the initial dispersion rapidly faded within 30min as CoO was forming (Fig. 10e).
Figure 10. TEM images of (a) 10nm Co particles (inset bar: 20nm) and (b) 10nm/3nm Co/MnFe2O4 particles. Initial hysteresis loops (c) of both types of particles. Stability test of both particles exposed in air at 70°C: (d) magnetic moment drop of dry powders, and (e) and (f) particle dispersions in aqueous PBS buffer. Reproduced with permission from Ref. 137, S. Peng et al., J. Solid State Chem. 181, 1560, (2008), Copyright @ Elsevier.
CoFe alloys have long been manufactured by physical approaches such as balling milling. Only until very recently, monodisperse CoFe nanoparticles were achieved by solution phase syntheses. It was first
294
Peng et al.
reported that CoFe/Fe3O4 core/shell particles were obtained through interfacial diffusion and reduction of CoFe2O4/Fe particles [138]. More conveniently, CoFe alloy particles were made by simultaneous decomposition of Fe(CO)5 and Co2(CO)8 in 1,2-dichlorobenzene [139], or reductive decomposition of Fe(CO)5 and organometallic Co(N(SiMe3)2)2 under a 3 bar H2 atmosphere [140]. Uniform 15nm FeCo particles with atomic ratio of ~Fe60:Co40 spontaneously formed largescale 3D densely packed superlattices (Fig. 11a). The as-prepared FeCo particles exhibited 160~180emu/gFeCo-1 saturation magnetizations and adopted a non-periodic structure with several atomic layers of surface oxidation. Further study revealed the metastable structure would readily convert to bcc-FeCo by e-beam irritation or thermal annealing in Ar at 500°C. The annealing resulted in an increased moment of 220emu/g FeCo-1.
Figure 11. SEM and TEM images of FeCo alloy particles: (a) 15nm (inset: HRTEM, bar: 5nm), (b) 20nm, and (c) 7nm (inset: HRTEM, arrows: graphitic shell, bar: 2nm). Reproduced with permission from Ref. 140, C. Desvaux et al., Nat. Mater. 4, 750, (2005), Copyright @ Nature Publishing Groups; Ref. 141, G. S. Chaubey et al., J. Am. Chem. Soc. 129, 7214, (2007), Copyright @ American Chemical Society; and Ref. 142, W. S. Seo et al., Nat. Mater. 5, 971, (2006), Copyright @ Nature Publishing Groups.
Polyol co-reduction of Fe(III)- and Co(II)-acetylacetonates in OA/OAm under FG (7%H2) atmosphere was also applied to make FeCo alloy nanoparticles [141]. The saturation magnetizations of 207 and 129emu/g were obtained for 20- (Fig. 11b) and 10-nm FeCo nanoparticles. Similarly, further annealing of the as-synthesized FeCo particles could protect them from gradual oxidation by providing a carbon shell and
Magnetic Nanoparticles
295
leads to an increase in the Ms to 230emu/g. Air-stable FeCo particles (Fig. 11c) with graphitic shell were prepared directly using chemical vapor deposition (CVD) method [142]. The resulted 7nm FeCo particles showed an Ms of 215emu/g, without any degradation over one month exposure in air. The particles were transferred into aqueous phase using phospholipid-poly(ethylene glycol) based molecules and applied the stable dispersion in MRI contrast enhancement studies. In addition to these efforts, Au or Ag shell has been applied to protect CoFe nanocubes via sputtering methods [143,144].
Figure 12. Our unpublished [145] TEM images of self-assembled superlattices of Fe/Fe3O4 particles: (a) hexagonal close packing, (b) cubic close packing, and (c) bodycentered cubic packing. Insets show schemes of corresponding lattices. The dotted circles in (a) illustrate a few missing particles. Three different layers of particles are highlighted in (b) by dotted, dashed and solid circles. The order of the layers is arbitrary here.
3.2. Self-Assembly, Shape-Controlled Synthesis of Fe and Co Spherical Fe particles discussed above could form various superlattices via self-assembly. When adding a third layer to a closepacked bilayer structure, two typical variations may occur depending on how it is stacked. If the third layer is exactly the same as the first, the particles would overlap in these two layers when projected vertically. Such assembly is a hexagonal close packing. Shown in Fig. 12a, several missing particles in the third layer would be exactly on top of the ones in the first layer if a perfect assembly, thus confirming the structure. On the other hand, the third layer could slide to overlap none of the first two layers, giving cubic close packing (shown in Fig. 12b). In both close
296
Peng et al.
packings, each particle has six equivalent most adjacent neighbors. Other non-close packing such as body centered cubic packing was also observed (Fig. 12c), in which each particle is surrounded by eight rather than six other particles [145].
Figure 13. TEM images of various Fe and Co nanomaterials: (a) 7nm Fe cubes forming a cubic assembly, bar 10nm (Ref. 108); (b) a 30nm Fe cube (inset: simulated magnetic vortex state) (Ref. 147); (c) 20nm hollow Fe frames (inset: the conversion from solid cube to hollow frame) (Ref. 148); (d) 20nm triangular Co plates, bar 100nm (Ref. 155); (e) 5×85nm Co rods, bar 30nm (Ref. 159); and (f) 20nm hollow Co particles (inset: a closer view) (Ref. 160). Reproduced with permission from Ref. 108, F. Dumestre et al., Science 303, 821, (2004), and Ref. 155, V. F. Puntes et al., Science 291, 2115, (2001), Copyrights @ AAAS; Ref. 147, E. Snoeck et al., Nano Lett. 8, 4293, (2008), and Ref. 148, D. Kim et al., J. Am. Chem. Soc. 129, 5812, (2007), Copyrights @ American Chemical Society; and Ref. 159, F. Dumestre et al., Angew. Chem. Int. Ed. 42, 5213, (2003), and Ref. 160, K. M. Nam et al., Angew. Chem. Int. Ed. 47, 9504, (2008), Copyright @ Wiley-VCH.
Fe nanocubes made from decomposition of Fe[N(SiMe3)2]2 at 150°C under H2 atmosphere in presence of hexadecylamine (HDA) and longchain acids form high-quality superlattices (Fig. 13a) [108]. The 7nm Fe nanocubes gave a saturation moment (212emu/gFe-1) close to the bulk value at 2K. Interestingly, a much higher blocking temperature (>50K)
Magnetic Nanoparticles
297
was observed probably due to the shape anisotropy and dipolar antiferromagnetic couplings. The synthesis was later extended to synthesize ultra-small Fe (<2nm) nanoparticles which had a dramatic increase in the effective anisotropy constant while keeping considerably high magnetic moments [146]. Very recently, 30nm Fe nanocubes were made with vortex state magnetic structures (Fig. 13b) [147]. Another approach to Fe nanocubes and hollow nanoframes was produced (Fig. 13c) by high temperature (380°C) reductive decomposition of iron(II)-stearate complex in the presence of the extra sodium oleate [148]. The formation of hollowness was not due to Kirkendall effect but because of severe molten salt corrosion at high temperature [148]. Self-assembled structures of Co nanoparticles have been well studied. Hexagonal close packing was frequently observed in assemblies of spherical Co particles. [122,149-151]. Square arrangement was observed in the assembly of small Co particle (4nm) because of the significant steric forces between surfactant molecules [152-154]. On the other hand, chain- and ring-like structures were resulted in large Co particle assembly, as the magnetic dipole interaction dominates the assembly process [154-156]. Stripes of Co particles were formed using microcontact printing on Si substrates [157,158]. Spindle-like structures have also been reported when a small magnetic field was applied during the assembly [127]. Co has more variations in morphology comparing to Fe. Large Co particles of up to 20nm with cube-like shape were yielded [156]. Disklike Co nanoplates were produced [130,155]. These nanoplates were easily confused with short nanorods (Fig. 13d), because the frequently observed face-to-face stacking tended to line the plates up vertically on their edges in order to minimize magneto-static energy while maximize the hydrophobic interaction of the surfactant tails. Monodisperse Co nanorods (Fig. 13e) were synthesized by decomposition of organometallic precursor [Co(η3-C8H13)(η4-C8H12)] in the presence of HDA and long chain acids [159]. By controlling the alkyl chain length of the acid surfactant, the length and uniformity of the resulted Co nanorods was tunable. Magnetically, the nanorods display a strong effective anisotropy (Keff) as a consequence of both their shape and crystalline
298
Peng et al.
structure, where the magnetic c axis in the hcp-phase Co was parallel to the rod axis. Particularly, the 8 × 128nm rods exhibited 142emu/gCo-1 Ms and 7200Oe Hc at 2K in a field up to 5T. Hollow Co nanoparticles (Fig. 13f) were made by simultaneous fast out-diffusion of CoO oxide species and surface reduction of the oxides by oleylamine [160]. The resulted 20nm single-crystalline fcc-Co nano-parallelepipeds displayed 36emu/g Ms and 120Oe Hc at room temperature. Hollow Fe70Co30 nanospheres were made by in situ galvanic replacement reaction using freshly prepared Fe particles as sacrificial templates [161]. 4. Tetragonal (L10-phase) Hard Magnetic FePt Nanoparticles and Their Applications Magnetic iron-platinum (FePt) nanoparticles made from solution phase chemical syntheses have shown great potentials for high performance permanent magnet [162-164], high density data storage [13, 165-166], and highly efficient biomedicine applications [11,167-168]. Their magnetic properties can be tuned not only by particle size, but also by Fe, Pt composition and Fe, Pt atomic arrangement in the FePt alloy structure [169]. The near-equiatomic FePt nanoparticles are known to have a chemically disordered face-centered cubic (fcc) structure or a chemically ordered face-centered (fct) structure [170]. The fcc-structured FePt has a small coercivity and is magnetically soft. The fully ordered fct-structured FePt can be viewed alternating atomic layers of Fe and Pt stacked along the [001] direction [11], fct-FePt nanoparticles are magnetically hard with large values of anisotropy constant K, which can reach as high as 107 J/m3 [13]. This large magnetic anisotropy K is originated from spin-orbit coupling and the hybridization between Fe 3d and Pt 5d electrons [171-174]. Due to the high magnetocrystalline anisotropy and high chemical stability, fct-FePt particles are particularly interesting as models for nanomagnetism study [175,176] and as building blocks for constructing single nanoparticle information storage media [13,165-166].
Magnetic Nanoparticles
299
4.1. General Chemical Syntheses of fcc-FePt Nanoparticles and the Phase Change via Thermal Treatment The fcc-FePt nanoparticles can be synthesized by thermal decomposition of Fe(CO)5 and reduction of Pt(acac)2 [165,177]. Fine tuning of the sizes of the FePt nanoparticles from 2 to 9nm was obtained by control of reaction parameters such as reaction time, temperature, heating rate, and the precursor/surfactant ratio. Polyol reduction of metal salts [178] and other reduction methods were also used to make fcc-FePt nanoparticles. By mixing and heating both an iron salt and a platinum salt with the reducing agent, monodisperse FePt particles were produced. For example, FeCl2 and Pt(acac)2 were reduced by LiBEt3H and diol to produce high quality FePt nanoparticles [179]. As-synthesized fcc-FePt particles could be transformed into fct-FePt by high temperature treatment (usually >550°C [180-182]) under inert or reductive atmosphere. 4.2. Shape Controlled FePt Nanoparticles and Their Self-Assembly Fabrication of ordered nanomagnet arrays with controlled magnetic alignment is an important goal in achieving high density information storage and high performance permanent magnets. Monodisperse FePt nanocubes were synthesized by mixing OA and Fe(CO)5 with benzyl ether/ODE solution of Pt(acac)2, and heating the mixture to 120°C for about 5 min before OAm was added, and further heating at 205°C for 2 h [183]. The addition sequence of OA and OAm keyed the shape control of the FePt particles. Different from the assembly of the spherical FePt particles, which showed 3D random structure orientation, self-assembly of the FePt nanocubes leads to a superlattice array with each FePt cube exhibiting (100) texture. Thermal annealing converted the chemically disordered fcc-FePt nanocubes to chemically ordered fct-FePt, and the annealed assembly showed a strong (001) texture in the directions both parallel and perpendicular to the substrate. The hysteresis loop was exactly the same in both parallel and perpendicular directions of the assembly confirming what was concluded from the XRD analysis (Fig. 14).
300
Peng et al.
(a)
(b)
Figure 14. (a) TEM image of 6.9nm Fe50Pt50 nanocubes and (b) XRD of thermally annealed FePt nanocube assembly on a Si surface. Reprinted with permission from Ref. 183, M. Chen et al., J. Am. Chem. Soc., 128, 7132 (2006), Copyright @ American Chemical Society.
Figure 15. TEM images of Fe55Pt45 nanowires and nanorods with a length of 200nm (a), 50nm (b), and 20nm (c). (d) HRTEM image of portions of two single 50-nm Fe55Pt45 NWs. Reprinted with permission from Ref. 185, C. Wang et al., Angew. Chem. Int. Ed., 46, 6333 (2007), Copyright @ Wiley-VCH.
The synthesis and self-assembly of FePt nanocubes suggest that elongated nanocrystals such as FePt nanowires and nanorods may be even more interesting to be applied to textured magnetic alignment. The synthesis of FePt rods using autoclave without stirring led to the
Magnetic Nanoparticles
301
formation of FePt nanorods [184]. As-synthesized FePt rods (the atomic percentage of Fe ranged 22-40%) showed the length of 11nm to 80nm with an average diameter of 2nm to 3nm, but the rods did not show a preferred crystalline orientation [184]. Alternatively, FePt nanowires or nanorods from 20 to 200nm can be made by thermal decomposition of Fe(CO)5 and reduction of Pt(acac)2 in a mixture of OAm and ODE at 160°C and the aspect ratio of the rods was controlled by the volume ratio of OAm/ODE [185]. For example, FePt nanowires with an average length of over 200nm were made when only OAm was used as both surfactant and solvent. An OAm/ODE ratio of 3:1 led to FePt wires of 100nm in length, while a 1:1 volume ratio of OAm/ODE gave 20nm FePt nanorods. Figure 15 shows the representative TEM images of the FePt nanowires or nanorods. Controlled evaporation of the carrier solvent from the hexane dispersion of the 50-nm Fe55Pt45 nanorods led to an Fe55Pt45 rod array with the rods parallel to each other. Thermal annealing in an argon atmosphere at 750°C transforms the fcc-FePt to fct-FePt and the annealed 200-nm Fe55Pt45 NW assembly shows a much stronger (001) peak, indicating a partial structural alignment with the (001) planes parallel to the substrate. The in-plane magnetic hysteresis loop of the assembly shows the better squareness with the coercivity reaching 9.5kOe [185]. 4.3. Synthesis of Dispersible fct-FePt Nanoparticles High temperature annealing needed for structure transformation of FePt nanoparticles also results in various degrees of particle aggregation and sintering. To avoid this aggregation/sintering problem, FePt nanoparticles are either coated with SiO2 [186,187], or grinded with a large excess of NaCl [188,189] before high temperature annealing is applied. In the SiO2 coated FePt, the conversion to the fct-FePt could be seen only above 600-700°C the uncoated nanoparticles began the structure transformation at 550°C [165]. The SiO2 coating seems to have a retarding effect for the structure transformation as found for FePt–SiO2 granular films [190]. The XRD patterns of the present samples annealed above 700°C matched well with that of the fct-FePt. The degree of Fe/Pt ordering, which could be estimated from the relative intensity of the
302
Peng et al.
peaks representing the fcc and fct phases, was found to be improved up to 900°C but no further above this temperature. Figure 16 shows the TEM images of the SiO2-coated FePt before and after thermal annealing at 900°C. It can be seen the nanostructure of the FePt nanoparticles is retained after this high temperature treatment. The SiO2 coating can be removed in the base solution, giving dispersible fct-FePt nanoparticles [186,187].
Figure 16. TEM images of the SiO2-coated FePt particles (a) before annealing and (b) after annealing at 900°C. Reprinted with permission from Ref. 186, S. Yamamoto et al., Appl. Phys. Lett. 87, 032503, (2005), Copyright @ American Institute of Physics.
Grinded sodium chloride (NaCl) is another robust coating materials used to protect FePt nanoparticles from sintering during the high temperature annealing process [188]. In the experiments, NaCl was first ball-milled for 24 h. The fine-grained NaCl was then suspended in hexane to mix with the as-synthesized fcc-FePt nanoparticles. After mixing and hexane evaporation, the mixture was annealed up to 700°C to complete the fcc to fct structure transition. The NaCl was removed by water washing, leaving fct-FePt nanoparticles that could be re-dispersed in hexane containing oleic acid and oleylamine. In a very recent protection effort, the fcc-FePt nanoparticles were coated with a layer of MgO, the controlled MgO coating process was developed [169,191]. In this process, MgO was coated on the FePt nanoparticle surface by decomposition of Mg(acac)2 in the presence of tetradecane-1,2-diol, OA, and OAm in benzyl ether at 298°C. The FePt/MgO can stand up to 800°C annealing without showing any observable FePt nanoparticle sintering. The MgO shell can be removed readily by a dilute acid washing, giving fct-FePt particles that are
303
Magnetic Nanoparticles
stabilized by a mixture of oleic acid and 1-hexadecanethiol and dispersed in hexane. The fct-FePt nanoparticles have a coercivity reaching 3.2T at 5K and 2T at 300K. This MgO coating approach is specifically interesting as it not only protects FePt particles from sintering at temperatures up to 800°C, but also provides a model system for studying magnetic tunneling between FePt particles with potentially much higher magneto-resistance than the thin film structure [192]. 5. Rare-earth Hard Magnets: Going into Nanoscale Rare-earth hard magnets, which usually compose of rare-earth elements and 3d transitional metals of Fe or Co, are the most technically important permanent magnets in the industry nowadays, because of their substantially superior performance comparing to conventional ferrite and alnico magnets. In such compounds, light lanthanides with incompletely filled f shells are the origin of large magnetic anisotropies while Fe or Co gives rise to the high magnetizations and high Curie temperatures. Such distinguished permanent magnets could achieve desired field strength with relatively small volumes and could work at high temperatures, and therefore have been widely applied in computer hard drives, audio speakers, auto motors and etc.
Table 2. Selected magnetic properties of several rare-earth hard magnets [12,24, 193-196]. Note anisotropic constant K ≈ K1 for hard magnetic materials. Hk is the anisotropic field which represents the upper limit for the intrinsic coercive field of the magnetic material. Rare-earth Compounds
Ms 3 (emu/cm )
K1 7 3 (10 ergs/ cm )
Hk (kOe)
Tc (°C)
SmCo5
910
11-20
240-400
720
Sm2Co17
1000
3.9
78
916
Nd2Fe14B
1270
4.6
73
312
Sm2Fe17N3
1220
8.9
146
477
L10-FePt
1140
6.6
116
477
304
Peng et al.
Two rare-earth systems: SmCo and NdFeB, developed in late ‘60s and early ‘80s respectively [193,196-198], have been the center of enormous research activities since their discoveries. Typical alloys are SmCo5, Sm2Co17 and Nd2Fe14B. As shown in Table 2, these rare-earth hard magnets have extremely large anisotropies. Particularly, among all known materials, SmCo5 has the largest K which is two orders greater than traditional ferrites and even one order greater than L10 phase Ptalloys. Thus it is expected that rare-earth compounds with nanoscale magnetic grains would possess enhanced coercivities to give even larger magnetic field once magnetized. Also to be noted, SmCo alloys have relatively low magnetizations comparing to Nd2Fe14B. A frequent solution both in research and in industry is via incorporation of soft phase Fe (Co) based high moment materials to enhance the magnetizations through exchange coupling [199]. Various aspects have been considered and it is predicted fine nanoscale mixture of exchange coupled hard-soft phases would lead to optimized rare-earth magnets with performance for future magnetic energy storage applications [200]. Traditional industry fabrication requires mixing of precursor oxides, high temperature reductive annealing followed by post-sintering heat treatment and final compaction [194,196]. More physical techniques have been employed, including melt spinning, magnetron sputtering, cluster gun fabrication and ball milling. These methods provide only limited control of the sizes of the final magnetic grains [201-210]. Solution based chemical synthesis has recently been introduced to achieve such goal. However , it still remains a major challenge until now to obtain nanostructured rare-earth magnets with properties comparable to bulk magnets using chemical approaches, because that (1) the huge reduction potential difference between Co2+ and Sm3+ (-0.29V versus -2.30V in aqueous phase) makes it extremely hard for simultaneous and homogenous reduction in solution, (2) nanoforms of rare-earth elements are very chemically instable and notoriously prone to fast oxidation, and (3) further understanding on the magnetic interface needs to be achieved in nanoscale systems. Here we discuss a few cases, not all successful studies, of fabricating nanostructured SmCo and NdFeB by chemical approaches.
Magnetic Nanoparticles
305
An early attempt to synthesize Sm-Co based nanoparticles in organic solution by decomposition of Sm(acac)3 and Co2(CO)8 in dioctyl ether gave poor-Sm particles exhibiting only 50Oe coercivity [211]. Later, the polyol process was incorporated into making SmCo5 particles with sizes below 10nm under inert atmosphere [212]. Magnetic measurements of the particles gave a coercivity of 2.2kOe at 5K and a calculated K of about 2.1×106 ergs/cm3 which was about 1% of the bulk SmCo5. A further modification of the synthetic procedure was extended to Sm(Co1−xFex)5/Fe3O4 core/shell composites via a second polyol process to create a Fe3O4 shell over the as-prepared SmCo-particles [213]. The Sm(Co1−xFex)5/Fe3O4 core/shell particles exhibited 2.5kOe coercivity at 10K, and showed a K of about 1.98×106 ergs/cm3. Polymer-assisted process has been applied to the synthesis of SmCo based nanoparticles as well. SmCo5 particles were prepared by heating a
Figure 17. HRTEM image (a) room temperature, and low temperature (10K) hystereses (b) of SmCo5 nanoblabe via PVP-process; and HRTEM image (c) and low temperature (100K) magnetic measurement (d) of nanocrystalline SmCo5 via high temperature reductive annealing of Co/Sm2O3 particles. Reproduced with permission from Ref. 216, C. N. Chinnasamy et al., Appl. Phys. Lett. 93, 032505, (2008), Copyrights @ American Institute of Physics; and Ref. 217, Y. Hou et al., Adv. Mater. 19, 3349, (2007), Copyrights @ Wiley-VCH.
306
Peng et al.
tetraethyleneglycol (TEG) solution of SmCl3 and Co(acac)3 in the presence of polyvinyl pyrrolidone (PVP) at 300°C [212]. The particles had a room temperature coercivity of 1.1kOe. Using noble metal Au to catalyze the reduction of Sm precursor, Sm-based magnetic particles were obtained with a kinked loop of 1.5kOe coercivity at room temperature [215]. Most recently, a significant improvement was made by using nitrate salts as precursors and produced 10 × 100nm blade-like rods (Fig. 17a) [216]. XRD revealed the air-stable SmCo-product was primarily hexagonal SmCo5 with slight Sm2Co17 content. The values of the intrinsic coercivity and the magnetization of the nanoparticles were 6.1kOe and 40emu/g at room temperature, and 8.5kOe and 44emu/g at 10K, respectively (Fig 17b). Though still not as good as bulk products, the SmCo nanoparticles prepared by this recipe shed bright light on future research of chemically synthesized SmCo. Alternative approaches were also developed to fabricate nanostructured SmCo. Recently, SmCo5 magnets were made by hightemperature reductive annealing of core/shell-structured Co/Sm2O3 nanoparticles [217]. The Co/Sm2O3 particles were chemically synthesized by decomposition of Sm(acac)3 over Co nanoparticle seeds. The Sm2O3-shell thickness and hence the Sm/Co ratio of the composite particles was controlled by varying the precursor and surfactant amounts. As-obtained core/shell particles with Sm/Co ratio 1:4.3 were reductively annealed in the presence of metallic Ca at 900°C to reduce Sm2O3 to Sm and to promote interface diffusion between Co and Sm, giving SmCo5. KCl as the dispersion medium during the annealing facilitated the reduction at relatively lower temperature comparing to bulk fabrication and prevented SmCo5 from sintering into large single-crystalline grains. XRD study confirmed hexagonal SmCo5, while HRTEM revealed nanocrystalline grains in the annealed product (Fig. 17c). The coercivity of the magnets reached 24kOe at 100K and 8kOe at room temperature with a remanent moment of 40-50emu/g (Fig. 17d). Sm2Co17 magnets were developed in the same way. Later, a similar process was used to produce exchange-coupled SmCo5/Fex nanocrystalline composites. A maximum coercivity of 11.6kOe was obtained at room temperature when x = 0.23, while an enhanced remanent moment of about 90emu/g was resulted when x = 1.5. Surfactant-assisted ball milling process has
Magnetic Nanoparticles
307
also been developed, giving Sm2Co17 particles ranging from 6 to 23nm [218]. Several chemical processes to produce Nd2Fe14B have been reported [219-221] even though the quality of the nanoparticle materials was far from perfect. 6. Summary and Outlook A dramatic burst in the research activities of chemically-synthesized magnetic nanoparticles has occurred in the past several years. We have only presented representative work on magnetic ferrites, 3d transitional metals, FePt-alloys and rare-earth nanocomposite magnets. We also illustrated the interesting magnetic properties observed in these materials and their potential biomedical applications. Future research directions may lie in: (1) controlled surface treatment of magnetic nanoparticles to ensure they are stable and biocompatible; (2) syntheses of magnetic materials with anisotropic morphology and controlled magnetic shapeanisotropy to form magnetically-aligned anisotropic structures; and (3) further understanding in magnetic interfaces and structure-property relationships at nanoscale. With these nanoscale controls, magnetic nanoparticle and their assembly will soon find promising applications in ultra-high density information storage, high performance permanent magnet and highly effective biomedicine. Acknowledgments The work at Brown University has been supported by DARPA/ARO, ONR/MURI, NSF/DMR, NIH/NCI and DOE/EPSCoR. References 1. 2. 3. 4. 5.
X. Batlle and A. Labarta, J. Phys. D: Appl. Phys. 35, R15 (2002). P. Buffat and J. P. Borel, Phys. Rev. A 13, 2287 (1976). Z. L. Wang, J. Phys. Chem. B 104, 1153 (2000). Z. L. Wang, Characterization of Nanophase Materials, Wiley-VCH, Weinheim, Germany (2000). S. C. Tjong, Nanocrystalline Materials: Their Synthesis-Structure-Property Relationships and Applications, Elsevier (2006).
308 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.
Peng et al. B. L. Cushing, V. L. Kolesnichenko and C. J. O’Connor, Chem. Rev. 104, 3893 (2004). Y.-W. Jun, J.-H. Lee, J.-S. Choi and J. Cheon, J. Phys. Chem. B 109, 14795 (2005). Y.-W. Jun, J.-S. Choi and J. Cheon, Angew. Chem. Int. Ed. 45, 3414 (2006). M. A. Tarascon, L. K. Kurihara, E. E. Carpenter, S. Calvin and V. G. Harris, Int. Mater. Rev. 49, 125 (2004). A. H. Morrish, The Physical Principles of Magnetism, John Wiley & Sons, New York (1965). S. Sun, Adv. Mater. 18, 393 (2006). D. Weller and A. Moser, IEEE. Trans. Mag. 35, 4423 (1999). D. Weller and M. F. Doerner, Annu. Rev. Mater. Sci. 30, 611 (2000). V. Skumryev, S. Stoyanov, Y. Zhang, G. Hadjipanayis, D. Givord and J. Nogues, Nature 423, 850 (2003). C. Xu and S. Sun, Polym. Int. 56, 821 (2007). S. Mornet, S. Vasseur, F. Grasset and E. Duguet, J. Mater. Chem. 14, 2161 (2004). E. Katz and I. Willner, Angew. Chem. Int. Ed. 43, 6042 (2004). T. Hyeon, Chem. Commun. 927 (2003). A. Gedanken and Y. Mastai, in The Chemistry of Nanomaterials ed. C. N. R. Rao, Wiley-VCH, Weinheim, Germany (2006), Chapter 6, p. 113-169. T. Matsunaga and Y. Okamura, Mater. Res. Soc. Symp. Proc. 724, N1.4 (2002). V. K. LaMer, R. H. Dinegar, J. Am. Chem. Soc. 72, 4847 (1950). C. B. Murray, C. R. Kagan and M. G. Bawendi, Annu. Rev. Mater. Sci. 30, 545 (2000). J. Park, J. Joo, S. G. Kwon, Y. Jang and T. Hyeon, Angew. Chem. Int. Ed. 46, 2 (2007). B. D. Cullity, Introduction to Magnetic Materials, Addison-Wesley, Reading, MA. (1972). R. C. O’Handley, Modern Magnetic Materials - Principles and Applications, John Wiley & Sons, New York (2000), pp. 126-132. T. Neuberger, B. Schopf, H. Hofmann, M. Hofmann and B. von Rechenberg, J. Magn. Magn. Mater. 293, 483 (2005). F. Sonvico, C. Dubernet, P. Colombo and P. Couvreur, Curr. Pharmaceut. Des. 11, 2091 (2005). A. K. Gupta and M. Gupta, Biomaterials 26, 3995 (2005). J. Dobson, Drug Dev. Res. 67, 55 (2006). C. J. Sunderland, M. Steiert, J. E. Talmadge, A. M. Derfus and S. E. Barry, Drug Dev. Res. 67, 70 (2005). S. Sun and H. Zeng, J. Am. Chem. Soc. 124, 8204 (2002). S. Sun, H. Zeng, D. B. Robinson, S. Raoux, P. M. Rice, S. X. Wang and G. Li, J. Am. Chem. Soc. 126, 273 (2004). J. Xie, S. Peng, N. Brower, N. Pourmand, S. X. Wang and S. Sun, Pure Appl. Chem. 78, 1003 (2006).
Magnetic Nanoparticles 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.
47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.
309
Y. Hou, Z. Xu and S. Sun, Angew. Chem. Int. Ed. 46, 6329 (2007). F. Redl, C. T. Black, G. C. Papaefthymiou, R. L. Sandstrom, M. Yin, H. Zeng, C. B. Murray and S. P. O’Brien, J. Am. Chem. Soc. 126, 14583 (2004). J. Rockenberger, E. C. Scher and A. P. Alivisatos, J. Am. Chem. Soc. 121, 11595 (1999). T. Hyeon, S. S. Lee, J. Park, Y. Chung and H. B. Na, J. Am. Chem. Soc. 123, 12798 (2001). Q. Guo, X. Teng, S. Rahman and H. Yang, J. Am. Chem. Soc. 125, 630 (2003). F. X. Redl, K.-S. Cho, C. B. Murray and S. O’Brien, Nature 423, 968 (2003). T. Hyeon, Y. Chung, J. Park, S. S. Lee, Y.-W. Kim and B. H. Park, J. Phys. Chem. B 106, 6831 (2002). J. Park, E. Lee, N. M. Hwang, M. Kang, S. C. Kim, Y. Hwang, J. G. Park, H. J. Noh, J. Y. Kim, J. H. Park and T. Hyeon, Angew. Chem. Int. Ed. 44, 2872 (2005). J. Park, K. An, Y. Hwang, J.-G. Park, H.-J. Noh, J.-Y. Kim, J.-H. Park, N.-M. Hwang and T. Hyeon, Nature Mater. 3, 891 (2004). N. Bao, L. Shen, Y. Wang, P. Padhan and A. Gupta, J. Am. Chem. Soc. 129, 12374 (2007). X. Wang, J. Zhuang, Q. Peng and Y. Li, Nature 437, 121 (2005). Y.-H. A. Wang, N. Bao, L. Shen, P. Padhan and A. Gupta, J. Am. Chem. Soc. 129, 12408 (2007). S. Sun, D. Weller and C. Murray, Self-Assembled Magnetic Nanoparticle Arrays, in The Physics of High Density Magnetic Recording, eds. M. Plumer, J. van Ek and D. Weller, Springer-Verlag, Chapter 9 (2001). S. Sun, Self-Assembled Nanomagnets, in Advanced Magnetic Nanostructures, eds. D. J. Sellmyer and R. Skomski, Klumer, Chapter 9 (2006). J. Cheon, J.-I. Park, J.-S. Choi, Y.-W. Jun, S. Kim, M. G. Kim, Y.-M. Kim and Y. J. Kim, PNAS 103(9), 3023 (2006). E. V. Shevchenko, J. B. Kortright, D. V. Talapin, S. Aloni and A. P. Alivisatos, Adv. Mater. 19, 4183 (2007). A. M. Kalsin, M. Fialkowski, M. Paszewski, S. K. Smoukov, K. J. M. Bishop and B. A. Grzybowski, Science 312, 420 (2006). E. V. Shevchenko, M. Ringler, A. Schwemer, D. V. Talapin, T. A. Klar, A. L. Rogach, J. Feldmann and A. P. Alivisatos, J. Am. Chem. Soc. 130, 3274 (2008). C. A. Mirkin, R. L. Letsinger, R. C. Mucic and J. Storhoff, Nature 382, 607 (1996). C. J. Loweth, W. B. Caldwell, X. Peng, A. P. Alivisatos and P. G. Schultz, Angew. Chem. Int. Ed. 38, 1808 (1999). F. Iskandar, L. Gradon and K. J. Okuyama, Colloid Interface Sci. 265, 296 (2003). H. Xu, L. Cui, N. Tong and H. Gu, J. Am. Chem. Soc. 128, 15582 (2006). I. Hussain, Z. X. Wang, A. I. Cooper and M. Brust, Langmuir 22, 2938 (2006). A. K. Boal, F. Ilhan, J. E. DeRouchey, T. Thurn-Albrecht, T. P. Russell and V. M. Rotello, Nature 404, 746 (2000). A. M. Kalsin and B. A. Grzybowski, Nano Lett. 7, 1018 (2007).
310
Peng et al.
59. 60. 61. 62. 63. 64. 65. 66. 67. 68.
P. Yang and F. Kim, Chem. Phys. Chem. 3, 503 (2002). C. T. Black, C. B. Murray, R. L. Sandstrom and S. Sun, Science 290, 1131 (2000). J.-S. Lee, E. V. Shevchenko and D. Talapin, J. Am. Chem. Soc. 130, 9673 (2008). J. Zhuang, H. Wu, Y. Yang and Y. C. Cao, J. Am. Chem. Soc. 129, 14166 (2007). J. Zhuang, H. Wu, Y. Yang and Y. C. Cao, Angew. Chem. Int. Ed. 47, 2208 (2008). J. Ge, Y. Hu, T. Zhang, T. Huynh and Y. Yin, Langmuir 24, 3671 (2008). J. Ge, and Y. Yin, Adv. Mater. 20, 3485 (2008). J. Ge, Y. Hu, and Y. Yin, Angew. Chem. Int. Ed. 46, 7428 (2007). J. Ge, and Y. Yin, J. Mater. Chem. 18, 5041 (2008). M. V. Kovalenko, M. I. Bodnarchuk, R. T. Lechner, G. Hesser, F. Schäffler and W. Heiss, J. Am. Chem. Soc. 129, 6352 (2007). H. Zeng, P. M. Rice, S. X. Wang and S. Sun, J. Am. Chem. Soc. 126, 11458 (2004). Q. Song and Z. J. Zhang, J. Am. Chem. Soc. 126, 6164 (2004). J. Cheon, N.-J. Kang, S.-M. Lee, J.-H. Lee, J.-H. Yoon and S. J. Oh, J. Am. Chem. Soc. 126, 1950 (2004). Y. Piao, J. Kim, H. B. Na, D. Kim, J. S. Baek, M. K. Ko, J. H. Lee, M. Shokouhimehr and T. Hyeon, Nature Mater. 7, 242 (2008). N. R. Jana, Y. Chen and X. Peng, Chem. Mater. 16, 3931 (2004). P. D. Cozzoli, E. Snoeck, M. A. Garcia, C. Giannini, A. Guagliardi, A. Cervellino, F. Gozzo, A. Hernando, K. Achterhold, N. Ciobanu, F. G. Parak, R. Cingolani and L. Manna, Nano Lett. 6, 1966 (2006). C.-J. Jia, L.-D. Sun, Z.-G. Yan, L.-P. You, F. Luo, X.-D. Han, Y.-C. Pang, Z. Zhang and C.-H. Yan, Angew. Chem. Int. Ed. 44, 4328 (2005). C. Wu, P. Yin, X. Zhu, C. Ouyang and Y. Xie, J. Phys. Chem. B 110, 17806 (2006). S. Peng and S. Sun, Angew. Chem. Int. Ed. 46, 4155 (2007). C. M. Wang, D. R. Baer, L. E. Thomas, J. E. Amonette, J. Antony, Y. Qiang and G. Duscher, J. Appl. Phys. 98, 094308 (2005). A. H. Latham, M. J. Wilson, P. Schiffer and M. E. Williams, J. Am. Chem. Soc. 128, 12632 (2006). D. Farrell, S. A. Majetich and J. P. Wilcoxon, J. Phys. Chem. B 107, 11022 (2003). A. Cabot, V. F. Puntes, E. Shevchenko, Y. Yin, L. Balcells, M. A. Marcus, S. M. Hughes and A. P. Alivisatos, J. Am. Chem. Soc. 129, 10358 (2007). J. Xie and S. Sun, Chemical Synthesis and Surface Modification of Monodisperse Magnetic Nanoparticles, in Nanomaterials: Inorganic and Bioinorganic Perspectives - Encyclopedia in Inorganic Chemistry, eds. C. M. Lukehart and R. A. Scott, John Wiley & Sons, Ltd. (2008). X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li, G. Sundaresan, A. M. Wu, S. S. Gambhir and S. Weiss, Science 307, 538 (2005). I. L. Medintz, H. T. Uyeda, E. R. Goldman and H. Mattoussi, Nature Mater. 4, 435 (2005).
69. 70. 71. 72. 73. 74.
75. 76. 77. 78. 79. 80. 81. 82.
83. 84.
Magnetic Nanoparticles 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109.
311
M. Platt, G. Muthukrishnan, W. O. Hancock and M. E. Williams, J. Am. Chem. Soc. 127, 15686 (2005). S. G. Grancharov, H. Zeng, S. Sun, S. X. Wang, S. O’Brien, C. B. Murray, J. R. Kirtley and G. A. Held, J. Phys. Chem. B 109, 13030 (2005). T. Pellegrino, L. Manna, S. Kudera, T. Liedl, D. Koktysh, A. L. Rogach, S. Keller, J. Radler, G. Natile and W. J. Parak, Nano Lett. 4, 703 (2004). J. W. Nah, Y. I. Jeong and C. S. Cho, J. Polymer Sci. Part B-Polymer Phys. 36, 415 (1998). V. P. Torchilin and V. S. Trubetskoy, Adv. Drug Delivery Rev. 16, 141 (1995). R. Gref, A. Domb, P. Quellec, T. Blunk, R. H. Muller, J. M. Verbavatz and R. Langer, Adv. Drug Delivery Rev. 16, 215 (1995). T. Verrecchia, G. Spenlehauer, D. V. Bazile, A. Murrybrelier, Y. Archimbaud and M. Veillard, J. Controlled Release 36, 49 (1995). B. Dubertret, P. Skourides, D. J. Norris, V. Noireaux, A. H. Brivanlou and A. Libchaber, Science 298, 1759 (2002). D. B. Robinson, H. H. J. Persson, H. Zeng, G. Li, N. Pourmand, S. Sun and S. X. Wang, Langmuir 21, 3096 (2005). C. Xu, K. Xu, H. Gu, R. Zheng, H. Liu, X. X. Zhang, Z. Guo and B. Xu, J. Am. Chem. Soc. 126, 9938 (2004). H. Gu, K. Xu, Z. Yang, C. Chang and B. Xu, Chem. Commun. 34, 4270 (2005). R. Hong, N. O. Fischer, T. Emrick and V. M. Rotello, Chem. Mater. 17, 4617 (2005). L. Wang, Z. Yang, J. Gao, K. Xu, H. Gu, B. Zhang, X. Zhang and B. Xu, J. Am. Chem. Soc. 128, 13358 (2006). J. Xie, K. Chen, H.-Y. Lee, C. Xu, A. R. Hsu, S. Peng, X. Chen and S. Sun, J. Am. Chem. Soc. 130, 7542 (2008). D. L. Huber, Small 1(5), 482 (2005). Q. A. Pankhurst, J. Connolly, S. K. Jones and J. Dobson, J. Phys. D: Appl. Phys. 36, R167 (2003). M. Green, Chem. Commun. 3002 (2005). T. Smith and D. Wychick, J. Phys. Chem. 84, 1621 (1980). C. Griffiths, M. Ohoro and T. Smith, J. Appl. Phys. 50, 7108 (1979). K. S. Suslick, S. B. Choe, A. A. Cichowlas and M. W. Grinstaff, Nature 353, 414 (1991). K. S. Suslick, M. Fang and T. Hyeon, J. Am. Chem. Soc. 118, 11960 (1996). D. Farrell, S. A. Majetich and J. P. Wilcoxon, J. Phys. Chem. B 107, 11022 (2003). D. Farrell, Y. Cheng, R. W. McCallum, M. Sachan and S. A. Majetich, J. Phys. Chem. B 109, 13409 (2005). F. Dumestre, B. Chaudret, C. Amiens, P. Renaud and P. Fejes, Science 303, 821 (2004). J. P. Wilcoxon and P. P. Provencio, J. Phys. Chem. B 103, 9809 (1999).
312
Peng et al.
110. S. Yamamuro, T. Ando, K. Sumiyama, T. Uchida and I. Kojima, Jpn. J. Appl. Phys. 43, 4458 (2004). 111. F. Li, C. Vipulanandan and K. Mohanty, Colloids Surf. A 223, 103 (2003). 112. A. Martino, M. Stoker, M. Hicks, C. Bartholomew, A. Sault and J. Kawola, Appl. Catal. A 161, 235 (1997). 113. C. Sei and C. O’Connor, Nanostruct. Mater. 12, 183 (1999). 114. G. N. Glavee, K. J. Klabunde, C. M. Sorensen and G. C. Hadjipanayis, Inorg. Chem. 34, 28 (1995). 115. L. C. Varanda, M. Jafelicci, P. Tartaj, K. O’grady, T. Gonzalez-Carreno, M. P. Morales, T. Munoz and C. J. Serna, J. Appl. Phys. 92, 2079 (2002). 116. H. Shao, H. Lee, Y. Huang, I. Ko and C. Kim, IEEE Trans. Magn. 41, 3388 (2005). 117. S. Peng, C. Wang, J. Xie and S. Sun, J. Am. Chem. Soc. 128, 10676 (2006). 118. J. Antony, Y. Qiang, D. Baer and C. Wang, J. Nanosci. Nanotechol. 6, 568 (2006). 119. S. Sun, C. B. Murray and H. Doyle, Mater. Res. Soc. Symp. Proc. 577, 395 (1999). 120. S. Sun and C. B. Murray, J. Appl. Lett. 85, 4325 (1999). 121. C. B. Murray, S. Sun, W. Gaschler, H. Doyle, T. A. Betley and C. R. Kagan, IBM J. Res. Dev. 45, 47 (2005). 122. C. T. Black, C. B. Murray, R. L. Sandstrom and S. Sun, Science 290, 1131 (2000). 123. U. Wiedwald, M. Spasova, M. Farle, M. Hilgendorff and M. Giersig, J. Vac. Sci. Technol. A 19, 1773 (2001). 124. V. F. Puntes, K. M. Krishnan and P. Alivisatos, Appl. Phys. Lett. 78, 2187 (2001). 125. D. P. Dinega, M. Bawendi, Angew. Chem. Int. Ed. 38, 1788 (1999). 126. A. C. S. Samia, K. Hyzer, J. A. Schlueter, C.-J. Qin, J. S. Jiang, S. D. Bader and X.-M. Lin, J. Am. Chem. Soc. 127, 4126 (2005). 127. C. B. Murray, S. Sun, H. Doyle and T. A. Betley, MRS Bull. 985 (2001). 128. S. Peng, Y. Lee, C. Wang, H. Yin, S. Dai and S. Sun, Nano Res. 1, 229 (2008). 129. C. Petit, A. Taleb and M. P. Pileni, J. Phys. Chem. B 103, 1805 (1999). 130. J.-I. Park, N. J. Kang, Y.-W. Jun, S. J. Oh, H.-C. Ri and J. Cheon, Chem. Phys. Chem. 6, 543 (2002). 131. Y. Bao and K. M. Krishnan, J. Magn. Magn. Mater. 293, 15 (2005). 132. J.-I. Park and J. Cheon, J. Am. Chem. Soc. 123, 5743 (2001). 133. J.-I. Park, M. G. Kim, Y.-W. Jun, J. S. Lee, W.-R. Lee, and J. Cheon, J. Am. Chem. Soc. 126, 9072 (2004). 134. W.-R. Lee, M. G. Kim, J.-R. Choi, J.-I. Park, S. J. Ko, S. J. Oh and J. Cheon, J. Am. Chem. Soc. 127, 16090 (2005). 135. H. Kim, M. Achermann, L. P. Balet, J. A. Hollingsworth and V. I. Klimov, J. Am. Chem. Soc. 127, 544 (2005). 136. M. L. Vadala, M. A. Zalich, D. B. Fulks, T. G. S. Pierre, J. P. Dailey and J. S. Riffle, J. Magn. Magn. Mater. 293, 162 (2005). 137. S. Peng, J. Xie and S. Sun, J. Solid State Chem. 181, 1560 (2008). 138. J. Li, H. Zeng, S. Sun, J. P. Liu and Z. L. Wang, J. Phys. Chem. B 108, 14005 (2004).
Magnetic Nanoparticles
313
139. A. Hütten, D. Sudfeld, I. Ennen, G. Reiss, K. Wojczykowski and P. Jutzi, Magn. Magn. Mater. 293, 93 (2005). 140. C. Desvaux, C. Amiens, P. Fejes, P. Renaud, M. Respaud, P. Lecante, E. Snoeck and B. Chaudret, Nature Mater. 4, 750 (2005). 141. G. S. Chaubey, C. Barcena, N. Poudyal, C. Rong, J. Gao, S. Sun and J. P. Liu, J. Am. Chem. Soc. 129, 7214 (2007). 142. W. S. Seo, J. H. Lee, X. Sun, Y. Suzuki, D. Mann, Z. Liu, M. Terashima, P. C. Yang, M. V. Mcconnell, D. G. Nishimura and H. Dai, Nature Mater. 5, 971 (2006). 143. J. Bai and J.-P. Wang, Appl. Phys. Lett. 87, 152502 (2005). 144. Y.-H. Xu and J.-P. Wang, Appl. Phys. Lett. 91, 233107 (2007). 145. S. Peng and S. Sun, unpublished results. 146. L.-M. Lacroix, S. Lachaize, A. Falqui, T. Blon, J. Carrey, M. Respaud, F. Dumestre, C. Amiens, O. Margeat, B. Chaudret, P. Lecante and E. Snoeck, J. Appl. Phys. 103, 07D521 (2008). 147. E. Snoeck, C. Gatel, L. M. Lacroix, T. Blon, S. Lachaize, J. Carrey and M. Respaud and B. Chaudret, Nano Lett. 8, 4293 (2008). 148. D. Kim, J. Park, K. An, N.-K. Yang, J.-G. Park and T. Hyeon, J. Am. Chem. Soc. 129, 5812 (2007). 149. S. I. Woods, J. R. Kirtley, S. Sun and R. H. Koch, Phys. Rev. Lett. 87, 137205 (2001). 150. M. Giersig and M. Hilgendorff, J. Phys. D: Appl. Phys. 32, L111 (1999). 151. C. Petit, A. Taleb and M. P. Pileni, Adv. Mater. 10, 259 (1998). 152. Y. Bao, M. Beerman and K. M. Krishnan, J. Magn. Magn. Mater. 272-276, e1367 (2004). 153. Y. Bao, M. Beerman and K. Krishnan, J. Magn. Magn. Mater. 266, L245 (2003). 154. Y. Gao, Y. Bao, M. Beerman, A. Yasuhara, D. Shindo and K. M. Krishnan, Appl. Phys. Lett. 84, 3361 (2004). 155. V. F. Puntes, K. M. Krishnan and P. Alivisatos, Science 291, 2115 (2001). 156. J. Gao, B. Zhang, X. Zhang and B. Xu, Angew. Chem. Int. Ed. 45, 1220 (2006). 157. S.-S. Bae, D. K. Lim, J.-I. Park, W.-R. Lee, J. Cheon and S. Kim, J. Phys Chem. B 108, 2575 (2004). 158. J.-I. Park, W.-R. Lee, S.-S. Bae, Y. J. Kim, K.-H. Yoo, J. Cheon and S. Kim, J. Phys. Chem. B 109, 13119 (2005). 159. F. Dumestre, B. Chaudret, C. Amiens, M. Respaud, P. Fejes, P. Renaud and P. Zurcher, Angew. Chem. Int. Ed. 42, 5213 (2003). 160. K. M. Nam, J. H. Shim, H. Ki, S.-I. Choi, G. Lee, J. K. Jang, Y. Jo, M.-H. Jung, H. Song and J. T. Park, Angew. Chem. Int. Ed. 47, 9504 (2008). 161. J. Huang, L. He, Y. Leng, W. Zhang, X. Li, C. Chen and Y. Liu, Nanotechnology 18, 415603 (2007) 162. E. F. Kneller, R. Hawig, IEEE Trans. Magn. 27, 3588 (1991). 163. R. Skomski and J. M. D. Coey, Phys. Rev. B 48, 15812 (1993). 164. H. Zeng, J. Li, Z. L. Wang, J. P. Liu and S. Sun, Nature 420, 395 (2002).
314
Peng et al.
165. S. Sun, C. B. Murray, D. Weller, L. Folks and A. Moser, Science 287, 1989 (2000). 166. A. Moser, K. Takano, D. T. Margulies, M. Albrecht, Y. Sonobe, Y. Ikeda, S. Sun and E. E. Fullerton, J. Phys. D: Appl. Phys. 35, R157 (2002). 167. Q. A. Pankhurst, J. Connolly, S. K. Jones and J. Dobson, J. Phys. D: Appl. Phys. 36, R167 (2003). 168. J. Gao, G. Liang, J. S. Cheung, Y. Pan, Y. Kuang, F. Zhao, B. Zhang, X. Zhang, E. X. Wu and B. Xu, J. Am. Chem. Soc. 130, 11828 (2008). 169. J. Kim, C. Rong, Y. Lee, J. P. Liu and S. Sun, Chem. Mater. 20, 7242 (2008). 170. O. Gutfleisch, J. Lyubina, K. H. Muller and L. Schultz, Adv. Eng. Mater. 7, 208 (2005). 171. T. Burkert, O. Eriksson, S. I. Simak, A. V. Ruban, B. Sanyal, L. Nordstrom and J. M. Wills, Phys. Rev. B: Condens. Matter Mater. Phys. 71, 134411 (2005). 172. G. Brown, B. Kraczek, A. Janotti, T. C. Schulthess, G. M. Stocks and D. D. Johnson, Phys. Rev. B. Condens. Matter Mater. Phys. 68, 052405 (2003). 173. O. Kitakami, S. Okamoto, N. Kikuchi and Y. Shimada, Jan. J. Appl. Phys. Part 2 42, 455 (2003). 174. J. B. Staunton, S. Ostanin, S. S. S. Razee, B. L. Gyorffy, L. Szunyogh, B. Ginatempo and E. Bruno, Phys. Rev. Lett. 93, 257204 (2004). 175. M. Ulmeanu, C. Antoniak, U. Wiedwald, M. Farle, Z. Frait and S. Sun, Phys. Rev. B 69, 054417 (2004). 176. C. Antoniak, J. Lindner, M. Spasova, D. Sudfeld, M. Acet, M. Farle, K. Fauth, U. Wiedwald, H. G. Boyen, P. Ziemann, F. Wilhelm, A. Rogalev and S. Sun, Phys. Rev. Lett. 97, 117201 (2006). 177. S. Sun, E. E. Fullerton, D. Weller and C. B. Murray, IEEE Trans. Magn. 37, 1239 (2001). 178. F. Fievet, J. P. Lagier and M. Figlarz, MRS Bull. 14, 29 (1989). 179. S. Sun, S. Anders, T. Thomson, J. E. E. Baglin, M. F. Toney, H. F. Hamann, C. B. Murray and B. D. Terris, J. Phys. Chem. B 107, 5419 (2003). 180. Z. R. Dai, S. Sun and Z. L. Wang, Nano Lett. 1, 443 (2001). 181. K. E. Elkins, T. S. Vedantam, J. P. Liu, H. Zeng, S. Sun, Y. Ding and Z. L. Wang, Nano Lett. 3, 1647 (2003). 182. T. S. Vedantam, J. P. Liu, H. Zeng and S. Sun, J. Appl. Phys. 93, 7184 (2003). 183. M. Chen, J. Kim, J. P. Liu, H. Fan and S. Sun, J. Am. Chem. Soc. 128, 7132 (2006). 184. M. Chen, T. Pica, Y. B. Jiang, P. Li, K. Yano, J. P. Liu, A. K. Datye and H. Fan, 129, 6348 (2007). 185. C. Wang, Y. Hou, J. Kim and S. Sun, Angew. Chem. Int. Ed. 46, 6333 (2007). 186. S. Yamamoto, Y. Morimoto, T. Ono and M. Takano, Appl. Phys. Lett. 87, 032503 (2005). 187. Y. Tamada, S. Yamamoto, M. Takano, S. Nasu and T. Ono, Appl. Phys. Lett. 90, 162509 (2007). 188. D. Li, N. Poudyal, V. Nandwana, Z. Jin, K. Elkins and J. P. Liu, J. Appl. Phys. 99, 08E911 (2006).
Magnetic Nanoparticles
315
189. C. Rong, N. Poudyal, G. S. Chaubey, V. Nandwana, Y. Liu, Y. Q. Wu, M. J. Kramer, M. E. Kozlov, R. H. Baughman and J. P. Liu, J. Appl. Phys. 103, 07E131 (2008). 190. T. K. Takahashi, T. Koyama, M. Ohnuma, T. Ohkubo and K. Hono, J. Appl. Phys. 95, 2690 (2004). 191. J. Kim, C. Rong, J. P. Liu and S. Sun, Adv. Mater. Online Available (2008). 192. S. S. P. Parkin, C. Kaiser, A. Panchula, P. M. Rice, B. Hughes, M. Samant and S. Yang, Nat. Mater. 3, 862 (2004). 193. D. Jiles, Introduction to Magnetism and Magnetic Materials, 2nd edition, Chapman & Hall, New York (1998). 194. M. McCaig, revisions A. G. Clegg, Permanent Magnets in Theory and Practice, 2nd edition, John Wiley & Sons, New York (1987). 195. D. J. Sellmyer and R. Skomski, Advanced Magnetic Nanostructures, Springer, New York (2006). 196. A. Menth, H. Nagel and R. S. Perkins, Ann. Rev. Mater. Sci. 8, 21 (1978). 197. E. A. Nesbitt, H. J. Williams and J. H. Wernick, J. Appl. Phys. 32, S342 (1961). 198. G. Hoffer and K. Strnat, IEEE Trans. Magn. 2, 487 (1966). 199. D. C. Jiles, Acta Materialia 51, 5907 (2003). 200. G. C. Hadjipanayis, J. Magn. Magn. Mater. 200, 373 (1999). 201. J. Ding, P. G. McCormick and R. Street, J. Alloys Compd. 228, 102 (1995). 202. S. K. Chen, J. L. Tsai and T. S. Chin, J. Appl. Phys. 79, 5964 (1996). 203. E. M. Kirkpatrick and D. L. Leslie-Pelecky, J. Appl. Phys. 87, 6734 (2000). 204. K. Gallagher, M. Venkatesan and J. M. Coey, IEEE Trans. Magn. 37, 2528 (2001). 205. L. Li, W. Y. Zhang, Y. Q. Zhou, J. Q. Li, B. G. Shen and J. Zhang, Appl. Phys. Lett. 80, 2660 (2002). 206. J. Zhou, R. Skomski and D. J. Sellmyer, J. Appl. Phys. 93, 6495 (2003). 207. A. Hsiao, S. Aich, L. Lewis and J. E. Shield, IEEE Trans. Magn. 40, 2913 (2004). 208. V. M. Chakka, B. Altuncevahir, Z. Q. Jin, Y. Li and J. P. Liu, J. Appl. Phys. 99, 08E912 (2006). 209. V. Pop, O. Isnard, I. Chicinas, D. Givord and J. M. LeBreton, J. Optoelectron. Adv. Mater. 8, 494 (2006). 210. S. Stoyanov, V. Skumryev, Y. Zhang, Y. Huang, G. Hadjipanayis and J. Nogués, J. Appl. Phys. 93, 7592 (2003). 211. K. Ono, Y. Kakefuda, R. Okuda, Y. Ishii, S. Kamimura, A. Kitamura and M. Oshima, J. Appl. Phys. 91, 8480 (2002). 212. H. Gu, B. Xu, J. Rao, R. K. Zheng, X. X. Zhang, K. K. Fung and C. Y. C. Wong, J. Appl. Phys. 93, 7589 (2003). 213. J. H. Hong, W. Kim, J. Lee and N. H. Hur, Solid State Commun. 141, 541 (2007). 214. N. Toshima, Macromol. Symp. 235, 1 (2006). 215. T. Matsushita, J. Masuda, T. Iwamoto and N. Toshima, Chem. Lett. 36, 1264 (2007).
316
Peng et al.
216. C. N. Chinnasamy, J. Y. Huang, L. H. Lewis, B. Latha, C. Vittoria and V. G. Harris, Appl. Phys. Lett. 93, 032505 (2008). 217. Y. Hou, Z. Xu, S. Peng, C. Rong, J. P. Liu and S. Sun, Adv. Mater. 19, 3349 (2007). 218. Y. Wang, Y. Li, C. Rong and J. P. Liu, Nanotechnology 18, 465701 (2007). 219. C. W. Km, Y. H. Km, H. G. Cha and Y. S. Kang, Phys. Scr. T129, 321 (2007). 220. Y. Haik, J. Chatterjee and C. J. Chen, J. Nanopar. Res. 7, 675 (2005). 221. Y. S. Kang and D. K. Lee, Int. J. Nanosci. 5, 315 (2006).
CHAPTER 7 RECENT DEVELOPMENT AND APPLICATIONS OF NANOIMPRINT TECHNOLOGY
Xing Cheng* and L. Jay Guo† Department of Electrical and Computer Engineering, Texas A&M University, College Station, TX 77843, USA *Email:
[email protected]; Department of Electrical Engineering and Computer Science, The University of Michigan, Ann Arbor, MI 48109, USA † Email:
[email protected]
Nanoimprint has emerged as a strong candidate of next-generation lithography techniques after more than a decade of intensive research and development. However, nanoimprint can be used as more than just a lithography technique. There are growing interests in using nanoimprint as an enabling fabrication technique for functional polymer processing and advanced multilayer polymer structures. In this review, recent applications of nanoimprint for nondestructive functional polymer patterning are summarized for polymer thin-film transistor, solar cell, and fuel cell applications. Improvement in reversal nanoimprint is presented. The improved technique is combined with optimized transfer-bonding technique for successful fabrication of three-dimensional multilayer polymer structures. Finally, a roll-to-roll nanoimprint technique is developed for low-cost high-throughput fabrication of micro- and nanostructures on a large scale.
1. Introduction In last decade nanoimprint lithography (NIL)1-6 has received a huge amount of attention in nanoscale patterning due to its characteristics such as sub-10 nm resolution,7-9 high throughput and low cost. Figure 1 shows the schematic of the originally proposed NIL process. A hard mold that contains nanoscale surface relief features is pressed into a polymer 317
318
Cheng and Guo
material cast on a substrate at controlled temperature and pressure, which creates thickness contrast in the polymer material. A thin residual layer is left underneath the mold protrusions acting as a soft cushion layer that prevents direct impact of the hard mold onto the substrate, which effectively protects the delicate nanoscale features on the mold surface. In most of the applications this residual layer needs to be removed by an anisotropic O2 plasma rective-ion-etching (RIE) process to complete the pattern definition. Since its inception, nanoimprint has received wide-spread attention due to its potential for high-resolution, low-cost and high-throughput nanopatterning. Current nanoimprint development mainly focus on two fronts: 1) developing nanoimprint as a lithography tool for micro- and nanoscale structure and device fabrication; and 2) developing nonconventional nanoimprint schemes and exploring novel applications for nanoimprint techniques. As to the principles of nanoimprint and its past development, Fig. 1. A schematic of nanoimprint. there have already been a number of excellent reviews published.3, 5, 6, 10-14 To avoid topical overlap, this review will focus on the authors’ recent work on the development and applications of nanoimprint lithography and can be regarded as a supplement to a previous review article.12 In nanoimprint, the micro- and nanostructures are first created in a soft material such as a polymer in thermal nanoimprint or a polymer precursor in UV-nanoimprint. Depending on specific applications, the patterned soft material can be removed eventually or remain as a permanent component in the final device structure. When used as a resist layer, the patterned polymer or polymer precursor will be used to transfer the pattern into another material by RIE or lift-off. This is the case for
Nanoimprint Technology
319
most current nanoimprint applications. However, there are growing interests in direct-patterning of soft materials by nanoimprint, particularly functional polymers. The patterned soft materials remain as a permanent structure or an active component in the final device structure. In this review, we present examples for nanoimprinting conjugated polymers for solar cell application and Nafion film for micro-fuel cell application. In functional polymer patterning, nanoimprint is used to create dense micro- and nanostructures in conjugated polymers to enhance their surface areas, which can increase the efficiency of organic solar cell devices. Not only nanoimprint can successfully pattern those functional polymers into micro- and nanostructures in a nondestructive fashion, but also it improves the properties of those functional polymers through flow-induced polymer chain orientation during nanoimprint. The origin and the factors that affect chain orientation in nanoimprinted polymer micro- and nanostructures are discussed. An example of improving organic electronics performance is presented to demonstrate the potential of manipulating chain orientation in polymer microstructures. In addition to functional polymer patterning, nanoimprint can also be used to pattern three-dimensional (3D) polymer structures for a variety of applications. Reversal nanoimprint and transfer-bonding techniques are improved and optimized for building 3D polymer structures in thermoplastic polymers in a layer-by-layer fashion. Critical factors that affect the fabrication yield of the polymer 3D structures are discussed in order to provide a general guideline in process development for specific applications. It is a consensus that nanoimprint is a very promising low-cost technique for replicating micro- and nanostructures with high-throughput. Since the cost of a technology often dictates its fate, it is highly desirable to further improve nanoimprint throughput to enhance its commercial impact. We discuss here the development of a roll-to-roll nanoimprint (R2RNIL) technique for micro- and nanostructure replication in a large scale and high speed fashion. It is envisioned that such R2RNIL will eventually make nanoimprint an economically viable solution towards a wide-range of commercial fabrications.
320
Cheng and Guo
2. Material Flow Behavior and the Associated Polymer Chain Alignment in NIL 2.1. Polymer Chain Alignment in Nanoimprinted Polymer Micro- and Nanostructures The formation of polymer patterns during nanoimprint is accomplished by pressure-induced polymer melt flow. It is widely known in polymer processing industry that polymer flow will lead to polymer chain orientation along the flow direction.15 Since nanoimprint involves polymer melt flow, polymer chains will be stretched and aligned along the flow direction and the chain orientation can be “frozen” into the polymer micro- and nanostructures during the cooldown stage (Fig. 2(a)) in thermal nanoimprint. The properties of the polymers often originate from inter-chain molecular interaction, thus chain orientation can significantly modify polymer properties because stretching and aligning polymer chains can greatly promote inter-chain interactions. For example, the mechanical strength of polymer films or ropes is significantly enhanced by stretching. The electrical and photophysical properties of conjugated polymers are highly dependent on film morphology and chain configuration.16 Chain orientation in polymer can result in novel optical properties, e.g. optically isotropic amorphous polymer film can be converted into birefringent films. In thermal nanoimprint, the direction and the extent of chain orientation depend on the flow pattern of the polymer melt. During nanoimprint, the polymer melt is forced to flow under a pressure and the flow stops at the point when mold cavities are completely filled. For a simple grating structure, the polymer flow field can be represented by the schematic shown in Fig. 2(b). The majority of chain orientation occurs under the mold protrusion area due to a large extensional flow of the polymer melt at that region. Optical birefringence is observed in all polymer micro- and nanostructures patterned by thermal nanoimprint (unpublished results). An example of optical birefringence in nanoimprinted semicrystalline polyvinylidene fluoride (PVDF) is shown in Fig. 3. Figure 3(a) is the optical microscope image of the PVDF structure nanoimprinted at 200°C. Figure 3(b) shows the same structure viewed between two-crossed
321
Nanoimprint Technology
polarizers. One of the polarization directions of the polarizer is parallel to the grating lines. When the sample is rotated by 45°, the dark lines in Fig. 3(b) turn into bright ones (Fig. 3(c)). This clearly indicates the PVDF patterns are optically birefringent and there is a refractive index difference between the parallel and perpendicular directions of the grating lines due to polymer chain orientation. (b)
(a)
Fig. 2. (a) A schematic of chain extension and orientation caused by polymer flow; (b) A schematic of polymer melt flow pattern during nanoimprint and the resultant optical axes of birefringence. (Reprinted with permission from Ref. 25, Copyright 2008 AIP)
(a)
(b)
(c) Fig. 3. (a) PVDF structures patterned by nanoimprint; (b) Same PVDF pattern viewed under polarizing microscope; (c) Sample rotated by 45° with respect to (b).
322
Cheng and Guo
The nature of the polymer materials can affect the chain orientation in polymer micro- and nanostructures. The preliminary results indicate poly(methyl methacrylate) (PMMA) microstructures have a smaller birefringence compared with PVDF structures. This may be explained by a more flexible chain backbone for PMMA than PVDF. Nanoimprinting PMMA gratings at different temperatures while keeping all other processing parameters constant reveals that nanoimprinted PMMA gratings at 120°C exhibit higher level of optical birefringence than those nanoimprinted at 175°C (unpublished result). At higher nanoimprint temperature, the PMMA chains have significant mobility. The random motion of the polymer chains at high temperature can destroy the chain alignment obtained during pattern formation. Rigid backbones, such as PVDF and conjugated polymers, have limited mobility and tend to have a much greater inter-chain interaction once stretched and aligned. Such interaction can lead to the stabilization of the chain orientation after nanoimprint. Depending on specific applications, chain orientation in polymer micro- and nanostructures can be either detrimental or beneficial. In polymer-based integrated optical devices, waveguides from optically isotropic polymer such as PMMA are often made by nanoimprint or hot embossing. The optical birefringence in this situation may be unwanted because it will affect the optical modes propagating in the waveguide. However, in most cases, chain alignment is highly desired for improving material performance, particularly in functional polymers such as piezoelectric, nonlinear optical, semiconducting and magnetic polymers. Since the polymer chain orientation is directly related to the polymer melt flow field in nanoimprint, it is thus possible to achieve desired chain orientation by designing specific mold patterns and tweaking nanoimprint processing parameters. This observation of chain alignment after nanoimprint may prove that nanoimprint can be used as not only just a patterning tool, but also a material processing tool that can improve material properties for functional polymers. Because the polymer flow field distribution in thermal nanoimprint is affected by many nanoimprint parameters, such as mold pattern size, shape, density and height,17, 18 polymer film thickness,17 nanoimprint temperature,19, 20 processing time, and polymer molecular weight,21, 22 the actual chain orientation in
Nanoimprint Technology
323
polymer micro- and nanostructures can be complicated. A systematic study is needed to investigate how each factor affects the chain orientation before it is possible to develop nanoimprint recipes to achieve maximum level of chain alignment along desired directions for device applications. 2.2. Improving the Performance of Polymer Electronics by Nanoimprint-Induced Chain Orientation When nanoimprinted polymer microstructures are used as a component of the final device, it is important to investigate the material property change in those microstructures after nanoimprint. The electrical and optical properties of polymer semiconductors are known to depend sensitively on the physical conformation of the polymer chains and interchain interactions in thin films16. The folding of polymer chains can change the average size of the conjugation segment on chain backbone. The way how the chains are packed inside the films affects the formation and the nature of various interchain ground and excited states that dominate the material properties, such as excimers, aggregates and polaron pairs,23, 24 Since nanoimprint can induce chain stretching and alignment in polymer micro- and nanostructures, it provides a simple yet versatile means to fine-tune the electrical and optical properties by manipulating the chain folding and packing in conjugated polymer films. The polarizing microscope images of regioregular poly(3hexylthiophene) (P3HT) structures patterned by nanoimprint are shown in Fig. 4. Figure 4(a) is the optical microscope image. Figures 4(b) and (c) are viewed between two crossed-polarizers, where the sample is rotated by 45° in (c) with respect to (b). Figures 4(d) and (e) are P3HT gratings nanoimprinted at 150°C and viewed between crossed-polarizers. The angles shown in Fig. 4(d) and (e) correspond to the angles between the grating lines and the polarization direction of the bottom polarizer. Dark images are observed when the grating lines are either parallel (0°) or perpendicular (90°) to the polarization direction of the incident light. As the P3HT gratings are rotated, images emerge gradually brighter. Maximum brightness is observed at 45°. The observations clearly demonstrate the optical axes of the birefringence in nanoimprinted P3HT
324
Cheng and Guo
gratings are parallel and perpendicular to the grating lines. This is consistent with the chain stretching direction in thermal nanoimprint (Fig. 2(b)) and indicates that the polymer melt flow is the cause of the observed optical birefringence. The conclusion that chain orientation originates from melt flow can be corroborated by the observation in Fig. 4(e), which shows the edge of the grating area after nanoimprint. The area without grating structures, which underwent exactly the same thermal history as the grating area, does not exhibit optical birefringence. The amorphous domains and the randomly oriented microcrystallines in regioregular P3HT make the film be optically isotropic, which appear as the dark area in Fig. 4(e). a
b
c
d
e
Fig. 4. (a) Optical image of nanoimprinted P3HT microstructures; (b) Same structure between two crossed polarizers; (c) Sample rotated by 45° with respect to (b). Largest circle is 50 µm in diameter; (d) and (e) correspond to 700 nm period P3HT grating nanoimprinted at 150°C. Numbers show the angle between the grating lines and the bottom polarization direction. Gray area in (e) has no grating pattern.
The interchain species formed after nanoimprint facilitate carrier transport in conjugated polymer films, thus enhance the performance of organic thin-film transistors (OTFT) based on those chain-aligned P3HT gratings. The drain current is significantly enhanced compared to a control device, which undergoes exactly the same processing and thermal
Nanoimprint Technology
325
history but without nanoimprint-induced chain orientation. Hole mobilities in the control device and the grating OTFT are 1.35×10-3 cm2/Vs and 1.62×10-2 cm2/Vs, respectively.25 Since the chain orientation is universal in all nanoimprinted polymer micro- and nanostructures, the potential impact of nanoimprint-induced chain orientation on organic electronics are far-reaching, including all conjugated polymer materials and devices such as polymer-based light-emitting diodes, thin-film transistors, solar cells and sensors. 3. Reversal Nanoimprint Lithography 3.1. Principles of Reversal Nanoimprint Many variations of nanoimprint have been developed in the last decade to expand the application range of the conventional nanoimprint technique. Reversal nanoimprint is one of such techniques.26-29 In reversal nanoimprint, the resist film is coated on top of the mold and then transfer-bonded to a substrate. However, placing a polymer film on top of a surfactant-coated mold is not a trivial task because conventional techniques such as spin-coating are seldom successful. Because the mold surface is typically coated with surfactant perfluorodecyltrichlorosilane (FDTS), which leaves the mold surface with very low surface energy, polymer solutions dewet and bead up on mold surface instead of spreading into a liquid film. When spinning starts, the polymer solution just roll off the mold surface and no polymer film can be coated on mold surface. In order to overcome this issue, several techniques can be used. The first one is to choose a surfactant and solvent combination such that the dewetting of polymer solution on mold surface can be prevented. The second method is to use highly viscous polymer solution because high viscosity can slow down the rate of dewetting and make it possible to leave a film on mold surface after spin-coating. Finally, it is possible to transfer the polymer film onto the mold surface in a normal thermal imprinting process by treating the substrate with a surfactant such as octadecyltrichlorosilane (OTS). OTS and other intermediate surface energy coating (20-40 dyn/cm) can reduce the adhesion between the
326
Cheng and Guo
polymer film and the substrate and also allow many solvents to be used for spin-coating a polymer film on a substrate. Although the mold surface is coated with FDTS (12 dyn/cm), the larger mold-polymer interface due to mold surface patterns can still yield greater moldpolymer adhesion than the polymer-substrate adhesion, which allows the polymer film to be detached from the substrate and remain on mold surface after nanoimprint. The yield of this process can be further improved by a twisting action during mold-substrate separation as shown in Fig. 5.
Fig. 5. A schematic of transferring polymer film onto mold surface for reversal nanoimprint by twist-demolding in nanoimprint.
3.2. Residual Layer Removal in Reversal Nanoimprint In traditional nanoimprint, there is always a residual layer that needs to be removed by a RIE process. Nanoimprint is a fast process that can be accomplished within a few minutes, which is critical for high throughput fabrication of nanostructures. The oxygen RIE step involves vacuum processing that can significantly increase the overall processing time and greatly lower nanoimprint throughput. Secondly and most importantly, oxygen RIE can degrade or even damage the polymer materials by breaking the chemical bonds in polymer chains with highly energetic oxygen plasma. This eliminates the possibility to create isolated micro- or nanostructures in functional polymers by nanoimprint for device and system applications, particularly for integrated polymer circuits where good isolations between individual devices are needed.
Nanoimprint Technology
327
Previously developed combined-nanoimprint-photolithography technique for residual layer removal is only suitable for UV-curable materials.30 It is thus highly desirable to develop a new technique to remove the residual layers without RIE for thermoplastic polymers.
Fig. 6. Schematics of (a) reversal nanoimprint to place a polymer film on a mold surface, and (b) solvent developing method, and (c) dewetting method for residual layer removal.
Once the polymer film is transferred to the mold surface in reversal nanoimprint, the residual layer, which is sitting on top of the mold protrusions, is exposed to the outmost layer and can be easily removed by two techniques (Fig. 6(b) and (c)). In Fig. 6(b), the residual layer is removed by soaking the mold in a solvent for a fixed amount of time. The soaking time can be determined based on the residual layer thickness and the dissolution rate of the polymer film in solvent. Since the mold surface is coated with low energy surfactant, dewetting usually occurs when heating the thin polymer film to above its Tg due to the incompatibility of the polymer film and the mold surface.31, 32 In Fig. 6(c), dewetting can cause the residual layer to rupture and the polymer melt of the residual layer will reflow to the trench areas of the mold. Once the residual layers are removed, the polymer structure can be transferred to the target substrate, which will yield polymer micro- and nanostructures without residual layers. Both techniques have been successfully applied to pattern isolated polymer micro- and nanostructures. Figure 7(a) shows
328
Cheng and Guo
5 µm poly[2-methoxy-5-(2'-ethyl-hexyloxy)-1,4-phenylene vinylene] (MEH-PPV) lines obtained by the solvent developing method and Fig. 7(b)
shows dewetted PMMA on a mold surface. The PMMA surface bulges out because the residual layer on mold protrusions are moved to the trench area by the dewetting process. The PMMA layer on the mold can be used to build 3D polymer microstructures as demonstrated in next section. In both cases, the dimension of the polymer structures are determined by the mold pattern size. It is possible to use these approaches to nondestructively pattern functional polymers into microand nanostructures for both fundamental research on the structureproperty relations of polymers at the nanoscale and practical development where the nanoscale functional polymers serve as the active components of the devices.
(a)
(b)
Fig. 7. (a) Optical microscope image of isolated MEH-PPV lines; (b) SEM image of a mold surface with residual layer removed by PMMA dewetting.
3.3. Building 3D Polymer Nanostructures 3D polymer microstructures have many applications in electronics, photonics and bioengineering. There is an increasing trend of using polymer materials and 3D structures in various MEMS applications due to their light weight, low cost, easy processing, and multiple functionalities. For example, 3D polymer structures can be used to build multilevel microfluidic channels, which can integrate more components into a lab-on-a-chip system to increase its functionality. 3D polymer
329
Nanoimprint Technology
structures on the order of a few microns to sub-microns can be directly used as diffractive optical components such as 3D photonic crystals.33 3D polymer structures can also be used as a sacrificial template to fabricate advanced inorganic 3D microstructures.34, 35 Polymer 3D scaffolds are also popular in biomedical applications such as tissue engineering. One of the biggest applications of the reversal nanoimprint is to build 3D polymer structures in a layer-by-layer fashion,36 as shown in Fig. 8. The bottom layer of the 3D structure can be created by a normal nanoimprint step (Fig. 8(a)). Additional layers can be added to this bottom layer by a reversal nanoimprint (Fig. 8(b)) and transfer bonding (Fig. 8(c)). Top layers can be successively added by repeating the process to build up multilayer structures (Fig. 8(d)). (a)
(b)
(c)
(d)
Fig. 8. Schematic of patterning process for 3D multilayer polymer structures. (a) Step 1: conventional nanoimprint lithography; (b) Step 2: layer transfer from OTS coated substrate to mold using twist-demolding; (c) Step 3: transfer-bonding; (d) Step 4: adding more polymer layers by repeating step 2 and step3. (Reprinted with permission from Ref. 38, Copyright 2008 AIP)
The most critical issue for 3D polymer structure fabrication is the adhesion between the polymer layers. In order to successfully transfer the polymer layer from the mold to an existent polymer layer on the substrate, the adhesion between the polymer layers must be greater than the adhesion between the mold and the polymer. This is difficult to
330
Cheng and Guo
achieve for most polymer materials, particularly thermoplastic polymers. To promote inter-polymer adhesion, several techniques can be used. The first one involves using a thin adhesion layer as shown in Fig. 9(a). Diluted SU-8 solution can be used for this purpose. Since the Tg of SU-8 is around 64°C,37 very good bonding between two PMMA layers can be achieved at 80°C because at this temperature SU-8 melts and forms good contact with the two PMMA layers to be bonded. Multilayer structures can be easily achieved by repeating the bonding process as shown in Fig. 9(b).38 The advantage of using the adhesion layer is that it can bond almost any combination of materials, such as polymer-polymer, polymerdielectric and polymer-metal. It provides a high-yield means to integrate different materials to form advanced 3D composite structures for many applications.
(a)
(b)
Fig. 9. (a) A schematic of fabricating multilayer polymer structures by bonding with a thin adhesive layer; (b) Three-layer PMMA microstructure bonded by a thin SU-8 adhesive layer. (Reprinted with permission from Ref. 38, Copyright 2008 AIP)
Although using adhesive layer such as SU-8 can greatly promote polymer bonding, it is still desirable to have the ability to bond the two polymer layers without any adhesion layer. Direct thermal bonding of thermoplastic layers can be performed at a temperature slightly below the Tg of the polymer materials. This takes advantage of a fact that polymer
331
Nanoimprint Technology
Tg at the surface is usually 5-10°C lower than that of the bulk.39, 40 Bonding at a temperature slightly below the bulk Tg allows the formation of a good interfacial mixing layer due to surface melting, while the bulk of the bottom polymer layer is still below Tg to preserve their pattern integrity. The successful fabrication of multilayer PMMA structures (Fig. 10) was achieved by bonding at 100°C and 3×106 Pa. Two-layer and three-layer PMMA gratings are shown in Fig. 10(a) and (b), respectively. This method can be extended to other thermoplastic polymers including many functional polymers, thus provides a route to create 3D multilayer structures in functional polymers for novel applications.
(a)
(b)
Fig. 10. SEM micrographs of 3D multilayer PMMA structures using direct thermal bonding. (a) Two-layered 10 µm period grating; (d) Three-layered 700 nm period grating. (Reprinted with permission from Ref. 38, Copyright 2008 AIP)
Due to the presence of the residual layer in reversal nanoimprint, the 3D polymer structures demonstrated so far contains structures within the layer and the structure between different layers are not connected. This is good for some applications such as advanced multilayer microfluidic devices where separation of fluidics between different layers is desired. However, in many applications it is preferred to have connectivity between layers in 3D structures.41 It is possible to combine the direct bonding method with the residual layer removal techniques to achieved 3D polymer structures with connectivity between layers. Before adding a layer by transfer bonding, the residual layer of the polymer film on the mold surface is removed by aforementioned solvent developing or dewetting methods (Fig. 11(a)). The polymer film without residual layer
332
Cheng and Guo
is then bonded to the existent polymer layers on the substrate, creating a multilayer structure without residual layers. The process can be repeated for more layers (Fig. 11(b)) and an example of three-layer PMMA structure with connectivity throughout the layers is shown in Fig. 11(c). Although here only grating structures are demonstrated, it is possible to vary polymer materials, pattern sizes and shapes in each layer based on the requirements of specific applications. Such flexibility provides almost unlimited combinations of scenarios for the 3D structures.
(a)
(b)
(c)
Fig. 11. Schematics of 3D polymer patterning process: (a) transfer-bonding a polymer grating without residual layer to another polymer grating, and (b) adding more polymer layers by repeating the transfer-bonding process. (c) An SEM image of 3D PMMA structures with 10 µm period gratings.
3.4. Process Yield of Reversal Nanoimprint For a fabrication technique to be practically useful, it must have good process yield. In building polymer 3D structures by reversal nanoimprint and transfer-bonding, the most important issue is the competition between the mold-polymer adhesion and the polymerpolymer adhesion. The mold-polymer adhesion depends on the type and the quality of the mold surfactant coating, the compatibility of the polymer and the surfactant, and the pattern depth and density. The total mold-polymer interfacial area, thus adhesion force, will increase with increasing mold pattern depth and density as shown in Fig. 12.
Nanoimprint Technology
333
Consequently, it is more difficult to transfer-bond high-density and high aspect-ratio nanostructures. The pattern in each polymer layer also has an impact on process yield. When transfer-bonding a polymer layer onto a bottom polymer grating instead of a uniform polymer film, the total contact area between the two polymer layers become much smaller. For higher process yield, it is preferred to have a larger ratio of the protrusion area to the recessed area in the bottom polymer layer.
Fig. 12. Larger mold-polymer adhesion for mold patterns of higher density. (Reprinted with permission from Ref. 38, Copyright 2008 AIP)
The polymer-polymer adhesion is an even more complicated issue and itself is a subject of intensive research with significant industrial impact.42, 43 In general, polymer-polymer adhesion strongly depends on the nature of the polymers and their interfacial interactions. For different polymer materials, interfacial interactions can be categorized into several types: van der Waals force, hydrogen bond, electrostatic attraction and covalent chemical bonds. Strong interactions, such as covalent bonding, can lead to excellent polymer bonding. Very high transfer-bonding yield can be achieved because the polymer-polymer adhesion is much greater than the mold-polymer adhesion. The opposite is true for weak interfacial interactions, such as van der Waals force, which often leads to failed transfer-bonding. Chemical bonding can usually form at polymer interface if the polymers are briefly treated by oxygen plasma before bonding. Forming intermixed layers at the interface by heating the polymers to near their Tg can also greatly enhance polymer-polymer adhesion. Finally, the mechanical strength, which is the magnitude of the cohesive force that holds the polymer chains together, also has an impact
334
Cheng and Guo
on the transfer-bonding yield. If both mold-polymer and polymerpolymer adhesion forces are greater than the mechanical strength of the polymer materials, it is inevitable that the polymer layer will be torn off – part of the polymer layer is transferred to the substrate and but the rest remains on mold. This may become a significant issue when bonding very thin polymer layers because of their weak mechanical strength. Due to the complexity of adhesion forces, typically it requires a case by case study to improve process yield based on the polymer materials used and the details of the patterns in each layer. In the situation where an adhesive layer, such as a thin SU-8 film, can be used, the process yield of transfer bonding can be very high (Fig. 13). Larger patterns (10 µm period grating) has higher bonding yield compared with finer patterns (700 nm period grating). For finer patterns, the increased sidewall interfacial contact between the mold and the polymer greatly enhances the mold-polymer adhesion, thus making it more difficult to achieve high yield. Though the yield for direct thermal bonding is lower, it is expected that further optimization of processing conditions such as fine-tuning of bonding temperature and pressure and improving pressure uniformity can improve the yield to a level high enough for practical applications. 100%
80%
60%
40%
20%
0%
SU-8(6%)
SU-8(1%)
Direct bonding
Fig. 13. Transfer-bonding yield of 3D PMMA structures with different bonding schemes and pattern sizes: 10 µm period grating (dash line) and 700 nm period grating (solid line). (Reprinted with permission from Ref. 38, Copyright 2008 AIP)
Nanoimprint Technology
335
4. Recent Applications of NIL As previously mentioned, NIL is not limited in producing lithographic patterns in resist materials, where the patterned polymer template is often used as a sacrificial material for further processes. A fascinating application of NIL is to create surface nanostructures in polymers with unique functionalities for device applications. In what follows, we describe two such applications in renewable energy area, one in conjugated polymer based organic solar cells and the second in micro fuel cells. 4.1. Organic Solar Cells with Imprinted Nanoscale Morphology Organic solar cells (OSC) have the merits of low cost, easy fabrication and therefore have been considered as a promising energy conversion platform for clean and carbon-neutral energy production. Conjugated polymers and organic semiconductors can possibly provide a practical solution due to their easy processability, light weight, capability with flexible substrates, and tunable optoelectronic properties of conjugated polymers.44 The basic structure of organic solar cells is composed of ITO anode, energy harvesting organic layer, electron acceptor layer, and cathode. Currently, the overall power conversion efficiency of a conjugated polymer solar cells is still substantially lower than that of silicon-based solar cells.45, 46 One of the reasons responsible for the limited attainable efficiency is that excited electrons in a conjugated polymer are strongly bound with holes and form the so-called excitons. An effective approach to “free” the charges from the excitons is by using two types of molecules with an energy offset between their LUMO levels. To produce photocurrent, the neutral excitons must diffuse to the interface between the two types of molecules where they dissociate as an electron in one material phase with lower LUMO level (the acceptor) and a hole in the other (the donor). But a bottleneck exists because the exciton diffusion length is only 10~20 nm in most of the organic materials.47, 48 In order for the electron and the hole to be separated within the lifetime of the exciton, the most effective approach so far is to blend the electron acceptors with the donor
336
Cheng and Guo
20 nm
(considering the excition diffusion length of ~10 nm)
Cathode
Donor
200 nm
Acceptor
For efficient energy harvesting
Anode
Fig. 14. An ideal organic photovoltaic cell structure with interdigitated interface between donor and acceptor layers.
conjugated polymer matrix and rely on phase separation to form donor and acceptor domains. This is referred to as bulk heterojunction solar cells. The incorporation of acceptor domains is essential for exciton charge separation, but unavoidably reduces the crystallinity of the donor polymer and consequently the hole mobility.49 Moreover the random distribution of donor and acceptor materials in such structures can lead to charge trapping at bottlenecks and cul-de-sacs in the conducting pathways to the electrodes. To address this bottleneck a promising approach is to create the ordered nanoscale interface between the donor and acceptor layers (Fig. 14), forming an interdigitated structure with domain size about twice that of the exciton diffusion length. Clearly the required domain size is within the resolution of NIL. As a first step toward achieving the ideal solar cell structure depicted in Fig. 14, Kim et al. have successfully nanoimprinted an active HO O
(a)
O
soluble
O
insoluble
(b)
S S
S S
n
TDPTD ITO coated PET cleaning
TDPTD Spincasting
PCBM
Demolding
n
PCBM Spincasting
Si mold
Imprinting with Si mold
Aluminum
Al evaporation
Fig. 15. (a) Device fabrication procedures. (b) Photograph of fabricated OPV on flexible substrate.
337
Nanoimprint Technology
conjugated polymer to produce flexible organic solar cells with controlled nanostructures.50 As pointed out earlier, the ordered continuous heterojunction having a large surface area allows efficient charge separation and the vertically oriented domains provides a line-ofsight pathway for rapid charge transport toward the electrodes while minimizing the probability of charge recombinations. Our preliminary results demonstrated that thermal nanoimprinting of conjugated polymers performed under vacuum condition preserves the optoelectronic properties of the conjugated polymer and that the energy conversion efficiency of the fabricated solar cell can be systematically enhanced by controlling the interfacial nanostructures in conjugated polymer based solar cells.
(a)
(b)
(c)
(d)
Fig. 16. SEM images. (a) Nanoimprinted TDPTD having 700 nm periodicity. (b) Nanoimprinted TDPTD having 510 nm periodicity. (c) After PCBM coating on the imprinted TDPTD of (b). (d) After Al coating on the PCBM of C. Scale bars shown are 2 µm. (Reprinted with permission from Ref. 50, Copyright 2007 AIP)
The overall device fabrication process is illustrated in Fig. 15(a). A thermally deprotectable polythiophene derivative (TDPTD) is used as electron donor material and [6,6]-phenyl C61-butyric acid methyl ester (PCBM) as electron acceptor, Al cathode and ITO-coated PET as a flexible substrate. The pristine form of this polymer is solution
338
Cheng and Guo
processable, and was spincast on the ITO-coated PET. The long branched alkyl side chain is removed by heating during the nanoimprint process at 180°C and the polymer becomes crosslinked. Therefore the imprinted structures will not be affected by the solvent in the PCBM to be spincoated over the imprinted surface.51 TDPTD was imprinted with two different molds with grating structures of 510 nm period and 700 nm period and 200 nm depth. The grating patterns of the molds were effectively transferred to TDPTD layer, as evidenced by the multicolor diffraction of light by the imprinted grating (Fig. 15(b)) and the cross sectional SEM images of imprinted structure (Fig. 16(a-b)). After imprinting, PCBM dissolved in chlorobenzene was spincast to conformally coat the imprinted TDPTD surface (Fig. 16(c)). Finally Al was deposited on PCBM layer to complete the device fabrication (Fig. 16(d)). The fabricated cells were characterized by using Oriel solar simulator at 56 mW/cm2 illumination intensity, which is about half of the sun light intensity. Flexible solar cells with three different structures were compared (control cells with a flat interface and nanostructured cells with 510 nm and 700 nm periodicity). The current density vs. voltage characteristics (J-V curves) clearly show that as the interface area between the donor and the acceptor increases short circuit current density increases linearly (a)
(b)
Fig. 17. (a) The control cell(■), the imprinted cells having the 700 nm(▲) and 510 nm( ) period under illuminated condition (56 mW/cm2 intensity). (b) The effect of the vertically oriented interface areas between TDPTD and PCBM is plotted. Solid line: linear fit of data, Dashed line: 1:1 line of Jsc/Jsco and A/Ao. (Reprinted with permission from Ref. 50, Copyright 2007 AIP)
◆
Nanoimprint Technology
339
(Fig. 17(a)). The effect of interface area on the short circuit current is plotted in Fig. 17(b). This result demonstrates that the interface area is directly related to amount of charge generation. Interestingly, the ratio of the increased short circuit current to the surface area gain (the slope of the Fig. 17(b) graph) is 1.39, larger than the 1:1 correlation from considering only the interface area between the donor and the acceptor layer. As described earlier the vertically oriented interfaces can provide a straight pathway for efficient charge transport once the excitons dissociate to electrons and holes at the interfaces. Due to this synergistic effects the power conversion efficiency of the nanostructured cells with 510 nm and 700 nm period are 3.20 and 2.64 times larger than that of the control cell, respectively, even though the actual interface area increases 1.75 and 1.5 times only. 4.2. Nanoimprinting Nafion® Film for Micro Fuel Cell Applications Fuel cells are an attractive alternative to batteries for portable electronic devices. Micro-fuel cells hold promise for being highly efficient with low cost. Several approaches have been reported for the design of micro-fuel cells using micromachining techniques. Most of these devices include only micromachined fuel cell components
Fig. 18. SEM cross-section of Nafion® nanostructured thin film; the insert is an enlarged SEM image of the nanograting imprinted in Nafion® thin film.
340
Cheng and Guo
combined with conventional macro membrane electrode assemblies (MEAs).52, 53 Previous work has demonstrated the possibility of completely microfabricated micro-fuel cells.54, 55 Towards improving the performance of electrode thin films, we used NIL technique to fabricate microelectrodes with increased surface areas for use in a micro fuel cell devices.56 The electrodes were formed using electron beam evaporation of Ti and Pt and was found to be electrochemically active. Moreover, we found that the spin casted Nafion® films and bulk Nafion® 117, commonly used proton exchange membranes, can be imprinted easily and remain active for operation in a fuel cell. To fabricate the micro fuel cell, a NIL mold with 700 nm period grating structure was used to hot emboss the Nafion film. The embossing of Nafion 117 is a fairly simple process. The molds were pressed into the substrates at 900 psi and 135°C for 5 min. Figure 18 demonstrates the results of imprinting nanostructures directly onto Nafion thin films. The features possess 700 nm period and exhibits consistent color diffraction. The surface edges of the embossed film appeared to be rounded. This may be due to the fact that the films were embossed immediately after casting without curing. The direct embossing of Nafion has the advantage of controlled surface modification without chemical contamination. Previously it was observed that chemicals used in modern micromachining processes (e.g. photoresist, photoresist developer, solvent, etc) can negatively impact the performance of an MEA. In our process a shadow mask was created to selectively deposit Pt over the imprinted nanostructured features, which eliminates the lift-off and post chemical treatment. The membrane was fabricated into an MEA and the performance was compared to an MEA prepared using conventional materials. The polarization curves are illustrated in Fig. 19. Although the peak power density of the nanoimprinted MEA was 123 mW/cm2, which was lower than that for the conventionally prepared MEA (410 mW/cm2), the Pt utilization for the former was 15,375 mW/mg Pt compared to 820 mW/mg Pt for the conventional electrode. These values were determined by dividing the peak power density by the Pt loadings for the anode (conventional MEA, 0.5 mg/cm2; MEA with nanoimprinted electrode, 8 µg/cm2). The added areas from the Pt on the sidewalls of the
Nanoimprint Technology
341
nanostructures could contribute to increased performance over a planar surface. For instance for the structure used in this study, the available added surface area is twice the amount of the planar surface.
Fig. 19. Polarization curves of standard MEA and nanoimprinted MEA. (Reprinted with permission from Ref. 54, Copyright 2007 Elsevier)
The embossing of nanostructures onto Nafion thin films holds promise for a variety of new micro fuel cell designs. In addition, micro fluidic devices that exploit the proton selectivity of Nafion for reactions and/or separations could be possible. We are presently investigating the viability of these options. 5. Roll-to-Roll Nanoimprint Lithography (R2RNIL) Though there has been steady progress in NIL technology in recent years, the current process and throughput (~several min or longer per Si wafer) is still far from meeting the demands of many practical applications, especially in the area of organic electronics, large area photonics, and biotechnologies. To meet these demands, a faster and more economical process is needed. In this regard, a continuous roll-toroll nanoimprint technique can provide a solution for high-speed
342
Cheng and Guo
large-area nanoscale patterning with greatly improved throughput; furthermore, it can overcome the challenges faced by conventional NIL in requiring large force, maintaining pressure uniformity and successful demolding in large area printing. Huge contact area between the mold surface and the imprinted nanostructures could produce significant adhesion force, making the mold-sample separation step very difficult or even impossible to achieve without damaging the substrate. In thermal NIL process, if the mold and substrate are made of materials with different thermal expansion coefficients such as Si mold and polymer substrate, stress can build up during a thermal cycle with such a magnitude that even destroys the Si mold during mold separation. R2RNIL provides a unique solution to these challenges encountered in the conventional wafer-level NIL process, because imprinting in the R2RNIL process takes place in a narrow region transverse to the web moving direction, thus requires much smaller force to replicate the patterns. Also since the mold used in R2RNIL is in the form of a roller, the mold-sample separation proceeds in a “peeling” fashion, which requires much less force and reduces the probability of defect generation.
Fig. 20. Schematic of a continuous roll-to-roll nanoimprinting setup, consisting of (a) a coating (a) and an imprinting module (b), followed by a metal deposition process (c). (Reprinted with permission from Ref. 57, Copyright 2008 Wiley-VCH Verlag)
Nanoimprint Technology
343
Figure 20 shows the overall configuration of a continuous R2RNIL process,57 which consists of three separate processing steps: 1) coating process, 2) imprinting and separating process, and 3) any of the subsequent processes. As an example, the last step in this schematic represents a continuous metal coating process for making structures such as metal wire-grid polarizers58 or transparent wire-grid electrode.59, 60 True roll-to-roll nanoimprinting has been a challenge to the community because it requires a complete set of solutions to address a number of interrelated material issues. Firstly, a special roller mold is required for continuous roll-to-roll imprinting of nanostructures. One approach is to use a NIL mold that is sufficiently flexible and can be wrapped on the roller surface. Meanwhile the mold should also have sufficient modulus and strength to be able to imprint other materials. Secondly, liquid resists should have good coating property and low viscosity to ensure fast imprinting; and they should be cured rapidly to maintain high-throughput and also should have minimal shrinkage. Therefore, conventional resist materials dissolved in solvents and require additional baking process could make the R2RNIL process more difficult to control and the imprinting prone to defect generation. The flexible mold can be made from metal or polymer materials. For example, a thin, electroplated Ni shim can be a durable mold for high volume manufacturing. Polymer molds, on the other hand, are much easier to replicate from original NIL masters. We introduce two types of flexible polymer molds that can be used for R2RNIL application. The first material is a commercially available fluoropolymer, ethylenetetrafluoroethylene (ETFE). ETFE has high modulus (~1.2 GPa) at room temperature but can be softened at elevated temperature. Therefore a ETFE mold can be easily replicated from an original Si mold by a thermal NIL process at 200°C. Moreover, the exceptional anti-sticking property of ETFE (surface energy of 15.6 dyn/cm, cf. PDMS ~19.6 dyn/cm) makes it easy to de-mold after imprinting without any mold surface treatment and without deterioration in surface properties over long imprinting cycles. The second material is a new fluorinated photocurable silsesquioxane (SSQ) resin,61 which possess outstanding properties for applications as nanoimprint material for nanoscale patterning. With an appropriate viscosity, this resin can be easily
344
Cheng and Guo
imprinted by an original Si NIL master using a low pressure nanoimprinting process. The resin has a sufficient modulus in its cured state, which makes it suitable for nanoimprinting other polymeric materials. Due to the high thermal stability and UV transparency of SSQ materials, such a stamp can be used for both UV and thermal nanoimprinting. Furthermore, the fluoroalkyl groups contained in the silsesquioxane resin provide the low surface energy necessary for easy demolding after nanoimprinting. Details on the material composition and the SSQ mold fabrication can be found in Ref. 61. Two types of liquid resists were exploited for patterning in the R2RNIL process. For thermal R2RNIL, a modified liquid PDMS resist with fast thermal-curing property was used.62 For higher speed R2RNIL, a low viscosity UV curable epoxysilicone63 was used as imprint resist. Also two types of coating methods, reverse and forward web coating, were used to provide a continuous and uniform resist coating on the plastic web from a resist container by using a coating roller. The resist thickness is controlled by using a doctor blade. The final thickness of the resist material is determined by the pressure in the R2RNIL process. The Imprint module is composed of an imprint roller, two backup rollers, a release roller and a curing section. The mold, in this case a flexible sheet of ETFE with surface relief patterns replicated from an original Si master, is wrapped around a stainless steel roller. In the demonstrated roller imprinting process the pressure is applied by means of web tension and forces from the back-up rollers. Moreover, the two backup rollers with spring system also guarantee non-slip motion in the rolling process, which is very important for successful pattern replication. Next, the imprinted resist precursor is cured either by convection heating or UV irradiation. Finally, PET substrate with roller-imprinted nanostructures continuously separates from the roller mold via the release roller. Figure 21 shows R2RNIL results of continuous grating pattern with 300 nm linewidth and 700 nm period imprinted in (a) the thermally cured PDMS and (b) UV cured epoxysilicone on PET substrate, respectively. Scanning electron microscopy (SEM) shows that the UV cured resist pattern has higher quality than the thermal-cured PDMS, likely due to the lower viscosity of the material that facilitates the fast filling of the mold
Nanoimprint Technology
345
cavity. Printing speed can be adjusted depending on the period of grating pattern and its aspect ratio. The fast curing of the resist material enabled a web speed of ~3.5 ft/min. Figure 21(c) show UV R2RNIL result of a 570 mm long, 700 nm period grating structure created on PET substrate with bright light diffraction.
Fig. 21. SEM micrographs of 700 nm period, 300 nm line width gratings imprinted on PET flexible substrate by using thermal-cured PDMS (a), and UV cured epoxysilicone (b). Photograph showing bright light diffraction from a section of a 570 mm PET strip imprinted with 700 nm grating pattern (c).
High aspect ratio or denser gratings results in larger contact area and therefore stronger adhesion force between the EFTE mold and the imprinted resist pattern. In order to produce these patterns successfully, the PET surface is plasma activated and chemically treated with adhesion promoter to increase its adhesion to the cured epoxysilicone. Figures 22(a) and (b) show the 700 nm period grating with aspect ratio of 5.4 on the original Si mold and the epoxysilicone grating pattern replicated from the ETFE mold. Very faithful pattern transfer can be observed, even down to the fine details at the bottom of the trenches. Figures 22(c) and (d) show
346
Cheng and Guo
the replicated 70 nm linewidth grating in epoxysilicone resist having 200 nm and 100 nm period, respectively. Such a high-speed nanoimprint capability could lead to many interesting applications of large area polymer nanostructures.
Fig. 22. SEM micrographs of high aspect ratio grating structures fabricated by UV R2RNIL: (a) the original Si mold, (b) epoxysilicone gratings replicated from the ETFE mold. (c) 200 nm period, 70 nm line width epoxysilicone grating pattern and (d) 100 nm period, 70 nm line width epoxysilicone pattern fabricated by UV R2RNIL.
In summary, the R2RNIL process maintains the high resolution characteristic of the NIL technique, but offers a speed that is at least an order of magnitude faster. It eliminates the need of substrate loading and unloading and reduces the total force required for imprinting. We anticipate that this new development could enable many practical applications of NIL, especially in the area of flat panel display, solid state lighting, solar cell, and filtration membrane, and superhydrophobic surfaces, to name an important few. 6. Conclusion In this paper we have summarized our recent efforts in nanoimprint development. We pay particular attention at expanding the capabilities of
Nanoimprint Technology
347
the conventional nanoimprint technique and exploring new applications that are enabled by this technique. In nanoimprint development, improvement in reversal nanoimprint and transfer-bonding are achieved. In reversal nanoimprint, schemes to remove residual layers without oxygen RIE are developed to improve nanoimprint throughput and enable nondestructive patterning of isolated micro- and nanostructures in functional polymers. The optimized transfer-bonding techniques open up a route toward high-yield fabrication of 3D polymer microstructures. The factors that affect the yield of transfer-bonding are discussed, which can be used as a guideline in process design. For example, if the polymer is not limited to a specific type for an application, it is advantageous to choose polymers that have strong interlayer bonding for high process yield. If the 3D multilayer structure is required to be built from a specific polymer, then the pattern shape and density and the process parameters can be fine-tuned to achieve higher yield. To improve the nanoimprint throughput, a R2RNL is developed for continuously imprinting microand nanostructures on rigid or flexible substrates. This will be particularly useful in high-speed fabrication of nanostructures over very large areas. It is even possible to combine R2RNL with reversal nanoimprint for large-scale commercial manufacturing of organic electronics in a fashion similar to gravure printing. Because nanoimprining creates surface relief structures by mechanical deformation rather than relying chemical etching, this technique is particularly suitable for creating structures in polymer materials with special functionalities that might be compromised by the chemical process. For functional polymers such as conjugated polymers, dense periodic structures greatly increase the total surface area, which can promote the performance of many devices in which actions occur on the surface or at the interface, such as chemical sensors, ordered bulk heterojunction solar cells and micro fuel cells. In additional to the applications demonstrated in this review, the new observation of chain alignment in nanoimprinted polymer micro- and nanostructures and the ability to build 3D polymer microstructures in a layer-by-layer fashion may have a far-reaching impact on miniaturized devices and systems based on functional polymers. Nondestructively patterning functional polymers while at the same time controlling chain
348
Cheng and Guo
orientation by careful design of mold pattern design and choice of process parameters will enable a wide range of new devices to utilize the superior material properties. By varying the polymer materials, the pattern sizes and shapes in each layer, 3D polymer structures can find numerous applications in photonics, sensors and bioengineering. It is even possible to integrate multiple functional polymers into one platform for all-polymer devices and systems that are low-cost and light-weight. With existent and future applications, we believe that nanoimprint can play an even greater role in many engineering fields where micro- and nanostructures are needed. The recent progress reviewed in this paper demonstrate that despite more than a decade of rapid advance in NIL, there remains many exciting and important areas for nanoimprint research and development. Acknowledgments This work has been supported in part by NSF grants ECS-0424204 and CMII 0700718, AFOSR grants F064-006-0084 and FA9550-04-10312, and NSFC grant 60528003. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
S. Y. Chou, P. R. Krauss, P. J. Renstrom, Science 272, 85 (1996). G. L. W. Cross, J. Phys. D: Appl. Phys. 39, R363 (2006). L. J. Guo, Adv. Mater. 19, 495 (2007). M. Heckele, W. K. Schomburg, J. Micromech. Microeng. 14, R1 (2004). H. Schift, J. Vac. Sci. Technol. B: Microelectron. Nanometer Struct. 26, 458 (2008). M. D. Stewart, C. G. Willson, Mrs Bulletin 30, 947 (2005). M. D. Austin, W. Zhang, H. Ge, D. Wasserman, S. A. Lyon, S. Y. Chou, Nanotechnology 16, 1058 (2005). S. Y. Chou, P. R. Krauss, W. Zhang, L. Guo, L. Zhuang, J. Vac. Sci. Technol. B 15, 2897 (1997). F. Hua, Y. Sun, A. Gaur, M. A. Meitl, L. Bilhaut, L. Rotkina, J. Wang, P. Geil, M. Shim, J. A. Rogers, A. Shim, Nano Lett. 4, 2467 (2004). M. Colburn, S. C. Johnson, M. D. Stewart, S. Damle, T. C. Bailey, B. Choi, M. Wedlake, T. B. Michaelson, S. V. Sreenivasan, J. G. Ekerdt, C. G. Willson, Emerging Lithographic Technologies III 3676, 379 (1999).
Nanoimprint Technology 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
349
H. C. Scheer, H. Schulz, Microelectron. Eng. 56, 311 (2001). L. J. Guo, J. Phys. D: Appl. Phys. 37, R123 (2004). G. L. W. Cross, J. Phys. D: Appl. Phys. 39, (2006). V. N. Truskett, M. P. C. Watts, Trends Biotechnol. 24, 312 (2006). G. Kumaraswamy, J. Macromol. Sci., Part C: Polym. Rev. 45, 375 (2005). B. J. Schwartz, Annu. Rev. Phys. Chem. 54, 141 (2003). H. D. Rowland, A. C. Sun, P. R. Schunk, W. P. King, J. Micromech. Microeng. 15, 2414 (2005). H. D. Rowland, W. P. King, A. C. Sun, P. R. Schunk, J. Vac. Sci. Technol. B 23, 2958 (2005). H. D. Rowland, W. P. King, J. Micromech. Microeng. 14, 1625 (2004). H. C. Scheer, N. Bogdanski, M. Wissen, T. Konishi, Y. Hirai, J. Vac. Sci. Technol. B 23, 2963 (2005). N. Chaix, C. Gourgon, S. Landis, C. Perret, M. Fink, F. Reuther, D. Mecerreyes, Nanotechnology 17, 4082 (2006). Y. Ding, H. W. Ro, J. F. Douglas, R. L. Jones, D. R. Hine, A. Karim, C. L. Soles, Adv. Mater. 19, 1377 (2007). W. Barford, Electronic and optical properties of conjugated polymers, Clarendon Press, Oxford (2005). M. Pope, C. E. Swenberg, Electronic processes in organic crystals and polymers, Oxford University Press, New York (1999). D. Cui, H. Li, H. Park, X. Cheng, J. Vac. Sci. Technol. B 26, 2404 (2008). X. D. Huang, L. R. Bao, X. Cheng, L. J. Guo, S. W. Pang, A. F. Yee, J. Vac. Sci. Technol. B 20, 2872 (2002). M. Nakajima, T. Yoshikawa, K. Sogo, Y. Hirai, Microelectron. Eng. 83, 876 (2006). B. Yang, S. W. Pang, J. Vac. Sci. Technol. B 24, 2984 (2006). W. Zhao, H. Y. Low, P. S. Suresh, Langmuir 22, 5520 (2006). X. Cheng, L. J. Guo, Microelectron. Eng. 71, 277 (2004). R. Seemann, M. Brinkmann, E. J. Kramer, F. F. Lange, R. Lipowsky, Proc. Nat. Acad. Sci. 102, 1848 (2005). K. Y. Suh, J. Park, H. H. Lee, J. Chem. Phys. 116, 7714 (2002). I. Divliansky, T. S. Mayer, K. S. Holliday, V. H. Crespi, Appl. Phys. Lett. 82, 1667 (2003). Q. Yan, X. S. Zhao, J. H. Teng, S. J. Chua, Langmuir 22, 7001 (2006). B. T. Holland, C. F. Blanford, A. Stein, Science 281, 538 (1998). L. R. Bao, X. Cheng, X. D. Huang, L. J. Guo, S. W. Pang, A. F. Yee, J. Vac. Sci. Technol. B 20, 2881 (2002). K. Pfeiffer, M. Fink, G. Gruetzner, G. Bleidiessel, H. Schulz, H. Scheer, Microelectron. Eng. 57-58, 381 (2001). H. Park, H. Li, X. Cheng, J. Vac. Sci. Technol. B 25, 2325 (2007).
350
Cheng and Guo
39.
J. A. Forrest, K. Dalnoki-Veress, J. R. Stevens, J. R. Dutcher, Phys. Rev. Lett. 77, 2002 (1996). Y. Yang, D. H. Liu, Y. B. Xie, L. J. Lee, D. L. Tomasko, Adv. Mater. 19, 251 (2007). H. Park, X. Cheng, submitted to Nanotechnology (2008). A. V. Pocius, D. A. Dillard, M. K. Chaudhury, Adhesion science and engineering, Elsevier, Amsterdam, Boston (2002). V. Kestelman, R. Veselovsky, R. A. Veselovsky, Adhesion of polymers, McGrawHill, New York (2002). C. J. Brabec, N. S. Sariciftci, J. C. Hummelen, Adv. Funct. Mater. 11, 15 (2001). W. Ma, C. Yang, X. Gong, K. Lee, A. J. Heeger, Adv. Funct. Mater. 15, 1617 (2005). G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 4, 864 (2005). N. S. Sariciftci, Primary photoexcitations in conjugated polymers: molecular exciton versus semiconductor band model, World Scientific Publishing, Singapore (1997). B. A. Gregg, M. C. Hanna, J. Appl. Phys. 93, 3605 (2003). K. M. Coakley, M. D. McGehee, Chem. Mater. 16, 4533 (2004). M.-S. Kim, J.-S. Kim, J. C. Cho, M. Shtein, L. J. Guo, J. Kim, Appl. Phys. Lett. 90, 123113 (2007). J. Liu, E. N. Kadnikova, Y. Liu, M. D. McGehee, J. M. J. Frechet, J. Am. Chem. Soc. 126, 9486 (2004). J. P. Meyers, H. L. Maynard, J. Power Sources 109, 76 (2002). S. C. Kelley, G. A. Deluga, W. H. Smyrl, AIChE J. 48, 1071 (2002). A. D. Taylor, L. T. Thompson, IEEE Solid State Sensors and Actuators Workshop, Hilton Head, SC (2004). J. D. Morse, A. F. Jankowski, R. T. Graff, J. P. Hayes, J. Vac. Sci. Technol. A 18, 2003 (2000). A. D. Taylor, B. D. Lucas, L. J. Guo, L. T. Thompson, J. Power Sources 171, 218 (2007). S. H. Ahn, L. J. Guo, Adv. Mater. 20, 2044 (2008). S. H. Ahn, J. S. Kim, L. J. Guo, J. Vac. Sci. Technol. B 25, 2388 (2007). M. G. Kang, L. J. Guo, Adv. Mater. 19, 1391 (2007). M.-G. Kang, M.-S. Kim, J. Kim, L. J. Guo, Adv. Mater. 20, 4408 (2008). C. Pina-Hernandez, P.-F. Fu, L. J. Guo, J. Vac. Sci. Technol. B 26, 2426 (2008). C. Pina-Hernandez, J. S. Kim, L. J. Guo, P. F. Fu, Adv. Mater. 19, 1222 (2007). X. Cheng, L. J. Guo, P. F. Fu, Adv. Mater. 17, 1419 (2005).
40. 41. 42. 43. 44. 45. 46. 47.
48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63.
CHAPTER 8 THREE-DIMENSIONAL NANOSTRUCTURE FABRICATION BY FOCUSED-ION-BEAM CHEMICAL-VAPOR-DEPOSITION
Shinji Matsui University of Hyogo 3-1-2 Koto, Kamigori, Ako, Hyogo, Japan
Three-dimensional nanostructure fabrication has been demonstrated by 30 keV Ga+ focused-ion-beam chemical-vapor-deposition (FIB-CVD) using a phenanthrene (C14H10) source as a precursor. Microstructure plastic arts is advocated as a new field using micro-beam technology, presenting one example of micro-wine-glass with 2.75 µm external diameter and 12 µm height. The deposition film is a diamondlike amorphous carbon. A large Young’s modulus that exceeds 600 GPa seems to present great possibilities for various applications. Producing of three-dimensional nanostructure is discussed. Micro-coil, nanoelectrostatic actuator, and nano-space-wiring with 0.1 µm dimension are demonstrated as parts of nanomechanical system. Furthermore, nanoinjector and nanomanipulator are also fabricated as a novel nano-tool.
1. Introduction Two-dimensional nanostructure fabrication using electron-beam (EB) and focused-ion-beam (FIB) has been achieved and applied to make various nanostructure devices such as single electron transistors and MOS transistors with nanometer gate-length. Ten-nm structures are able to be formed by using a commercial available EB or FIB system with 5-10 nm beam diameter and high-resolution resist [1]. In this way, it is considered that the technique of two-dimensional nanostructure fabrication has been established. Outlook on three-dimensional
351
352
Matsui
fabrication, there are three techniques using laser, EB, and FIB Chemical Vapor Deposition (CVD). Compared to three-dimensional fabrication of laser-CVD, FIB and EB-CVD are superior to laser-CVD [2] in point of a spatial resolution and a beam-scan control. Koops et al. demonstrated some applications such as AFM tip and field emitter by using EB-CVD [3]. Blauner et al. demonstrated pillars and walls with high aspect ratios by using FIB-CVD [4]. The deposition rate of FIB-CVD is much higher than that of EBCVD due to factors such as the difference of mass between electron and ion. Furthermore, a smaller penetration-depth of ion compared to electron allows to make a complicated 3-dimensional nanostructures. For example, when we make a coil nanostructure with 100 nm linewidth, electrons with 10-50 keV pass the ring of coil and reach on the substrate because of large electron-range (over a few µm), so it may be difficult to make a coil nanostructure by EB-CVD. On the other hand, as ion range is less than a few ten-nm, ions stop inside the ring. So far the complicated nanostructures using FIB-CVD have not been reported. This presents a complicated 3-dimensional nanostructure fabrication using FIB-CVD. 2. Three-Dimensional Nanostructure Fabrication We used a commercially available two FIB systems (SMI9200, SMI2050, SII Nanotechnology Inc.) with a Ga+ ion beam operating at 30 keV. The FIB-CVD was done using a precursor of phenanthrene (C14H10) as the source material. The beam diameter of SMI9200 was about 7 nm and that of SMI2050 was about 5 nm. The SMI9200 system was equipped with two gas sources in order to increase the gas pressure. The top of the gas nozzles faced each other and were directed at the beam point. The nozzles were set at a distance of 40 µm from each other and positioned about 300 µm above the substrate surface. The inside diameter of a nozzle was 0.3 mm. The phenanthrene gas pressure during pillar growth was typically 5x10-5 Pa in the specimen chamber, but, the local gas pressure at the beam point was expected to be much higher. The crucible of the source was heated to 85°C. The SMI2050 system, on the other hand, was equipped with a single gas nozzle. The FIB is scanned to
Three-Dimensional Nanostructure Fabrication
353
write the desired pattern by a computer control and the ion dose is adjusted to deposit a film of the desired thickness. The experiments were carried out at room temperature on a silicon substrate. The characterization of deposited film was performed by observation of transmission electron microscope (TEM) and measuring of Raman spectra. A carbon thin film with 200 nm thickness was deposited on a silicon substrate by 30 keV Ga+ FIB using a phenanthrene precursor gas. The cross-section structures and electron diffraction patterns were observed by using a 300 kV TEM. As a result, there were no crystal structures in TEM images and diffraction patterns. It is concluded that the deposited film is amorphous carbon (a-C). Raman spectra of a-C films were measured at room temperature with 514.5 nm line of an argon ion laser. The Raman spectra were recorded by a monochromator equipped with a CCD multi-channel detector. Raman spectra were measured at 0.1-1.0 mW to avoid thermal decomposition of the samples. A relatively sharper Raman band at 1550 cm-1 and a broad shoulder band at around 1400 cm-1 are observed in the spectra excited by a 514.5 nm line. Two Raman bands were plotted after the Gaussian line shape analysis. Raman bands at 1550 cm-1 and 1400 cm-1 originate in trigonal (sp2) bonding structure of graphite and tetrahedral (sp3) bond structure of diamond. This result indicates that a-C film deposited by FIB-CVD is diamondlike amorphous carbon (DLC) which have attracted attention because of their hardness, chemical inertness, and optical transparency. 2.1. Fabrication Process Beam-induced chemical vapor deposition (CVD) is widely used in the electrical device industry in repairing chips and masks. This type of deposition mainly done on two-dimensional (2D) patterning features, but it can also be used to fabricate a three-dimensional (3D) object. Koops et al. demonstrated a nano-scale structure 3D construction [3] using electron-beam-induced amorphous carbon deposition applied to a micro vacuum tube. In contrast, focused-ion-beam (FIB) induced CVD seems to have big advantages and potential in the fabrication of 3D nanostructures [4-6]. The key issue in making such 3D-work is the short
354
Matsui
penetration depth of the ions (a few nm) in the target material, where the penetration depth of the ions is much shorter compared to that of the electrons (several hundreds of microns). This short penetration depth reduces the dispersion area of the secondary electrons, and thus the deposition area is tightly limited to within about several tens nanometers. Usually, a 3D structure contains overhang structures and hollows. Gradual position-scanning of the ion beam during the CVD process causes the position of the preferentially growing region around the beam point to shift. When the beam point reaches the edge of the wall, secondary electrons appear at the side of the wall and just below the top surface. The DLC then starts to grow laterally. The width of the vertical growth is also about 80 nm. Therefore, combining the lateral growth mode with the rotating beam scanning, 3D structures having a rotational symmetry like a wineglass are obtained.
Fig. 1. Fabrication process for three-dimensional nanostructure by FIB-CVD.
Three-dimensional structure fabrication process by FIB-CVD is illustrated in Fig. 1 [7]. In FIB-CVD processes, beam is scanned at digital mode. First, a pillar is formed on the substrate by fixing a beam-position (position 1). After that, the beam-position is moved within a diameter of pillar (position 2) and then fixed until the deposited terrace thickness exceeds an ion-range which is a few ten nm. This process is repeated to make three-dimensional structures. The key point to make threedimensional structures is to adjust a beam-scan-speed as remaining ion-
Three-Dimensional Nanostructure Fabrication
355
beam within the deposited terrace which means that the terrace thickness exceeds an ion-range. The growth conditions of x and y-directions are controlled by both beam-deflectors. The growth of z-direction is determined by a deposition rate, that is, a height of structure is proportional to an irradiation-time when a deposition rate is constant.
Fig. 2. (a) Micro-wine-glass with 2.75 µm external diameter and 12 µm height. (b) Micro-coil with 0.6 µm coil-diameter, 0.7 µm coil-pitch, and 0.08 µm linewidth. (c) Micro colossem.
Fig. 3. Micro-wine-glass with 2.75 µm external diameter and 12 µm height on a human hair.
356
Matsui
We intend to open up microstructure plastic arts as a new field using FIB-CVD. To demonstrate the possibility, a micro-wine-glass was created on a Si substrate and a human hair as a work of microstructure plastic arts as shown in Figs. 2(a) and 3. A micro-wine-glass with 2.75 µm external diameter and 12 µm height was formed. Fabrication time was 600 s at 16 pA beam current. The beautiful micro-wine-glass gives us expectations of opening up microstructure plastic arts. The microcoliseum and leaning tower of Pisa were also fabricated on a Si substrate as shown in Figs. 2(c) and 4.
Fig. 4. Leaning tower of Pisa.
Various micro-system parts were fabricated by FIB-CVD. Figure 2(b) shows a micro-coil with 0.6 µm coil-diameter, 0.7 µm coil-pitch, and 0.08 µm linewidth. Exposure time was 40 s at 0.4 pA beam current. A coilpitch is able to change by controlling a growth speed with ease. Reducing a diameter of micro-coil, a micro-drill was formed. A diameter, pitch, and height of the micro-coil are 0.25, 0.20, and 3.8 µm, respectively. Exposure time was 60 s at 0.4 pA beam-current. The results show that FIB-CVD is one of the promising techniques to make parts of micro-system, although those mechanical performances have to be measured. 2.2. Three-Dimensional Pattern Generating System To fabricate the 3-D structure, we used deposition of a source gas by ion beam assist. The 3-D structure fabricates as a multi layer structure. In
Three-Dimensional Nanostructure Fabrication
357
this 3-D pattern-generating system, a 3-D model designed by a 3-D CAD system (3-D DXF format) is needed to manufacture the 3-D structure as the first step. There are no specializations of structure shape without cernuous structure shaped like pendulum. The 3-D CAD model, which is a surface model, is cut into several slices, as shown in Fig. 5. The thickness of slices depends on the resolution of the z direction (vertical direction). Second, the slice data are divided in the x and y directions (horizontal directions) to create the scan data (voxel data). To fabricate the overhang structure, ion beam must be irradiated in optimum order. If the ion beam is irradiated to a voxel located in midair without a support layer, the voxel deposits on the substrate. Therefore, the priority of irradiation is determined as number 1 to number 7 of Fig. 5.
Fig. 5. Data flow of 3-D pattern-generating system for FIB-CVD.
The scan data and blanking signal are then made from the scan order of priority, set dwell time, interval time, and irradiation pitch. These parameters are calculated from beam diameter, x-y resolution, and z resolution of fabrication. The z resolution is proportional to dwell time and inverse proportional to irradiation pitch squared. The scan data are input to the beam-deflector of the FIB-CVD in synchronization with the blanking data. The blanking signal controls the dwell time and interval time of the ion beam. Figure 6 shows a 3-D CAD model and SIM image of the Star Trek spaceship Enterprise NCC-1701D fabricated by FIB-CVD at 10~20 pA
358
Matsui
[8]. The nano-spaceship is 8.8 µm long and about a 1:100,000,000 scale on silicon substrate. The dwell time (td), interval time (ti), irradiation pitch (p), and total process time (tp) were 80 µs, 150 µs, 2.4 nm, and 2.5 h, respectively. The horizontal overhang structures was fabricated successfully.
Fig. 6. Star Trek, spaceship Enterprise NCC-1701D’s micro model, 8.8 µm long.
Figure 7 shows the artificial nano “T4 Bacteriophage”, which is a virus like the robot in the living body, fabricated by FIB-CVD on Si surface. Size of the artificial nano “T4 Bacteriophage” is about ten times as large as the real virus.
Three-Dimensional Nanostructure Fabrication
359
Fig. 7. T-4 Bacteriophage
3. Nanoeletromechanics 3.1. Young’s Modulus Measurement An evaluation of the mechanical characteristics of such nanostructures are needed as the basis of the material physics. Buks and Loukes reported a simple but essential technique [9] for measuring the resonant frequency of nano-scale objects by using a scanning electron microscope (SEM). The detector of secondary electrons in the SEM can respond up to around 4 MHz, thus the vibration of the sample is measured as the oscillatory output signals of the detector. Buks and Loukes used this technique to evaluate the Casimir force attracted between the two parallel beams fabricated on an nano scale. We evaluated the mechanical characteristics of DLC pillars in terms of the Young’s modulus determined by using resonant vibration and the SEM monitoring technique [10, 11]. The system set-up for monitoring mechanical vibration is shown in Fig. 8(b). There were two ways of measuring the pillar vibrations. One was active measurement, where the mechanical vibration was induced by a thin piezo-electric device, 300 µm thick and 3 mm square. The piezo device was bonded to the sidewall of the SEM’s sample holder with silver-paste. The sample holder was designed to observe cross sections in the SEM (S5000, Hitachi) system. Therefore, the pillar’s vibration was
360
Matsui
observed as a side view image as shown in Fig. 8(a). The vibrating frequency was the range of 10 kHz up to 2 MHz, which is much faster than the SEM raster scanning speed. Thus the resonant vibration of the pillars can be taken as the trace of the pillar’s vibration in the SEM image. The resonant frequency and amplitude were controlled by adjusting the power of the driving oscillator.
Fig. 8. (a) SEM image of the vibration. The resonant frequency was 1.21 MHz. (b) Schematic diagram of the vibration monitoring system.
The other way of measuring pillars vibrations is passive measurement using a spectrum analyzer (Agilent, 4395A), where the vibration seemed to mainly be induced by a environmental noise from rotary pumps and air conditioners. Some parts of the vibration would be resulted from the spontaneous vibration associated with thermal excitations [9]. Because of such excitation and residual noise, pillars on the SEM sample holder always vibrated at a fundamental frequency, even if the noise isolation is done in the SEM system. The amplitude of such spontaneous vibration was in the order of a few nanometer at the top of the pillar, and the high-resolution SEM can easily detect it at a magnification typically of 300,000.
Three-Dimensional Nanostructure Fabrication
361
We arranged several pillars that had varying diameters and lengths. The DLC pillars having the smallest diameter of 80 nm were grown using point irradiation. While we used two FIB systems for pillar fabrication, slight differences in the beam diameter of the two systems did not affect the diameter size of the pillars. Larger diameter pillars were fabricated using an area-limited raster scan mode. The raster scan in a 160-nm-square region produced a pillar having about a 240-nm-square cross section, and a 400-nm-square scan resulted in a pillar having a 480-nm-square cross section. The typical SEM image during resonance is shown in Fig. 8(a). The FIB-CVD pillars seemed very durable against the mechanical vibration. This kind of measurement usually requires at least 30 min including a spectrum analysis and photo-recording, but the pillars still survived without any change in the resonant characteristics. This durability in DLC pillars will be useful in nano-mechanical applications. The resonant frequency f of the pillar is defined by Eq. (1) for a pillar with a square cross-section and Eq. (2) for that with a circular cross-section: f square =
aβ 2 E 2 2π L 12 ρ
(1)
fcircular =
aβ 2 E 2 2π L 16 ρ
(2)
where a is the width of the square pillar and/or the diameter of the circular-shaped pillar, L is the length of the pillar, ρ is the density, and E is Young’s modulus. The coefficient of β defines the resonant mode and the β=1.875 at the fundamental mode. We used Eq. (1) for pillars 240 nm wide and 480 nm wide, and Eq. (2) for pillars grown by point-beam irradiations. The resonant frequency in terms of Young’s modulus depending on the ratio of the pillar diameter divided by the squared length is summarized in Fig. 4. All of the pillars evaluated in this figure were fabricated using the SMI9200 FIB system in rapid growth conditions. Typical growth rates were about 3 µm/min to 5 µm/min. for
362
Matsui
the 100-nm-diameter and 240-µm-wide pillars, and 0.9 µm/min. for the 480-nm-wide pillars. In the calculation of Fig. 9, we assumed that the density of the DLC pillars was about 2.3 g/cm3, which is almost identical to that of graphite and quartz. The inclination of the line in Fig. 9 indicated the Young’s modulus for each pillar. The Young’s modulus of each pillar was distributed in a range from 65 GPa to 140 GPa, which is almost identical to that of normal metals. A wider pillar tended to have a larger Young’s modulus.
Fig. 9. Resonant frequency dependence on the pillar length.
We found that the stiffness becomes significantly stiffer as the local gas pressure decreased as shown in Fig. 10. While the absolute value of the local gas pressure at the beam point is very difficult to determine, we found the growth rate can be useful as a parameter in terms of the local gas pressure to describe the pressure dependence on the Young’s modulus. All data points indicated in the Fig. 10 were obtained by pillars grown using point irradiation. Thus the pillar diameters were slightly distributed around 100 nm but did not exceed 5%. A relatively lower gas pressure maintaining good uniformity was obtained using a single gas nozzle and gas reflector. We use a cleaved side wall of Si tips as the gas reflector, which was placed 10-50 µm away from the beam point so as to be facing to the gas nozzle. The growth rate was controlled by changing the distance to the wall. While there is a large distribution of data points, the
Three-Dimensional Nanostructure Fabrication
363
stiffness of the pillar tended to become stiffer as the growth rate decreased. Two curves in Fig. 10 represented data points obtained under beam current of 0.3 pA (open circle) and 1 pA (solid circle), respectively. Both curves showed the same tendency where the saturated upper levels of the Young’s modulus was different for each ion current at lower gas pressure(lower growth rate). It should be noted that some of the pillar’s Young’s modulus exceeded 600 GPa, which is of the same order of tungsten-carbide. In addition, those estimation assumed the pillar density to be 2.3 g/cm3, however, a finite amount of Ga was incorporated with the pillar growth. If the calculation will take account of the increase of pillar density by the Ga concentration, Young’s modulus will exceed 800 GPa. Such high Young’s modulus is almost closed to that of carbon nano-tube and natural diamond crystal. We think that such high Young’s modulus is presumably due to surface modification caused by the direct ion impact.
Fig. 10. Young’s modulus dependence on the growth rate.
In contrast, when the gas pressure was high enough to achieve a growth rate of more than 3 µm/min, pillars became soft but the change of the Young’s modulus was small. The uniformity of Young’s modulus as shown in Fig. 9, presumably resulted from the fact that the growth condition was in this insensitive region, where the supplement of source gas limited the pillar growth.
364
Matsui
3.2. Free-Space-Nanowiring All experiments were carried out in a commercially available FIB system (SMI9200: SII NanoTechnology Inc) using a beam of 30 kV Ga+ ions. The beam is focused to a spot size of 7 nm at a 0.4 pA beam current, and it is incident perpendicular to the surface. The pattern drawing system of CPG (CPG-1000: Crestec Co) was added to the FIB apparatus to draw any patterns Using the CPG, a beam scanning control is possible such as scanning speed, x-y direction, and blanking of a beam, and the 3D free-space-nanowiring can be fabricated [12].
Fig. 11. Fabrication process of DLC free-space-wiring using both FIB-CVD and CPG.
Figure 11 illustrates the free-space-nanowiring fabrication process using both FIB-CVD and CPG. When phenanthrene (C14H10) gas or tungsten hexacarbonyl [W(CO)6] gas, which is a reactant organic gas, are evaporated from a heated container and injected into the vacuum chamber by a nozzle located 300 µm above the sample surface at an angle of about 45 deg with respect, the gas density of the C14H10 or W(CO)6 molecules increases on a substrate near a gas nozzle. The nozzle system served to create a local high-pressure region over the surface. The base pressure of the sample chamber is 2×10-5 Pa and the chamber pressure after introducing C14H10 and W(CO)6 as a source gas are 1×10-4
Three-Dimensional Nanostructure Fabrication
365
+ and 1.5×10-3 Pa, respectively. If a Ga ion beam is irradiated onto the substrate, C14H10 or W(CO)6 molecules adsorbed on the substrate surface are decomposed, and carbon (C) are mainly deposit on the substrate surface. The growth direction of deposition can be freely determined by controlling the scanning direction of a beam. Deposited material using C14H10 gas was diamondlike carbon, which was confirmed by Raman spectra and it had a very large Young’s modulus of 600 GPa [7, 10]. After two walls were formed in Fig. 11, free-space-nanowiring was grown by adjusting a beam scanning speed. The ion beam was used at 30 kV Ga+ FIB, and the amount of irradiation current was 0.8-2.3 pA. The x and y scanning directions and beam scanning speed were controlled by CPG. The z direction height was proportional to an irradiation time. A growth of deposition occurs horizontally by scanning a beam at a certain fixed speed in the direction of a plane. However, if the beam scanning speed is faster than the nanowiring growth speed, it grows downward or fall, and conversely if the scanning speed is slower, it grows up slantingly. That is, it is very important for growing up a nanowiring into a horizontal direction to control the beam scanning speed. It turns out that the optimal beam scanning speed to make the nanowiring, which grows up to be a horizontal direction using two C14H10 gas guns, is about 190 nm/s. The expected pattern resolution by FIB-CVD is around 80 nm, because both primary Ga+ ion and secondary-electron scattering are found around 20 nm [10, 13]. Figures 12 and 13 show the examples of free-space-nanowirings fabricated by FIB-CVD and CPG. All structures were fabricated using C14H10 gas as a precursor gas. Figure 12(a) shows nano-bridge free-space-wirings. The growth time was 1.8 min, and the wiring width was 80 nm. Figure 12(b) shows freespace-wirings of parallel resistances. The growth time was 2.8 min, and the wiring width was also 80 nm. Figure 13(a) shows free-space-wiring grown into sixteen directions from the center. Figure 13(b) shows a scanning-ion-microscope (SIM) image of an inductance (L), a resistance (R), and a capacitor (C) parallel circuit structure with free-space-nanowirings. A coil structure was fabricated by a circle scanning of Ga+ FIB. These growth times of L, R, and C structures were about 6, 2, and 12 min, and the all nanowiring
366
Matsui
width was about 110 nm. From these structures, it is possible to fabricate arbitrary nanowirings at arbitrary places by using FIB-CVD and CPG. And these results indicate that various circuit structures can be formed by combining L, C, and R.
Fig. 12. (a) DLC free-space-wiring with a bridge shape. (b) DLC fre-space-wiring with parallel resistances.
Fig. 13. (a) Radial DLC free-space-wiring grown into 16 directions from the center. (b) Scanning ion microscope (SIM) micrograph of an inductance (L), a resistance (R), and a capacitor (C) structures.
Three-Dimensional Nanostructure Fabrication
367
Free-space-wiring structures were observed by 200 keV TEM. The analyzed area was 20 nm . Figures 14(a) and 14(b) show TEM images of DLC free-space-wiring and pillar. It became clear from these EDX measurements that the dark part (A) of Fig. 14(a) corresponds to the Ga core, and the outside part (B) of Fig. 14(a) corresponds to amorphous carbon. In this way, a free-space-wiring consists of amorphous carbon containing a Ga core in the wiring. The center position of the Ga core is located below the center of the wiring. However, in the case of the DLC pillar, the Ga core is located in the center of the pillar. This result indicates that a center position of the Ga core is different between the DLC free-space-wiring and pillar. To evaluate the difference, the Ga core distribution in free-space-wiring was observed in detail by TEM. The center position of the Ga core is about 70 nm from the top, which is 20 nm below the center of the free-space-wiring. We calculated an ion range of 30 kV Ga ions into amorphous carbon by TRIM (Transport of Ions in Matter), of 20 nm. The calculation indicates that the displacement of the Ga core center position corresponds to the ion range.
Φ
Fig. 14. TEM images of (a) DLC free-space-wiring and (b) DLC pillar.
The electrical properties of free-space-nanowiring fabricated by FIBCVD using a mixture gas of C14H10 and W(CO)6 were measured. Nanowirings fabricated on Au electrode by using C14H10 and W(CO)6 as a source gas. Au electrodes were formed on a 0.2 µm-thick SiO2 on Si
368
Matsui
substrate by EB lithography and lift-off process. Two-terminal electrode method was used to measure the electrical resistivity of the nanowiring. Figure 15(a) shows the nanowiring fabricated by using only C14H10 source gas. This growth time was 65 s and the wiring width was 100 nm. Next, W(CO)6 gas was added to C14H10 gas as a mixture gas containing a metal to obtain a lower electrical resistivity. Figures 15(b), 15(c), and 15(d) correspond to the order of increasing W(CO)6 content in a mixture gas. The W(CO)6 content rate was controlled by sublimation temperature of C14H10 gas. Increasing W(CO)6 content, the nanowiring growth time and width become longer: (b) was 195 s and 120 nm, (c) was 237 s and 130 nm, and (d) was 296 s and 140 nm. Finally, we tried to fabricate a free-space-nanowiring using only W(CO)6, but did not obtain a continuous wiring, because the deposition rate in the case of using W(CO)6 source gas was very slowly.
Fig. 15. Electrical resistivities measurement for nanowirings. Electrical resistivity ρ was calculated by I–V curve. Elemental contents C, Ga, W were measured by SEM-EDX.
The electrical resistivity of Fig. 15(a) fabricated by using only C14H10 source gas was 1×102 cm. The elemental contents were 90% C and 10% Ga, which were measured by a spot beam of SEM-EDX. I–V curves (b), (c), and (d) correspond to the order of increasing W(CO)6 content in a mixture gas. Increasing W(CO)6 content, the electrical resistivity
Ω
369
Three-Dimensional Nanostructure Fabrication
decreases as shown in I–V curves (b)–(d). Moreover, Ga content rate was also increasing because nanowirings growth time became slower, that is, the irradiation time of Ga+ FIB became longer. The electrical resistivity of I–V curves (b), (c), and (d) were 16, 4×10-2, and 2×10-2 cm, respectively. The electrical resistivity of (e), which was fabricated by using only W(CO)6 source gas was 4×10-4 cm. The increasing of Ga and W metallic content corresponds to decreasing of electrical resistivity as shown in SEM-EDX measurement results of Fig. 15. These results indicate that a lower resistivity is caused by increasing metallic content. Electron holography is useful technology for direct observation of electrical and magnetic fields at nanoscale, and also has an efficient property of showing useful information by detecting the phase shift of the electron wave due to the electrical and magnetic field. The technique necessarily needs an electron biprism, which plays an important role of dividing electron wave into reference wave and objective wave. The biprism is composed of one thin filament and two ground electrodes.
Ω
Ω
Fig. 16. Electron biprism fabricated by FIB-CVD.
It is important to fabricate a filament as narrow as possible to obtain an interference fringe with a high contrast and good fringe quality. However, fabricating the filament with a diameter below 500 nm is very difficult, because a conventional electron biprism is fabricated by pulling
10 µm
370
Matsui
a melted glass rod by hand.To overcome this problem, we introduce a new fabrication technique of the electron biprism using FIB-CVD, and evaluate the characteristics of the new biprism. Figure 16 shows an SEM micrograph of the FIB-CVD biprism. We successfully fabricated DLC wiring with smooth surface in between W rods by free-space-wiring fabrication technology of FIB-CVD. The 80-nm DLC thin wiring works as the filament of the biprism. The diameter and length of the filament are 80 nm and 15 µm, respectively.
Fig. 17. Interference fringes and corresponding fringe profiles. (a) obtained using the biprism with diameter of 80 nm, and (b) obtained using the biprism with diameter of 400 nm.
Figure 17 shows the interference fringes obtained using the biprism with a filament of (a) 80-nm diameter and (b) 400-nm diameter, and corresponding fringe profiles. The applied-prism voltages were 20 V, respectively. The filament with 400-nm-diameter, close to the standard size used in the conventional electron biprism, was fabricated by Ptsputter coating onto the 80-nm-diameter filament. The interference
Three-Dimensional Nanostructure Fabrication
371
fringes were successfully obtained. Moreover, an interference region of the fringe obtained using the biprism with the 80-nm-diameter filament is larger than that of the fringe obtained using the biprism with the 400-nmdiameter filament. These results demonstrate an adequacy of the thin filament fabricated by FIB-CVD, and the new biprism will be very useful for an accurate observation with a high contrast and good fringe quality in electron holography.
Fig. 18. Fabrication process of 3D nano-electrostatic actuators.
3.3. Nanoelectrostatic Actuator The fabrication process of 3-D nano-electrostatic actuators and manipulators is very simple [14]. Figure 18 shows the fabrication process. First, a glass capillary (GD-1: Narishige Co.) was pulled using a micropipette puller (PC-10: Narishige Co.). The dimensions of the glass capillary are 90 mm in length and 1 mm in diameter. In this process, we obtained a 1-µm-diameter tip of the glass capillary. Next, we carried out Au coating on the glass capillary surface by DC sputtering. Au thickness was approximately 30 nm. This Au coating serves as the electrode that controls the actuator and manipulator. Then, the 3-D nano-
372
Matsui
electrostatic actuators and manipulators were fabricated by FIB-CVD. This process was carried out in a commercially available FIB system (SIM9200: SII NanoTechnology Inc.) with a Ga+ ion beam operating at 30 keV. FIB-CVD was carried out using a precursor of phenanthrene (C14H10) as a source material. The beam diameter was about 7 nm. The inner diameter of each nozzle was 0.3 mm. The phenanthrene gas pressure during growth was typically 5x10-5 Pa in the specimen chamber. The Ga+ ion beam could be controlled by transmitting CAD data of the arbitrary structures to the FIB system.
Fig. 19. Coil-type electrostatic actuator. (a) SIM image of a coil-type electrostatic actuator fabricated on the tip of Au-coated glass capillary. (b) Illustration of moving principle.
A coil-type electrostatic actuator was fabricated by FIB-CVD. Figure 19(a) shows the SIM image of the coil-type electrostatic actuator fabricated at 7 pA and 10 min exposure time. Figure 19(b) shows the movement principle of this actuator. The movement principle of this actuator is very simple. The driving force is the repulsive force induced by electric charge accumulation. This electric charge can be stored in this coil structure by applying the voltage onto a glass capillary. This coil
Three-Dimensional Nanostructure Fabrication
373
structure expands and contracts due to charge repulsion, as shown in Fig. 19(b). Figure 20 shows the applied voltage dependence of coil expansion. The length of the coil expansion is defined as the distance “a” in the in set of Fig. 20. The result revealed that the expansion could be controlled in the applied voltage range from 0 to 500 V.
Fig. 20. Applied voltage dependence of coil expansion.
4. Nanooptics: Brilliant Blue Observation from a Morpho-Butterfly-Scale Quasi-Structure The Morpho-butterfly has mysteriously brilliant blue wings, and the source of this color has been an interesting scientific problem for a long time. Through an intriguing optical phenomenon, the scales reflect interfered brilliant blue color for any incidence angle of white light. This color is called a structural color, meaning that it is not caused by pigment reflection [15]. When we observed the scales with a scanning electron microscope (SEM) (Fig. 21(a)), we found three-dimensional (3D) nanostructures with 2-µm height, 0.7-µm width, and a 0.22 µm grating pitch on the scales. These nanostructures caused the optical phenomenon in the same way as the play of color is produced in an opal and iridescence is produced by a jewel beetle. We fabricated the Morpho-butterfly-scale quasi-structure with a commercially available FIB system (SMI9200: SII Nanotechnology Inc.)
374
Matsui
using a Ga+ ion beam operating at 30 kV [16]. The beam diameter was about 7 nm at 0.4 pA. The FIB-CVD was done using a precursor of phenanthrene (C14H10).
Fig. 21. Morpho-butterfly scales. (a) Top view optical microscope image of Morphobutterfly. Cross-sectional view SEM image of Morpho-butterfly scales. (b) Inclined-view SIM images of Morpho-butterfly-scale quasi-structure fabricated by FIB-CVD.
In this experiment, we used a computer-controlled pattern generator, which converted 3-D computer-aided design (CAD) data into a scanning signal, as an FIB scanning apparatus to fabricate a 3-D mold [8]. The scattering range of Ga primary ion is about 20 nm and secondly electron range induced by Ga ion beam is about 20 nm, therefore the expected pattern resolution of the FIB-CVD was about 80 nm. Figure 21(b) is a scanning ion microscope (SIM) image of the Morpho-butterfly quasi-structure fabricated by FIB-CVD using 3-D CAD data. This result demonstrates that FIB-CVD can be used to freely fabricate the quasi-structure. We measured the reflection intensity from Morpho-butterfly scales and the Morpho-butterfly-scale quasi-structure through optical
Three-Dimensional Nanostructure Fabrication
375
measurement. In this measurement system, white light from a halogen lamp was directed onto a sample with incident angles ranging from 5 to 45°. The reflection was concentrated by an optical microscope and analyzed by a commercially available photonic multi-channel spectral analyzer system (PMA-11: Hamamatsu Photonics K.K.). The intensity of incident light from the halogen lamp had a wavelength with peak intensity close to 630 nm. The Morpho-butterfly-scale quasi-structure was made of DLC. The reflectivity and transmittance of a 200-nm-thick DLC film deposited by FIB-CVD, measured by the optical measurement system at a wavelength close to 440 nm (the reflection peak wavelength of the Morpho-butterfly), were 30% and 60%, respectively. The measured data thus indicated that the DLC film had high reflectivity near 440 nm, which is important for fabrication of an accurate Morpho-butterfly-scale quasi-structure.
Fig. 22. Intensity curves of reflection spectra. (a) Morpho-butterfly scales. (b) Morphobutterfly-scale quasi-structure.
376
Matsui
We measured the reflection intensities of Morpho-butterfly scales and the quasi-structure with an optical measurement system, and compared their characteristics. Figures 22(a) and 22(b) respectively show the reflection intensity from Morpho-butterfly scales and the quasistructure. Both had a wavelength whose peak intensity was near 440 nm and showed very similar reflection intensity spectra for the various incidence angles. We have thus successfully demonstrated that a Morpho-butterflyscale quasi-structure fabricated by FIB-CVD can show nearly the same optical characteristics as real Morpho-butterfly scales.
5. Nanobiology 5.1. Nanoinjector Three-dimensional nanostructures on a glass capillary have a number of useful applications such as manipulators and sensors in the various microstructures. We have demonstrated the fabrication of a nozzle nanostructure on a glass capillary for a bio injector by 30 keV Ga+ focused-ion-beam assisted deposition with a precursor of phenanthrene vapor and etching [17]. It has been demonstrated that nozzle nanostructures with various shapes and sizes have been successfully fabricated. An inner tip diameter of 30 nm on a glass capillary and a tip shape with an inclined angle have been realized. We reported that diamond-like carbon (DLC) pillars grown by FIB-CVD with a precursor of phenanthrene vapor have very large Young’ modulus that exceeds 600 GPa, which gives great possibilities for various applications [10]. These characteristics are very useful for various biological device fabrications. In this experiment, a nozzle nanostructure fabrication for biological nanoinjector research has been studied. The tip diameters of conventional bio-injectors are over 100 nm and tip shapes cannot be controlled. A bionanotool with various nanostructures on the top of a glass capillary has the following feature usages shown in Fig. 23: (1) injection of various reagents into a specific organelle in a cell, (2) selective manipulation of a
Three-Dimensional Nanostructure Fabrication
377
specific organelle out side of a cell by using the nanoinjector as an aspirator, (3) reduction of the mechanical stress when operating into the cell by controlling the shape and the size of the bio-nanoinjector, and (4) measurment of the electric potential of a cell, an organelle, and an ion channel exiting on a membrane by fabricating an electrode. Thus far, 3D nanostructure fabrications on a glass capillary have not been reported. We presents nozzle nanostructure fabrication on a glass capillary by FIBCVD and etching to confirm the possibility of bio-nanoinjector fabrication.
Fig. 23. Usages of bio-nanoinjector.
The nozzle structures of the nano injector were fabricated using a function generator (Wave Factory: NF Electronic Instruments). Conventional microinjectors are fabricated by pulling a glass capillary (GD-1: Narishige Co.) using a micropipette puller (PC-10: Narishige Co.). The dimensions of the glass capillary are 90 mm in length and 1 mm in diameter. Conventionally, the tip-shape control of a microinjector made by pulling a glass capillary, which is used as an injector into a cell, is carried out without or with mechanical grinding. However, the reliability of tipshape control is very poor and depends on a personal experience.
378
Matsui
Fig. 24. SIM images of bio-nanoinjector fabricated on a glass capillary by FIB-CVD. (a) before FIB-CVD, (b) after FIB-CVD, and (c) cross section of (b).
Fig. 25. Injection into an egg cell (Ciona intestinalis) using a bio-nanoinjector.
A bio-nanoinjector tip was fabricated on a glass capillary by FIBCVD as shown in Figs. 24(a)-24(c). First, FIB etching makes the tip surface of the glass capillary smooth. And then, a nozzle structure was fabricated on the tip by FIB-CVD. Figure 24(a) shows the tip –surface smoothed at 120 pA and 30 s exposure time by FIB-etching with an inner hole diameters of 870 nm. The nozzle structure fabricated by FIB-CVD with an inner hole diameters of 220 nm was shown in Fig. 24(b).
Three-Dimensional Nanostructure Fabrication
379
Figure 24(c) corresponds to a cross section of Fig. 24(b). These results demonstrate that a bio-nanoinjector could be successfully fabricated by a 3D nanostructure fabrication using FIB-CVD. The bio-nanoinjector was used to inject dye into a egg cell (Ciona intestinalis) as shown in Fig. 25.
5.2. Nanomanipulator An electrostatic 3-D nano-manipulator that can perform inclusion of nano parts and cell operation has been developed by FIB-CVD. This 3-D nano manipulator has four fingers in order to catch the target of various shapes certainly. The movable principle is that an electric charge is accumulated in the structure by applying voltage to four fingers structure and it move by repulsion of the electric charge. Furthermore, we succeeded to catch the micro-sphere (polystyrene latex with a diameter of 1 µm) by using this 3-D nano-manipulator with four fingers [18].
Fig. 26. SIM image of the 3D electrostatic nano-manipulator with four fingers before manipulation.
First, pulling of a glass capillary (GD-1; NARISHIGE CO.) was performed by using micropipette puller (PC-10; NARISHIGE CO.). About 1.0 µm diameter tip of a glass capillary could be obtained in this
380
Matsui
process. Second, Au coating of the glass capillary surface was carried out in order to fabricate an electrode for nano-manipulator control. Au thickness that carried out coating at this time was about 30 nm. Finally, 3-D nano-manipulator structure with four fingers (Fig. 26) was fabricated by FIB-CVD on the tip of the glass capillary with single electrode. A sphere can also be caught easily and certainly by doing so. That is, although it is expected that it is difficult to catch a sphere by a pair of chopsticks, it will become less difficult to catch a sphere, if the manipulator has several fingers like man’s hand.
Fig. 27. Illustration of 1 µm polystyrene micro-sphere manipulation by using 3-D electrostatic nano-manipulator with four fingers.
Micro-sphere (polystyrene latex with a diameter of 1 µm) manipulation was carried out under the optical microscope by using 3-D nano-manipulator with four fingers. The illustration describing the situation of a manipulation experiment is shown as Fig. 27. By connecting the manipulator fabricated by FIB-CVD to a commercial manipulator (MHW-3; NARISHIGE CO.), the movement of the direction of an X-axis, a Y-axis, and Z-axis was controlled. And the micro-sphere that is a target was fixed to the side of a glass capillary, and the situation of manipulation was observed from the top with the optical microscope.
Three-Dimensional Nanostructure Fabrication
381
Fig. 28. In-situ observation of 1 µm polystyrene micro-sphere manipulation by using 3-D electrostatic nano-manipulator with four fingers.
Fig. 29. SIM image of the 3-D electrostatic nano-manipulator with four fingers after manipulation.
And, optical microscope image of Fig. 28 shows the situation during manipulation. First, the 3-D nano-manipulator was made to approach a micro-sphere without applying voltage. Next, four fingers were opened by applying 600V before the micro-sphere and the micro-sphere was caught by turning off voltage in the position which can catch the micro-sphere.
382
Matsui
And then, the 3-D nano-manipulator was taken off from the side of a glass capillary. At this time, voltage is not applied to the manipulator and the force of catching the micro-sphere is the elastic force of a manipulator’s own structure. We succeeded to catch the micro-sphere as shown in SIM image of Fig. 29.
6. Summary Three-dimensional nanostructure fabrication has been demonstrated by 30 keV Ga+ FIB-CVD using a phenanthrene (C14H10) source as a precursor. The characterization of deposited film on a silicon substrate was performed by a transmission microscope and Raman spectra. This result indicates that the deposition film is a diamondlike amorphous carbon (DLC) which have attractive attention because of their hardness, chemical inertness and optical transparency. A large Young’s modulus that exceeds 600 GPa seems to present great possibilities for various applications. Nanoelectrostatic actuator, and nano-space-wiring with 0.1 µm dimension were fabricated and evaluated as parts of nanomechanical system. Furthermore, nanoinjector and nanomanipulator were fabricated as a novel nano-tool for manipulation and analysis of subcellular organelles. These results demonstrate that FIB-CVD is one of key technologies to make 3D nanostructure devices in the field of electronics, mechanics, optics and biology.
References 1. 2. 3. 4.
5. 6. 7.
S. Matsui, Proc. IEEE 85, 629 (1997). O. Lehmann, F. Foulon and M. Stuke, NATO ASI Ser. E 265, 91 (1994). H. W. Koops, Jpn. J. Appl. Phys. 33, 7099 (1994). A. Wargner, J. P. Levin, J. L. Mauer, P. G. Blauner, S. J. Kirch and P. Long, J. Vac. Sci. Technol. B8, 1557 (1990)., P.G. Blauner, Proc. 1991 Int. MicroProcess Conf. p.309. I. Utke, P. Hoffmann, B. Dwir, K. Leifer, E. Kapon and P. Doppelt, J. Vac. Sci. Technol. B18, 3168 (2000). A. J. DeMarco and J. Melngailis, J. Vac. Sci. Tecnol. B 17, 3154 (1999). S. Matsui, T. Kaito, J. Fujita, M. Komuro, K. Kanda and Y. Haruyama, J. Vac. Sci. Tecnol. B18, 3181 (2000).
Three-Dimensional Nanostructure Fabrication 8.
9. 10. 11. 12.
13. 14. 15. 16. 17. 18.
383
T. Hoshino, K. Watanabe, R. Kometani, T. Morita, K. Kanda, Y. Haruyama, T. Kaito, J. Fujita, M. Ishida, Y. Ochiai and S. Matsui, J. Vac. Sci. & Technol. B21, 2732 (2003). E. Buks and M. L. Roukes, Phys. Rev. B63, 033402 (2001). J. Fujita, M. Ishida, T. Sakamoto, Y. Ochiai, T. Kaito and S. Matsui, J. Vac. Sci. Technol. B19, 2834 (2001). M. Ishida, J. Fujita, and Y. Ochiai, J. Vac. Sci. Technol. B20, 2784 (2002). T. Morita, R. Kometani, K. Watanabe, K. Kanda, Y. Haruyama, T. Hoshino, K. Kondo, T. Kaito, T. Ichihashi, J. Fujita, M. Ishida, Y. Ochiai, T. Tajima and S. Matsui, J. Vac. Sci. Technol. B21, 2737 (2003). J. Fujita, M. Ishida, Y. Ochiai, T. Ichihashi, T. Kaito and S. Matsui, J. Vac. Sci. Technol. B20, 2686 (2002). R. Kometani, T. Hoshino, K. Kondo, K. Kanda, Y. Haruyama, T. Kaito, J. Fujita, M. Ishida, Y. Ochiai and S. Matsui, J. Appl. Phys. 43, 7187 (2004). P. Vukusic and J. Roy. Sambles, Nature 424, 852 (2003). K. Watanabe, T. Hoshino, K. Kanda, Y. Haruyama, and S. Matsui, Jpn. J. Appl. Phys. 44, L48 (2005). R. Kometani, T. Morita, K. Watanabe, K. Kanda, Y. Haruyama, T. Kaito, J. Fujita, M. Ishida, Y. Ochiai and S. Matsui, Jpn. J. Appl. Phys. 42, 4107 (2003). R. Kometani, T. Hoshino, K. Kondo, K. Kanda, Y. Haruyama, T. Kaito, J. Fujita, M. Ishida, Y. Ochiai and S, Matsui, J. Vac. Sci. Tecnol. B23, 298 (2005).
CHAPTER 9 DYE-SENSITIZED SOLAR CELLS BASED ON NANO-STRUCTURED ZINC OXIDE
Qifeng Zhang and Guozhong Cao* Department of Materials Science and Engineering, University of Washington, Seattle, WA 98195 USA *Corresponding author, Email:
[email protected]
This review focuses on recent developments in the use of ZnO nanostructures for dye-sensitized solar cells (DSSCs) applications. It will show that nano-structured ZnO photoelectrode films can significantly enhance the solar cell performance by offering large surface areas, direct electron pathways, effective light scattering centers, and when combined with TiO2, produce a core-shell structure that reduces the combination rate. The limitations of ZnO-based DSSCs are also discussed. Based on the achievement of ZnO aggregates that exhibit double increase in the conversion efficiency than the film that only consists of nanocrystallites, several possible methods are proposed so as to expand this idea to TiO2, motivating further improvement in the power conversion efficiency of DSSCs.
1. Introduction Increasing demand for fossil fuels and environmental impact of their use are continuing to exert pressure on an already stretched world energy infrastructure. Significant progress has been made in the development of renewable energy technologies such as solar cells, fuel cells, and biofuels. However, although these alternative energy sources have been marginalized in the past, it is expected that new technology could make them more practical and price competitive with fossil fuels, enabling eventual transition away from fossil fuels as our primary energy sources. 385
386
Zhang and Cao
Solar energy is considered the ultimate solution to the energy and environmental challenge as a carbon-neutral energy source. The conversion from solar energy to electricity is fulfilled by the devices of solar cells based on photovoltaic effect. Many photovoltaic devices have already been developed over the past five decades. However, a wide-spread use is still limited by two significant challenges, namely the conversion efficiency and the cost. One of traditional photovoltaic devices is single crystalline silicon solar cells, which were invented more than 50 years ago and currently make up 94% of market. Single crystalline silicon solar cells operate on the principle of p-n junctions formed by joining the p-type and n-type semiconductors. The electrons and holes are photogenerated at the interface of p-n junctions, separated by the electrical field across the p-n junction, and collected through the external circuits. In principle, the single crystalline silicon semiconductors can reach 92% of the theoretical attainable conversion, with 20% conversion efficiency in commercial designs. However, because of the considerably high materials cost, thin film solar cells have attracted wide attention. Amorphous thin film silicon is a good candidate because the defect energy level can be controlled by hydrogenation and the band gap can be reduced so that the light absorption efficiency can be much higher than crystalline silicon. The problem is that amorphous silicon tends not to be stable and can loss up to 50% efficiency within the first hundred hours. Today commercial roof products are available in operating at ~15% efficiency. Bridging the gap between the single crystalline silicon and amorphous silicon is the polycrystalline silicon film, for which the conversion efficiency of around 15% is obtained. Compound semiconductors such as gallium arsenide (GaAs), cadmium telluride (CdTe) and copper indium gallium selenide (CIGS) received much attention because they present direct energy gap, can be doped to either p-type or n-type, have band gaps matching the solar spectrum and high optical absorbance. The devices have demonstrated the conversion efficiencies of 16-32%. Although those photovoltaic devices built on silicon or compound semiconductors have been achieving high efficiency for practical use, they still require major breakthroughs to meet the long term goal of very low cost ($0.4/kWh).[1-6]
ZnO Dye-Sensitized Solar Cells
387
Figure 1. A dye-sensitized solar cell based on electrochemical system. (a) Schematic of the construction and operational principle of device, (b) SEM image of oxide (TiO2) electrode film with nanocrystallites (~20 nm in diameter), and (c) Electron transport in nanocrystalline oxide electrodes, in which photoexcited electrons are injected from the dye to the conduction band (denoted as “c.b.”) of the nanocrystallite (1), the dye is regenerated by electron transfer from a redox couple in the electrolyte (3), a recombination may take place between the injected electrons and the dye cation (2) or redox couple (4). (4) is normally believed to be the predominant loss mechanism. Electron trapping in the nanocrystallites (5) is also a mechanism that causes energy loss. LUMO and HOMO represent the lowest unoccupied molecular orbital and the highest occupied molecular orbital of the dye, respectively.[7, 16, 17]
To aim at further lowering the production costs, dye-sensitized solar cells (DSSCs) based on oxide semiconductors and organic dyes have recently emerged as promising approach to efficient solar energy conversion. The DSSCs are a photo-electrochemical system, which incorporates a porous-structured oxide film with adsorbed dye molecules as the photosensitized anode. A platinum-coated silicon wafer acts as the counter electrode (i.e., cathode), and a liquid electrolyte that traditionally contains I-/I3- redox couples serves as a conductor to electrically connect the two electrodes.[7-12] During operation, photons captured by the dye monolayer create excitons that are rapidly split at the nanocrystallite surface of oxide film; electrons are injected into the oxide film and holes are released by the redox couples in the liquid electrolyte (Fig. 1a). Compared with the conventional single crystal silicon-based or compound semiconductor thin film solar cells, the DSSCs are thought to be advantageous as a photovoltaic device possessing both practicable high efficiency and cost effectiveness. To date, the most successful DSSC was procured on TiO2 nanocrystalline film (Fig. 1b) combined with ruthenium-polypyridine complex dye as first reported by O’Regan and Grätzel in 1991.[13] Following this idea, a certified overall conversion
388
Zhang and Cao
efficiency of 10.4% was achieved on TiO2-RuL’(NCS)3 (namely “black dye”) system, in which the spectral response of the complex dye was extended into the near-infrared region so as to absorb far more of the incident light.[7, 14-16] The porous nature of nanocrystalline TiO2 films drives their use in DSSCs due the high surface area available for dye molecule adsorption. Meanwhile, the suitable relative energy levels at the semiconductor-sensitizer interface, i.e., the position of the conduction band edge of TiO2 being lower than the excited-state energy level of the dye, allow for the effective injection of electrons from the dye molecules to the semiconductor.[17] The achievement of acceptable conversion efficiency put much confidence to DSSCs in the ability to challenge the high costs of commercially available solar cells based on silicon or compound semiconcuctors. However, a further increase in conversion efficiency has limited by the energy loss due to the recombination between electrons and either the oxidized dye molecules or electron accepting species in the electrolyte during the charge transport process.[18-20] Such a recombination is predominately derived from the lack of depletion layer on TiO2 nanocrystallite surface, and it become significantly serious when the thickness of photoelectrode film is increased. To figure out this issue, DSSC technology based on ZnO has been explored extensively. ZnO is a wide-band-gap semiconductor that possesses the energy band structure and physical properties similar as those of TiO2 (Table 1), but has higher electronic mobility that would be favorable for electron transport with reduced recombination loss when used in DSSCs. A lot of studies have already been reported on the use of ZnO material for application in DSSCs. Although the conversion efficiencies of 0.4~5.8% obtained for ZnO are much lower than that of 11% for TiO2, ZnO is still thought of as a distinguished alternative to TiO2 due to the ease of crystallization and anisotropic growth. These natures provide ZnO with the ability to produce a wide variety of nanostructures presenting unique properties for electronics, optics, or photocatalysis.[21-28] Recent studies on ZnO nanostructure-based DSSCs have delivered many new concepts, leading to a better understanding of the photoelectrochemically-based energy conversion. This, in turn, would speed up the development of DSSCs that are associated with TiO2.
389
ZnO Dye-Sensitized Solar Cells Table 1. A comparison of physical properties between ZnO and TiO2. ZnO
TiO2
Ref.
Crystal structure
rocksalt, zinc blende, and wurtzite
rutile, anatase, and brookite
[29, 30]
Energy band gap (eV)
3.2-3.3
3.0-3.2
[29-31]
Electron mobility (cm2/Vs)
205-300 (bulk ZnO), 1000 (single nanowire)
0.1-4
Refractive index
2.0
2.5
[33]
Electron effective mass (me)
0.26
9
[31]
Relative dielectric constant
8.5
170
[31]
Electron diffusion coefficient (cm2/s)
5.2 (bulk ZnO), 1.7×10-4 (nanoparticulate film)
0.5 (bulk TiO2), ~10-8-10-4 (nanoparticulate film)
[29, 30, 32]
[34, 35]
One of defining features of nanostructures is their basic units on the nanometer (10-9 m) scale. This, first of all, provides the nanostructures with a large specific surface area. It may also result in many particular behaviors in electron transport or light propagation in view of the surface effect, quantum confinement effect, or photon localization.[36-38] Those nanostructural forms of ZnO developed during the past several decades mainly include nanoparticles,[39, 40] nanowires (or nanorods),[41, 42] nanotubes,[43] nanobelts,[44] nanosheets[41, 45] and nanotips.[21, 25] The production of these structures can be achieved through sol-gel synthesis,[40] hydrothermal/solvothermal growth,[41] physical or chemical vapor deposition,[42, 44] low temperature aqueous growth,[43, 46] chemical bath deposition,[47] or electrochemical deposition.[45, 48, 49] In this article, recent developments in ZnO nanostructures particularly for application in DSSCs are reviewed. It will show that photoelectrode films with nanostructured ZnO can significantly enhance solar cell performance by offering a large surface area for dye adsorption, direct transport pathways for photoexcited electrons, and efficient scattering centers for enhanced
390
Zhang and Cao
light harvesting efficiency. In addition, ZnO may be combined with TiO2 via core-shell structure for reduced combination rates. The limitations of ZnO-based DSSCs are also discussed. In the outlook, several attempts to expand ZnO concepts to TiO2 are presented to motivate further improvement in the conversion efficiency of DSSCs. 2. Nanostructures Offering Large Specific Surface Area A large internal surface area is the foremost requirement of the photoelectrode film in DSSCs, so that sufficient dye molecules would be adsorbed to act as an “antenna” for the capture of incident photons. Nano-materials can satisfy this requirement due to the formation of a porous interconnected network in which the specific surface area may be increased by more than 1000 times when compared with bulk materials. The abundant forms of ZnO nanostructures provide a great deal of opportunities to obtain high surface-to-volume ratios, which in turn contributes to the dye adsorption as well as the light harvesting in DSSCs. 2.1. ZnO Nanoparticulate Films ZnO films with nanoparticles for application in DSSCs have been extensively studied, partially due to the direct availability of porous structures with assembled nanoparticles and the simplicity in the synthesis of nanoparticles via chemically-based solution methods. Due to the mechanism of charge transfer between the dye and semiconductor, the strategy for studying DSSCs with nanoparticulate ZnO films is almost the same as that adopted for nanocrystalline TiO2. It can be somewhat regarded as a straight implantation of TiO2-based Grätzel-type solar cell technology, in which the TiO2 nanoparticles are replaced by ZnO nanoparticles while the dye sensitizer and electrolyte remain unchanged. 2.1.1. Sol-Gel Derived Nanocrystalline Films A traditional method of synthesizing ZnO nanoparticles is achieved by the preparation of ZnO sols in the liquid phase from homogeneous
391
ZnO Dye-Sensitized Solar Cells
ethanolic solutions with the precursors of lithium hydroxide and zinc acetate.[50-52] The resulting sol contains ZnO nanoparticles with the average diameter in a range of ten to several tens nm. A further process is performed on the sol to yield film by spin-coating or dip-coating, during which the system is converted from liquid sol into solid wet gel. With a following drying and heat treatment, the gel is thermalized so as to generate a porous structured film on the substrate. Doctor-blade method[53] is also a frequently adopted approach for the preparation of ZnO nanocrystalline film, during which Triton X-100 (1% by volume) is conventionally added to the sol to facilitate the film formation. Those residual organics are removed by a following heat treatment typically at temperatures between 300 and 400°C. For DSSC applications, the film is sensitized by immersing it in ruthenium-polypyridine complex dyes (so-called N3, N719 or black dye that is commercially available) for a given time. The as-sensitized film, for use as working electrode, is constructed to a cell device by assembling it with a counter electrode consisting of a thin platinum layer deposited on a silicon wafer. An electrolyte solution containing KI (0.5 M) and I2 (0.03 M) in a mixed solvent with ethylene carbonate and acetonitrile (60:40% by volume) is then introduced to the interval of two electrodes. The performance of the solar cell is characterized by recording the current-voltage (I-V) behavior when the cell is irradiated under AM 1.5 type simulated sunlight. The overall power conversion efficiency, η, can be calculated by
η (%) =
100 × FF × VOC ( mV ) × I SC ( mA cm 2 ) , Pin (mW cm 2 )
(1)
and FF, i.e., fill factor, is defined as FF =
Vmax (mV ) × I max (mA cm 2 ) , VOC (mV ) × I SC (mA / cm 2 )
(2)
where Pin is the input power density, i.e., the intensity of incident light, VOC is the open-circuit voltage, ISC is the short-circuit current density,
392
Zhang and Cao
Vmax and Imax are the voltage and current at maximum power output, respectively. With the use of sol-gel derived ZnO nanoparticle films and at the early developmental stage, the conversion efficiencies of DSSCs are reported fairly low, with values generally around 0.4-2.22%.[52-54] The scattered distribution of efficiency implys that the solar cell performance is sensitive to many factors, including the size and shape of the nanoparticles, the porosity of film, the technique of film fabrication, and the post-treatment to the photoelectrode. In regards to the improvement of conversion efficiency in ZnO DSSCs, Keis et al. reported a compression method for the film fabrication to achieve a highly active ZnO surface for dye adsorption.[55, 56] In this process, ZnO nanoparticles with an average size of ~150 nm were synthesized via a sol-gel route, and particularly the film was prepared by compressing the nanoparticle powder under a pressure of 1000 kg/cm2. When the obtained film was used in a DSSC under the illumination with an intensity of 10 mW/cm2, an overall conversion efficiency as high as 5% was achieved.[57] Comparing this result with a 2% conversion efficiency attained for ZnO films with 15-nm nanoparticles prepared using the aforementioned doctor-blade method without compression,[58, 59] the authors explained that the improvement in solar cell performance was due to the increased interfacial kinetics inherent in the compression method that didn’t involve a heat treatment process. It had been demonstrated that such an increased interfacial kinetics could result in a more efficient injection of electrons from the dye molecules to ZnO semiconductor. In this work, although light scattering effect is not addressed as a predominant reason that promoted the light harvesting, it is likely the ZnO nanoparticles with a size of ~150 nm play a role in the generation of light scattering, leading to additional contribution to the light harvest. 2.1.2. Electrostatic Spray Deposition Technique for Nanoparticulate Films Highly efficient conversion can be also obtained for nanoparticulate ZnO films fabricated by using an electrostatic spray deposition (ESD)
ZnO Dye-Sensitized Solar Cells
393
technique developed by Jansen et al. for coating electrode films.[60, 61] The basic principle of ESD is the generation of an aerosol using organic solvents containing inorganic or organometallic precursors under the influence of a high voltage. An aerosol is defined as a dispersion of solid particles or liquid droplets in a gaseous ambient. In order to obtain such an aerosol of micron-sized droplets, the precursor liquid is pumped through a metal nozzle. Usually a spherical droplet is formed at the tip of a nozzle, but if a high voltage is applied (typically between 6-15 kV) between the nozzle and a grounded substrate, this droplet at the tip of the nozzle transforms into a conical shape and fans out to form a spray of highly charged droplets. The generated spray droplets are attracted by the grounded and heated substrate as a result of the applied potential difference. Consequently, the droplets impinge onto the heated substrate, where they lose their charge. After complete solvent evaporation, a thin layer consisting of the inorganic product is left on the substrate surface.[60] The advantages of the ESD technique are (1) the great variety in tailoring the film morphology, (2) the possibility of controlling the chemical composition of the deposited coatings, and (3) the high deposition efficiency since the electric field directs the charged droplets to the substrate. As for a typical preparation of nanoparticulate ZnO films with the ESD technique,[62] a colloidal solution of precursor was prepared through hydrolysis of zinc salts in the presence of an amine and then mixed with polyvinyl alcohol (PVA) that was dissolved in a mixture of ethanol and water. The obtained precursor solution containing nanoparticles was subsequently used for the ESD deposition of ZnO film. PVA was employed to prevent the agglomeration of ZnO during high temperature calcinations process. With this fabrication, the porosity as well as the specific surface area of the films can be controlled by adjusting the weight ratio of nanoparticles to PVA. It has been reported that a ratio of 1.8 may yield a maximum specific surface area of 12.55 m2/g and a peak DSSC conversion efficiency of 3.4%. Recently, this technique was also exploited in the synthesis of TiO2 nanofibers, achieving a DSSC conversion efficiency around 5%.[63, 64]
394
Zhang and Cao
2.2. Nanoporous Structured ZnO Films Besides nanoparticulate films, nanoporous structured ZnO films were also studied as photoelectrodes in DSSCs due to their high porosity. Many techniques have been reported for the preparation of nanoporous ZnO films.[48, 49, 65, 66] Among these methods, electrochemical deposition and chemical bath deposition are optimal in fabricating nanoporous films with upstanding growth on the substrate, which is thought of as a structure favorable for electron transport from the point of generation to the collection electrode in DSSCs. 2.2.1. Electrochemical Deposition Fabrication Electrochemical deposition is a simple and low-cost route for the preparation of nanoporous films possessing a large surface area. The advantage of this method is that the growth orientation, morphology and thickness of the films can be modestly controlled by adjusting the deposition parameters (deposition voltage, current density, temperature, etc). A typical fabrication process for the electrochemical deposition of ZnO nanoporous films at low temperature was described by Xi et al.[48] In this fabrication, a two-electrode system consisting of a Zn sheet as the anode and glass coated with indium-tin-oxide (ITO) as the cathode was used. The electrolyte contained 0.04 M zinc nitrate and 0.04 M hexamethylenteramine (HMT) in a solvent formed by a mixture of distilled water and ethanol with a volume ratio of 1 to 1. HMT was employed in the electrolyte as a buffer to obtain a pH value around 6. The additional ethanol in the electrolyte acted as a wetting agent to improve the thickness uniformity and surface coverage. The deposition of porous ZnO films was carried out at 40°C in a water bath with a bias voltage of 0.5 ~ 0.7 V. The obtained films presented a sheet-like morphology comprised of interconnected small particles of 250 nm in diameter (Fig. 2a). It has been demonstrated that the film thickness is correlated to the bias voltage and deposition time. Typically, a thickness of about 2.6 µm can be obtained at 0.5 V for 30 min, 2.4 µm at 0.7 V for 20 min, and 3.4 µm at 0.7 V for 30 min. The film structure is also dependent on the temperature of the water bath. Increasing the
ZnO Dye-Sensitized Solar Cells
395
temperature to 70°C would result in a ZnO film with compact structure, showing a much lower DSSC conversion efficiency than the porous film due to the decreased internal surface area. In electrochemical deposition, the nucleation rate can significantly affect the film structure, especially in terms of the porosity. Through the use of a shape-control reagent or surfactant to mediate the nucleation, more porous ZnO films with an abundance of nanostructures have been developed by electrochemical deposition. For example, it was reported that ethylenediaminetetraacedic acid (EDTA) as a reagent could enable the morphology of ZnO films to evolve from a uniform planar foam structure to a bricklike or spherelike foam with increasing current density.[67] A typical SEM image of a ZnO nanoporous film with a bricklike structure consisting of ~20-nm-thick nanosheets is shown in Figure 2b. The film was synthesized by electrochemical deposition in solution of 0.05 M ZnCl2 + 0.01 M EDTA with a current density of 0.05 mA/cm2. In the electrochemical synthesis of ZnO nanoporous films, the use of sodium laurylsulfate as a surfactant was also studied.[68] The results showed that the growth rate of ZnO nanocrystals could be affected by the concentration of sodium laurylsulfate. Only slight effects were observed when the concentration of sodium laurylsulfate was lower than 400 µM. However, if the concentration was high enough (600 µM, up to 1 mM), a strongly increased growth rate due to catalysis by sodium laurylsulfate would enable the formation of micelles and assembly of these micelles on the charged electrode surface. A nanoporous structured ZnO film was achieved after extraction of the sodium laurylsulfate with ethanol. Polyvinylpyrrolidone (PVP) has been reported as a highly efficient surfactant that creates very impressive DSSC conversion efficiency on ZnO nanoporous films synthesized by means of electrochemical deposition.[69] As-obtained films possess a structure of interconnected crystal grains. It was demonstrated that the grain size, and thus the surface morphology of ZnO films, could be strongly influenced by the PVP concentration, i.e., increased PVP concentration would result in reduced grain size. Typically, when the PVP concentration was 4 g/L, the grain size was reduced to 20-40 nm. However, no ZnO film could be formed if the concentration was larger than 6 g/L. The DSSCs with ZnO
396
Zhang and Cao
films synthesized by using this method exhibit an extremely high conversion efficiency of 5.08%, while the photoelectrode consists of electrochemically deposited double-layer ZnO films typically containing 8-µm-thick nanoporous film on a 200-nm-thick compact nanocrystalline film. 2.2.2. Chemical Bath Deposition Fabrication Chemical bath deposition (CBD), developed for producing ceramic films from aqueous solutions at low temperatures, has also been used for the fabrication of ZnO nanoporous films.[70, 71] A typical process for the fabrication of ZnO films by CBD technique can be described in the following steps: (1) the formation of layered basic zinc acetate (LBZA), Zn5(OH)8(CH3COO)2·2H2O, by hydrolysis of zinc acetate dihydrate in methanol, (2) the deposition of the LBZA film on a substrate at 60°C through heterogeneous nucleation, and (3) a heat treatment at the temperature above 150°C so as to transform the LBZA film into crystalline ZnO. Film synthesized through this method show a nest-like morphology and porous structure (Fig. 2c).[72] More detailed studies pointed out that films fabricated by CBD were composed of sheet-like grains with ~11 nm ZnO crystallites, pore sizes of ~12 nm, and surface areas of about 25 m2/g.[73, 74]
Figure 2. SEM images of ZnO nanoporous films. (a) Electrochemically deposited film without using reagent,[48] (b) electrochemically deposited film with the use of EDTA reagent,[67] and (c) chemical bath deposited nest-like ZnO film.[72]
Unlike previously reported on a formation of Zn2+/dye complex, which is nonconductive and hinder the electron injection from dye
ZnO Dye-Sensitized Solar Cells
397
molecules to semiconductor, a critical feature of ZnO films produced by CBD is their remarkable stability in acidic dye.[75] That is, even for the prolonged dye-loading time (6 h), there is only a monolayer adsorption of dye molecules on the ZnO surface and no Zn2+/dye complex is formed. This has resulted in an overall conversion efficiency as high as 4.1%, while the photoelectrode of a DSSC consists of a 20-µm-thick ZnO film prepared via CBD and sensitized with N719 dye. Moreover, for this cell, the dye-loading amount has been estimated to be about 1.4×10-7 mol/cm2, which is comparable to the 1.3×10-7 mol/cm2 obtained for a 10-µm-thick nanocrystalline TiO2 photoanode (η = 10%) sensitized with N3 dye.
2.3. Other Nano-Structured ZnO Films ZnO nanostructures with other morphologies such as nanosheets, nanobelts and nanotetrapods have also been studied for DSSC applications on account of their large specific surface area. However, for these nanostructures, the specific surface area is not the only factor that determines the adsorption amount of the dye. Solar cell performance is also believed to be significantly affected by the geometrical structure of the photoelectrode films that are particular in the electron transport and/or light propagation. 2.3.1. Nanosheets ZnO nanosheets are quasi two-dimensional structures that can be fabricated by a re-hydrothermal growth process.[76] A film with dispersed ZnO nanosheets (Fig. 3a) used in a DSSC has been shown to possess a relatively low conversion efficiency, 1.55%, possibly due to an insufficient internal surface area. Spherical agglomeration of ZnO nanosheets (Fig. 3b) can significantly increase this internal surface area and result in an improvement of the conversion efficiency up to 2.61%.[77] It has been demonstrated that the performance of the solar cell also benefits from sheet-spheres that have a high degree of crystallinity and therefore, present low resistance with regards to the electron transport.
398
Zhang and Cao
2.3.2. Nanobelts ZnO films with nanobelt arrays prepared through an electrodeposition method were also studied for DSSC applications.[78] In fabricating these nanobelts, polyoxyethylene cetylether was added in the electrolyte as a surfactant. The obtained ZnO nanobelt array shows a highly porous stripe structure (Fig. 3c) with a nanobelt thickness of 5 nm, a typical surface area of 70 m2/g, and a DSSC conversion efficiency as high as 2.6%.
Figure 3. SEM images of nano-structured ZnO films. (a) Dispersed nanosheets,[76] (b) nanosheet-assembled spheres,[77] (c) nanobelt array,[78] (d) a ZnO tetrapods formed by three-dimensional structure with four arms extending from a common core,[79] and (e) networked film with interconnected ZnO tetrapods.[80]
2.3.3. Tetrapods A ZnO tetrapod possesses a three-dimensional structure consisting of four arms extending from a common core (Fig. 3d).[79, 80] The length of the arms can be adjusted within the range of 1-20 µm, while the diameter can be tuned from 100 nm to 2 µm through changing the substrate
ZnO Dye-Sensitized Solar Cells
399
temperature and oxygen partial pressure during vapor deposition.[81] Multiple-layer deposition can result in tetrapods connected to each other so as to form a porous network with a large specific surface area (Fig. 3e). The films with ZnO tetrapods used in DSSCs have achieved overall conversion efficiencies of 1.20~3.27%.[79, 80] It was reported that the internal surface area of tetrapod films could be further increased by incorporating these films with ZnO nanoparticles, leading to significant improvement in the solar cell performance.[79]
3. Nanostructures with Direct Pathway for Electron Transport Electron transport in DSSCs based nanoparticulate films has been proposed to occur either by a series of hopping events between trap states on neighboring particles[82] or by diffusive transport within extended states slowed down by trapping/detrapping events.[8, 17] An electron is estimated to cross 103-106 particles when traveling in the photoelectrode film.[82] While an 11% efficiency has been achieved on dye-sensitized TiO2 nanocrystalline films, the nanoparticulate oxides do possess deficiencies. Electron transport becomes more difficult with an increase in photocurrent due to the also increased recombination between the electrons and the oxidized dye or redox mediator in electrolyte.[83] Recombination, intrinsically, arises from the absence of a depletion layer at the TiO2 nanoparticle/electrolyte interface. In DSSCs with the film in a form of nanoparticles, the recombination has been addressed to be a crucial factor that causes energy loss and limits the further increase of conversion efficiency. In the past few years a variety of one-dimensional nanostructures based on different oxides have been developed so as to reduce the recombination rate in DSSCs. These one-dimensional nanostructures serve to increase the electron diffusion length by providing a direct pathway for electron transport from the point of generation to the collection electrode. Among all studies, one-dimensional ZnO nanostructures have been most reported due to both the ease of controllable anisotropic growth and the development of physical or chemical techniques that allows for a flexible tailoring of ZnO
400
Zhang and Cao
morphology (for example, shape, diameter, length, and density et al.). Typical one-dimensional ZnO nanostructures used for DSSCs involve types of nanowires/nanorods, nanotubes, nanotips, and other derivatives, synthesized through vapor-liquid-solid processes,[42] metal organic chemical vapor deposition (MOCVD),[84, 85] thermal evaporation,[85, 86] chemical solution synthesis at low temperature,[87] electrodeposition,[88, 89] and chemical spray pyrolysis.[90] In general, the vapor or physical methods possess an inherent advantage in forming both thin and long one-dimensional nanostructures with a high degree of crystal quality while affording the ability to precisely control the growth rate. Chemical solution methods own the capability of large-scale mass production at low temperatures and therefore, a flexible substrate can be used.
3.1. ZnO Nanowires ZnO nanowire arrays were first used in DSSCs by Law et al. in 2005 with the intent of replacing the traditional nanoparticle film with a consideration of increasing the electron diffusion length.[91, 92] A schematic of the construction of a nanowire DSSC is shown in Fig. 4a. Arrays of ZnO nanowires were synthesized in an aqueous solution using a seeded growth process. This method employed fluorine-doped tin oxide (FTO) substrates that were thoroughly cleaned by acetone/ethanol sonication. A thin film of ZnO quantum dots (dot diameter = 3-4 nm, film thickness = 10-15 nm) was deposited on the substrates via dipcoating in a concentrated ethanol solution. Nanowires were grown by immersing the seeded substrates in aqueous solutions containing 25 mM zinc nitrate hydrate, 25 mM hexamethylenetetramine and 5-7 mM polyethylenimine (PEI) at 92°C for 2.5 h. After this period, the substrates were repeatedly introduced to fresh solution baths in order to obtain continued growth until the desired film thickness was reached. The use of PEI, a cationic polyelectrolyte, is particularly important in this fabrication, as it serves to enhance the anisotropic growth of nanowires. As a result, nanowires synthesized by this method possessed aspect ratios in excess of 125 and densities up to 35 billion wires per square centimeter. The longest arrays reached 20-25 µm with nanowire diameter that varied from 130 to 200 nm. These arrays featured a surface
ZnO Dye-Sensitized Solar Cells
401
area up to one-fifth as large as a nanoparticle film. Figure 4b shows a typical SEM cross-section image of an array of ZnO nanowires. It was measured that the resistivity values of individual nanowires ranged from 0.3 to 2.0 Ω cm, with an electron concentration of 1-5×1018 cm–3 and a mobility of 1-5 cm2V-1s-1. Consequently, the electron diffusivity could be calculated as 0.05-0.5 cm2s-1 for a single nanowire. This value is several hundred times larger than the highest reported diffusivities for TiO2 or ZnO nanoparticle films in operating DSSCs, thus demonstrating the very good electrical conductivity of ZnO nanowires.[34, 93] At a full sun intensity of 100 ± 3 mW/cm2, the highest-surface-area devices with ZnO nanowire arrays were characterized by short-circuit current densities of 5.3-5.85 mA/cm-2, open-circuit voltages of 610-710 mV, fill factors of 0.36-0.38 and overall conversion efficiencies of 1.2-1.5%. The superiority of ZnO nanowires as a direct pathway for electron transport has been illustrated by Fig. 4c, in which the short-circuit current densities as a function of the internal roughness factor (defined as the ratio of actual surface area to the projected surface area) are compared for the cells with ZnO nanowires, TiO2 nanoparticles and ZnO nanoparticles. This plot shows that a rapid saturation and a subsequent decline in the short-circuit current density can be observed on the cells built with either 12-nm TiO2 nanoparticles or 30-nm and 200-nm ZnO nanoparticles. This confirms that the transport efficiency of nanoparticle films falls off above a certain film thickness due to the critical recombination. However, the nanowire films show a nearly linear increase in the short-circuit density that maps almost directly onto the TiO2 data even though the films are as thick as ~ 25 µm, manifesting a highly efficient transport of electrons in nanowires with decreased recombination rate. In addition, the nanowire cells generate considerably higher current densities than the ZnO nanoparticle cells over the accessible range of roughness factors (approximately 55-75% higher at a roughness of 200 as shown in Fig. 4c). This is also a confirmation that the nanowires offer better electron transport when compared to nanoparticles. The mechanism has been attributed to both the high crystallinity of nanowires and an internal electric field within the nanowires that can assist carrier collection by separating injected electrons from the surrounding electrolyte and sweeping them towards
402
Zhang and Cao
the collection electrode. Similar investigations can be else found in literature regarding the application of ZnO nanowires in DSSCs and with the attainment of higher conversion efficiencies.[94-99] However, it seems that the results are quite dependent on the experimental conditions, specifically in the synthesis method, substrate treatment, growth parameters, geometrical structure of the nanowires and array (pattern, diameter, length, density etc.), the measuring method (sensitization process, electrolyte composition, light source intensity, and active area of photoelectrode film etc.).
Figure 4. ZnO nanowire dye-sensitized solar cells.[91] (a) Schematic diagram of the cell with a photoelectrode comprised of the ZnO nanowire array, (b) Cross-sectional SEM image of the ZnO nanowire array, and (c) Comparative performance of nanowire and nanoparticle cells.
Nanowire arrays comprise a structure that provides a direct pathway for electron transport in DSSCs, however, their insufficient surface area has been demonstrated to be a limitation for higher conversion efficiency.[97] Many attempts have been made to solve this problem such as reducing the diameter size of nanowires thus increasing the density of array. It was reported that a 10 fold increase in the DSSC conversion efficiency could be achieved for high-density vertically aligned ZnO nanowire arrays when compared with randomly oriented nanowire films, indicating the potential space for DSSCs with nanowire array.[100] By blending the nanowires with nanoparticles, ZnO nanowire/nanoparticle composite films have been also proved to be effective in optimizing the surface area. Ku et al. reported a significant promotion in the DSSC overall conversion efficiency from 0.84% to 2.2% when ZnO nanoparticles with diameters of 5-30 nm were added to ZnO
ZnO Dye-Sensitized Solar Cells
403
nanowires.[101] As for the fabrication of nanowire/nanoparticle composite films, a pre-treatment performed by immersing the ZnO nanoparticle powder into a methanol solution containing 2% titanium isoproxide and 0.02 M acetic acid has shown to be favorable for the attachment of nanoparticles onto nanowire surface.[102]
3.2. ZnO Nanotubes Nanotubes differ from nanowires in that they typically have a hollow cavity structure. An array of nanotubes possesses high porosity and may offer larger surface area than that of nanowires. The synthesis of ZnO nanotube arrays can be achieved by using a modified method for the aqueous growth of ZnO nanowires at low temperatures.[43, 103] An overall conversion efficiency of 2.3% has been reported for DSSCs with ZnO nanotube arrays possessing a nanotube diameter of 500 nm and a density of 5.4×106 per square centimeter.[104] ZnO nanotube arrays can be also prepared by coating anodic aluminum oxide (AAO) membranes via atomic layer deposition (ALD), however it yields a relatively low conversion efficiency of 1.6%, primarily due to the modest roughness factor of commercial membranes.[105]
3.3. ZnO Nanotips By using MOCVD processing methods, ZnO nanotip arrays with different lengths can be synthesized. The DSSC performance of these nanotips has been investigated in previous studies.[106, 107] The results confirmed that the energy conversion efficiency of the cells increased with the length of the ZnO nanotips due to the increase in surface area of the photoelectrode film. An overall conversion efficiency of 0.55% was obtained for 3.2-µm-length ZnO nanotips. It has been reported that ZnO nanotips present a maximum in overall conversion efficiency at higher light intensities than TiO2 nanoparticles. This implies a nontrap-limited electron transport in the respect that the nanotips provide a faster conduction pathway for electron transport. This feature allow for the use of ZnO nanotips in the fabrication of more stable and efficient DSSCs under high illumination. It has also been demonstrated that the overall
404
Zhang and Cao
conversion efficiency could be increased to 0.77% by combining the ZnO nanotips with a Ga-doped ZnO film as a transparent conducting layer.[107]
3.4. ZnO Nanoflowers
Figure 5. Flower-like ZnO nanostructures.[108] (a) A schematic to indicate the difference in light harvest for films with nanowires and nanoflowers, and (b) SEM images of a ZnO nanoflower film.
The use of ZnO films with nanoflowers consisting of upstanding nanowires and outstretched branches have been also reported for application in DSSCs. Nanowires alone may not capture the photons completely due to the existence of intervals inherent in the morphology. Nanoflowers structures howver have nano-scaled branches that stretch to fill these intervals and, thus, provide both a larger surface area provided and a direct electron transport pathway along the nanowire network (Fig. 5a). Nanoflower films can be grown by a hydrothermal method at low temperatures, typically by employing a 5 mM zinc chlorine aqueous solution with a small amount of ammonia.[108] These as-synthesized nanoflowers, shown in Fig. 5b, have the dimensions of about 200 nm in diameter. The solar cell performance of ZnO nanoflower films was characterized by a current density and fill factor of 5.5 mA/cm2 and 0.53, respectively. These values are higher than the 4.5 mA/cm2 and 0.36 for films of nanowires with comparable diameter and array density.[109]
ZnO Dye-Sensitized Solar Cells
405
3.5. Dendritic ZnO Nanowires
Figure 6. Dendritic ZnO nanowires for DSSC application.[84] (a) Cross-sectional SEM image of film with dendritic ZnO nanowires (after two generations of growth), and (b) Semi-log plot of short circuit current (filled squares) and efficiency (open circles) vs. nanowire growth generation, with insets showing images of nanowire morphology for generations 0-2.
Dendritic ZnO nanowires, which possess a fractal structure more complicated than that of nanoflowers, are formed by a nanowire backbone with outstretched branches, on which the growth of smallersized nanowire backbones and branches is reduplicated. Baxter et al. described a MOCVD fabrication for dendritic ZnO nanowires by using a route of so-called multiple generation growth.[84] They first grew 100nm-diameter ZnO nanowires with 20-nm secondary nanowire branches that nucleated and grew from the primary nanowire backbone. This substrate with nanowires was then used to continue the nanowire growth, called “secondary generation” growth. During the growth of a second generation of nanowires, the outstretched nanowire branches act as new nucleation sites for nanowire growth. The growth can also be continued for third and fourth generations for the attainment of a dendrite-like branched ZnO nanostructure (Fig. 6a). A DSSC characterization showed that the short-circuit density increased with increasing growth generation due to the larger surface area, which in turn lead to increased adsorption of dye molecules. A total improvement of over 250 times in current density and over 400 times in efficiency has been observed when the film morphology was changed from smooth nanowires to branched
406
Zhang and Cao
second-generation nanowires (Fig. 6b). The efficiencies obtained using fourth-generation dendritic nanowire films with a branched nanostructure and a 10 µm thickness displayed an overall conversion efficiency of 0.5%, a more than 600-fold improvement over smooth nanowires. By integrating ZnO nanoparticles within the film of dendritic nanowires, the specific surface area was increased, leading to an improved conversion efficiency of 1.1% for these cells.
4. Core-shell Structures with ZnO Shell for Reduced Recombination Rate
Figure 7. Schematic of a typical TiO2-ZnO core-shell structure used in DSSCs. The ZnO shell provides an energy barrier at the interface between the TiO2 and the dye or electrolyte to reduce the recombination of electrons with oxidized dye molecules or those [110] accepting species in the electrolyte.
Core-shell structures are a designed configuration for electrode films in DSSCs to reduce the recombination rate at the electrode/electrolyte interface. A core-shell nanostructured electrode usually consists of a nanoporous TiO2 matrix that is covered with a shell of another metal oxide or salt (Fig. 7).[110] The conduction band potential of the shell should be more negative than that of the core semiconductor (TiO2). This establishes an energy barrier which hinders the reaction of electrons in the core with the oxidized dye or redox mediator in the electrolyte.[111] Several shell materials such as ZnO, Al2O3, SiO2, Nb2O5, WO3, MgO, SrTiO3, and CaCO3 have been reported to form an energy barrier (or a
ZnO Dye-Sensitized Solar Cells
407
surface dipole layer) on TiO2 enhancing the solar cell performance by providing a 35% increase in the overall conversion efficiency.[112-123] Among all those materials studied, ZnO has been attracting the most attention in view of its low electron affinity, high isoelectric point, and more negative conduction band edge when compared to TiO2.
4.1. Fabrication of Core-Shell Structures and Influence of Shell Thickness There have been several reports on the fabrication of ZnO shells for the TiO2 surface. In general, the fabrication approach of core-shell structures can be classified into two cases. One involves the initial synthesis of core-shell structured TiO2-ZnO nanoparticles and then applying them onto a conducting substrate so as to obtain a film.[115, 124] The consequence of this fabrication method is the formation of an energy barrier not only at the electrode/electrolyte interface but also between the individual TiO2 particles. Thus, at least conceptually, resistance to the transport of photoinjected electrons through the TiO2 network should increase. In a second approach, a nanoporous TiO2 electrode film is fabricated. This film then serves as a matrix for coating of a thin ZnO shell layer.[125, 126] As a result, the core particles are connected directly to each other, allowing for electron transport through TiO2 network. Wang et al. reported a hydrothermal method for the formation of TiO2-ZnO nanoparticles with a core-shell structure.[115] The researchers mixed 0.5 mol % ZnCl2 with 2M TiCl4 aqueous solution and adjusted the pH value to 5 with KOH. After hydrothermal growth at 170°C and annealing of the precipitates at 450°C, 0.46 mol % ZnO was found in the TiO2-Zn product. TEM images revealed the TiO2 nanoparticles to be about 10 nm in diameter and the covering layer of ZnO had no influence on the growth of TiO2 nanoparticles. XRD analysis indicated that the ZnO-covered TiO2 was still in pure anatase phase and no separated ZnO phase (or other phase) was detected. Therefore the possibility of Zn2+ substituting the lattice position of Ti4+ was exclude and it was inferred that the Ti4+ ions were first precipitated followed by the precipitation of Zn2+ on the TiO2 surface. The Zn2+ forms ZnO after annealing. Based on a similar mechanism, Kim et al. also reported a simple route to obtain a
408
Zhang and Cao
ZnO shell on TiO2 nanoparticles by directly soaking the TiO2 powder (Degussa, P25) in a 10 mM solution of ZnCl2 in absolute ethanol. This suspension solution was then sprayed into liquid N2. A powder sample was finally obtained after a freeze-drying process, followed by sintering at 500°C.[124] The thickness of the ZnO layer synthesized by this method was measured to be ~0.50 nm, which was thin enough for electron tunneling in the process of particle-to-particle transport. By first preparing the TiO2 nanocrystalline film and then depositing the ZnO shell using a layer-by-layer technique, Han et al. made a series of systematic studies on the construction of ZnO shells with varying thicknesses.[125, 127] The influence of this parameter was evaluated with regards to DSSC performance. It was found that, for TiO2 photoelectrodes coated with a 30-nm-thick ZnO layer, the overall conversion efficiency reached 4.51%, while it was 3.31% for bare TiO2 without modification. The thickness of the ZnO shell could significantly affect the short-circuit density, which became decreased as the shell thickness larger than 30 nm, partially because of the small effective mass of electrons (~ 0.3me) in ZnO.[128, 129] RF-magnetron sputtering method was also reported for the deposition of a ZnO coating layer on the TiO2 nanocrystalline films.[130] It was demonstrated that the conversion efficiency of TiO2 DSSCs was improved from 4.76% to 6.55% due to the ZnO modification.
4.2. The Role of ZnO Shell As for core-shell structures in DSSC applications, the role of the ZnO shell has most often been demonstrated to provide an energy barrier at the interface between the TiO2 and the dye or electrolyte to reduce the recombination of electrons with oxidized dye molecules or those accepting species in the electrolyte. Figure 8 presents an energy-level diagram that schematically indicates the function of the ZnO shell in a TiO2-ZnO core-shell structure.[127] That is, the bottom of conduction band (-4.0 eV) and the top of valence band (-6.8 eV) of ZnO are lower than the lowest unoccupied molecular orbital (LUMO, -3.8 eV) and the highest occupied molecular orbital (HOMO, -5.4 eV) energy levels of the dye, and meanwhile are higher than those of -4.2 eV and -7.4 eV for
ZnO Dye-Sensitized Solar Cells
409
TiO2, respectively. Those photogenerated electrons in the dye molecules with high kinetic energy can readily tunnel through the ZnO shell and inject into TiO2. However, the transport of electrons in the inverse direction may be blocked due to the presence of an energy barrier provided by the ZnO shell, thereby suppressing the recombination rate of photogenerated electrons.
Figure 8. Schematic of energy levels of a core-shell (TiO2-ZnO) structured photoelectrode in DSSCs.[127]
Although the energy barrier model can sufficiently explain the enhanced DSSC performance by ZnO-coated TiO2, Zaban et al.[111, 117] and Bandara et al.[131] believe that this enhancement results from the conduction band shift mechanism rather than the formation of an energy barrier at the surface. The shift of the TiO2 conduction band is attributed to the existence of a dipole layer at the electrode/electrolyte interface due to the differences between the shell and core materials with respect to the higher isoelectric point and lower electron affinity of ZnO than those of TiO2. Thus, the TiO2 conduction band is moved towards a relatively negative direction resulting in higher open-circuit voltage for the DSSCs. Some arguments insist that the enhancement of the ZnO shell with regards to the DSSC performance is a result of the prolonged lifetime of photogenerated electrons in ZnO than in TiO2. Hagfeldt et al. studied the
410
Zhang and Cao
charge transport properties in the nanostructured ZnO thin film electrode-electrolyte system with time resolved laser flash induced photocurrents and compared with the results on TiO2. They observed that the electrons were existed longer in ZnO than in TiO2 and thus, the electron losses to acceptors in the electrolyte or deep traps were less for ZnO than for TiO2.[132, 133] In addition, the ZnO shell has been proposed to increase the free electron concentration in the conduction band with respect to pure TiO2 by the blocking of surface states so as to prevent the loss of photogenerated electrons in virtue of electrical traps. It is therefore thought to favor the electron transport.[115] However, opinions that deny the efficiency improvement of TiO2-based DSSCs by forming outer shell structures of insulator or semiconductor materials are also shown in literature.[110, 134] These viewpoints are based on the experimental observations that show that the ZnO modification of nanocrystalline TiO2 may increase the open-circuit voltage indeed, but may also causes more of a decrease in the shortcircuit current density, ultimately resulting in a reduced conversion efficiency. This has been explained by either poor dye adsorption to zinc sites on the TiO2 surface or a low electron injection efficiency due to the buildup of a thin insulating ZnO surface layer. Thusly, the role of the ZnO shell on the TiO2 surface is still under discussion. The difference in the observed effects of the ZnO shell is perhaps due to the dye adsorption and the electron injection process that are both sensitive to the surface/interface of semiconductors. Also, the status of ZnO shell is quite dependent on the fabrication method as well as other experimental variables.
5. Light Scattering Enhancement Effect In DSSCs, the dynamic competition between the generation and recombination of photoexcited carriers has been clarified to be a bottleneck for developing higher conversion efficiency, i.e., the film thickness was expected to be larger than the light absorption length for capturing more photons. Meanwhile, the film thickness was impelled to be smaller than the electron diffusion length with respect to avoiding or reducing the recombination.[18, 19] The aforementioned approaches
ZnO Dye-Sensitized Solar Cells
411
outlined above serve mostly to overcome the recombination by either using one-dimensional nanostructures that provide a direct pathway for electron transport or using core-shell structures with an oxide coating on TiO2 to minify the recombination rate. Besides those approaches, a series of methods that address the generation of photoexcited carriers by combining nano-structured films with optical effects (light scattering or optical confinement)[135] has also been demonstrated to be effective in enhancing the light-harvesting capability of the photoelectrode film so as to promote the DSSC performance. Usami,[136] Ferber and Luther,[137] and Rothenberger et al.[138] ever theoretically demonstrated that the optical absorption of dye sensitized TiO2 nanocrystalline films could be promoted by additionally admixing large sized TiO2 particles as the light scattering centers. The light scattering efficiency has been shown to correlate with both the size of the scattering centers and the wavelength of incident light.[139] The scattering reaches a maximum when the size of the scattering centers is about kλ, where k is a constant and λ is the wavelength. Experimentally, it has been verified that the performance of DSSCs can be significantly improved when the TiO2 nanocrystalline films are combined with largesized SiO2, Al2 O3, or TiO2 particles.[140-144] By coupling a photonic crystal layer to conventional TiO2 nanocrystalline films for light scattering, Nishimura et al.[145] and Halaoui et al.[146] also succeeded in enhancing the light-harvesting capability of the photoelectrode. However, the introduction of large-sized particles into nanocrystalline films has the unavoidable effect of lowering the internal surface area of the photoelectrode film. This serves to counteract the enhancement effect of light scattering on the optical absorption, whereas the incorporation of a photonic crystal layer may lead to an undesirable increase in the electron diffusion length and, consequently, increase the recombination rate of photogenerated carriers. A recently reported hierarchically structured film with ZnO aggregates, which provides the photoelectrode with both a large surface area and efficient light scattering centers, can, to some extent, resolve such an inconsistency. When this type of nanostructured film was used for DSSCs, a very impressive enhancement in the overall conversion efficiency had been observed.[147-149]
412
Zhang and Cao
5.1. ZnO Aggregates Hierarchically structured ZnO films consist of submicron-sized ZnO aggregates. The synthesis of ZnO aggregates can be achieved by a hydrolysis of zinc salt in a polyol medium at 160°C.[147] By adjusting the heating rate during synthesis and using a stock solution containing ZnO nanoparticles of 5-nm in diameter, the ZnO aggregates with an either monodisperse or polydisperse distribution in size can be prepared.[148, 149] Figure 9 shows the morphology of a hierarchically structured ZnO film and the structure of the aggregates. It can be seen that the film is well packed by ZnO aggregates with a highly disordered stacking, while the spherical aggregates are formed by the numerous interconnected nanocrystallites that have sizes ranging from several tens to several hundreds of nanometers (Fig. 9b and c). The structural features of the aggregates are their possession of a porosity and geometrical size comparable with the wavelengths of visible light. Four kinds of ZnO films, differing in the degree of aggregation have been prepared for a comparison of their DSSC performance. Sample 1 is comprised of wellpacked polydisperse ZnO aggregates, sample 2 consists of aggregates with slight distortion of the spherical shape, sample 3 includes parts of aggregates and nanocrystallites, and sample 4 is constructed with no aggregates but dispersed nanocrystallites alone. It has been demonstrated that all these samples present approximately the same crystallite size of about 15 nm and similar specific surface area of ~ 80 m2/g. However, their photovoltaic behaviors exhibit a significant difference in the shortcircuit current density, resulting in a difference of overall conversion efficiency. Typically, a maximum short-circuit current density of 19 mA/cm2 and conversion efficiency of 5.4% are observed for sample 1, while minimum values of 10 mA/cm2 and 2.4% respectively are observed for sample 4. Intermediate current densities and efficiencies are found for samples 2 and 3 (Fig. 9d). An obvious trend is that the overall conversion efficiency becomes decreased as the degree of the spherical aggregation is gradually destructed. In other words, the aggregation of ZnO nanocrystallites is favorable for achieving a DSSC with high performance.
ZnO Dye-Sensitized Solar Cells
413
Figure 9. ZnO aggregate dye-sensitized solar cells.[148, 149] (a) Cross sectional SEM image of a ZnO aggregate film, (b) a magnified SEM image of an individual ZnO aggregate, (c) a schematic diagram illustrating the microstructure of aggregated ZnO comprised of closely packed nanocrystallites, (d) photovoltaic behaviors and (e) optical absorption spectra of N3 dye adsorbed ZnO film samples with difference in the degree of aggregation of nanocrystallites, (f) schematic of light scattering and photon localization within a film consisting of submicron-sized aggregates, and (g) dependence of overall conversion efficiency on the size and size distribution of aggregates in dye-sensitized ZnO solar cells.
Figure 9e shows the optical absorption spectra of the four kinds of ZnO films previously discussed after sensitized with N3 dye. All the ZnO samples exhibit an intrinsic absorption with similar absorption intensity below 390 nm, caused by the semiconductor of ZnO with the electron transfer from valence band to conduction band. However, the
414
Zhang and Cao
absorption at wavelengths above 400 nm varies significantly, revealing the highest intensity for sample 1, a lower intensity for samples 2 and 3, and minimum in intensity for sample 4. In the spectra shown in Fig. 9e, only sample 4 presents an absorption peak centered around 520 nm. This corresponds to the visible t2→π* metal-to-ligand charge transfer (MLCT) in N3 dye,[75] but with a slight blue-shift than adsorption of pure N3 due to the electronic coupling between N3 and ZnO. All the other samples show a monotonic increase in the absorption intensity as the wavelength varies from visible to ultraviolet. This portion of absorption is contributed by the dye molecules that are adsorbed on the ZnO surface, and the difference in the absorption intensity implies that the absorption is structure-related. It has been suggested that the difference in the absorption of samples 1-3 arises from the submicron-sized aggregates that cause the light scattering. This weakens the transmittance of films and causes a pseudo absorption in the spectra to be displayed. The solar cell performance is improved due to the presence of light scattering in hierarchically structured ZnO films, by which the traveling distance of light within the photoelectrode film can be significantly extended (Fig. 9f). As such, the opportunities of incident photons being captured by the dye molecules are increased. The difference in the optical absorption of the four kinds of films implies that an improvement in the degree of aggregation of nanocrystallites would induce more effective light scattering in the visible region. A photon localization effect may also occur on these films due to their highly disordered structure that confines the light scattering in closed loops. Further studies reinforce the light scattering mechanism. It has been demonstrated the performance of DSSCs with hierarchically structured ZnO films can be significantly affected by either the average size or the size distribution of aggregates.[149] The films with polydisperse aggregates, which result in more disordered structure and achieve better packing, establish higher conversion efficiencies than those with monodisperse aggregates. The enhancement effect becomes more intensive when the maximum size of the aggregates in the polydisperse films or the average size of aggregates in monodisperse films increases to be as larger as or comparable to the wavelength of visible light (Fig. 9g). These results confirm the rationality of enhanced solar cell
ZnO Dye-Sensitized Solar Cells
415
performance arising from light scattering, generated by hierarchically structured ZnO films and promoting the light capturing ability of photoelectrode. Compared with those large-sized TiO2 particles or photonic crystal layer ever reported elsewhere,[141, 145, 146] the method using controlled ZnO aggregation of nanocrystallites in DSSCs presents obvious advantages in the generation of light scattering meanwhile without causing detrimental loss in internal surface area of the photoelectrode film. That is, the size of these aggregates is on submicron scale comparable to the wavelength of visible light, so that an efficient scattering to the incident light would be established within photoelectrode film and thus extends the traveling distance of photons being absorbed by dye molecules with more opportunities. Meanwhile, the film with aggregates consisting of interconnected nano-sized crystallites provides a highly porous structure, ensuring large specific surface area for dye adsorption, unlike the large-sized TiO2 particles with solid core that consume the internal surface area or the photonic crystal layer that is added to generate light scattering but with increased film thickness as well as electron diffusion length.
5.2. One-Dimensional ZnO Nanostructures for Light Scattering Recently, the enhancement effect of light scattering on DSSC performance was also observed in films consisting of one-dimensional ZnO nanostructures.[150] The films were typically fabricated using a spray deposition technique, where 10-nm-diameter ZnO nanoparticles dispersed in 1-butanol were prepared as a precursor. Based on such precursor, various nanostructures (nanorods, nanoflakes, or nanobelts) had been produced by an electric-field-induced self-assembly process, which could cause dipole-dipole interaction of nanoparticles. Films fabricated through this technique displayed large surface areas. Moreover, it was speculated that these highly disordered structures contributed to the light harvesting efficiency of photoelectrode by causing random multiple light scattering, as well as possible photon localization because of the formation of optical traps. Typically, an overall conversion
416
Zhang and Cao
efficiency of 4.7% was achieved on films with randomly oriented nanorods.
6. Limitation on ZnO-Based DSSCs It has been demonstrated that ZnO has approximately the same band gap and band position as TiO2. Furthermore, ZnO possesses a high electron mobility, low combination rate, and good crystallization into an abundance of nanostructures. Many efforts have been made on ZnObased DSSCs with either nanocrystalline films or the films consisting of various nanostructures, which are special in electron transport and/or photon capture. However, so far the results obtained for dye-sensitized ZnO solar cells have still shown relatively low overall conversion efficiencies when compared with TiO2-based systems. Limiting performance in ZnO-based DSSCs may be explained by the instability of ZnO in acidic dye (which causes results in the formation of excessive Zn2+/dye agglomerates) and the slow electron injection kinetics from dye to ZnO.
6.1. Instability of ZnO in Acidic Dyes 6.1.1. Formation of Zn2+/Dye Complex Commercially available dyes such as N3, N719 or “black” dye derived from Ruthenium-polypyridine complex have been widely used as sensitizer for TiO2-based DSSCs. The molecules of these dyes have carboxyl groups that connect with TiO2. However, direct use of these dyes with ZnO is difficult in that the surface structure of the ZnO crystals may be destroyed when they are soaked in an acidic dye solution containing Ru-complex for an extended period of time. By preparing an authentic Zn2+-N3 film and the N3-adsorbed ZnO nanoparticle films and comparing the absorption spectrum and fluorescence spectroscopy of those films immersed in dye solution for different durations of time, Horiuchi et al. studied the formation of the Zn2+/dye complex layer on ZnO nanoparticle surface.[151] They found that such a complex layer could always be observed if the immersing time was longer than 3 h.
ZnO Dye-Sensitized Solar Cells
417
Below an immersion time of 10 min, no complex was formed. Westermark et al.[152] and Chou et al.[153] verified that a short sensitization time may be favorable to avoid the formation of Zn2+/dye complex. However, the formation of Zn2+/dye complex seems very sensitive to the experimental conditions, resulting in a considerable difference in the optimization of sensitization time. The Zn2+/dye complex is nonconductive and can agglomerate to form thick covering layer instead of a monolayer, and it is therefore inactive for electron injection. That means, the sensitization time of ZnO in acidic dyes is limited by its stability, resulting in an insufficient dye adsorption and thus the poor performance of DSSCs. The spatial images of both the distribution of dye adsorbed on semiconductor surface and the electron injection process from an excited dye into nanocrystallites can be reflected by transient absorption microscopy, a newly developed technique for studying heterogeneous photochemical systems.[154] A typical example of using transient absorption microscopy for the study of N3-adsorbed ZnO films is based on the experimental observations that 1) ZnO films immersed in dye for an extremely short time (3 min) can obtain a monolayer adsorption of dye molecules on the surface, whereas a long time immersion (12 h) leads to serious agglomeration of Zn2+/dye complex, 2) the images of ground-state absorption not be homogeneous across the films, indicating that the ground-state absorption is not due solely to N3 dye directly adsorbed on the ZnO surface, but also to the dye in the agglomerates, 3) fluorescence emission can only be obtained for conc. N3/ZnO film instead of dil. N3/ZnO film, indicating that the fluorescence emission originates from the excited Zn2+/dye agglomerates; this is consistent with the observed heterogeneous image, and 4) the image of transient absorption, which intensity is proportional to the efficiency of electron injection, shows to be homogeneous in the film region. That is, in spite of the heterogeneity of the distribution of Zn2+/dye agglomerates, the electron injection is distributed homogeneously on the ZnO film. In other words, those Zn2+/dye agglomerates are inactive for electron injection. Accordingly, a model is proposed to exhibit the structure of dye adsorption on ZnO and the distribution of Zn2+/dye agglomerates (Fig. 10). With this model, a monolayer of dye molecules is formed
418
Zhang and Cao
homogeneously on the ZnO nanoparticle surface. These dyes are active and contribute to electron injection. On the exterior of monolayer, agglomeration of Zn2+/dye complex is yielded even to form micron-sized crystals. These agglomerates are inactive since those photogenerated electrons in the agglomerates are separated from the ZnO semiconductor by the layer of active dyes.
Figure 10. A proposed structure of a N3 dye-sensitized ZnO film with Zn2+/dye agglomerates.[154]
The formation of Zn2+/dye complex has been attributed to the dissolution of surface Zn atoms by the protons released from the dye molecules in an ethanolic solution. The instability of ZnO results from its surface properties in acidic dyes. In general, in a solution, the surface of oxide is predominantly positively charged at a pH below the point of zero charge and negatively charged above this value, while the point of zero charge of metal oxides is defined as the pH where the concentrations of protonated and deprotonated surface groups are equal. For the ZnO sensitization process with Ru-complex dye, the pH (pH = 5) is much lower than the point of zero charge of ZnO (≈ 9). That means the ZnO surface is positively charged. Thus, the protons adsorbed on the ZnO surface will dissolve the ZnO.[155] Bahnemann has shown that dissolution of ZnO colloids occurs below pH 7.4.[156] Theoretical investigations also indicate that the bond length between Zn and O atoms on the ZnO (ı0ī0) surface increases upon the adsorption of formic acid, making the Zn-O bond weaker and prone to a Zn atom dissolution.[157]
ZnO Dye-Sensitized Solar Cells
419
Recent study else reports that the temperature may have an effect on the adsorption and aggregation of dye on the surface of ZnO.[158] A higher temperature can effectively reduce the occurrence of dye aggregation and increase the amount of dye adsorbed on the ZnO. 6.1.2. Passivation of ZnO Surface for Improved Stability In order to avoid the formation of Zn2+/dye complex, many core-shell structures have been developed to improve the stability of ZnO in acidic dye solution by coating a buffer layer on the ZnO surface. SiO2 has been demonstrated to be one of very effective shell materials on ZnO, preventing the generation of Zn2+/dye agglomerates due to the strong interaction between Si4+ and O2- ions. Typically, when 5-nm-diameter ZnO nanoparticles were coated with SiO2 layer (molar ration of Si/Zn around 0.2), for a film in 13-µm thick, an overall conversion efficiency of 5.2% was achieved.[159] The role of the SiO2 shell was demonstrated to 1) suppress the formation of Zn2+/dye agglomerates and establish adequate dye adsorption on the electrode surface, and 2) reduce the recombination by decreasing the surface traps. TiO2 modification of the ZnO surface is also a representative method to prevent the surface Zn atoms from being dissolved and forming Zn2+/dye agglomerates. A simple route for coating TiO2 layer on ZnO can be completed by directly soaking or dip-coating ZnO film in a solution of 10 mM titanium alkoxide (Ti(OBu)4) dissolved in 2-propanol and followed with heat treatment at 400°C.[160] It can also be achieved by electrochemical deposition in solution containing 0.15 M LiNO3, 0.005 M Zn(NO3)2 hydrate, and 0.05 M ZnCl2 in propylene carbonate.[161] However, these methods are not without their drawbacks on account that the film thickness is hard to be controlled. Atomic layer deposition (ALD) is a newly developed technique for thin film fabrication and has been also reported for coating TiO2 layer on the surface of nano-structured ZnO.[162] With ALD technique, the growth rate as well as the shell thickness can be precisely controlled on the order of 1 nm. Figure 11 shows typical transmission electron microscopy (TEM) images of ZnO-TiO2 core-shell nanowires and nanotube-structured TiO2 shell prepared by ALD method.[163] The crystal phase of deposited TiO2
420
Zhang and Cao
shell has been demonstrated to be dependent on its thickness, being completely amorphous for less than 5 nm and converted to polycrystalline anatase for thicker film. While like shells composed of other oxides can provide a passivated surface with increased stability, the TiO2 shells on ZnO can else lead to significant promotion in both the open-circuit voltage and fill factor, resulting in even doubly improved overall conversion efficiency. Typically, it was reported that the conversion efficiency was increased from 0.85% for bare ZnO nanowire array to 1.7-2.1% with a crystalline TiO2 coating layer (10-35 nm thick) on ZnO.[163] (a)
(b) )
(d) )
(c) )
Figure 11. A characterization of ZnO-TiO2 core-shell structured nanowire and nanotubestructured TiO2 shell prepared by ALD.[163] (a) TEM image of a core-shell structure with ZnO nanowire core and TiO2 shell, (b) selected area electron diffraction pattern of the core-shell structured nanowire, (c) energy dispersive spectroscopy (EDS) elemental profile along the dashed line in part (a), and (d) TEM image of a nanotube-structured TiO2 shell, where the ZnO nanowire core was removed by using 1 M aqueous HCl.
6.2. Low Electron Injection Efficiency Electron injection efficiency describes the probability of photogenerated electrons to transfer from the dye molecules to semiconductor. In term of DSSC performance, the electron injection efficiency φinj and the incident photonto-current conversion efficiency (IPCE) are related by IPCE(λ)=LHE(λ)×φinj×ηC, where LHE(λ) is the light harvesting efficiency, λ is the wavelength of incident light, and ηC
ZnO Dye-Sensitized Solar Cells
421
is the collecting efficiency for injected electrons at the back contact.[75] The electron injection efficiency is, above all, determined by the electronic coupling between the dye and semiconductor, and it is also judged by the relative energy levels of the dye and semiconductor, the residential lifetime of photogenerated electrons in the dye molecules, and the density of electron accepting states in semiconductor. Particularly, for ZnO sensitized with Ru-complex acidic dyes, it has been demonstrated that the injection process may else be influenced by the formation of Zn2+/dye complex agglomerates, resulting in low electron injection efficiency. Either for ZnO or TiO2, the injection of electrons from Ru-based dyes to semiconductor shows a similar kinetics that includes a fast component less than 100 fs and slower components on picosecond time scale.[164-169] Such a biphasic kinetics are caused by competition processes between the ultrafast electron injection and molecular relaxation.[170] However, for ZnO with Ru-based dyes, the electron injection is dominated by slow components, whereas for TiO2 it is dominated by fast component, leading to the difference more than 100 times in injection rate constant. Based on a two-state injection model[166] and using ultrafast infrared transient absorption spectroscopy, a quantitative study has been given to describe the electron injection dynamics from Ru polypyridyl complexes to nanocrystalline ZnO or TiO2 film, revealing that the injection time scales from unthermalized and relaxed excited states to ZnO are estimated to be 1.5 and 150 ps, respectively, both of which are an order of magnitude slower than to TiO2.[164] The different injection dynamics most likely originates from the conduction band structures of semiconductors, i.e., the ZnO conduction bands are largely derived from empty s and p orbitals of Zn2+, while the TiO2 conduction band is comprised primarily of empty 3d orbitals from Ti4+.[170] The difference in the band structure results in different density of states and possibly different electronic coupling strength with the adsorbate. Enright et al. estimated that the density of conduction band states near the band edge is as much as two orders of magnitude higher in TiO2 by reason of the lager effective mass of the conduction band electron in TiO2 (5-10me) than that in ZnO (~0.3me).[129]
422
Zhang and Cao
The slower electron injection dynamics for ZnO with Ru-based dyes is also proposed as a result that the electron injection proceeds stepwise via intermediate states, as described by[171] hν ~ 150 ps < 100 fs N3 + ZnO → N3 * + ZnO → intermediate → N3+ + e
− CB
where N3* and N3+ represent the excited states and oxidized states, respectively, and eCB- indicates a conducting electron in ZnO. The origin of intermediate states has been ascribed to the interaction between the photoexcited dye and localized surface states on ZnO surface. The electron transfer is considered to be occurred slowly due to the existence of intermediate states and therefore results in low injection efficiency.
6.3. New Types of Photosensitizers for ZnO In view of the instability of ZnO in acidic dyes, the development of new types of photosensitizers for use in ZnO DSSCs has already been a subject that is widely concerned. These photosensitizers are expected to be chemically bonded to the ZnO semiconductor, be charge transferable with high injection efficiency, and be effective for light absorption in a broad wavelength region. New types of dyes have already developed with the aim of fulfilling these criteria. Examples include heptamethinecyanine dyes adsorbed on ZnO for absorption in the red/near-infrared (IR) region,[172, 173] and unsymmetrical squaraine dyes with deoxycholic acid, which increase photovoltage and photocurrent by suppressing the back electron transport.[174] Mercurochrome (C20H8Br2HgNa2O) is one of the newly developed photosensitizers that, to date, is most suitable for ZnO, offering an incident photon-to-current efficiency (IPCE) as high as 69% at 510 nm and overall conversion efficiency of 2.5%.[175, 176] It was also reported that mercurochrome photosensitizer could provide ZnO DSSCs with a fill factor significantly larger than what obtained with N3 dye, where the latter device was believed possess a higher degree of interfacial electron recombination due to the higher surface trap density in the N3 dye-adsorbed ZnO.[177] Eosin Y is also a very efficient dye to ZnO-based DSSCs, with 1.11% conversion efficiency for nanocrystalline films.[52] When eosin Y is combined with nanoporous film, overall
ZnO Dye-Sensitized Solar Cells
423
conversion efficiencies of 2.0-2.4% have been obtained.[178, 179] Recently, Senevirathne et al. reported that acriflavine (1,6diamino-10methylacridinium chloride) as photosensitizer for ZnO could generate photocurrents that are an order of magnitude higher than in the case of TiO2.[180] As an alternative of organic dyes, semiconductor quantum dots have also been studied as photosensitizers for application in DSSCs, which are accordingly sometimes called semiconductor-sensitized solar cells (SSSCs). Quantum dots take the form of nanocrystals of compound semiconductors and therefore, offer the advantage of stability over the organometallic or even pure organic dyes. Moreover, quantum dots present the ability to match solar spectrum better because their absorption wavelength range can be tailored by size quantization. More recently, the quantum dots were reported the ability to generate multiple electron-hole pairs per photon, which would be greatly beneficial to the photovoltaic devices with extremely high conversion efficiency.[181, 182] Up to now, the highest overall conversion efficiency of TiO2 sensitized with quantum dots is about 2.8%.[183, 184] A few of attempts are also made on ZnO. One example is to combine ZnO nanowires with CdSe quantum dots for photoelectrochemical cell, which has demonstrated an internal quantum efficiency of 50-60%, comparable to the results obtained for similar ZnO nanowires sensitized with Ru-complex dye.[185]
7. Conclusion and Outlook Nanostructures in the instance of ZnO have been demonstrated to be advantageous for use as the photoelectrode film in DSSCs. To review, these nanostructures offer large specific surface areas for dye adsorption, direct pathways for electron transport, and light scattering effects that extend the traveling distance of light within the photoelectrode film. Core-shell structures typically with ZnO shell on TiO2 core is demonstrated for reduced recombination rate. Table 2 summarizes the majority of recent results obtained for DSSCs based on ZnO nanostructures. Among these nanostructures, ZnO aggregates feature a structure with the aggregation of nano-sized crystallites and thus combine a large surface area with efficient light scattering centers. As a
424
Table 2. Summary of DSSCs based on ZnO nanostructures. Structure
Nanoparticles
Photosensitizer
Efficiency
Ref.
Structure
Photosensitizer
Efficiency
Ref.
N719
0.44%, 2.1%, 2.22%
[54, 55, 96]
Nanotips
N719
0.55%, 0.77%
[106, 107]
N719
5% (0.1 sun)
[56, 155]
Nanotubes
N719
1.6%, 2.3%
[104, 105]
N3
0.4%, 0.75%, 3.4%
[53, 62, 76]
Nanobelts
N719
2.6%
[78]
heptamethine cyanine
0.16%, 0.67
[172, 173]
N719
2.61%, 3.3%
[74, 77]
N3
1.55%
[76]
Nanosheets
Nanoporous films
1.5%
[174]
eosin-Y
1.11%
[52]
Nano-tetrapods
N719
1.20%, 3.27%
[79, 80]
acriflavine
0.588%
[180]
Nanoflowers
N719
1.9%
[108]
mercurochrome
2.5%
[175, 176]
Core-shell nanoparticles (ZnO-TiO2)
N719
1.78%
[160]
N3
5.08%
[69]
Core-shell nanoparticles (ZnO-SiO2)
N719
5.2%
[159]
N719
4.1%
[73]
Core-shell nanoparticles (TiO2-ZnO)
N3
4.3%, 4.51%, 9.8%
[115, 126, 127]
Zhang and Cao
unsymmetrical squaraine
[49]
N719
6.55%
[130]
eosin-Y
2.0%, 2.4%
[178, 179]
N3
1.21% (0.2 sun, flexible substrate)
[124]
eosin-Y
3.31% (0.1 sun)
[178]
Core-shell nanoporous films (ZnO-TiO2)
N719
1.02%
[48]
N3
0.73%, 2.1%, 2.4%, 4.7%
[94, 98, 150, 177]
Core-shell nanowires (ZnO-TiO2)
N719
2.25%
[163]
Nanowires
N719
0.3%, 0.6%, 0.9%, 1.5%, 1.54%
[91, 96, 97, 99, 100]
Core-shell nanowires (ZnO-Al2O3)
N719
↓
[163]
QDs (CdSe)
0.4%
[185]
Core-shell nanotubes (TiO2-ZnO)
N719
0.704%
[161]
Dendritic nanowires
N719
0.5%, 1.1%
[84, 109]
Aggregates
N3
3.51%, 4.4%, 5.4%
[147-149, 153]
Nanowire/nan oparticle composite films
N719
0.9% (flexible substrate)
[102]
Mercurochrome
2.2%
[101]
ZnO Dye-Sensitized Solar Cells
0.23% (hybrid ZnO/N719)
N719
425
426
Zhang and Cao
result, ZnO aggregates possess the highest conversion efficiency of 5.4%, which is more than double to 2.4% obtained for dispersed ZnO nanocrystallites. This is desirable scenario found in ZnO has indicated that the performance of DSSCs bansed on nanoparticulate films can be further improved via the route of aggregates. As such, it would be anticipated that this approach can also work for nanocrystalline TiO2 films, which have already achieved the maximum conversion efficiency of around 11%. However, when a similar strategy is employed with TiO2, it is hampered by the availability of synthesis methods in the production of appropriate nanostructures. Although the mechanism of growth in ZnO aggregates has been proposed to be a result of the dipole nature of ZnO nanocrystallites in a supersaturation colloidal solution,[186] it seems to be unsuitable for TiO2. As an outlook, we would like to propose several possible methods that are intended for the synthesis of TiO2 aggregates, including 1) surface modification of ZnO aggregates using TiO2, 2) hydrothermal growth of TiO2 nanoparticle aggregates, 3) emulsionassisted aggregation of TiO2 nanostructures, 4) electrostatic spray deposition using colloidal TiO2, and 5) synthesis of porous-structured TiO2 spheres.
7.1. Surface Modification of ZnO Aggregates – An Indirect Method for TiO2 Aggregates To modify ZnO aggregates with TiO2 is a consideration that can directly utilize the already achieved aggregate structure of ZnO serving as light scattering centers and meanwhile attain a TiO2 surface for dye adsorption. Many physical or chemical deposition methods can be used for this purpose, however the recently ALD technique distinguishes itself by offering atomic level thickness as well as large area uniformity.[187] ALD is actually a self-limiting process of vapor-solid deposition, by which film formation takes place in a cyclic manner through a series of saturative surface reactions between the adsorbed precursor and the species left on the surface, and moreover, multilayer adsorption is, by definition, excluded. In this method, film thickness can be precisely controlled by adjusting the deposition cycles and growth rate.[163] As for
427
ZnO Dye-Sensitized Solar Cells
the deposition of TiO2, many titanium alkoxides such as titanium chloride (TiCl4),[188] titanium isopropoxide (Ti(OiPr)4),[189, 190] titanium ethoxide (Ti(OEt)4),[191, 192] and titanium methoxide (Ti(OMe)4)[193] can be adopted as the titanium precursor due to the fact that they are liquid with a moderate vapor pressure convenient to handle. It has been reported that both the crystal structure and phase of TiO2 are predominated by the deposition temperature.[193] For example, when using Ti(OMe)4 as the titanium precursor, the TiO2 films typically deposited at 200°C are amorphous, whereas those deposited at 250°C or above are polycrystalline with anatase phase.[194]
7.2. Hydrothermal Growth of TiO2 Nanoparticle Aggregates
(a)
(b)
Figure 12. The morphology and structure of flower-like TiO2 nanostructures.[195] (a) SEM image of spherical aggregates with an average diameter of ~300 nm, and (b) TEM image revealing the aggregates that consist of needle-like nanoparticles.
A recent study reported on the synthesis of flower-like TiO2 nanostructures by a hydrothermal treatment of titanium tetrabutoxide (Ti(OBu)4) aqueous solution under strongly acidic condition.[195] These flower-like nanostructures were aggregated by needle-like TiO2 nanoparticles and possessed a spherical shape with a size of about 300 nm (Fig. 12). These structures appear to be promise as efficient light scatterers for DSSCs. The formation of flower-like nanostructures has been ascribed to the hydrolysis of Ti(OBu)4 with high concentration, which leads to a large amount of needle-like nanoparticles of
428
Zhang and Cao
Ti(OC4H9)4-n(OH)n. These nanoparticles are homogenously adsorbed at the active sites around the surface of TiO2 nuclei generated at the primary stage of hydrolysis, resulting in a radial growth from the center. It was mentioned that a major proportion of the as-synthesized nanostructures were rutile-phase TiO2 due to the strongly acidic condition in the presence of Cl-.[196]
7.3. Emulsion-Assisted Synthesis of TiO2 Nanostructure Aggregation Emulsion-assisted synthesis is a facile route to creating colloidal aggregates by means of water-in-oil emulsion droplets. In this technique, the colloidal nanoparticles are encapsulated in water droplets that provide confined geometries. When the water is removed from the droplets by drying, the colloidal nanoparticles assemble spontaneously to form spherical aggregates (Fig. 13).[197-201] The size of the aggregates can be readily controlled by tuning the tip size of the micropipette that emulsifies the suspension or changing the weight fraction of nanoparticles in suspension. Kim et al. illustrated the preparation of TiO2 microspheres by using the emulsion-assisted method.[197] It was demonstrated that the size of the microspheres, dm, could be estimated by a relation d m ∝ d tφw1 3 , where dt is the inner diameter of the micropipette, and φw is the weight fraction of TiO2 nanoparticles in suspension.
Figure 13. Schematic of the emulsion-assisted method for the synthesis of spherical TiO2 aggregates with nanoparticles. (TRITC: tetramethylrhodamine isothiocyanate)[197]
ZnO Dye-Sensitized Solar Cells
429
Compared with other methods of preparing colloidal aggregates, the emulsion-assisted method possesses the advantage that it does not have any requirement towards the properties of the constituent materials provided that they are dispersed in a liquid medium on the nanometer scale in a stable manner. The technological importance of this method is its viability for the spherical aggregation of other TiO2 nanomaterials, such as TiO2 nanotubes, which exhibit a huge potential for highly efficient DSSCs.[202, 203]
7.4. Electrostatic Spray Deposition Fabrication of TiO2 Aggregates Electrostatic spray deposition (ESD) is a process in which the colloidal solution (containing nanoparticles, binder, and solvent) is atomized into charged droplets through a capillary needle by an electrohydrodynamic force and the film is formed on the substrate with a heat treatment for solvent evaporation (See Sec. 2.1.2.).[62, 204] This technique allows to package nanoparticles into droplets, and the size of droplets can be adjusted in a wide range from a few nanometers up to hundreds of micrometers. It is therefore a promising method that may be employed for the fabrication of TiO2 aggregates consisting of nanocrystallites.
7.5. Synthesis of Porous-Structured TiO2 Spheres Porous-structured TiO2 spheres with sub-micrometer sizes have a structure similar to that of nanocrystallite aggregates, thus enable efficient light scattering and large surface areas to be attained simultaneously. The synthesis of porous-structured TiO2 spheres can be achieved by using a template-free method via a sol-gel process combined with the inverse miniemulsion technique. In this method, droplets are generated by a dispersed aqueous phase that consists of a precursor material of titania mixed with a continuous organic phase containing surfactant. During the hydrolysis and condensation, each droplet acts as a nanoreactor for the encapsulation of colloidal TiO2. Unlike conventional emulsion in which diffusion processes take place and lead to an exchange of reactants, a miniemulsion is stabilized against diffusion and
430
Zhang and Cao
coalescence so that both the droplet size and size distribution can be well controlled.[205] Rossmanith et al. reported on a fabrication of porous-structured TiO2 by using glycol-modified titanate (bis(2hydroxyethyl)titanate, EGMT) as the precursor material and amphiphilic block copolymer (P(E/B-b-EO)) as the surfactant.[206] The size of the synthesized TiO2 spheres is typically about 200 nm in diameter. Impressively, these spheres are formed by interconnected nanocrystallites with the size of several nanometers, leading to a high specific surface area of more than 300 m2/g.
Acknowledgments This work is supported in part by US Department of Energy (DEFG02-07ER46467), Air Force Office of Scientific Research (AFOSRMURI, FA9550-06-1-032), and National Science Foundation (DMI0455994 and DMR-0605159). This research is also funded by grants from Washington Research Foundation, Washington Technology Center, Intel Corporation, and EnerG2.
References 1. 2. 3. 4. 5. 6. 7. 8. 9.
10.
J. Liu, G. Z. Cao, Z. Yang, D. Wang, D. Dubois, X. Zhou, G. L. Graff, L. R. Pederson, J.-G. Zhang, ChemSusChem 2008, 1, 22. D. E. Carlson, C. R. Wronski, Applied Physics Letters 1976, 28, 671. A. Goetzberger, C. Hebling, H. W. Schock, Materials Science & Engineering R-Reports 2003, 40, 1. A. Goetzberger, C. Hebling, Solar Energy Materials and Solar Cells 2000, 62, 1. J. S. Ward, K. Ramanathan, F. S. Hasoon, T. J. Coutts, J. Keane, M. A. Contreras, T. Moriarty, R. Noufi, Progress in Photovoltaics 2002, 10, 41. K. L. Chopra, P. D. Paulson, V. Dutta, Progress in Photovoltaics 2004, 12, 69. M. Gratzel, Nature 2001, 414, 338. J. Bisquert, D. Cahen, G. Hodes, S. Ruhle, A. Zaban, Journal of Physical Chemistry B 2004, 108, 8106. J. M. Kroon, N. J. Bakker, H. J. P. Smit, P. Liska, K. R. Thampi, P. Wang, S. M. Zakeeruddin, M. Gratzel, A. Hinsch, S. Hore, U. Wurfel, R. Sastrawan, J. R. Durrant, E. Palomares, H. Pettersson, T. Gruszecki, J. Walter, K. Skupien, G. E. Tulloch, Progress in Photovoltaics 2007, 15, 1. C. Longo, M. A. De Paoli, Journal of the Brazilian Chemical Society 2003, 14, 889.
ZnO Dye-Sensitized Solar Cells 11. 12. 13. 14.
15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
431
A. B. F. Martinson, T. W. Hamann, M. J. Pellin, J. T. Hupp, Chemistry-a European Journal 2008, 14, 4458. M. Gratzel, Progress in Photovoltaics 2000, 8, 171. B. Oregan, M. Gratzel, Nature 1991, 353, 737. M. K. Nazeeruddin, P. Pechy, T. Renouard, S. M. Zakeeruddin, R. Humphry-Baker, P. Comte, P. Liska, L. Cevey, E. Costa, V. Shklover, L. Spiccia, G. B. Deacon, C. A. Bignozzi, M. Gratzel, Journal of the American Chemical Society 2001, 123, 1613. M. Gratzel, Journal of Photochemistry and Photobiology C-Photochemistry Reviews 2003, 4, 145. M. Gratzel, Philosophical Transactions of the Royal Society a-Mathematical Physical and Engineering Sciences 2007, 365, 993. J. Nelson, R. E. Chandler, Coordination Chemistry Reviews 2004, 248, 1181. J. Nissfolk, K. Fredin, A. Hagfeldt, G. Boschloo, Journal of Physical Chemistry B 2006, 110, 17715. M. Gratzel, Inorganic Chemistry 2005, 44, 6841. M. Gratzel, Journal of Photochemistry and Photobiology a-Chemistry 2004, 164, 3. Z. L. Wang, Journal of Physics-Condensed Matter 2004, 16, R829. L. Schmidt-Mende, J. L. MacManus-Driscoll, Materials Today 2007, 10, 40. J. Tornow, K. Schwarzburg, Journal of Physical Chemistry C 2007, 111, 8692. A. B. Djurisic, Y. H. Leung, Small 2006, 2, 944. U. Pal, J. G. Serrano, P. Santiago, G. Xiong, K. B. Ucer, R. T. Williams, Optical Materials 2006, 29, 65. M. L. Curri, R. Comparelli, P. D. Cozzoli, G. Mascolo, A. Agostiano, Materials Science & Engineering C-Biomimetic and Supramolecular Systems 2003, 23, 285. J. Bandara, K. Tennakone, P. P. B. Jayatilaka, Chemosphere 2002, 49, 439. J. Tornow, K. Ellmer, J. Szarko, K. Schwarzburg, Thin Solid Films 2008, 516, 7139. U. Ozgur, Y. I. Alivov, C. Liu, A. Teke, M. A. Reshchikov, S. Dogan, V. Avrutin, S. J. Cho, H. Morkoc, Journal of Applied Physics 2005, 98. H. Tang, K. Prasad, R. Sanjines, P. E. Schmid, F. Levy, Journal of Applied Physics 1994, 75, 2042. G. Oskam, Z. S. Hu, R. L. Penn, N. Pesika, P. C. Searson, Physical Review E 2002, 66. H. S. Bae, M. H. Yoon, J. H. Kim, S. Im, Applied Physics Letters 2003, 83, 5313. K. Keis, A. Roos, Optical Materials 2002, 20, 35. V. Noack, H. Weller, A. Eychmuller, Journal of Physical Chemistry B 2002, 106, 8514. J. J. Wu, G. R. Chen, C. C. Lu, W. T. Wu, J. S. Chen, Nanotechnology 2008, 19. A. S. Arico, P. Bruce, B. Scrosati, J. M. Tarascon, W. Van Schalkwijk, Nature Materials 2005, 4, 366. Y. N. Xia, P. D. Yang, Y. G. Sun, Y. Y. Wu, B. Mayers, B. Gates, Y. D. Yin, F. Kim, Y. Q. Yan, Advanced Materials 2003, 15, 353.
432
Zhang and Cao
38. 39. 40. 41. 42.
K. Bittkau, R. Carius, Superlattices and Microstructures 2007, 42, 47. E. A. Meulenkamp, Journal of Physical Chemistry B 1998, 102, 5566. M. Vafaee, M. S. Ghamsari, Materials Letters 2007, 61, 3265. S. Kar, A. Dev, S. Chaudhuri, Journal of Physical Chemistry B 2006, 110, 17848. P. D. Yang, H. Q. Yan, S. Mao, R. Russo, J. Johnson, R. Saykally, N. Morris, J. Pham, R. R. He, H. J. Choi, Advanced Functional Materials 2002, 12, 323. Q. C. Li, V. Kumar, Y. Li, H. T. Zhang, T. J. Marks, R. P. H. Chang, Chemistry of Materials 2005, 17, 1001. X. D. Wang, Y. Ding, C. J. Summers, Z. L. Wang, Journal of Physical Chemistry B 2004, 108, 8773. M. Fu, J. Zhou, Q. F. Xiao, B. Li, R. L. Zong, W. Chen, J. Zhang, Advanced Materials 2006, 18, 1001. M. Wang, C. H. Ye, Y. Zhang, G. M. Hua, H. X. Wang, M. G. Kong, L. D. Zhang, Journal of Crystal Growth 2006, 291, 334. H. H. Wang, C. S. Xie, Journal of Crystal Growth 2006, 291, 187. Y. Y. Xi, Y. F. Hsu, A. B. Djurisic, W. K. Chan, Journal of the Electrochemical Society 2008, 155, D595. K. Nonomura, T. Yoshida, D. Schlettwein, H. Minoura, Electrochimica Acta 2003, 48, 3071. L. Spanhel, M. A. Anderson, Journal of the American Chemical Society 1991, 113, 2826. K. Keis, L. Vayssieres, H. Rensmo, S. E. Lindquist, A. Hagfeldt, Journal of the Electrochemical Society 2001, 148, A149. S. Rani, P. Suri, P. K. Shishodia, R. M. Mehra, Solar Energy Materials and Solar Cells 2008, 92, 1639. G. Redmond, D. Fitzmaurice, M. Graetzel, Chemistry of Materials 1994, 6, 686. L. Y. Zeng, S. Y. Dai, W. W. Xu, K. J. Wang, Plasma Science & Technology 2006, 8, 172. K. Keis, J. Lindgren, S. E. Lindquist, A. Hagfeldt, Langmuir 2000, 16, 4688. K. Keis, E. Magnusson, H. Lindstrom, S. E. Lindquist, A. Hagfeldt, Solar Energy Materials and Solar Cells 2002, 73, 51. Note that the employed illumination intensity in this work is 10 mW/cm2, which is one tenth of a standard intensity of 100 mW/cm2 conventionally used elsewhere. A weaker illumination maybe results in slightly boosted conversion efficiency. A. I. Kontos, A. G. Kontos, D. S. Tsoukleris, M. C. Bernard, N. Spyrellis, P. Falaras, Journal of Materials Processing Technology 2008, 196, 243. K. Keis, L. Vayssieres, S. E. Lindquist, A. Hagfeldt, Nanostructured Materials 1999, 12, 487. S. Leeuwenburgh, J. Wolke, J. Schoonman, J. Jansen, Journal of Biomedical Materials Research Part A 2003, 66A, 330. M. C. Siebers, X. F. Walboomers, S. C. G. Leeuwenburgh, J. G. C. Wolke, J. A. Jansen, Biomaterials 2004, 25, 2019.
43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57.
58. 59. 60. 61.
ZnO Dye-Sensitized Solar Cells 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75.
76. 77. 78. 79. 80.
81. 82. 83. 84. 85. 86. 87.
433
X. Sheng, Y. Zhao, J. Zhai, L. Jiang, D. Zhu, Applied Physics a-Materials Science & Processing 2007, 87, 715. M. Y. Song, D. K. Kim, K. J. Ihn, S. M. Jo, D. Y. Kim, Nanotechnology 2004, 15, 1861. R. Jose, A. Kumar, V. Thavasi, S. Ramakrishna, Nanotechnology 2008, 19. P. K. Basu, N. Saha, S. Maji, H. Saha, S. Basu, Journal of Materials ScienceMaterials in Electronics 2008, 19, 493. B. Pal, M. Sharon, Materials Chemistry and Physics 2002, 76, 82. G. R. Li, C. R. Dawa, Q. Bu, X. H. Lu, Z. H. Ke, H. E. Hong, F. L. Zheng, C. Z. Yao, G. K. Liu, Y. X. Tong, Journal of Physical Chemistry C 2007, 111, 1919. E. Michaelis, D. Wohrle, J. Rathousky, M. Wark, Thin Solid Films 2006, 497, 163. Z. G. Chen, Y. W. Tang, L. S. Zhang, L. J. Luo, Electrochimica Acta 2006, 51, 5870. T. P. Niesen, M. R. De Guire, Journal of Electroceramics 2001, 6, 169. K. Kakiuchi, E. Hosono, T. Kimura, H. Imai, S. Fujihara, Journal of Sol-Gel Science and Technology 2006, 39, 63. E. Hosono, S. Fujihara, T. Kimura, H. Imai, Journal of Colloid and Interface Science 2004, 272, 391. K. Kakiuchi, E. Hosono, S. Fujihara, Journal of Photochemistry and Photobiology a-Chemistry 2006, 179, 81. K. Kakiuchi, M. Saito, S. Fujihara, Thin Solid Films 2008, 516, 2026. M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphrybaker, E. Muller, P. Liska, N. Vlachopoulos, M. Gratzel, Journal of the American Chemical Society 1993, 115, 6382. A. E. Suliman, Y. W. Tang, L. Xu, Solar Energy Materials and Solar Cells 2007, 91, 1658. M. S. Akhtar, M. A. Khan, M. S. Jeon, O. B. Yang, Electrochimica Acta 2008, 53, 7869. C. F. Lin, H. Lin, J. B. Li, X. Li, Journal of Alloys and Compounds 2008, 462, 175. Y. F. Hsu, Y. Y. Xi, C. T. Yip, A. B. Djurisic, W. K. Chan, Journal of Applied Physics 2008, 103. W. Chen, H. F. Zhang, I. M. Hsing, S. H. Yang, A New Photoanode Architecture of Dye Sensitized Solar Cell Based on ZnO Nanotetrapods, 2008, The Hong Kong University of Science and Technology (In press). H. Q. Yan, R. R. He, J. Pham, P. D. Yang, Advanced Materials 2003, 15, 402. K. D. Benkstein, N. Kopidakis, J. van de Lagemaat, A. J. Frank, Journal of Physical Chemistry B 2003, 107, 7759. A. Hagfeldt, M. Gratzel, Chemical Reviews 1995, 95, 49. J. B. Baxter, E. S. Aydil, Solar Energy Materials and Solar Cells 2006, 90, 607. W. Lee, M. C. Jeong, J. M. Myoung, Acta Materialia 2004, 52, 3949. A. Umar, B. K. Kim, J. J. Kim, Y. B. Hahn, Nanotechnology 2007, 18. Y. Tak, K. J. Yong, Journal of Physical Chemistry B 2005, 109, 19263.
434
Zhang and Cao
88.
J. Cembrero, A. Elmanouni, B. Hartiti, M. Mollar, B. Mari, Thin Solid Films 2004, 451-52, 198. C. H. Liu, J. A. Zapien, Y. Yao, X. M. Meng, C. S. Lee, S. S. Fan, Y. Lifshitz, S. T. Lee, Advanced Materials 2003, 15, 838. M. Krunks, A. Katerski, T. Dedova, I. O. Acik, A. Mere, Solar Energy Materials and Solar Cells 2008, 92, 1016. M. Law, L. E. Greene, J. C. Johnson, R. Saykally, P. D. Yang, Nature Materials 2005, 4, 455. L. E. Greene, B. D. Yuhas, M. Law, D. Zitoun, P. D. Yang, Inorganic Chemistry 2006, 45, 7535. N. Kopidakis, K. D. Benkstein, J. van de Lagemaat, A. J. Frank, Journal of Physical Chemistry B 2003, 107, 11307. M. Guo, P. Diao, S. M. Cai, Applied Surface Science 2005, 249, 71. J. H. Noh, S. H. Lee, S. Lee, H. S. Jung, Electronic Materials Letters 2008, 4, 71. K. S. Kim, Y. S. Kang, J. H. Lee, Y. J. Shin, N. G. Park, K. S. Ryu, S. H. Chang, Bulletin of the Korean Chemical Society 2006, 27, 295. J. B. Baxter, A. M. Walker, K. van Ommering, E. S. Aydil, Nanotechnology 2006, 17, S304. M. Guo, P. Diao, X. D. Wang, S. M. Cai, Journal of Solid State Chemistry 2005, 178, 3210. A. J. Cheng, Y. H. Tzeng, Y. Zhou, M. Park, T. H. Wu, C. Shannon, D. Wang, W. W. Lee, Applied Physics Letters 2008, 92. R. Zhang, S. Kumar, S. Zou, L. L. Kerr, Crystal Growth & Design 2008, 8, 381. C. H. Ku, J. J. Wu, Applied Physics Letters 2007, 91. C. Y. Jiang, X. W. Sun, K. W. Tan, G. Q. Lo, A. K. K. Kyaw, D. L. Kwong, Applied Physics Letters 2008, 92. L. Vayssieres, K. Keis, A. Hagfeldt, S. E. Lindquist, Chemistry of Materials 2001, 13, 4395. M. Guo, P. Diao, S. M. Cai, Chinese Chemical Letters 2004, 15, 1113. A. B. F. Martinson, J. W. Elam, J. T. Hupp, M. J. Pellin, Nano Letters 2007, 7, 2183. A. Du Pasquier, H. H. Chen, Y. C. Lu, Applied Physics Letters 2006, 89. H. H. Chen, A. Du Pasquier, G. Saraf, J. Zhong, Y. Lu, Semiconductor Science and Technology 2008, 23. C. Y. Jiang, X. W. Sun, G. Q. Lo, D. L. Kwong, J. X. Wang, Applied Physics Letters 2007, 90. J. B. Baxter, E. S. Aydil, Applied Physics Letters 2005, 86. N. G. Park, M. G. Kang, K. M. Kim, K. S. Ryu, S. H. Chang, D. K. Kim, J. van de Lagemaat, K. D. Benkstein, A. J. Frank, Langmuir 2004, 20, 4246. Y. Diamant, S. Chappel, S. G. Chen, O. Melamed, A. Zaban, Coordination Chemistry Reviews 2004, 248, 1271.
89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111.
ZnO Dye-Sensitized Solar Cells
435
112. D. B. Menzies, Q. Dai, L. Bourgeois, R. A. Caruso, Y. B. Cheng, G. P. Simon, L. Spiccia, Nanotechnology 2007, 18. 113. J. C. Guo, C. X. She, T. Q. Lian, Journal of Physical Chemistry C 2007, 111, 8979. 114. S. G. Chen, S. Chappel, Y. Diamant, A. Zaban, Chemistry of Materials 2001, 13, 4629. 115. Z. S. Wang, C. H. Huang, Y. Y. Huang, Y. J. Hou, P. H. Xie, B. W. Zhang, H. M. Cheng, Chemistry of Materials 2001, 13, 678. 116. K. Tennakone, P. V. V. Jayaweera, P. K. M. Bandaranayake, Journal of Photochemistry and Photobiology a-Chemistry 2003, 158, 125. 117. Y. Diamant, S. G. Chen, O. Melamed, A. Zaban, Journal of Physical Chemistry B 2003, 107, 1977. 118. E. Palomares, J. N. Clifford, S. A. Haque, T. Lutz, J. R. Durrant, Journal of the American Chemical Society 2003, 125, 475. 119. N. Okada, S. Karuppuchamy, M. Kurihara, Chemistry Letters 2005, 34, 16. 120. Z. S. Wang, M. Yanagida, K. Sayama, H. Sugihara, Chemistry of Materials 2006, 18, 2912. 121. P. Cheng, C. S. Deng, X. M. Dai, B. Li, D. N. Liu, J. M. Xu, Journal of Photochemistry and Photobiology a-Chemistry 2008, 195, 144. 122. S. Wu, H. W. Han, Q. D. Tai, J. Zhang, S. Xu, C. H. Zhou, Y. Yang, H. Hu, B. L. Chen, B. Sebo, X. Z. Zhao, Nanotechnology 2008, 19. 123. S. J. Wu, H. W. Han, Q. D. Tai, J. Zhang, S. Xu, C. H. Zhou, Y. Yang, H. Hu, B. L. Chen, X. Z. Zhao, Journal of Power Sources 2008, 182, 119. 124. S. S. Kim, J. H. Yum, Y. E. Sung, Journal of Photochemistry and Photobiology a-Chemistry 2005, 171, 269. 125. R. S. Mane, W. J. Lee, H. M. Pathan, S. H. Han, Journal of Physical Chemistry B 2005, 109, 24254. 126. K. E. Kim, S. R. Jang, J. Park, R. Vittal, K. J. Kim, Solar Energy Materials and Solar Cells 2007, 91, 366. 127. S. J. Roh, R. S. Mane, S. K. Min, W. J. Lee, C. D. Lokhande, S. H. Han, Applied Physics Letters 2006, 89. 128. K. Tennakone, J. Bandara, P. K. M. Bandaranayake, G. R. A. Kumara, A. Konno, Japanese Journal of Applied Physics Part 2-Letters 2001, 40, L732. 129. B. Enright, D. Fitzmaurice, Journal of Physical Chemistry 1996, 100, 1027. 130. S. J. Wu, H. W. Han, Q. D. Tai, J. Zhang, B. L. Chen, S. Xu, C. H. Zhou, Y. Yang, H. Hu, X. Z. Zhao, Applied Physics Letters 2008, 92. 131. J. Bandara, H. C. Weerasinghe, Solar Energy Materials and Solar Cells 2005, 88, 341. 132. A. Solbrand, K. Keis, S. Sodergren, H. Lindstrom, S. E. Lindquist, A. Hagfeldt, Solar Energy Materials and Solar Cells 2000, 60, 181. 133. A. Solbrand, H. Lindstrom, H. Rensmo, A. Hagfeldt, S. E. Lindquist, S. Sodergren, Journal of Physical Chemistry B 1997, 101, 2514.
436
Zhang and Cao
134. K. M. P. Bandaranayake, M. K. I. Senevirathna, P. Weligamuwa, K. Tennakone, Coordination Chemistry Reviews 2004, 248, 1277. 135. C. Rockstuhl, F. Lederer, K. Bittkau, R. Carius, Applied Physics Letters 2007, 91. 136. A. Usami, Chemical Physics Letters 1997, 277, 105. 137. J. Ferber, J. Luther, Solar Energy Materials and Solar Cells 1998, 54, 265. 138. G. Rothenberger, P. Comte, M. Gratzel, Solar Energy Materials and Solar Cells 1999, 58, 321. 139. H. J. Koo, J. Park, B. Yoo, K. Yoo, K. Kim, N. G. Park, Inorganica Chimica Acta 2008, 361, 677. 140. C. Anderson, A. J. Bard, Journal of Physical Chemistry B 1997, 101, 2611. 141. C. J. Barbe, F. Arendse, P. Comte, M. Jirousek, F. Lenzmann, V. Shklover, M. Gratzel, Journal of the American Ceramic Society 1997, 80, 3157. 142. S. Hore, P. Nitz, C. Vetter, C. Prahl, M. Niggemann, R. Kern, Chemical Communications 2005, 2011. 143. W. W. So, K. J. Kim, J. K. Lee, S. J. Moon, Japanese Journal of Applied Physics Part 1-Regular Papers Short Notes & Review Papers 2004, 43, 1231. 144. L. Yang, Y. Lin, J. G. Jia, X. R. Xiao, X. P. Li, X. W. Zhou, Journal of Power Sources 2008, 182, 370. 145. S. Nishimura, N. Abrams, B. A. Lewis, L. I. Halaoui, T. E. Mallouk, K. D. Benkstein, J. van de Lagemaat, A. J. Frank, Journal of the American Chemical Society 2003, 125, 6306. 146. L. I. Halaoui, N. M. Abrams, T. E. Mallouk, Journal of Physical Chemistry B 2005, 109, 6334. 147. T. P. Chou, Q. F. Zhang, G. E. Fryxell, G. Z. Cao, Advanced Materials 2007, 19, 2588. 148. Q. F. Zhang, T. R. Chou, B. Russo, S. A. Jenekhe, G. Z. Cao, Angewandte ChemieInternational Edition 2008, 47, 2402. 149. Q. F. Zhang, T. P. Chou, B. Russo, S. A. Jenekhe, G. Z. Cao, Advanced Functional Materials 2008, 18, 1654. 150. A. R. Rao, V. Dutta, Nanotechnology 2008, 19. 151. H. Horiuchi, R. Katoh, K. Hara, M. Yanagida, S. Murata, H. Arakawa, M. Tachiya, Journal of Physical Chemistry B 2003, 107, 2570. 152. K. Westermark, H. Rensmo, H. Siegbahn, K. Keis, A. Hagfeldt, L. Ojamae, P. Persson, Journal of Physical Chemistry B 2002, 106, 10102. 153. T. P. Chou, Q. F. Zhang, G. Z. Cao, Journal of Physical Chemistry C 2007, 111, 18804. 154. R. Katoh, A. Furube, Y. Tamaki, T. Yoshihara, M. Murai, K. Hara, S. Murata, H. Arakawa, M. Tachiya, Journal of Photochemistry and Photobiology aChemistry 2004, 166, 69. 155. K. Keis, C. Bauer, G. Boschloo, A. Hagfeldt, K. Westermark, H. Rensmo, H. Siegbahn, Journal of Photochemistry and Photobiology a-Chemistry 2002, 148, 57. 156. D. W. Bahnemann, Israel Journal of Chemistry 1993, 33, 115.
ZnO Dye-Sensitized Solar Cells
437
157. P. Persson, L. Ojamae, Chemical Physics Letters 2000, 321, 302. 158. R. Zhang, J. Pan, E. P. Briggs, M. Thrash, L. L. Kerr, Solar Energy Materials and Solar Cells 2008, 92, 425. 159. Y. J. Shin, J. H. Lee, J. H. Park, N. G. Park, Chemistry Letters 2007, 36, 1506. 160. Y. J. Shin, K. S. Kim, N. G. Park, K. S. Ryu, S. H. Chang, Bulletin of the Korean Chemical Society 2005, 26, 1929. 161. S. H. Kang, J. Y. Kim, Y. Kim, H. S. Kim, Y. E. Sung, Journal of Physical Chemistry C 2007, 111, 9614. 162. L. E. Greene, M. Law, B. D. Yuhas, P. D. Yang, Journal of Physical Chemistry C 2007, 111, 18451. 163. M. Law, L. E. Greene, A. Radenovic, T. Kuykendall, J. Liphardt, P. D. Yang, Journal of Physical Chemistry B 2006, 110, 22652. 164. N. A. Anderson, X. Ai, T. Q. Lian, Journal of Physical Chemistry B 2003, 107, 14414. 165. J. B. Asbury, E. Hao, Y. Q. Wang, H. N. Ghosh, T. Q. Lian, Journal of Physical Chemistry B 2001, 105, 4545. 166. J. B. Asbury, N. A. Anderson, E. C. Hao, X. Ai, T. Q. Lian, Journal of Physical Chemistry B 2003, 107, 7376. 167. D. Kuciauskas, J. E. Monat, R. Villahermosa, H. B. Gray, N. S. Lewis, J. K. McCusker, Journal of Physical Chemistry B 2002, 106, 9347. 168. G. Benko, J. Kallioinen, J. E. I. Korppi-Tommola, A. P. Yartsev, V. Sundstrom, Journal of the American Chemical Society 2002, 124, 489. 169. J. B. Asbury, Y. Q. Wang, T. Q. Lian, Journal of Physical Chemistry B 1999, 103, 6643. 170. N. A. Anderson, T. Lian, Coordination Chemistry Reviews 2004, 248, 1231. 171. A. Furube, R. Katoh, K. Hara, S. Murata, H. Arakawa, M. Tachiya, Journal of Physical Chemistry B 2003, 107, 4162. 172. M. Matsui, Y. Hashimoto, K. Funabiki, J. Y. Jin, T. Yoshida, H. Minoura, Synthetic Metals 2005, 148, 147. 173. A. Otsuka, K. Funabiki, N. Sugiyama, H. Mase, T. Yoshida, H. Minoura, M. Matsui, Chemistry Letters 2008, 37, 176. 174. A. Otsuka, K. Funabiki, N. Sugiyama, T. Yoshida, Chemistry Letters 2006, 35, 666. 175. K. Hara, T. Horiguchi, T. Kinoshita, K. Sayama, H. Sugihara, H. Arakawa, Chemistry Letters 2000, 316. 176. K. Hara, T. Horiguchi, T. Kinoshita, K. Sayama, H. Sugihara, H. Arakawa, Solar Energy Materials and Solar Cells 2000, 64, 115. 177. J. J. Wu, G. R. Chen, H. H. Yang, C. H. Ku, J. Y. Lai, Applied Physics Letters 2007, 90. 178. E. Hosono, S. Fujihara, T. Kimura, Electrochimica Acta 2004, 49, 2287. 179. W. J. Lee, A. Suzuki, K. Imaeda, H. Okada, A. Wakahara, A. Yoshida, Japanese Journal of Applied Physics Part 1-Regular Papers Short Notes & Review Papers 2004, 43, 152.
438
Zhang and Cao
180. M. K. I. Senevirathne, P. Pitigala, V. Sivakumar, P. V. Jayaweera, A. G. U. Perera, K. Tennakone, Journal of Photochemistry and Photobiology a-Chemistry 2008, 195, 364. 181. R. J. Ellingson, M. C. Beard, J. C. Johnson, P. R. Yu, O. I. Micic, A. J. Nozik, A. Shabaev, A. L. Efros, Nano Letters 2005, 5, 865. 182. V. I. Klimov, Journal of Physical Chemistry B 2006, 110, 16827. 183. G. Hodes, Journal of Physical Chemistry C 2008, 112, 17778. 184. A recent paper using CdS on TiO2 nanotubes in sulfide electrolyte reported a cell efficiency of 4.15%. Sun, W.-T.; Yu, Y.; Pan, H.-Y.; Gao, X.-F.; Chen, Q.; Peng, L.-M J. Am. Chem. Soc. 2008, 130, 1124. However, it has been demonstrated to be not a true conversion efficiency since the photovoltage was measured relative to the Ag/AgCl reference electrode rather than to the counter electrode in the sulfide electrolyte. The real cell efficiency would be ca. 1%. (Ref. 183). 185. K. S. Leschkies, R. Divakar, J. Basu, E. Enache-Pommer, J. E. Boercker, C. B. Carter, U. R. Kortshagen, D. J. Norris, E. S. Aydil, Nano Letters 2007, 7, 1793. 186. Y. Y. Tay, S. Li, F. Boey, Y. H. Cheng, M. H. Liang, Physica B-Condensed Matter 2007, 394, 372. 187. R. L. Puurunen, Chemical Vapor Deposition 2005, 11, 79. 188. H. Kumagai, M. Matsumoto, K. Toyoda, M. Obara, M. Suzuki, Thin Solid Films 1995, 263, 47. 189. M. Ritala, M. Leskela, L. Niinisto, P. Haussalo, Chemistry of Materials 1993, 5, 1174. 190. A. Rahtu, M. Ritala, Chemical Vapor Deposition 2002, 8, 21. 191. M. Ritala, M. Leskela, E. Rauhala, Chemistry of Materials 1994, 6, 556. 192. J. Aarik, J. Karlis, H. Mandar, T. Uustare, V. Sammelselg, Applied Surface Science 2001, 181, 339. 193. V. Pore, A. Rahtu, M. Leskela, M. Ritala, D. Sajavaara, J. Keinonen, Chemical Vapor Deposition 2004, 10, 143. 194. N. G. Park, J. van de Lagemaat, A. J. Frank, Journal of Physical Chemistry B 2000, 104, 8989. 195. H. B. Li, X. C. Duan, G. C. Liu, X. B. Jia, X. Q. Liu, Materials Letters 2008, 62, 4035. 196. Y. W. Wang, L. Z. Zhang, K. J. Deng, X. Y. Chen, Z. G. Zou, Journal of Physical Chemistry C 2007, 111, 2709. 197. S. H. Kim, Y. S. Cho, S. J. Jeon, T. H. Eun, G. R. Yi, S. M. Yang, Advanced Materials 2008, 20, 3268. 198. G. R. Yi, V. N. Manoharan, E. Michel, M. T. Elsesser, S. M. Yang, D. J. Pine, Advanced Materials 2004, 16, 1204. 199. Y. S. Cho, G. R. Yi, S. H. Kim, D. J. Pine, S. M. Yang, Chemistry of Materials 2005, 17, 5006. 200. G. R. Yi, V. N. Manoharan, S. Klein, K. R. Brzezinska, D. J. Pine, F. F. Lange, S. M. Yang, Advanced Materials 2002, 14, 1137.
ZnO Dye-Sensitized Solar Cells
439
201. S. H. Kim, S. Y. Lee, G. R. Yi, D. J. Pine, S. M. Yang, Journal of the American Chemical Society 2006, 128, 10897. 202. G. K. Mor, O. K. Varghese, M. Paulose, K. Shankar, C. A. Grimes, Solar Energy Materials and Solar Cells 2006, 90, 2011. 203. B. Tan, Y. Y. Wu, Journal of Physical Chemistry B 2006, 110, 15932. 204. J. H. Moon, G. R. Yi, S. M. Yang, D. J. Pine, S. Bin Park, Advanced Materials 2004, 16, 605. 205. B. Erdem, E. D. Sudol, V. L. Dimonie, M. S. El-Aasser, Journal of Polymer Science Part a-Polymer Chemistry 2000, 38, 4441. 206. R. Rossmanith, C. K. Weiss, J. Geserick, N. Husing, U. Hormann, U. Kaiser, K. Landfester, Chemistry of Materials 2008, 20, 5768.
CHAPTER 10 NANOCOMPOSITES AS HIGH EFFICIENCY THERMOELECTRIC MATERIALS
Suraj Joottu Thiagarajan, Wei Wang and Ronggui Yang* Department of Mechanical Engineering, University of Colorado, Boulder, CO 80309, USA *Email:
[email protected]
The phenomenon of thermoelectricity provides a means of directly converting electricity to a temperature gradient and vice versa, using the electrical carriers, electrons and holes, in the solid state devices as the working fluids with no moving parts. This offers many distinguished characteristics such as that they are environmentally friendly, quiet, compact and scalable. However the inherent low efficiency of thermoelectric devices based on conventional materials has restricted their use only to niche applications, such as power generators for space exploration and temperature control for some laboratory instruments. In the last 15 years, significant progress has been made in developing higher efficiency thermoelectric materials and this has led to a renewed interest in the field of thermoelectric energy conversion. Central to this advancement is the recognition that nanostructured materials make it possible to effectively decouple the Seebeck coefficient, electrical conductivity and thermal conductivity, all of which are intimately connected in conventional materials, and vary each of these somewhat independently. This review describes some recent developments in the field with emphasis on the development of bulk nanostructured materials, i.e. nanocomposites. After a brief review of the theoretical foundation of high efficiency thermoelectric nanocomposites and the synthesis methods for nanocomposites, we review the different classes of materials recently developed for various applications in different operating temperature ranges.
441
442
Thiagarajan et al.
1. Introduction to Thermoelectricity With the rising demand for energy, the impending energy scarcity and problems of global warming due to fossil fuels, it is incumbent on us to develop new and more efficient ways of utilizing energy. This has motivated the development of cleaner and renewable energy technologies such as solar photovoltaics, wind power, tidal power, among others. Thermoelectric phenomenon is the direct conversion between electric and thermal energy, and offers a convenient means for heating and cooling materials and direct electricity generation from thermal sources. The most familiar example is a thermocouple, where the open circuit voltage of a junction between two dissimilar conductors is determined by the temperature of the junction. If instead of open-circuit operation, the thermocouple is allowed to do work across an electrical load, then the device operates as a thermoelectric power generator (Fig. 1(a)). Alternatively, if the load is replaced with an electrical power source to reverse the current flow, then the device operates as a refrigerator or heat pump (Fig. 1(b)). Compared to traditional refrigerators and heat engines, thermoelectric energy converters have the advantages of simplicity, reliability, no vibrations, and scalability. Furthermore, because they use no refrigerants or working fluids, thermoelectric devices may be expected to have negligible direct emissions of greenhouse gases over their lifetime, likely reducing their contribution to global warming compared to conventional technologies. Due to this combination of desirable qualities, they are ideal for applications such as household refrigeration and recovery of heat energy in automobile and industrial exhaust gas that is otherwise dumped to the atmosphere, or integrated solar-thermoelectric systems for harnessing the thermal energy in the solar spectrum. Despite the possibility of such attractive applications, however, thermoelectric technology is not widespread due to the fact that the efficiency of devices made of conventional thermoelectric materials is very low. Thus, it has been limited to such niche applications as radioisotope thermoelectric power generators (RTGs) for space probes, heating or cooling car seats in luxury cars, and temperature control of some laboratory instruments.
443
Nanocomposites as High Efficiency Thermoelectric Materials
The efficiency of heat to electrical energy conversion in a thermoelectric generator, as shown in Fig. 1(a), depends on many parameters, including the material properties Seebeck coefficient S, electrical conductivity σ and thermal conductivity κ, the (absolute) temperatures of the hot side TH and the cold side TC, and the load resistance. The thermal conductivity of a typical material has contributions of both electrons and phonons, denoted as κe and κp, respectively. Joule heating and heat conduction inside the device cause irreversible energy losses, leading to lower efficiency. The efficiency can be calculated by considering it as a thermodynamic heat engine, and taking into account the different energy losses. Under optimized load conditions, assuming that the material properties are constant within each thermoelectric leg, and that the contact resistances are negligible compared to the total resistance in the arms, the efficiency η of power generation is given by [1, 2]
η=
TH − TC ZT + 1 − 1 ⋅ TH ZT + 1 + TC TH
(1)
where T is the mean temperature in each arm.
(a)
(b)
Figure 1. A thermoelectric couple with semiconducting n- and p-type arms configured for (a) power generation and (b) refrigeration.
444
Thiagarajan et al.
If the device is configured for refrigeration, then the coefficient of performance (COP) can be obtained similarly as, T TC ZT + 1 − H TC COP = (TH − TC ) ZT + 1 + 1
(2)
In both of the above equations which quantify the maximum achievable performance of thermoelectric power generators and refrigerators, the materials properties S, σ, and κ appear in the combination S 2σ κ , making it convenient to define this quantity as the thermoelectric figure of merit of the material. This provides a means for the assessment of a material for its suitability for the use in a thermoelectric device. The figure of merit can be defined nondimensionally as
ZT =
S 2σ
κ
T
(3)
A good thermoelectric material is the one with a high Seebeck coefficient, a high electrical conductivity, and a low thermal conductivity. As all the three properties are interconnected, it has proved impossible to raise the value of the ZT to much over 1 in bulk materials till recently. The best thermoelectric materials are found in heavily doped semiconductors. Insulators have poor electrical conductivity and metals have low Seebeck coefficient. In semiconductors, the phonon contribution to the thermal conductivity can be reduced without much reduction in electrical conductivity. A proven approach to reduce the phonon thermal conductivity is through alloying proposed in later 1950’s [3]. The mass difference scattering in an alloy reduces the lattice (phonon) thermal conductivity significantly without much degradation to the electrical conductivity. Figure 2 shows the figure of merit vs. temperature of some of the common bulk thermoelectric materials, in their respective operating temperatures. Usually we could categorize the
Nanocomposites as High Efficiency Thermoelectric Materials
445
thermoelectric materials according to the temperature range where their peak ZT occurs as low temperature materials (200K-400K), mediate temperature materials (400K-800K), and high temperature materials (above 800K). For example, the commercially available thermoelectric materials for room temperature operation are from the (Bi1-xSbx)2 (Se1-yTey)3 alloy family that reach ZT~1 around room temperature. This class of materials currently dominates in temperature control and thermal management applications. High temperature materials such as silicon germanium alloys are heavily investigated and used for space exploration. Medium temperature range thermoelectric materials could have significant impacts on waste heat recovery and solar thermal utilization.
Figure 2. Dimensionless figure of merit ZT at different temperature ranges of conventional materials.
Figure 3 shows that the efficiency of thermoelectric power generation with Th /Tc=2.5 is about 13% (TH ≈ 480°C, TC ≈ 30°C). This situation might apply to waste heat scavenging from automobile exhaust. However, to be able to achieve efficiencies that are competitive with conventional technologies for power generation and refrigeration, and thus to achieve widespread utilization of thermoelectrics, we need to
446
Thiagarajan et al.
develop materials with ZT of about 3 or higher, that can be manufactured in large quantities inexpensively. If thermoelectric materials could have a ZT=3, the efficiency of the same waste heat scavenging application almost doubles to 24%. This vastly improved efficiency could open up many more potential applications for thermoelectric energy conversion by lowering the operating cost, although the important issue of the capital cost of the materials still needs to be addressed.
Thermal-to-Electric Conversion Efficiency (%)
30
25 ZTave = 2
20
ZTave = 1
15
10 ZTave = 0.5
5 Tcold = 300K 0 300
500
700
900
1100
1300
Hot Side Temperature (K)
Figure 3. Efficiency of power generation vs. hot side temperature using a thermoelectric generator for materials of various ZTs. The cold side is kept at room temperature.
Figure 4 shows the historic progress of high efficiency thermoelectric materials. The maximum ZT has stayed stagnant at around 1 for all temperature range over 50 years since the important advancements using alloying approach. In the 1990s, two parallel approaches were proposed for the enhancement of the ZT of thermoelectric materials. We have witnessed heightened interests in
Nanocomposites as High Efficiency Thermoelectric Materials
447
thermoelectrics due to the significant ZT improvements published around year 2000. The first of these approaches is based on new categories of advanced bulk materials [4-6], with crystal structures that contain weakly bound atoms or molecules with large vibrational amplitudes (called ratters) at partially filled structural sites acting as effective phonon scatterers. The notion of phonon glass-electron crystal (PGEC) enunciated by Glen Slack [7] has been useful in guiding the efforts in this direction. Material systems, such as skutterudites (e.g., CoSb3) [8-10], clathrates (e.g., Ba8Ga16Ge30) [11-13] and Zintl phases [14] belong to this category.
Figure 4. Dimensionless figure of merit of some of the recent materials showing the quantum size effects and low thermal conductivity due to selective phonon scattering.
The second approach is using low-dimensional materials (such as quantum well superlattices, quantum wires and quantum dots) [15, 16] that would result in an enhancement of the ZT by two mechanisms: (i) nanoscale features that introduce quantum confinement effects in the material lead to an enhancement of the power factor S2σ , and (ii) the use of the numerous interfaces in the nanostructures that scatter phonons more than the electrons, based on the difference in their respective
448
Thiagarajan et al.
scattering lengths, and thus reducing the thermal conductivity without adversely affecting the electrical conductivity as much [17, 18]. The first phase of the investigation of low-dimensional thermoelectric materials was focused on the development of these concepts and on their experimental proof-of-concept verification. This approach has proved to be of great value to present research directions where composite materials are being specially designed and synthesized for superior thermoelectric performance. The quantities S, σ, and κ for conventional bulk crystalline systems (3-dimensional) are interrelated in such a way that it is very difficult to control these variables independently so that ZT could be increased. This is because an increase in S (by lowering carrier concentration) usually results in a decrease in σ, and an increase in σ produces an increase in the electronic contribution to κ, according to the Wiedemann–Franz law. However, if the dimensionality of the material is decreased, the new variable of length scale becomes available for the control of materials properties. It is possible to induce dramatic change in the density of electronic states, allowing new opportunities to vary S, σ, and κ quasi-independently when the length scale is small enough to give rise to quantum-confinement effects as the number of atoms in any direction (x, y, or z) becomes small (e.g., less than ~100). In addition, as the dimensionality is decreased from 3D crystalline solids to 2D (quantum wells) to 1D (quantum wires) and finally to 0D (quantum dots), new physical phenomena, such as metal-semiconductor transition (as demonstrated in Bi nanowire composites [19]), are also introduced and these phenomena may also create new opportunities to vary S, σ, and κ independently. Furthermore, the introduction of many interfaces, which scatter phonons more effectively than electrons, or serve to filter out the low-energy electrons at the interfacial energy barriers, allows the development of nanostructured materials with enhanced ZT, suitable for thermoelectric applications [18]. Based on the theoretical study of how low-dimensional materials such as quantum wells [15] and quantum wires [16] could be utilized to enhance ZT of a material, Harman et al. grew superlattices of PbTe with embedded nanodots of PbSe by molecular beam epitaxy, and these materials exhibited a very high ZT of ~1.7 and 3.5 at 300 K and around
Nanocomposites as High Efficiency Thermoelectric Materials
449
570 K respectively [20-25]. The lattice thermal conductivity in the superlattices is as low as ~0.33 W/mK, which is a 6 times reduction from that of bulk PbTe (~2.4 W/mK). Later experiments showed that the increase in the ZT is wholly due to the reduction of the thermal conductivity and not by an enhancement of the Seebeck coefficient [26]. Venkatasubramanian et al. [27] prepared p-type superlattice of Bi2Te3Sb2Te3 by molecular beam epitaxy. Compared with normal bulk materials, the thin film materials with the superlattice structure also showed very low lattice thermal conductivity leading to high ZT values of up to 2.4 at 300 K. These early works have experimentally demonstrated that it is possible to raise the ZT of materials much beyond 1 by the use of nanostructures although the understanding of the responsible mechanisms in ZT enhancement has taken quite long time. An additional effect that could be realized by the use of low dimensional materials (like superlattices) is electron filtering, which would result in a concomitant increase in the Seebeck coefficient. It was Moyzhes and Nemchinsky who first proposed that increasing the power factor using potential barrier scattering may be useful not only for a film material but also for a bulk material [28]. They proposed that the formation of a structure with potential barriers in a bulk material, such as the grain-boundary structure of a film, can result in an effective filtering of electrons of energy lower than the barrier height, thus allowing only the high energy electrons to contribute to electrical current. This in turn, will increase the Seebeck coefficient, thus leading to an enhancement of the power factor. Their proposal has been theoretically supported by calculations performed by other groups [29, 30]. Following this, Zide et al. demonstrated an increase in the Seebeck coefficient by means of electron filtering in superlattice composite material made of In0.53Ga0.47As with In0.53Ga0.28Al0.19As barriers [31, 32]. Thus, in the past few years, numerous avenues have been explored in search of new physical phenomena that could lead to better thermoelectric performance, and novel materials that could exhibit these new phenomena have been developed. Some recent reviews on the different approaches include Refs. [18, 33] on nanoscale thermoelectricity, Ref. [34] on thermoelectric materials with complex unit cells and Ref. [35] on the chemical problems associated with the
450
Thiagarajan et al.
design of new thermoelectric materials. In the following sections, we show why nanocomposite materials are a promising new class of materials, describe the various synthesis routes being explored and highlight some of the latest achievements in developing high ZT nanocomposites materials.
2. Nanocomposites as Highly Efficient Thermoelectric Materials As described in the previous section, many studies have shown that it is possible to enhance the thermoelectric performance of materials beyond what is possible in bulk materials by the use of nanostructures, either through the quantum low-dimensional effects on the change of electronic energy states, the selective filtering of low energy electrons across a barrier, or by interfacial scattering of phonons to reduce the thermal conductivity. However, these are to be considered proof-ofconcept studies and significant challenges exist in applying these superlattices thin films or nanowires to commercial applications, as they are far too expensive to fabricate, can only be made in small quantities, and are difficult to make high performance devices [36]. To achieve commercial usage, new types of nanostructured materials that preserve the advantages of the low-dimensional materials including thermal conductivity reduction and potentially power factor increase, while at the same time be cost-effective and susceptible to large-scale batch production need to be developed. Chen and co-workers studied carefully the thermal conductivity reduction mechanisms in the aforementioned high efficiency thermoelectric superlattices [37]. They found that the periodicity of superlattices is not a necessary condition for thermal conductivity reduction. The reduced thermal conductivity in superlattices comes from the sequential interface scattering of phonons rather than the coherent superposition of phonon waves [38]. This conclusion leads naturally to the idea of using nanocomposites as potentially a cheap alternative to superlattices in the quest for high ZT materials [18, 39]. Such nanocomposites can be in the form of nanoparticles and nanowires embedded in a host material, or mixtures of two different kinds of nanoparticles [40]. Indeed nano-inclusions for thermoelectric materials
Nanocomposites as High Efficiency Thermoelectric Materials
451
have been attempted before [41]. For example, by the addition of BN and B4C nanoparticles into a Si-Ge alloy, it was founded that the thermal conductivity can be reduced appreciably. Unfortunately, the inclusion also reduces the electrical conductivity and thus the net gain in ZT was not large. This is because the added inert particles have a large bandgap and thus introducing a high electric potential barrier that scatters electrons. This indicates that one should carefully choose the materials with matched electronic properties. Recent experimental results [23, 27] show no significant reduction in the electrical conductivity was observed for current flow perpendicular to the interface of Bi2Te3/Sb2Te3 superlattices and along the interface of PbTe/PbSeTe quantum-dot superlattices. This demonstrates that by properly choosing the mismatch in electronic properties, the electron transport properties can be maintained at a level comparable to bulk materials or even enhanced using interfaces as energy filters or energy quantization barriers. At the present time a number of research groups are developing nanocomposite materials with a potential for scale-up and practical applications. Section 4 highlights some of recent achievements. The overarching goals for designing these nanocomposites materials are to introduce many interfaces that are specially chosen to: 1) reduce the thermal conductivity more than the electrical conductivity conduction by interface scattering, and 2) to increase Seebeck coefficient (for example, by carrier-energy filtering or by quantum confinement) more than decreasing the electrical conductivity, thereby yielding an increase in power factor, with both goals helping to increase ZT. Nanocomposite materials offer a promising approach for the preparation of bulk samples with nanostructured constituents. As reviewed in Sec. 3, a variety of materials synthesis processes and approaches have been suggested by various research groups, involving different materials systems and processing methods, utilizing a number of common fundamental concepts. Such nanocomposites can be easily handled for both material property measurements and characterization; they can also be assembled into a variety of shapes for device applications, and can be scaled up for commercial applications. The question is can nanocomposites replicate the enhancements obtained in samples made by atomically precise methods such as molecular beam
452
Thiagarajan et al.
expitaxy grown superlattices? Based on an effective medium theory they developed, Bergman and Levy [42] showed that figure of merit of a composite of two materials A and B can never exceed the highest of either A or B (though the power factor can [43]). Nevertheless, in view of the fact that this theory was based on a phenomenological description without the details of the transport mechanisms, it is conceivable that composites made with nanoscale structures that influence the electron and phonon transport in different ways than bulk composites will show better thermoelectric performance. In the past, some theoretical and modeling studies have been attempted to investigate phonon and electron transport in low dimensional structures such as superlattices and nanocomposites. In the following, we describe some of the modeling studies that have been performed in the recent past to study the effect of nanostructures on thermal and electrical transport in nanocomposites. In Sec. 4, we will describe a selected number of recent developments that have taken place in the recent past with emphasis on advances made since the last review in the subject was written [18].
2.1. Modeling of Phonon Transport Considering that the phonon wavelength for dominant phonon heat carriers is about 1 nm and the phonon mean free path could be in the order of 100 nm [44] and thermoelectric nanocomposites often are made of nanoparticles or nanowires with a characteristic length in tens of nanometers, there would be significant challenges in modeling phonon transport using electron or phonon wave mechanics while the effective medium theory based on Fourier heat conductivity is not valid. To study the thermal conductivity of thermoelectric nanocomposites, Yang and Chen heavily relied on statistical mechanics description of thermal transport and developed deterministic solution of phonon Boltzmann equation to study periodic two-dimensional nanowire composites [45, 46], and Monte Carlo simulation of phonon transport for the thermal conductivity in periodic and random (3D) nanoparticle composites [47]. Following assumptions are made for their modeling studies: (1) The phonon wave effect can be excluded. (2) The frequency-dependent
Nanocomposites as High Efficiency Thermoelectric Materials
453
scattering rate in the bulk medium is approximated by an average phonon mean free path. (3) The interface scattering is diffuse. It can be a daunting problem to model or to simulate the transport properties of nanocomposites since the distribution of the nanoparticle size and location can vary a lot. To accurately model the transport, the simulation domain should be as large as possible, or even the same size as the sample. The memory and computational time requirements for such a multiscale problem are very demanding. Simulation of the properties of a periodic structure often gives physical insights of materials even in their random form. Instead of treating the whole structure, Yang and Chen [40, 46, 47] simplified the problem by dealing with periodic nanocomposites that can be constructed by a periodic stack of a unit cell. A unit cell might consist of one nanoparticle/nanowire or many nanoparticles and nanowires. If the unit cell consists of only one nanoparticle or one nanowire, the repeating structure is a simple stack of a periodic nanocomposite. If the unit cell consists of many nanoparticles and nanowires inside and the distribution inside the unit cell (simulation box) is random, the nanocomposite is then semi-periodic, i.e, long range periodic but random inside the unit cell. The study shows that the prevailing approach to model thermal conductivity of nanocomposites, which includes the interface thermal resistance, or Kapitza resistance, with the Fourier heat conduction theory, underpredicts the effect of interface for thermal conductivity reduction since the Fourier heat conduction theory is based on the diffusion picture and is not applicable when the phonon mean free path is longer than the characteristic length of the nanocomposites such as the particle diameter and/or interparticle separation distance. Figure 5 shows the size effect on the thermal conductivity of Si1-xGex nanocomposites with Si nanoparticles embedded in Ge matrix. First of all, for fixed size of silicon nanoparticles, the less the atomic percentage of germanium, which has lower thermal conductivity than silicon, the lower is the effective thermal conductivity of the nanocomposites. This is very different from macroscale composites, in which the effective thermal conductivity increases with the decreasing volumetric fraction of the lower thermal conductivity component. This is caused by the ballistic transport of phonons in both the host material and
454
Thiagarajan et al.
the nanoparticles, and the interface resistance between the host material and the nanowires.
Figure 5. Size effects on the thermal conductivity of nanocomposites with Si nanoparticles embedded in Ge matrix. Reprinted with permission from Ref. [47]: M. S. Jeng, R. G. Yang, D. Song, and G. Chen, Journal of Heat Transfer-Transactions of the ASME 130, 042410 (2008) Copyright @ American Society of Mechanical Engineers.
The comparison of the thermal conductivity of the nanocomposites with 50 nm silicon particles and 10 nm silicon particles simply aligned in germanium matrix shows that the thermal conductivity decreases as the size of the nanoparticles decreases. The comparison of thermal conductivity of nanocomposites with the corresponding alloy value also demonstrates that nanocomposite can be an effective approach to reduce the thermal conductivity and thus to develop high-efficiency thermoelectric material. Jeng et al. [47] also compared the thermal conductivity of periodic and random nanocomposites and found out that the randomness either in particle size or in particle location distribution causes only slight fluctuation but is not a dominant factor for thermal conductivity reduction. Based on the fact that the phonon-interface scattering dominates the thermal conductivity reduction for nanocomposites, Yang et al. proposed to use interfacial area per unit volume (interface density) as a unified parameter to replace the nanoparticle/nanowire size and the atomic
Nanocomposites as High Efficiency Thermoelectric Materials
455
composition and to correlate the wide spreading thermal conductivity data. Figure 6 shows that the thermal conductivity data of nanoparticle/nanowire composites falls nicely onto one curve as a function of interfacial area per unit volume. The randomness either in particle size or position distribution causes slight fluctuation but is not a dominant factor for thermal conductivity reduction. The key for thermal conductivity reduction is to have high interface density where nanoparticle composites can have much higher interface density than simple 1-D stacks such as those expensive periodic superlattices, thus nanocomposites benefits ZT enhancement in terms of thermal conductivity reduction.
Figure 6. Thermal conductivity of nanocomposites as a function of interfacial area per unit volume (interface density). The thermal conductivity data of nanoparticle composites falls into one curve as a function of interfacial area per unit volume. At sufficiently high interface densities, the thermal conductivity of the nanocomposite reaches a value lower than that of an alloy of the same composition. Reprinted with permission from Ref. [47]: M. S. Jeng, R. G. Yang, D. Song, and G. Chen, Journal of Heat Transfer-Transactions of the ASME 130, 042410 (2008) Copyright @ American Society of Mechanical Engineers.
Figure 7 shows the temperature-dependent thermal conductivity of nanoparticle composites. Boundary scattering results in very different temperature dependence of the thermal conductivity of nanocomposites
456
Thiagarajan et al.
comparing to their bulk counterpart where at high temperature the thermal conductivity is dominated by the Umklapp phonon-phonon scattering process. The thermal conductivity of Si-Ge nanocomposites with 10 nm particles in the germanium matrix is almost temperature independent.
Figure 7. Temperature-dependent thermal conductivity of Si-Ge nanoparticle composites. Reprinted with permission from Ref. [47]: M. S. Jeng, R. G. Yang, D. Song, and G. Chen, Journal of Heat Transfer-Transactions of the ASME 130, 042410 (2008) Copyright @ American Society of Mechanical Engineers.
To conclude, thermal conductivity of nanocomposites can be effectively reduced which renders nanocomposite approach as potentially a cheap alternative to superlattices for high ZT material development. The challenge is to properly choose the mismatch in electronic properties between the constituent materials so that the electron transport properties can be maintained or even enhanced.
2.2. Modeling of Electron Transport The modeling tool for electron transport in nanocomposites is relatively rare. In theory, electron transport in nanocomposites can be modeled similarly to that of phonons, such as developing the Boltzmann
Nanocomposites as High Efficiency Thermoelectric Materials
457
equation solver or Monte Carlo simulation for electron transport, as long as quantum confinement is negligible, which is often the case for random nanocomposites. However, Monte Carlo simulation of thermoelectric transport of electrons in nanocomposites is considerably more challenging because of the possibility of nonequilibrium between electrons and phonons at the interfacial region and the requirement of solving concurrently the Poisson equation to determine the electrostatic potential. Simplified models where electron transport properties are calculated using standard Boltzmann equation expressions in the relaxation time approximation have previously been attempted [48, 49]. Yang and Chen [50] extended this model to study the thermoelectric transport properties of electrons in SiGe nanocomposites, where the nanocomposite is made by compacted SiGe alloys nanoparticles. The formulation of the transport properties are written similarly as that for bulk materials [49], which is relatively easy to implement, with an inclusion of interface scattering model. The interface scattering can be viewed as electron energy filters [29, 51]. At the boundaries of two different grains (nanoparticles), low energy electrons are reflected and high energy electrons pass through. Thus the relaxation time due to boundary scattering can be written as,
1
τb 1
τb
=
v
δ
=
6 2E / m when E < Eb d
= 0 when E > Eb
(4)
where d is the size of nanoparticles, Eb is the energy barrier height. Figure 8 shows the temperature dependent electrical conductivity, Seebeck coefficient, and power factor for compacted nanoparticle composites made of Si0.8Ge0.2 alloy as a function of energy barrier height. The nanoparticle diameter is assumed to be 20 nm and the doping concentration is assumed to be 1.0x1020 cm-3. As shown in Fig. 8(a), the electrical conductivity decreases with energy barrier height and the Seebeck coefficient increases with energy barrier height due to the low
458
Thiagarajan et al.
energy carrier filtering. Figure 8(b) shows that there exists an optimum barrier height for the power factor enhancement. Overall the enhancement is effective at low temperature and becomes less effective at high temperature, which is very similar to experiment observations [52, 53]. This simplified model could be a good tool to guide the material synthesis since it predicts the dependence of thermoelectric transport properties on carrier concentration, temperature, grain (nanoparticle) size, and energy barrier height after the input parameters are optimized with experimental data.
(a)
(b)
Figure 8. The temperature dependent electrical conductivity and Seebeck coefficient (a) and power factor (b) for Si0.8Ge0.2 alloy compacted nanoparticle composites as a function of energy barrier height Eb. Reprinted with permission from Ref. [50]: R. Yang and G. Chen, in SAE World Congress (Society of Automotive Engineers, 2006), Article # 2006-01-0289. Copyright @ Society of Automotive Engineers
Faleev and Leonard [54] developed a model for predicting the Seebeck coefficient, electrical conductivity and ZT of materials with nanoscale metallic inclusions, using the idea of the band bending at the metal-semiconductor interface acting as energy filters. They found that the Seebeck coefficient of the nanocomposite material is always enhanced compared to the inclusion-free system, and that the smaller the
Nanocomposites as High Efficiency Thermoelectric Materials
459
nanoinclusion, the greater the enhancement. Similar to Yang and Chen [50], they also found that the power factor is optimized for certain values of the boundary potential. The enhancement of the ZT is dominated by the reduction in the lattice thermal conductivity for low carrier density, while at high carrier density the electronic contribution becomes important. In summary, electrical and energy transport in nanostructures differs significantly from macrostructures because of classical and quantum size effects on energy carriers. Both thermal conductivity reduction and the possibility to maintain—and even enhance—the electronic power factor in nanocomposites render cost-effective random nanocomposites as a promising alternative to expensive superlattices for high ZT material development. The key to thermal conductivity reduction is to have high interface density where nanocomposites can have much higher interface density than simple 1D stacks such as superlattices, thus nanocomposites benefits ZT enhancement in terms of thermal conductivity reduction. In the meantime, the interfaces can be viewed as energy filters for electrons which allow only electrons having higher energy to pass through the barrier, and thus enhance the Seebeck coefficient. Overall there exists an optimum barrier height and nanoparticle (grain) size for the electronic power factor enhancement due to the electrical conductivity reduction at the same time.
3. Synthesis of Thermoelectric Nanocomposites A thermoelectric nanocomposite is a composite constructed by incorporating thermoelectric nanostructures in a matrix of a bulk thermoelectric material or compacting various thermoelectric nanostructures into bulk form. Several methods for the preparation of thermoelectric nanocomposites have been exercised. These methods to obtaining bulk samples with nanoscale features can be broadly classified into two categories: (i) compaction of nanoscale constituents (nanoparticles, nanowires, etc.) into bulk samples, (ii) in situ precipitation of nanoscale constituents by means of phase separation. In the following, the two routes are briefly described.
460
Thiagarajan et al.
3.1. Preparation of Nanocomposites by Compaction Techniques Several different compaction techniques have been utilized in the recent past to obtain bulk thermoelectric samples from nanoscale constituents that are synthesized by an array of physical and chemical methods. The essence of all compaction techniques is to apply high pressure for densification, and often a rather high temperature to soften the material so that plastic deformation allows better filling and material flow by diffusion to remove the remaining porosity. The challenge is in achieving high density (and low porosity) without losing the nanoscale microstructure and keeping the material chemically pure. 3.1.1. Compaction Methods Cold compaction is a process in which powder materials are compressed in a temperature range where high temperature deformation mechanics like dislocation or diffusional creep can be neglected. Cold compressing is the most important compaction method in powder metallurgy. Thus, cold sintering offers the potential for retaining the metastable nanoscale constituents. Despite this, the nanopowders may not bond very well, leading to lower carrier mobility and therefore low ZT. A more common way of consolidation of nanopowders is hot pressing, where, in addition to the high pressure, moderate to high temperature is applied to the sample simultaneously. This results in a better particle-particle bonding, and higher carrier mobility in the final sample. However, it is a challenge to retain the nanometer-sized crystal grains in the final sample because the grains can grow significantly. Figure 9 shows the TEM images of nanocomposites of BixSb2-xTe3 prepared by ball milling and hot pressing [55]. Evidently, under the right conditions hot pressing can preserve the nanostructure and lead to enhanced thermoelectric performance.
Nanocomposites as High Efficiency Thermoelectric Materials
461
Figure 9. TEM images showing the microstructure of a hot pressed nanocomposite bulk sample of Bi-Sb-Te. (A) Low magnification image showing the nanograins. (B) High magnification image showing the nanosize, high crystallinity, random orientation and clean grain boundaries. The nanostructure is seen to be preserved even after hot pressing. Reprinted with permission from Ref. [55] B. Poudel, Q. Hao, Y. Ma, Y. Lan, A. Minnich, B. Yu, X. Yan, D. Wang, A. Muto, D. Vashaee, X. Chen, J. Liu, M. S. Dresselhaus, G. Chen, and Z. Ren, Science 320, 634 (2008) Copyright @ American Association for the Advancement of Science.
Spark Plasma Sintering (SPS), also known as Field Assisted Sintering Technique or Pulsed Electric Current Sintering, is a novel sintering technique which is gaining increasing popularity for making thermoelectric nanocomposites [56]. In the SPS technique, the sample is heated by pulsed electric current which flows through the punch-die-sample-assembly under a low voltage. It is expected that due to the high current, at the comparatively small gaps between the powder particles, electrical discharges will occur. These discharges result in microscopic electric arcs, leading to high temperatures and pressures locally, forming a good contact between the particles. And additional advantage is that, gases and moisture that have been adsorbed on surfaces of the nanoparticles are eliminated, and oxide layers can be broken due to the arcs. Subsequently, Joule heating occurs in the compact due to the current flow, especially at spots of high electrical resistance. This temporarily overheats the sample while the overall sintering temperature is relatively low. As the heat is generated internally in the SPS, in contrast to the conventional hot pressing where the heat is provided by external heating elements, very high heating rates
462
Thiagarajan et al.
(>300°C/min) and short sintering times in the range of a few minutes can be achieved resulting in a fast throughput. Also, the high speed of the process ensures it has the potential of densifying powders with nanosize or nanostructure while avoiding coarsening which accompanies standard densification routes. Densities very close to theoretical densities and excellent thermoelectric performances have been achieved in samples treated by SPS process. In view of these advantages over the other compaction methods, it is being increasingly utilized for making thermoelectric nanocomposites [57-59]. 3.1.2. Synthesis of Thermoelectric Nanostructures There are numerous techniques available to synthesize the nanoscale constituents, such as nanoparticles, nanoplates, nanowires, nanobelts and nanotubes etc. Some of the popular techniques for synthesizing thermoelectric nanostructures are described below. Mechanical Attrition Mechanical attrition is one of the most popular methods for synthesis of nanostructures from bulk raw materials, due not only to the convenience and minimal requirement for complex equipment, but also, to the versatility in terms of the number of different systems of materials that can be prepared this way. Mechanical attrition produces its nanostructures by the structural decomposition of coarse grains into finer structures as a result of plastic deformation and can be carried out at room temperature. The process can be performed on high energy mills, centrifugal type mill and vibratory type mill, and low energy tumbling mill. Nanoparticles, of sizes ranging from 200 nm to as low as 5-10 nm, can be prepared by the use of attritors, vibratory mills and horizontal ball mills [60]. As the process is sensitive to contamination from the milling environment, tight atmospheric control is essential to maintain the purity of the material, in particular to avoid oxidation. Consequently, an argon or nitrogen gas atmosphere is used for preparation of thermoelectric materials. Contamination from wear debris of the milling media is also a problem with mechanical attrition that may negatively impact the quality
Nanocomposites as High Efficiency Thermoelectric Materials
463
of the alloy, requiring a judicious selection of processing time and milling speed [2]. Mechanical attrition has been used for the preparation of nanopowders of Fe-Si alloys [61], Si-Ge alloys [52, 62], PbTe [63, 64] and PbSbTe alloys[65], BiSbTe alloys [55, 66, 67], MgSiSn[68], CoSb3 [69], and materials such as La3-xTe4 [70] that are challenging to synthesize using melt synthesis and other traditional methods. Wet Chemistry Synthesis Wet chemistry method is the powerful tool to generate various nanostructures in different shapes. For example, solvothermal (including hydrothermal) method synthesizes the nanostructures by using the solubility in water (or a suitable solvent) of inorganic precursors at elevated temperatures (above the critical point of the solvent) and selfformed pressures in an autoclave, and the subsequent crystallization of the dissolved material from the fluid. Compared with other synthesis routes performed at atmospheric pressure, the increased reaction temperature in the solvothermal technique may lead to an accelerated crystal growth accompanied by a narrow particle size distribution and better crystallinity. Another advantage of this method is that nanostructures of different morphologies such as nanopowders, nanorods, polygonal nanosheets, polyhedral nanoparticles and sheet-rods can be synthesized. Also, as most materials can be dissolved in the solvent by heating and pressurizing close to the critical point, this approach is suitable for synthesizing nanostructures of a wide variety of solid materials. Hydrothermal synthesis has been used to obtain nanostructures of Bi2Te3 [71-73], Sb2Te3 [74], PbX (X=S, Se, Te) [75], CoSb3 [76, 77], etc. Figure 10 shows TEM images of Bi2Te3 nanotubes synthesized by hydrothermal processs. On the other hand, ambient solution phase method can be operated in mild conditions to fabricate different kinds of nanostructures through an anisotropic growth process by adding different surfactants or tuning reaction conditions, such as temperature, pH value etc. Bi2Te3 nanoplates and nanorods have been successfully fabricated using this method [78, 79], and using a two step process, Te/Bi2Te3 coreshell nanowires can be obtained [80]. Furthermore, rough silicon
464
Thiagarajan et al.
nanowires can be synthesized through a wet etching process and were recently reported with enhanced thermoelectric performance [81].
Figure 10. TEM photos of hydrothermally synthesized Bi2Te3 nanotubes. Reprinted with permission from Ref. [71] X. B. Zhao, X. H. Ji, Y. H. Zhang, T. J. Zhu, J. P. Tu, and X. B. Zhang, Applied Physics Letters 86, 062111 (2005), Copyright @ American Physical Society.
Electrochemical Deposition Electrochemical deposition provides a facile and effective route to fabricate various nanostructured metal alloys for thermoelectric applications [82, 83]. Stacy’s group made a breakthrough by fabricating high quality Bi2Te3 nanowire arrays for the first time using the porous anodic alumina (PAA) template assisted electrodeposition process. This technique has been quickly developed as a popular method to obtain various thermoelectric nanowire arrays, such as: Bi2Te3, Sb2Te3, Bi-SbTe, Bi-Te-Se, CoSb3, PbTe etc. [84-89]. A high degree of control in the diameter and length of the nanowires can be exercised in this method. The diameter of as-obtained nanowires ranges from 20 to 300 nm and is related to the template pore size, while the length depends on the electrodeposition time. Moreover, the alloy composition can be adjusted by changing the content of electrolyte solution [90], and the orientation of the nanowire arrays can be changed
Nanocomposites as High Efficiency Thermoelectric Materials
465
by tuning the deposition potential or using pulsed electrodeposition process [91]. Uniformity of growth can also be achieved by electroplating in low temperature [92]. Furthermore, in some special electrodeposition conditions, novel hollow thermoelectric nanostructures can be obtained. Li et al. reported the successful fabrication of Bi nanotube arrays [93], and Zhu’s group was able to synthesize Bi2Te3 and relative compounds nanotube arrays [94]. On the other hand, even without the assistance of templates, one-dimensional chinelike Bi-Sb nanostructure was fabricated through a template-free electrodeposition process by Zhou et al. [95], and PbTe cubes can be directly deposited on the polycrystalline gold substrate by Xiao et al. [96]. Using the cyclic electrodeposition/stripping method, significant amounts of long polycrystalline Bi-Te nanowires were obtained on highly oriented pyrolytic graphite (HOPG) surface [97].
Figure 11. (a) and (b) show the TEM images of multilayered Bi2Te3/Sb nanowires deposited in different conditions labeled in the bottom of each corresponding figure. (c) and (d) are the corresponding high magnification TEM images. Reprinted with persmission from Ref. [100]: W. Wang, G. Q. Zhang, and X. G. Li, Journal of Physical Chemistry C 112, 15190 (2008) Copyright @ American Chemical Society.
466
Thiagarajan et al.
Electrodeposition can also be used to synthesize various heterostructured nanomaterials. Using template assisted pulsed electrodeposition, different kinds of thermoelectric materials can be deposited alternately and periodically, by applying different deposition potentials [98, 99]. Wang et al. reported a detailed study of Bi-Sb-Te system and successfully manipulated the growth of Bi2Te3/Sb heterostructure nanowires with desired period, and the minimum period of as-synthesized Bi2Te3/Sb heterostructure nanowires to as low as 10 nm (Fig. 11) [100]. They also fabricated Bi2Te3/Te heterostructured nanowire arrays through a nanoconfined precipitation process [101]. Inert Gas Condensation Inert gas condensation is a versatile process in use today for synthesizing experimental quantities of nanostructured metallic and intermetallic powders. A feature of the process is its ability to generate non-agglomerated nanopowders, which can be sintered at relatively low temperatures. An evaporative source is used to generate the powder particles, which are convectively transported to and collected on a cold substrate. The nanoparticles develop in a thermalizing zone just above the evaporative source, due to interactions between the hot vapor species and the much colder inert gas atoms (typically 1-20 mbar pressure) in the chamber. Recently, this method has been utilized for making Si-Ge nanocomposites [102]. Sonochemical Synthesis The underlying mechanism of sonochemistry arises from the acoustic cavitation phenomenon, that is the formation, growth and implosive collapse of bubbles in a liquid medium due to irradiation with ultrasonic waves. Extremely high temperatures (>5000 K), pressures (>20 MPa), and very high cooling rates (>107 K/s) can be attained locally during acoustic cavitation that lead to many unique properties in the irradiated solution [103]. The remarkable advantages of this method include a rapid reaction rate, the controllable reaction condition and the ability to form nanoparticles with uniform shapes, narrow size distributions and high
Nanocomposites as High Efficiency Thermoelectric Materials
467
purities. Sonochemical synthesis method has been used to obtain nanocrystals of Bi2Se3 [104], Bi2Te3 and intermediate compounds [105], and other metal tellurides and selenides [106]. Figure 12 shows a TEM micrograph of Bi2Se3 nanocrystals made by sonochemical synthesis.
Figure 12. TEM image and a plot of the size distribution of nanocrystals of Bi2Se3 prepared by sonochemical synthesis. The scale of the TEM image is 100 nm. Reprinted with permission from Ref. [104] X. F. Qiu, J. J. Zhu, L. Pu, Y. Shi, Y. D. Zheng, and H. Y. Chen, Inorganic Chemistry Communications 7, 319 (2004) Copyright @ Elsevier.
In addition to the above methods, chemical vapor deposition [107], and sol–gel process [108] have also been explored to synthesize thermoelectric nanostructures.
3.2. Synthesis of Nanocomposites by Phase Separation The phase separation method of synthesizing nanostructures in situ in a bulk sample is inspired by precipitation hardening of aluminum. Basically, in this process, different kinds of metals will be heated up to the liquid phase, and then quenched to obtain a homogenous solid solution. According to the miscibility gap in the phase diagram, the asobtained metastable solid solution will decompose into different phases A and B (or phases rich in A and B during the spinodal decomposition) after a nucleation and growth process by annealing at certain duration, and thus forms the embedded precipitates in bulk matrix. The size of the precipitates increases as the duration and the temperature of the annealing process increase [109-114]. For example, according to the pseudo-binary PbTe-Sb2Te3 phase diagram (Fig. 13), Ikeda et al. were able to produce the self-assembled lamellae PbTe and Sb2Te3 with
468
Thiagarajan et al.
epitaxy-like interfaces by annealing the metastable Pb2Sb6Te11 alloy. Such spontaneous formation of nanoscale features is desirable because it minimizes the possibility of oxidation and the introduction of other forms of impurities, which would lead to degradation of electrical performance.
Figure 13. Pseudo-binary phase diagram of PbTe-Sb2Te3, with the Pb2Sb6Te11 phases shown as a metastable phase. The region near the eutectic composition is enlarged in (b). Reprinted with permission from Ref. [114] : T. Ikeda, L. A. Collins, V. A. Ravi, F. S. Gascoin, S. M. Haile, and G. J. Snyder, Chemistry of Materials 19, 763 (2007) Copyright @ American Chemical Society.
The phase separation method has also been used to obtain PbTe nanocomposites with Ag, Pb, and Sb nanoprecipitates [110, 115], AgSbTe2 in PbTe [111] and in PbSnTe [112], and PbS in PbTe [113]. Figure 14(a) shows a TEM micrograph of LAST-18 sample (AgPb18SbTe20) obtained by phase separation process. The sample shows nano-sized region of the crystal structure that is Ag-Sb–rich in composition. The surrounding structure is epitaxially related to this feature, but is Ag-Sb–poor in composition, and closer to that of PbTe. Figure 14(b) shows the TEM image of the lamellar nanostructure
469
Nanocomposites as High Efficiency Thermoelectric Materials
formed spontaneously by the separation of PbSbTe into Sb2Te3-rich and PbTe-rich phases.
(a)
(b)
Figure 14. (a) TEM image of a AgPb18SbTe20 sample showing a nano-sized region (a “nanodot” shown in the enclosed area) of the crystal structure that is Ag-Sb–rich in composition. The surrounding structure, which is epitaxially related to this feature, is Ag-Sb–poor in composition, and closer to that of PbTe. Reprinted with permission from Ref. [111] : K. F. Hsu, et al. Science 303, 818 (2004) Copyright @ American Association for the Advancement of Science. (b) Microstructure of metastable phase Pb2Sb6Te11 transformed into self-assembled lamellae of Sb2Te3 and PbTe regions by annealing. The lighter regions are PbTe, and the darker regions are Sb2Te. Reprinted with permission from Ref. [114] : T. Ikeda, L. A. Collins, V. A. Ravi, F. S. Gascoin, S. M. Haile, and G. J. Snyder, Chemistry of Materials 19, 763 (2007) Copyright @ American Chemical Society.
4. Recent Achievements in Thermoelectric Nanocomposites As the Seebeck coefficient, the electrical conductivity and thermal conductivity are strongly temperature dependent, any thermoelectric material is suitable for operation over a limited temperature range. Corresponding to the conventional bulk thermoelectric materials as shown in Fig. 2, nanostructured bulk thermoelectric materials with enhanced ZT have been developed over the past 5-10 years. Figure 15 summarizes some nanostructured bulk thermoelectric materials appropriated for different temperature ranges.
470
Thiagarajan et al.
Figure 15. Plots showing the temperature dependence of the dimensionless figure of merit ZT of several nanocomposite materials. In comparison with Fig. 2, it may be noted that that many of the nanocomposites show enhanced ZTs significantly higher than 1.
4.1. Bi2Te3-Based Nanocomposites for Low Temperature Applications Thermoelectric materials that operate in the range 200 K to 400 K are considered as low temperature materials. The primary application of these materials is refrigeration and temperature control of laboratory instruments. Another application of materials that operate at this temperature range lies in the recovery of low-quality waste heat from automobile radiators (~400 K) or even from electronic chips. Currently, most of the thermoelectric devices commercially available and commonly used for applications around room temperature are based on Bi2Te3-Sb2Te3 alloys, due to the fact they have the highest ZT (~1) among any bulk materials around room temperature. However, the temperature range over which these devices can efficiently operate is rather small (-20°C to 100°C) due to the fast deterioration of thermoelectric properties with variation of temperature. Recently, Poudel
Nanocomposites as High Efficiency Thermoelectric Materials
471
et al. [55] made nanostructured samples of BixSb2-xTe3 by hot pressing nanopowders obtained by ball milling of crystalline ingots of Bi2Te3 and Sb2Te3 under inert conditions. Figure 9 shows the micrograph of a sample obtained using TEM. Nanoscale crystalline features, randomly oriented with each other can be clearly seen. As shown in Fig. 15, ZT of about 1.2 has been reached at room temperature, and as high as 0.8 at 250°C. Comparing this data with the ZT of bulk BixSb2-xTe3 materials shown in Fig. 2, it can be seen that the nanocomposite has extended the operational range of the material to a considerable extent, making it useful for both cooling and power generation applications. The high ZT is the result of low lattice thermal conductivity, due to the increased phonon scattering with the interfaces of nanostructures and dislocations. The nanocomposites also show a comparable or higher power factor throughout the temperature range than the bulk ingots. These samples do not suffer from cleavage problem that is common in ingots prepared by traditional zone-melting, which leads to easier device fabrication and integration. Alternative to starting from the alloyed crystalline ingots of BixSb2-xTe3 as in Poudel’s work, nanocomposites could also be made by starting from elemental chunks of Bi, Sb and Te, which are ball milled to get nanopowders [116]. These nanopowders are then hot pressed to obtain samples that show similarly high ZT. The direct route from elements to nanostructured alloy-compounds is more cost-effective and environmentally friendly. The ZT obtained by this method are only about 10% lower than those obtained by using the compounds of Bi2Te3 and Sb2Te3 as the starting materials, apparently due to some microstructural differences and absence of minority elements like Zn, Cd. Hydrothermal method has also been used to synthesize Bi2Te3 nanocomposites [73, 117-119]. Ni et al. synthesized nanopowders of Bi2Te3 using this method, which were then hot pressed with zone-melted alloy in a 10:90 ratio [73]. It was found that the nanosized powders reduce the thermal conductivity much stronger than the electrical conductivity, which results in an enhanced thermoelectric figure of merit of a nanocomposite. ZT value of up to 0.83 has been obtained. Further improvement on the figure of merit of the nanocomposites should be
472
Thiagarajan et al.
possible by appropriate doping of the nanopowders and optimization the composition of the base alloys. Cao et al. [117] synthesized nanosized binary Bi2Te3 and Sb2Te3 powders by hydrothermal route, which were then hot pressed in 1:1, 1:3 and 1:7 ratios. TEM images show that the composites have a laminated structure composed of Bi2Te3 and Sb2Te3 nanolayers with the thickness varying alternately between 5 and 50 nm. The nanoscale laminated structure improves the thermoelectric performance in comparison with bulk samples of similar compositions, reaching a high ZT of 1.47 at 450 K for the nanocomposite of 1:1 composition. Tang et al. prepared bulk p-type Bi2Te3 materials with layered nanostructures combining melt spinning with spark plasma sintering [120]. The lattice thermal conductivity measured was up to 60% lower than zone melt ingot, and ZT is enhanced up to 70%. Figure 16 shows the lattice thermal conductivity and electrical conductivity of these samples. While the lattice thermal conductivity of all the SPS samples was lower than that of the ingot, one of the samples showed a higher electrical conductivity than the ingot, resulting in the highest ZT of about 1.35 at 300 K.
Figure 16. Electrical conductivity and thermal conductivity of layered nanostructure of Bi2Te3 using melt spinning combined with spark plasma sintering, in comparison with the zone melt ingot. The numbers in the sample name indicate the speed of the roller during the melt spin process in m/s. Reprinted with permission from Ref. [120] X. Tang, W. Xie, H. Li, W. Zhao, Q. Zhang and M. Niino, Appl. Phys. Lett. 90, 012102, (2007) Copyright @ American Physical Society.
Nanocomposites as High Efficiency Thermoelectric Materials
473
Theoretical studies briefly reviewed in Sec. 2 show that nanostructuring of a bulk material can lead to up to an order of magnitude decrease in the lattice thermal conductivity, as seen in Si-Ge material system for instance. However, in Bi2Te3 thin films, the lattice thermal conductivity could only be reduced by a factor of 2 with respect to bulk at room temperature. The reason is likely due to the presence of structural modulations (natural nanostructures) and dislocations in bulk Bi2Te3 which already reduce the lattice thermal conductivity [121], and make the effect of further scattering of phonons less pronounced.
4.2. Medium Temperature Materials Several nanocomposites materials with high ZT in the medium temperature range (400 K to 800 K) have been discovered. These materials could have significant impacts in waste heat recovery for both transportation sectors and industrial exhaust heat. The most prominent ones are based on alloys of PbTe, Mg2Si, skutterudites, etc. 4.2.1. Pbte-Based Nanocomposites Heremans et al. [122] prepared PbTe nanopowders by ball-milling, and then sintered the powders into bulk samples. The nanocomposites showed a slight increase in the Seebeck coefficient over bulk PbTe of the same carrier concentration. It was also found that the scattering parameter showed a slight increase as well, implying the possibility that the nanostructure was responsible for electron energy filtering. In a separate study [110], bulk PbTe samples were prepared in which nanoparticles of excess Pb or Ag metal were precipitated within the PbTe matrix by a tempering anneal process. These samples showed a remarkable enhancement (by up to 100%) in the Seebeck coefficient, and a simultaneous increase in the scattering parameter (which went from < 1 for bulk to about 3-4 in the nanoprecipitates samples). Though the origin of this increase in the scattering parameter is not clear, the effect probably is energy filtering of the electrons, resulting in the high Seebeck coefficient. On the other hand, because the mobility of the
474
Thiagarajan et al.
electrons was too low, no increase in the power factor and ZT was obtained. More recently, Sootsman et al. [115] prepared PbTe with nanoprecipitates of both Pb and Sb simultaneously. This resulted in large enhancement in the power factor over that of bulk PbTe. Remarkably, and rather inexplicably, when the concentration of Sb was 3% and Pb was 2% in the nanocomposite, the electron mobility actually increased with temperature (between 300 K and 450 K). Thought ionic impurity scattering can result in a rising mobility, it is not expected to be dominant at these temperatures. Moreover, simple nanostructuring with either Pb or Sb did not result in the enhancement. Thus, co-nanostructuring seems to result in a novel effect that could probably be extended to other material systems. Some of the highest values of the figure of merit in the medium temperature range have been obtained in the AgSbTe2-(PbTe)m (LAST-m) family of thermoelectric materials [111]. These materials have NaCl structure, with the tellurium occupying the Cl positions, and silver, lead and antimony occupying the Na positions. Thus the anions carry a net charge of -2, while each of the cations carries a net charge of +2. (One pair of Ag+ and Sb3+ may be considered to iso-electronically substitute for two Pb2+ ions in the lattice). Originally, the LAST compounds were considered to be solid solutions of AgSbTe2 and PbTe. Although, according to X-ray diffraction data such as shown in Fig. 17, bulk Ag1-xPbmSbTem+2 specimens with m from 6 to 18 are single-phase, the results of electron diffraction and HRTEM suggest that different microscopic phases co-exist in these specimens [123], which differ in composition. It is always found that a minority phase rich in Ag and Sb is endotaxially embedded in the majority phase poor in Ag and Sb (and rich in Pb). Thus, contrary to the previous understanding [124], AgSbTe2-PbTe do not form solid solutions but exhibit extensive nanostructures caused by compositions fluctuations. High ZT in the order of 2 or more has been demonstrated in the LAST materials at high temperatures. This enhancement of the figure of merit is the result of a very low lattice thermal conductivity, without much loss in the Seebeck coefficient and electrical conductivity. The spontaneously developed nanoscale inhomogeneities act as embedded nanoparticles that scatter phonons, thus reducing the lattice thermal
Nanocomposites as High Efficiency Thermoelectric Materials
475
conductivity. The low lattice thermal conductivity is caused by the increased phonon scattering due to the distribution of three types of atoms with different masses over the lattice positions of one kind. For different m, the compounds AgPbmSbTem+2 demonstrated close values of thermal conductivity, namely, below 0.5 W/mK at 700 K and 1.3 W/mK at room temperature. All compounds of the LAST family exhibit semiconductor properties with a narrow band gap of ~0.25 eV. The electrical conductivity of compounds increases with an increase in m (i.e. the PbTe content), and reaches a maximum at m=18. The LAST materials demonstrated are n-type. Electrons are the predominant charge carriers; hence, the Seebeck coefficient is negative. However p-type materials can be obtained by use for Na in place of Ag [125], or by using Sn in addition to the Ag, Pb, Sb and Te [112, 126].
Figure 17. Typical powder X-ray diffraction pattern obtained for LAST-m samples showing a single phase rock salt-like lattice structure. However, according to electron diffraction and HRTEM, depending upon m and the processing conditions, both nanophase separation and long-range atomic ordering within the nanophases exist. From Ref. [123]: E. Quarez, K. F. Hsu, R. Pcionek, N. Frangis, E. K. Polychroniadis, and M. G. Kanatzidis, Journal of the American Chemical Society 127, 9177 (2005).
476
Thiagarajan et al.
Ab initio electronic structure calculations [127] show that the high power factor of the LAST compounds can be explained by the appearance of resonance states in the lower part of the conduction band and in the upper part of the valence band. Ag atoms introduce new electronic states near the top of the valence band of PbTe; isolated Sb atoms introduce resonant electronic states near the bottom of the PbTe conduction band. The Ag–Sb pairs result in an increase in the density of states right around the band gap, compared to that of pure PbTe. As a result, the Seebeck coefficient and the power factor are increased. However, it is also found that the increase in the power factor is small when compared with the values typical of pure lead telluride; hence, the good thermoelectric ZT values for the LAST compounds were largely due to the nanostructure-induced thermal conductivity reduction [128]. The LAST materials were originally synthesized by mixing the constituent elements, melting them and then cooling slowly to room temperature, leading to the formation of the nanoscale features by phase separation. Later works have shown that it is possible to obtain similar crystallographic structure and thermoelectric performances by using a mechanical alloying and annealing [58, 129] or preparing nanoparticles by hydrothermal synthesis and then compacting via pressure-less sintering, hot pressing and spark plasma sintering [130]. As PbS is immiscible in PbTe, it is possible to use a similar method as used for the preparation of the LAST compounds to obtain phase separated PbTe-PbS alloys. Androulakis et al. [113] prepared (PbTe)1-x(PbS)x and (Pb0.95Sn0.05Te)1-x(PbS)x. These materails were found to contain nanoscale features rich in PbS, resulting in lattice thermal conductivity as low as ~0.4 W/mK at room temperature. As the mobility of the carriers stayed reasonably high (of the order of 100 cm2/Vs), the ZT reached 1.5 at 642 K for the sample with x=0.08. Ikeda et al. performed extensive microstructural studies in the immiscible PbTe-Sb2Te3 system. It was found that rapid solidification of off- and near-eutectic compositions yield a variety of microstructures, from dendritic to lamellar [131]. Starting with the metastable composition Pb2Sb6Te11 close to the eutectic, they were able to obtain nanometer lamellar structures that resemble thin film superlattices [114, 132]. Figures 13 and 14 (b) show the pseudo-binary phase diagram of the
Nanocomposites as High Efficiency Thermoelectric Materials
477
PbTe-Sb2Te3 system, and the naturally formed nanoscale multilayers, respectively. It was also shown that by adjusting the temperature and rate of the transformation process, it is possible to control the lamellar spacing. 4.2.2. Mg2Si-Based Nanocomposites An ideal thermoelectric material should not only have a high ZT, but should also be composed of elements that are abundant, non-toxic and light. That is why Mg2(Si,Sn) based materials have attracted much attention lately [133]. In fact, a reasonably high ZT value of ~1.1 was obtained at 800 K [134] in MgSi0.4Sn0.6 solid solutions, which is comparable to that of PbTe and filled skutterudites. Zhang et al. [135] undertook a microstructure study of high ZT Mg2Si0.4-xSn0.6Sbx alloys. The lattice thermal conductivity of these samples are about 1.5-2.1 W/mK at 300 K, as compared to 7.9 W/mK of Mg2Si and 5.9 W/mK of Mg2Sn. Interestingly, the samples showed in situ formed nanodots by phase separation, similar to that observed in the LAST materials. These naturally formed nanoscale compositional/structural modulations are believed to be responsible for the low value of thermal conductivity in these samples.
4.3. High Temperature Materials Thermoelectric devices that operate in the temperature range of above 800 K are primarily of interest to power generation modules in probes for deep space exploration. Silicon-germanium alloys have been used for making the space-exploration generators. 4.3.1. Si-Ge Nanocomposites Alloys of Si and Ge, which represent a solid solution SixGe1-x are among the very few thermoelectric materials that operate at temperatures of above 1000 K. Elemental silicon and germanium are crystallized in the diamond-like structure. As a result of the rigid and symmetric crystal structure, they exhibit thermal conductivity too high to become good
478
Thiagarajan et al.
thermoelectric materials (150 and 63 W/mK, respectively, at room temperature). However, their thermal conductivity can be reduced to approximately 5-10 W/mK by the formation of a solid solution alloying [136]. The chemical stability of SixGe1-x solid solutions at high temperatures, particularly, against oxidation, and the high figure of merit (<~1) provide the prerequisites for the high-temperature (1000-1200 K) use of thermoelectric materials based on them. Recently much theoretical and experimental effort has been devoted to improving the rather modest figure of merit of Si-Ge alloys, with some success. With nanostructuring by means of ball milling, the thermal conductivity has been lowered than the bulk alloy, and the ZT of both p-type and n-type Si-Ge alloys have been shown to be enhanced [52, 53]. The ZT of n-type Si-Ge nanocomposite, in particular, has exceeded 1 at around 1100 K, for the first time in this system. Figure 18 shows the power factor and thermal conductivity of the Si80Ge20 nanocomposite samples prepared by ball milling and hot pressing. 4.3.2. Lanthanum Chalcogenides Heavily doped lanthanum telluride and other rare-earth chalcogenides have been extensively studied in the past as potential thermoelectric materials due to their excellent thermal stability and high ZT [136]. They were made by either solid-state diffusion or by melt synthesis or a combination of both. However, the high temperature and pressure required for these synthesis processes led to inhomogeneities and a lack of stoichiometric reproducibility. Recently, La3-xTe4 alloys of specified composition were made by May et al. [70] using mechanical alloying and hot pressing. By utilizing much lower temperatures, while maintaining high diffusion rate by the use of mechanically alloyed starting elements (La and Te), they were able to obtain pure and homogeneous alloys of La3-xTe4. With an average crystallite size of about 20-30 nm after hot pressing, the resulting samples showed ZT in excess of 1.1 at 1273 K, which is comparable to the best ZT achieved in n-SiGe alloy.
Nanocomposites as High Efficiency Thermoelectric Materials
479
Figure 18. Temperature dependent power factor and thermal conductivity of p-type Si80Ge20 nanocomposites prepared by ball milling, hot pressing and annealing, in comparison with p-type bulk alloy used in RTGs for space missions. Reprinted with permission from Ref. [52] G. Joshi, H. Lee, Y. Lan, X. Wang, G. Zhu, D. Wang, R. W. Gould, D. C. Cuff, M. Y. Tang, M. S. Dresselhaus, G. Chen, and Z. Ren, Nano Letters 8, 4670 (2008) Copyright @ American Chemical Society.
5. Summary Thermoelectricity has gained much from the recent developments in nanotechnology. Thermoelectric figure of merit ZT has been doubled over the past 15 years from the 50-years non-changing unity to 2 and beyond after the introduction of nano-thermoelectrics. Though still not competitive with traditional mechanical energy conversion technologies, thermoelectricity could have significant impact in energy sectors, such as waste heat recovery for automobiles and solar-thermal utilization where thermal energy are free or very low cost, considering the rising demand for energy and the global warming challenges. A few factors including quantum size effects and electron filtering for enhancing the power factor have been accounted for ZT enhancement in nanostructures, while most experiments show that the dominant mechanisms come from thermal conductivity reduction due to phononinterface scattering. Careful study on thermal conductivity reduction mechanisms in nanostructures has led to the belief that nanocomposites will bring in a paradigm shift in low-cost high efficiency thermoelectric
480
Thiagarajan et al.
materials search. These nanocomposites are expected to have very low thermal conductivity but maintain or even enhance the thermoelectric power factor. Models and simulation tools for electron and phonon transport have been developed to study thermoelectric performances of nanocomposites, primary on semiconductor nanocomposites. These models could be used to qualitatively explain and provide some guidelines for nanocomposite materials design. A quantitative prediction model is highly desirable, but does not exist yet. Various approaches have been practiced for the synthesis of nanocomposites and their constituent nanostructures with some success. Some of these methods combine the state-of-the-art nanoparticle and nanowires synthesis techniques with traditional cold and hot pressing and the recently developed spark plasma sintering. Phase-separation method has also been used to synthesize thermoelectric composites with in situ grown nanostructures. For all the temperature ranges, from room temperature to 1000 K, nanocomposites with enhanced ZT, within a range of 1.2-2, have been successfully synthesized. The major gain in the figure of merit of these nanocomposites came from a suppression of the lattice thermal conductivity. However, most of these recent developments in nanostructured materials have focused on starting from well-known thermoelectric materials. More revolutionary ideas could be practiced using nanocomposites approach and will be needed for low-cost high efficiency thermoelectric materials.
Acknowledgments The authors acknowledge the support from the National Science Foundation through awards CMMI 0729520 and CBET 0846521, and the Air Force Office of Scientific Research through DCT grant FA9550-081-0078 and MURI grant FA9550-06-1-0326. The authors also acknowledge the discussions with Professor Gang Chen and Professor Mildred Dresselhaus at MIT.
Nanocomposites as High Efficiency Thermoelectric Materials
481
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
G. S. Nolas, J. Sharp, and H. J. Goldsmid, Thermoelectrics: basic principles and new materials developments (Springer, Berlin, New York, 2001). D. M. Rowe ed., Thermoelectrics handbook: macro to nano (CRC/Taylor & Francis, Boca Raton, 2006). A. F. Ioffe, Semiconductor thermoelements, and Thermo-electric cooling (Infosearch Ltd., London, 1957). D. T. Morelli, T. Caillat, J. P. Fleurial, A. Borshchevsky, J. Vandersande, B. Chen, and C. Uher, Physical Review B 51, 9622 (1995). T. Caillat, A. Borshchevsky, and J. P. Fleurial, Journal of Applied Physics 80, 4442 (1996). G. S. Nolas, G. A. Slack, and S. B. Schujman, in Recent Trends in Thermoelectric Materials Research I (Academic Press Inc, San Diego, 2001), Vol. 69, p. 255. D. M. Rowe ed., CRC Handbook of Thermoelectrics (CRC Press, Boca Raton, FL, 1995). G. S. Nolas, D. T. Morelli, and T. M. Tritt, Annual Review of Materials Science 29, 89 (1999). B. C. Sales, D. Mandrus, and R. K. Williams, Science 272, 1325 (1996). B. C. Sales, D. Mandrus, B. C. Chakoumakos, V. Keppens, and J. R. Thompson, Physical Review B 56, 15081 (1997). G. S. Nolas, J. L. Cohn, G. A. Slack, and S. B. Schujman, Applied Physics Letters 73, 178 (1998). N. P. Blake, L. Mollnitz, G. Kresse, and H. Metiu, Journal of Chemical Physics 111, 3133 (1999). B. B. Iversen, A. E. C. Palmqvist, D. E. Cox, G. S. Nolas, G. D. Stucky, N. P. Blake, and H. Metiu, Journal of Solid State Chemistry 149, 455 (2000). S. M. Kauzlarich, S. R. Brown, and G. J. Snyder, Dalton Trans., 2099 (2007). L. D. Hicks and M. S. Dresselhaus, Physical Review B 47, 12727 (1993). L. D. Hicks and M. S. Dresselhaus, Physical Review B 47, 16631 (1993). G. Chen, Physical Review B 57, 14958 (1998). M. S. Dresselhaus, G. Chen, M. Y. Tang, R. G. Yang, H. Lee, D. Z. Wang, Z. F. Ren, J. P. Fleurial, and P. Gogna, Advanced Materials 19, 1043 (2007). J. P. Heremans, C. M. Thrush, D. T. Morelli, and M. C. Wu, Physical Review Letters 88, 216801 (2002). T. C. Harman, D. L. Spears, and M. J. Manfra, Journal of Electronic Materials 25, 1121 (1996). T. C. Harman, D. L. Spears, and M. P. Walsh, Journal of Electronic Materials 28, L1 (1999). T. C. Harman, P. J. Taylor, D. L. Spears, and M. P. Walsh, Journal of Electronic Materials 29, L1 (2000). T. C. Harman, P. J. Taylor, M. P. Walsh, and B. E. LaForge, Science 297, 2229 (2002).
482
Thiagarajan et al.
24.
T. C. Harman, M. P. Walsh, B. E. Laforge, and G. W. Turner, Journal of Electronic Materials 34, L19 (2005). T. C. Harman, R. E. Reeder, M. P. Walsh, B. E. LaForge, C. D. Hoyt, and G. W. Turner, Applied Physics Letters 88, 243504 (2006). C. J. Vineis, T. C. Harman, S. D. Calawa, M. P. Walsh, R. E. Reeder, R. Singh, and A. Shakouri, Physical Review B 77, 235202 (2008). R. Venkatasubramanian, E. Siivola, T. Colpitts, and B. O’Quinn, Nature 413, 597 (2001). B. Moyzhes and V. Nemchinsky, Applied Physics Letters 73, 1895 (1998). K. Nishio and T. Hirano, Japanese Journal of Applied Physics 36, 170 (1997). L. W. Whitlow and T. Hirano, Journal of Applied Physics 78, 5460 (1995). J. M. O. Zide, D. Vashaee, Z. X. Bian, G. Zeng, J. E. Bowers, A. Shakouri, and A. C. Gossard, Physical Review B 74, 205335 (2006). W. Kim, S. L. Singer, A. Majumdar, D. Vashaee, Z. Bian, A. Shakouri, G. Zeng, J. E. Bowers, J. M. O. Zide, and A. C. Gossard, Applied Physics Letters 88, 242107 (2006). J. P. Heremans, Acta Physica Polonica A 108, 609 (2005). G. J. Snyder and E. S. Toberer, Nature Materials 7, 105 (2008). A. V. Shevelkov, Uspekhi Khimii 77, 3 (2008). G. Chen, IEEE Transactions on Components and Packaging Technologies 29, 238 (2006). B. Yang and G. Chen, Physical Review B 67, 195311 (2003). G. Chen, Recent Trends in Thermoelectric Materials Research III 71, 203 (2001). R. G. Yang and G. Chen, Physical Review B 69, 195316 (2004). W. Tian and R. Yang, Journal of Applied Physics 101, 054320 (2007). N. Scoville, C. Bajgar, J. Rolfe, J. P. Fleurial, and J. Vandersande, Nanostruct Mater 5, 207 (1995). D. J. Bergman and O. Levy, Journal of Applied Physics 70, 6821 (1991). D. J. Bergman and L. G. Fel, Journal of Applied Physics 85, 8205 (1999). G. Chen, D. Borca-Tascuic, and R. G. Yang, in Encyclopedia of Nanoscience and Nanotechnology, edited by H. S. Nalwa (American Scientific Publishers, 2004), Vol. 7, p. 429. R. G. Yang and G. Chen, Physical Review B 69 (2004). R. G. Yang, G. Chen, and M. S. Dresselhaus, Physical Review B 72, 125418 (2005). M. S. Jeng, R. G. Yang, D. Song, and G. Chen, Journal of Heat TransferTransactions of the ASME 130, 042410 (2008). G. A. Slack and M. A. Hussain, Journal of Applied Physics 70, 2694 (1991). C. B. Vining, W. Laskow, J. O. Hanson, R. R. Vanderbeck, and P. D. Gorsuch, Journal of Applied Physics 69, 4333 (1991). R. Yang and G. Chen, in SAE World Congress (Society of Automotive Engineers, 2006), Article Number 2006-01-0289.
25. 26. 27. 28. 29. 30. 31. 32.
33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.
45. 46. 47. 48. 49. 50.
Nanocomposites as High Efficiency Thermoelectric Materials 51. 52. 53.
54. 55.
56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73.
483
Y. I. Ravich, in CRC Handbook of Thermoelectrics, edited by D. M. Rowe (CRC Press, Boca Raton, FL, 1995), p. 65. G. Joshi, H. Lee, Y. Lan, X. Wang, G. Zhu, D. Wang, R. W. Gould, D. C. Cuff, M. Y. Tang, M. S. Dresselhaus, G. Chen, and Z. Ren, Nano Letters 8, 4670 (2008). X. W. Wang, H. Lee, Y. C. Lan, G. H. Zhu, G. Joshi, D. Z. Wang, J. Yang, A. J. Muto, M. Y. Tang, J. Klatsky, S. Song, M. S. Dresselhaus, G. Chen, and Z. F. Ren, Applied Physics Letters 93, 193121 (2008). S. V. Faleev and F. Leonard, Physical Review B 77, 214304 (2008). B. Poudel, Q. Hao, Y. Ma, Y. Lan, A. Minnich, B. Yu, X. Yan, D. Wang, A. Muto, D. Vashaee, X. Chen, J. Liu, M. S. Dresselhaus, G. Chen, and Z. Ren, Science 320, 634 (2008). Z. A. Munir, U. Anselmi-Tamburini, and M. Ohyanagi, Journal of Materials Science 41, 763 (2006). J. Jiang, L. D. Chen, S. Q. Bai, Q. Yao, and Q. Wang, Materials Science and Engineering B-Solid State Materials for Advanced Technology 117, 334 (2005). H. Wang, J. F. Li, C. W. Nan, M. Zhou, W. S. Liu, B. P. Zhang, and T. Kita, Applied Physics Letters 88, 092104 (2006). J. Tani and H. Kido, Physica B-Condensed Matter 364, 218 (2005). A. I. Gusev and A. A. Rempel, Nanocrystalline Materials (Cambridge Int Science Publishing, 2004). M. Umemoto, Materials Transactions JIM 36, 373 (1995). R. M. Davis and C. C. Koch, Scripta Metallurgica 21, 305 (1987). N. Bouad, R. M. Marin-Ayral, and J. C. Tedenac, Journal of Alloys and Compounds 297, 312 (2000). K. Kishimoto and T. Koyanagi, Journal of Applied Physics 92, 2544 (2002). N. Bouad, M. C. Record, J. C. Tedenac, and R. M. Marin-Ayral, Journal of Solid State Chemistry 177, 221 (2004). P. Pierrat, A. Dauscher, B. Lenoir, R. Martin-Lopez, and H. Scherrer, Journal of Materials Science 32, 3653 (1997). H. C. Kim, T. S. Oh, and D. B. Hyun, Journal of Physics and Chemistry of Solids 61, 743 (2000). J. Schilz, M. Riffel, K. Pixius, and H. J. Meyer, Powder Technology 105, 149 (1999). J. Y. Yang, Y. H. Chen, J. Y. Peng, X. L. Song, W. Zhu, J. F. Su, and R. G. Chen, Journal of Alloys and Compounds 375, 229 (2004). A. F. May, J. P. Fleurial, and G. J. Snyder, Physical Review B 78, 125205 (2008). X. B. Zhao, X. H. Ji, Y. H. Zhang, T. J. Zhu, J. P. Tu, and X. B. Zhang, Applied Physics Letters 86, 062111 (2005). N. Gothard, X. Ji, J. He, and T. M. Tritta, Journal of Applied Physics 103, 054314 (2008). H. L. Ni, X. B. Zhao, T. J. Zhu, X. H. Ji, and J. P. Tu, Journal of Alloys and Compounds 397, 317 (2005).
484
Thiagarajan et al.
74.
W. Wang, W. Wang, B. Poudel, J. Yang, D. Z. Wang, and Z. F. Ren, Journal of the American Chemical Society 127, 13792 (2005). S. H. Yu, J. Yang, Y. S. Wu, Z. H. Han, J. Lu, Y. Xie, and Y. T. Qian, Journal of Materials Chemistry 8, 1949 (1998). J. L. Mi, X. B. Zhao, T. J. Zhu, J. P. Tu, and G. S. Cao, Journal of Alloys and Compounds 417, 269 (2006). J.-L. Mi, X.-B. Zhao, T.-J. Zhu, and J.-P. Tu, Journal of Inorganic Materials 23, 715 (2008). F. L. A. Purkayastha, S. Kim, T. Borca-Tasciuc, G. Ramanath, Advanced Materials 18, 496 (2006). W. Lu, Y. Ding, Y. Chen, Z. L. Wang, and J. Fang, Journal of the American Chemical Society 127, 10112 (2005). W. W. Genqiang Zhang, Xiaoguang Li, Advanced Materials 20, 3654 (2008). A. I. Hochbaum, R. K. Chen, R. D. Delgado, W. J. Liang, E. C. Garnett, M. Najarian, A. Majumdar, and P. D. Yang, Nature 451, 163 (2008). F. Xiao, C. Hangarter, B. Y. Yoo, Y. W. Rheem, K. H. Lee, and N. V. Myung, Electrochimica Acta 53, 8103 (2008). G. J. Snyder, J. R. Lim, C. K. Huang, and J. P. Fleurial, Nature Materials 2, 528 (2003). L. J. Chen, H. N. Hu, Y. X. Li, G. F. Chen, S. Y. Yu, and G. H. Wu, Chemistry Letters 35, 170 (2006). M. Martin-Gonzalez, A. L. Prieto, R. Gronsky, T. Sands, and A. M. Stacy, Advanced Materials 15, 1003 (2003). C. G. Jin, X. Q. Xiang, C. Jia, W. F. Liu, W. L. Cai, L. Z. Yao, and X. G. Li, Journal of Physical Chemistry B 108 1844 (2004). C. G. Jin, G. Q. Zhang, T. Qian, X. G. Li, and Z. Yao, Journal of Physical Chemistry B 109, 1430 (2005). M. Martin-Gonzalez, G. J. Snyder, A. L. Prieto, R. Gronsky, T. Sands, and A. M. Stacy, Nano Letters 3, 973 (2003). W. F. Liu, W. L. Cai, and L. Z. Yao, Chemistry Letters 36, 1362 (2007). W. Wang, J. F. Qu, X. L. Lu, G. Q. Zhang, G. Li, and X. G. Li, Materials Science Forum 546-549, 2171 (2007). L. Li, Y. W. Yang, X. H. Huang, G. H. Li, and L. D. Zhang, Nanotechnology 17 1706 (2006). L. Trahey, C. R. Becker, and A. M. Stacy, Nano Letters 7, 2535 (2007). L. Li, Y. W. Yang, X. H. Huang, G. H. Li, R. Ang, and L. D. Zhang, Applied Physics Letters 88 103119 (2006). X. H. Li, B. Zhou, L. Pu, and J. J. Zhu, Crystal Growth & Design 8, 771 (2008). B. Zhou, X. H. Li, and J. J. Zhu, Crystal Growth & Design 7, 2276 (2007). F. Xiao, B. Y. Yoo, K. N. Bozhilov, K. H. Lee, and N. V. Myung, Journal of Physical Chemistry C 111, 11397 (2007).
75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96.
Nanocomposites as High Efficiency Thermoelectric Materials 97. 98. 99. 100. 101. 102.
103. 104. 105. 106. 107. 108. 109.
110. 111. 112. 113.
114. 115. 116. 117.
485
E. J. Menke, Q. Li, and R. M. Penner, Nano Letters 4, 2009 (2004). F. H. Xue, G. T. Fei, B. Wu, P. Cui, and L. D. Zhang, Journal of the American Chemical Society 127 15348 (2005). B. Yoo, F. Xiao, K. N. Bozhilov, J. Herman, M. A. Ryan, and N. V. Myung, Advanced Materials 19 296 (2007). W. Wang, G. Q. Zhang, and X. G. Li, Journal of Physical Chemistry C 112, 15190 (2008). W. Wang, X. L. Lu, T. Zhang, G. Q. Zhang, W. J. Jiang, and X. G. Li, Journal of the American Chemical Society 129, 6702 (2007). M. S. Dresselhaus, G. Chen, M. Y. Tang, R. G. Yang, H. Lee, D. Z. Wang, Z. F. Ren, J. P. Fleurial, and P. Gogna, in Materials and Technologies for direct Thermal-to-Electric Energy Conversion, MRS Symposium Proceedings, edited by J. Yang, T. P. Hogan, R. Funahashi and G. S. Nolas, Pittsburgh, PA, 2005, Vol. 886, p. 3. K. S. Suslick, Science 247, 1439 (1990). X. F. Qiu, J. J. Zhu, L. Pu, Y. Shi, Y. D. Zheng, and H. Y. Chen, Inorganic Chemistry Communications 7, 319 (2004). Y. Y. Zheng, T. J. Zhu, X. B. Zhao, J. P. Tu, and G. S. Cao, Materials Letters 59, 2886 (2005). J. P. Ge and Y. D. Li, Journal of Materials Chemistry 13, 911 (2003). B. Zhang, J. He, and T. M. Tritt, Applied Physics Letters 88, 043119 (2006). Y. Chu, X. Tang, W. Zhao, and Q. Zhang, Crystal Growth & Design 8, 208 (2008). J. P. Heremans, in Materials and Technologies for direct Thermal-to-Electric Energy Conversion, MRS Symposium Proceedings, edited by J. Yang, T. P. Hogan, R. Funahashi and G. S. Nolas, Pittsburgh, PA, 2005, Vol. 886, p. F04. J. P. Heremans, C. M. Thrush, and D. T. Morelli, Journal of Applied Physics 98, 063703 (2005). K. F. Hsu, S. Loo, F. Guo, W. Chen, J. S. Dyck, C. Uher, T. Hogan, E. K. Polychroniadis, and M. G. Kanatzidis, Science 303, 818 (2004). J. Androulakis, K. F. Hsu, R. Pcionek, H. Kong, C. Uher, J. J. DAngelo, A. Downey, T. Hogan, and M. G. Kanatzidis, Advanced Materials 18, 1170 (2006). J. Androulakis, C. H. Lin, H. J. Kong, C. Uher, C. I. Wu, T. Hogan, B. A. Cook, T. Caillat, K. M. Paraskevopoulos, and M. G. Kanatzidis, Journal of the American Chemical Society 129, 9780 (2007). T. Ikeda, L. A. Collins, V. A. Ravi, F. S. Gascoin, S. M. Haile, and G. J. Snyder, Chemistry of Materials 19, 763 (2007). J. R. Sootsman, H. Kong, C. Uher, J. J. D’Angelo, C. I. Wu, T. P. Hogan, T. Caillat, and M. G. Kanatzidis, Angewandte Chemie-International Edition 47, 8618 (2008). Y. Ma, Q. Hao, B. Poudel, Y. C. Lan, B. Yu, D. Z. Wang, G. Chen, and Z. F. Ren, Nano Letters 8, 2580 (2008). Y. Q. Cao, X. B. Zhao, T. J. Zhu, X. B. Zhang, and J. P. Tu, Applied Physics Letters 92, 143106 (2008).
486
Thiagarajan et al.
118. X. Ji, B. Zhang, T. M. Tritt, J. W. Kolis, and A. Kumbhar, Journal of Electronic Materials 36, 721 (2007). 119. X. H. Ji, J. He, Z. Su, N. Gothard, and T. M. Tritt, Journal of Applied Physics 104, 034907 (2008). 120. X. F. Tang, W. J. Xie, H. Li, W. Y. Zhao, Q. J. Zhang, and M. Niino, Applied Physics Letters 90, 012102 (2007). 121. N. Peranio, O. Eibl, and J. Nurnus, Journal of Applied Physics 100, 114306 (2006). 122. J. P. Heremans, C. M. Thrush, and D. T. Morelli, Physical Review B 70, 115334 (2004). 123. E. Quarez, K. F. Hsu, R. Pcionek, N. Frangis, E. K. Polychroniadis, and M. G. Kanatzidis, Journal of the American Chemical Society 127, 9177 (2005). 124. F. D. Rosi, E. F. Hockings, and N. E. Lindenblad, RCA Review 22, 82 (1961). 125. P. F. R. Poudeu, J. D’Angelo, A. D. Downey, J. L. Short, T. P. Hogan, and M. G. Kanatzidis, Angewandte Chemie-International Edition 45, 3835 (2006). 126. A. Kosuga, K. Kurosaki, H. Muta, and S. Yamanaka, Journal of Alloys and Compounds 416, 218 (2006). 127. D. Bilc, S. D. Mahanti, E. Quarez, K. F. Hsu, R. Pcionek, and M. G. Kanatzidis, Physical Review Letters 93, 146403 (2004). 128. D. I. Bilc, S. D. Mahanti, and M. G. Kanatzidis, Physical Review B 74, 125202 (2006). 129. M. Zhou, J. F. Li, and T. Kita, Journal of the American Chemical Society 130, 4527 (2008). 130. K. F. Cai, C. Yan, Z. M. He, J. L. Cui, C. Stiewe, E. Müller, and H. Li, Journal of Alloys and Compounds (In Press). 131. T. Ikeda, S. M. Haile, V. A. Ravi, H. Azizgolshani, F. Gascoin, and G. J. Snyder, Acta Materialia 55, 1227 (2007). 132. T. Ikeda, V. A. Ravi, L. A. Collins, S. M. Haile, and G. J. Snyder, Journal of Electronic Materials 36, 716 (2007). 133. G. S. Nolas, D. Wang, and M. Beekman, Physical Review B 76, 235204 (2007). 134. V. K. Zaitsev, M. I. Fedorov, E. A. Gurieva, I. S. Eremin, P. P. Konstantinov, A. Y. Samunin, and M. V. Vedernikov, Physical Review B 74, 045207 (2006). 135. Q. Zhang, J. He, T. J. Zhu, S. N. Zhang, X. B. Zhao, and T. M. Tritt, Applied Physics Letters 93, 102109 (2008). 136. C. Wood, Reports on Progress in Physics 51, 459 (1988).
CHAPTER 11 NANOSTRUCTURED MATERIALS FOR HYDROGEN STORAGE
Saghar Sepehri and Guozhong Cao Department of Materials Science and Engineering, University of Washington, Seattle, WA 98195, USA Email:
[email protected],
[email protected]
Hydrogen generated from clean and renewable energy sources has been considered as an alternate energy carrier for several decades. Although many advances in hydrogen production and usage have been made, storing hydrogen remains a significant challenge. Many drawbacks including energy intensive processes, low volumetric densities, and safety concerns are associated with storing hydrogen as compressed or liquefied. Solid state hydrogen storage is considered to be the most promising method as a safe and effective storage option, but there is still no material or method that satisfies the requirements for a practical approach. A feasible hydrogen storage media should address several issues including targeted storage capacities, thermodynamics and hydrogen sorption kinetics, and safety. Nanostructured materials with their unique properties can provide new capabilities through tailormade structures by offering improved bonding environment for storing hydrogen in molecular or atomic form. This chapter describes some of the recent developments in application of nanostructures for solid state hydrogen storage.
1. Introduction During the recent years, significant progress has been made in the development of alternative energy technologies. Hydrogen has the potential to be a good energy carrier candidate in a carbon-free emission
487
488
Sepehri and Cao
cycle. It is abundant in nature, has high energy density per mass and burns cleanly. However, hydrogen combines easily with other elements and is usually found as bonded form in chemical compounds, therefore, releasing hydrogen requires energy. Many questions must be addressed before hydrogen can be used as the primary energy carrier through a ‘hydrogen economy’. Figure 1 shows the three components of the hydrogen economy: production, storage, and usage. [1]
Fig. 1. The hydrogen economy network is consisted of primary energy sources linked to multiple end uses through hydrogen as an energy carrier. [G.W. Crabtree, M.S. Dresselhaus, and M.V. Buchanan, Phys. Today 57, 39 (2004)]
Hydrogen can be produced from primary energy sources (including fossil fuels, nuclear and renewable energies) and used in fuel cells. Effective and safe hydrogen storage is a challenging step in the hydrogen technology and considerable efforts have been made in synthesizing and investigating novel materials for hydrogen storage in the past decade. Ideally, hydrogen should be stored in such a way to attain high storage capacity under near the ambient conditions to be safe and economical, and perform rapid and reversible hydrogenation and dehydrogenation process for practical applications. Although various techniques and materials have been used or studied to store hydrogen, there is neither method nor material that satisfies all the requirements for perceived hydrogen economy. [2]
Nanostructured Materials for Hydrogen Storage
489
Hydrogen can be stored as gas, liquid, or solid. Storing hydrogen as compressed gas needs high pressure and heavy containers to support such pressure. Compressed hydrogen is highly flammable. Liquification of hydrogen needs energy and consumes more than 20% of the recoverable energy. Also strong cryogenic containers that stand the high pressure should be used decrease the hydrogen boil-off. Storing hydrogen in solids may offer the best option to avoid the hazard potential in compressed and energy losses in liquefied hydrogen. Solid state hydrogen storage can happen through two basic mechanisms: physisorption (or physical adsorption) and chemisorptions (or chemical adsorption). In physisorption, molecular hydrogen is adsorbed by weak intermolecular (van der Waals) forces. Examples of physisorption include storing of hydrogen in carbon structures and organic or inorganic frameworks. In chemisorption, hydrogen molecules and chemical bonding of the hydrogen atoms dissociate by integration in the lattice of a metal, an alloy, or by formation of a new chemical compound. Metal, chemical and complex hydrides are examples of chemisorption. Each principle has its own prospects and limitations. Chemisorption may provide high volumetric and gravimetric storage capacities, but the chemical bonds need to split or recombine to release hydrogen. Storing hydrogen by physisorption is not subject to this limitation, because the hydrogen stays in its molecular form, but the challenge is to provide materials with a sufficient amount of bonding sites for the hydrogen per volume to achieve high storage capacity. One of the major differences between physisorption and chemisorption is their binding energies (Fig. 2). Physisorption binding is usually too weak (<10 kJ/mol), thus demands cryogenic temperatures for significant storage capacity. Chemisorption shows a stability that is too high (>50 kJ/mol) and demands high desorption temperatures. [3] To achieve an ideal binding energy (in the range of 10–60 kJ/mol), we need to increase the physisorption binding energy or reduce the chemisorption binding energy.
490
Sepehri and Cao
Fig. 2. Bond strengths for physisorption and chemisorption and the desirable range of binding energies that allow hydrogen release around room temperature. [V. Bérubé, G. Radtke, M. Dresselhaus, and G. Chen, Int. J. Energy Res. 31, 637 (2007)]
Nanostructured materials with their unique physical, chemical, thermodynamic and kinetics properties can provide effective and sufficient ways to address the challenges involve in hydrogen storage. [4] At nanoscale materials can have noticeably different properties than their bulk analogs. Nanostructured materials can affect the thermodynamics and kinetics of hydrogen adsorption and dissociation by increasing diffusion rate and decreasing the diffusion path. [5] They also have the potential for high surface areas and hybrid structures that allow multifunctional performances for hydrogen storage systems. This chapter gives an overview of the current achievements in developing nanostructured materials and methods for hydrogen storage. 2. Hydrogen Storage by Physisorption Physisorption is a principle where the forces involved are weak intermolecular forces, therefore; in general it is associated with fast kinetics and reversibility. But the challenge with the physisorption of hydrogen also results from these weak forces. H2 is the smallest molecule and only has two electrons, hence it is hard to polarize and in the absence of relatively strong polarizing centers, interaction between the adsorbent and the non-polar hydrogen molecules relies on the weak dispersion forces which created by temporarily induced dipoles, and are typically of the order of 3–6 kJ/mol. [6] Thus, significant hydrogen adsorption often takes place only at a cryogenic temperature. Nanostructured materials may offer advantages for molecular hydrogen storage by providing high surface areas, or by encapsulation or trapping hydrogen in microporous
Nanostructured Materials for Hydrogen Storage
491
media. Using porous nanostructured materials, in general, can reduce the gravimetric and the volumetric storage densities. However, the increased surface area and porosity in nanostructures will offer additional binding sites on the surface and in the pores that could increase storage mainly through physisorption. The possibility of storing a significant amount of hydrogen on high surface area materials has been a key driver in the investigation of hydrogen sorption properties of various porous materials including, nanostructured carbons (nanotubes, graphite sheets, and template ordered porous carbons), zeolites, metal-organic frameworks, clathrates, and polymers with intrinsic microporosity (pore sizes <2 nm). 2.1. Nanostructured Carbon Carbon materials with high surface area, good chemical stability, and low density have received considerable attraction. Each carbon atom has six electrons which occupy 1s2, 2s2, and 2p2 ground-state atomic orbitals. Carbon can hybridize into a sp, sp2, or sp3 configuration by the promotion of the electron in 2s orbital to the empty p orbital. The various bonding states are connected with certain structural arrangements, so that sp bonding gives rise to chain structure, sp2 bonding to planar structures, and sp3 bonding to tetrahedral structures. [7] The structural and practical properties of carbon critically depend on the ratio between the number of sp2 (graphite-like) and sp3 (diamond-like) bonds, resulting in various hybridized states of nanostructured carbon materials including carbon nanofibers (CNF), graphitic nanofibers (GNF), multiwalled carbon nanotubes (MWNT), single walled carbon nanotubes (SWNT), carbon nanorods, and carbon aerogels. Examples of carbon- based nanomaterials are shown in Fig. 3. [8]
Fig. 3. Hybridization states of carbon-based nanomaterials with novel and distinct properties. [M. S. Mauter and M. Elimelech, Environ. Sci. Technol. 42, 5843 (2008)]
492
Sepehri and Cao
Carbon nanotubes are made from graphitic layers or graphitic- like materials. Early reports [9, 10] on hydrogen storage in carbon nanotubes and graphitic nanofibers, proposed high storage capacities (to 67 wt.%) and started an extensive worldwide surge of research. Since then many succeeding experiments were carried out with different methods, but such high values have not yet been reproduced by other groups. [11] Furthermore, no hypothesis could support the unusually high storage capacities and the high storage capacity results are believed to be related to the faults of experiment. [12, 13] Nevertheless, hydrogen adsorption on carbon materials is still an attractive and improving filed. Pores in a porous carbon are classified with respect to their width, as micropores (<2 nm), mesopores (2-50 nm), and Macropores (>50 nm). The result of several investigations proposes that the amount of adsorbed hydrogen is proportional to the specific surface area and the amount of subnanometer diameter pores (moicropore volume) in the carbon material. [14, 15] For surface greater than 4000 m2/g a possible gravimetric density of 6 wt.% is suggested but not yet reached. [14] In case of activated carbons and activated carbon fibers, the hydrogen absorption of 5 wt.% is obtained at low temperature (77 K) and high pressure (30 to 60 bar). [16] For GNF, SWNT, and MWNT, the reversible hydrogen uptake of 1.5 wt.% per 1000 m2/g under ambient conditions is reported. [17]. Maximum hydrogen adsorption for carbon nanotubes is found to be about 4 wt.% at 77 K, and less than 1 wt.% at ambient temperatures. [18] Ordered porous carbon can be obtained by using an ordered porous solid as a template. For template ordered porous carbon with surface area of 3200 m2/g, hydrogen capacity of 7 wt.%, at 77 K and 20 bar, is reported. [19] Carbon aerogels (CAs) are another class of amorphous porous carbon structures with high surface area and tunable porous structure. Recent studies on carbon aerogels (CAs) shows more hydrogen storage densities for CAs with higher surface areas with a 5 wt.% of hydrogen adsorption for surface area of 3200 m2/g at 77 K and pressure range of 20-30 bar (Fig. 4). [20] Increasing the surface area and micropore volume can improve the hydrogen uptake. Latest research in hydrogen storage by physisorption on nanostructured carbon material involves efforts on increasing surface
Nanostructured Materials for Hydrogen Storage
493
area of carbon to provide more binding sites, and incorporating functional groups (dopants) in carbon to increase the binding energy between hydrogen and carbon surface. [21]
Fig. 4. Adsorption isotherms at 77K for the carbon aerogels show the linear dependency of hydrogen adsorption to the surface area. [H. Kabbour, T. F. Baumann, J. H. Satcher, Jr., A. Saulnier, and C. C. Ahn, Chem. Mater. 18, 6085, (2006)]
2.2. Zeolites Zeolite is an inorganic porous material consisted of hydrated aluminosilicate mineral with highly regular structure that exhibit reversible occlusion of gases. However, because of the high density of the aluminosilicate framework, which contains Si, Al, O, and heavy cations, a high gravimetric hydrogen storage density might not be achieved. On the other hand, zeolites can be ideal choice in studying the hydrogen binding because of their well known crystal structure and easy ion exchange and those studies may provide insight valuable for work on other hydrogen adsorbents. The working principle of hydrogen storage in zeolites is that the guest molecules under high temperature and pressure are forced into the cavities of the molecular sieve host. Upon cooling to room temperature or below, hydrogen is trapped inside the cavities and it can be released again by raising the temperature. The amount of
494
Sepehri and Cao
encapsulated hydrogen in zeolite is related to the size of the exchanged cation and higher storage capacity is observed for zeolites with high number of small cavities in their structure. [22] It has been showed that the storage capacity of zeolite is strongly dependent on the temperature, the type of framework and cation, micropore surface area and micropore volume. [23] Hydrogen uptake of 1.81 wt.% (at 15 bar and 77K) was obtained for NaY zeolite. [24] The theoretical maximum possible hydrogen storage capacity of zeolites is less than 3 wt.%. [25] Large mass of the zeolite framework is a limiting factor in the storage capacity of zeolites, therefore using light elements in the framework and enhancing the interaction energy between hydrogen and the framework can enhance the storage capacity. 2.3. Metal–Organic Frameworks Metal–organic frameworks (MOFs) are crystalline solid compounds consisting of organic ligands connecting metal ions or clusters that form a cage network. Most MOFs have a three-dimensional interconnected porous framework with uniform pores that provides an ordered network of channels. MOFs can be synthesized using a self–assembly process that allows different combinations of organic linkers and provides a wide range of different functionality and pore size. [26] Figure 5 shows an example of MOF structure, consisting of Zn4O(L)3 with Zn4O clusters at the corners linked by linear carboxylates L. [4]
Fig. 5. The prototypical MOF structure of Zn4O(L)3, Zn4O clusters located at the corners are linked by linear carboxylates L resulting in cubic framework. The large void inside MOF is shown by the sphere. [M. Fichtner, Adv. Eng. Mater. 7, 443 (2005)]
Nanostructured Materials for Hydrogen Storage
495
MOFs can provide light porous framework with high surface areas and pore volumes. Surface area of MOFs are usually in the range of 500–3000 m2/g while values higher than 5000 m2/g are also attainable. [27] Large pore volumes (1.1 cm3/g) is observed for some MOFs. [28] Similar to carbon nanostructures, hydrogen storage capacity of MOFs increases with the surface area and micropores (less than 2 nm) volume. [5] Hydrogen adsorption capacity in MOFs is temperature dependent. At low temperature (77 K) and high pressure (70-90 bar) MOFs with hydrogen adsorption of 7 wt.% are reported but at 298 K and 90 bar the maximum observed hydrogen adsorption is only 1.4 wt.%. [29] One approach to improve the temperature dependence of adsorption can include adsorption of hydrogen at high pressures followed by storing at lower pressures. [30] Increasing the interaction of hydrogen with the organic ligands and metal centers in MOFs can improve the hydrogen adsorption at ambient temperature. Several approaches including obtaining MOFs with more polarized cations, impregnation, catalyzed dissociative adsorption, and optimization of pore size are being investigated. [31, 32] 2.4. Clathrates Clathrates are crystalline structures consisting of hydrogen-bonded water framework as the ‘host’ lattice providing cavities which hold ‘guest’ molecules. Several natural gases including methane and carbon dioxide are well-known to form water clathrates or clathrate hydrate. [33] Since the first reports on hydrogen clathrates, they have been attracted attention as possible hydrogen storage materials. [34] However, there are several barriers to practical hydrogen storage application of clathrates, including: slow kinetics for enclathration, relatively low hydrogen storage capacities, and high pressure required for formation and stabilization of clathrates. Hydrogen-bonded water molecules can produce polyhedral small and large cages around guest molecules to form solid clathrate hydrates. When these cages are empty, they are not stable and may collapse into ice crystal structure, but inclusion of gas molecules can stabilize the cages. Three common types of gas hydrate structures can form
496
Sepehri and Cao
depending on the size of the guest molecules: sI hydrate, consisting of 46 water molecules that form two pentagonal dodecahedron and six tetrakaidecahedron, that gives 512 and 512612 cages in a unit cell the sII hydrate, consisting of 136 water molecules tht form sixteen 512 and eight 51264 cages in a unit cell; and the sH hydrate consisting of 36 water molecules that form three 512, two 435663, and one 51268 cages in a unit cell. Formation and stability of H2/H2O clathrates need high pressure and low temperatures, therefore, researches on facilitating the formation and stabilizing the structures are necessary. [35] A binary hydrogen-water clathrate is reported to contain 5.3 wt.% hydrogen at 250 K and very high pressure (2 kbar), while comparing the size of H2 cluster and the volume of the cage cavities suggests that two molecules of H2 were located in each of the 512 cages and four molecules in each of 51264 cages (Fig. 6). [35, 36]
Fig. 6. sII clathrate hydrate: (A) in pure hydrogen hydrate H2 occupies all cage, (B) in H2/THF hydrate H2 occupies the smaller 512 cages and THF occupies the larger 51264 cages, (C) H2/THF hydrate with H2 is in both 512 and 51264 and THF in 51264 cages. [Y. H. Hu, and E. Ruckenstein, Angew. Chem., Int. Ed. 45, 2011 (2006)].
To ease the fabrication process and reduce the synthesis pressure, a second guest component (such as tetrahydrofuran) can be used to fill the cages of clathrate, Tetrahydrofuran acts as stabilizer and can reduced the pressure for hydrogen enclathration significantly but this approach decreases the hydrogen storage capacity of the clathrate to less than 4 wt.% (at 270 K and 120 bar). [37]
Nanostructured Materials for Hydrogen Storage
497
Although, a system that combines high hydrogen storage capacities, moderate pressure, and reasonable kinetic of clathrate formation is yet to be found, clathrates have opened up an interesting field for further research. 2.5. Polymers with Intrinsic Microporosity Most organic polymers posses sufficient conformational flexibility to pack space efficiently and thus do not offer microporosity and high surface areas. However, polymers with intrinsic microporosity (PIMs) are composed wholly of fused-ring sub-units producing a highly rigid and contorted macromolecular structure that is unable to pack space efficiently. PIMs are microporous materials with interconnected porosity and large accessible surface areas (500-2000 m2/g), which makes them potential candidate for hydrogen storage application. [38] Similar to other organic polymers, PIMs can be synthesized by various methods. They can be made by dioxane-forming reaction as insoluble network polymers (including hexaazatrinaphtylene (HATNPIM), cyclotricatechylene (CTC-PIM), porphyrin (Porph-PIM), and triptycene (Trip-PIM) subunits), or as solvent processable non-network polymers such as PIM-1 and PIM-7 (Fig. 7). [39] PIMs can also be fabricated through intensive corsslinking of sovent-swollen, chloromethylated polystyrene beads which results in HyperCrosslinked Plymers (HCP) with microporous structure and high surface area (up to 2000 m2/g). [39] The hydrogen adsorption of PIMs is found to depend on the micropore distribution, with higher concentration ultramicropores (smaller than 0.7 nm) results in improved hydrogen uptake at 77 K. For network-PIM and hyper-cross linked polymer a maximum hydrogen adsorption (at 77 K and 10-15 bar) of 3 wt.% is observed. [40, 39] However, hydrogen adsorption at near ambient temperature can be much lower than at 77 K due to the weak interaction between the hydrogen and polymer. In order to improve the hydrogen adsorption on PIMS they must be optimized further to improve their porosity and increase the micropores concentration.
498
Sepehri and Cao
Fig. 7. Examples of molecular structures of PIMs made by dioxane- forming reactions as network (HATN-PIM, CTC-PIM, Porph-PIM, and Trip-PIM) or non-network (PIM-1 and PIM-7) polymers. [P. M. Budd, A. Butler, J. Slbie, K. Mahmood, N. B. McKeown, G. Ghanem, K. Msayib, D. Book, and A. Walton, Phys. Chem. Chem. Phys. 9, 1802 (2007)].
3. Hydrogen Storage by Chemisorption Chemisorption is the adsorption of a particle with the formation of a chemical bond. Hydrogen can be stored in hydrides (hydrogen rich materials) by chemisorptions to offer high storage capacity at ambient conditions. Different hydrides (metal, chemical, and complex) with high hydrogen densities have been studied as hydrogen storage materials. The volumetric and gravimetric hydrogen densities of some selected hydrides are compared with other hydrogen storage methods in Fig. 8. [41] Although hydrides are good candidates for hydrogen storage but various hydrides suffer from a range of drawbacks such as poor reversibility, poor thermal conductivity, and relatively high dehydrogenation temperature. [42] Developing of new hydrides has been a very active research topic. Nanostructures can be used to improve the hydrogen storage properties of hydrides. They can change the thermodynamic properties of hydride which define the theoretical
Nanostructured Materials for Hydrogen Storage
499
working parameters, i.e. the pressure and temperature that hydrogen can be absorbed and desorbed. They can also change the kinetic properties that determine the rate of hydrogen release. This section discusses the recent finding and developments in this field.
Fig. 8. Hydrogen storage density in compressed gas, liquid, adsorbed monolayer (physisorbed), and selected chemical compounds, as a function of the hydrogen mass fraction. [A. Züttel, Mater. Today 6, 24 (2003)].
3.1. Metal and Complex Hydrides Metal hydrides are solid alloys which are typically composed of metal atoms with a host lattice and hydrogen atoms that are trapped in the interstitial sites forming a single-phase compound between a metal host and hydrogen. Binary hydrides can essentially be classified into three categories depending on the nature of the bonding between hydrogen and the metal host. Ionic or saline hydrides (e.g. MgH2, NaH, and CaH2) are formed by alkali and alkaline earth atoms and exhibit ionic bonding between the hydrogen and metal atoms. Covalent hydrides are formed by nonmetal elements like S, Si, C or B. Metallic hydrides
500
Sepehri and Cao
(e.g. LaNi5H6, PdH0.6, FeTiH2) originate from the metallic bonding between hydrogen and either a transition metal or a rare earth metal. In addition, group IA, IIA, and IIIA light metals form metal–hydrogen complexes [AlH4-], BH4-, which form covalent or ionic bonds with a cation, giving rise to highly stable complex hydrides (e.g. NaAlH4, Mg(AlH4)2). Hydrides have higher hydrogen storage density than hydrogen in gas or liquid form. For example the hydrogen density of MgH2 is 6.5 H atoms/cm3, while those of gas and liquid hydrogen are 0.99 H atoms/cm3 and 4.2 H atoms/cm3, respectively. [43] Some metal hydrides absorb and desorb hydrogen at near ambient temperature and pressure, and demonstrate very high hydrogen density. However, all the reversible hydrides working around ambient temperature and atmospheric pressure consist of heavy transition metals; therefore, even in nanocrystalline form, the gravimetric hydrogen density of metal hydrides is limited to less than 3 wt.%. [44] One way to improve the hydrogen capacity of metal hydrides is to use light weight materials such as magnesium. [42] Another challenge in hydrogen desorption from metal hydrides is their stability that demands elevated temperatures for release of hydrogen. Heat transfer is yet another challenge. In general, the formation of metal hydrides is an exothermic reaction. Efficient heat removal (for absorption) and heat addition (for desorption) has proven extremely difficult to achieve in metal hydride based systems. [45] Synthesis of new hydrides, particularly complex hydrides, has the potential to develop materials with superior hydrogen storage properties. Complex metal hydrides demonstrate higher gravimetric hydrogen capacities than simple metal hydrides. However, some of them show poor reversibility and once hydrogen is released they need high pressure to adsorb hydrogen again. [46] Moreover, due to the localization of the hydrogen and the slow diffusion rate of the metals in the solid, hydrogen sorption reactions are slow. Also, similar to metal hydrides, most complex hydrides suffer from high thermal stability. [47] Destabilization approaches can be used to improve the hydrogen storage properties of stable hydrides materials. [48] Thermodynamic destabilization of light-metal hydrides is achieved by using additives that reacts with metals to form new compounds (an intermediate state) during
Nanostructured Materials for Hydrogen Storage
501
dehydrogenation and lowers the enthalpy and hydrogen release temperature. Figure 9 shows the decreased dehydrogenation enthalpies for LiBH4 by adding destabilization agents. However, it should be mentioned that the dehydrogenation temperature for destabilized LiBH4 is still high (higher than 350°C) and the kinetics are slow. Also, adding any additive may increase the mass and reduce the hydrogen capacity, for example, the hydrogen capacity decreases from 13.6 wt.% in pure LiBH4 to 11.4 wt.% after adding MgH2, while the enthalpy is lowered by 25 kJ/molH2. [49]
Fig. 9. Enthalpy diagram for destabilization of LiBH4 shows the decreased dehydrogenation enthalpies after adding destabilization agent MgX (X = H2,F2, S, Se). [J. J. Vajo and G. L. Olson, Scripta Materialia 56, 829 (2007)].
Improving the reaction kinetics by decreasing the size of hydride may offer an interesting approach without increasing the mass. Increased surface area in nanosized hydrides can augment their surface energy and reduce the dehydrogenation enthalpy drastically. The increased surface area of hydrides facilitates the dissociation of hydrogen atoms by offering a larger number of dissociation sites and allowing fast gaseous diffusion. Different methods (including laser ablation, vapor condensation, sputtering, and ball milling) can be used to reduce the size of metal hydride particles. The most common method is ball milling in which metal hydride is placed in an inert grinding media (such as balls)
502
Sepehri and Cao
inside a rotating cylinder or conical mill. Reducing the particle and crystallite size is shown to enhance the hydrogen sorption kinetics in aluminum and magnesium based hydrides. [50, 51] Figure 10 shows improved hydrogen density and absorption kinetics for MgH2 ananoparticles doped with 5 wt.% TiF3 at different temperatures as compared to micrometer MgH2 particles doped with TiF3. [52, 30]
Fig. 10. Increased hydrogen absorption for nanoparticle MgH2 as compared to micropaticles MgH2 shows improved hydrogen absorption and kinetics. [U. Sahaym, and M. G. Norton, J. Mater. Sci. 43, 5395 (2008)].
In conclusion, although the metal and complex hydrides are considered to be potential candidates for hydrogen storage, significant fundamental research should be performed to obtain hydrides with practical hydrogen storage properties. Dehydrogenation temperatures should be decreased and kinetics of reaction should be improved. Other issues that need to be addressed include thermal management, reversibility and durability. [53] 3.2. Chemical Hydrides Chemical hydrides store hydrogen as M-H bonds where M is a light main group element such as C, B, N, or O. [54] They can release
Nanostructured Materials for Hydrogen Storage
503
hydrogen through a chemical reaction which is typically not easily reversible. Sometimes, metal and complex hydrides are also categorized under chemical hydrides, but those classes often refer to reversible dehydrogenation. Common reactions to release hydrogen from chemical hydrides involve the reaction with water (hydrolysis) or alcohols (alcoholysis), and thermal decomposition (pyrolysis). In all these methods, several issues such as controllability of reaction and regeneration energy should be considered A number of chemical hydrides, with both exothermic and endothermic dehydrogenation through different reaction are currently under investigation. Moreover, new chemical hydrides with high hydrogen densities can offer promising approaches for hydrogen storage. One of the early chemical hydrides studied, Ammonia (NH3) has been used in the fuel cells and power plants for more than 40 years. [55, 56] Anhydrous ammonia has high gravimetric hydrogen density of 17.5 wt.% and the byproduct of the hydrogen dissociation process is nitrogen that have no adverse environmental effects. However, decomposition (cracking) of ammonia is an endothermic reaction that happens efficiently at temperatures higher than 500°C, with an enthalpy of +46 kJ/mol. Therefore, it takes energy to gain hydrogen from ammonia. There are also safety and toxicity issues such as propensity for reacting with water, reactivity with container materials, and high toxicity of the vapor if released into the air. These drawbacks should be addressed before using ammonia as a hydrogen storage material, however, because of high hydrogen density and well-established technology, ammonia is being considered as a means for delivering hydrogen. Several boron hydrides have high hydrogen content. Ammoniaborane (AB), also known as borazane or by formula NH3BH3, has been of great interest as a hydrogen storage material. At ambient temperature and pressure, AB is a stable, white, crystalline solid (orthorhombic at lower temperatures and tetragonal more than -50°C), with low molecular weight (30.8 g/mol) with high gravimetric hydrogen capacity (19.6 wt.%). [57] There have been several experimental studies on the multi-step thermal decomposition of AB. [58-60] It was found that AB releases one
504
Sepehri and Cao
mole of hydrogen (per mole of AB) around 110°C and a second mole of hydrogen at 150°C. Different methods, including milling and catalysts have been used to improve the kinetics of AB dehydrogenation reactions and lower its decomposition temperature. [61] These findings are encouraging, however, future research should address several issues including reducing the dehydrogenation temperature, minimizing the formation of volatile products and developing economically viable methods for regeneration of AB. [62] 3.3. Nanocomposites Nanocomposites refer to materials consisting of at least two phases with one dispersed in another that is called matrix and forms a threedimensional network. [63] At the nano-scale, materials can show distinctly different properties than those of their bulk analogs. New fabrication techniques have offered new opportunities to design materials with specific structure to achieve desired properties. In hydrogen storage studies, nanocomposites have observed to significantly improve the thermodynamics and kinetics of hydrogen sorption by providing high surface area and hybrid structures that offers multifunctional performance. Decreasing particle size increases surface/volume ratio, resulting in enhanced surface energies, and alters the hydrogen release mechanisms. Using nanoporous scaffold as structure-directing agents to host hydrides can facilitate the formation of nano-size hydrides within the scaffold while reduces the hydrogen diffusion distances. It has been shown that infusing ammonia-borane (AB) in nanoporous silica scaffold (SBA-15, pore volume 1.2 cc/g, surface area 900 m2/g and pore size 7.5 nm), decreases the activation barrier for the hydrogen release, significantly improves the dehydrogenation kinetics, lowers the dehydrogenation temperature, and suppresses unwanted volatile products such as borazine (Fig. 11). [64]
Nanostructured Materials for Hydrogen Storage
505
Fig. 11. Mass spectrometry (at 1°C/min) of volatile products generated by heating neat AB and AB infused in porous silica (AB: SBA-15) for m/e=2 (H2, top) and m/e=80 (borazine, c-(NHBH)3, bottom) shows lower dehydrogenation temperature and suppression of borazine (unwanted byproduct) for AB inside SBA-15 compared to neat AB. [Gutowska, L. Li, Y. Shin, C.M. Wang, X.S. Li, J.C. Linehan, R.S. Smith, B.D. Kay, B. Schmid, W. Shaw, M. Gutowski, and T. Autrey, Angew. Chem. Int. Ed. 44, 3578 (2005)].
In other studies, the improved hydrogen sorption kinetics of nanosized NaAlH4 supported on surface-oxidized carbon nanofiber was observed (Fig. 12), which can be attributed to the minimized solid-state diffusion path length during hydrogen sorption in nano-NaAlH4. [53]
Fig. 12. The improved hydrogen desorption for sodium alanate (NaAlH4) supported on carbon nanofibers. [C. P. Baldé, B. P. C. Hereijgers, J. H. Bitter, and K. P. de Jong, J. Am. Chem. Soc. 130, 6761(2008)].
Using scaffolds with high pore volume and low weight can minimize the gravimetric and volumetric penalties associated with nanocomposites.
506
Sepehri and Cao
Carbon aerogels (CAs) and carbon cryogels (CCs) are porous carbons with tunable densities and pore sizes can provide high pore volume and surface area to serve this purpose. CAs and CCs can be prepared from hydrogels generated by sol-gel polycondensation of organic monomers such as resorcinol and formaldehyde in aqueous solution in the presence of a polymerization catalyst following by different drying methods and pyrolysis. [65] Aerogels are supercritically dried while cryogels are made by freeze drying. Because of the low density and high porosity, CAs and CCs can accommodate a large fraction of hydrides with little addition of weight. Moreover, the extremely high surface area facilitates an intimate contact between hydrides and the carbon network. Incorporation of LiBH4 into CAs with has shown to enhance the dehydrogenation kinetics and lower the dehydrogenating temperature of LiBH4. Figure 13 shows the thermo-gravimetric analysis for hydrogen release from LiBH4 confined in aerogels with pore sizes of 13 and 26 nm, microporous activated carbon, and a non porous graphite control sample. This study shows that incorporation of LiBH4 into the CA accelerates the dehydrogenation, reduces the energy barrier to release of hydrogen, and decreases the hydrogen release temperature, with lower dehydrogenation temperature observed for CA with smaller pore size. [66]
Fig. 13. Thermogravimetric analysis of LiBH4 dehydrogenation, shows that the reaction temperatures decrease with decreasing scaffold pore size. [A. F. Gross, J. J. Vajo, S. L. Van Atta, and G. L. Olson, J. Phys. Chem. C112, 5651, (2008).]
Nanostructured Materials for Hydrogen Storage
507
It has been shown that infusing AB inside CC, making CC-AB nanocomposites, results in the releasing of more than 5 wt.% H2 (including material) in one exothermic event at decreased dehydrogenation temperatures while suppressing undesirable volatile products such as borazine. Schematic of the CC-AB, consisted of nanocrystallite AB confined in the porous CC matrix, is shown in Fig. 14 (left). Figure 14 (right) shows the changes in the pore size distribution (PSD) of a CC with average peak pore size of 7 nm when loaded with AB (CC-AB), and after dehydrogenation through thermal reaction (reacted). Loading CCs with AB fills some of the mesopores resulting in an appreciable reduction in pore volume and a shift in PSD toward smaller pore sizes, which can be indicative of uniform loading of AB throughout the CCs. The dehydrogenation partially empties some of the filled pores and increases pore volume but porous structure of the CC is maintained after thermal reaction and PSD does not show appreciable shift. [67, 68]
Fig. 14. (left) Schematic of CC-AB nanocomposites showing the nanocrystallite AB dispersed and confined in the CC network, (right) pore size distribution of the CC sampel with peak pore size of 7 nm, and of CC-AB nanocomposite before and after dehydrogenation (reacted) showing the decrease in the pore volume and size after loading CC with AB and the increased pore volume after dehydrogenation while porous structure of CC is sustained. [S. Sepehri, B. B. Garíca, G. Z. Cao, and G. Z. Cao, J. Mater. Chem. 18, 4034 (2008)].
Figure 15 depicts cross sectional scanning electron microscopy (SEM) micrographs of CC before and after loading with AB and also after dehydrogenation. CC composed of interconnected porous network (Fig. 15A), loading AB in the CC results in the distribution of AB within
508
Sepehri and Cao
the interconnected CC network and coating the CC scaffold (Fig. 15B). After the dehydrogenation the porous structure of CC is sustained and fiber-like residue of AB is scattered throughout the CC (Fig. 15C). [68]
Fig. 15. (A) Interconnected porous structure of CC, (B) loading CC with AB coats the CC pores with AB in CC-AB, and (C) after thermal reaction of CC-AB the residual AB is scattered throughout CC (scale bar 100 nm). [S. Sepehri, B. B. Garíca, G. Z. Cao, and G. Z. Cao, J. Mater. Chem. 18, 4034 (2008)].
Mass spectrometry analysis shows the release of hydrogen in one step, lower dehydrogenation temperature, and suppression of borazine in CC-AB (Fig. 16). [67]
Fig. 16. Mass spectrometry results showing: (A) release of H2 release at 90°C in a CC-AB and at 110 and 150°C in neat AB; (B) the suppression of borazine in the CC-AB nanocomposite (heating rate 1°C/min). [A. M. Feaver, S. Sepehri, P. Shamberger, A. Stowe, T. Autrey, and G. Z. Cao, J. Phys. Chem. B 111, 7469 (2007)].
Nanostructured Materials for Hydrogen Storage
509
Studying the effect of pore size on the dehydrogenation temperature of CC-AB nanocomposites, using diffraction scanning calorimetry (DSC), shows that dehydrogenation temperature are lower in CC-ABs of different pore sizes as compared to that of neat AB, also, the dehydrogenation temperatures of CC-ABs decreases with reducing the pore size of the CCs (Fig. 17). [68] Inside the the CC matrix, nanocrystallite AB possesses a huge surface to volume ratio and a significantly larger surface energy which destabilizes the hydrogen bonding network of AB and lowers the barrier to hydrogen release.
Fig. 17. Dehydrogenation temperatures decreases with reducing pore size in CC-ABs (left) DSC exotherms for CC-AB nanocomposites (peak pore sizes of 7, 9, and 16 nm) and neat AB at similar heating rate (5°C/min), shows dehydrogenation temperatures in CC-ABs are lower than that of neat AB and reducing with pore size, (right) DSC dehydrogenation peaks for CC-ABs vs. their pore sizes, the dehydrogenation temperature of neat AB is shown with the dashed line, solid line connects the data points for visual guidance. [S. Sepehri, B. B. Garíca, G. Z. Cao, and G. Z. Cao, J. Mater. Chem. 18, 4034 (2008)].
Moreover, chemical modification of CCs (such as dispersion of B and N in CCs) has demonstrated appreciable impacts on the porous structure and surface chemistry of CCs, as well as, dehydrogenation temperatures of modified CC-AB nanocomposites. Catalytic effects of chemical modification of CCs can further promote the destabilization of AB, and thus lowers its activation energy, as compared to unmodified CC scaffold of the same pore size (Fig. 18). [70] These studies provides
510
Sepehri and Cao
further experimental evidence that the dehydrogenation properties of hydrides can be readily controlled by the porous structure and surface chemistry of nanoscaffold.
Fig. 18. DSC exotherms (heating rate 5°C/min) for CC-AB and boron-nitrogen modifiedCC-AB (BNCC-AB) nanocomposites of the same pore size (16 nm) and neat AB, shows reduced dehydrogenation temperature for nanocomposites while dehydrogenation temperature is lower in BN-CC-AB as compared to CC-AB indicating the effect of chemical modification on dehydrogenation. [S. Sepehri, B. B. García, and G. Z. Cao, accepted by Eur. J. Inorg. Chem.].
Using nanocomposites for hydrogen storage opens up the possibility of designing functional systems, where an external matrix could act as multi-functional destabilization system for hydride. It can decrease the hydride size, and increase surface energy, increase the heat transfer, and chemically catalyzing the dehydrogenation to achieve desirable thermodynamic and kinetic properties of the hydride. Fabrication of a light and thermally conductive nanoporous material (such as porous carbon) provides a compelling approach. Porous carbon–hydride nanocomposites have shown impressive results in enhancing the dehydrogenation kinetics by reducing the diffusion distance, increase the reaction surface area, and destabilizing the hydride. Incorporation of hydrides into functional frameworks or matrices with desired chemical and physical properties should be investigated more intensively. The tunable structure of nanoporous carbon offers different ways to catalyze
Nanostructured Materials for Hydrogen Storage
511
the dehydrogenation reaction. Increasing porosity of the nanoscaffold to attain large accessible pore volume will improve the storage capacity. Thermal properties of the nanoscaffold play an important role in the reaction kinetics. Modifying structure of carbon scaffold by adding other elements can increase the heat transfer while catalyzing the dehydrogenation furthermore. These approaches are expected to generate new ideas that can lead to fundamental innovations in hydrogen storage technology. 4. Summary Effective storage of hydrogen is a major challenge in implementing hydrogen economy. Storing hydrogen in solids is a promising and appealing approach but there is still no material or method that meets the requirements for an ideal storage system. Recent developments in the nanotechnological approach and nanostructured materials can open new doors to more development by addressing hydrogen storage challenges including: hydrogen storage capacities, hydrogen sorption near ambient conditions, and reasonable thermodynamics and kinetics properties. Nanostructured materials with their unique properties offer new opportunities to address the challenges involve in physicsorption and chemisorptions storage of hydrogen, Nanostructured materials can improve the interaction between the hydrogen and surface of the storage medium because of their large surface area and enhance the hydrogen physisorption. Enhanced hydrogen storage properties are observed for nanosized and nanocomposite hydrides which are related to size and catalytic effects. Optimizing these effects can improve the storage density. Acknowledgments This work has been supported in part by National Science Foundation (DMI-0455994 and DMR-0605159), and Air Force Office of Scientific Research (AFOSR-MURI, FA9550-06-1-032). This work has also been supported by Washington Technology Center and EnerG2.
512
Sepehri and Cao
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.
G. W. Crabtree, M. S. Dresselhaus, and M. V. Buchanan, Phys. Today 57, 39 (2004). S. Satyapal, J. Petrovic, C. Read, G. Thomas, and G. Ordaz, Catal. Today 120, 246 (2007). V. Bérubé, G. Radtke, M. Dresselhaus, and G. Chen, Int. J. Energy Res. 31, 637 (2007). M. Fichtner, Adv. Eng. Mater. 7, 443 (2005). A. W. C. van den Berg and C. O. Areán, Chem. Commun 668 (2008). K. M. Thomas, Catal. Today 120, 389 (2007). M. S. Dresselhaus, G. Dresselhaus, and P. C. Elkund, Science of Fullerenes and Carbon Nanotubes, Academic Press, Boston (1996). M. S. Mauter and M. Elimelech, Environ. Sci. Technol. 42, 5843 (2008). A. Chambers, C. Park, R. T. K. Baker, and N. M. Rodriguez, J. Phys. Chem. B102, 4253 (1998). A. C. Dillon, K. M. Jones, T. A. Bekkedahl, C. H. Kiange, D. S. Bethune, and M. J. Heben, Nature 386, 377 (1997). Q. Wang and J. K. Johnson, J. Phys. Chem. B103, 277 (1999). A. Züttel, Naturwissenschaften 91, 157 (2004). P. Kowalczyk, R. Holyst, M. Terrones, and H. Terrones, Phys. Chem. Chem. Phys. 9, 1786 (2007). R. Ströbel, J. Garche, P. T. Moseley, L. Jörisen, and G. Wolf, J. Power Sources 159, 781 (2006). H. G. Schimmel, G. J. Kearley, M. G. Nijkamp, C. Visser, K. de Jong, and F. M. Mudler, Chem. Eur. J. 9, 4764 (2003). L. Zhou, Y. P. Zhou, and Y. Sun, Int. J. Hydrogen Energy 29, 319 (2004). A. Zuttel, P. Sudan, Ph. Mauron, T. Kiyobayashi, Ch. Emmenegger, and L. Schlapbach, Int. J. Hydrogen Energy 27, 203 (2002). B. Panella, M. Hirscher, and B. Ludescher, Microporous Mesoporous Mater. 103, 230 (2007). Z. Yang, Y. Xia, and R. Mokaya, J. Am. Chem. Soc. 129, 1673 (2007). H. Kabbour, T. F. Baumann, J. H. Satcher, Jr., A. Saulnier, and C. C. Ahn, Chem. Mater. 18, 6085 (2006). T. C. M chaung, Y. Jeong, Q. Chen, A. Kleinhammers, and Y. Wu, J. Am. Chem. Soc. 130, 6668 (2008). J. Weitkamp, M. Fritz, and S. Ernst, Int. J. Hydrogen Energy, 20, 967 (1995). M. K. Song and K. T. No, Catal. Today 120, 374 (2007). H. W. Langmi, A. Walton, M. M. AL-Mamouri, S. R. Johnson, D. Book, J. D. Speight, P. P. Edwards, I. Gameson, P. A. Anderson, and I. R. Harris, J. Alloys Compd. 356, 710 (2003).
Nanostructured Materials for Hydrogen Storage 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.
37. 38.
39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50.
513
J. G. Vitillo, G. Ricchiardi, G. Spoto, and A. Zecchina, Phys. Chem. Chem. Phys. 7, 3948 (2005). H. Park, J. F. Britten, U. Mueller, J. Y. Lee, J. Li, and J. B. Parise, Chem. Mater. 19, 1302 (2007). J. L. C. Roswell, A. R. Millward, K. S. Park, and O. M. Yaghi, J. Am. Chem. Soc. 126, 5666 (2004). S. Ma, D. Sun, M. W. Ambrogio, J. A. Fillinger, S. Parkin, and H. C. Zhou, J. Am. Chem. Soc. 129, 1858 (2007). A. G. Wong-Foy, A. J. Matzger, and O. M. Yaghi, J. Am. Chem. Soc. 128, 3494 (2006). U. Sahaym, and M. G. Norton, J. Mater. Sci. 43, 5395 (2008). D. J. Collins and H. C. Zhou, J. Mater. Chem. 17, 3154 (2007). M. Dinca, A. F. Yu, and J. R. Long, J. Am. Chem. Soc. 128, 8904 (2006). E. D. Sloan, Clathrate Hydrates of the Natural Gases, Dekker, New York (1997). Y. A. Dyadin, E. G. Larionov, E. Y. Aladko, A. Y. Manakov, F. V. Zhurko, T. V. Mikina, V. Y. Komarov, and E. V. Grachev, J. Struct. Chem. 40, 790 (1999). Y. H. Hu, and E. Ruckenstein, Angew. Chem., Int. Ed. 45, 2011 (2006). W. L. Mao, H. K. Mao, A. F. Goncharov, V. V. Struzhkin, Q. Z. Guo, J. Z. Hu, J. F. Shu, R. J. Hemley, M. Somayazulu, and Y. S. Zhao, Science 297, 2247 (2002). D. Y. Kim, Y. Park, and H. Lee, Catal. Today 120, 257 (2007). N. B. McKeown, B. Ganhem, K. J. Msayib, P. M. Budd, C. E. Tattershall, K. Mahmood, S. Tan, D. Book, H. W. Langmi, and A. Walton, Angew. Chem. Int. Ed. 45, 1804 (2006). P. M. Budd, A. Butler, J. Slbie, K. Mahmood, N. B. McKeown, G. Ghanem, K. Msayib, D. Book, and A. Walton, Phys. Chem. Chem. Phys. 9, 1802 (2007). G. S. Ghanem, K. J. Msayib, N. B. McKewon, K. D. M. Harris, Z. Pan, P. M. Budd, A. Butler, J. Selbie, D. Book, and A. Walton, Chem. Commun. 67 (2007). A. Züttel, Mater. Today 6, 24 (2003). A. M. Seayad, D. M. Antonelli, Adv. Mater. 16, 765 (2004). B. Sakintuna, F. Lamari-Darkrim, and M. Hirscher, Int. J. Hydrogen Energy 32, 1121 (2007). L. Li and J. A. Hurley, Int. J.Hydrogen Energy 32, 6 (2007). V. Bérubé, G. Radtke, M. Dresselhaus, and G. Chen, Int. J. Energy Res. 31, 637 (2007). Y. Nakamori and S. J. Orimo, J. Alloys Compd. 370, 271(2004). A. Züttel, P. Wenger, P. Sudan, P. Mauron, and S. Ormio, Mater. Sci. Eng. B108, 9 (2004). W. Grochala and P. P. Edwards. Chem. Rev. 104, 1283 (2004). J. J. Vajo and G. L. Olson. Scripta Materialia 56, 829 (2007). M. Zhu, H. Wang, L. A. Ouyang, and M. Q. Zeng, Int. J. Hydrogen Energy 31, 251 (2006).
514
Sepehri and Cao
51.
M. Dorenheim, N. Eigen, G. Barkhordarian, T. Klassen, and R. Bormann, Adv. Eng. Mater. 8, 377 (2006). L. Xie, Y. Liu, Y. T. Wang, J. Zheng, and X. G. Li, Acta Mater. 55, 4585 (2007). C. P. Baldé, B. P. C. Hereijgers, J. H. Bitter, and K. P. de Jong, J. Am. Chem. Soc. 130, 6761 (2008). E. Fakioğlu, Y. Yurum, and T. N. Veziroğlu, Int. J. Hydrogen Energy 29, 1371 (2004). E. L. Simons, E. J. Cairns, and D. J. Surd, J. Electrochem. Soc. 116, 556 (1969). G. Strickland, Int. J. Hydrogen Energy 9, 759 (1984). F. P. Hoffmann, G. Wolf, and L. D. Hansen, R. Soc. Chem. Cambridge 514 (1997). C. R. Miranda and G. Ceder, J. Chem. Phys. 126, 184703 (2007). J. Baumann, F. Baitalow, and G. Wolf, Thermochim. Acta. 430, 9 (2005). A. Stowe, W. Shaw, J. Linehan, B. Schmid, and T. Autrey, Phys. Chem. Chem. Phys. 9, 1831 (2007). S. D. Benedetto, M. Carewska, C. Cento, P. Gislon, M. Pasquali, S. Scaccia, and P. P. Prosini, Thermochim. Acta 441, 184 (2006). T. B. Marder, Angew. Chem. Int. Ed. 46, 8116 (2007). G. Z. Cao, Nanostructures and Nanomaterials, Synthesis, Properties and Applications, Imperial College Press, London (2004). A. Gutowska, L. Li, Y. Shin, C. M. Wang, X. S. Li, J. C. Linehan, R. S. Smith, B. D. Kay, B. Schmid, W. Shaw, M. Gutowski, and T. Autrey, Angew. Chem. Int. Ed. 44, 3578 (2005). S. A. Al-Muhtaseb and J. A. Ritter, Adv. Mater. 15, 101 (2003). A. F. Gross, J. J. Vajo, S. L. Van Atta, and G. L. Olson, J. Phys. Chem. C112, 5651 (2008). A. M. Feaver, S. Sepehri, P. Shamberger, A. Stowe, T. Autrey, and G. Z. Cao, J. Phys. Chem. B 111, 7469 (2007). S. Sepehri, B. B. Garíca, G. Z. Cao, and G. Z. Cao, J. Mater. Chem. 18, 4034 (2008). S. Sepehri, A. M. Feaver, W. J. Shaw, C. J. Howard, Q. Zhang, T. Autrey, and G. Z. Cao, J. Phys. Chem. B 111, 14285 (2007). S. Sepehri, B. B. García, and G. Z. Cao, accepted by Eur. J. Inorg. Chem.
52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64.
65. 66. 67. 68. 69. 70.
CHAPTER 12 RECENT ADVANCES IN THE CHARACTERIZATION OF MESOPOROUS MATERIALS BY PHYSICAL ADSORPTION
Matthias Thommes Quantachrome Instruments, 1900 Corporate Drive, Boynton Beach Fl-33426, USA Email:
[email protected]
In recent years, major progress has been achived in the understanding of the adsorption and phase behavior of fluids in highly ordered mesoporous materials with simple pore geometries (e.g., M41S materials). This has led to major advances in the structural characterization by physical adsorption, also because of the development and availability of advanced theoretical approaches based on statistical mechanics (e.g., non local density functional theory (NLDFT) and molecular simulation). However, there are still many open questions concerning the structural characterization of more complex porous systems. Within this context, this chapter provides an overview of the major underlying mechanisms associated with the adsorption, pore condensation and hysteresis behavior of fluids in ordered mesoporous systems with simple geometries, ordered pore networks, and novel micro-mesoporous materials with hierarchical pore structure. Fluids adsorbed in hierarchically structured micromesoporous materials can exhibit very complex, but very interesting pore condensation and hysteresis behavior. A combination of phenomena such as delayed pore condensation, advanced pore condensation, pore blocking/percolation and cavitation induced evaporation can be observed, which is reflected in characteristic types of adsorption hysteresis. These complex hysteresis loops introduce of course a considerable complication for pore size analysis, but if interpreted correctly, provide important information about the pore structure/network, which is crucial for obtaining a comprehensive and accurate textural analysis of advanced micro-mesoporous materials. 515
516
Thommes
1. Introduction In recent years, major progress has been made concerning the synthesis of highly ordered mesoporous materials with tailored pore size and structure, controlled surface functionality and their applications [1-4]. Advances have also been made in the synthesis and structural characterization of micro-mesoporous materials such as mesoporous zeolites (see [5, 6] for reviews) and hierarchically ordered pore structures [e.g. 7, 8]. Typical microporous materials (e.g. zeolites, activated carbons), consist of very narrow pores (i.e. pore width smaller than 1.5 nm) leading to a small effective diffusivity which limits the reaction rate, and hampers therefore the application of these material, hence, the targeted introduction of mesoporosity is desirable. Further, a comprehensive characterization of these porous materials with regard to pore size, surface area, porosity and pore size distribution is required in order to select and optimize the performance of mesoporous and micro-mesoporous materials in many industrial applications (e.g., in heterogeneous catalysis, separation, drug delivery, gas storage). Major advancements in materials characterization and practical utilization, have been governed by the progress made in the development of various experimental techniques, such as gas adsorption, X-ray diffraction (XRD), small angle x-ray and neutron scattering (SAXS and SANS), mercury porosimetry, electron microscopy (scanning and transmission), thermoporometry, NMR-methods, and others. Each method has a limited length scale of applicability for pore size analysis, and an overview of different methods for pore size characterization and their application range was given by IUPAC [9]. In order to explore the nature of the adsorption and phase behavior of fluids confined to more complex porous systems (e.g. micromesoporous zeolites, hierarchically structured porous materials), it is necessary to combine various experimental methods (e.g. coupling adsorption experiments with SAXS and SANS i.e. in-situ scattering, and high resolution transmission electronic microscopy) [10-231]. However, among all these methods, gas adsorption (i.e. here physical adsorption) is still the most popular one because it allows assessing a wide range of pore sizes (from 0.35 nm up to 100 nm) including the complete range of
Characterization of Mesoporous Materials
517
micro- (pore width <2 nm), meso- (pore width: 2-50 nm) and to some extend even macropores (pore width >50 nm)] [24]. Furthermore, gas adsorption techniques are convenient to use and are less cost-intensive than some of the other methods. In order to obtain surface area, pore size, pore size distribution, pore volume, porosity, and other information from the analysis of gas adsorption isotherms, one needs to apply proper theoretical models, which capture the important underlying adsorption mechanisms. The theoretical background of physical adsorption and its significance for textural characterization has recently been reviewed and described in various books [e.g., 25-31]. In particular during the last decade, significant progress has been achieved in the understanding of the underlying mechanisms of adsorption in highly ordered mesoporous materials with simple pore geometries and consequently, in elaborating the theoretical foundations of adsorption characterization. These advances are to a large extent related to (i) the discovery of novel ordered mesoporous materials, such as MCM-41, MCM-48, SBA-15, which exhibit a uniform pore size with periodically ordered structure and can therefore be used as model adsorbents to test theories of gas adsorption; (ii) carefully performed adsorption experiments coupled with complimentary experimental techniques (iii) the development and application of microscopic approaches such as the nonlocal density functional theory (NLDFT) of inhomogeneous fluids (and methods based on molecular simulation), which allows one to describe adsorption and phase behavior of fluids in pores on a molecular level [32-36]. These modern methods (which are based on statistical mechanics) describe the configuration of the adsorbed fluid (i.e. the adsorbate) on a molecular level, in contrast to classical methods which are based on macroscopic, thermodynamic assumptions, such as Kelvin equation based approaches. Direct experimental tests of the validity of the Kelvin equation were made possible by using for instance MCM-41 silica as a model material, which consists of an array of independent cylindrical pores (of the same diameter in the range 2 nm to 10 nm). Because of the high degree of order, the pore diameter can be derived by independent methods (based on X-ray diffraction, high-resolution transmission electron microscopy etc.). It has been found that methods based on the modified Kelvin
518
Thommes
equation such as the Barett-Joyner-Halenda (BJH) methods can underestimate the pore size by up to 20-30% for pores smaller than 10 nm, if not properly corrected [30, 31] (review see [28] and references therein]. 0.3
560
490
N2 (77 K): ads N2 (77 K): des
0.25
BJH-Pore size distribution DFT-Pore size distribution
420
Dv(d) [cc/Å/g]
Volume [10-6 m3/g]
DFT-Fitting
350
280
0.15
0.1
0.05
210
140 0
0.2
0.2
0.4
0.6
0.8
0 15
1
23
31
RELATIVEPRESSUREp/p0
39
47
55
Pore Diameter [Å]
(a)
(b)
Figure 1. (a) Nitrogen sorption isotherm (at 77 K) in MCM-41 silica and NLDFT fit. Figure; (b) BJH and NLDFT pore size distribution curves obtained from the sorption isotherm shown in Fig. 1a. [26]
0.036
360
0.03
240
D(v) [cm3/Å/g]
Volume [cm3/g] STP
300
Ads (Run1) Des (Run1) Ads (Run2) Des (Run2)
180
120
0.018 0.012 0.006
60
0 0
0.024
0.2
0.4
0.6
P/P0
(a)
0.8
1
0 0
60
120
180 240 Pore Diameter[Å]
300
360
(b)
Figure 2. (a) High-resolution nitrogen adsorption (77.4 K) into a hierarchically ordered micro/mesoporous silica (KLE/IL silica), and (b) NLDFT pore size distribution for KLE/IL silica calculated from the adsorption branch by applying a novel NLDFT method. [23]
Characterization of Mesoporous Materials
519
This is demonstrated in Fig. 1, which compares the BJH pore size distribution curves (PSD) calculated from a N2(77 K) isotherm obtained on a MCM-41 silica with the NLDFT pore size distribution. An important advantage of the NLDFT method is, that it allows to perform pore size analysis over the complete micro-mesopore size range. This is most important for the textural characterization of hierarchically ordered porous structures and mesoporous zeolites in which many adsorptives exhibit very interesting pore condensation and hysteresis behavior (for a detailed discussion see Sec. 3). Figure 2a shows a nitrogen (at 77.4 K) adsorption isotherm obtained on a hierarchically structured porous silica materials. The observed very wide hysteresis loop is caused by delayed condensation (due to metastable pore fluid) as well as by cavitation induced evaporation which affects the position of the desorption branch (see Sec. 3). The combined micro-mesopore size distribution was calculated from the adsorption branch by applying a novel NLDFT method which assumes (in agreement with the structure of the material) a cylindrical pore model for the micropore range and a spherical pore model for the mesopore range in which hysteresis is observed. This hybrid NLDFT kernel takes also correctly into account that pore condensation in the mesopores occurs delayed due to metastable pore fluid. These pore size data have been compared with data from independent SANS and TEM measurements and very good agreement had been found [19]). Methods for pore size analysis based on NLDFT and molecular simulation are widely applied, featured in an ISO standard [38] and commercially available for many important adsorptive/adsorbent systems including hybrid methods which assume different pore geometries for the micro- and mesopore size range, as it can be found for materials with hierarchical pore structures. A drawback of standard NLDFT and GCMC methods is that they do not take sufficiently into account the chemical and geometrical heterogeneity of the pore walls, i.e. usually a structureless (i.e. chemically and geometrically smooth) pore wall model is assumed. While NLDFT has been demonstrated to be a reliable method for characterization of many ordered and hierarchically structured mesoporous silica materials [e.g. 30, 37], the pore size analysis of
520
Thommes
materials where the pore walls exhibit pronounced roughness and chemical inhomogeneity (e.g. as pronounced in porous carbon materials with broad pore size distribution curves) is more problematic. These deficiencies, are currently being addressed by various scientific groups [39, 40], and very recently a novel DFT method, namely QSDFT (quenched solid density functional theory has been put forward which quantitatively accounts for the surface geometrical inhomogeneity in form of a roughness parameter. It has been demonstrated that this novel method significantly improves the pore size distribution of micromesoporous carbon materials [41] and QSDFT is currently also being employed to assess the heterogeneity and roughness in ordered mesoporous silica materials [21]. This chapter should be considered an update of the author’s review article from 2004 on the same topic [30]. Hence, for an extensive indepth description of the theoretical background of the adsorption and phase behavior of fluids in mesoporous materials (and consequences for pore size analysis), the reader is referred to this review [30]. In this chapter, we focus on a discussion and interpretation of adsorption phenomena (e.g., hysteresis) which are of importance for obtaining a comprehensive and accurate textural characterization of advanced mesoporous (and micro-mesoporous) materials. In the next section (i.e., Sec. 2) we describe briefly some general experimental and theoretical aspects concerning physical adsorption characterization of mesoporous materials. Section 3 provides a discussion on the theoretical background of pore condensation and hysteresis, but focuses mainly on some very recent advances in the understanding of hysteresis in more complex pore structures (e.g. hierarchically structured porous materials). The correct interpretation of hysteresis is crucial in order to obtain a reliabel pore size analysis. Currenty available methods applied for the interpretation of experimental adsorption data for pore size distribution calculations will be discussed in Sec. 4. Concluding remarks will be given in Sec. 5.
Characterization of Mesoporous Materials
521
2. General Aspects of Surface and Pore Size Analysis by Physisorption Physisorption (physical adsorption) occurs whenever a gas (the adsorptive) is brought into contact with the surface of a solid (the adsorbent). The matter in the adsorbed state is known as the adsorbate, as distinct from the adsorptive, which is the gas or vapor to be adsorbed. The forces involved in physisorption always include the long-range London dispersion forces and the short-range intermolecular repulsion. These combined forces give rise to nonspecific molecular interactions. Various types of specific interactions come into play when polar molecules are adsorbed on ionic or polar surfaces. However, as long as there is no formation of chemical bonding, the process is still regarded as physisorption. In order to obtain surface area, pore size, pore size distribution, pore volume, porosity, and other information from the analysis of gas adsorption isotherms, one needs to apply proper theoretical models that capture the important underlying adsorption mechanisms. This topic has been recently extensively reviewed in the literature and in appropriate textbooks [25, 27, 28, 30]. Physisorption in porous materials is governed by the interplay between the strength of fluid-wall and fluid-fluid interactions as well as the effect of restricted pore space (in narrow pores) on the state of confined fluids. This is reflected in the shape or type of the adsorption isotherm. Within this context the International Union of Pure and Applied Chemistry (IUPAC) has published a classification of six types of adsorption isotherms [24] and proposed to classify pores by their internal pore width (the pore width is defined as the pore diameter in case of a cylindrical pore and as the distance between opposite walls in case of a slit pore), i.e., Micropore: pore of internal width less than 2 nm; Mesopore: pore of internal width between 2 and 50 nm; Macropore: pore of internal width greater than 50 nm. The sorption behavior in macropores is distinct from that of mesopores and micropores. Whereas macropores are so wide that they can be considered as nearly flat, the adsorption behavior in micropores is dominated almost entirely by the interactions between fluid molecules and the pore walls; in fact the adsorption potentials of the opposite pore
522
Thommes
walls are overlapping; as a consequence micropores fill through a continuous process. In contrast, the sorption behavior in mesopores depends not only on the fluid-wall attraction, but also on the attractive lateral interactions between the fluid molecules. This leads to the occurrence of multilayer adsorption and capillary (pore) condensation. The adsorbed amount as a function of pressure can be obtained by volumetric (manometric) and gravimetric methods, carrier gas and calorimetric techniques, nuclear resonance as well as by a combination of calorimetric and impedance spectroscopic measurements (for an overview see Refs. [28, 31, 42-46]). However, the most frequently used techniques are the volumetric- (manometric) and the gravimetricmethods. The gravimetric method is based on a sensitive microbalance and a pressure gauge. The adsorbed amount (i.e. the surface excess) can be measured directly, but a pressure dependent buoyancy correction is necessary. The gravimetric method is convenient to use for the study of vapor adsorption not too far from room temperature, whereas the volumetric (manometric) has advantages for the measurement of nitrogen, argon and krypton adsorption at cryogenic temperatures (77.4 K and 87.3 K), which are mainly used for surface area and pore size characterization. The volumetric method is based on calibrated volumes and pressure measurements by applying the general gas equation. The adsorbed amount (i.e. the surface excess, which however at cryogenic temperature and very low gas densities corresponds to the total adsorbed amount) is calculated by determining the difference of the total amount of gas admitted to the sample cell with the adsorbent and the amount of gas in the free space (also referred to as the void volume). Generally, nitrogen adsorption at liquid nitrogen temperature (77.4 K) is used for surface and pore size characterization. Krypton adsorption at 77.4 K is more or less exclusively used for low surface area analysis of materials such as thin films [28, 47, 48]. If applied at 87.3 K, Krypton adsorption also allows to obtain the pore size distribution of thin mesoporous silica films with pore diameters ranging from below 1 nm up to 10 nm [47]. Despite the fact that at 87.3 K, krypton is ca. 30 K below the bulk triple point temperature, if this adsorbate is confined in silica mesopores with pore diameters <10 nm, it is still in a supercooled liquid state which has also been confirmed in a molecular simulation study [49].
Characterization of Mesoporous Materials
523
The use of argon adsorption at 87.3 K is advantageous for the pore size analysis of zeolites and other microporous materials because the filling of pores of dimension 0.5-1 nm occurs at much higher relative (and total) pressure as compared to nitrogen adsorption [28, 31, 46]. This phenomenon is related to the fact, that the nitrogen molecules possess a quadrupole moment, which leads to specific fluid-wall interactions. Contrary to argon adsorption at 87.3 K, argon adsorption at 77.4 K is less than optimal for both porosity and surface area analysis, because: (i) a combined and complete micro- and mesopore size analysis with argon is not possible at 77.4 K (which is ca. 6.5 K below the triple point temperature of bulk argon), i.e. the pore size analysis by argon adsorption at 77.4 K is limited to pore diameters smaller than ca. 16 nm (for details, please see [50, 51, 30]); (ii) There is some evidence that the structure of the argon monolayer (at 77.4 K) is highly dependent on the surface chemistry of the adsorbent [e.g., 25, 27]. The latter problem also affects the determination of the surface area by the BET method [52]. With regard to this nitrogen is usually considered the “standard adsorptive”, also because of the availability of liquid nitrogen. A key-issue for the BET analysis is the assumption of a cross-sectional area, i.e. the area occupied by an adsorbed molecule in a complete monolayer. Usually for nitrogen at its boiling point of 77.4 K. one assumes a cross-sectional area of 0.162 nm2. This cross-sectional area of 0.162 nm2 is based on the assumption that at 77.4 K the nitrogen monolayer is in a close-packed “liquid state”, which appears to be quite accurate for instance in the case of adsorption on carbon surfaces. On the other hand, the quadrupole moment of the N2 molecule can lead to specific interactions with polar surface groups (e.g, hydroxyl groups) which can result in an orientating effect of the adsorbed nitrogen molecules [53]. An accurate cross-sectional area value (i.e. 0.135 nm2), valid for nitrogen adsorbed on surfaces with polar groups (e.g, hydroxylated silica surfac), was obtained by measuring the volume of N2 adsorbed on silica spheres of known diameter [54]. Indeed, recent experimental sorption studies on highly ordered mesoporous silica materials such as MCM-41 suggest strongly that the cross-sectional area of nitrogen on a hydroxylated surface might differ from the commonly
524
Thommes
adopted value of 0.162 nm2 [55]. If one uses the standard cross-sectional area (0.162 nm2 the BET surface area of hydroxylated silica surfaces can be overestimated by ca. 20%). As already explained, in contrast to nitrogen, argon has no quadrupole moment and the above-mentioned problems do not occur when argon is used as the adsorptive. Hence, if the surface chemistry is not well know (i.e. this is the case for many oxidic surfaces) argon adsorption at 87.3 K is good alternative for an accurate surface area determination. 3. Pore Condensation and Hysteresis in Mesoporous Materials 3.1. Pore Condensation In case of complete wetting, the pore walls are covered by a multilayer adsorbed film at the onset of pore condensation. The stability of this film is determined by the attractive fluid-wall interactions, the surface tension and curvature of the liquid-vapor interface. Multilayer adsorption can for instance be described in the spirit of the Frenkel–Halsey–Hill theory [56-58]. One of the basic assumptions is that the (sufficiently thick) adsorbed multilayer film can be considered as a slab of liquid, which reveals the same properties (i.e., density etc.) as the bulk liquid would have at this temperature. The only modification to its free energy arises from the interaction with the solid, i.e., the adsorption forces (dispersion forces). For a planar surface it is expected that the thickness of the adsorbed film increases without limit by approaching P/P0 = 1. However, in pores, the film thickness cannot grow to unlimited thickness. The stability of the adsorbed multilayer film for instance in a cylindrical pore is determined by the long-range van der Waals interactions, and by the surface tension and curvature of the liquid-vapor interface. For small film thickness the adsorption potential dominates. However, when the adsorbed film becomes thicker, the adsorption potential becomes less important, whereas surface tension/curvature effects become significant. At a certain critical thickness tc, the multilayer film cannot be stabilized anymore, and pore
Characterization of Mesoporous Materials
525
condensation occurs in the core of the pore, controlled by intermolecular forces in the core fluid. Pore condensation represents a phenomenon whereby gas condenses to a liquid-like phase in pores at a pressure less than the saturation pressure P0 of the bulk fluid. It represents an example of a shifted bulk transition under the influence of the attractive fluid-wall interactions. For pores of uniform shape and width (ideal slit-like or cylindrical mesopores) pore condensation can be classically described on the basis of the Kelvin equation [59, 60], i.e., the shift of the gas-liquid phase transition of a confined fluid from bulk coexistence, is expressed in macroscopic quantities like the surface tension γ of the bulk fluid, the densities of the coexistent liquid ρl and gas ρg (∆ρ = ρl - ρg) and the contact angle θ of the liquid meniscus against the pore wall. For cylindrical pores the modified Kelvin equation [60] is given by: ln(P/P0) = -2γcosθ/RT∆ρ(rp – tc)
(1)
where R is the universal gas constant, rp the pore radius and tc the thickness of an adsorbed multilayer film, which is formed prior to pore condensation. The occurrence of pore condensation is expected as long as the contact angle is below 90°. A contact angle of 0° (i.e. complete wetting) is usually assumed in case of nitrogen and argon adsorption at 77.4 K and 87.3 K, respectively. The Kelvin equation provides a relationship between the pore diameter and the pore condensation pressure, and predicts that pore condensation shifts to a higher relative pressure with increasing pore diameter and temperature. Hence, the modified Kelvin equation (Eq. 1) serves as the basis for many methods applied for mesopore analysis, including the Barett–Joyner–Halenda method (BJH) [61], which is widely used. However, the validity of macroscopic, thermodynamic concepts such as the Kelvin equation and related methods becomes questionable for narrow mesopores (i.e. pore diameter smaller than ca. 15 nm). In order to account for the pre-adsorbed multilayer film, the Kelvin equation is combined with a standard isotherm or a so-called t-curve, which usually refers to adsorption measurements on a nonporous solid of a surface similar to that of the sample under consideration.
526
Thommes
Accordingly, the pre-adsorbed multilayer film is assessed by the statistical (mean) thickness of an adsorbed film on a nonporous solid of a surface similar to that of the sample under consideration. However, in particular for narrow pores of widths < ca.10 nm this mean thickness does not reflect the real thickness of the pre-adsorbed multilayer film, because curvature effects are not taken into account. Further, it is assumed that the pore fluid has essentially the same thermophysical properties as the correspondent bulk fluid. For instance, the surface tension of the pore-liquid is thought to be equal to the properties of the corresponding bulk liquid, but the surface tension of the pore-liquid depends on the radius of curvature, although deviations from the bulk surface tension are however expected to occur in narrow mesopores [25]. It is further evident that the Kelvin concept fails to describe correctly the peculiarities of the critical region and the confinement-induced shifts of the phase diagram (i.e. critical point shifts, freezing point and triple point shifts, etc) of the pore fluid. The thermodynamic state and the thermophysical properties of the adsorbed pore fluids differs significantly from the bulk fluid, and this has a pronounced effect on the shape of the adsorption isotherm; e.g., the disappearance of hysteresis with decreasing pore size (at given temperature), or increasing temperature (for a given pore size) cannot be described by the Kelvin equation [30, 32]. However, microscopic methods based on statistical mechanics which can describe the configuration of the adsorbed phase on a molecular level (DFT, molecular simulation) take this into account, i.e. these methods allow a more accurate description of hysteresis behavior in isolated pores and certain porous networks. 3.2. Interpretation of Adsorption Hysteresis 3.2.1. Classification of Hysteresis Loops Capillary condensation is very often accompanied by hysteresis, and it is widely accepted that there is a correlation between the shape of the hysteresis loop and the texture of the adsorbent. An empirical classification of hysteresis loops was given by IUPAC [22], which is shown in Fig. 3. According to the IUPAC classification, type H1 is often
Characterization of Mesoporous Materials
527
associated with porous materials exhibiting a narrow distribution of relatively uniform (cylindrical-like) pores. In principle, an H1 hysteresis loop is mainly determined by the effect of delayed condensation (due to metastable pore fluid) and not by network effects. The desorption branch reflects the liquid-vapor equilibrium transition, and can be used for calculation of the pore size distribution. Materials that give rise to H2 hysteresis contain a more complex pore structure in which network effects (e.g. pore blocking/percolation) are important.
Figure 3. IUPAC classification of hysteresis loops. [24]
In this case, any simple analysis of the desorption branch is likely to give a misleading picture of the pore size and it is recommended that the pore size distribution should be calculated from the adsorption branch – if possible, with allowance made for delayed condensation. Isotherms with type H3 hysteresis do not exhibit any limiting adsorption at high P/P0. They are given by non-rigid aggregates of plate-like particles or assemblages of slit-shaped pores and in principle should not be expected to provide a reliable assessment of either the pore size distribution or the total pore volume. H4 hysteresis loops are generally observed with
528
Thommes
complex materials containing both micropores and mesopores.A particular feature of many H2, H3 and H4 hysteresis loops is a very steep region of the desorption branch. This characteristic step-down is dependent on the properties of the adsorptive (i.e. instability of the condensate due to cavitation) rather than with the evaporation of pore fluid from a certain class of pores (the origin of this phenomenon will be discussed in Sec. 3.3.3). Hence, the calculations from the desorption branch may lead to an artificial narrow peak in the pore size distribution. The following sections of this review will focus in details on factors contributing to hysteresis, which can be observed in single pores as well as in pore networks. Generally, hysteresis is being considered: (i) on the level of a single pore of a given shape, (ii) cooperative effects due to the specifics of connectivity of the pore network, and (iii) in highly disordered, and inhomogenous porous materials one has to take into account a combination of kinetic and thermodynamic effects spanning the complete disordered pore system. Progress has been achieved in understanding the underlying internal dynamics of hysteresis in disorderd pore systems (see [62] and references therein), however a discussion of this topic is beyond the scope of this chapter. 3.2.2. Type H1 Hysteresis: Pore Condensation and Hysteresis in Single Pores (of Cylindrical or Slit-Like Geometry) and Ordered Pore Networks On the pore level (or independent pore model), adsorption hysteresis is considered as an intrinsic property of the vapor-liquid phase transition in a finite volume system. A classical scenario of capillary condensation implies that the vapor-liquid transition is delayed due to the existence of metastable adsorption films and hindered nucleation of liquid bridges [63-67]. In open uniform cylindrical pores of finite length, these metastabilities occur only on the adsorption branch. Indeed, in an open pore filled by liquid-like condensate, the liquid-vapor interface is already present, and evaporation occurs without nucleation, via a receding meniscus. That is (as indicated before), the desorption process is associated with the equilibrium vapor-liquid transition, whereas hysteresis is caused by the fact that condensation occurs delayed due to
Characterization of Mesoporous Materials
529
the metastabilities associated with the nucleation of liquid bridges. Typically, one observes a hysteresis loop of type H1, i.e. hysteresis that exhibits parallel adsorption and desorption branches. As indicated before, modern, microscopic approaches such as nonlocal density functional theory (NLDFT) and molecular simulation (e.g, Grand Canonical Monte-Carlo simulation) are capable of qualitatively and quantitatively predicting the pore condensation and hysteresis behavior of fluids in ordered mesoporous materials, such as MCM-41 silica. NLDFT correctly predicts (i) the positions of equilibrium vaporliquid transition which is associated with the desorption branch of the isotherm in a pore of given size and geometry; (ii) the pressure where capillary condensation occurs by taking into account delayed condensation due to the metastability associated with the nucleation of liquid bridges (the resulting NLDFT method/kernel is based on so-called metastable adsorption isotherms) [65, 66]. Hence, if the hysteresis is caused solely by the delayed condensation effect, the pore sizes calculated from the adsorption branch (by applying the kernel of metastable adsorption isotherms) and desorption branch (by applying the kernel of equilibrium isotherms) should be in agreement. This was indeed found for MCM-41, SBA-15 silicas [65, 66, 30], which clearly confirmed the applicability of the so-called single (or independent) pore model for these materials. Hysteresis in pore networks is expected to be more complex and very often hysteresis loops which reflect the shapes of types H2 to H4 are observed. However, some novel mesoporous materials such as MCM-48 and KIT-6 silica, which consist of ordered 3D pore networks reveal perfect type H1 adsorption hysteresis, which indicates that pore channels do not exhibit constrictions which would otherwise give rise to type H2 hysteresis due to pore blocking/percolation effects (please see Sec. 3.2.3) and therefore would lead to deviations from type H1 hysteresis. However, the question here is whether the pore condensation and hysteresis behavior in such materials can still be fully described with the independent pore model [69-74]. The availability of novel mesoporous molecular sieves such as MCM-41, SBA-15, MCM-48, KIT-6 also allows to investigate and answer these questions, and indeed it appears that connectivity – even in the absence of typical network effects (e.g.
530
Thommes
pore blocking) can have an impact on the width of the hysteresis loop. It has been observed that hysteresis loops associated with the pore condensation of fluids (e.g. nitrogen, argon) into the three dimensional pore systems of KIT-6 silica are generally more narrow as compared to the hysteresis observed in pseudo-one-dimensional systems, such as certain SBA-15 silica materials [73, 74]. A typical example is shown Fig. 4, which compares the pore condensation and hysteresis behavior of nitrogen (77.4 K) in a SBA-15 silica (pore size of 7.3 nm) with the behavior in a KIT-6 silica sample with the same pore diameter. SBA-15 consists (similar to MCM-41 silica) of a hexagonal packing of cylindrical channels, but depending on details of the synthesis procedure this material can possess intrawall micromesopores, which connect adjacent channels [75]. KIT-6 is a novel type of large-pore mesoporous silica with a cubic Ia3d structure which was synthesized by using triblock copolymer Pluronic P123 as structuredirecting agent [76]. This mesoporous silica material is composed of two interwoven mesoporous networks similar as in MCM-48, but can be synthesized with much larger mean pore diameters. The nitrogen isotherm of the SBA-15 sample reveals a wellpronounced hysteresis loop of type H1. The pore condensation and hysteresis behavior in this SBA-15 sample can be completely described within the independent pore model. This is demonstrated clearly in Fig. 4b, i.e. the pore size distribution curves obtained from the adsorption branch (by applying the NLDFT-kernel of metastable adsorption isotherms) and desorption branch (by applying the NLDFT equilibrium method) are in perfect agreement. This confirms that sorption hysteresis in this SBA-15 silica sample is more or less entirely caused by delayed condensation (i.e., by metastable states of the pore fluid occurring during adsorption/condensation) [30, 66]. The perfect agreement between the pore size distribution curves obtained from the adsorption and desorption branches also confirms that the microporosity of the SBA-15 sample does not affect the pore condensation and hysteresis behavior. This was expected because the micropores (which connect the free-accessible cylindrical-like SBA-15 meso-channels) have been filled at pressures much smaller than the pressure needed for condensation into the mesopores, and in addition micropore filling and
531
Characterization of Mesoporous Materials
pore condensation are essentially two different phenomena, that is, micropores fill via a continuous process whereas a phase transition is involved in mespores filling. 700
0.12
500
0.1
Dv(d) [cm3/Å/g]
Volume [cm3 g-1] STP
600
0.14
SBA-15(7.3nm) KIT-6
400 300
0.08 0.06
200
0.04
100
0.02
0 0
0.2
0.4
0.6
P/P0
(a)
0.8
1
KIT-6(Ads) KIT-6(Des) SBA-15(Ads: 7.3 nm) SBA-15(Des: 7.3 nm)
0 20
40
60
80
100
120
Pore Diameter [Å]
(b)
Figure 4. (a) Nitrogen physisorption isotherms measured at 77.4 K on SBA-15 and KIT-6 silica, and (b) NLDFT pore size distribution curve calculated from nitrogen physisorption isotherms at 77.3 K.
Figure 4 reveals clearly that the hysteresis loop obtained for KIT-6 is smaller in width as compared to SBA-15 silica. The position of both desorption branches agree very well (indicating that both samples have essentially the same pore size which is demonstrated in the PSD shown in Fig. 4b), whereas pore condensation occurs at a somewhat lower relative pressure (i.e. the maximum of the PSD curve of KIT-6 calculated using the adsorption branch is shifted to smaller pore size, indicating that hysteresis in KIT-6 silica cannot be entirely described within the single pore model.) This could indicate that as a consequence of pore connectivity, the pressure range over which metastable pore fluid exists is reduced in the KIT-6 pore network. As a consequence advanced capillary condensation (the effective nucleation barrier associated with the nucleation of the liquid phase is reduced in a pore network) occurs with the consequence that the hysteresis loop is smaller in width as compared to the pseudoone-dimensional SBA-15 system. This observation agrees with the
532
Thommes
predictions of classical theoretical work [77], i.e. one expects that the interconnectivity of pores may accounts for advanced condensation, i.e. the initiation of condensation in wider pores appears after condensation in narrow pores adjacent to them. In qualitative agreement with the KIT-6/SBA-15 result is the observation that argon 77 K hysteresis loops in 3D-MCM-48 silica are smaller in width than in pseudo-one-dimensional MCM-41 silica of the same pore size [70]. One would further expect that the connectivity of pores (in the absence of pore blocking effects) does not have an influence on the position of the capillary evaporation, i.e. desorption. Hence, one can calculate the pore size distribution for KIT-6 silica from the desorption by applying methods based on the single pore model. Indeed, one has found very good agreement between the NLDFT pore size from desorption and the pore diameter obtained using a geometrical model based on unit cell and wall thickness as derived from XRD modeling [73]. Similar observations were also made for large pore materials obtained at high temperature (e.g. 130°C), but now both isotherms of SBA-15 and KIT-6 seem to be consistent with each other (with regard to the width of the hysteresis loop), supporting the hypothesis that SBA-15 materials aged at high temperatures become highly interconnected in agreement with previous reports [78-80]. The schematics shown in Fig. 5 summarize the situations (i.e. different types of pore systems) where type H1 hysteresis can be observed. Figure 5a describes the situation in materials such as MCM-41, which consists of truly independent single mesopores. An example of a material which would correspond to Fig. 5b would be SBA-15 silica of a pore size around 7 nm (corresponding to aging temperatures at around 100 C), i.e. for such a sample the well defined mesopores are connected by micropores [77, 79]. The microporosity has no effect on the width of hysteresis loop. However, in case the connection between pore are mesoporous and are of the same or similar size than the main pore channels (Fig. 5c) one observes advanced condensation which leads to a hysteresis loop which is more narrow than one observed in system with independent pores of the same size (Fig. 5a). The situation depicted in Fig. 5c corresponds to a connected mesopore system with well defined pores (without constrictions) such as can be found for materials such a
Characterization of Mesoporous Materials
533
MCM-48 silica, KIT-6 silica, and large-pore SBA-15 materials which were synthesized at higher temperatures (e.g. 130°C) [78-80].
Figure 5. Pore condensation and hysteresis (type H1) in single pores and ordered pore networks.
3.2.3. Type H2- H4 Hysteresis: Pore Condensation in Ink-Bottle Pores, Complex and/Disorderd Pore Networks In order to describe hysteresis phenomena in complex and disordered adsorbents the network models have been developed [e.g. 81-91]. The network models attribute hysteresis to the so-called pore blocking, or percolation effect. This effect is expected to occur if a pore has access to
534
Thommes
the external surface only through a narrower neck, as it can be found in an ink-bottle pore [81-83]. Some typical pore structures and networks where one expects that pore blocking occurs are shown in Fig. 6. In a network of ink-bottle pores, evaporation of the capillary condensate is obstructed by the pore necks. The vapor pressure, at which a pore body empties, depends on the size of the necks, the connectivity of the network, and the state of the neighbouring pores. The pore network empties when the relative pressure is below a characteristic percolation threshold associated with the onset of a continuous cluster of pores open to the surface [85-90]. In such case, the desorption branch of the hysteresis loop is significantly steeper than the adsorption branch; that results in a triangular hysteresis loop of type H2 according to the IUPAC classification (see Fig. 6a). In this case, any simple analysis of the desorption branch is likely to give a misleading picture of the pore size and it is recommended that the pore size distribution should be calculated from the adsorption branch – if possible, with allowance made for delayed condensation due to metastable pore fluid. For example, there is some evidence that pore blocking/percolation effects contribute significantly to the the observed type H2 hysteresis for fluids adsorbed in porous Vycor glass and some mesoporous silicas produced in a solgel process is associated with the occurence of percolation effects [e.g., 14, 15]. The same type of conventional type H2 hysteresis will also occur in the case of a wide distribution of independent pores with the same or similar neck size, or in a network where the necksize distribution is much more narrow than the size distribution of the main cavities. Recently, a different type of hysteresis loop, which looks like an inverse type H2 hysteresis, has been associated with the occurrence of pore blocking as well (see Fig. 6b). In this case the desorption branch is less steep than the adsorption branch. Such hysteresis could be observed in materials consisting of pores of uniform size, but with varying pore entrance diameters. Inverse type H2 hysteresis has been observed for instance in mesoporous foam consisting of polyhedral foam cells of 60-70 nm diameter, interconnected by cylindrical access channels with several characteristic sizes for the latter [91], or in materials such as FDU-1 silica [92] or KIT-5 silica [93], where the entrances to the
535
Characterization of Mesoporous Materials
spherical pores had been widened by either calcination and or hydrothermal treatment, respectively.
(a)
H2 Hysteresis
(b) “Inverse” H2 Hysteresis
Figure 6. Pore condensation hysteresis in ink-bottle pores and pore networks. (a) Typical pore structures associated with conventional type H2 hsyteresis, and (b) Pore structures associated with inverse-type H2 hsyteresis.
In this case, the distribution of necks/constrictions is much wider than the distribution of main pore cavities, therefore the adsorption/condensation branch is much steeper than the desorption branch. Hence, the distribution of neck sizes can be obtained from an
536
Thommes
analysis of the desorption branch, whereas the pore/cavity size distribution is only available from an analysis of the adsorption branch. Adsorption/desorption mechanisms in ink-bottle pores were recently revisited with the aid of model materials, which contained well-defined ink-bottle pores, theoretical approaches and molecular simulation [94-100, 19, 20]. These studies have revealed that if the neck diameter is smaller than a critical size (estimated to be ca. 5 nm for nitrogen at 77 K), the mechanism of desorption from the larger region involves cavitation (i.e. the spontaneous nucleation of gas bubble). Thus, for a given adsorptive and temperature, the neck diameter of an ink-bottle pore determines the mechanism of evaporation from the pore body. This situation is illustrated schematically in Fig. 7.
Figure 7. Schematic illustration of pore blocking and cavitation controlled evaporation. Adapted from Ref. [19].
As indicated, in the case of pore blocking, the neck diameter is larger than the critical width (∼5 nm for N2(77K) and evaporation occurs at the equilibrium pressure of the corresponding meniscus. Therefore, information about the neck size can be obtained from the desorption
Characterization of Mesoporous Materials
537
branch of the isotherm. In the case, the neck size is smaller than the critical width, desorption from the pore body occurs via cavitation. The pore body can empty by diffusion, while the pore neck remains filled. It is very important to note that if the pore neck is below the critical neck size, the actual width of the pore neck appears not to play any role for the pressure where cavitation occurs, i.e. the cavitation pressure depends solely on the thermophysical properties of the fluid in the main pore cavity. Cavitation controlled evaporation can for instance be found in materials such as SBA-16 [101, 102] and silicas with hierarchical pores structures such as KLE, and KLE/IL silica [19, 20], mesoporous zeolites [e.g., 10] and some clays [e.g., 106]. In particular, KLE/IL serves as model material to study the mechanisms of condensation and evaporation/cavitation [19, 20]; KLE/IL silica consists of uniform spherical mesopores of ca. 14 nm in diameter distributed on a cubic lattice (templated by so-called “KLE” block copolymers), connected through worm-like pores of 2-3 nm (generated by a certain ionic-liquid surfactant) and also by a minor fraction of micropores. The nitrogen (77.4 K) adsorption isotherm, which exhibits a very wide hysteresis, is shown together with the NLDFT pore size distribution curve in Fig. 2. From the shape of adsorption/desorption iostherm shown in Fig. 2, it is obvious that after the large spherical cavities (pore diameter of 14 nm), empy via cavitation giving rise to a steep desorption branch at rel. pressures 0.5, the connecting smaller mesopores (at 3 nm) and micropores remain filled. The hierarchical pore structure had been confirmed by SAXS, TEM systematic physical adsorption experiments, as well as by in-situ SANS during nitrogen physisorption [103, 20]. These recently performed small-angle neutron scattering (SANS) experiments in combination with in-situ nitrogen adsorption at 77 K did not only confirm the structure (i.e. that indeed all large mesopores are connected through the smaller ones), but also provided invaluable insights into details of the adsorption and phase behavior of fluids in such hierarchically sructure materials. In fact, the results provided in agreement with the adsorption data, direct experimental evidence for the occurrence of cavition controlled evaporation from the large mesopores
538
Thommes
in KLE/IL silica which confirms the conclusions drawn from the systematic gas adsorption experiments reported in [19]. As mentioned before, it is expected that at a given temperature, the neck size controls whether pore blocking or cavitation occurs. Above a certain critical neck size pore blocking occurs, and below this cavitation controlled evaporation takes place. Hence, by varying the neck size/entrances to the main pore system, one should be able to observe such a transition from cavitation induced evaporation to pore blocking. Indeed such results have been reported for SBA-16 silica [101], FDU-1 silica [92], and KIT-5 silica [93]. In case of KIT-5 silica (KIT-5 silica is a highly ordered large-cage mesoporous silica with Fm3m close packed structure), the sample was hydrothermally treated for prolonged times, which resulted in an increase in the neck diameter.
Figure 8. Change from cavitation induced evaporation to pore blocking for N2(77K) desorption for hydrothermally treated KIT-5 silica.
Characterization of Mesoporous Materials
539
As shown in Fig. 5, this increase in neck diameter allowed to see a transition from cavitation (visible for the adsorption iostherm obtained after one day of hydrothermal treatment) to pore blocking ontrolled evaporation (note here the appearance of “inverse type H2 hysteresis”) which occurs after a hydrothermal treatment of KIT-5 silica for five days. Furthermore, for a given pore system (i.e. fixed entrance/neck and pore cavity size) one also expects to see a transition from pore blocking to cavitation by increasing the temperatures. This had been predicted by NLDFT [19] and had been experimentally confirmed again for hydrothermally treated KIT-5 silica [93]. This suggests a possibility of estimating the neck size distribution from the desorption isotherm by tuning the experimental conditions, such as tempertaure. Another possibility is to use various probe molecules, i.e. it has been demonstrated that adsorption experiments performed with different adsorptives (e.g., nitrogen and argon at 77 K and 87 K, respectively) allow for detecting and separating the effects of pore blocking/percolation and cavitation in the course of evaporation [19]. This can help to determine neck sizes of materials, which is crucial for a comprehensive pore structure characterization, and is in particular important in many application, because the necks controls the accessibility of molecules into the porosity. In addition to the methods described above, which are based on tuning the experimental conditions for adsorption experiments, neck sizes in ordered materials with cagelike pore structures (e.g. SBA-1, SBA-6 and SBA-16) can be determined for some systems by high-resolution electron crystallography [104]. It was also suggested to determine the entrance sizes by post synthesis surface modification by employing the reaction of the pore wall surface with a series of monofunctional organosilances of gradually increasing ligand sizes in combination with gas adsorption [105]. In addition to m hierarchically structured materials (e.g. KLE/IL silica), and micro-mesoporous zeolites, plugged hexagonal templated silica (PHTS) with combined micro- and mesopores and a tunable amount of both open and inkbotle pores received recently some attention [106, 107].
540
Thommes
Figure 9. Typical nitrogen (77.4 K) adsorption isotherm for a plugged hexagonal templated SBA-15 silica material. The pore size distribution was calculated from the adsorption branch by applying a proper NLDFT method (i.e. metastable adsorption branch kernel). [106]
A typical adsorption isotherm and hysteresis loop is shown in Fig. 9. The adsorption/desorption isotherm is consistent with a structure which exists of both open and blocked cylindrical mesopores. The two-sep desorption isotherm indicates the occurrence of pore blocking/cavitation effects. The high pressure desorption is associated with the evaporation of liquid from open pores. On the other hand, blocked mesopores remain filled until the pressure is lowered to P/P0 = 0.45, after which cavitation of the condensed N2 occurs, which leads to a spontaneous emptying of closed mesopores. Based on the examples discussed here, it appears that cavitation induced evaporation appears to be important for many micro/mesoporous solids and is responsible for the often observed characteristic step down in the desorption isotherm at relative pessure range from 0.4–0.5 (giving rise to type H3 and H4 isotherms). In the past, this characteristic step down was discussed within the framework of the so-called tensile strength hypothesis [109-112, 25] which was associated with the lower closure point of hysteresis. In this classical approach it was supposed that the tensile stress limit of condensed fluid, which is indicated by cavitation, does not depend on the nature and pore structure of the
Characterization of Mesoporous Materials
541
adsorbent, yet is a universal feature of the adsorptive. However, the observed cavitation induced stepwise desorption in ink-bottle pores occcurs at appreciably larger pressures (up to rel. pressures of 0.5) than the “universal” lower closure point of hysteresis mentioned above, which for nitrogen (at 77.4 K) and argon (at 87.3 K) was assumed to be at rel. pressures of 0.42 and 0.38, respectively. Although there is a weak dependency of the cavitation pressure with the cavity size for pore diameters smaller than ca. 12 nm (for spherical pores) [8, 104, 113], the position of the cavitation transitions is mainly determined by the thermophysical state of the pore fluid. Hence, pore size analysis from the desorption would lead to artifical spikes in the pore size range of ca. 4 nm (this has been dicussed recently in the following two review articles [28, 112]). Therefore the pore size distribution for KLE/IL silica (see Fig. 2) was calculated from the adsorption branch by applying a hybrid NLDFT kernel based on a cylindrical pore model for the micropore range and a spherical pore model for the mesopore range in which hysteresis is observed [19].
Figure 10. Examples for cavitation induced evaporation: Argon (87.3 K) adsorption in a pillared clay material.
542
Thommes
It should be clearly noted that based on the current understanding cavitation induced evaporation will not occur in isolated cylindrical or slit-like pores. However, the occurence of cavitation cleary confirms that mesopores can only be accessed by narrow meso-and micropores, or that larger mesoporous cavities are embedded in a micoporous matrix, as it can be found for instance in various micro/mesoporoys active carbons or in some pillared clay materials as shown in Fig. 10. Accordingly, there is no appreciable hysteresis observed when these large mesoporus cavities remain unfilled, which follows from the partial argon adsorption isotherm which was measured up to a relative pressure of 0.8. 4. Comments to Mesopore Size Analysis 4.1. Classical Methods The modified Kelvin equation (Eq. 1) serves as the basis for many methods applied for mesopore analysis, including the Barett-Joyner Halenda method (BJH) [61], which is widely used. The BJH-approach The BJH approach is based on the modified Kelvin equation and the accuracy of the calculated PSD depends on the applicability and the deficiencies of the Kelvin equation (see Sec. 3). As discussed in Sec. 3 and in detail in Ref. [28], the development of model mesoporous silicas (e.g, M41S materials, SBA-15 etc.) over the past decade allowed for the first time direct experimental tests of the validity of the modified Kelvin equation and the BJH method. Because of the high degree of order, the pore diameter of such model substances can be derived by independent methods (methods based on X-ray diffraction, high-resolution transmission electron microscopy etc.). As already indicated in the Introduction of this chapter it was found that the BJH method and related procedures may underestimate the pore size by up to 20-30% for pores smaller than 10 nm if not properly corrected (see for instance Fig. 1). Some improvement could be achieved by improving the classical Broekhoff-de Boer approach [114-116] and by calibrating the Kelvin equation using a series of MCM-silicas of known pore
Characterization of Mesoporous Materials
543
diameter (obtained from XRD interplanar spacing and the mesopore volume). In this manner, a relation between capillary condensation pressures and pore size can be established and used to obtain an empirically corrected Kelvin equation. Such an approach was originally introduced by Kruk-Jaroniec-Sayari and Sayari (calibrated range pore diameter range: 2 nm–7 nm [103]); their approach has been very recently extended by Soloyov and Jaroniec up to 10 nm [22, 23]. But it needs to be stressed that approaches based on calibration are strictly only valid over a limited pore size range. In contrast, modern, microscopic approaches such as NLDFT (and molecular simulation methods allow for an accurate pore size analysis over the complete range of micro- and mesopores for various adsorptive/adsorption pairs assuming slit, cylindrical, spherical or hybrid pore models. In case that the sorption isotherm exhibits a distinct plateau the total specific pore volume can be obtained by converting the amount adsorbed after the pore filling step into liquid volume assuming that the density of the adsorbate is equal to the bulk liquid density at saturation (so-called Gurvich rule) [25]. However, this does not allow for the fact that the degree of molecular packing in small pores and narrow mesopores is dependent on both pore size and pore shape. This has also been addressed by applying for instance methods based on statistical mechanics such as density functional theory and molecular simulation, which also allows one to differentiate between the micropore volume and mesopore volume. Assessing the pore volumes of mesopores and micropores in polymer templated mesoporous silicas and organosilicas is also possible by combining XRD structure modeling and gas adsorption [117]. Gas adsorption technique allows of course only to determine the volume of open pores. Closed porosity cannot be accessed (scattering techniques can here be applied), but can be derived if the true density and particle (bulk) density of the materials are known. 4.2. Pore Size Analysis by Non Local Density Functional Theory (NLDFT) As already discussed NLDFT describes the configuration of the adsorbed phase in pores on a molecular level and thus provides detailed
544
Thommes
information about the local fluid structure near curved solid walls as compared to the bulk fluid. To practically apply the theory for the calculation of the pore size distributions from the experimental adsorption isotherms, theoretical model isotherms have to be calculated using methods of statistical mechanics. In essence, these isotherms are calculated by integration of the equilibrium density profiles, ρ (r), of the fluid in the model pores. A set of isotherms calculated for a set of pore sizes in a given range for a given adsorbate is called a kernel, and can be regarded as a theoretical reference for a given adsorption system. Such a kernel can be used to calculate pore size distributions from adsorption isotherms measured for the corresponding systems. It is important to realize that the numerical values of a given kernel depend on a number of factors, such as the assumed geometrical pore model, values of the gasgas and gas-solid interaction parameters, and other model assumptions. The calculation of pore size distribution is based on a solution of the Integral Adsorption Equation (IAE), which correlates the kernel of theoretical adsorption/desorption isotherms with the experimental sorption isotherm (for details see [65, 66]). The IAE equation is given as: N ( p p0 ) =
∫
WMAX
WMIN
N ( p p0 ,W ) f (W ) d W
(2)
where N(p/p0) = adsorbed volume data (from the experimental sorption isotherm), W = pore width (distance between opposite walls of slit; diameter of cylindrical and spherical pores), N(p/p0,W) = kernel of the theoretical isotherms in pores of different widths, f(W) = pore size distribution function. The IAE equation reflects the assumption that the total isotherm consists of a number of individual “single pore” isotherms multiplied by their relative distribution, f(W), over a range of pore sizes. The set of N(p/p0,W) isotherms (kernel) for a given system (adsorptive/adsorbent) can be obtained by either Density Functional Theory or by Monte Carlo computer simulation. The pore size distribution is then derived by solving the IAE equation numerically. A way to confirm the validity of the calculation, one can compare the calculated NLDFT (fitting)
Characterization of Mesoporous Materials
545
isotherm with the experimental sorption isotherm. The procedure applied to obtain a DFT pore size distribution from an experimental adsorption isotherm is illustrated in Fig. 11.
Figure 11. Schematic representation of the procedure applied to calculate a DFT pore size distribution curve from experimental adsorption/desorption isotherms.
While NLDFT has been demonstrated to be a reliable method for characterization of ordered silica materials, the pore size analysis of materials where the pore walls exhibit pronounced roughness and chemical inhomogeneity is still under investigation [39-41]. A drawback of the standard NLDFT is that they do not take sufficiently into account the chemical and geometrical heterogeneity of the pore walls, i.e. usually a structureless (i.e. chemically and geometrically smooth) pore wall model is assumed. The consequence of this mismatch between the theoretical assumption of a smooth and homogeneous surface and the experimental situation is that theoretical adsorption isotherms exhibit multiple steps associated with layering transitions related to the formation of a monolayer, second adsorbed layer, and so on. Experimentally, stepwise adsorption isotherms are observed only at low
546
Thommes
temperatures for fluids adsorbed onto molecularly smooth surfaces, such as mica or graphite. However, in amorphous porous materials layering transitions are hindered due to inherent energetic and geometrical heterogeneities of real surfaces. The layering steps on the theoretical isotherms can cause artificial gaps on the calculated pore size distributions, because the computational scheme, which fits the experimental isotherm as a linear combination of the theoretical isotherms in individual pores, attribute a layering step to a pore filling step in a pore of a certain size. The problem is enhanced in many porous carbon materials, which exhibit broad PSD’s, and here the artificial layering steps obtained in the theoretical isotherms cause artificial gaps on the calculated pore size distributions. This problem has been addressed in the so-called QSDFT (quenched solid density functional theory), which allows one to take into account pore wall heterogeneity whithin a one-dimensional DFT approach [40]. The application of QSDFT improves significantly the method of adsorption porosimetry for heterogenous porous carbons, the pore size distribution (PSD) functions do not possess anymore the artificial gaps in the regions of ~1 nm and ~2 nm. [41]. Furthermore, the application of QSDFT method to assess the effect of heterogeneity/roughness on pore condensation and hysteresis in porous silica material is in progress [21]. Major progress was also made in theoretical and molecular simulation based approaches to develop more realistic adsorbent models (e.g. for ordered mesoporous silicas such as MCM-41 and SBA-15, but also for disordered materials such a porus glasses) in order to overcome the limitations of the wide used single pore model [e.g., 118, 119].
4.3. Hysteresis and Pore Size Analysis The application of microscopic methods such as NLDFT has also led to major progress in the understanding of adsorption hysteresis. As discussed extensively in Sec. 3, it is now possible to obtain reliable information from both the adsorption and desorption branches of the hysteresis loop. Current NLDFT are capable of qualitatively and quantitatively predicting the pore condensation and hysteresis behavior of fluids in ordered micro/mesoporous materials. If hysteresis is caused
Characterization of Mesoporous Materials
547
solely by the delayed condensation effect, one observes type H1 hysteresis. Pore sizes calculated from the adsorption branch (by applying the kernel of so-called metastable adsorption isotherms) and pore sizes calculated from the desorption branch (by applying the kernel of equilibrium isotherms) must be consistent (see for instance the pore size distribution curves calculated for the SBA-15 sample in Fig. 3). However, type H1 hysteresis was also found for pore condensation into ordered 3D pore networks such as KIT-6, but here on observes so-called advanced pore condensation, hence the position of the pore condensation pressure cannot be quantitavely predicted by an approach which is based on single pore model (for details see Sec. 3.2.2). On the other hand, it appears that if type H1 hysteresis is observed, capillary evaporation/desorption reflects in any case (i.e. in systems consisting of singles pore or ordered pore network) the thermodynamic equilibrium transition. Hence, mesopore connectivity (in absense of pore contrictions) may well effect the condensation pressure, but not the pressure where desorption and evaporation occurs. Therefore one can obtain the pore size distribution of samples which give rise to H1 hysteresis safely from desorption branch. The situation is of course different for systems which consists of inkbottle pores, hierarchically structured and disorderd pore networks. One observes here hysteresis loops of types H2, H3 of H4 (or related hysteresis loop shapes). In this case, the evaporation/desorption of condensed pore fluids from the main pore bodies occurs delayed, due to either a classical pore blocking mechanism or cavitation. A reliable pore size analysis is not possible from the desorption branch, but could be calculated from the adsorption branch (see Fig. 2). It is of advantage here to apply hybrid NLDFT methods which take into account that the pores in such a material are of different geometry and that capillary condensation occurs delayed in the region of hysteresis. By using various probe molecules (e.g., nitrogen and argon at 77.4 K, and 87.3 K), i.e. it is possible to deconvolute the contribution of pore blocking/percolation and cavitation to hysteresis. In case, pore blocking is dominant one can obtain information about the necksize (or distribution of pore entrances) from an analysis of the desorption branch, which is particulary important in many application, because the necks
548
Thommes
controls the accessibility of molecules into the porosity. In case cavitation is present (e.g. for a nitrogen (77.4) adsorption/desorption iostherm, quantitative pore size information cannot be derived from the desorption branch, however the occurrence of cavitation induced evaporation confirmes the existence of a connected micro-mesopore systems, with pore entrances smaller than 5 nm.
5. Summary and Conclusion Summarizing, during recent years major progress has been achived in the understanding of the adsorption, pore condensation and hysteresis behavior of fluids in novel ordered mesoporous materials. This has led to major advances in the structural characterization by physical adsorption, also because of the development and availability of advanced theoretical procedures based on statistical mechanics (e.g. Non-Local Density Functional Theory (NLDFT) and molecular simulation. Contrary to classical, macroscopic thermodynamic approaches, these microscopic methods describe the configuration of the adsorbed phase on a molecular level. The validity of these advanced models (in particular NLDFT) for pore size analysis could be confirmed with the help of ordered mesoporous molecular sieves of known pore size and structure. NLDFT is meanwhile widely used for pore size analysis, featured in an ISO standard and commercially available. While NLDFT has been demonstrated to be a reliable method for characterization of many ordered and hierarchically structured mesoporous silica materials, the pore size analysis of materials where the pore walls exhibit pronounced roughness and chemical inhomogeneity (e.g. as pronounced in porous carbon materials with broad pore size distribution curves). These deficiencies are currently being addressed by various scientific groups, and a novel DFT method, namely QSDFT (quenched solid density functional theory) accounts for the surface geometrical in-homogeneity in form of a roughness parameter. More recently, the focus has shifted towards the structural analysis of advanced micro-mesoporous materials (e.g. micro-mesoporous zeolites, and hierarchically structured porous materials), which have
Characterization of Mesoporous Materials
549
many potential applications (e.g. in catalysis, separations, etc.). Fluids adsorbed in hierarchically structured micro-mesoporous materials can exhibit very complex, but interesting adsorption, pore condensation and hysteresis behavior. A combination of phenomena such as delayed pore condensation, advanced condensation, pore blocking/percolation and cavitation induced evaporation can be observed, which is reflected in characteristic types of adsorption hysteresis. These complex hysteresis loops introduce of course a considerable complication for pore size analysis, but if interpreted correctly, also allow to obtain important and unique information about the pore structure of such advanced micromesoporous material. Its has been demonstrated that for this purpose the use of additional probe molecules (e.g. argon adsorption at 87 K) in addition to nitrogen at 77 K is extremely beneficial, not only to check for consistency, but also to obtain more accurate and comprehensive surface area, pore size and pore structure information. Within this context one also needs to point out the importance of coupling gas adsorption with other experimental techniques (e.g. x-ray and neutron scattering based techniques, electronic microscopy, and others) for studying details of the adsorption and phase behavior of fluids in such hierarchically sructure materials and to arrive at a comprehensive structural characterization of such complex pore networks. Despite the progress made in theoretical and molecular simulation based approaches to develop more realistic adsorbent models, there are still major problems in the characterization of disordered porous materials and mesoporus with imhomogeneous surface chemistry (incl. materials with chemically functionalized surfaces). New challenges are also associated with just emerging new types of micro/mesoporous materials, such as mesoporous metal-organic framework materials (mesoporous MOFs ) mesoporous covalent organic frameworks (COFs), as well as materials with non-rigid pore structures. This needs to be addressed in future experimental and theoretical work with advanced theoretical, computational and experimental approaches, and well chosen model substances.
550
Thommes
Acknowledgment The author would like to thank Maritza Roman for help with the preparation of the graphics. Many thanks also to Dr. Peter Weidler for providing a pillared clay sample.
References 1.
2.
3. 4.
5. 6.
7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
D. Zhao and Y. Wang, In: Introduction to Zeolite Science and Practice (3rd revised Edition), (Cejka, J., van Bekkum, H., Corma, A., Schüth, F., eds.), Stud. Surf. Sci. Catal. 168, Chapter 8, Elsevier, 241 (2007). F. Kleitz, Handbook of Heterogeneus Catalysis, 2nd edition, G. Ertl, H. Knoetzinger, F. Schueth and J. Weitkamp (Eds), Wiley-VCH Verlag GmbH & Co, KgaA, Weinheim, 178 (2008). W. J. Roth, and J. C. Vartuli, Progress and Prospects, Studies in Surface Science and Catalysis, J. Cejka, H. van Bekkum (Eds), 91 (2005). T. J. Barton, M. Bull, W. G. Klemperer, D. A. Loy, B. McEnaney, M. Misono, P. A. Monson, G. Pez, G. W. Scherer, J. C. Vartuli and M. O. Yagi, Chem. Mater. 11, 2633 (1999). J. Cejka, In: Zeolites and Ordered Mesoporous Materials, Studies in Surface Science and Catalysis, J. Cejka and H. van Bekkum (Eds) 157, 111 (2005). S. Mintova and J. Cejka In: Introduction to Zeolite Science and Practice (3rd revised Edition), (Cejka, J., van Bekkum, H., Corma, A., Schüth, F., eds.), Stud. Surf. Sci. Catal. 168 (Chapter 15), Elsevier, 495-525 (2007). W. Guo, L. Huang, P. Deng, Z. Xue, Q. Li, Microp. Mesop. Mater, 44 (2001). O. Sel, D. Kuang, M. Thommes, B. Smarsly, Langmuir 22, 2311 (2006). J. Rouquerol, D. Avnir, C. W. Fairbridge, D. H. Everett, J. H. Haynes, N. Pernicone, J. D. F Ramsay. K. S. W Sing, Pure Appl. Chem. 66, 1739 (1994). A. H. Janssen, A. J. Koster, K. P. de Jong, J. Phys. Chem. B 106, 11905 (2002). D. Vargas-Florencia, I. Furó, R. W. Corkery, Langmuir 24, 4827 (2008). S. Polarz, B. Smarsly, J. Nanosc. Nanotechn. 6(2), 581 (2002). G. A. Zickler, S. Jähnert, W. Wagermaier, S. S. Funari, G. H. Findenegg, O. Paris, Phys. Rev. B 73, 184109 (2006). J. H Page, L. Liu, B. Abeles, E. Herbolzheimer, H. W. Deckmann and D. A. Weitz, Phys. Rev. E 54, 6557 (1996). E. Hoinkis, B. Roehl-Kuhn, Langmuir 21, 7366 (2005). L. A. Solovyov, O. V. Belousov, R. E. Dinnebier, A. N. Shmakov, S. D. Kirik, J. Phys. Chem. B 109, 3233 (2005). N. Muroyama, T. Ohsuna, R. Ryoo, Y. Kubota, O. Terasaki, J. Chem. B 110, 10630 (2006). A. I. Sagidullin, I. Furó, Langmuir 24, 4470 (2008).
Characterization of Mesoporous Materials 19. 20. 21.
22. 23. 24. 25. 26.
27. 28.
29. 30.
31.
32. 33.
34.
35.
36.
551
M. Thommes, B. Smarsly, M. Groenewolt, P. I. Ravikovitch and A. V. Neimark, Langmuir 22, 756 (2006). O. Sel, A. Brandt, D. Wallacher, M. Thommes and B. Smarsly, Langmuir 23, 4724 (2007); N. Smarsly et al., manuscript in preparation (2009). N. Muroyama, A. Yoshimura, Y. Kubota, K. Myasaka, T. Ohsuna, R. Ryoo, P. I. Ravikovitch, A. V. Neimark, M. Takata, and O. Terasaki, J. Phys. Chem C 112, 10803 (2008). M. Kruk, M. Jaroniec and A. Sayari, J. Phys. Chem. B 101, 583 (1997). M. Jaroniec and A. L. Solovyov, Langmuir 22, 6757 (2006). K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Mouscou, R. A Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem. 57, 603 (1985). S. J. Gregg, K. S. W. Sing, Adsorption, Surface Area and Porosity, Academic Press, London (1982). A. V. Neimark, K. S. W. Sing, and M. Thommes, In Handbook of Heterogeneous Catalysis, 2nd edition, G. Ertl, H. Knoetzinger, F. Schueth and J. Weitkamp (Eds), Wiley-VCH Verlag GmbH & Co, KgaA, Weinheim, 721 (2008). F. Rouquerol, J. Rouquerol, K. Sing, Adsorption by Powders and Porous Solids, Academic Press (1999). S. Lowell, J. Shields, M. A. Thomas, M. Thommes, Characterization of Porous Solids and Powders: Surface Area, Pore Size and Density, Springer, The Netherlands (2004). A. Neimark, K. S. W. Sing, M. Thommes, Surface Area and Porosity 2nd Edition, Ertl, Koezinger, Schueth, & Weitkamp (Eds), Verlag Chemie (2006). M. Thommes, Physical Adsorption Characterization of Ordered and Amorphous Mesoporous Materials, in Nanoporous Materials: Science and Engineering, G. Q. Lu & X. S. Zhao (Eds), Imperial College Press, Oxford, 317 (2004). M. Thommes; In: Introduction to Zeolite Science and Practice (3rd revised Edition), (Cejka, J., van Bekkum, H., Corma, A., Schüth, F., eds.), Stud. Surf. Sci. Catal. 168, Chapter 15, Elsevier, 495 (2007). L. D. Gelb, K. E. Gubbins, R. Radhakrishnan and M. Sliwinska-Bartkowiak, Phase Separation in Confined Systems, Rep. Prog. Phys. 62, 1573 (1999). R. Evans, U. M. B. Marconi, and P. Tarazona, J. Chem. Soc. Faraday Trans. 82, 1763 (1986); (b) P. Tarazona, U. M. B. Marconi and R. Evans, Mol. Phys. 60, 573 (1987). K. E. Gubbins, Theory and Simulation of Adsorption in Micropores, In Physical Adsorption: Experiment, Theory and Application (J. Fraissard, Kluwer, Dordrecht, Eds), 65 (1997). (a) A. V. Neimark, Langmuir 11, 4183 (1995); (b) A. V. Neimark, P. I. Ravikovitch, M. Grün, F. Schüth and K. K. Unger J. Colloid, Interface Sci. 207, 159 (1998); (c) P. I. Ravikovitch, A. V Neimark, Langmuir 18, 1550 (2002). J. P. Olivier, W. B. Conklin and V. Szombathley, M., Stud. Surf. Sci. Catal. 87, 81 (1994).
552
Thommes
37.
M. Thommes, B. Smarsly, P. I. Ravikovitch, A. V. Neimark, Langmuir 22, 756 (2006). ISO-15901-3: Pore size distribution and porosity of solid materials by mercury porosimetry and gas adsorption – Analysis of micropores by gas adsorption, to be published in (2007); ISO-15901-2: Pore size distribution and porosity of solid materials by mercury porosimetry and gas adsorption – Analysis of mesopores by gas adsorption, to be published in (2007). E. A. Ustinov, D. D. Do, and V. B Fenelonov, Carbon 44, 653 (2006). P. Ravikovitch and A. V. Neimark, Langmuir 160, 11171 (2006). P. I Ravikovitch, M. Thommes, A. V. Neimark, Carbon, Seattle WA(USA), (2007) The American Carbon Society (ed), CD ROM, Omnipress, Madison WI (USA), (2007), B141(K); A. V. Neimark, M. Thommes; A. V. Neimark, Y. Li, P. I. Ravikovitch, M. Thommes, Carbon (2009), in press. J. U. Keller, R. Staudt, Gas Adsorption Equilibria, Springer (2004). R. S. H. Mikhail and E. Robens, Microstructure and Thermal Analysis of Solid Surfaces, Wiley, Chichester (1983). K. Kaneko, T. Ohba, Y. Hattori, M. Sunaga, H. Tanaka and H. Kanoh, Stud. Surf. Sci. Catal. 144, 11 (2002). J. Rouquerol, In Physical Adsorption: Experiment, Theory and Applications (J. Fraissard and C. W. Conner, Eds) Kluwer, Dordrecht, 17 (1997). W. C. Conner, In Physical Adsorption: Experiment, Theory and Applications (J. Fraissard and W. C. Conner, Eds) Kluwer, Dordrecht, 33 (1997). M. Thommes, S. Tanaka and N. Nishyama, Stud. Surf. Sci. Catal. 165, 551 (2007). S. T. Pauporté, J. Rathousky, J. Phys. Chem. C 111, 7639 (2007). F. Hung, Bhattacharya, B. Coasne, M. Thommes and K. E. Gubbins, Argon and krypton adsorption on templated mesoporous silicas: molecular simulation and experiment, Adsorption 13, 425 (2007). M. Thommes, R. Koehn and M. Froeba, Appl. Surf. Sci. 196(1-4), 239 (2002). M. Thommes, R. Köhn, M. Fröba, J. Phys. Chem. B 104, 793 (2000). S. Brunauer, P.H. Emmett, E. Teller, J. Amer. Chem. Soc. 60, 309 (1938). F. Rouquerol, J. Rouquerol, C. Peres, Y. Grillet and M. Boudellal, In: Characterization of Porous Solids (S. J. Gregg, K. S. W. Sing, H. F. Stoeckli, Eds), The Society of Chemical Industry, 107 (1978). L. Jelinek and E. S. Kovats, Langmuir 10, 4225 (1994). A. Galarneau, D. Desplantier, R. Dutartre and F. Di Renzo, Microp. Mesop. Mater. 27, 107 (1999). J. Frenkel, Kinetic Theory of Liquids, Oxford University Press (1946). G. D. Halsey, J. Chem. Phys. 16, 931(1948). T. L. Hill, Advances in Catalysis IV, Academic Press, N.Y, 236 (1952). W. T. Thompson, Philos. Mag. 42, 448 (1871); (b) Z. Zsigmondy, Anorg. Chem. 71, 356 (1911).
38.
39. 40. 41.
42. 43. 44. 45. 46. 47. 48. 49.
50. 51. 52. 53.
54. 55. 56. 57. 58. 59.
Characterization of Mesoporous Materials 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77.
78. 79. 80. 81. 82. 83. 84. 85. 86.
553
L. H. J. Cohan, J. Am. Chem. Soc. 60, 433 (1938); (b) L. H. Cohan, J. Am. Chem. Soc. 66, 98 (1944). E. P. Barrett, L. G. Joyner and P. P. Halenda, J. Am. Chem. Soc. 73, 373 (1951). R. Valiullin, S. Naumov, P. Galvosas, J. Kaerger, H-J. Woo, F. Porcheron, and P.A. Monson, Nature 443, 965 (2006). D. H Everett, In: The Solid-Gas Interface, (E. A. Flood, Ed), Marcel Decker, N.Y., 2 (1967). P. C. Ball, R. Evans, Langmuir 5, 714 (1989); (b) R. J. Evans, Phys. Condens. Matter. 2, 8989 (1990). A. V. Neimark, P .I. Ravikovitch, A. Vishnyakov, Phys. Rev. E 62, 1493 (2000). (a) A. V. Neimark, P. I. Ravikovitch, Microp. Mesop. Mater. 44, 697 (2001); (b) P. I. Ravikovitch, A. V. Neimark, Colloids Surf., A 187, 11 (2001). P. A. Monson, Langmuir 24, 12295 (2008). M. Thommes, R. Skudas, K. K. Unger, and D. Lubda, J. Chromatog. 1191, 57 (2008). P. I. Ravikovitch, A. V. Neimark, Langmuir 16, 2419 (2000). M. Thommes, R. Koehn, M. Froeba, Stud. Surf. Sci. Catal. 142, 1695 (2002). M. Morishige, K. Tateishi, S. Fukuma, J. Phys. Chem B. 107, 5177 (2003). B. Coasne, A. Galarneau, F. Di. Renzo, R. M. Pellenq, Langmuir 22, 11097 (2006). F. Kleitz, F. Berube, C-M. Yang and M. Thommes, Stud. Surf. Sci. Catal 170, 1843 (2007). M. Thommes, F. Kleitz et al., manuscript in preparation (2009). D. Y. Zhao, J. L Feng., Q. S. Huo., N. Melosh, G. Frederickson., B. F. Chemelka, and G. D. Stucky, Science 279, 548 (1998). F. Kleitz, S. H. Choi, and R. Ryoo, Chem. Commun., 2136 (2003). A. V. Neimark, In Characterization of Porous Solids III (J. Rouquerol, F. Rodriguez-Reinoso, K. S. W. Sing, K. K. Unger, Eds.), Studies in Surface Science and Catalysis 87, Elsevier, Amsterdam, 67 (1994). A. Galarneau, H. Cambon, F. Di Renzo, R. Ryoo, M. Choi, F. Fajula, New J. Chem. 27, 73 (2003). E. B. Celer and M. Jaroniec, J. Am. Chem. Soc. 128, 14408 (2006). R. Nicolas, F. Berube, T.W. Kim, M. Thommes, F. Kleitz, Stud. Surf. Sci. Catal. 174, A 141 (2008). J. H. De Boer, The Structure and Properties of Porous Materials, Butterworths, London (1958). E. O. Kraemer, In Treatise on Physical Chemistry, D.Van Nostrand, N.Y. (1931). J. W. J. McBain, J. Am. Chem. Soc. 57, 699 (1935). G. Mason, J. Coll. Interf. Sci. 88, 36 (1982). A. V. Neimark, Rep. Acad. Sci., USSR (Phys. Chem.) 273, 867 (1983); (b) A. V. Neimark, Colloid J., USSR 46, 1004 (1986). L. Z. Zhang, N. A. Seaton, Langmuir 9, 2576 (1993); (b) H. Liu, N. A. Seaton, Chem. Eng. Sci. 49, 1869 (1994).
554
Thommes
87.
F. Rojas, I. Kornhauser, C. Felipe, J. M. Esparza, S. Cordero, A. Dominguez, J. L. Riccardo, Phys. Chem. Phys. 4, 2346 (2002). A. V. Neimark, Percolation theory of capillary hysteresis phenomena and characterization of porous solids, Stud. Surf. Sci. Catal. 62, 67 (1991). M. Parlar and Y. C. Yortsos, Percolation theory of vapor adsorption-desorption processes in porous materials, J. Coll. Interf. Sci. 124, 162 (1988). H. Zhu, L. Zhang and N. A. Seaton, Sorption hysteresis as a probe of pore structure, Langmuir 9, 2576 (1993). D. Vargas-Florencia, I. Furo, R. W. Corkery, Langmuir 24, 4827 (2008). M. Kruk, E. B. Celer, J. Matos, S. Pikus, M. Jaroniec, J. Phys. Chem. B 109, 3838 (2005). K. Morishige, M. Tateishi, F. Hirose, K. Aramaki, Langmuir, 9220 (2006). H. J. Woo, L. Sarkisov, P. A. Monson, Langmuir 17, 7472 (2001). L. Sarkisov, P. A. Monson, Langmuir 17, 7600 (2001). (a) P. I. Ravikovitch, A. V. Neimark, Langmuir 18, 1550 (2002); (b) P. I. Ravikovitch, A. V. Neimark, Langmuir 18, 9830 (2002). A. Vishnyakov, A. V. Neimark, Langmuir 19, 3240 (2003). B. Libby, V. Monson, Langmuir 20, 6482 (2004). B. Smarsly, M. Thommes, P. I. Ravikovitch, A. V. Neimark, Adsorption 11, 65 (2005). M. Thommes, B. Smarsly, P. I. Ravikovitch, A. V. Neimark, to be published, (2009). T. W. Kim, R. Ryoo, M. Kruk, K. P. Gierszal, M. Jaroniec, S. Kamyia, O. Terasaki, J. Phys. Chem. B 108, 11480 (2004). F. Kleitz, T. Czuryszkiewicz, L. A. Solovyov, M. Linden, Chem. Mater. 18, 5070 (2006). M. Thommes, B. Smarsly, O. Sel, A. Brandt, D. Wallacher, BENSC experimental for proposal Nr. PHY-04-1396, 2007; B. Smarsly et al., manuscript in preparation (2009). Y. Sakamoto, M. Kaneda, O. Terasaki, D. Y. Zhao, J. M. Kim, G. D. Stucky, H. J. Shin, and R. Ryoo, Nature 408, 449 (2000). C. Yu, Y. Yu, D. Zhao, Chem. Commun., 575 (2000); (b) M. Kruk, V. Antochshuk, J. R. Matos, L. P. Mercuri, M. J. Jaroniec, Am. Chem. Soc. 124, 768 (2002). P. Van Der Voort, P. I. Ravikovitch, K. P. De Jong, A. V. Neimark, A. H. Janssen, M. Benjelloun, E. Van Bavel, P. Cool, B. M. Weckhuysen, E. F. Vansant, Chem. Commun., 1010 (2002). S.-E. Park and N. Jiang, In Zeolites: From Model Materials to Industrial Catalysis (J. Cejka, J. Perez-Pariente and W. J. Roth, Eds), 81 (2008). C. Bisio, G. Gatti, E. Boccaleri, L. Marchese, L. Bertinetti, S. Coluccia, Langmuir 24, (2008). (a) O. Kadlec, M. M. J. Dubinin, Coll. Interf. Sci. 31, 479 (1969); (b) C. G. V. Burgess, D. H. Everett, J. Coll. Interf. Sci. 33, 611 (1970).
88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103.
104. 105. 106.
107. 108. 109.
Characterization of Mesoporous Materials 110. 111. 112. 113. 114.
115. 116. 117. 118. 119.
555
A. Schreiber, S. Reinhardt, G. H. Findenegg, Stud. Surf. Sci. Catal. 144, 177 (2002). C. G. Sonwane and S. K. Bhatia, Langmuir 15, 5347 (1999). J. C. Groen, L. A. Peffer, J. Perez-Ramirez, Microp. Mesop. Mater. 60, 1 (2003). M. Thommes, A. V Neimark, C. Rasmussen, B. Smarsly, mansucript in preparation (2009). J. C. P. Broekhoff and J. H. de Boer, J. Catal, 9 (1967) 8; (b) J. C. P. Broekhoff and J. H. de Boer, J. Catal, 9 (1967) 15; (c) J. C. P. Broekhoff and J. H. de Boer, J. Catal. 10 (1968) 377; (d) J. C. P. Broekhoff and J. H. de Boer, J. Catal, 10 (1968) 368. K. P. Gierszal, T.-W. Kim, R. Ryoo, M. Jaroniec, J. Phys. Chem. B 109, 23263 (2005). P. Kowalczyk, M. Jaroniec, A. P. Terzyk, K. Kaneko, D. D. Do, Langmuir 21, 1827 (2005). M. Jaroniec, L. A. Solovyov, Chem. Commun., 2242 (2006). B. Coasne, F. R. Hung, R. J. M. Pellenq, F. R. Siperstein, and K. E. Gubbins, Langmuir 22, 194 (2006). F. Porcheron, M. Thommes, R. Ahmad and P. A. Monson, Langmuir 23, 3372 (2007).