Advanced Polymer Nanoparticles: Synthesis and Surface Modifications 1439814430, 9781439814437

Polymer latex particles continue to become increasingly important in numerous commercial applications. Advanced synthesi

308 94 7MB

English Pages 384 Year 2010

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover Page......Page 1
Title Page......Page 3
ISBN 9781439814437......Page 4
Contents......Page 5
Preface......Page 7
Editor......Page 9
Contributors......Page 10
1.1 Introduction......Page 12
1.2 Emulsion Polymerization......Page 13
1.3 Controlled Polymerization and its Use in Emulsion Polymerization Processes......Page 20
1.4 Conventional and Controlled Miniemulsion Polymerization......Page 27
1.5 Generation of Copolymer or Core-Shell Particles......Page 31
References......Page 36
2.1 Introduction......Page 40
2.2 Equilibrium and Nonequilibrium Morphologies......Page 41
2.2.1 Equilibrium Morphologies......Page 42
2.2.2 Nonequilibrium Morphologies......Page 44
2.3.1 Emulsion Polymerization......Page 46
2.3.2 Miniemulsion Polymerization......Page 54
2.3.4 Dispersion Polymerization......Page 58
2.3.5 Suspension Polymerization......Page 59
2.3.6 Other Techniques......Page 60
2.4.1 Transmission Electron Microscopy......Page 63
2.4.4 Additional Techniques Used for Particle Characterization......Page 65
References......Page 66
3. Advanced Polymer Nanoparticles with Nonspherical Morphologies......Page 72
3.1 Introduction......Page 73
3.2 Overview of Approaches to Nonspherical Polymer Nanoparticles Synthesis......Page 74
3.3.1 Phase Separation and Seeded Emulsion Polymerization......Page 75
3.3.2 Solvent Evaporation and Seeded Dispersion Polymerization......Page 87
3.4 Physical Posttreatment Approaches......Page 89
3.4.1 Stretching......Page 90
3.4.2 Compression......Page 97
3.4.3 Self-Assembly......Page 98
3.5 Summary......Page 103
References......Page 104
4. Block, Graft, Star and Gradient Copolymer Particles......Page 108
4.1 Introduction......Page 109
4.2 Polymerization Techniques for Complex Macromolecular Architectures......Page 110
4.2.1 Block Copolymer Topology......Page 111
4.2.2 Graft Copolymer Topology......Page 112
4.2.4 Gradient Copolymer Topology......Page 114
4.3 Polymer Nanoparticles in Heterogeneous Polymerization......Page 115
4.3.1 Approaches for Successful Application of CRP in Heterogeneous Media......Page 116
4.3.2 Challenges with the Application of CRP in Heterogeneous Media......Page 118
4.3.3 Characteristics of the Individual CRP Techniques in Miniemulsion Polymerization......Page 119
4.3.4 Nanoparticles with Complex Architecture Synthesized via CRP in (Mini)emulsion......Page 121
4.4.1 Synthesis of Amphiphilic Copolymers......Page 124
4.4.2 Nanostructures via Self-Assembly of Copolymers......Page 125
4.4.3 Nanoparticles Synthesis by Cross-Linking of Self-Assembled Structures......Page 126
4.4.5 Potential Applications of Cross-Linked Nanoparticles......Page 132
4.5 Conclusions......Page 134
References......Page 135
5. Polymer Nanoparticles by Reversible Addition-Fragmentation Chain Transfer Microemulsion Polymerization......Page 144
5.1 Introduction......Page 145
5.2 Uncontrolled Microemulsion Polymerization......Page 146
5.2.1 Impact of Monomer Solubility......Page 147
5.2.2 Microemulsion Copolymerization......Page 149
5.2.3 Multiple Addition and Semicontinuous Microemulsion Polymerization......Page 150
5.3 Reversible Addition-Fragmentation Chain Transfer Polymerization......Page 151
5.4 Reversible Addition-Fragmentation Chain Transfer Microemulsion Polymerization......Page 154
5.4.1 Microemulsion Polymerization Kinetics......Page 155
5.4.2 Molecular Weight and Polydispersity......Page 164
5.4.3 Latex Particle Size......Page 172
References......Page 174
6.1 Introduction......Page 180
6.2 pH-Responsive Polymer Micelle Particles......Page 183
6.3 pH-Responsive Cross-Linked Micelle Particles......Page 188
6.4 pH-Responsive Microgel Particles......Page 190
6.5 pH-Responsive Branched Copolymer Particles......Page 193
6.6 Polymer Nanoparticles with pH-Responsive Surfaces......Page 194
6.7 Applications of pH-Responsive Polymer Particles......Page 196
6.8 Outlook......Page 200
References......Page 201
7.1.1 Thermo-Responsive Polymer......Page 208
7.1.2 Phase Transition of Thermo-Responsive Polymer......Page 209
7.1.3 Thermo-Responsive PNIPAAm Nanoparticles......Page 210
7.2 Thermo-Responsive Nanoparticle Synthesis and Characterization......Page 211
7.3 Thermo-Responsive Nanoparticle Size Control......Page 216
7.4 Thermo-Responsive Nanoparticle Applications......Page 219
7.5 Conclusions and Future Development......Page 225
References......Page 226
8.1 Introduction......Page 234
8.2 Architecture of Polymer Particles Possessing Radical Initiating Sites on the Surface by Emulsion Copolymerization and Synthesis of Core-Brush Structures by Photoinduced ATRP Approach......Page 236
8.3 Synthesis of Silica Particles Coated with High-Density Polymer Brushes......Page 239
8.4 Synthesis of Polymer Brushes Encapsulated Silica Particles by DC-Mediated Living Radical Polymerization......Page 241
8.5 Synthesis of Silica Hybrid Nanoparticles Modified with Photofunctional Polymers and Construction of Colloidal Crystals......Page 246
8.6 Architecture of Polymer Particles Composed of Brush Structure at Surface and Application for Structural Color Materials......Page 250
8.7 Surface Modification of Polymer Particles via RAFT Polymerization......Page 260
8.8 Surface Modification of Silica Nanoparticles via Nitroxide-Mediated Radical Polymerization......Page 264
References......Page 266
9. Effects of Nano-Sized Polymerization Locus on the Kinetics of Controlled/Living Radical Polymerization......Page 274
9.1 Introduction......Page 275
9.2 Origin of Livingness in Controlled/Living Radical Polymerization (CLRP)......Page 276
9.3.1 Basic Concept......Page 278
9.3.2 SFRP and ATRP (Radical Long Life-ization Process)......Page 279
9.3.3 RAFT and DT......Page 293
9.4.1 Fundamental Distribution in CLRP......Page 304
9.4.2 SFRP and ATRP in Dispersed Systems......Page 308
9.4.3 RAFT Polymerization in Dispersed Systems (Normal Lifetime Process)......Page 309
9.5 Summary......Page 313
References......Page 314
Appendix: Formulation of Various Average Times......Page 316
10.1 Introduction......Page 318
10.2 Particle Nucleation......Page 319
10.3 Functional Particles by Surfactant-Free Polymerization......Page 320
Acknowledgments......Page 335
References......Page 336
11.1 General Introduction......Page 340
11.2 Historical Introduction......Page 341
11.3 Specific Introduction......Page 344
11.4 Role of Initiators in Aqueous Emulsion Polymerization......Page 345
11.5 Initiation and Interfaces......Page 349
11.6 Surface Active Initiators–Inisurfs......Page 352
11.7 Polymers as Initiators for Emulsion Polymerization......Page 358
References......Page 364
B......Page 372
C......Page 373
D......Page 374
G......Page 375
L......Page 376
M......Page 377
P......Page 378
S......Page 380
T......Page 382
Z......Page 383
Back Page......Page 384
Recommend Papers

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications
 1439814430, 9781439814437

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

ADVANCED POLYMER NANOPARTICLES Synthesis and Surface Modifications

ADVANCED POLYMER NANOPARTICLES Synthesis and Surface Modifications

Edited by

Vikas Mittal

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2011 by Taylor and Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number: 978-1-4398-1443-7 (Hardback) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright. com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Advanced polymer nanoparticles : synthesis and surface modifications / [edited by] Vikas Mittal. p. cm. “A CRC title.” Includes bibliographical references and index. ISBN 978-1-4398-1443-7 (hardcover : alk. paper) 1. Polymerization. 2. Nanoparticles. 3. Polymers--Surfaces. I. Mittal, Vikas. II. Title. TP156.P6A38 2011 620.1’92--dc22 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

2010020564

Contents Preface .................................................................................................................... vii Editor........................................................................................................................ix Contributors ............................................................................................................xi 1. Polymer Latex Technology: An Overview .................................................1 V. Mittal 2. Synthesis of Polymer Particles with Core-Shell Morphologies .......... 29 Claudia Sayer and Pedro Henrique Hermes de Araújo 3. Advanced Polymer Nanoparticles with Nonspherical Morphologies................................................................................................. 61 Yongxing Hu, Jianping Ge, James Goebl, and Yadong Yin 4. Block, Graft, Star, and Gradient Copolymer Particles .......................... 97 H. Matahwa, E. T. A. van den Dungen, J. B. McLeary, and B. Klumperman 5. Polymer Nanoparticles by Reversible Addition-Fragmentation Chain Transfer Microemulsion Polymerization .................................. 133 J. O’Donnell and E. Kaler 6. pH-Responsive Polymer Nanoparticles ................................................. 169 Jonathan V. M. Weaver 7. Smart Thermo-Responsive Nanoparticles ............................................ 197 Peng Tian and Qinglin Wu 8. Surface Tailoring of Polymer Nanoparticles with Living Polymerization Methods ...........................................................................223 Koji Ishizu and Dong Hoon Lee 9. Effects of Nano-Sized Polymerization Locus on the Kinetics of Controlled/Living Radical Polymerization ........................................... 263 Hidetaka Tobita 10. Functional Polymer Particles by Emulsiier-Free Polymerization ... 307 V. Mittal

v

vi

Contents

11. Polymer Nanoparticles with Surface Active Initiators and Polymer Initiators ....................................................................................... 329 Klaus Tauer Index ..................................................................................................................... 361

Preface Polymer latex particles are a very important class of polymeric materials, which are used for a large number of commercial applications. These particles are synthesized in the aqueous dispersion phase by numerous synthesis methodologies such as emulsion, miniemulsion, microemulsion, dispersion, suspension, inverse emulsion (in organic phase), polymerization, etc. Over the years, signiicant enhancement in the techniques dealing with the synthesis and surface tailoring of polymer particles has been achieved, which has also resulted in the widening of the application spectrum of these particles. These advances include use of advanced controlled polymerization means such as nitroxide-mediated polymerization, atom transfer radical polymerization, radical addition fragmentation transfer polymerization, etc., as well as use of advanced stabilizers, surface modiiers, etc. These advances have made it possible to achieve polymer particles with speciic sizes consisting of polymer chains of speciic molecular weights and tailorable chemical compositions or properties according to the requirement. Because the advanced synthesis techniques are the key to achieve new functional properties in the polymer nanoparticles, and the surface modiications of these particles are required to ensure their use for speciic applications, it is of immense importance to bring readers up-to-date on recent advances in these ields. This information will enable readers to design the required particle systems. This book thus serves the purpose of summarizing the developments in the synthesis and surface modiication techniques to generate advanced polymer particles, and the contents have been accordingly organized. Chapter 1 introduces polymer latex technology with an overview of the various conventional and recent synthesis methodologies. Synthesis and characterization of particles with core-shell morphologies have been focused on in Chapter 2. Chapter 3 reports the generation of nonspherical polymer particles by following different synthetic routes. The generation of speciic architectures such as block, star, graft, and gradient copolymer particles has been detailed in Chapter 4. Microemulsion polymerization using reversible addition-fragmentation chain transfer controlled radical polymerization is the subject of Chapter 5. In Chapter 6, pH-responsive nanoparticles have been described, whereas the synthesis of smart thermally responsive particles has been reported in Chapter 7. Surface tailoring of various organic and inorganic nanoparticles by polymers is the subject of Chapter 8. Theoretical studies on the kinetics of controlled radical polymerization techniques have been explained in Chapter 9. Chapter 10 reports the synthesis of functional nanoparticles by using the surfactant-free emulsion polymerization vii

viii

Preface

approach. Chapter 11 describes various surface-active initiators as well as polymeric stabilizers developed for polymer nanoparticles in recent years. At this juncture, I would like to express my heartfelt thanks to Taylor & Francis Group for their kind support during the project. I am equally thankful to Professor Massimo Morbidelli at the Swiss Federal Institute of Technology, Zurich, Switzerland, who has been my guide in polymer latex technology. I am indebted to my family, especially my mother, whose continuous support and motivation have made this work feasible. I dedicate this book to my dear wife Preeti, for her valuable help in coediting the book as well as for her efforts in improving the quality of the book. Vikas Mittal Ludwigshafen, Germany

Editor Dr. Vikas Mittal studied chemical engineering at Punjab Technical University in Punjab, India. He later obtained his master of technology in polymer science and engineering from the Indian Institute of Technology, Delhi, India. Subsequently, he joined Professor U. W. Suter’s polymer chemistry group at the Department of Materials at the Swiss Federal Institute of Technology, Zurich, Switzerland, where he worked for his doctoral degree with a focus on the subjects of surface chemistry and polymer nanocomposites. He also jointly worked with Professor M. Morbidelli at the Department of Chemistry and Applied Biosciences on the synthesis of functional polymer latex particles with thermally reversible behaviors. After completion of his doctoral research, Dr. Mittal joined the Active and Intelligent Coatings section of Sun Chemical Group Europe in London. He worked for the development of water- and solvent-based coatings for food-packaging applications. He later joined BASF Polymer Research in Ludwigshafen, Germany, as a polymer engineer, where he is currently working as a laboratory manager responsible for the physical analysis of organic and inorganic colloids. His research interests include organic–inorganic nanocomposites, novel iller surface modiications, thermal stability enhancements, polymer latexes with functionalized surfaces, etc. He has authored more than 40 scientiic publications, book chapters, and patents on these subjects.

ix

Contributors

Pedro Henrique Hermes de Araújo Department of Chemical Engineering Federal University of Santa Catarina Florianópolis, Brazil

Dong Hoon Lee Department of Organic Materials and Macromolecules Tokyo Institute of Technology Tokyo, Japan

Jianping Ge Department of Chemistry University of California–Riverside Riverside, California

H. Matahwa Department of Chemistry and Polymer Science University of Stellenbosch Matieland, South Africa

James Goebl Department of Chemistry University of California–Riverside Riverside, California Yongxing Hu Department of Chemistry University of California–Riverside Riverside, California Koji Ishizu Department of Organic Materials and Macromolecules Tokyo Institute of Technology Tokyo, Japan E. Kaler Stony Brook University Stony Brook, New York B. Klumperman Department of Chemistry and Polymer Science University of Stellenbosch Matieland, South Africa and Lab of Polymer Chemistry Eindhoven University of Technology Eindhoven, the Netherlands

J. B. McLeary Plascon Research Centre University of Stellenbosch Matieland, South Africa V. Mittal Polymer Research BASF SE Ludwigshafen, Germany and Department of Chemistry and Applied Biosciences Institute of Chemical and Bioengineering ETH Zurich Zurich, Switzerland J. O’Donnell Iowa State University Ames, Iowa Claudia Sayer Department of Chemical Engineering Federal University of Santa Catarina Florianópolis, Brazil xi

xii

Contributors

Klaus Tauer Department of Colloid Chemistry Max Planck Institute of Colloids and Interfaces Golm, Germany

E. T. A. van den Dungen Department of Chemistry and Polymer Science University of Stellenbosch Matieland, South Africa

Peng Tian School of Renewable Natural Resources Louisiana State University Baton Rouge, Louisiana

Jonathan V. M. Weaver Department of Chemistry University of Liverpool Liverpool, United Kingdom

Hidetaka Tobita Department of Materials Science and Engineering University of Fukui Fukui, Japan

Qinglin Wu School of Renewable Natural Resources Louisiana State University Baton Rouge, Louisiana

Yadong Yin Department of Chemistry University of California–Riverside Riverside, California

1 Polymer Latex Technology: An Overview*

V. Mittal CONTENTS 1.1 Introduction ....................................................................................................1 1.2 Emulsion Polymerization .............................................................................2 1.3 Controlled Polymerization and its Use in Emulsion Polymerization Processes .............................................................................9 1.4 Conventional and Controlled Miniemulsion Polymerization............... 16 1.5 Generation of Copolymer or Core-Shell Particles ................................... 20 References............................................................................................................... 25

1.1 Introduction Polymer nanoparticles ind use in a number of applications like coatings, adhesives, paints, etc. The applications of these nanoparticles are signiicantly affected by their physical properties as well as surface morphology, which can be controlled by the synthesis process used to generate such particles. Emulsion polymerization and its modiied methodologies are the most commonly used techniques to achieve polymer nanoparticles. These techniques also allow the generation or surface functionalization of the particles either in situ or by following separate speciic steps. Polymerization of monomer by emulsion polymerization offers signiicant advantages in the whole polymerization process as compared to bulk and solution polymerization methods. It allows better control of the heat and viscosity of the system, and emulsion polymerization allows the achievement of an increase in molecular weight of the polymer chains without negatively impacting the rate of polymerization [1]. In emulsion polymerization, most of the monomer is present as monomer droplets in the aqueous phase, which diffuses to the polymerizing particles during the course of polymerization. The diffusion of the monomer is *

The work was carried out at Institute of Chemical and Bioengineering, Department of Chemistry and Applied Biosciences, ETH Zurich, Zurich, Switzerland.

1

2

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

possible when the monomer is partially water soluble. Thus, emulsion polymerization is not very effective with extremely hydrophobic and extremely hydrophilic monomers. The extremely hydrophobic monomers would always stay in the monomer droplets, leading to no polymerization, whereas the hydrophilic monomers would polymerize mainly by homogenous polymerization and not micellar polymerization. To circumvent these dificulties, miniemulsion polymerization is used [2,3]. In this technique, the diffusion of the monomer molecules through the aqueous phase is not required, as the monomer droplets are directly polymerized. Therefore, such a technique has no problem in achieving the polymerization of even extremely hydrophobic monomers. To polymerize very hydrophilic monomers, inverse miniemulsion can be used. Combination of controlled polymerization methods like nitroxide-mediated polymerization, atom transfer radical polymerization, and reversible addition fragmentation chain transfer polymerization with the emulsion and miniemulsion polymerization methods has further enhanced the possibilities of achieving functional polymer particles [4]. By using these techniques, synthesis of functional block copolymer or graft copolymer particles can be achieved, which is not possible by using conventional emulsion polymerization techniques owing to the very short lifetime of the radicals. The surface morphologies of the particles can also be eficiently controlled or tuned by using such controlled polymerization methods, which expands the spectrum of application of these particles. This chapter aims to provide an overview of the conventional emulsion polymerization methods and the more advanced methods of synthesizing polymer particles.

1.2 Emulsion Polymerization Emulsion polymerization is a heterogeneous polymerization technique that uses water as dispersion medium for the polymerization of water-insoluble monomers in the form of suspended particles. Styrene, methyl methacrylate, butyl acrylate, etc. are examples of the most commonly used monomers for the generation of polymers by emulsion polymerization. The surfactants are generally used to provide colloidal stability to the system. The surfactant can be cationic, anionic, or nonionic, and its amount exceeds the critical micelle concentration signiicantly. The surfactants form micelles in the system in which the polymerization takes place. Thus, this process can be visualized as a bulk polymerization in each of the suspended particles. Polymerization by this mode helps to circumvent the problems of heat and viscosity control generally associated with bulk polymerization. By changing the amount of surfactant, the molecular weight of the polymer chains can be increased without decreasing the polymerization rate, which is not possible in other modes of polymerization. The presence of a signiicant amount of surfactant

3

Polymer Latex Technology: An Overview

in the system can lead to certain disadvantages; however, many applications of particles are not affected by the presence of surfactants. Surfactant-free polymerization can also be used to generate polymer particles in order to circumvent the problems associated with the use of emulsiier, but in this case, the mode of polymer nucleation is completely different. As mentioned previously, the amount of the surfactant exceeds the critical micelle concentration in the emulsiied emulsion polymerization process. The micelles formed as a result of this excess amount have a size in the range of 10 nm, and one micelle generally consists of 100–200 surfactant molecules [1]. Surface tension of the solution decreases with the addition of surfactant at critical micelle concentration. A host of other solution properties are also affected at critical micelle concentration of the surfactant, which include conductivity, turbidity, osmotic pressure, etc. [5]. However, in the emulsion polymerization process, it is the reduction in the surface tension of the aqueous phase that is of prime importance. Because surfactants are amphiphilic molecules containing one hydrophobic part and one hydrophilic part, in the micelle they orient themselves in a way so that the hydrophobic part forms the inner part of the micelle and the hydrophilic part radiates away from this inner part of the micelle into the aqueous phase. The resulting hydrophobic space inside the micelle owing to the self-assembly of the surfactant molecules is an ideal place for the hydrophobic monomer to reside and also provides an ideal environment for the radicals to enter the micelle. Figure 1.1 shows the representation of the association of the surfactant molecules after the critical micelle concentration of the surfactant is reached [6]. When the inverse emulsion polymerization is used, then the hydrophilic part of the surfactant forms the inner part of the micelles and the hydrophilic chains intermix with the organic dispersion phase.

(a)

(b)

FIGURE 1.1 Organization of the surfactant molecules (a) below and (b) above the critical micelle concentration of the surfactant in the aqueous solution.

4

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

The monomers used for emulsion polymerization are water insoluble (water soluble for inverse emulsion polymerization). However, the monomer should have some extent of water solubility in order to diffuse through the aqueous phase as required during the course of polymerization. When the monomer is added to the system, a part of the monomer enters the micelles and a part is dissolved in the aqueous phase owing to partial water solubility. However, the majority of the monomer is present in the form of monomer droplets. The size of the monomer droplets is much larger than that of micelles; however, their number is much lower as compared to the micelles. Water-soluble initiators are generally used to initiate the polymerization reaction. The initiator generates the radicals in the aqueous phase owing to thermal dissociation. The generated radicals have the possibility of entering either the micelles or the monomer droplets. However, experimental evidence proves the absence of droplet polymerization. The radicals do not enter the monomer droplets, as the radical entities are hydrophilic in nature whereas the monomer droplets are hydrophobic. Also, because the number of monomer droplets is much smaller than the number of micelles, it is micelles that capture the majority of the radicals. Also, the unique architecture of the micelles provides attractive conditions for the radicals to enter. Figure 1.2a shows the mechanism of the micellar nucleation for the generation of polymer particles. This mode of nucleation is also termed heterogeneous nucleation. The homogenous mode of particle nucleation is also possible when (a) the amount of surfactant is below its critical micelle concentration, (b) no surfactant is used during the polymerization, or (c) the monomer is signiicantly water soluble. In this mode of nucleation, the generated radicals in the aqueous phase start reacting with the dissolved monomer molecules. However, after adding a few monomer units in the chains, these chains no longer remain water soluble and come out of the solution. These chains are not stable on their own and keep collapsing with each other in order to attain stability. They also adsorb a certain amount of surfactant from either the micelles or the aqueous phase itself. Partial stability is also provided by the negative charges from the initiator moieties. In the case of the surfactant-free polymerization, the initiator charges are the only source of colloidal stability of the particles. Figure 1.2b shows the process of homogenous nucleation. The emulsion polymerization process is generally divided into three intervals. The irst is the particle formation interval. The radicals are generated in the aqueous phase after the thermal dissociation of the initiator. These radicals start entering the micelles and initiate polymerization. These active micelles where the polymerization starts to take place are then referred to as polymer particles. The number of particles in this interval keeps increasing owing to the continuous entry of the generated radicals in the micelles. This also leads to continuous increase in the rate of the polymerization. As the polymerization of the monomer in the particles proceeds, the size of the particles keeps increasing and the amount of monomer in the particles keeps depleting. However, this depletion of the monomer is replenished by the absorption of

Polymer Latex Technology: An Overview

5

(a)

(b) FIGURE 1.2 (a) Representation of micellar nucleation mechanism for the generation of poly mer particles. (b) Homogenous nucleation mechanism for the synthesis of poly mer particles.

the monomer from the aqueous phase. The aqueous phase in turn absorbs more monomer from the monomer droplets. Therefore, a mass transfer from the monomer droplets to the polymer articles keeps taking place during the course of polymerization. For this diffusion process to take place, the monomer is required to be partially soluble in water. As the polymer particles become bigger in size and their surface area increases as a function of time or

6

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

FIGURE 1.3 Schematic of various intervals of the emulsion poly merization process.

monomer conversion, they require more amount of surfactant to remain stable. The surfactant dissolved in the aqueous phase is continuously adsorbed on the surface of the polymer particles, leading to the reduction of the surfactant amount in the solution to lower than the critical micelle concentration. This in turn destabilizes the remaining micelles and these micelles disappear, providing their surfactant for the stabilization of the polymer particles. Thus at the end of the irst interval, no micelle is left and most of the surfactant is used to stabilize the polymer particles. It has to be noted that the inal number of polymer particles is much lower than the original number of micelles. Also, roughly 15% of the monomer is polymerized by the end of the irst interval [1]. Figure 1.3 represents the various intervals of emulsion polymerization. Once excess surfactant is no longer present in the system, no new particles nucleate. This marks the beginning of the second interval of emulsion poly mer i zation. Because no new particles nucleate, the amount of the particles remains almost constant; this also leads to an almost constant poly mer i zation rate in this interval. The size of the poly mer particles, however, keeps increasing as a function of conversion. The monomer present in the monomer droplets continues to replenish the monomer in the aqueous phase as well as monomer-swollen poly mer particles. After a certain extent of conversion, the monomer droplets also disappear. This also signals the start of the inal interval of the emulsion poly mer i zation process. The concentration of the monomer in the poly mer particles keeps decreasing, and as a result the rate of poly mer i zation also steadily decreases. Because the monomer is almost consumed, the poly mer i zation rate virtually falls to zero. Figure 1.4 shows the evolution of the particle size as a function of conversion.

7

Polymer Latex Technology: An Overview

300 nm

200 nm (a)

(b)

400 nm (c) FIGURE 1.4 (a–c) Increase of the size of the particles as a function of conversion.

Emulsion polymerization can also lead to the generation of various different surface morphologies of the polymer particles as well as particle sizes or families. Figure 1.5 shows the examples of monomodal, bimodal, or multimodal polymer particles along with morphologies like planar, orange-peel, strawberry or surface craters, etc. Polymers synthesized with emulsion polymerization are not always homopolymers, but most of the time are copolymers. When more than one monomer is polymerized together, the reactivity of the monomers deines the resulting morphology of the particles. Different reactivities of the monomer lead to totally different copolymer composition in the polymer particles, leading to a gradient in the concentration of the monomers with radius. This occurs because of the faster polymerization of the reactive monomer thus accumulating near the center of the particles, followed by the poly merization of the lesser reactive monomer, which then is present in a majority near the

8

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

300 nm

200 nm (a)

(b)

250 nm (c)

300 nm (d)

FIGURE 1.5 (a–d) Various morphologies of particles achieved with emulsion poly merization.

outer surface of the particles. Differences in the water solubilities of the monomers can also lead to the generation of speciic morphologies of the particles. As an example, in Figure 1.6 are shown the copolymer particles of styreneco-N-isopropylacrylamide synthesized by the surfactant-free approach, that is, by the homogenous nucleation method. N-isopropylacrylamide being hydrophilic in nature starts to polymerize irst, followed by the polymerization of more hydrophobic styrene. But because the polymer chains from styrene are also hydrophobic in nature, they push the hydrophilic chains of poly(N-isopropylacrylamide) away to the surface, leading to the morphology as shown in Figure 1.6.

9

Polymer Latex Technology: An Overview

300 nm

250 nm

(a)

(b)

300 nm (c) FIGURE 1.6 (a–c) Scanning electron microscopy (SEM) micrographs of copolymer particles of styrene-coN-isopropylacrylamide.

1.3 Controlled Polymerization and its Use in Emulsion Polymerization Processes The conventional radical polymerization is limited as a technique in that the control in the molecular weight or its distribution is dificult to achieve. It is also not easy to achieve well-deined morphologies in the particles like block copolymers because the life of the radical is too short, and uncontrolled termination reactions take place very fast. Controlled living polymerization techniques, on the other hand, can circumvent the aforementioned limitations in the emulsion polymerization process [4]. In these techniques, the chains

10

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TEMPO mediated nitroxide polymerization (NMP)

Atom transfer radical polymerization (ATRP)

Activator regenerated by elecron transfer (ARGET) atom transfer radical polymerization (ATRP)

SG1 mediated nitroxide polymerization (NMP) Degenerative transfer

Living polymerization

Activator generated by elecron transfer (AGET) atom transfer radical polymerization (ATRP)

Reversible additionfragmentation chain transfer (RAFT) polymerization

Reverse atom transfer radical polymerization (ATRP)

FIGURE 1.7 Representation of various living poly merization techniques. (Reprinted from V. Mittal, Advances in Polymer Latex Technology, New York: Nova Science Publishers, 2009. With permission.)

are terminated but only irreversibly, and after a short period of time become active again to propagate the polymer chains. In such processes, the termination reactions are effectively eliminated, and the controlled molecular weight distributions as well as advanced morphologies in the polymer particles can be achieved. There have been many techniques developed in the last years to achieve controlled polymerization, and these techniques are generally classiied into two categories: those based on reversible termination and those based on reversible transfer. Figure 1.7 is the representation of the various controlled polymerization techniques. In the category of reversible termination, nitroxide-mediated polymerization (NMP) and atom transfer radical polymerization (ATRP) are the most studied approaches. ATRP has also been further modiied into techniques like reverse atom transfer radical polymerization, activator generated by electron transfer ATRP, etc. In the category of reversible transfer, techniques like reversible addition-fragmentation chain transfer (RAFT) polymerization and degenerative transfer are mostly reported. During the polymerization, the concentration of dormant species continues to increase as compared to the active chains. At the end of polymerization, the dormant species may be present in amounts six times higher than the active chains. This effectively leads to elimination of termination and allows much longer lifetimes for the radicals.

Polymer Latex Technology: An Overview

11

NMP, where nitroxides are used to irreversibly terminate the polymer chains, has been used in two different ways. In one case, a conventional free radical initiator and a separately added nitroxide are added to control the polymerization. The two most commonly used nitroxides for this purpose are 2,2,6,6-tetramethyl-1-piperidinoxyl (TEMPO) and N-ter-butyl-1diethylphosphono-2,2’-dimethylpropyl (SG1). The nitroxides were initially developed for the polymerization of styrene; however, a number of other nitroxides have been developed that are also suitable for the polymerization of acrylates. In the other case, an alkoxyamine is used, the decomposition of which leads to the generation of two radicals: one reactive and one stable. This radical pair then controls the polymerization and thus does not require the addition of conventional free radical initiator. Figures 1.8 and 1.9 represent the polymerization of lauryl methacrylate and styrene by using nitroxides and alkoxyamines, respectively. The only disadvantage of the nitroxide-mediated stable free radical polymerization was the requirement of a high temperature for the polymerization reaction, which is sometimes not feasible for thermally sensitive systems; however, various nitroxides have been developed that also allow use at lower reaction temperatures. The initially carried-out reactions with styrene in emulsion led to poor colloidal stability, which resulted in a large amount of coagulum generated in the polymerization reactions. It was claimed that the particle nucleation as well as polymerization in droplets were a few reasons, among others, for this behavior. The seed method has also been described for emulsion poly merization with SG1 as nitroxide [7]. In this case, a seed is generated irst with low solid content, and the seed particles are then swollen in monomer and followed by subsequent polymerization of these seed particles. This helps to avoid the generation of monomer droplets and thus polymerization in droplets. It was also possible to achieve the previously described reaction as a single step. After the synthesis of seed as before, a certain amount of monomer is added without cooling the seed latex and the reaction is run until high conversion is achieved. The formed latexes were very stable in nature and no coagulum was generated. It is also important to monitor the progress of the reaction especially at high conversion, as at very high conversions, chains start to terminate each other and the polydispersity in the chain length as well as molecular weight increases. Therefore, it is always beneicial to stop the polymerization reaction a little below the full conversion. It was also conirmed that the alkoxyamines based on SG1 are more optimally operatable for achieving controlled polymerization of a wide range of monomers as compared to TEMPO nitroxide. ATRP represents another technique based on the principle of reversible termination, and in this process, an organic halide is used to irreversibly terminate the propagating chains. This technique has been very successful for the controlled polymerization of styrenics as well as acrylates and methacrylates. It also does not require very high temperatures as compared

12

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

O C

CH2

O O

O

CH C

+

C

O

OC12H25

O C

CH2

O

CH C

• CH

CH2 O

C

OC12H25

CH2

CH

+

O

O

C

OC12H25

OC12H25

CH3 H3C •O

OC2H5

CH3 N

CH

H3 C

P CH3

OC2H5 O

CH3 CH3 O C

H3 C O

CH2

CH C

CH2 O

OC12H25

CH C

O O

OC12H25

CH3 N

CH

H3C

OC2H5 P CH3

OC2H5 O

CH3

FIGURE 1.8 Nitroxide-mediated controlled radical poly merization of lauryl methacrylate with SG1 nitroxide.

to NMP, and in many cases can also be undertaken at room temperature. Figure 1.10a shows the schematic of the ATRP process. Cuprous salt forms a complex with ligand, L (amines of different chemical architectures), which makes it more soluble in the solvents [1]. Initiation of the reaction takes place by the dissociation of the halide atom from the initiator and leading to the generation of a free reactive radical. The bromide atom is captured by cuprous halide ligand complex and it forms CuBr2 ligand complex. This compound is very stable and hence is called deactivator. The generation of this compound thus leads to reduction in the concentration of the free radical species in the system. The growing radical continues to add monomer units to the polymer chain, and at some point it comes in contact with CuBr2

13

Polymer Latex Technology: An Overview

CH H3 C

H3 C

• CH

• CH

N

O

• O

+

H3C

CH3

CH2

CH3

H3 C

CH3

H3C

N H3C CH3

CH

+

CH2

CH

H3C

CH

H3 C

H3 C • O

CH3

N H3C CH3

CH3

H3 C CH

CH2

CH

N

O

H 3C

H3C

CH2

CH3

CH

n

H3 C CH H3C

CH2

CH

CH2 n

CH

O

CH3

N H3 C

CH3

FIGURE 1.9 Nitroxide-mediated free radical poly merization of styrene by using alkoxyamine as nitroxide as well as initiator.

ligand complex and is temporarily terminated by the formation of RMn+1-Br compound. It is also possible to carry out the reverse ATRP process similarly (Figure 1.10b). In this process, a conventional free radical initiator like AIBN or benzoyl peroxide is used to initiate the polymerization reaction, which is controlled by the addition of CuBr2 ligand complex. The radicals add few monomer units, and during this process come in contact with this complex to form the dormant species. One limitation of such a technique is the presence of transition metal in the inal particles, which though possible to wash off adds another processing step in the synthesis process. Another limitation

14

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

P

P• +

Br + CuBr/L

CuBr2/L

M PM• nM PM•n+1

RMn+1 – Br +

CuBr/L

(a) O C

O

O O

O

+

C

CuBr2/L

C

O

Br

M O C

O O

M•

nM

C

O

M•n+1

+

CuBr2/L

O C

O

M•n+1

Br

+

CuBr/L

(b) FIGURE 1.10 Schematic of (a) ATRP and (b) reverse ATRP processes for controlled living poly merization. (Adapted from G. Odian, Principles of Polymerization, Hoboken, NJ: John Wiley & Sons, 2004; and V. Mittal, Advances in Polymer Latex Technology, New York: Nova Science Publishers, 2009.)

is the reaction of the copper compounds with the other constituents of the system. One example is the reaction of these compounds with the emulsiier used in the polymerization system, leading to the poisoning of the initiator, which subsequently results in no or little polymerization. Therefore, it is possible to work in the emulsion with the ATRP when there is no surfactant or surfactants with no interaction with the initiator are chosen. ATRP in emulsion processes also faced problems similar to those in NMP. In one reported study, ethyl 2-bromoisobutyrate was used as an ATRP initiator, and copper bromide was complexed with 4,4’-dinonyl-2,2’-bipyridyl to form the catalyst system [8]. Nonionic surfactant Tween 85 was used. The polymerization was achieved by irst mixing together copper salts with 4,4’-dinonyl-2,2’-bipyridyl, to which the monomer was added. The solution was allowed to mix and was added with surfactant. To this solution, water was added under vigorous

15

Polymer Latex Technology: An Overview

stirring to form the emulsion. To this emulsion was then added the initiator to initiate the polymerization reaction. The reaction conditions needed to be very accurately controlled, and the reaction was overall very sensitive to minor changes in the reaction parameters. Seeded polymerization similar to that used with NMP was also used in this case. The use of cationic surfactants was also reported for the ATRP processes in emulsion. Dodecyl trimethyl ammonium bromide and myristyl trimethyl ammonium bromide were used as cationic surfactants, and their effect on the latex stability, amount of coagulum, and the polydispersity in the molecular weight was quantiied. The irst surfactant though gave a good control of polydispersity; however, the whole system was observed to coagulate after the initiation of polymerization. In the case of the second surfactant, the latex stability was better, but the polydispersity in the molecular weight or chain lengths was very high. RAFT is a controlled polymerization technique based on the principle of reversible transfer. The core of this process is a RAFT agent that contains dithioester groups. The living polymerization takes place because the transferred end group in the polymeric dithioester is as labile as the dithioester group in the starting RAFT agent. The initiator for the polymerization can be the conventional initiators like AIBN or benzoyl peroxide. Figure 1.11 explains the principle of RAFT polymerization. Though it is one of the versatile techniques for the polymerization of a large number of monomers, it also has its own limitations, such as the presence of remainder RAFT agent and the commercial unavailability of the RAFT agents. Similar to NMP and ATRP, initial trials with RAFT also were faced with dificulties of coagulum generation. The RAFT agent was dificult to be transported to the polymer particles through the aqueous phase. RAFT polymerization of styrene in emulsion was reported by Szkurhan et al. [9]. The process named nanoprecipitation was carried out by forming nano-sized particles by precipitation of the acetone solution of macro RAFT agent in the aqueous poly(vinyl alcohol) solution. The macro RAFT agent was prepared by conventional free R1

Mn + S

SR2

MnS

S



SR2

SMn

MmS



+ R•2 S

MnS R1

R1

R1

M•m +

R1

R1

SMn

MmS

+ M•n S

FIGURE 1.11 Mechanism of reversible transfer processes used in reversible addition-fragmentation chain transfer processes. (Adapted from G. Odian, Principles of Polymerization, Hoboken, NJ: John Wiley & Sons, 2004; and V. Mittal, Advances in Polymer Latex Technology, New York: Nova Science Publishers, 2009.)

16

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

radical polymerization. The formed nanoparticles were subsequently swollen with monomer and were polymerized in the living manner. This nanoprecipitation method was also named seeded polymerization because, in this case, the nano-sized particles formed by precipitation act as seeds to form polymer particles. Both water- and oil-soluble initiators were used. When the initiator used was oil soluble, it was premixed with the RAFT agent, whereas the water-soluble initiator was dissolved in PVA solution. With both watersoluble and oil-soluble initiators, the rate of polymerization was quite slow, and increasing the reaction temperature was not helpful in increasing the rate of polymerization. Another study reported the synthesis of polymer particles in emulsion by RAFT without the problems of loss of colloidal stability and the molecular weight control [10]. Trithiocarbonate RAFT agents were used in the study to form short stabilizing blocks from a hydrophilic monomer, from which diblocks were created by the subsequent polymerization of a hydrophobic monomer. These diblocks self-assembled to form micelles, and subsequent polymerization could be carried out. Figure 1.12 demonstrates the particles generated by this technique.

1.4 Conventional and Controlled Miniemulsion Polymerization In surfactant-aided emulsion polymerization, the goal is to achieve the micellar nucleation and to avoid the droplet nucleation as much as possible. But the polymerization of extremely hydrophobic monomers by conventional emulsion polymerization is not possible because of their inability to diffuse from the monomer droplets through the aqueous phase to the polymer particles. To achieve polymerization of such systems, miniemulsion polymerization has proved to be a versatile method [2,3]. The mode of polymerization is based on the droplet polymerization principle. The monomer droplets are generated by shearing the system with high energy along with the addition of costabilizer (with the surfactant), which needs to be hydrophobic in order to avoid the collapse of the monomer droplets by Ostwald ripening when the shearing of the system is stopped. Thus in this mode of polymerization, it is important to avoid the micellar nucleation; therefore, the amount of surfactant is below the critical micelle concentration. The particles in the general size range of 50–500 nm can be synthesized by using miniemulsion polymerization. These are similar in size to the monomer droplets in the beginning of the polymerization. The initiators used for the polymerization are water soluble as in the case of emulsion polymerization. The initiator on dissociation generates radicals in the aqueous phase, and these radicals enter the droplets and initiate polymerization. In conventional emulsion polymerization, also called macroemulsion polymerization, the micellar nucleation is very sensitive and is affected by a large number of factors like surfactant amount, initiator amount,

17

Polymer Latex Technology: An Overview

Conventional core-shell: different chains in the core and shell

Triblock polymer: the same chains extend from the shell into the core (a)

0.20 µm X 19000

(b) FIGURE 1.12 (a) Schematic representation of RAFT-based triblock copolymer chains forming the core and shell of the particles and their comparison with conventional core-shell particles. (b) TEM micrograph of poly(acrylic acid)-b-poly(butyl acrylate)-b-polystyrene particles. (Reprinted from C. J. Ferguson, R. J. Hughes, D. Nguyen, T. T. Pham, R. G. Gilbert, A. K. Serelis, C. H. Such, and B. S. Hawkett, Macromolecules 38: 2191–2204, 2005. With permission from the American Chemical Society.)

18

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

FIGURE 1.13 Schematic of the miniemulsion poly merization process. The molecules with black and gray color represent the surfactant and costabilizer, respectively.

agitation, reaction temperature, etc. However, that is not the case in miniemulsion polymerization. Figure 1.13 also represents the scheme of polymerization in miniemulsion. The monomer is added to the surfactant and costabilizer in the aqueous phase followed by homogenization under high shear to break the bigger monomer droplets into droplets with the size range of 10–500 nm [2,3,11]. The amount of the surfactants and the shearing translate into the size of the monomer droplets. As mentioned earlier, in emulsion polymerization, the polymerization is driven by the diffusion of the monomer through the aqueous phase. In this process, monomer droplets disappear and the micelles convert into the polymer particles, whereas in miniemulsion polymerization, the monomer droplets directly translate into polymer particles. As a result, the rate of polymerization is also different for these two processes. The rate of polymerization in the emulsion polymerization irst increases owing to the generation of the particles and reaches a constant phase after the disappearance of the micelles. The rate then decreases owing to the depletion of the monomer in the particles. As there is no diffusion of monomer in the miniemulsion polymerization, the constant rate period is absent. The rate irst increases owing to the nucleation of the particles and then decreases when the monomer is consumed in the particles. Not only hydrophobic monomers, but also extremely hydrophilic monomers can be polymerized using the miniemulsion polymerization method. In this case, one has to use the inverse miniemulsion polymerization. Also in this case, hydrophobic reaction medium is used along with a lipophobe used as costabilizer. One must be clear that the addition of costabilizer stops the conversion of a miniemulsion into a conventional emulsion; however, the addition of a costabilizer to conventional emulsion does not automatically convert it into a miniemulsion. It is only after the addition of high shearing energy that it becomes a stable miniemulsion. The costabilizers should be hydrophobic, soluble in monomer, and have a low molecular weight. However, the use of conventional costabilizers like cetyl alcohol and hexadecane pose a potential hazard owing to their volatility, and the presence of these even in the minor amounts in the polymer particles may not be acceptable for many applications. Therefore, a lot of research effort has been focused in the direction of generation of more compatible costabilizers. Polymeric stabilizers from the polymer of the monomer

Polymer Latex Technology: An Overview

19

to be polymerized have been used in some of the reported studies [3,12,13]. These polymers are soluble in their own monomers and thus allow better intermixing at the interphase. They also eliminate the use of volatile costabilizers, providing more acceptability to the system for the commercial applications. Monomeric costabilizers have also been developed in recent years that can copolymerize with the monomer under polymerization [14]. These costabilizers form copolymer chains that are bound inside the particles, and thus the possibility of the diffusion of low molecular weight components out of the particles is eliminated. The potential diffusion of the low molecular weight components, especially substances like cetyl alcohol or hexadecane, can pose health hazards when the polymer particles are used in application with food contact. Similarly, the other components of the polymerization system like initiator and chain transfer agent have also been used as costabilizers [15,16]. Therefore, the materials act as dual-role components. They not only perform their function as initiator or chain transfer agent, respectively, but also help to achieve the stability for the monomer droplets and polymer particles. Miniemulsion polymerization has also been proved to be advantageous in living polymerization systems. The various living polymerization methods like NMP, ATRP, and RAFT have been shown to be beneicial in miniemulsion polymerization to generate specialty polymers or polymers with special architecture like block copolymers. Colloidal stability and ease of poly merization process are also better in the case of miniemulsion poly merization than conventional emulsion polymerization. For NMP, nitroxide-capped polymer chains have been used as initiator as well as nitroxide. The use of such nitroxide-capped polymer chains for the initiation and reaction control allows one to properly estimate the number of starting chains in the system, helping to achieve better control of the molecular weight. The use of nitroxide-capped poly mer also helps to partition the nitroxide solely in the organic phase owing to the hydrophobicity. In such reported studies with polystyrene terminated with TEMPO as initiator as well as nitroxide, hexadecane as costabilizer, and DOWFAX 8390 as surfactant, it was reported that changing the amount of surfactant led to the generation of different particle sizes, but the rate of polymerization was not affected, which is different from the behavior seen in conventional emulsion polymerization [17,18]. Figure 1.14 demonstrates the transmission electron microscopy (TEM) images of the polystyrene particles generated in nitroxide-mediated miniemulsion polymerization using different amounts of TEMPO-terminated oligomers of polystyrene as macroinitiator. A great deal of process development has been reported for ATRP in miniemulsion. A number of studies have been reported that apply the direct or forward ATRP in miniemulsion, but reverse ATRP, in which conventional free radical polymerization initiator like AIBN can be used with the transition metal compound in its higher oxidation state, was observed to be more suitable for miniemulsion polymerization. This eliminates the use of air-sensitive Cu(I) species and requires only the use of Cu(II) species, which

20

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(a)

(b)

FIGURE 1.14 TEM images of the polystyrene particles prepared by nitroxide-mediated miniemulsion polymerization using different amounts of TEMPO-terminated oligomers of polystyrene (TTOPS) as macroinitiator. (a) 5% TTOPS and (b) 20% TTOPS (100 nm). (Reprinted from G. Pan, E. D. Sudol, V. L. Dimonie, and M. S. El-Aasser, Macromolecules 34: 481–88, 2001. With permission from the American Chemical Society.)

is more stable in air. A narrow molecular weight distribution as well as linear increase in the molecular weight as a function of conversion was reported, and the inal latexes were stable over a period of time. In one such study on reverse ATRP processes [19], Brij 98 surfactant and CuBr2/dNbpy (4,4’-di[5nonyl]-4,4’-bipyridine) complex were used along with hexadecane costabilizer. Both water-soluble as well as oil-soluble initiators were used for the polymerization. It was observed that the polymerization rate was independent of the size and number of particles and the amount of surfactant. The shear forces were able to inluence only the size of the particles and not the polymerization rate. In the case of oil-soluble initiator, the polymerization was observed to proceed by droplet nucleation mode because the monomersoluble initiator was already present in the droplet during the miniemulsion; whereas when water-soluble initiator was used, both micellar as well as droplet nucleation were reported to take place. Similarly RAFT has also been used successfully in miniemulsion for the polymerization of styrene as well as water-soluble monomers like acrylamide [11,20–22].

1.5 Generation of Copolymer or Core-Shell Particles The generation of well-deined copolymer morphologies like block copolymer particles is dificult to achieve by the use of conventional emulsion polymerization because of the uncontrolled free radical polymerization and the short life of the radicals. To achieve certain control on the copolymerization, monomer reactivity ratios and monomer-feeding methodology must be considered.

21

Polymer Latex Technology: An Overview

300 nm

300 nm

(a)

(b)

300 nm

300 nm (c)

(d)

FIGURE 1.15 (a–d) Different morphologies of copolymer particles generated by conventional emulsion polymerization.

The monomers have different reactivity ratios; therefore, if the monomers are added together, the more reactive monomer starts to polymerize irst followed by the polymerization of the less reactive monomer. This creates a gradient of concentration of the monomer units in the polymer particles as a function of radius. Figure 1.15 provides some examples of various copolymer particles that can be achieved by emulsion polymerization like core-shell grafted particles, core-shell particles with hydrophilic shell and hydrophobic core, copolymer particles with different surface morphologies, etc. Apart from reactivity ratios, mode of addition of the monomers during the course of polymerization is also of utmost importance to achieve control on the particle characteristics. Batch addition of the monomers does not lead to generation of the structured latexes; therefore, semibatch addition of the monomers is generally preferred. This mode of addition can be achieved by either looded

22

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

addition or starved addition of the monomers. In looded addition, monomers are added at a rate higher than their rate of consumption. This mode of addition leads to buildup of the monomers in the polymerization reaction and may lead to the generation of secondary nucleation of the particles. Starved addition of the monomers, on the other hand, is the addition of monomers at a rate slower than their polymerization rate, and this allows one to retain the chemical composition of the polymer chains equal to the monomer ratios in the feed or according to the requirement. The starved conditions eliminate the possibility of secondary nucleation, though one has to be careful about the control of the amount of the surfactant in the system as well as charges on the surface. As an example, if copolymerization of a hydrophilic monomer and a hydrophobic monomer is considered, initially the surface may be hydrophobic, but as the chains rich in hydrophilic monomer content get pushed out to the surface of the particles during the course of polymerization, the surface becomes hydrophilic. This would lead to a change in the surface properties of the particles and would allow the release of the surfactant from the surface of the particles owing to the hydrophilicity. This also results in the nonentry of the hydrophobic monomer into the polymer particles, causing monomer concentration drift in particles or secondary nucleation. Interesting studies have been reported for the generation of copolymer particle latexes by emulsion polymerization. Batch polymerization of copolymer particles of polystyrene and poly(methyl methacrylate) were reported without the use of initiators [23]. These copolymer particles of poly(methyl methacrylateco-styrene) were prepared by thermally initiated emulsion copolymerization. It was observed that totally different particle morphologies like hemispherical, sandwich-like, core-shell, inverted core-shell particle morphologies, etc. were obtained depending on the polymerization conditions. It was reported that the incorporation of the initiator fragments to one end of the chains allows the polystyrene chains to become more hydrophilic, changing the surface nature of the polymer particles. In another study to generate copolymer particles, poly(methyl methacrylate) seed was used to generate the copolymer particles of poly(methyl methacrylate) with polystyrene [24]. It was observed that by using the oil-soluble initiators, an inverted core-shell morphology of the particles was obtained, in which the polystyrene chains were present in the core of the particles and the poly(methyl methacrylate) covered the particles owing to its hydrophilicity. In the case of water-soluble initiator, the morphology was less affected by the hydrophilicity of the polymers, but was more affected by the initiator concentration and polymerization temperature. Controlled living polymerization methods provide much better possibilities to generate the structured latex particles with different morphologies of chemistries owing to the prolonging of the radical age either by reversible termination or by reversible transfer. In one such study to generate triblock copolymers using emulsion polymerization, water-soluble SG1based bifunctional alkoxyamine (Figure 1.16; sodium salt of alkoxyamine was used owing to water solubility) and Dowfax 8390 surfactant were used

23

Polymer Latex Technology: An Overview

O N HOOC

O P O O

O CH2 CH2 O 3

O

O

t-butanol T = 80 – 100°C

HOOC

O O P

COOH

O CH2 CH2 O

N O

O N

3

O

O

O

O

(1) NaOH (2) R

R = Ph, COOBu

Emulsion T = 112°C

COO– +Na

Na+ –OOC R O O P O

N O

O P O

R O CH2 CH2 O

x

3

O

x

O

O N

O P O O

FIGURE 1.16 Synthesis and use of SG1-based water-soluble bifunctional alkoxyamine. (Reprinted from J. Nicolas, B. Charleux, O. Guerret, and S. Magnet, Macromolecules 38: 9963–73, 2005. With permission from the American Chemical Society.)

[11,25]. When a bifunctional alkoxyamine was used, two functional ends of this alkoxyamine could be used to generate triblock copolymers. Thus, in order to generate polystyrene-b-poly(butyl acrylate)-b-polystyrene triblock copolymer particles, a seed was irst generated from butyl acrylate particles. The seed was further swollen with butyl acrylate to form central poly(butyl acrylate) block in the emulsion particles. The particles were then added with styrene to form two blocks of styrene around the central poly(butyl acrylate) block to form the triblock copolymer. In another study using ATRP, the seeded-polymerization approach was used to synthesize block copolymers of poly(i-butyl methacrylate) and polystyrene using ethyl 2-bromoisobutyrate as initiator and CuBr/4,4’-dinonyl-2,2’-dipyridinyl as catalyst ligand

24

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

800 nm

300 nm (a)

(b)

100 nm (c)

250 nm (d)

FIGURE 1.17 (a, b) SEM and TEM micrographs of latex particles functionalized with an ATRP initiator. (c, d) The grafted brushes of thermally responsive poly mer poly(N-isopropylacrylamide) from the surface of the ATRP initiator functionalized particles.

complex [26]. Tween 80 (polyoxyethylene sorbitan monooleate) was used as surfactant. First a seed of poly(i-butyl methacrylate) end-capped with ATRP initiator was prepared to which a batch of styrene then was added to form a block copolymer. Thermally responsive polymer particles were also reported by the use of ATRP [27–29]. The process included the synthesis of seed particles, functionalization of seed particles by ATRP initiator, and the grafting of the poly(N-isopropylacrylamide) chains from the surface. Figure 1.17 shows these functional particles [28]. In another study using RAFT polymerization, core-shell functional particles were achieved by using o-ethylxanthyl

Polymer Latex Technology: An Overview

25

pTA/RuO4 cryo

Observation at –150°C

FIGURE 1.18 Core-shell particles of polystyrene-b-poly(butyl acrylate)/poly(acetoacetoxy ethyl methacrylate). The black core represents polystyrene whereas the soft shell is poly(butyl acrylate)/ poly(acetoacetoxy ethyl methacrylate) component. (Reprinted from M. J. Monteiro and J. de Barbeyrac, Macromolecules 34: 4416–23, 2001. With permission from the American Chemical Society.)

ethyl propionate as RAFT agent [30]. The polymerization reactions were carried out in the presence of poly(methyl methacrylate) seed particles of predetermined number and size distribution. The seed was added irst with styrene monomer to form polystyrene block, which was then added with butyl acrylate. Styrene was added in batch conditions, whereas butyl acrylate was added slowly to the emulsion system so as to avoid the buildup of high concentration of monomer in this system. Similarly, core-shell particles consisting of block copolymer of polystyrene-b-poly/butyl acrylate)/ poly(acetoacetoxy ethyl methacrylate) were also prepared by using xanthates as RAFT agents. Figure 1.18 shows the TEM image of such core-shell particles [31]. Miniemulsion polymerization has also been extensively used for the synthesis of functional latex particles [32–38].

References 1. Odian, G. 2004. Principles of polymerization. Hoboken, NJ: John Wiley & Sons. 2. Landfester, K. 2001. Polyreactions in miniemulsions. Macromolecular Rapid Communications 22:896–936.

26

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

3. Schork, F. J., Luo, Y., Smulders, W., Russum, J. P., Butté, A., and K. Fontenot. 2005. Miniemulsion polymerization. Advances in Polymer Science 175:129–255. 4. Matyjaszewski, K., and T. P. Davis. 2002. Handbook of radical polymerization. Hoboken, NJ: John Wiley & Sons. 5. Hiemenz, P. C., and R. Rajagopalan. 1997. Principles of colloid and surface chemistry. New York: Marcel Dekker. 6. Mittal, V. 2009. Advances in polymer latex technology. New York: Nova Science Publishers. 7. Nicolas, J., Charleux, B., and S. Magnet. 2006. Multistep and semibatch nitroxide-mediated controlled free-radical emulsion polymerization: A signiicant step toward conceivable industrial processes. Journal of Polymer Science, Part A: Polymer Chemistry 44:4142–53. 8. Eslami, H., and S. Zhu. 2005. Emulsion atom transfer radical polymerization of 2-ethylhexyl methacrylate. Polymer 46:5484–93. 9. Szkurhan, A. R., Kasahara, T., and M. K. Georges. 2006. Reversible-addition fragmentation chain transfer radical emulsion polymerization by a nanoprecipitation process. Journal of Polymer Science, Part A: Polymer Chemistry 44:5708–18. 10. Ferguson, C. J., Hughes, R. J., Nguyen, D., Pham, B. T. T., Gilbert, R. G., Serelis, A. K., Such, C. H., and B. S. Hawkett. 2005. Ab initio emulsion polymerization by RAFT-controlled self-assembly. Macromolecules 38:2191–2204. 11. Cunningham, M. F. 2008. Controlled/living radical polymerization in aqueous dispersed systems. Progress in Polymer Science 33:365–98. 12. Reimers, J. L., and F. J. Schork. 1996. Predominant droplet nucleation in emulsion polymerization. Journal of Applied Polymer Science 60:251–62. 13. Reimers, J., and F. J. Schork. 1996. Robust nucleation in polymer-stabilized miniemulsion polymerization. Journal of Applied Polymer Science 59:1833–41. 14. Reimers, J. L., and F. J. Schork. 1996. Miniemulsion copolymerization using water-insoluble comonomers as cosurfactants. Polymer Reaction Engineering 4:135–52. 15. Reimers, J. L., and F. J. Schork. 1997. Lauroyl peroxide as cosurfactant in miniemulsion polymerization. Industrial Engineering Research 36:1085–87. 16. Mouran, D., Reimers, J., and F. J. Schork. 1996. Miniemulsion polymerization of methyl methacrylate with dodecyl mercaptan as cosurfactant. Journal of Polymer Science, Part A: Polymer Chemistry 34:1073–81. 17. Pan, G., Sudol, E. D., Dimonie, V. L., and M. S. El-Aasser. 2002. Surfactant concentration effects on nitroxide-mediated living free radical miniemulsion polymerization of styrene. Macromolecules 35:6915–19. 18. Pan, G., Sudol, E. D., Dimonie, V. L., and M. S. El-Aasser. 2001. Nitroxide-mediated living free radical miniemulsion polymerization of styrene. Macromolecules 34:481–88. 19. Matyajaszewski, K., Qiu, J., Tsarevsky, N. V., and B. Charleux. 2000. Atom transfer radical polymerization of n-butyl methacrylate in an aqueous dispersed system: A miniemulsion approach. Journal of Polymer Science, Part A: Polymer Chemistry 38:4724–34. 20. Moad, G., Chiefari, J., Chong, Y. K., Krstina, J., Mayadunne, R. T. A., Postma, A., Rizzardo, E., and S. H. Thang. 2002. Living free radical polymerization with reversible addition-fragmentation chain transfer (the life of RAFT). Polymer International 49:993–1001.

Polymer Latex Technology: An Overview

27

21. Butté, A., Storti, G., and M. Morbidelli. 2001. Miniemulsion living free radical polymerization by RAFT. Macromolecules 34:5885–96. 22. Qi, G., Jones, C. W., and F. J. Schork. 2007. RAFT inverse miniemulsion polymerization of acrylamide. Macromolecular Rapid Communications 28:1010–16. 23. Du, Y. Z., Ma, G. H., Ni, H. M., Nagai, M., and S. Omi. 2002. Morphological studies in thermally initiated emulsion (co)polymerization without conventional initiators. Journal of Applied Polymer Science 84:1737–48. 24. Cho, I., and K. W. Lee. 1985. Morphology of latex particles formed by poly(methyl methacrylate)-seeded emulsion polymerization of styrene. Journal of Applied Polymer Science 30:1903–26. 25. Nicolas, J., Charleux, B., Guerret, O., and S. Magnet. 2005. Nitroxide-mediated controlled free-radical emulsion polymerization using a difunctional watersoluble alkoxyamine initiator. Toward the control of particle size, particle size distribution, and the synthesis of triblock copolymers. Macromolecules 38:9963–73. 26. Okubo, M., Minami, H., and J. Zhou. 2004. Preparation of block copolymer by atom transfer radical seeded emulsion polymerization. Colloid and Polymer Science 282:747–52. 27. Mittal, V., Matsko, N. B., Butté, A., and M. Morbidelli. 2007. Functionalized polystyrene latex particles as substrates for ATRP: Surface and colloidal characterization. Polymer 48:2806–17. 28. Mittal, V., Matsko, N. B., Butté, A., and M. Morbidelli. 2007. Synthesis of temperature responsive polymer brushes from polystyrene latex particles functionalized with ATRP initiator. European Polymer Journal 43:4868–81. 29. Mittal, V., Matsko, N. B., Butté, A., and M. Morbidelli. 2008. Swelling deswelling behavior of PS-PNIPAAM copolymer particles and PNIPAAM brushes grafted from polystyrene particles & monoliths. Macromolecular Materials and Engineering 293:491–502. 30. Smulders, W., and M. J. Monteiro. 2004. Seeded emulsion polymerization of block copolymer core-shell nanoparticles with controlled particle size and molecular weight distribution using xanthate-based RAFT polymerization. Macromolecules 37:4474–83. 31. Monteiro, M. J., and J. de Barbeyrac. 2001. Free-radical polymerization of styrene in emulsion using a reversible addition-fragmentation chain transfer agent with a low transfer constant: Effect on rate, particle size, and molecular weight. Macromolecules 34:4416–23. 32. Farcet, C., and B. Charleux. 2002. Nitroxide-mediated miniemulsion polymerization of n-butyl acrylate: Synthesis of controlled homopolymers and gradient copolymers with styrene. Macromolecular Symposia 182:249–60. 33. Tortosa, K., Smith, J.-A., and M. F. Cunningham. 2001. Synthesis of polystyrene-block-poly(butyl acrylate) copolymers using nitroxide-mediated living radical polymerization in miniemulsion. Macromolecular Rapid Communications 22:957–61. 34. Keoshkerian, B., MacLeod, P. J., and M. K. Georges. 2001. Block copolymer synthesis by a miniemulsion stable free radical polymerization process. Macromolecules 34:3594–99. 35. Li, M., Jahed, N. M., Min, K., and K. Matyjaszewski. 2004. Preparation of linear and star-shaped block copolymers by ATRP using simultaneous reverse and normal initiation process in bulk and miniemulsion. Macromolecules 37:2434–41.

28

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

36. Min, K., Li, M., and K. Matyjaszewski. 2005. Preparation of gradient copolymers via ATRP using a simultaneous reverse and normal initiation process. I. Spontaneous gradient. Journal of Polymer Science, Part A: Polymer Chemistry 43:3616–22. 37. Min, K., Gao, H., and K. Matyjaszewski. 2005. Preparation of homopolymers and block copolymers in miniemulsion by ATRP using activators generated by electron transfer (AGET). Journal of the American Chemical Society 127:3825–30. 38. Luo, Y., and X. Liu. 2004. Reversible addition-fragmentation transfer (RAFT) copolymerization of methyl methacrylate and styrene in miniemulsion. Journal of Polymer Science, Part A: Polymer Chemistry 42:6248–58.

2 Synthesis of Polymer Particles with Core-Shell Morphologies Claudia Sayer and Pedro Henrique Hermes de Araújo CONTENTS 2.1 Introduction .................................................................................................. 29 2.2 Equilibrium and Nonequilibrium Morphologies ...................................30 2.2.1 Equilibrium Morphologies ............................................................. 31 2.2.2 Nonequilibrium Morphologies ...................................................... 33 2.3 Synthesis of Core-Shell Particles ............................................................... 35 2.3.1 Emulsion Polymerization ............................................................... 35 2.3.1.1 Synthesis of Core-Shell Particles (CS) ............................ 37 2.3.1.2 Synthesis of Inverted Core-Shell Particles (ICS) ........... 41 2.3.2 Miniemulsion Polymerization .......................................................43 2.3.3 Microemulsion Polymerization ..................................................... 47 2.3.4 Dispersion Polymerization ............................................................. 47 2.3.5 Suspension Polymerization ............................................................ 48 2.3.6 Other Techniques ............................................................................. 49 2.4 Characterization of Core-Shell Particles................................................... 52 2.4.1 Transmission Electron Microscopy ............................................... 52 2.4.2 Scanning Electron Microscopy ......................................................54 2.4.3 Atomic Force Microscopy ...............................................................54 2.4.4 Additional Techniques Used for Particle Characterization .......54 References............................................................................................................... 55

2.1 Introduction Core-shell polymer particles have received a great deal of industrial and academic interest in the last decades. These multicomponent particles with controlled morphology create a versatile class of materials in which the inal properties depend not only on the composition of each polymer phase but also on the morphology of these particles. This characteristic opens the

29

30

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

possibility for tailor-made properties for each application as, for instance, soft core–hard shell results in particles suitable to act as impact modiiers, and hard core–soft shell latex particles result in paints with low ilm formation temperature. In addition, via core-shell polymerization, it is also possible to get incompatible polymers into one particle or to add functionality either into the core or into the shell (Koskinen and Wilén 2009). Structured particles can be obtained with different morphologies: welldeined core-shell structure, inverted core-shell, interface with a gradient of both core and shell, interface with microclusters, and multiple or irregularly shaped shells (Sundberg and Durant 2003). The inal morphology depends on both thermodynamic and kinetic aspects, as quite frequently the equilibrium morphology may not be achieved due to kinetic control of the morphology development. The poly mer i zation techniques play a major role in the particle size as well as in the kinetic control of the poly mer i zation. Several heterogeneous poly mer i zation techniques such as emulsion, miniemulsion, microemulsion, dispersion, and suspension poly mer i zations could be employed to obtain poly mer particles with core-shell structures. The irst three techniques lead to the formation of submicrometric particles (10–800 nm), whereas the two last are used, respectively, to prepare small (1–30 µm) and large (50–1500 µm) micrometric particles. The end use properties of the structured particles depend on the design and control of particle morphology; therefore, it is necessary to understand how this morphology can be controlled and which are the main features of each poly mer i zation technique related to particle morphology control. The purpose of this chapter is to describe the factors that will lead to a certain particle morphology and to discuss the heterogeneous polymerization techniques that could be employed to obtain those particles. Section 2.2 introduces the basics of equilibrium and nonequilibrium morphologies. Section 2.3 deals with the different heterogeneous polymerization techniques and the main features related to morphology control. Section 2.4 discusses briely the characterization techniques, as any work in this area relies on the need to adequately characterize particle morphology.

2.2 Equilibrium and Nonequilibrium Morphologies The formation of core-shell particles is a challenging issue of poly mer reaction engineering in dispersed media. In principle, the most stable particle morphology is determined by thermodynamics according to the minimum interfacial energy (González-Ortiz and Asua 1995), as given by Equation (2.1):

31

Synthesis of Polymer Particles with Core-Shell Morphologies

3

1 φ= 2

3

∑∑ a σ ij

ij

(2.1)

i=1 j=1 j≠i

where aij and σij are, respectively, the interfacial area and the interfacial tension between phases i and j. Nevertheless, frequently the equilibrium morphology is not achieved since particle morphology depends on the interplay between thermodynamics, which establishes the equilibrium morphology, and kinetics. If kinetic control prevails, nonequilibrium-type (metastable or kinetically stable) structures may be formed. In the following paragraphs, equilibrium morphologies and the main factors that establish these morphologies will be discussed, followed by nonequilibrium ones. 2.2.1 Equilibrium Morphologies Figure 2.1 shows four different equilibrium morphologies for a twocomponent system based on the polymer-polymer and polymer-aqueous phase interfacial tensions, which determine the interfacial energy and, consequently, also the equilibrium morphology for a given system. The equilibrium morphologies are: • Core-shell (CS), in which the second-stage polymer 2 (●) forms a continuous shell around the seed polymer 1 (○) dispersed in the aqueous phase 3. This equilibrium morphology might be achieved when either (a) polymer 1 is more hydrophobic than polymer 2 (σ13 > σ23) and polymer 2 has more afinity with polymer 1 than with the aqueous phase (σ12 < σ23) or (b) polymer 1 has more afinity with polymer 2 than with the aqueous phase (σ12 < σ13) and polymer 2 has less afinity with polymer 1 than with the aqueous phase (σ12 > σ23). • Inverted core-shell (ICS), in which the seed polymer 1 forms a continuous shell around the second-stage polymer 2. This equilibrium morphology might be obtained when polymer 2 is more hydrophobic than polymer 1 (σ23 > σ13) and polymer 2 has more afinity with polymer 1 than with the aqueous phase (σ12 < σ23). • Hemisphere, snowman-like, Janus, half-moon, occluded, partially engulfed, depending on the different degrees of protrusion and on the different curvatures of the polymer/polymer interface and the coverage of one polymer upon the other (Sundberg and Durant 2003). This equilibrium morphology might be achieved under different conditions: (a) when polymer 2 has similar afinities with polymer 1 and with the aqueous phase (σ12 ≈ σ23); (b) when polymer 2 has more





○●

32

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(σ23 – σ12) σ13

10

1

0.1 0.1

1

σ12 σ23

10

FIGURE 2.1 Equilibrium morphologies for a two-component system: ○ poly mer 1 (seed), ● poly mer 2 (produced by the poly merization of the second-stage monomer). σ12: interfacial tension between polymers 1 and 2; σ13: interfacial tension between poly mer 1 and aqueous phase; σ23: interfacial tension between poly mer 2 and aqueous phase. (Reprinted from V. Herrera, R. Pirri, J. R. Leiza, and J. M. Asua, Macromolecules 39: 6969–74, 2006. With permission.)

○●

afinity with polymer 1 than with the aqueous phase (σ12 < σ23) and both polymers have similar afinities with the aqueous phase (ǀσ23 – σ12ǀ ≤ σ13); (c) when polymer 2 has less afinity with polymer 1 than with the aqueous phase (σ12 > σ23) and polymer 1 has similar afinities with polymer 2 and the aqueous phase (ǀσ23 – σ12ǀ ≤ σ13). • Separate particles of the different polymers. This equilibrium morphology might be obtained when both polymers have more afinity with the aqueous phase than with each other (σ13 < σ12) and (σ23 < σ12). As shown in Figure 2.1, equilibrium morphologies of two-component systems are established basically by three interfacial tension values. Notwithstanding, several factors inluence these three interfacial tensions and may, therefore, be used for particle morphology control purposes. Besides polymer types, which determine the interfacial tension between the polymers and affect the interfacial tension between each polymer and the aqueous phase, polymer-polymer interfacial tension can be inluenced substantially by compatibilizing agents, and the interfacial tensions between each polymer and the aqueous phase are also affected by the types and amounts of surfactant and initiator. In addition, it has been shown that the equilibrium morphology may also be affected by the level of cross-linking of polymer 1, which inluences the free energy (Sundberg and Durant 2003), and by the molecular weights of the polymers, since the interfacial tension between the polymer phases, which is much lower than that between the polymer and the aqueous phase, depends on the molecular weight (Tanaka

Synthesis of Polymer Particles with Core-Shell Morphologies

33

et al. 2008). Finally, the ratio between polymers 1 and 2 may affect the curvature of the polymer-polymer interface in the hemisphere morphology (Sundberg and Durant 2003), and the CS and ICS morphologies may not be achieved if the amount of the shell forming polymer (polymer 2) in CS and polymer 1 in ICS is not enough to form a continuous shell with a minimum thickness. If copolymers are to be considered, the analysis becomes more complex since the surface and interfacial tensions depend on the copolymer compositions. 2.2.2 Nonequilibrium Morphologies As mentioned at the beginning of this section, quite frequently the equilibrium morphology is not achieved due to kinetic control of the morphology development. In this case, nonequilibrium-type structures may be formed. Three main processes have been used by González-Ortiz and Asua (1995, 1996a, 1996b) to describe the morphology development: 1. The formation of polymer chains occurs at a given position in the polymer particle. 2. Incompatible polymer chains cause phase separation leading to the formation of clusters. 3. Clusters migrate toward the equilibrium morphology in order to minimize the Gibbs free energy. During this migration the size of the clusters may increase by (i) polymerization of monomer inside the cluster, (ii) diffusion of polymer chains into the cluster, and (iii) coagulation with other clusters. The rates of process (ii) and (iii) depend strongly on the particle viscosity. The Sundberg group has studied the effect of several factors on the morphology development during two-stage emulsion polymerizations, especially those involving the less hydrophobic copolymer of methyl methacrylate and methyl acrylate as seed polymer and polystyrene as the more hydrophobic second-stage polymer. Durant et al. (1997) veriied the inluence of different amounts of cross-linking monomer (EGDMA) during the syntheses of the PMMA cores on the morphology of PMMA/PS particles. It was observed that 0.015 wt% EGDMA was enough to shift the particle morphology from ICS (second-stage PS in the core) toward CS. At 0.2 wt% EGDMA, the particles was essentially of the CS morphology. Cross-linking during the second stage, on the other hand, was observed by Stubbs and Sundberg (2006) to have very little, if any, effect on morphology, though it enhances the mechanical stability of the shell. The effect of the feed rate of the second-stage more hydrophobic monomer (styrene) when less hydrophobic high-Tg seed polymers (poly[methyl methacrylate]) are used was studied by Stubbs et al. (1999). Fast second-stage monomer addition resulted in CS

34

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

particles, whereas slower addition increased the number of occlusions of the hydrophobic second-stage polymer. When less hydrophobic low-Tg seed polymer (poly[methyl acrylate]) was used, ICS was obtained independently of the second-stage monomer feed rate. Ivarsson et al. (2000) and Karlsson et al. (2003) veriied that it is possible to keep the more hydrophobic second-stage polymer (styrene) at the shell of the particles if the reaction temperature is less than 15°C above Tg of the seed copolymer (poly[methyl methacrylate]/ poly[methyl acrylate]). Stubbs and Sundberg (2004) observed that, though ionic initiators that are able to anchor the chains of the second-stage polymer to the particle surface make it more likely to obtain CS morphologies with poly(methyl methacrylate)/poly(methyl acrylate) core and polystyrene shell under some conditions, this effect is not dominant under most conditions. Based on these results, Stubbs and Sundberg (2008) proposed the decision lowchart shown in Figure 2.2 to be used for morphology prediction of latex particles obtained by emulsion polymerization. Stubbs and Sundberg (2008) considered three main questions for the prediction of the morphology of composite particles: the irst one is whether radicals may penetrate during the second stage of the synthesis, the second is about phase separation, and the last is related to phase consolidation. The theory of radical penetration considers that in emulsion polymerization, radicals are typically created in the water phase, and thus enter latex particles at the outer particle surface.

1

Penetration possible?

Yes No

Yes Phase separation possible? Yes Phase consolidation possible? Yes Extent large or small?

No

2

σ13 > σ23 No

4 No

Core-shell

Gradient or mixed phase Small occlusions

Yes

Tg2 < Treaction No

Core-shell 3

5

Small

Large occlusions

Large

Equilibrium morphology

Lobed particle

6

7

FIGURE 2.2 Adaptation of the decision tree lowchart for predicting morphology development in multiphase particles proposed by Stubbs and Sundberg (2008). σ13: interfacial tension between polymer 1 and aqueous phase; σ23: interfacial tension between polymer 2 and aqueous phase; Tg2: glass transition temperature of poly mer 2. (Reprinted from J. M. Stubbs and D. C. Sundberg, Progress in Organic Coatings 61: 156–65, 2008. With permission.).

Synthesis of Polymer Particles with Core-Shell Morphologies

35

The extent of radical penetration will depend on the effective Tg of the seed polymer. The effective Tg considers that the particle will be partially swollen with second-stage monomer during the polymerization and this will lower its glass transition temperature below that of the pure poly mer. When penetration is possible, a second-stage polymer chain will ind itself inside the particle and fully entangled with the seed polymer chains. Phase separation then requires chain diffusion in order to get multiple chains together, and this process may be so slow that phase separation is not possible. The driving force for morphology rearrangement is the minimization of interfacial free energy, and the system will evolve toward the equilibrium morphology if given suficient time. However, the process of phase consolidation requires an increased extent of polymer mobility compared to the previous two processes of oligomeric radical penetration and polymer phase separation.

2.3 Synthesis of Core-Shell Particles Particles with core-shell morphologies may be synthesized by a number of heterogeneous polymerization techniques such as emulsion, miniemulsion, microemulsion, dispersion, and suspension poly merizations. The irst three techniques lead to the formation of submicrometric particles (10–800 nm), whereas the two last are used, respectively, to prepare small (1–30 µm) and large (50–1500 µm) micrometric particles. Nevertheless, it must be kept in mind that in any of these techniques, the application of a two-stage strategy to build up a shell of the second-stage polymer onto the core of the irst-stage polymer core will not necessarily lead to the formation of particles with coreshell morphology (Rajatapiti et al. 1997). In the next sections, the synthesis of particles with core-shell structure by these different techniques will be described. A detailed description of these polymerization techniques can be found in several excellent books involving emulsion polymerization (Piirma 1982; Gilbert 1995; Lovell and El-Aasser 1997; Van Herk 2005), as well as in recent book chapters about emulsion (de la Cal et al. 2005; Nomura et al. 2005), miniemulsion (Schork et al. 2005), microemulsion (Chow and Gan 2005), dispersion (Kawagushi and Ito 2005), suspension (Brooks 2005), and heterogeneous (Van Herk and Monteiro 2002) polymerization techniques. 2.3.1 Emulsion Polymerization Emulsion polymerization is a heterogeneous polymerization system composed of water, an initiator (usually water soluble), surfactant (usually above the critical micelle concentration [CMC]), and monomer with low water solubility, which under stirring forms droplets with diameters ranging from 1 to

36

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

10 µm. The main locus of polymerization is not in these monomer droplets, but within the submicrometric monomer-swollen polymer particles (60–800 nm). In ab initio emulsion polymerizations, these polymer particles are formed at the beginning of the polymerization by the entry of radicals into micelles, if the surfactant concentration is above the CMC (micellar nucleation mechanism) and/or by the precipitation of growing oligomers in the aqueous phase (homogeneous nucleation mechanism). Due to the relatively large size of the monomer droplets compared to the size of monomer-swollen micelles (10–20 nm), the surface area of the monomer droplets is orders of magnitude smaller than that of the micelles and, consequently, the radical entry into monomer droplets (droplet nucleation) is insigniicant. Since monomer droplet nucleation is insigniicant and monomer-swollen poly mer particles, instead of monomer droplets, are the main poly merization locus, monomer droplets act as monomer reservoirs and the monomer must be transported from these droplets by diffusion through the aqueous phase to allow the growth of the polymer particles by polymerization. In seeded emulsion polymerizations, polymer particles (seeds) are added at the beginning of the polymerization and, usually, these reactions are conducted in the absence of micelles to avoid secondary particle nucleation. Seeded emulsion polymerization is by far the most applied technique for the synthesis of structured polymer particles with core-shell morphology. And two operation modes, batch and semicontinuous, with or without preswelling of the seed particles, are commonly used in these seeded emulsion polymerizations. When direct CS particles with a irst-stage polymer core and a secondstage polymer shell are to be obtained, usually the second-stage monomer is continuously fed at a prespeciied rate in order to allow the second-stage polymer to build up upon the surface of the seed particles, forming a uniform and continuous shell. In this case, reaction conditions that avoid or minimize phase consolidation are often required, especially in those cases in which the CS morphology is not the equilibrium morphology. When, on the other hand, ICS particles with a second-stage polymer core and a irst-stage polymer shell are to be obtained, usually seed particles are swelled by the second-stage monomer prior to either batch or looded semicontinuous polymerization of the second monomer, since conditions that allow phase consolidation are required for phase inversion leading to the equilibrium ICS morphology. The choice of the best operation mode relies strongly on the polymer types and whether the core-shell morphology is to be achieved directly (CS) or through phase inversion (ICS). In the following, the syntheses of CS and ICS particles via emulsion polymerization will be discussed. This also includes some works in which the irst-stage polymer seeds were synthesized via miniemulsion poly merization and the second-stage monomer poly merization was performed via emulsion polymerization.

Synthesis of Polymer Particles with Core-Shell Morphologies

37

2.3.1.1 Synthesis of Core-Shell Particles (CS) In this strategy the second-stage polymer should build up upon the surface of the seed particles, forming a uniform and continuous and, consequently, concentric shell. As a consequence, CS particles may be formed independently of this being the equilibrium morphology. If CS is not the equilibrium morphology, reaction conditions must be carefully adjusted to avoid and/or minimize phase consolidation. Several conditions may favor the formation of CS particles: • High supericial area of the irst-stage polymer. This can be achieved by using small seed particles and/or high seed contents. • Reaction temperature below or close to the glass transition temperature of the polymer particles. If the Tg of the seed polymer is not too low, this might be achieved by starved second-stage monomer feed and/or high initiation rate to keep second-stage monomer concentration, and its plasticizing effect, low in the particles. • Cross-linked core polymer to reduce diffusion of the second-stage monomer and, especially, of the radicals into seed particles. • Enhanced hydrophobicity of seed polymer through its synthesis using an oil-soluble initiator and/or copolymerization with a hydrophobic monomer. • Aqueous phase initiator that is able to anchor the second-stage radicals at the particle surface. • If the incompatibility between the irst- and second-stage polymers is too high to allow the formation of particles with a CS structure, the use of a compatibilizing agent is advised to reduce the polymer-polymer interfacial tension. This compatibilizing agent can be produced in situ by the incorporation of a proper macromonomer (Nelliappan et al. 1996), comonomer (Sherman and Ford 2005), or CRP agent (Herrera et al. 2006) during the synthesis of the seed polymer. Though this strategy might seem the most straightforward for the synthesis of CS particles with hydrophobic cores and hydrophilic shells, in some cases it may result in considerable secondary nucleation instead of the formation of CS particles, since several points listed previously, as for instance high aqueous phase initiator concentration, may lead to homogenous particle nucleation. A quite complete study was presented by Ferguson et al. (2002), who succeeded in synthesizing PS/PVAc core-shell latexes with small cores. In this case the small size of the PS seed particles (unswollen diameter of 88 nm) resulted in a high enough supericial area to capture PVAc radicals formed in the aqueous phase. When, on the other hand, PS seed particles with unswollen diameter of 400 nm were used, excessive new particle formation occurred and

38

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

no PVAc shells could be detected. Numerous strategies for overcoming this were evaluated through simulations with a simpliied nucleation model and/or implemented experimentally, as for instance using an organic-phase initiator in the seeded polymerization to avoid homogeneous nucleation or addition of surfactant during the seeded polymerization in order to maintain the surface charge density and keep the surfactant concentration below the CMC, but always either extensive secondary nucleation occurred or the system became colloidally unstable. Similar results were obtained by Sherman and Ford (2005). The authors used a cross-linked 80/20 PS/PMMA seed 70 nm in diameter to form cationic PS/PMMA CS particles with acentric cores with 530 nm in three steps of PMMA growth, using starved semicontinuous addition of MMA to avoid secondary nucleation and a single initial addition of aqueous phase initiator to provide fast radical generation. The increase of the diameter of the core latex, on the other hand, led to secondary particle nucleation. Due to the incompatibility of PS and PMMA, the equilibrium morphology of the PS/PMMA system is as a double-ball structure. Consequently, when PS homopolymer seeds are used without any procedure to reduce the interfacial tension between both polymers, the CS morphology usually is not attained. A good illustration of the effect of the operation mode on the morphology of PS/PMMA particles was shown in the work of Jönssin et al. (1991). On one extreme, in the seeded batch polymerization of MMA, the authors observed the formation of composite particles with large cluster sizes. Whereas when the seeded semicontinuous polymerization of MMA was conducted under starved conditions, composite particles with small cluster sizes were formed. Both reduction of particle viscosity and the rate of initiation lead to structures with increased cluster sizes. Postpolymerization treatment with solvents leads to the equilibrium morphology (double-ball). In a recent study, Jönssin et al. (2007) veriied the morphology of PMMA (irst-stage)/PS (second-stage) particles prepared by semicontinuous seeded emulsion polymerization of styrene in the presence of polar PMMA seed particles, using different conditions of nonpolar styrene feed rate, radical formation rate, seed particle concentration, and temperature of polymerization. The authors observed that, depending on the set of process conditions used, the morphology of the resulting two-phase particles varied from that of a PMMA/PS CS structure, over intermediate structures in which a shell of PS surrounded a PMMA core containing an increasing number of PS clusters, to a structure in which the entire PS phase was present as discrete PS phase clusters, more or less evenly distributed in a matrix of PMMA. Figure 2.3 (a–d) shows transmission electron microscopy (TEM) micrographs of the microtomed sections of samples of reactions A, B, and C conducted with the same formulation and under the same conditions except for the second-monomer (styrene) feed rate, which was increased from A to C and of reaction D, carried out with the same formulation and conditions of reaction C, but with a higher initiator (KPS) feed rate. The higher the feed rate of the second monomer, the higher is its plasticizing effect and the lower is the Tg of the polymer

Synthesis of Polymer Particles with Core-Shell Morphologies

(a)

(b)

(c)

(d)

39

FIGURE 2.3 (a–d) TEM micrographs of thin sections taken from particles obtained in experiments A–D. (e) Rate of poly merization for experiments A–D. (f) The amount of styrene accumulated in the reactor as calculated from the poly merization curves from experiments A–D, F, and G. (Reprinted from J.-E. L. Jönsson, O. L. Karlsson, H. Hassander, and B. Törnell, European Polymer Journal 43: 1322–32, 2007. With permission.)

40

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Rate of Polymerization (W)

1.2 D

1

C

0.8

B

0.6

A 0.4 0.2 0 0

100

200

300 400 Time (min.) (e)

500

600

Accumulated Styrene (g/g)

0.25 C

0.2 0.15

G 0.1

B

D

A 0.05 F 0 0

100

200

300 400 Time (min.) (f )

500

600

FIGURE 2.3 (continued).

particles; consequently, the higher becomes the diffusivity of the entering oligo radicals, increasing the number and distributing more homogeneously the PS clusters in the PMMA matrix. Results of reaction D, conducted with the same styrene feed rate as reaction C, but with a higher initiator feed rate, indicate the formation of a direct CS structure; this is attributed to the high reaction rate (Figure 2.3e), leading to low monomer concentrations in the polymer particles (Figure 2.3f) and, consequently, increasing the Tg of the polymer particles to only a few degrees below the reaction temperature. Consequently, contrary to the expected equilibrium morphologies, twophase particles with a true core-shell structure were obtained in experiments where the estimated glass transition temperature of the PMMA phase was only a few degrees below the polymerization temperature, independently

Synthesis of Polymer Particles with Core-Shell Morphologies

41

of the feed rate of the second monomer, provided that the rate of initiation, using an aqueous phase initiator, was properly adjusted. The formation of particles with nonequilibrium morphologies was also observed by Zhao et al. (2004), who studied the morphological development of PVAc (irst-stage)/PBuA (second-stage) in seeded semicontinuous starved emulsion poly merization using an aqueous phase initiator in the synthesis of both seed and second-stage poly mer. The authors observed that, despite the low Tgs of both seed and second-stage polymers, the PBuA was formed at the outside of the PVAc seed. After storage for 1.5 years at room temperature, latex morphology evolved to the equilibrium morphology, which for this system is ICS with PBuA core and a PVAc (more hydrophilic) shell. 2.3.1.2 Synthesis of Inverted Core-Shell Particles (ICS) As mentioned in the previous section (2.3.1.1), sometimes the synthesis of structured polymer particle presenting the core-shell morphology may be very dificult, either because the CS morphology is not favored thermodynamically or because secondary particle nucleation is dificult to avoid. Often in these cases, the synthesis of ICS particles formed by a core of the secondstage polymer and a shell of irst-stage polymer is an alternative that should not be discarded. For the synthesis of inverted core-shell particles, conditions must be carefully chosen to favor phase inversion. Consequently, contrary to CS particles, ICS particles may be formed only if this is the equilibrium morphology. Several conditions may favor the formation of ICS particles: • High supericial area of the irst-stage polymer to avoid secondary particle nucleation. This can be achieved by using small seed particles and/or high seed contents. • Reaction temperature much above (Ivarsson et al. 2000; Karlsson et al. 2003) glass transition temperature of the polymer particles. This might be achieved by either batch or looded semicontinuous second-stage polymerization with preswelling and/or low initiation rate to keep second-stage monomer concentration, and its plasticizing effect, high in the particles. • Chain transfer agent (CTA) to reduce the molecular weight of the polymers. • Enhanced hydrophilicity of seed polymer through its synthesis using an aqueous phase initiator and/or copolymerization with a hydrophilic monomer. • Enhanced hydrophobicity of the second-stage polymer through its synthesis using an oil-soluble initiator that is not able to anchor the second-stage radicals at the particle surface.

42

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

• If the incompatibility between the irst- and second-stage polymers is too high to allow the formation of particles with an ICS structure, the use of a compatibilizing agent is advised to reduce the polymer-polymer interfacial tension. This compatibilizing agent can be produced in situ by the incorporation of a proper macromonomer (Nelliappan et al. 1996), comonomer (Sherman and Ford 2005), or CRP agent (Herrera et al. 2006) during the synthesis of the seed polymer. Most of the factors just listed were mentioned by Lee and Ishikawa (1983), who found that when hydrophobic monomers are polymerized in the presence of highly hydrophilic polymer seed particles, the second-stage hydrophobic polymers form cores surrounded by the irst-stage hydrophilic polymers, resulting in inverted core-shell latexes. They also observed that the formation of the core-shell morphology by this inversion process is favored by higher hydrophilicity, lower interfacial tension, lower molecular weight of hydrophilic irst-stage polymer, and greater phase separation between both polymers, which is expected to increase with greater incompatibility between these polymers, lower glass transition temperatures of both polymers, lower molecular weights and cross-linking, higher polymerization temperatures, and more looded conditions. In the case of the soft polymer pair systems, the formation of inverted core-shell morphology was equally complete, regardless of the molecular weight of the hydrophilic polymer molecules, whereas in the case of the hard polymer pair systems, the eficiency of inversion was dependent on the molecular weights of both hydrophilic and hydrophobic polymers. In the work of Ferguson et al. (2002), the synthesis of PS(irst stage)/ PVAc(second stage) core-shell particles with large cores was hampered by secondary nucleation. Nevertheless, in a different attempt, Ferguson et al. (2003) showed that it is possible to synthesize PS/PVAc core-shell particles with large PS cores through an inverse core-shell synthesis using PVAc seed particles in a batch styrene polymerization. For this system, the inverse coreshell morphology is favored thermodynamically, and undesired secondary particle nucleation is avoided by the low aqueous phase solubility of styrene. Nevertheless, the authors observed that conditions had to be chosen carefully in order to minimize the kinetic control on the morphology, allowing the achievement of the ICS morphology. To keep it short, the fast diffusion of the hydrophilic seed polymer had to be ensured, for instance by using a chain transfer agent (to reduce the molecular weight and, especially, the degree of branching), an initiator able to minimize grafting between both polymers, and a comonomer in the seed polymer to increase its hydrophilicity. An important aspect of ICS syntheses mentioned in this work is that even for those systems in which the ICS is thermodynamically favored and kinetically achievable, there is no driving force for this core to be concentric

Synthesis of Polymer Particles with Core-Shell Morphologies

43

100 nm

100 nm 100 nm (a) Latex 1 (with SG1)

(b) Latex 2 (without SG1)

(c) Latex 3 (with CTA)

FIGURE 2.4 TEM micrographs of PS/PMMA composite particles stained with PTA and RuO4. (Reprinted from V. Herrera, R. Pirri, J. R. Leiza, and J. M. Asua, Macromolecules 39: 6969–74, 2006. With permission.).

in an ICS latex (as long as the hydrophobic core is shielded from the particlewater interface region). The high incompatibility of the polymer phases may be overcome by the incorporation of a compatibilizing agent. Herrera et al. (2006) added during the miniemulsion polymerization for the formation of the PMMA seed a small amount of a CRP agent, so that at the end of the seed formation some of the polymer chains were capped with the CRP agent. Batch emulsion polymerization of the second-stage monomer (styrene, with previous swelling) in the presence of additional initiator led to the in situ formation of block copolymer chains. When PMMA seeds were synthesized in the presence of the CRP agent, ICS structures were formed (Figure 2.4a). On the other hand, in similar reactions with PMMA seeds synthesized without CRP agent, the hemisphere morphology was obtained, independently of the molecular weight of the seed PMMA polymer (PMMA seeds formed without, Figure 2.4b, or with chain transfer agent, Figure 2.4c). This result indicated that the in situ formation of block copolymer chains allowed the modiication of particle morphology through the reduction of the interfacial tension between the two polymers, helping to increase their compatibility and, consequently, allowing the formation of the ICS morphology. 2.3.2 Miniemulsion Polymerization Miniemulsion polymerization differs from a conventional emulsion polymerization in that the monomer droplets, formed with the aid of highshear devices (rotor-stator system, soniier, high-pressure homogenizer), are in a submicrometric range (50–500 nm) forming a miniemulsion that is

44

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

thermodynamically unstable, but kinetically metastable (stable for a period ranging from hours to months) due to the addition of a surfactant to minimize coalescence and a costabilizer to retard droplet diffusion degradation (Ostwald ripening). Free surfactant concentrations in the aqueous phase are usually kept below the CMC, and the relatively small size of the monomer droplets favors droplet nucleation. Thus, droplet nucleation and poly merization within the submicrometric miniemulsion droplets are aimed in a miniemulsion polymerization. Either oil- or water-soluble initiators may be used. In miniemulsion polymerization, the irst-stage polymer may be dissolved in the second-stage monomer, and this solution, with or without the addition of a costabilizer, is dispersed in the aqueous phase containing a surfactant using an adequate homogenization equipment; the formed miniemulsion of submicrometric organic phase droplets is subsequently polymerized, maximizing droplet nucleation. This approach has the advantage that the irststage polymer must not be in the form of nanoparticles, being well suited for the synthesis of hybrid polymer particles composed of a polymer produced by step-growth polymerization and a polymer produced by free radical polymerization (Wang et al. 1996; Wu et al. 1999; Barrere and Landfester 2003; Li et al. 2005; Wang et al. 2005; Guyot et al. 2007; Jahanzad et al. 2007). The drawback of this approach is that the amount of irst-stage polymer is limited by the viscosity increase of the organic phase, which depends strongly on the molecular weight of the irst-stage polymer. However, this amount can reach up to 50 wt% related to the organic phase as shown by López et al. (2008), who succeeded in preparing miniemulsions with droplet sizes around 100 nm using irst-stage polymers (alkyd resins and poly[ε-caprolactone]) dissolved in acrylic monomers. It has also been shown that hybrid polymer particles may also be synthesized by simultaneous polymerization in miniemulsion, as in Barrere and Landfester (2003), who synthesized hybrid polystyrene/polyurethane and poly(butyl acrylate)/polyurethane particles via miniemulsion polymerization by combining polyaddition and radical polymerization. The hybrid particles obtained by miniemulsion poly merization will show only CS or ICS structures if these are the equilibrium morphologies and if they are kinetically achievable. This is the case of the work of Rodríguez et al. (2008), where it was shown that most particles obtained by silicon-acrylic miniemulsion polymerization presented nonreactive polydimethylsiloxane divinyl terminated (PDMS)-rich core morphologies, as PDMS is highly hydrophobic. In the case of the ICS structures, the ratio (irst-stage polymer):(second-stage monomer) must be high enough to allow the formation of a continuous shell. An increasing number of publications involve the synthesis via miniemulsion polymerization of CS particles with a liquid core (nanocapsules). The procedure is quite the same as for the synthesis of hybrid particles using a

Synthesis of Polymer Particles with Core-Shell Morphologies

45

100 nm (a) FIGURE 2.5 (a) TEM micrograph of PS (styrene poly merized with 10 wt% acrylic acid) nanocapsules with hexadecane core in a 1:1 (PS:HD) ratio prepared by miniemulsion poly merization. (Reprinted from K. Tiarks, K. Landfester, and M. Antonietti, Langmuir 17: 908–18, 2006. With permission.) (b) TEM and SEM (insert) micrographs of PMMA nanocapsules with Neobee M-5 core in a 1:1 (PMMA:Neobee M5) ratio obtained by miniemulsion poly merization. Arrows with indexes TEC, PEC, and NEC indicate, respectively, totally enguling capsules, partially enguling capsules, and nonenguling particles. (Reprinted from A. P. Romio, C. Sayer, P. H. H. Araújo, M. Al-Haydari, L. Wu, and S. R. P. da Rocha, Macromolecular Chemistry and Physics 210:747–51, 2009. With permission.)

preformed polymer. One difference in this latter case is that phase separation is facilitated by the lower viscosity of the liquid compared to that of a preformed polymer. When hydrophobic liquids are to be encapsulated by a polymer shell, direct miniemulsion polymerization may be applied (Tiarks et al. 2001; Crespy et al. 2006, 2007; Luo and Gu 2007; Romio, Bernardy et al. 2009; Romio, Sayer et al. 2009). In this case, the hydrophobe and monomer form a homogenous organic phase, which is miniemulsiied in the aqueous phase, and phase separation in the organic phase occurs throughout the polymerization. Figure 2.5 (a and b) shows micrographs of polystyrene (Tiarks et al. 2001) and poly(methyl methacrylate) (Romio, Sayer et al. 2009) nanocapsules synthesized via miniemulsion polymerization. A special case of core-shell particles are the onion-like morphologies. This kind of structure can be obtained when block copolymers (of two immiscible homopolymers) are formed. Kagawa et al. (2005) synthesized

46

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

2 µm

PEC

TEC NEC

1 µm (b) FIGURE 2.5 (continued).

(b)

(a)

200 nm

200 nm

FIGURE 2.6 TEM micrographs of ultrathin cross sections of PiBMA-b-PS particles, stained with RuO4 vapor for 30 min, prepared by seeded-ATRP with PiBMA-Br particles prepared by miniemulsion ATRP at Tween 80 concentrations of (a) 3 and (b) 6 wt% based on iBMA. (Reprinted from Y. Kagawa, H. Minami, M. Okubo, and J. Zhou, Polymer 46: 1045–49, 2005. With permission.)

submicron-sized poly(i-butyl methacrylate)-block-polystyrene particles with an onion-like multilayered structure (Figure 2.6) using a two-step atom transfer radical polymerization (ATRP) in aqueous media: ATRP in miniemulsion to obtain the PiBMA seed particles followed by ATRP in seeded styrene emulsion polymerization.

Synthesis of Polymer Particles with Core-Shell Morphologies

47

2.3.3 Microemulsion Polymerization In the same way as emulsion polymerization, two-stage microemulsion polymerization may also be used to synthesize CS particles. In this poly merization technique, extremely small monomer droplets (10–30 nm) forming a thermodynamically stable microemulsion are obtained by the combination of high surfactant (and cosurfactant) and relatively low monomer concentrations. The advantage of this two-stage microemulsion polymerization technique is related to the small sizes and, consequently, high supericial area of the particles that favors radical and monomer entry in detriment to undesired nucleation during the second stage of the polymerization. Nonetheless, CS particle synthesis by this technique shows the same drawbacks related to the low monomer content (0–10 wt%) and low monomer:surfactant ratio of other microemulsion polymerizations. This technique has been used by Jang and Ha (2002) to prepare CS particles of 30 nm in size formed by PMMA core and PS shell with triblock copolymers of poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) [(EO)x(PO)y(EO)x] as surfactant. These CS particles were subsequently treated with a selected solvent to remove the PMMA in order to obtain nanosized hollow polystyrene microlatexes. 2.3.4 Dispersion Polymerization Seeded dispersion polymerization is another technique that may be applied for the synthesis of particles with core-shell structures with average particle diameters in the range of 1 to 30 μm. In dispersion polymerization, the monomer is completely miscible with the continuous phase that is also composed by one or more nonreactive components, usually an alcohol as methanol or ethanol with or without water and that contains a stabilizer, as for instance poly(vinyl pyrrolidone) and the initiator. As a consequence, dispersion polymerization starts as a homogeneous polymerization, but, since the polymer formed throughout the polymerization is not soluble in the continuous phase, it precipitates, forming colloidal particles that are stabilized by the added stabilizer. Thus, polymerization takes place in both continuous and particle phases, in degrees that may be varied, depending on the partitioning of monomer and radicals between these phases. Okubo and Izumi (1999) synthesized PMMA/PS CS particles by seeded dispersion polymerization and showed that, when the reaction temperature is below the glass transition temperature of the seed particles, this technique has the advantage of producing core-shell polymer particles in which layers accumulate in their order of production (kinetic control of morphology), regardless of the hydrophobicity of polymers forming core and shell, even if the morphology is thermodynamically unstable. This is attributed to the enhanced solubility of the second-stage monomers in the continuous phase

48

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(due to the presence of an alcohol) combined with the high glass transition temperature of the seed polymer and fast adsorption of polymer radicals and inactive chains formed in the continuous phase onto the seed particles. When, however, the rate of adsorption of polymer radicals and inactive chains formed in the continuous phase onto the seed particles is slow, the second-stage polymer has the tendency to form clusters on the surface of the seed particle, and particles with different morphologies, egg-, snowman-, and confetti-like, may be formed (Fujibayashi et al. 2008). 2.3.5 Suspension Polymerization Suspension polymerization is a process in which the polymerization of relatively water-insoluble monomer droplets formed by vigorous stirring in the presence of a steric stabilizer leads to an aqueous dispersion of polymer particles in the size range of 50 to 1500 µm. To obtain CS particles by suspension polymerization, a two-step procedure should be employed, where the seed particles are obtained by suspension polymerization and the second-stage monomer is added afterward. The large particle size of the suspension polymerization particles, and the consequently low supericial area, do not favor second-stage monomer absorption; as a consequence, undesired secondary particle nucleation is hard to avoid (Gonçalves et al. 2008a). In addition, if large amounts of second-stage monomer are used to swell seed particles, shear forces inside the reactor may lead to the undesired breakage of the swollen seed particles. And, even with small amounts, if this second-stage monomer concentrates at the outer layers of the seed particles due to slow monomer diffusion inside the particles, this would turn the particle surface very sticky and could lead to particle agglomeration. As a consequence, the synthesis of CS particles via suspension poly merization is a quite dificult and challenging task. Gonçalves et al. (2008b) synthesized large polystyrene (PS)/poly(methyl methacrylate) (PMMA) CS-structured particles by seeded (PS seeds) suspension polymerization of MMA in which the shell was composed by PMMA clusters densely dispersed in the PS matrix. The size and concentration of these clusters (shell) decreased from the outer shell to the core, and particle morphology was largely controlled by limitations to the diffusion of MMA through the polymer particle. As a consequence, thicker shells with a higher amount of incorporated PMMA were obtained by increasing the swelling time. Figure 2.7 shows a TEM micrograph and infrared (IR) data of the cross section of a particle. IR data conirm that the concentration of PMMA, denoted by the intensity of carbonyl absorption bands at 1740 cm–1, reached maximum values near the particle surface and decreased along the particle radius. An alternative for the synthesis of large particles with CS structures is the combination of suspension and emulsion polymerization to synthesize, respectively, the core and shell of the particles (Lenzi et al. 2003; Zhenqian

49

Synthesis of Polymer Particles with Core-Shell Morphologies

5 µm

5 µm

5 µm

(a) FIGURE 2.7 (a) TEM micrographs and the approximated representation of the cross section (stained with RuO4, PMMA appears as light gray and PS as dark gray; 6000x). (b) FTIR spectra of the cross section of a typical core-shell particle obtained with swelling time of 0 min and 0.23 mol% of BPO. (Reprinted from O. H. Gonçalves, J. M. Asua, P. H. H. Araújo, and R. A. F. Machado, Macromolecules 41: 6960–64, 2008. With permission.)

et al. 2009). In this case, the shell is composed by small aggregated emulsion polymerization particles, and the CS particles did not present a smooth spherical shape. 2.3.6 Other Techniques Several other techniques have also been described for the formation of CS particles. For instance, the preparation of preformed polymer blends in nanoor microparticles through the dissolution of both polymers in a common solvent followed by either the Shirasu porous glass membrane (SPG) emulsiication technique (Ma et al. 1999, 2003), emulsiication (Tanaka et al. 2008), miniemulsiication using high-energy devices (Kietzke at al. 2007), or the nanoprecipitation technique always requires subsequent solvent evaporation

50

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

0.12 0.10

0.06 0.04

w

0.02

e av

1820 1800 1780 1760 1740 1720 1700 1680 1660 1640

intensity

0.08

P2

P4

P5

P6

0.00 P8

–1

cm r(

be

m

nu

P1

P3

P7

)

(b) FIGURE 2.7 (continued).

after the dispersion of the organic phase in the aqueous phase. The two former methods lead to the formation of hybrid microparticles, whereas the two latter methods may be used to form hybrid nanoparticles. In all four cases, particles with CS structures may be attained only if this is the equilibrium morphology. The SPG emulsiication technique with dichloromethane (DCM) as solvent for the preformed polymers, lauryl alcohol (LOH) as costabilizer, and poly(vinyl alcohol) and sodium lauryl sulfate in the aqueous phase has been reported for the preparation of uniform composite polystyrene/poly(methyl methacrylate) microspheres with diameters around 10 μm (Ma et al. 1999). The morphology of the composite particles was varied by the amount of LOH and by the PMMA:PS ratio, as shown in Figure 2.8. Polymer microcapsules have also been produced by applying microluidic devices using capillary instability-driven break-up of a liquid jet formed by two immiscible luids (oil and monomer) followed by photopolymerization of the monomeric shells (Nie et al. 2005), resulting in precise control of core number, location, and diameter, as well as shell thickness. Finally, particles having hollow structures, either core-shell or multihollow, have also been synthesized by water-in-oil-in-water (W/O/W) emulsion poly merization (Kim et al. 1999) and by a modiied emulsion polymerization with a watermiscible alcohol and a hydrocarbon nonsolvent for the polymer (McDonald et al. 2000).

Synthesis of Polymer Particles with Core-Shell Morphologies

51

(b) 402

(a) 398

10 µm

10 µm

(c) 396

(d) 398

1 µm

10 µm

(e) 402

1 µm

(f ) 396

1 µm

FIGURE 2.8 Optical micrographs of droplets (a–c) and TEM micrographs of embedded cross section with RuO4 vapor staining of particles (d–f) as a function of the PMMA:PS ratio when 11.9 ml of DCM and 0.1 ml of LOH were used for 0.24 g of poly mer (PMMA and PS). PMMA/PS (w/w): (a, d) 5/5, (b, e) 4, 6, and (c, f) 1/9. (Reprinted from G. H. Ma, M. Nagai, and S. Omi, Journal of Colloid Interface Science 219: 110–28, 1999. With permission.)

52

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

2.4 Characterization of Core-Shell Particles The characterization of core-shell particles includes determining the particle shape and the composition of the particle surface and the particle interior. Although microscopy data is required to make deinite conclusions about particle morphology, according to Stubbs and Sundberg (2008), no single analytical technique is able to provide all the relevant details about the particle structure, but different techniques are required to adequately characterize the morphology of a given system, even for relatively simple systems. There are several relevant characterization techniques used for particle morphology determination including the direct microscopic techniques such as: TEM (transmission electron microscopy), AFM (atomic force microscopy), and SEM (scanning electron microscopy). Other indirect analyses could be performed by NMR (nuclear magnetic resonance), DSC (differential scanning calorimetry), and IR spectroscopy. And to characterize the latex particle surface, conductometric titration, among others, could also be used. The choice of the techniques used to characterize the core-shell latexes depends on whether the polymer particle is being characterized with respect to shape, structure, and/or surface. 2.4.1 Transmission Electron Microscopy TEM is the most widely used technique to characterize core-shell latexes (Gaillard et al. 2007) because a direct image of the structure can be obtained. However, the quality of the results obtained by TEM is a direct result of the sample preparation methods employed and the fact that this technique requires a high level of operator skill (Stubbs and Sundberg 2005b). TEM imaging presents several dificulties due to ilm formation during sample preparation of latex of low Tg, the weak contrast between the different phases and/or between the particles and the background, and their poor electron beam resistance, causing the deformation of the polymer particle during the analysis (Egerton et al. 2004). Most polymer analyses are carried out at accelerating voltages of 80 or 100 kV, although several polymers, like acrylic polymers, are more susceptible to radiation damage. Therefore, some limitations of TEM analysis are imposed by the polymer system. According to Gaillard et al. (2007), these dificulties could be overcome by using low temperature and/or low exposure to strong electron beam techniques and specimen preparation methods, like chemical staining, avoiding ilm formation, and stabilizing and enhancing the contrast between the different polymer phases. Cold-stage holders could be used to investigate the particle morphology of low-Tg latex or microgels’ core-shell particles (Ballauff and Lu 2007) because

Synthesis of Polymer Particles with Core-Shell Morphologies

53

dispersions can be frozen, placed into TEM, and imaged. The low temperature also has the advantage of reducing radiation damage that occurs with all polymers (Dimonie et al. 1997). Chemical staining of the polymer particles generally has the advantage of hardening many types of polymer, which may facilitate specimen preparation and observation. To enhance the contrast between the particles and the background, “negative” staining with phosphotungstic acid (PTA) or uranyl acetate (UAc) is largely applied where the “negative” stain chemically interacts with the carbon ilm that covers the TEM grid. “Negative” staining is preferably used with latex particles not having reactive groups and that are not stable in the microscope or that are ilm forming (Karlsson and Schade 2005) like poly(vinyl acetate). “Positive” staining occurs when the chemical stain interacts directly with the polymer specimen by diffusion or reaction; the main use of selective “positive” staining with heavy elements (Ru, Os, tungsten, and uranium) is to increase high-angle Rutherford scattering, generating contrast between the stained and the unstained regions of the polymer particle (Ferguson et al. 2002). The most common “positive” chemical staining agents are: osmium tetra oxide, OsO4, which is applied for unsaturated polymers that present residual allylic double bonds such as natural rubber-based latexes or polybutadiene; and ruthenium tetroxide, RuO4, which is more reactive than OsO4 and will react with double bonds in phenyl rings and with amine groups besides allylic double bonds (Trent 1984). RuO4 is particularly useful in morphological studies of multicomponent particles containing polystyrene in one of the phases (Stubbs et al. 1999). To avoid distortion of ilm-forming latex particles during drying, staining can be done in the liquid phase by adding the positive stain solution to a diluted latex dispersion (Karlsson et al. 1995). The combination of “positive” and “negative” staining methods is also useful to reveal internal structures and to ixate ilm-forming latexes like PS-PVAc core-shell latex where UAc is used for negative staining and RuO4 for staining the PS phase (Ferguson et al. 2002). Another way to increase the contrast between the polymer phases is to resolve one of the phases under exposure of a strong electron beam (200 kV) due to preferential degradation of one of the polymer phases. This is particularly true for PMMA/PS core-shell particles (Okubo and Izumi 1999) because PMMA homopolymers are signiicantly more beam sensitive than styrenic polymers; therefore, there will be a preferential beam-induced mass loss in the MMA-rich copolymer phase. Embedding and sectioning are sample preparation methods that enable one to cut the sample suficiently thin for the microscope to increase its resolution (Karlsson and Schade 2005). In this technique, dried latex particles are embedded in epoxy resins and cured at room temperature. The cured epoxy/particle blend is then ultramicrotomed to obtain sections in the order of 50–100 nm.

54

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

2.4.2 Scanning Electron Microscopy Compared to TEM, the resolution of SEM is lower but detailed images of particle surface can be obtained resulting in information regarding size, shape, and structure. High-resolution ield emission SEM (FESEM) presents a much higher resolution than conventional SEM using a thermionic emitter (Karlsson and Schade 2005). Typically, the polymer dispersion is dried down onto a stub; and to increase the resolution of nonconductive polymer, specimen is sputter coated with a ~10- to 20-nm-thick layer of gold or palladium. To observe the different phases in a poly mer blend, one of the polymer phases could be selectively chemically etched off to achieve topographical differences that could be imaged by SEM. However, this is normally not employed on particles obtained by emulsion poly merization due to the submicrometric dimensions of these particles. 2.4.3 Atomic Force Microscopy AFM can be used to determine the surface structure of polymer particles because it presents a much higher resolution than SEM, making it a very useful tool to study ilm formation of polymer dispersions. No vacuum is needed and no dificult preparation is involved. However, the direct use of AFM for studying single particles has not been widely employed, partly because not all dispersions are suitable for morphology studies by AFM as the poly mer particle should present at least one accessible low-Tg phase in order to reveal internal morphologies (Karlsson and Schade 2005). 2.4.4 Additional Techniques Used for Particle Characterization Other techniques can be used to complement the information obtained from the direct microscopic techniques. Many of these techniques may be applied in ilms made from structured latex particles. NMR methods (1H spin-diffusion measurements) can be used in latex ilms to characterize the interface of two immiscible polymer phases and its thickness (Landfester et al. 1996). Careful interpretation of thermal analysis data from DSC can indicate the relative extent of phase separation within the structured latex particles by comparing the size of the ΔCp transitions in DSC for ilms of latex particles to the transitions for the same pure bulk polymer (Stubbs and Sundberg 2005a). Because IR spectroscopy is sensitive to polymer composition, it can be used to investigate the structure of core-shell latexes and to deduce the thickness of the interfacial layer from spectral characteristics (Lange et al. 1988). Besides that, when polymer particles are large enough, as particles produced by suspension polymerization, IR spectroscopy could be employed to quantify polymer composition along the particle radius (Gonçalves et al. 2008b).

Synthesis of Polymer Particles with Core-Shell Morphologies

55

References Ballauff, M., and Y. Lu. 2007. “Smart” nanoparticles: Preparation, characterization and applications. Polymer 48:1815–23. Barrere, M., and K. Landfester. 2003. High molecular weight polyurethane and polymer hybrid particles in aqueous miniemulsion. Macromolecules 36:5119–25. Brooks, B. W. 2005. Free-radical polymerization: Suspension. In Handbook of polymer reaction engineering, ed. T. Meyer, and J. Keurentjes, 213–47. Berlin: Wiley-VCH Verlag. Chow, P. Y., and L. M. Gan. 2005. Microemulsion polymerizations and reactions. In Advances in polymer science, vol. 175, Polymer particles, ed. M. Okubo, 257–98. Berlin-Heidelberg: Springer-Verlag. Crespy, D., Musyanovych, A., and K. Landfester. 2006. Synthesis of polymer particles and nanocapsules stabilized with PEO/PPO containing polymerizable surfactants in miniemulsion. Colloid and Polymer Science 284:780–87. Crespy, D., Stark, M., Hoffmann-Richter, C., Ziener, U., and K. Landfester. 2007. Polymeric nanoreactors for hydrophilic reagents synthesized by interfacial polycondensation on miniemulsion droplets. Macromolecules 40:3122–35. De la Cal, J. C., Leiza, J. R., Asua, J. M., Butte, A., Storti, G., and M. Morbidelli. 2005. Emulsion polymerization. In Handbook of polymer reaction engineering, ed. T. Meyer and J. Keurentjes, 249–322. Berlin: Wiley-VCH Verlag. Dimonie, V. L., Daniels, E. S., Shaffer, O. L., and M. S. El-Aasser. 1997. Control of particle morphology. In Emulsion polymerization and emulsion polymers, ed. P. A. Lovell and M. S. El-Aasser. New York: John Wiley & Sons. Durant, Y. G., Sundberg, E. J., and D. C. Sundberg. 1997. Effects of cross-linking on the morphology of structured latex particles. 2. Experimental evidence for lightly cross-linked systems. Macromolecules 30:1028–32. Egerton, R. F., Li, P., and M. Malac. 2004. Radiation damage in the TEM and SEM. Micron 35:399–409. Ferguson, C. J., Russel, G. T., and R. G. Gilbert. 2002. Synthesis of latices with polystyrene cores and poly(vinyl acetate) shells. 1. Use of polystyrene seeds. Polymer 43:6371–82. ———. 2003. Synthesis of latices with hydrophobic cores and poly(vinyl acetate) shells. 2. Use of poly(vinyl acetate) seeds. Polymer 44:2607–19. Fujibayashi, T., Komatsu, Y., Konishi, N., Yamori, H., and M. Okubo. 2008. Effect of polymer polarity on the shape of “golf ball-like” particles prepared by seeded dispersion polymerization. Industrial Engineering and Chemistry Research 47:6445–49. Gaillard, C., Fuchs, G., Plummer, C. J. G., and P. A. Stadelmann. 2007. The morphology of submicronsized core-shell latex particles: An electron microscopy study. Micron 38:522–35. Gilbert, R. G. 1995. Emulsion polymerization: A mechanistic approach. London: Academic Press. Gonçalves, O. H., Machado, R. A. F., Araújo, P. H. H., and J. M. Asua. 2008a. Secondary particle formation in seeded suspension polymerization. Polymer 50:375–81.

56

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Gonçalves, O. H., Asua, J. M., Araújo, P. H. H., and R. A. F. Machado. 2008b. Synthesis of PS/PMMA core-shell structured particles by seeded suspension polymerization. Macromolecules 41:6960–64. González-Ortiz, L. J., and J. M. Asua. 1995. Development of particle morphology in emulsion polymerization. 1. Cluster dynamics. Macromolecules 28:3135–45. ———. 1996a. Development of particle morphology in emulsion polymerization. 2. Cluster dynamics in reacting systems. Macromolecules 29:383–89. ———. 1996b. Development of particle morphology in emulsion polymerization. 3. Cluster nucleation and dynamics in polymerizing systems. Macromolecules 29:4520–27. Guyot, A., Landfester, K., Schork, F. J., and C. Wang. 2007. Hybrid polymer latexes. Progress in Polymer Science 32:1439–61. Herrera, V., Pirri, R., Leiza, J. R., and J. M. Asua. 2006. Effect of in-situ-produced block copolymer on latex particle morphology. Macromolecules 39:6969–74. Ivarsson, L. E., Karlsson, O. J., and D. C. Sundberg. 2000. Inluence of glass transition temperature on latex particle morphology. Macromolecular Symposia 151:407–12. Jahanzad, F., Karatas, E., Saha, B., and B. W. Brooks. 2007. Hybrid polymer particles by miniemulsion polymerization. Colloids and Surfaces A: Physicochemical Engineering Aspects 302:424–29. Jang, J., and H. Ha. 2002. Fabrication of hollow polystyrene nanospheres in microemulsion polymerization using triblock copolymers. Langmuir 18:5613–18. Jönsson, J.-E. L., Hassander, H., Jansson, L. H., and B. Törnell. 1991. Morphology of two-phase polystyrene/poly(methyl methacrylate) latex particles prepared under different polymerization conditions. Macromolecules 24:126–31. Jönsson, J.-E. L., Karlsson, O. L., Hassander, H., and B. Törnell. 2007. Semi-continuous emulsion polymerization of styrene in the presence of poly(methyl methacrylate) seed particles. Polymerization conditions giving core-shell particles. European Polymer Journal 43:1322–32. Kagawa, Y., Minami, H., Okubo, M., and J. Zhou. 2005. Preparation of block copolymer particles by two-step atom transfer radical polymerization in aqueous media and its unique morphology. Polymer 46:1045–49. Karlsson, L. E., Karlsson, O. J., and D. C. Sundberg. 2003. Nonequilibrium morphology development in seeded emulsion polymerization. II. Inluence of seed polymer Tg. Journal of Applied Polymer Science 90:905–15. Karlsson, O. J., Hassander, H., and B. Wesslen. 1995. Particle size measurements of heterogeneous ilm-forming latexes. Colloid and Polymer Science 273:496–504. Karlsson, O., and B. E. H. Schade. 2005. Particle analysis: Particle size, particle shape and structure and surface characterization. In Chemistry and technology of emulsion polymerization, ed. A. van Herk. Oxford: Blackwell Publishing, pp 186–225. Kawaguchi, S., and K. Ito. 2005. Dispersion polymerizations. In Advances in polymer science, vol. 175, Polymer particles, ed. M. Okubo, 299–328. BerlinHeidelberg: Springer-Verlag. Kietzke, T., Neher, D., Kumke, M., Ghazy, O., Ziener, U., and K. Landfester. 2007. Phase separation of binary blends in polymer nanoparticles. Small 3:1041–48. Kim, J. W., Joe, Y. G., and K. D. Suh. 1999. Poly(methyl methacrylate) hollow particles by water-in-oil emulsion polymerization. Colloid and Polymer Science 277:252–56. Koskinen, M., and C.-E. Wilén. 2009. Preparation of core-shell latexes for paper coatings. Journal of Applied Polymer Science 112:1265–70.

Synthesis of Polymer Particles with Core-Shell Morphologies

57

Landfester, K., Boeffel, C., Lambla, M., and H. W. Spiess. 1996. Characterization of interfaces in core-shell polymers by advanced solid-state NMR methods. Macromolecules 29:5972–80. Lange, J., Hergeth, W.-D., and S. Wartewig. 1988. Infrared spectroscopy of polymeric core-shell latexes. Acta Polymerica 39:479–81. Lee, D. I., and T. Ishikawa. 1983. The formation of “inverted” core-shell latexes. Journal of Polymer Science, Polymer Chemistry Edition 21:147–54. Lenzi, M. K., Silva, F. M., Lima, E. L., and J. C. Pinto. 2003. Semibatch styrene suspension polymerization. Journal of Applied Polymer Science 89:3021–38. Li, M., Daniels, E. S., Dimonie, V., Sudol, E. D., and M. S. El-Aasser. 2005. Preparation of polyurethane/acrylic hybrid nanoparticles via a miniemulsion polymerization process. Macromolecules 38:4183–92. López, A., Chemtob, A., Milton, J. L., Manea, M., Paulis, M., Barandiaran, M. J., Theisinger, S., Landfester, K., Hergeth, W. D., Udagama, R, Mckenna, T., Simal, F., and J. M. Asua. 2008. Miniemulsiication of monomer-resin hybrid systems. Industrial Engineering and Chemistry Research 47:6289–97. Lovell, P. A., and M. S. El-Aasser. 1997. Emulsion polymerization and emulsion polymers. New York: John Wiley & Sons. Luo, Y., and H. Gu. 2007. Nanoencapsulation via interfacially conined reversible addition fragmentation transfer (RAFT) miniemulsion polymerization. Polymer 48:3262–72. Ma, G. H., Nagai, M., and S. Omi. 1999. Effect of lauryl alcohol on morphology of uniform polystyrene-poly(methyl methacrylate) composite microspheres prepared by porous glass membrane emulsiication technique. Journal of Colloid and Interface Science 219:110–28. Ma, G. H., Su, Z.G., Omi, S., Sundberg, D., and J. Stubbs. 2003. Microencapsulation of oil with poly(styrene-N,N-dimethylaminoethyl methacrylate) by SPG emulsiication technique: Effects of conversion and composition of oil phase. Journal of Colloid and Interface Science 266:282–94. McDonald, C. J., Bouck, K. J., and A. B. Chaput. 2000. Emulsion polymerization of voided particles by encapsulation of a nonsolvent. Macromolecules 33:1593–1605. Nelliappan, V., El-Aasser, M. S., Klein, A., Daniles, E. S., and J. E. Roberts. 1996. Compatibilization of PBA/PMMA core/shell latex interphase. II. Effect of PMMA macromonomer. Journal of Polymer Science, Part A: Polymer Chemistry 34:3183–90. Nie, Z., Xu, S., Seo, M., Lewis, P. C., and E. Kumacheva. 2005. Polymer particles with various shapes and morphologies produced in continuous microluidic reactors. Journal of American Chemical Society 127:8058–63. Nomura, M., Tobita, H., and K. Suzuki. 2005. Emulsion polymerization: kinetic and mechanistic aspects. In Advances in polymer science, vol. 175, Polymer particles, ed. M. Okubo, 1–128. Berlin-Heidelberg: Springer-Verlag. Okubo, M., and J. Izumi. 1999. Synthesis of micron-sized monodispersed, core-shell composite polymer particles by seeded dispersion polymerization. Colloids and Surfaces A: Physiochemical Engineering Aspects 153:297–304. Piirma, I. 1982. Emulsion polymerization. New York: Academic Press. Rajatapiti, P., Dimonie, V. L., and M. S. El-Aasser. 1997. Latex particle morphology: The role of macromonomers as compatibilizing agents. In Polymeric dispersions: Principles and applications, ed. J. M. Asua, 189–202. Drodrecht: Kluwer Academic Publishers.

58

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Rodríguez, R., Barandiaran, M. J., and J. M. Asua. 2008. Polymerization strategies to overcome limiting monomer conversion in silicone-acrylic miniemulsion polymerization. Polymer 49:691–96. Romio, A. P, Bernardy, N., Lemos-Senna, E., Araújo, P. H. H., and C. Sayer. 2009. Polymeric nanocapsules via miniemulsion polymerization using redox initiation. Materials Science and Engineering 29:514–18. Romio, A. P, Sayer, C., Araújo, P. H. H., Al-Haydari, M., Wu, L., and S. R. P. da Rocha. 2009. Nanocapsules by miniemulsion polymerization with biodegradable surfactant and hydrophobe. Macromolecular Chemistry and Physics 210:747–51. Schork, F. J., Luo, Y., Smulders, W., Russum, J. P., Butté, A., and K. Fontenot. 2005. Miniemulsion polymerization. In Advances in polymer science, vol. 175, Polymer particles, ed. M. Okubo, 129–255. Berlin-Heidelberg: Springer-Verlag. Sherman, R. L., Jr., and Ford, W. T. 2005. Small core/thick shell polystyrenepoly(methyl methacrylate) latexes. Industrial and Engingeering Chemistry Research 44: 8538–41. Stubbs, J. M., Karlsson, O., Jönsson, J.-E., Sundberg, E., Durant, Y., and D. C. Sundberg. 1999. Non-equilibrium particle morphology development in seeded emulsion polymerization. 1. Penetration of monomer and radicals as a function of monomer feed rate during second stage polymerization. Colloids and Surfaces A: Physiochemical and Engineering Aspects 153:255–70. Stubbs, J. M., and D. C. Sundberg. 2004. Nonequilibrium morphology development in seeded emulsion polymerization. III. Effect of initiator end group. Journal of Applied Polymer Science 91:1538–51. ———. 2005a. Measuring the extent of phase separation during polymerization of composite latex particles using modulated temperature DSC. Journal of Polymer Science, Part B: Polymer Physics 43:2790–2806. ———. 2005b. A round robin study for the characterization of latex particle morphology—multiple analytical techniques to probe speciic structural features. Polymer 46:1125–38. ———. 2006. Nonequilibrium morphology development in seeded emulsion polymerization. V. The effect of crosslinking agent. Journal of Applied Polymer Science 102:2043–54. ———. 2008. The dynamics of morphology development in multiphase latex particles. Progress in Organic Coatings 61:156–65. Sundberg, D. C., and Y. G. Durant. 2003. Latex particle morphology, fundamental aspects: A review. Polymer Reaction Engineering 11:379–432. Tanaka, T., Nakatsuru, R., Kagari, Y., Saito, N., and M. Okubo. 2008. Effect of molecular weight on the morphology of polystyrene/poly(methyl methacrylate) composite particles prepared by the solvent evaporation method. Langmuir 24:12267–71. Tiarks, K., Landfester, K., and M. Antonietti. 2001. Preparation of polymeric nanocapsules by miniemulsion polymerization. Langmuir 17:908–18. Trent, J. S. 1984. Ruthenium tetraoxide staining of polymers: New preparative methods for electron microscopy. Macromolecules 17:2930–31. van Herk, A. M. 2005. Chemistry and technology of emulsion polymerization. Oxford: Blackwell Publishing. van Herk, A. M., and M. Monteiro. 2002. Heterogeneous systems. In Handbook of radical polymerization, ed. K. Matyjaszewski and T. P. Davis, 301–2. New York: John Wiley & Sons.

Synthesis of Polymer Particles with Core-Shell Morphologies

59

Wang, C., Chu, F., Graillat, C., Guyot, A., Gauthier, C., and J. P. Chapel. 2005. Hybrid polymer latexes: acrylics-polyurethane from miniemulsion polymerization: Properties of hybrid latexes versus blends. Polymer 46:1113–24. Wang, S. T., Schork, F. J., Poehlein, G. W., and J. W. Gooch. 1996. Emulsion and miniemulsion copolymerization of acrylic monomers in the presence of alkyd resins. Journal of Applied Polymer Science 60:2069–76. Wu, X. Q., Schork, F. J., and J. W. Gooch. 1999. Hybrid miniemulsion polymerization of acrylic/alkyd systems and characterization of the resulting polymers. Journal of Polymer Science, Part A: Polymer Chemistry 37:4159–68. Zhao, K., Sun, P., Liu, D., and G. Dai. 2004. The formation mechanism of poly(vinyl acetate)/poly(butyl acrylate) core/shell latex in two-stage seeded semi-continuous starved emulsion polymerization process. European Polymer Journal 40:89–96. Zhenqian, Z., Yongzhong, B., Zhiming, H., and W. Zhixue. 2009. Preparation of polystyrene/poly(methyl methacrylate) core-shell composite particles by suspension-emulsion combined polymerization. Journal of Applied Polymer Science 111:1659–69.

3 Advanced Polymer Nanoparticles with Nonspherical Morphologies Yongxing Hu, Jianping Ge, James Goebl, and Yadong Yin CONTENTS 3.1 3.2

Introduction .................................................................................................. 62 Overview of Approaches to Nonspherical Polymer Nanoparticles Synthesis........................................................................................................63 3.3 Synthesis of Nonspherical Polymer Particles via Polymerization ........64 3.3.1 Phase Separation and Seeded Emulsion Polymerization ..........64 3.3.1.1 Mechanism .........................................................................64 3.3.1.2 Thermodynamics of Swelling and Phase Separation........................................................................... 66 3.3.1.3 Kinetics of Phase Separation ........................................... 68 3.3.1.4 Shape Control .................................................................... 69 3.3.1.5 Complex Structures through Multistep Polymerization................................................................... 70 3.3.1.6 Particles with Anisotropic Properties ............................ 73 3.3.2 Solvent Evaporation and Seeded Dispersion Polymerization ................................................................................. 76 3.4 Physical Posttreatment Approaches .......................................................... 78 3.4.1 Stretching .......................................................................................... 79 3.4.1.1 Simple 1D Stretching ........................................................ 79 3.4.1.2 Modiied Stretching Scheme ........................................... 81 3.4.1.3 2D Stretching .....................................................................83 3.4.2 Compression ..................................................................................... 86 3.4.3 Self-Assembly ................................................................................... 87 3.4.3.1 Self-Assembly in Soft Templates ..................................... 87 3.4.3.2 Self-Assembly in Hard Template .................................... 91 3.5 Summary....................................................................................................... 92 References............................................................................................................... 93

61

62

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

3.1 Introduction Polymer particles have been the subject of research in various ields for several centuries. The discovery of rubber latex, a class of naturally occurring colloidal dispersions of cis-polyisoprene particles, is usually regarded as the starting point of the development of modern polymer science (Whitby et al. 1933; Hohenstein et al. 1946). Driven by the need for rubber polymers, in the early twentieth century extensive research was carried out in order to ind synthetic routes to polymers by mimicking Mother Nature’s process for producing natural rubber (Hofmann 1912; Gottlob 1910, 1912; Pond 1914). In 1912, Kurt Gottlob invented the irst process that enabled the preparation of synthetic rubber latex by polymerization of isoprene in viscous aqueous solutions. The process involves the use of naturally occurring polymers such as gelatin, egg albumin, and starch to stabilize a dispersion of monomer droplets, which are then polymerized into solid particles. This process, now widely known as “emulsion polymerization,” has since become one of the dominant methods for synthesizing polymer particles from a wide variety of monomer materials. In most cases, the particles produced using emulsion polymerization or related processes have sizes within the range of 100 nm to 1 µm, although it is not uncommon to ind recipes for making particles beyond this range. In this chapter, the term “nanoparticle” is used to broadly cover particles with sizes of a few nanometers to several micrometers. Polymer nanoparticles play an important role in worldwide commerce as well as scientiic studies. Because of their versatile properties and applications, they are omnipresent parts of our daily life, with uses in paper making, paints and coatings, adhesives, textile processing, and also medical and pharmaceutical products (Xia et al. 2000). In fundamental research, poly mer nanoparticles have been used as an interesting model system for studying the behavior of atoms because many of the forces that govern the structure and behavior of matter, such as excluded volume interactions or electrostatic forces, govern the structure and behavior of polymer particle suspensions. For example, the phase transitions in liquid are analogous to those in polymer nanoparticle suspensions, which can be studied conveniently in real time using optical techniques. Due to the ease of making uniform samples with well-controlled sizes, recently poly mer nanoparticles have become popular candidates for sacriicial templates in fabrication of functional nanostructures such as core-shell and hollow spheres. Many procedures have been developed to coat polymer particles with various materials such as metal, semiconductor, or inorganic-organic hybrid layers to imbue them with new properties while at the same time retaining the sample uniformity. The high monodispersity that is achievable with polymer nanoparticles has also gained them an important role in the exciting ield of colloidal crystals. Similar to silica colloids, uniform polymer particles can self-assemble

Advanced Polymer Nanoparticles with Nonspherical Morphologies

63

into highly ordered three-dimensional arrays over a very long range (Sutera et al. 1980) to form artiicial crystals that appear analogous to their atomic or molecular counterparts. When the periodicity of the array is of the same order of magnitude as the wavelength of visible light, such colloidal crystals display brilliant iridescence, which can be attributed to the constructive interference of light waves as the result of interaction with the particle arrays. Macroporous materials have been produced by combining colloidal crystals with suitable structure-forming precursors, followed by selective removal of the polymer nanoparticle template from the solid composite by chemical or thermal methods. The resulting products are inverse replicas of the particle arrays, which can be viewed as three-dimensional structures of interconnected macropores with regular spacing and long-range order. Driven by the minimization of interfacial energy, the majority of poly mer nanoparticles have been limited to spherical shapes. Spherical colloids have been the dominant subject of research in colloidal science for many decades because of the ease of production as monodispersed samples. However, it has been recently realized that spherical particles are not necessarily the best option for fundamental studies or practical applications that are associated with these materials. For example, they have limited use for modeling the interparticle interactions and hydrodynamic behaviors of various irregular colloidal particles that are commonly found in industrial products. When they are used as building blocks in fabricating three-dimensional colloidal crystals, they can provide only a very limited set of crystal structures. Crystals assembled from spherical colloids can have only incomplete photonic bandgaps, which do not allow full control of the propagation of light in all three dimensions (Lu et al. 2001). In these regards, nonspherical colloidal particles are believed to offer some immediate advantages over their spherical counterparts in applications that require lower symmetries and higher complexities. For example, core-shell particles produced using nonspherical polymer cores will inherit the anisotropic shape, which in many cases adds an additional parameter for tuning the physical properties of the material. In the case of colloidal photonic crystals, it has also been remarked theoretically that a face-centered cubic colloidal crystal composed of nonspherical particles could possess a complete bandgap if the relative orientation of the nonspherical particles met certain requirements (Lu et al. 2001, 2002).

3.2 Overview of Approaches to Nonspherical Polymer Nanoparticles Synthesis Since a sphere is the energetically most favorable shape for liquid droplets, nonspherical particles are rarely seen as products in the standard

64

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

emulsion poly mer i zation processes. Interestingly, however, emulsion poly mer i zation still serves as the starting point for the synthesis of most nonspherical poly mer particles. Generally, there are two important approaches for synthesizing nonspherical poly mer nanoparticles: one refers to the direct synthesis, which involves swelling and phase separation processes in an emulsion poly mer i zation system; the other refers to the indirect approach, which utilizes the physical posttreatment of preformed spherical poly mer particles. Strictly speaking, the irst approach is also indirect because the initial products in the emulsion polymerization are spherical particles, which transform to nonspherical shapes following the additional swelling and polymerization steps. We classify it as a “direct synthesis” because in most cases these steps can be completed in one pot without the need for signiicant changes in the chemical environment. The transformation is due to differential interfacial tension between the preexisting cross-linked polymer particles and the absorbed monomer swelled in the polymeric matrix. By controlling the degree of cross-linking and the swelling processes, nonspherical particles with different asymmetries can be prepared. In the second approach, the spherical poly mer precursors are irst synthesized using standard emulsion processes and then transformed into different nonspherical shapes through posttreatment, which involves either physical deformation or controlled assembly processes. The posttreatment of preexisting structures includes a group of eficient methods that are able to yield particles with complex shapes/structures that are usually inaccessible by direct synthesis methods. In one of the typical indirect methods, preprepared particles irst can be heated above their glass transition temperature (Tg) and then deformed to form anisotropic particles with distinct shapes using physical techniques such as mechanical stretching. By controlling the direction and degree of deformation, a variety of shapes can be obtained. Another typical indirect method utilizes self-assembly approaches to organize polymer nanoparticles to form anisotropic structures and then applies a post physical or chemical method to ix the structures.

3.3 Synthesis of Nonspherical Polymer Particles via Polymerization 3.3.1 Phase Separation and Seeded Emulsion Polymerization 3.3.1.1 Mechanism Emulsion polymerization is a powerful method for the production of uniform spherical polymer particles. It is typically a radical polymerization

Advanced Polymer Nanoparticles with Nonspherical Morphologies

65

process that occurs in an emulsion incorporating water, monomer, and surfactants. The most common type of emulsion polymerization is an oil-inwater emulsion, in which droplets of monomer (the oil) are emulsiied with surfactants in a continuous phase of water. Since its discovery, this process has been widely employed for the preparation of monodispersed spherical nanoparticles from a wide range of polymer materials. However, its application in the synthesis of nonspherical particles is very dificult because the spherical shape is more favorable in terms of free energy. This limitation has been overcome by a seeded emulsion polymerization method, which was discovered by Ugelstad et al. (Skjeltorp et al. 1986) and further developed by Sheu, El-Aasser, and Vanderhoff (1990a, 1990b) and then Weitz, and Gilbert et al. (Kim et al. 2006, 2007; Mock et al. 2006). After ~30 years of development, it is now an effective method for synthesizing various polymer particles with anisotropic characteristics in both shape and physical properties. Essentially, this method utilizes the phase separation of swelling monomers from the cross-linked polymer seeds followed by polymerization at elevated temperature to produce nonspherical particles. A general mechanism for the phase separation and domain formation is discussed in this section, followed by a detailed discussion of its thermodynamics and kinetics, to present a clear irst image of the major procedures (Sheu et al. 1990a, 1990b). As shown in Figure 3.1, the synthesis generally starts with cross-linked polymer particle seeds, also called latex interpenetrating poly mer networks (IPNs), which can be prepared by conversional emulsion poly merization methods (Stage 1). Upon mixing with monomers, the poly mer seeds are gradually swollen until an equilibrium state is reached at room temperature, at which the monomer-polymer mixing force is balanced with the elastic force of the polymer network and the particle-liquid interfacial tension 1

2 ∆

+

6

3

5

4

FIGURE 3.1 Schematic illustration of the monomer swelling and phase separation in a seeded emulsion poly merization process. (Reprinted from H. R. Sheu, M. S. Elaasser, and J. W. Vanderhoff, Journal of Polymer Science, Part A: Polymer Chemistry 28: 629, 1990. With permission.)

66

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(Stage 2). With increasing temperature, the balance is broken due to the increased elastic-retractive force of the polymer chains, the decreased viscosity of the polymer phase, and thereby the enhanced mobility of monomers and short-chain linear polymers. As a result, the monomers and linear polymers are expulsed from the IPN upon heating and form a new phase domain adhered to the seeds (Stage 3). The new phase domain grows by absorption of monomer from the seed domain and is simultaneously polymerized until the forces are balanced again for both domains (Stages 3–5). Finally, nonspherical polymer particles with two well-deined domains are formed at the end of the polymerization (Stage 6). 3.3.1.2 Thermodynamics of Swelling and Phase Separation It has been demonstrated that raised temperature will enhance the phase separation of monomers and linear polymers from an interpenetrating polymer network, thus forming a liquid bulge adhered to the seed particle and further polymerizing into nonspherical particles. Sheu and El-Aasser et al. have developed a thermodynamic model to explain the swelling and phase separation through a comprehensive consideration of the monomer-polymer mixing force, the elastic retraction from the stretching of chains in crosslinked polymer networks, and the surface tension at the particle-solution interface (Sheu et al. 1990a). This model is built based on a physical fact that the chemical potential ΔGm,p of monomer in the particle phase is equal to that ΔGm,a in the aqueous phase when the swelling is at the equilibrium state. For monomers like styrene, whose solubility in water is extremely low, ΔGm,a is often negligible such that ΔGm,p approximately equals zero at the balanced state. The chemical potential of monomer in the particle phase can be considered as the sum of three contributions, including the monomer-polymer mixing force ΔGm, the elastic force of cross-linked polymer network ΔGel, and the surface tension at the particle-water interface ΔGt. ∆G m , p = ∆G m + ∆G el + ∆Gt = 0

(3.1)

Here, ΔGm = RT[ln(1 – νp) + νp + χmpνp2] according to the Flory-Huggins expression, ΔGel = RTNVm(νp1/3 – νp/2) according to the the Flory-Rehner equation, and ΔGt = 2Vmγ/r according to the Morton equation, R is the gas constant, T is the absolute temperature, νp is the volume fraction of polymer in the swollen particle, χmp is the monomer-polymer interaction parameter, N is the effective number of chains per unit volume, Vm is the molar volume of the monomer, γ is the particle-water interfacial tension, and r is the radius of the swollen particles. In a typical instance, where a cross-linked polystyrene (PS) seed 5.2 μm in diameter is swelled with styrene, the values of ΔGm/RT, ΔGel/RT, and

Advanced Polymer Nanoparticles with Nonspherical Morphologies

67

0.01 0 ∆Gt RT

∆Gm/RT, ∆Gel/RT, and ∆Gt/RT

–0.01 ∆Gm RT

–0.03

∆Gel RT

–0.05

–0.07

–0.09

–0.11 0

0.10

0.30

0.20

0.40

0.50

VP FIGURE 3.2 Relationship between the poly mer volume fraction and the changes of each partial molar free energy (ΔGm/RT, ΔGel/RT, ΔGt/RT). (Adapted from H. R. Sheu, M. S. El-Aasser, and J. W. Vanderhoff, Journal of Polymer Science, Part A: Polymer Chemistry 28: 629, 1990. With permission.)

ΔGt/RT can be plotted as a function of polymer volume fraction νp, respectively (Figure 3.2). The mixing force ΔGm makes a negative contribution to the overall chemical potential, promoting the particle expansion and thereby further swelling of monomers. On the contrary, both the elastic force ΔGel and interfacial force ΔGt make positive contributions, which restrain the stretching of polymer chains and expansion of particles. For particles larger than 2 μm, the interfacial force is usually negligible compared to the elastic force, so that the swelling and phase separation are actually determined by the other two parameters. This thermodynamic model qualitatively explains the driving force of monomer swelling and phase separation. When the polymer seeds are mixed with a saturated solution of monomers, the latter is continuously absorbed into the former because the high initial polymer volume fraction νp in the polymer seeds results in negative chemical potential ΔGm,p that encourages the polymer seeds’ expansion. With further progress of swelling, the polymer volume fraction νp decreases, and eventually the swelling stops when the chemical potential ΔGm,p reaches zero at room temperature. As the temperature is raised, the chemical potential ΔGm,p immediately increases and exceeds zero due to the increased retractive elastic force ΔGel, which contracts the swollen network and expels the monomer to form a new phase domain attached to the seed. The reduction of viscosity and enhancement of

68

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

mobility upon heating in the polymer network also contribute to this phase separation process. The small new domain contains monomers at a higher concentration than that of the seed domain, and thereby possesses a higher polymerization rate and higher polymer volume fraction νp. The resulting low free energy of mixing and negative chemical potential ΔGm,p for the new domain facilitates the migration of monomer and further polymerization. Finally, the transportation of monomer ceases because of the decreased difference in monomer concentration in both domains, eventually resulting in polymer doublets. 3.3.1.3 Kinetics of Phase Separation Sheu and El-Aasser et al. have proven that the kinetics of phase separation are different for PS seeds with different degrees of cross-linking (Sheu et al. 1990a), as they noticed that lightly cross-linked seeds generally produced doublets, whereas highly cross-linked ones produced triplets and multiplets. In the case of lightly cross-linked PS seed (0.2% divinylbenzene [DVB]), the particle swollen with monomer retains its spherical shape throughout the swelling process, and its size stops increasing after 24 h (Figure 3.3). After the temperature is raised to 70°C for 2 min, a new domain appears due to the heat-induced retraction of polymer chains and expulsion of monomers. The new domain keeps growing while the seed shrinks, and they reach the same size after 177 min, forming a symmetric doublet. After 738 min, the new domain grows larger than the seed domain, which shrinks to a size slightly larger than its original dimensions. Finally, the polymerization is completed after 1440 min with 100% conversion, producing asymmetric doublets (Figure 3.3h). By substituting monomer with toluene, Sheu and El-Aasser et al. conirmed that the phase separation is initiated by the a

b

c

d

e

f

g

h

20 µm

FIGURE 3.3 Optical microscopy images of 5.2-μm PS particles cross-linked with 0.20% DVB and swollen with styrene at 23.2°C for 1440 min (a), and the nonspherical particles after poly merization at 70°C for (b) 2 min, (c) 77 min, (d) 97 min, (e) 177 min, (f) 390 min, (g) 738 min, and (h) 1440 min. (Reprinted from H. R. Sheu, M. S. El-Aasser, and J. W. Vanderhoff, Journal of Polymer Science, Part A: Polymer Chemistry 28: 629, 1990. With permission.)

Advanced Polymer Nanoparticles with Nonspherical Morphologies

69

increase in temperature at the beginning of the polymerization, which is a quick process, and is further enhanced by the conversion from monomer to polymer in the new domain as the reaction proceeds. These two processes can be summarized as elastic-driven and polymerization-driven phase separation, respectively (Kim et al. 2006). Therefore, the entire phase separation can be considered as a burst in new domain generation, followed by further domain growth due to the monomer transportation and polymerization. In the case of highly cross-linked PS seeds, there are two remarkable characteristics that differ from the kinetic process with lightly cross-linked seeds. One is the early phase separation during the swelling process, which is believed to be caused by a progressive contraction of the polymer network as the chains are highly cross-linked. The second feature is the formation of multiple domains and multiplets. Once the temperature is raised, the new phase domain continues growing and some particles form additional domains. After polymerization, a mixture including some doublets, triplets, multiplets, and small free particles is produced. The formation of multiple domains is promoted when the degree of cross-linking in seeds is increased. This difference is attributed to the larger gel fraction and the formation of inhomogeneities in the highly cross-linked seed particles, which also localize the contraction because the diffusion of monomer is not fast enough to uniformize the monomer distribution in the polymer network before the creation of new domains. 3.3.1.4 Shape Control Various parameters including the degree of seed cross-linking, the monomer:polymer swelling ratio, temperature, seed size, and cross-linker level in swelling monomer contribute to the control over the shape of the nonspherical polymer particles (Sheu et al. 1990b). As summarized by Sheu and El-Aasser et al., the inal products may vary from uniform spheres to ellipsoids, asymmetric and symmetric doublets, and multiplets through the manipulation of these parameters. Degree of seed cross-linking. As mentioned previously, the degree of seed cross-linking determines the elastic retraction and polymer chain conformation, which is critical for the proceeding of phase separation and homogeneity of swelling. Seeds with minimal cross-linking (0.06%) grow into larger spherical particles with no phase separation because the slightly increased elastic retraction upon heating is not strong enough to expel the monomer to form new phase domains. As the cross-linking increases (0.2–1.5%), the elastic retraction enhances upon heating and leads to phase separation, which eventually produces symmetric or asymmetric doublets. Besides these thermodynamic considerations, in a kinetic sense, a higher cross-link density indicates a larger fraction of the gel portion and a shorter polymer chain relaxation time, which also promote phase separation. As discussed earlier, when the seeds are highly cross-linked (2.0–6.0%), inhomogeneous swelling and localized contraction lead to the production of multiplets.

70

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Monomer:polymer swelling ratio. The degree of phase separation and the size of new phase domains increase with higher monomer:polymer swelling ratios, since the new domains grow by the absorption and polymerization of monomers transported from the seed domain until the chemical potentials of monomer in both domains become balanced. With a monomer:polymer ratio of 1:1, no phase separation is observed and spherical particles are produced. A monomer:polymer ratio of 2:1 gives asymmetric doublets with a smaller new domain. Increasing the monomer:polymer ratio to 3:1 and 4:1 will produce symmetric doublets and asymmetric doublets with a larger new domain, respectively. Temperature. The reaction temperature is also a critical parameter for phase separation and polymerization. From a thermodynamic point of view, a higher polymerization temperature will cause a great increase in the retractive elastic force upon heating and a larger separated phase domain. In the kinetic sense, a higher temperature will increase the polymer chain relaxation time and lower the viscosity to accelerate the formation and growth of a new phase domain. It has indeed been observed that particles polymerized at 50°C grow to asymmetric doublets with smaller new domains, while those formed at 70°C contain an enlarged second domain. Size of seed particles. The size of the seed particles is important to the thermodynamics of phase separation because it directly affects the surface tension at the interface of the seed and the aqueous solution. According to the Morton equation ΔGt = 2Vmγ/r, a larger seed particle size leads to a smaller interfacial tension, which certainly enhances the phase separation. The 0.6-μm seed particles grow to spherical particles without phase separation, while 1.9-μm, 5.2-μm, and 8.1-μm seed particles change to uniform ellipsoids, symmetric doublets, and asymmetric doublets, respectively. Note that even for submicron scale seeds, it is still possible to create nonspherical particles when the degree of cross-linking of seed particles and the monomer:polymer swelling ratio are suficiently high. Cross-linker level of swelling monomer. Polymerization with no cross-linker (0% DVB) generates a latex semi-IPN, and with cross-linker produces a latex full-IPN as both the seed domain and new domain are cross-linked. The new domain shrinks as the cross-linker level increases in the monomer mixture, thus producing doublets with a smaller new domain. This occurs because the new domain with higher cross-linking has higher viscosity and stronger elastic retraction upon expansion, both of which limit the transportation of monomers from the seed domain. 3.3.1.5 Complex Structures through Multistep Polymerization In the previous section, the synthesis of basic nonspherical particles through swelling, phase separation, and poly merization was discussed. However, it is of great interest to chemists and materials scientists to explore the possibility of fabricating more complex nonspherical poly mer structures

Advanced Polymer Nanoparticles with Nonspherical Morphologies

71

through this “seeded emulsion poly merization” scheme. Recently, Weitz et al. advanced this technique by synthesizing a variety of uniform complex particles on a large scale through the tight control of the cross-linking density of the poly mer network (Kim et al. 2007). Their synthesis starts with symmetric polymer doublets with controllable degrees of cross-linking in each domain, and utilizes the gradients in cross-linking density as the driving force to manipulate the direction of phase transfer, thus leading to the production of rod-, cone-, triangle-, and diamond-shaped particles after poly merization. The entire synthesis comprises three continuous seeded emulsion poly merizations, and the cross-linking degree of the seed and second domain of the doublets are adjusted by changing the concentration of cross-linker in the poly merization of spherical seeds ([DVB]a) and doublets ([DVB]b), respectively. Since the IPN can be perfectly swelled in toluene, one can observe the swelling of dimers in toluene using optical microscopy and obtain the crosslinking density (ν) from the ratio of unswelled volume to the swelled volume (φ). According to the theory of Flory and Rehner, the cross-linking density can be calculated by the following equation, where NA is Avogadro’s number, V is the molar volume of the solvent, and χ is the polymer-solvent interaction parameter. The dependence of cross-linking gradient (Δν) on the cross-linker ratio ([DVB]b:[DVB]a) can be plotted to investigate the inluence of cross-linking on the shape. It should be noted that in dimer particles, the cross-linking degree in the new domain (bulb b) can be estimated based on the concentration of cross-linker ([DVB]b) in the second reaction, where a small deduction must be considered as the polymerization also occurs in the original domain (bulb a). Accordingly, the cross-linking degree in the original domain (bulb a) is slightly larger than the one quantiied with [DVB]a. ν=

NA  ln(1 − φ) + φ + χφ 2  φ 1/3 − φ/2 V 

(3.2)

As shown in Figure 3.4, the gradient of cross-linking density (Δν) in dimers is critical to the shape of multiplets. IPN with high cross-linking density has been proven to generate stronger elastic retraction upon heating, thus facilitating the phase separation and new domain formation. When the degree of cross-linking of the second domain is as large as that of the original (ν b ≈ νa, Δν = 0.9 mol m–3), the elastic retractions in both domains are close enough that the new separated domain grows perpendicularly between the two bulbs, forming triangle-shaped particles. Decreasing the cross-linking density of the second domain in dimer seeds (ν b < νa, Δν = 15 mol m–3) will break the balance of elastic retraction upon heating, as the driving force of pushing the monomer mixture out of the original domain is stronger than that in the second domain. In other words, the gradient in cross-linking density drives the monomer mixture to low down the gradient, leading to linear growth

72

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(a)

(b)

(c)

(d)

45 ∆ν = νa – νb (mol  m–3)

40 35

5 µm

30 25 20 15 10 5 0 0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

[DVB]b/[DVB]a (e) FIGURE 3.4 (a–d) SEM images of (a) triple-rod particles prepared at [DVB]b/[DVB]a = 0.7; (b) triangular particles at [DVB]b/[DVB]a = 1.1; (c) cone particles made similar to the triple-rod particles except with a higher volume of monomer mixtures in the inal swelling stage; (d) diamond-shaped particles made similar to the triangular particles but with 20% more total cross-linker; and (e) cross-linking density gradient (Δν) tuned by changing the concentration of DVB in each swelling step ([DVB]b and [DVB]a). Insets show the optical microscope images of dimer particles before (upper) and after (lower) swelling at different [DVB]b:[DVB]a ratio, and the schematic images of the inal particles grown from these dimers. (Adapted from J. W. Kim, R. J. Larsen, and D. A. Weitz, Advanced Materials 19: 2005, 2007. With permission.)

Advanced Polymer Nanoparticles with Nonspherical Morphologies

73

and rod-shaped particles. Further decreasing the cross-linking density of the second domain in dimer seeds (ν b < νa, Δν = 37 mol m–3) will seriously weaken the elastic retraction, such that no phase separation occurs and asymmetric dimers are produced. When Δν is tuned between these critical values, mixtures of two types of particles are generated after poly merization. This procedure represents a general method for the preparation of uniform and high-yield polymer triplets in a one-batch reaction, and also holds the potential for chemical fabrication of more complex nonspherical polymer particles. Modiications to the recipes for rod- and triangle-shaped particles will produce additional complex particles. For instance, the cone-shaped particles can be produced similarly to the rod particles, but with a higher volume of monomer mixtures in the inal swelling stage; while the diamondshaped particles can be synthesized similarly to the triangular particles, but with higher cross-link densities in both domains of dimer seeds ([DVB]a = 1.2 vol%, [DVB]b = 1.32 vol%). 3.3.1.6 Particles with Anisotropic Properties The preceding sections have discussed seeded emulsion polymerization as an effective strategy to produce various nonspherical polymer particles. Here, we will introduce polymer colloids with not only nonspherical shapes but also anisotropic physical properties. By surface modiication and other chemical treatments, or incorporation of inorganic functional materials, this method can be extended to the preparation of polymeric particles with anisotropic properties, such as amphiphilic particles and magnetically anisotropic particles. Particles with chemical rather than shape anisotropy play an important role in recognizing speciic molecules, self-assembling, and emulsion stabilization. One particular kind of useful anisotropy of chemical property is “amphiphilicity,” which is generally induced by organic groups or polymeric blocks. Recently, Weitz et al. have reported a lexible approach, based on phase separation and seeded emulsion poly merization, for the synthesis of uniform and amphiphilic nonspherical particles (Kim et al. 2006). The experimental procedure includes the synthesis of cross-linked poly(styrene-co-glycidyl methacrylate) (PS-co-PGMA) spherical particles, PS-co-PGMA/PS symmetric polymer doublets, and surface treatment with poly(ethylene imine) (PEI). Glycidyl methacrylate is copolymerized to the original PS seed to provide epoxy rings on the surface, which offer reactive sites for hydrophilic chemicals that have active hydrogen. Through swelling by styrene monomer solution, phase separation upon heating, and thermal polymerization, the spherical seeds turn into PS-co-PGMA/PS doublets. It is important that the PGMA chains stay in the original domain during phase separation so that the distribution of surface epoxy rings is extremely asymmetric in the doublets. The half-sphere functionalized with the epoxy rings is then reacted with hydrophilic PEI, turning the hydrophobic doublets into

74

a

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

b

c

Hydrophilic PEI PS PEI Hydrophobic PS 10 µm

10 µm

FIGURE 3.5 (a) SEM image of amphiphilic PS doublets obtained by the reaction of surface epoxy rings with PEI. (b) Fluorescence microscope image of luorescein-labeled amphiphilic PS doublets. (c) Bright-ield microscope image of the assembly of amphiphilic PS doublets at water/octanol interface. (Reprinted from J. W. Kim, R. J. Larsen, and D. A. Weitz, Journal of the American Chemical Society 128: 14374, 2006. With permission.)

amphiphilic structures. The PEI-coated part can then be labeled with luorescein sodium salt and imaged using luorescence microscopy to conirm the exact amphiphilic structure (Figure 3.5). Compared with the spherical amphiphilic Janus particle, these particles with both a nonspherical shape and amphiphilic properties may be more suitable to stabilize the interfaces and emulsion droplets. As shown in Figure 3.5, the water droplets inside 1-octanol can be stabilized by these colloid surfactants. The interface curvature and droplet size are expected to be conveniently tuned by controlling the geometry of the doublets. Besides selective surface modiication to the premade nonspherical particles, incorporation of functional content into the polymer particle through core-shell fabrication is an alternative way to realize anisotropic properties. A series of spherical and nonspherical magnetite-PS composite colloids with anisotropic magnetic properties has been reported (Ge et al. 2007). In a typical process, 3-(methacryloyloxy)propyl trimethoxysilane (MPS)-modiied Fe3O4@SiO2 particles are used as seeds to prepare concentric and eccentric Fe3O4@SiO2@PS core-shell particles through seeded emulsion polymerization. The location of the Fe3O4@SiO2 core is determined by the degree of crosslinking in the polymer layer: the eccentric particles are prepared without the addition of DVB while concentric particles are obtained when the polymers are cross-linked. The formation of eccentric structure is due to the interfacial tension between the hydrophilic seed particle and the hydrophobic monomer (Park 2001; Minami et al. 2003; Mock et al. 2006). At the initial stage of polymerization, a thin layer of polystyrene shell is deposited in the form of small particles on the silica surface through copolymerization with the surface double bonds offered by MPS molecules. After absorbing hydrophobic monomers, the polystyrene chains in the shell tend to contract and reduce the surface area as a result of interfacial tension, leading to asymmetric distribution of polymer around the seeds (Figure 3.6a). Modiication with MPS renders the silica surfaces considerable compatibility with the swollen polymer shell so that the inal products still maintain an overall spherical shape

Advanced Polymer Nanoparticles with Nonspherical Morphologies

a

b

c

d

e

f

g

75

h

400 nm

FIGURE 3.6 TEM images of (a) eccentric and (b) concentric spherical particles produced in one-step emulsion poly merization, and evolution of doublets (c–f) prepared by swelling with particle b, phase separation, and poly merization. Self-assembly of (g) concentric and (h) eccentric particles into chain structures in a magnetic ield. (Reprinted from J. P. Ge, Y. X. Hu, T. R. Zhang, and Y. D. Yin, Journal of the American Chemical Society 129: 8974, 2007. With permission.)

with the majority part of the inorganic core engulfed in the polymer phase. The high viscosity of the monomer-swollen shell, achieved by introducing a cross-linker of DVB during polymerization, limits the transportation of monomers and leads to the formation of a concentric core-shell structure of Fe3O4@SiO2@PS (Figure 3.6b). The spherical symmetry of the external interface can be broken as a result of phase separation when monomer concentration is increased, forming ellipsoidal colloids with a magnetic core eccentrically located at one end. The considerable effects of swelling and phase separation are responsible for the ellipsoidal shape of the colloids. At the early stages of poly merization, the particles have a spherical and concentric morphology similar to that in Figure 3.6b. As the thickness of shell polymer increases, more excessive styrene monomers are absorbed into the cross-linked polymer networks. In the inal stage, the elastic stress driven by the entropy change of the swollen networks causes phase separation of the monomer from the network and eventually forms additional bulges attached to the original particles. As discussed in previous sections, phase separation is negligible when the starting monomer concentration is low so that the inal products have a spherical shape. Further increasing the concentration of monomers induces phase separation and forms ellipsoidal particles. When the swelling and phase separation processes are performed in two separate steps, nonspherical colloids can be produced with a higher degree of control over the shape asymmetry. Starting with the cross-linked concentric core-shell particles, a series of nonspherical poly mer particles with controllable size of the new domains can be prepared by taking advantage of the well-developed swelling and phase separation processes in seeded emulsion polymerization, where the relative size of the new domain to the original

76

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

inorganic sphere changes dramatically as the monomer:seed ratio increases (Figure 3.6c–f). The combination of anisotropic shapes and magnetic properties makes these composite colloids new types of building blocks for constructing secondary structures. The composite spheres with concentric and eccentric superparamagnetic cores display distinct assembly behaviors when they are subjected to an external magnetic ield, as shown in Figure 3.6g and h. The concentric particles form simple linear chains, which are expected for building blocks with simple isotropic conigurations, while more complicated zigzag chains have been observed when the eccentric particles are assembled under similar conditions. The zigzag coniguration is a direct result of the eccentric structure: the strong magnetic dipole interactions that pull the magnetic cores in the neighboring colloids to the closest positions. 3.3.2 Solvent Evaporation and Seeded Dispersion Polymerization The combination of seeded dispersion polymerization (SDP) and solvent evaporation represents a unique kinetic-based technique for the preparation of nonspherical polymer particles, whose glass transition temperature (Tg) usually exceeds the polymerization temperature so that the viscosity within the particles is high enough to prevent equilibrium morphologies from being obtained. Recently, Okubo et al. have developed an SDP method for the preparation of nonspherical particles with the aid of organic solvents (Fujibayashi et al. 2007; Okubo et al. 2002). The solvent molecules are selectively absorbed into certain particle domains, whose volume decreases dramatically as the solvent evaporates after polymerization, leading to the formation of golfball-like, polyhedral, and disk-like particles. A typical system involves the SDP of 2-ethylhexyl methacrylate (EHMA) with PS seeds in the presence of various hydrocarbon solvents. The key to controlling the shape of the polymer particles lies in the absorption of hydrocarbon solvent into certain domains and their coalescence during the reaction. As the dispersion polymerization proceeds, both the PS and PEHMA domain absorb the hydrocarbon solvent, but the PEHMA/solvent domain grows much larger than the PS/solvent domain. The absorption of hydrocarbon solvent signiicantly increases the particle volume and decreases the viscosity of the polymer network, which accelerates the coalescence of the PEHMA/ solvent domain by collision and Ostwald ripening. Thermodynamically, the composite particle prefers to contain as few PEHMA/hydrocarbon domains as possible because the reduction of the interface between PEHMA/solvent and PS phase will decrease the interfacial free energy. However, the practical degree of domain coalescence is always affected by several kinetic considerations, such as the viscosity of the PS/solvent domain and the reaction time. Finally, the solvent evaporates and the PEHMA/solvent domain shrinks to produce nonspherical particles. The degree of coalescence of the PEHMA/ solvent domain determines whether the products are spheres, hemispheres,

77

Advanced Polymer Nanoparticles with Nonspherical Morphologies

(a)

(b)

2 µm (d)

(c)

2 µm (e)

2 µm

2 µm (f )

2 µm

2 µm

FIGURE 3.7 SEM images of PS particles after extraction of PEHMA and decane, prepared by seeded dispersion poly merization at various conversions of (a) 5%, (b) 13%, (c) 34%, (d) 49%, (e) 57%, and (f) 69%. (Reprinted from T. Fujibayashi and M. Okubo, Langmuir 23: 7958, 2007. With permission.)

golf balls, polyhedra, or disks. Any experimental parameter with inluence on the absorption of hydrocarbon and the coalescence of the PEHMA/solvent domain can be used to control the morphology of the nonspherical particles (Fujibayashi et al. 2007; Okubo et al. 2002). Reaction time. Particles with various percent conversions of EHMA collected at different reaction times show dramatically different morphologies (Figure 3.7). Initially, spherical particles with dents are observed after 5% conversion of EHMA. The number of dents on the surface increases as the polymerization proceeds, which leads to the formation of golf-ball-like particles at a conversion of 13%. This result can be explained by the increasing number of PEHMA/solvent domains as the EHMA is gradually polymerized. The particles then keep on changing their shape to polyhedron at an EHMA conversion of 34 to 57%, as the PEHMA/solvent domains coalesce into several larger domains. Upon further coalescence, the PEHMA/solvent domains eventually combine into two large hemispherical domains at after 69% conversion, leaving a PS disc-like domain sandwiched in between to form a hamburger-like structure. Time-dependent hydrocarbon absorption and domain coalescence have great inluence on the particle shapes produced by SDP. Type of hydrocarbon. PS/EHMA composite particles with different shapes can also be synthesized by introducing various hydrocarbon solvents. By

78

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

decreasing the length of the alkyl chain of the hydrocarbons, the particle shapes change from golf ball (hexadecane) to polyhedron (tetradecane, dodecane), and then to disks (decane, octane, hexane) at the same reaction conditions. Gas chromatography studies indicate that the amount of hydrocarbon absorbed by the PS particles increases in the order of hexadecane, tetradecane, dodecane, decane, octane, and hexane, suggesting that shortchain hydrocarbons have a higher afinity for the PS. DSC heating curves of swollen PS particles also show that the Tg shifts to a lower zone as the alkyl chain length decreases, which means the viscosity of the PS/solvent domain decreases accordingly. For PS particles swollen with short-chain hydrocarbons, both the increased amount of absorbed solvent and decreased viscosity promote the enlargement and coalescence of PEHMA/solvent domains, leading to the shape transformation from golf balls to disks. Methanol:water ratio. The aforementioned conclusions are consistent with the results when particles are prepared in methanol/water solutions with different ratios. A high methanol:water ratio (90:10 w/w) enhances the absorption of hydrocarbons so that a disk is the preferential shape. On the other hand, a low methanol:water ratio (70:30 w/w) leads to polyhedral particles. Amount of solvents. The inluence of the amount of solvent has been investigated through a different reaction system, where particles are prepared by the SDP of poly(methyl methacrylate) (PMMA) seeds, styrene, and decalin (solvent droplet) in a mixture of methanol and water. In this case, all the particles are golf-ball-like and no transformation to disks is observed, probably due to thermodynamic reasons for the speciic domains of PMMA/ decalin and PS/decalin. By increasing the decalin, the number of surface dents decreases, and the domains enlarge by the coalescence of neighboring domains. This result is consistent with the expectation that increasing the amount of solvent will enhance the solvent absorption and domain coalescence, making the dents on the golf balls fewer in number but larger in size.

3.4 Physical Posttreatment Approaches In addition to direct chemical synthesis, indirect approaches by posttreatment of preexisting structures have also been widely used to prepare particles with complex nonspherical shapes, including those that are usually inaccessible by direct methods. Spherical polymer particles are usually synthesized irst using chemical methods such as emulsion polymerization, and then posttransformed into nonspherical structures by either physical or chemical means such as stretching and self-assembly (Champion et al. 2007; Courbaron et al. 2007; Hu et al. 2008; Jiang et al. 2001).

Advanced Polymer Nanoparticles with Nonspherical Morphologies

79

3.4.1 Stretching 3.4.1.1 Simple 1D Stretching The most straightforward indirect method involves one-dimensional (1D) physical deformation of poly mer nanoparticles embedded in polymeric matrices at elevated temperatures. It was irst developed by Felder while stretching a polyvinyl alcohol (PVA) ilm to produce monodisperse ellipsoidal dye-containing polyvinyl acetate microparticles embedded in the matrix (Felder 1966, 1968). Figure 3.8 shows the schematic procedure of this indirect method that employs uniaxial viscoelastic deformation to convert colloidal particles from spherical to ellipsoidal shapes (Keville et al. 1991). In this method, monodisperse, spherical particles of a polymer—for example, PS, PMMA, and poly(vinyl toluene)—are irst dispersed in a precursor solution of elastic matrix such as PVA or un-cross-linked poly(dimethylsiloxane) PDMS. An elastic ilm is formed after solvent evaporation or poly merization of the matrix solution. By stretching the ilm at a temperature higher than the Tg of the poly mer latexes, spherical particles can be transformed into uniform ellipsoids. The softened poly mer spheres deform in the elongated matrix ilm and become ellipsoids whose aspect ratio can be controlled by changing the magnitude of uniaxial elongation and the ratio between the elastic moduli of the matrix and the poly mer spheres (Ho et al. 1993). In the case of PS particles in PVA, the particle deformation is enabled by the strong adhesion arising from the hydrogen bond between the hydroxyl groups on the PVA and the surface groups on the PVA PS

M

at

rix

di ss

ol

ut

io

n

St re tch

FIGURE 3.8 Schematic illustration of stretching polyvinyl alcohol ilms with embedded particles.

80

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

latex particle surface (Champion et al. 2007; Ho et al. 1993). The ellipsoidal shape resulting from the viscoelastic deformation can be completely retained by cooling down the sample below the Tg of the poly mer sphere, while the matrix remains constrained in the elongated state. These ellipsoidal particles can then be collected as a stable dispersion by dissolving the matrix with an appropriate solvent. To obtain well-deined microspheroids, the polymer particles should possess good chemical and physical stability to be able to withstand the subsequent heating/solvent treatment and be able to remain colloidally stable in the media used in the various synthesis steps. The matrix, which consists of cross-linked elastomer or thermoplastics with controlled elastic properties and high ultimate elongation, can determine the maximum aspect ratio of the obtained ellipsoids. The elastic modulus of the matrix can be adjusted by controlling the cross-linking density. With higher cross-linking density, the ultimate elongation is decreased and stress relaxation is less pronounced. A number of thermoplastics or elastomers have been found to be good elastic matrixes with typical examples including PVA and cross-linked PDMS. The difference between an elastomer and a thermoplastic is that an elastomer tends to relax back to the unstretched state, while a thermoplastic will remain in the stretched state forever once the sample has been cooled. In the ideal case, if the elastic properties of the particles and the matrix are identical and the materials are pure elastic, the aspect ratio could be predicted by using the equation suggested by Eshelby et al.: 3

1 = (1 + e) 2 , d

(3.3)

where l is the length of the major axis of the spheroid, d is the diameter of its cross section normal to the axis, and e is the axial strain e=(l-l0)/l0 (Eshelby et al. 1957). If the elastic properties of the particles are not identical to that of the matrix, the resulting aspect ratio will also depend on the ratio of elastic moduli, which will be predicted as

1 = d

  1 1+     = 0.4(1 − r ) 

e

1   1 + (1 + e) 2 − 1 1 − 0.4(1 − r )  

,

(3.4)

where r is the moduli ratio, r=G1/G, and G and G1 are the elastic moduli of the matrix and the particles. For matrix materials with nonlinear elastic or viscoelastic properties, or when contribution of interfacial energy is signiicant, the aspect ratio will deviate from the predictions. Moreover, it should be

Advanced Polymer Nanoparticles with Nonspherical Morphologies

81

taken into account that the properties of the matrix embedded with polymer nanoparticles might be different from those of the pure material. The inal shape of the poly mer particles depends on a few key parameters, including the thickness of the ilm, the viscosity of a liqueied droplet, and the temperature. A practical problem of this simple stretching approach is that the uniaxial stretching is not always uniform throughout the whole ilm due to the formation and propagation of necks at the center of the poly mer strip. To improve the uniformity of the recovered particles, strips of the ilm around the center with lengths corresponding to the expected draw ratio are cut out from the ilm and used for recovery of the particles. This indirect 1D deformation approach has been investigated by a number of research groups for generating small quantities of micrometer-sized ellipsoids of polymers for light scattering, rheological, or other fundamental studies. Ellipsoidal PS particles of a few micrometers with predesigned axial ratios were reported by Nagy and Keller through the afine deformation of uniform spherical PS beads in a deformable matrix (Ho et al. 1993). PMMA spheroids produced by deforming an elastic matrix of PDMS were later fully studied by Keville et al. (1991). Lu et al. (2002) also exploited the use of this method in generating long-range ordered three-dimensional (3D) colloidal crystals with polymer ellipsoids as building blocks. Xu et al. (2007) prepared hollow ellipsoids with desired aspect ratios by stretching a PVA ilm embedded with PS hollow spheres. The hollow ellipsoids can be used further as templates for the synthesis of composite ellipsoids. 3.4.1.2 Modified Stretching Scheme The 1D stretching scheme outlined previously allows only for the production of simple ellipsoidal particles. Modiication of the stretching/heating sequence or using a liquefaction method makes it possible to produce particles with highly complex shapes. As demonstrated by Champion et al. (2007) in their recent reports, this modiied stretching protocol can be used to generate particles of more than 20 distinct shapes including ellipsoids with lat ends, rods with circular cross section, rec tangular disks, and sharp-ended worm-like particles. Figure 3.9 shows the general procedure, which can be categorized into two schemes. The irst modiication involves the liquefaction of particles by using toluene instead of heating, and then solidiication by solvent evaporation after stretching. As indicated previously, the particle viscosity is a critical parameter that dictates whether a deformed particle will have pointed or lat ends and whether the original width is preserved. Since the liquefaction of the particles using toluene leads to a much lower viscosity than when they are heated under otherwise similar conditions, it is possible to produce nonspherical particles with different shapes. With 1D stretching under this condition,

82

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Scheme A

Liquefy

Stretch

Solidify

Stretch

Liquefy

Solidify

Scheme B

FIGURE 3.9 Methods used for making particles with different shapes can be categorized into two general schemes. (Reprinted from J. A. Champion, Y. K. Katare, and S. Mitragotri, Proceedings of the National Academy of Sciences of the United States of America 104: 11901, 2007. With permission.)

thin elliptical disks with curved ends can be obtained, while lat circular disks can be fabricated with two-dimensional (2D) stretching. The second modiication involves changing the stretching/heating sequence. In this case, the elastic matrix embedded with particles is irst deformed so that voids are created around the particles. Liquefaction of the particles by controlled heating above the Tg in the presence of voids allows the production of many additional complex shapes (Figure 3.10). During liquefaction, softened PS will ill the voids by wetting the ilm (Figure 3.10e–f). At low liquefying temperature (~130°C), the high viscosity of PS inhibits complete illing of the voids and leads to barrel-like particles. At higher temperatures (~140°C), PS wets the PVA matrix better due to its decreased viscosity, and leads to the illing of one end of the void, eventually forming bullet-like structures. Moreover, by liquefying the PS particles with toluene, more different shapes will be formed as a result of the even lower viscosity: conducting 1D stretching in air can produce pill-like particles while moderate 2D stretching leads to pulley shapes. More extensive stretching forms biconvex lenses. The combination and/or repetition of multiple steps of stretching, heating, and toluene liquefaction can lead to even more unusual shapes including ribbon-like particles with curled ends, bicone, diamond disks, emarginated disks, lat pills, elongated hexagonal disks, ravioli, and tacos (Figure 3.10g–h). With further modiication of the surface texture by relaxing the stretched

Advanced Polymer Nanoparticles with Nonspherical Morphologies

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

83

FIGURE 3.10 SEM images of shapes made by using (a–d) scheme A; (e–f) scheme B; (g–h) a combination of schemes A and B. (Adapted from J. A. Champion, Y. K. Katare, and S. Mitragotri, Proceedings of the National Academy of Sciences of the United States of America 104: 11901, 2007. With permission.)

ilm ahead of liquefaction, wrinkled prolate ellipsoids, wrinkled oblate ellipsoids, and porous elliptical disks can be formed. 3.4.1.3 2D Stretching While the majority of works involve simple uniaxial stretching, which yields prolate ellipsoids with deined aspect ratios, only a few reports have explored the possibility of the deformation of spheres using 2D stretching. As demonstrated by Champion et al. (2007), a 2D stretching method produces oblate shapes that are dificult to realize in the 1D stretching scheme. Since biaxial stretching with orthogonal forces may lead to nonuniform deformation of the ilm, particularly at high stretching ratios, few reports have detailed the possibility of the uniform deformation of spheres into well-controlled diskshaped structures. A simple blown ilm process has been developed that can uniformly deform polymer spheres across the blown ilm and allows the fabrication of disk-shaped ellipsoids with good control (Hu et al. 2008).

84

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Oil bath

PS/PVA film

PS oblate spherioids

PS spheres

FIGURE 3.11 Schematic outline of the blown ilm process. (Reprinted from Y. X. Hu, J. P. Ge, T. R. Zhang, and Y. D. Yin, Advanced Materials 20: 4599, 2008. With permission.)

Figure 3.11 schematically outlines the procedure. A thin composite PS/PVA ilm is irst prepared and then mounted on top of the wide opening of a ceramic iltering funnel, and sealed with a steel clamp. The system is then immersed into a silicon oil bath preheated at 160°C. As this composite ilm softens and becomes deformable at this temperature, air is then injected into the funnel through the small opening using a syringe, allowing the ilm to blow into a dome-like shape. The funnel is removed from the oil bath and cooled in air with the pressure maintained so that the stretched PVA ilm does not show apparent shrinkage upon cooling. The liqueied PS particles experience a uniform 2D stretching force because of their small sizes compared to the PVA ilm, and eventually evolve into disk shapes during the bubble expansion. The transformation from spheres to disks can be controlled by the extent of bubble expansion, which can be represented simply by the ratio of the height of the bubble (H) to the original ilm size (D). Practically, the height of the bubble can be controlled by the amount of air injected into the funnel. Figure 3.12 shows that the major axis of the disks increases dramatically with the degree of expansion of the ilm (H/D). By assuming that both the particle and the matrix materials are purely elastic with identical elastic properties, a sphere will deform into a disk-shaped ellipsoid under biaxial tension by following the relation of Rd R0 = 2

H D

.

Here Rd is the major axis of the disks, R0 the radius of the original spheres, H the height of the bubble, and D the original ilm diameter. In fact, the length of the major axis increases with the degree of bubble expansion at a

85

Advanced Polymer Nanoparticles with Nonspherical Morphologies

(e) 2.2 2.0 1.8 1.6 1.4 1.2 1.0

b

Rd/Ro

a

0.0 c

0.2 0.4 H/D

d

FIGURE 3.12 (a) SEM image of the starting PS spheres. SEM images of PS disks with controllable diameters by injecting different amounts of air: (b) 716 nm; (c) 903 nm; (d) 1149 nm; (e) the stretching ratio Rd:R0 of PS disks as a function of the degree of bubble expansion (H/D). The insets in b–d show a gradual decrease in the minor axis during the stretching. Scale bars in insets are 500 nm and the others are 1 mm. (Reprinted from Y. X. Hu, J. P. Ge, T. R. Zhang, and Y. D. Yin, Advanced Materials 20: 4599, 2008. With permission.)

rate higher than the value predicted with this simple model, which could result from the viscoelastic nature of the materials, the difference in elastic moduli between the particles and the matrix, and the interfacial energy contribution. The blown ilm process can also be applied to transform nanoparticleembedded polymer spheres into disk-shaped composite structures. Typically, the nanoparticles are dispersed in the polymer spheres without strong interactions between each other. As a result, they can move with the hydrodynamic low during the transformation of the polymer particles. We have prepared magnetic microdisks using iron oxide–doped PS beads as the starting materials, and demonstrated the manipulation of their spatial orientation by using external magnetic ields. More complex structures can be achieved by stretching colloidal spheres containing multiple components with different elastic properties. For example, lying saucer–shaped particles can be produced when SiO2@PS core-shell composite particles are used

86

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(a)

(b)

(c)

(d)

FIGURE 3.13 TEM images of lying saucers (a) without magnetic core; (c) with magnetic core; SEM images of (b) lying saucers and (d) hollow disk-shaped PS. (Images (a) and (b) adapted from Y. X. Hu, J. P. Ge, T. R. Zhang, and Y. D. Yin, Advanced Materials 20: 4599, 2008. With permission.)

as the starting materials (Figure 3.13). When such core-shell composite colloids are subjected to blown ilm stretching, only the PS shell can be deformed into disks. The silica core remains spherical because of its rigid nature. As a result of the good adhesion between the PS and the silica surface, the original spherical core-shell particles are transformed into lying saucers with a thin polymer layer covering silica. If anisotropic composite particles are used as the starting materials, more geometrically complex structures can be fabricated through this blown ilm process. In addition, by incorporating active materials such as magnetic particles to the silica cores or using hollow polymer particles, it is possible to produce multifunctional colloids with complex shapes for various new applications (Figure 3.13c–d). 3.4.2 Compression In principle, a thin elastic ilm containing polymer particles can also be deformed by mechanical compression. However, such efforts have been rarely reported probably because of the requirement for a high uniformity in ilm thickness and pressing force in order to obtain uniform deformation. Nevertheless, literature still contains some elegant works that utilize mechanical compression for nonspherical particle fabrication. For example, Sun et al. (2005) recently reported the use of thermal pressing of colloidal crystals for preparing nonspherical particles with well-deined shapes such

Advanced Polymer Nanoparticles with Nonspherical Morphologies

87

as polyhedrons. In this case, a crystalline array of monodispersed polymer particles is sandwiched between two solid substrates and then compressed while heating the system at a temperature slightly lower than the Tg of the polymer. Heating at this temperature allows the softening of polymer chains to facilitate particle deformation upon pressing, and at the same time prevents the colloidal crystals from fusing into continuous ilms. After the thermal pressing treatment, the spherical particles are transformed into polyhedrons in accordance to the structure of the colloidal crystal, and the long-range order of the 3D lattice is essentially preserved (Figure 3.14). In addition to the shape change of the polymer particles, the thermal pressing also leads to obvious changes in the optical properties of colloidal crystals. Another method that involves mechanical compression was reported by Courbaron et al. (2007), who produced disk-shaped poly mer particles by compressing an elastic aqueous gel containing oil-in-water emulsion droplets. An emulsion containing various polymerizable oil droplets is irst mixed with a hot solution of two polysaccharides. By lowering the temperature below the gelling point of the polysaccharide solution, the continuous phase of the emulsion becomes a gel and acquires the elastic properties. Upon compression, the emulsion droplets of polymerizable oil are deformed due to the local strain caused by the macroscopic stress applied to the gel. UV irradiation initiates the polymerization of the emulsion droplets and produces solid polymer particles. Heating the sample above 80°C melts the gel and releases the solid disk-shaped particles. Ellipsoids of a variety of poly mer materials can be obtained by incorporating different polymerizable precursors inside the emulsion droplets. However, the particles are highly polydisperse due to the inhomogeneous nature of emulsion droplets. Another limit of this process is the dificulty in achieving higher aspect ratios as the gel ruptures easily when the elongation ratio is above 1.55. 3.4.3 Self-Assembly Besides posttreatment approaches that involve physical deformation, selfassembly represents another type of method that can produce polymer particles with complex structures. Generally speaking, these self-assembly methods always start with presynthesized uniform polymer particles and then use templates as physical coninement to assemble a limited number of particles into the desired geometric structures. Two types of templates have been used in the self-assembly processes: one variety is soft templates such as droplets in an emulsion system, and the other is hard templates such as patterned thin photo resist ilms. 3.4.3.1 Self-Assembly in Soft Templates In this method, a three-phase emulsion system is employed to form complex-shaped aggregates of colloidal particles (Velev et al. 2000). Suspended

88

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

A)

(a)

(b)

(d)

Evaporation

(c)

B)

hermalpressing

(b)

(a)

250 nm

1 µm (c)

(d)

1 µm

250 nm

FIGURE 3.14 A: Schematic illustration of the procedure used to prepare NSCCs. B: (a) Morphology of the CCs fabricated by two substrate vertical deposition. (b–d) Typical SEM images of the NSCCs. (Images adapted from Z. Q. Sun, X. Chen, J. H. Zhang, Z. M. Chen, K. Zhang, X. Yan, Y. F. Wang, W. Z. Yu, and B. Yang, Langmuir 21: 8987, 2005. With permission.)

emulsion droplets are used as templates within which the colloidal particles such as polymers can be gradually concentrated by drying to form aggregates with unique packings, yielding different kinds of microstructured asymmetric assemblies (Yi et al. 2002, 2004; Cho et al. 2005a, 2005b). The shapes of the as-synthesized anisotropic clusters can be controlled by the

Advanced Polymer Nanoparticles with Nonspherical Morphologies

89

size of suspended template droplets, the number of colloids per droplet, and the surface properties of the colloidal particles. Moreover, composite assemblies can be achieved by varying the components of the colloidal beads. This soft-template approach involving encapsulation and shrinkage is eficient at creating polymeric clusters containing a small number of spheres (n = ~12–15) and is a high-throughput method capable of producing a massive number of clusters, approximately 108–1010 in lab-scale experiments. The method was irst reported by Manoharan et al. in 2003, and involves polymer particles such as PS dispersed in an oil-in-water emulsion system with each oil droplet containing a small number (N) of spheres. By preferentially removing the liquid from a droplet containing particles absorbed onto its surface or conined in its core, compact clusters with different shapes can be obtained as a result of capillary forces and Van der Waals interactions during the drying process. Interestingly, the inal particle packings are unique for all N < 11 and correspond to the clusters with minimized second moments of mass distribution, M = ∑i(ri–r0)2 (r0 is the center of mass of the cluster). Subsequently, clusters of different aggregation numbers are fractionated by density gradient centrifugation. However, this method lacks control of the emulsiication conditions, leading to polydisperse emulsions and severely limited yields of different types of clusters. Since the size of the emulsion droplet is one of the crucial factors that determine the inal shape of the polymer clusters, extensive research has been conducted to generate uniform emulsion droplets. A series of novel emulsiication strategies have been exploited, including for example microluidic devices, micropipette injection, and electrospray. All of these methods involve a droplet break-off technique to produce monodispersed droplets (Umbanhowar et al. 2000). The size of droplets detaching from a tip immersed in a colowing stream has already been investigated (Yi et al. 2004). When the low rate in the capillary is low, a balance between the interfacial tension and the drag force on the droplet at the end of the tip results in a droplet diameter dd given by  γ  dd = dti 1 + ,  3µηc 

(3.5)

where dti is the inner diameter of the capillary tip, γ the interfacial tension between the water and oil, ηc the viscosity of the continuous phase, and µ the velocity of the continuous phase (the shear rate). The droplet sizes of oilin-water emulsions increases when the viscosity ratio between the continuous and dispersed phases is increased because their high internal viscosity resists fragmentation. Also, a low to moderate shear rate can generate the highest proportions of aggregates containing a small number of particles. In

90

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

all, by controlling various parameters, droplets with different sizes can be obtained, thus generating different packing numbers in the emulsion template assisted assembly. The surface properties of these particles cannot be neglected when controlling the structures of the packings. The particles inside the droplets pack according to the capillary force acting upon them. The geometrical structure of clusters can be predicted as a function of constituent particle number through theoretical considerations of minimizing Lennard-Jones potential, second moment of the mass distribution, or Coulomb potential (Cho et al. 2005). By optimizing these functions respectively, three types of sphere packings or cluster structures have been proposed (Livshits et al. 1999; Maranas et al. 1992; Sloane et al. 1995; Wille 1986). The conigurations of clusters predicted by the three models remain the same when the packing number is smaller than seven. When a large number of particles are assembled, the structure is sensitive to small differences in the interparticle forces, and therefore various conigurations of clusters can be achieved when particles with different surface properties are used and consequently different particle interactions are involved. For example, it has been reported that 11 PS clusters assemble into a nonconvex structure in oil while the same number of PMMA spheres pack in a convex shape (Yi et al. 2004). The different packing behavior is attributed to the different swelling characteristics of these two types of particles. The van der Waals attraction, which varies between systems, also contributes to the structure determination in the assembly of large number particles by ixing them at different points in the consolidation process before all of the oil evaporates. Overall, the packing constraints imposed by the particle interaction and the inward pressure of the shrinking oil droplets determine the inal conigurations when a large number of particles are considered. Apparently, the number of spheres per droplet is another important parameter, which can be easily adjusted by controlling the volume fraction of polymer particles in the oil phase (Zerrouki et al. 2006). Largervolume fractions of the particles in the oil phase will result in anisotropic clusters with higher packing numbers while smaller-volume fractions generate clusters with lower packing numbers. Further control of the complex composite assemblies can be attained by controlling the original drop components, for example, by mixing two or more different types of particles (Cho et al. 2005a, 2005b; Reculusa et al. 2004; Rabideau et al. 2004). The segregation of two types of colloids with different sizes and physical properties during assembly yields complex anisotropic particles. For example, with a mixture of regular PS beads and paramagnetic particles, anisotropic particles can be produced by gravitational separation of these components or by the application of a constant magnetic ield (Rabideau et al. 2004). Suspensions containing large latex particles (~270 nm) and small gold particles can form unconventional “glazed doughnut” shapes during the drying (Velev et al. 2000).

91

Advanced Polymer Nanoparticles with Nonspherical Morphologies

3.4.3.2 Self-Assembly in Hard Template The soft-template method can create nonspherical complex assemblies in relatively large quantities. However, more effort is still needed to control the uniformity of the size and structural arrangement of the assemblies. By using hard templates that are lithographically patterned with designed shapes, homogeneous complex aggregates (such as polygonal and polyhedral clusters) assembled from monodispersed spherical colloids can be obtained with well-controlled sizes and near identical structures (Yin et al. 2001a, 2001b). Figure 3.15 outlines the schematic procedure of self-assembly with designed hard templates. The key strategy of this assembly approach is the dewetting of an aqueous dispersion of monodispersed spherical colloidal particles across a surface that contains template holes. As shown schematically in Figure 3.15a, the assembly is performed in a glass cell, which is composed of two glass substrates. The surface of the bottom substrate contains a 2D array of cylindrical holes lithographically patterned in a thin ilm of photo resist. When the particle dispersion is allowed to dewet slowly across the cell, the capillary force exerted on the rear edge of this liquid slug drags the spherical particles across the surface of the bottom substrate until they are physically trapped in the arrays of cylindrical holes. The structural arrangement of these beads can be controlled by the maximum number of polymer beads that can be possibly retained in each hole, which can be determined by the ratios between the dimensions (diameter and height) of the holes and the diameter of the polymer beads. A range of uniform, polygonal or polyhedral clusters can be formed in such 2D arrays of holes. When the solvent is

(a)

Gla

Water

Fe

Flow

(b)

c(c)

(d)

(e)

ss on

cti dire

Fc Fg Patterned photoresist

2 µm

FIGURE 3.15 (a) Schematic illustration of the procedure to assemble spherical colloids into well-controlled clusters under the physical coninement exerted by a 2D array of cylindrical holes patterned in a thin ilm of photo resist. The size, shape, and structure of these clusters can all be easily controlled by changing the ratios between the dimensions of the holes and the diameter of the poly mer beads. (b–e) SEM images of dimers, trimers, squares, and pentamers formed in photo resist templates.

92

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

completely evaporated, the polymer beads trapped in each hole always tend toward physical contact as a result of attractive capillary forces among them. Once the solvent is evaporated, clusters in each hole can be welded into a permanent form via thermal treatment at a temperature slightly higher than the glass transition temperature of the polymer. The extent of viscoelastic deformation of colloidal beads during thermal jointing depends on a number of parameters, in particular, the Tg of the beads. The anisotropic clusters such as dimers, trimers, or squares of PS beads can then be released by dissolving the photo resist ilm with ethanol under sonication. This self-assembly approach can be conveniently extended to the generation of asymmetric dimers from two types of monodispersed, spherical colloids that could be different in size, chemical composition, surface functionality, density or sign of surface charges, bulk properties, or a combination of characteristics (Yin et al., 2001a). Templates other than cylindrical holes have also been demonstrated to generate aggregates of spherical colloids with more complex geometric shapes and internal conigurations (Yin et al. 2001a, 2001b; Xia et al., 2003).

3.5 Summary In this chapter, we have discussed the preparation of nonspherical polymer nanoparticles through either direct seeded polymerization or indirect posttreatment. Nonspherical particles with relatively simple structures can be prepared by controlling the swelling and phase separation processes in seeded polymerization procedures, while more complicated anisotropic polymeric particles can be obtained through indirect approaches such as the stretching of a matrix embedded with polymer particles. Self-assembly processes, as one type of unique posttreatment approach, do not rely on any physical or chemical deformation techniques; instead, they involve using templates to organize the precursor nanoparticles into secondary structures. The starting materials used in the indirect approaches are not necessarily limited to spherical particles. For example, nonspherical particles obtained by seeded emulsion polymerization might also be employed as precursors to fabricate even more complicated structures through self-assembly. Nonspherical polymer nanoparticles can be used as interesting model systems for studying the behavior of molecules in solution, or modeling the behavior of irregular colloidal particles that can be found everywhere in our daily lives. They also have great potential in a wide range of technological applications including the fabrication of three-dimensional photonic colloidal crystals.

Advanced Polymer Nanoparticles with Nonspherical Morphologies

93

References Champion, J. A., Katare, Y. K., and S. Mitragotri. 2007. Making polymeric micro- and nanoparticles of complex shapes. Proceedings of the National Academy of Sciences of the United States of America 104:11901–4. Cho, Y. S., Yi, G. R., Kim, S. H., Pine, D. J., and S. M. Yang. 2005a. Colloidal clusters of microspheres from water-in-oil emulsions. Chemistry of Materials 17:5006–13. Cho, Y. S., Yi, G. R., Lim, J. M., Kim, S. H., Manoharan, V. N., Pine, D. J., and S. M. Yang. 2005b. Self-organization of bidisperse colloids in water droplets. Journal of the American Chemical Society 127:15968–75. Courbaron, A. C., Cayre, O. J., and V. N. Paunov. 2007. A novel gel deformation technique for fabrication of ellipsoidal and discoidal polymeric microparticles. Chemical Communications: 628–30. Eshelby, J. D. 1957. The determination of the elastic ield of an ellipsoidal inclusion, and related problems. Proceedings of the Royal Society A 241:376–96. Felder, B. 1966. Dependence of light-absorption on particle size in heterogenous systems. II. Experimental investigation using model particles. Helvetica Chimica Acta 49:440–53. ———. 1968. Effect of particle geometry on the optical properties of strongly absorbing pigment particles. Helvetica Chimica Acta 51:1224–34. Fujibayashi, T., and M. Okubo. 2007. Preparation and thermodynamic stability of micron-sized, monodisperse composite polymer particles of disc-like shapes by seeded dispersion polymerization. Langmuir 23:7958–62. Ge, J. P., Hu, Y. X., Zhang, T. R., and Y. D. Yin. 2007. Superparamagnetic composite colloids with anisotropic structures. Journal of the American Chemical Society 129:8974–75. Gottlob, K. 1910. Caoutchouc or caoutchouc-like masses. Patent DE 249947. ———. 1912. Isoprene and other hydrocarbons. Patent DE 269240. Ho, C. C., Hill, M. J., and J. A. Odell. 1993. Morphology of ellipsoidal latex-particles. Polymer 34:2019–23. Ho, C. C., Keller, A., Odell, J. A., and R. H. Ottewill. 1993. Preparation of monodisperse ellipsoidal polystyrene particles. Colloid and Polymer Science 271:469–79. Hofmann, F. 1912. Synthetic caoutchouc from technical standpoint. Angewandte Chemie 25:1462–67. Hohenstein, W. P., and H. Mark. 1946. Polymerization of oleins and dioleins in suspension and emulsion. Journal of Polymer Science 1:549–80. Hu, Y. X., Ge, J. P., Zhang, T. R., and Y. D. Yin. 2008. A blown ilm process to diskshaped polymer ellipsoids. Advanced Materials 20:4599–4602. Jiang, P., Bertone, J. F., and V. L. Colvin. 2001. A lost-wax approach to monodisperse colloids and their crystals. Science 291:453–57. Keville, K. M., Franses, E. I., and J. M. Caruthers. 1991. Preparation and characterization of monodisperse polymer microspheroids. Journal of Colloid and Interface Science 144:103–26. Kim, J. W., Larsen, R. J., and D. A. Weitz. 2006. Synthesis of nonspherical colloidal particles with anisotropic properties. Journal of the American Chemical Society 128:14374–77.

94

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

———. 2007. Uniform nonspherical colloidal particles with tunable shapes. Advanced Materials 19:2005–9. Livshits, A. M. L., and E. Yu. 1999. Coulomb clusters on a sphere: Topological classiication. Chemical Physics Letters 314:577–83. Lu, Y., Yin, Y. D., Li, Z. Y., and Y. N. Xia. 2002. Colloidal crystals made of polystyrene spheroids: Fabrication and structural/optical characterization. Langmuir 18:7722–27. Lu, Y., Yin, Y. D., and Y. N. Xia. 2001. Three-dimensional photonic crystals with nonspherical colloids as building blocks. Advanced Materials 13:415–20. Maranas, C. D., and C. A. Floudas. 1992. A global optimization approach for LennardJones microclusters. Journal of Chemical Physics 97:7667–78. Minami, H., Wang, Z., Yamashita, T., and M. Okubo. 2003. Adsorption of styrene on micron-sized, monodisperse, cross-linked polymer particles in a snowmanshaped state by utilizing the dynamic swelling method. Colloid and Polymer Science 281:246–52. Mock, E. B., De Bruyn, H., Hawkett, B. S., Gilbert, R. G., and C. F. Zukoski. 2006. Synthesis of anisotropic nanoparticles by seeded emulsion polymerization. Langmuir 22:4037–43. Okubo, M., Takekoh, R., and A. Suzuki. 2002. Preparation of micron-sized, monodisperse poly(methyl methacrylate)/polystyrene composite particles having a large number of dents on their surfaces by seeded dispersion polymerization in the presence of decalin. Colloid and Polymer Science 280:1057–61. Park, J. M. 2001. Core-shell polymerization with hydrophilic polymer cores. Korea Polymer Journal 9:51–65. Pond, F. J. 1914. A review of the pioneer work on the synthesis of rubber. Journal of the American Chemical Society 36:165–99. Rabideau, B. D., and R. T. Bonnecaze. 2004. Computational study of the selforganization of bidisperse nanoparticles. Langmuir 20:9408–14. Reculusa, S., Mingotaud, C., Bourgeat-Lami, E., Duguet, E., and S. Ravaine. 2004. Synthesis of daisy-shaped and multipod-like silica/polystyrene nanocomposites. Nano Letters 4:1677–82. Sheu, H. R., El-Aasser, M. S., and J. W. Vanderhoff. 1990a. Phase-separation in polystyrene latex interpenetrating polymer networks. Journal of Polymer Science, Part A: Polymer Chemistry 28:629–51. ———. 1990b. Uniform nonspherical latex-particles as model interpenetrating polymer networks. Journal of Polymer Science, Part A: Polymer Chemistry 28:653–67. Skjeltorp, A. T., Ugelstad, J., and T. Ellingsen. 1986. Preparation of nonspherical, monodisperse polymer particles and their self-organization. Journal of Colloid and Interface Science 113:577–82. Sloane, N. J., Hardin, R. H., Duff, T. D. S., and J. H. Conway. 1995. Minimal-energy clusters of hard spheres. Discrete and Computational Geometry 14:237–59. Sun, Z. Q., Chen, X., Zhang, J. H., Chen, Z. M., Zhang, K., Yan, X., Wang, Y. F., Yu, W. Z., and B. Yang. 2005. Nonspherical colloidal crystals fabricated by the thermal pressing of colloidal crystal chips. Langmuir 21:8987–91. Sutera, S. P., and C. W. Boylan. 1980. A nearly monodisperse population of prolate ellipsoidal particles potentially useful for colloidal research. Journal of Colloid and Interface Science 73:295–97. Umbanhowar, P. B., Prasad, V., and D. A. Weitz. 2000. Monodisperse emulsion generation via drop break off in a colowing stream. Langmuir 16:347–51.

Advanced Polymer Nanoparticles with Nonspherical Morphologies

95

Velev, O. D., Lenhoff, A. M., and E. W. Kaler. 2000. A class of microstructured particles through colloidal crystallization. Science 287:2240–43. Whitby, G. S., and M. Katz. 1933. Synthetic rubber. Industrial and Engineering Chemistry 25:1204–1210. Wille, L. T. 1986. Searching potential energy surfaces by simulated annealing. Nature 324:46–48. Xia, Y. N., Gates, B., Yin, Y. D., and Y. Lu. 2000. Monodispersed colloidal spheres. Old materials with new applications. Advanced Materials 12:693–713. Yin, Y. D., Lu, Y., Gates, B., and Y. N. Xia. 2003. Template-assisted self-assembly of spherical colloids into complex and controllable structures. Advanced Functional Materials 13:907–18. Xu, C., Wang, Q., Xu, H., Xie, S., and Z. Yang. 2007. General synthesis of hollow composite ellipsoids. Colloid and Polymer Science 285:1471–78. Yi, G. R., Manoharan, V. N., Klein, S., Brzezinska, K. R., Pine, D. J., Lange, F. F., and S. M. Yang. 2002. Monodisperse micrometer-scale spherical assemblies of polymer particles. Advanced Materials 14:1137–40. Yi, G. R., Manoharan, V. N., Michel, E., Elsesser, M. T., Yang, S. M., and D. J. Pine. 2004. Colloidal clusters of silica or polymer microspheres. Advanced Materials 16:1204–8. Yin, Y. D., Lu, Y., Gates, B., and Y. N. Xia. 2001. Template-assisted self-assembly: A practical route to complex aggregates of monodispersed colloids with welldeined sizes, shapes, and structures. Journal of the American Chemical Society 123:8718–29. Yin, Y. D., Lu, Y., and Y. N. Xia. 2001a. A self-assembly approach to the formation of asymmetric dimers from monodispersed spherical colloids. Journal of the American Chemical Society 123: 771–72. ———. 2001b. Self-assembly of monodispersed spherical colloids into 1D chains with well-deined lengths and structures. Journal of Material Chemistry 11:987–89. Yin, Y. D., and Y. N. Xia. 2001. Self-assembly of monodispersed spherical colloids into complex aggregates with well-deined sizes, shapes, and structures. Advanced Materials 13:267–71. Zerrouki, D., B. Rotenberg, S. Abramson, J. Baudry, C. Goubault, F. Leal-Calderon, D. J. Pine, and J. Bibette. 2006. Preparation of doublet, triangular, and tetrahedral colloidal clusters by controlled emulsiication. Langmuir 22:57–62.

4 Block, Graft, Star, and Gradient Copolymer Particles H. Matahwa, E. T. A. van den Dungen, J. B. McLeary, and B. Klumperman CONTENTS 4.1 4.2

Introduction .................................................................................................. 98 Polymerization Techniques for Complex Macromolecular Architectures ................................................................................................ 99 4.2.1 Block Copolymer Topology .......................................................... 100 4.2.2 Graft Copolymer Topology .......................................................... 101 4.2.3 Star Copolymer Topology ............................................................. 103 4.2.4 Gradient Copolymer Topology .................................................... 103 4.3 Polymer Nanoparticles in Heterogeneous Polymerization ................. 104 4.3.1 Approaches for Successful Application of CRP in Heterogeneous Media ................................................................... 105 4.3.2 Challenges with the Application of CRP in Heterogeneous Media ............................................................................................... 107 4.3.3 Characteristics of the Individual CRP Techniques in Miniemulsion Polymerization ..................................................... 108 4.3.4 Nanoparticles with Complex Architecture Synthesized via CRP in (Mini)emulsion ................................................................. 110 4.4 Polymer Nanoparticles via Self-Assembly ............................................. 113 4.4.1 Synthesis of Amphiphilic Copolymers ....................................... 113 4.4.2 Nanostructures via Self-Assembly of Copolymers ................... 114 4.4.3 Nanoparticles Synthesis by Cross-Linking of SelfAssembled Structures ................................................................... 115 4.4.4 Functionalization of Cross-Linked Micelles .............................. 121 4.4.5 Potential Applications of Cross-Linked Nanoparticles............ 121 4.5 Conclusions................................................................................................. 123 References............................................................................................................. 124

97

98

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

4.1 Introduction This chapter describes routes that have been utilized in the synthesis of advanced polymer nanoparticles. The progress that has been made in the ield of controlled radical polymerization (CRP) for the synthesis of nanoparticles consisting of advanced topology and morphology is highlighted. Advanced polymer topologies that will be discussed include block, graft, star, and gradient copolymers (Scheme 4.1). Heterogeneous polymerization techniques are most commonly used to synthesize polymer nanoparticles. Polymerization in heterogeneous media is relatively simple and high rates of polymerization are obtained, leading to high conversions and high molecular weight polymers in relatively short times. Latex polymers are predominantly synthesized via a free radical polymerization mechanism. This leads to a wide variety of latex products with a large variation of physical and mechanical properties. Polymerizations in heterogeneous media, where the continuous medium is water, provide an environmentally friendly approach for the synthesis of materials, which is necessary due to the increasing restrictions for solvent-based systems. Because water is the medium in which the polymerization takes place, easy removal of exothermic reaction heat is facilitated and the inal latex is a low-viscosity, easy-to-handle product.

AB diblock copolymer

A-B graft copolymer A-B star copolymer

A-BC heterograft copolymer

A-B gradient copolymer

A-BC miktoarm star copolymer

SCHEME 4.1 Various topologies that can be obtained via the employment of various CRP techniques in homogeneous and heterogeneous media. The synthesis of nanoparticles consisting of these topologies is discussed in this chapter.

Block, Graft, Star, and Gradient Copolymer Particles

99

The successful application of CRP techniques in heterogeneous polymerization would provide better control over the molecular weight and give narrow molecular weight distribution polymers. More importantly, due to the retention of the mediating agent as an end group at the polymer chain end, the synthesis of (co)polymers comprising advanced macromolecular architectures, such as shown in Scheme 4.1, is easily obtained. Synthesis of polymer nanoparticles via self-assembly and cross-linking of amphiphilic copolymers, and their modiication for various applications are also discussed. Several approaches to achieving self-assembly of complex architectures can be used. Amphiphilic copolymers can readily self-assemble into nanostructures when a common solvent is replaced by a selective solvent. Additionally, response to stimuli (temperature, pH, or ionic strength) can be used to drive self-assembly of the copolymers. The stable nanoparticles are then obtained by cross-linking of the self-assembled nanostructures. Various chemistries for modiication of the nanoparticles and possible applications are highlighted. Advantages, disadvantages, and the feasibility of the various approaches for industrial application are also discussed.

4.2 Polymerization Techniques for Complex Macromolecular Architectures The invention of so-called CRP techniques provided similar lexibility as observed in conventional free radical poly merization, but with additional advantages such as control over molecular weight and molecular weight distribution and the ability to easily synthesize structures of complex architecture [1–3]. This is attributed to the long lifetime of the polymer chains, which grow throughout the reaction. The polymer chains are initiated fast at the early stages of the polymerization, and all polymer chains grow simultaneously throughout the course of polymerization. Due to the involvement of the polymer chains in a reversible activation-deactivation mechanism, the majority of polymer chains are in a dormant state at any given instant. This means that the active propagating radical concentration is low and that between each activation-deactivation cycle, only a few monomer units are incorporated in the chain. The low active propagating radical concentration also leads to a signiicantly reduced rate of termination and consequently formation of dead chains. In conventional free radical poly merization, all polymer chains synthesized are involved in termination reactions leading to so-called “dead” polymer chains that are not capable of further growth. In CRP on the other hand, at the end of polymerization the polymer chains are end-capped with an activator that allows chain extension for the synthesis of structures with complex topology (Scheme 4.1).

100

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

O

NMP

kact

N

Pn•

kdeact

Pn

O• N

+

kp M

kact ATRP

Pn—X

Mnt —X/L

+

kdeact

Mn+1 t X2/L

+

P n• kp M

RAFT-mediated polymerization

Pm

S P n• kp

+

kdeact

S Z

kact

M

Pn

S

• C

S

Z chain-transfer

Pm

kdeact kact

Pn

S Pm• kp

+

S Z

M

SCHEME 4.2 Reversible termination mechanism for NMP and ATRP and degenerative transfer mechanism for RAFT-mediated poly merization.

The three main CRP techniques that are employed in polymer science nowadays are nitroxide-mediated polymerization (NMP), atom transfer radical polymerization (ATRP) [4,5], and reversible addition-fragmentation chain transfer (RAFT)-mediated polymerization [6,7]. NMP and ATRP are examples of reversible termination, whereas RAFT-mediated polymerization is an example of degenerative transfer (reversible chain-transfer), as illustrated in Scheme 4.2. Both NMP and ATRP are controlled by the persistent radical effect (PRE), which means that every radical-radical termination event leads to an irreversible accumulation of deactivator, shifting the equilibrium toward the dormant species and consequently decreasing the rate of termination. RAFT-mediated polymerization requires a conventional radical initiator to start polymerization, whereas both ATRP and NMP are self-initiating polymerization techniques that initiate polymerization via a halogen-containing ATRP initiator or an alkoxyamine, respectively. Scheme 4.3 shows examples of RAFT agents, ATRP ligands used in Cu(I)-mediated ATRP, and the nitroxides 2,2,6,6-tetramethylpiperidinyl-1oxy (TEMPO) and N-tert-butyl-N-(1-diethylphosphono-2,2-dimethylpropyl) nitroxide (SG1, also called DEPN). 4.2.1 Block Copolymer Topology CRP results in chains that are end-capped with a mediating agent, and chain extension is possible to form block copolymers by adding another monomer followed by initiation. The formation of homopolymer during chain extension experiments in RAFT-mediated polymerization is inevitable and reduces

101

Block, Graft, Star, and Gradient Copolymer Particles

RAFT agents S

S

CH3

S C

S C

N

H

S

CH3 cumyl dithiobenzoate

CH3 2-cyanoprop-2-yl dithiobenzoate

H benzyl dithiobenzoate

S N

O

S N

O

S

S C

OH

S C

O —poly(ethylene-co-butylene)

S PMMA

CH3 2-cyanopentanoic acid dithiobenzoate S H3C

S

CH3

CH3 poly(ethylene-co-butylene) macroRAFT agent

H CH3 O S C

S

PMMA macroRAFT agent

S

H

OH S S H3C H S-butyl-S’-benzyl trithiocarbonate

CH3 S-butyl-S’-2-methylpropanoic acid trithiocarbonate

S

S H

CH3

S

S

S

H3C

H

S

H dibutyl-1,4-phenylenebis(methylene)bistrithiocarbonate

ATRP ligands C18H37 N

N

H3C

N

N

N bis(2-pyridylmethyl)octadecylamine (BPMODA)

tris[(2-pyridyl)methyl]amine (TPMA) C4H9

CH3

N,N,N’,N”,N”-pentamethyldiethylenetriamine (PMDETA)

C4H9 O• N

2-2'-bipyridine (bpy)

N

NMP mediators

H9C4

H9C4 N

N CH3

N

N

CH3

CH3

N N

N

N

4,4'-di(5-nonyl)-2,2'-bipyridine (dNbpy)

O

O• N

CH P OC2H5 OC2H5 SG1

TEMPO

SCHEME 4.3 RAFT agents, ligands for complexation with Cu catalyst in ATRP, and NMP mediators employed in the various controlled radical poly merization techniques.

the purity of the block copolymer. This leads to formation of block copolymer in conjunction with the formation of homopolymer, which increases the molecular weight distribution of the inal product. In addition, chains that do not contain the RAFT moiety as an end group after the synthesis of the irst block cannot be chain extended during the synthesis of the second block and therefore will remain as an impurity in the inal polymer product. In case the monomer conversion during the synthesis of the irst block does not reach 100%, chain extension of the irst block with a second monomer will lead to a second block consisting of a random copolymer instead of a homopolymer (provided excess monomer is not removed). 4.2.2 Graft Copolymer Topology Graft copolymers are a type of branched copolymer in which the side chains are structurally distinct from the main chain. The side chains can

102

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

extend from polymer backbones [8–12] or ([in]organic nanoparticle) surfaces [13–21]. CRP controls the molecular weight and polydispersity index (PDI) of the backbone and the side chains, and either one may consist of a homopolymer, block, alternating, or random copolymer topology. Three approaches for the synthesis of graft copolymers include “grafting from,” “grafting onto,” and “grafting through” [1,2,10]. “Grafting from” involves the creation of a predetermined number of initiating sites on the polymer backbone or surface, and the subsequent polymerization from these initiating sites. This technique inevitably leads to the formation of homopolymer impurity, which requires an extra puriication step, unless the side chains are polymerized via ATRP or NMP, which are self-initiating CRP techniques that are exclusively initiated by halogen-containing ATRP initiators or alkoxyamine initiators, respectively. “Grafting onto” involves the covalent attachment of a preformed polymer chain onto a polymer backbone/ surface via radical reaction or reaction between complementary functional groups. A disadvantage is the low grafting eficiency due to steric hindrance. “Grafting through” involves the polymerization of the pendant vinyl groups of a macromonomer. Generally the degree of polymerization of the backbone is low due to the low concentration of vinyl groups and the high steric hindrance around the chain end propagating radical. A facile method to synthesize graft copolymers via CRP is to functionalize the repeating units of the polymer backbone with a RAFT agent or ATRP initiator (Scheme 4.4) [22]. After polymerization of the backbone, graft copolymers can be synthesized via CRP of the side chains. RAFT-mediated polymerization allows for two different approaches. The R-group approach Grafting from a surface using Z-group approach in RAFT-mediated polymerization O

O OH

S

pyridine

+ Cl

surface

S

S

O

R

O

S S

S

R

MA

O

S S

R

S

PMA

RAFT agent

Grafting from a polymer backbone via ATRP CH3

H2C

O O

n

ATRP

O

O

Br

n

pyridine

O

O

n

Br

ATRP styrene

Br OH

OH HEMA

PHEMA

O O

O

O

O

Br

O

O

Br Br ATRP initiating site on side chain

PS Br

SCHEME 4.4 Synthesis of graft copolymers with the “grafting from” approach via CRP.

Block, Graft, Star, and Gradient Copolymer Particles

103

involves anchoring of the leaving group to the polymer backbone from which the chains are grown. Two macroradicals allow for termination of the polymer chains (and cross-linking between side chains) and loss of RAFT agent functionality. The Z-group approach involves anchoring of the stabilizing group to the polymer backbone. In this case, the degree of polymerization may be limited due to steric hindrance of the side chains in approaching the RAFT moiety close to the backbone. 4.2.3 Star Copolymer Topology Star polymer molecules can be deined as macromolecules containing a single branch point from which linear chains emanate. Polymerization techniques such as conventional free radical polymerization and anionic polymerization have been used to synthesize star polymers. Star polymers are best prepared by coupling of anionic living polymers with multifunctional electrophilic coupling agents. Multifunctional chloromethylated benzene derivatives or multifunctional chlorosilane compounds are often used as the electrophilic coupling reagents. CRP provided another approach for the synthesis of star polymers with control over molecular weight and PDI [23]. The basic requirement for star polymer synthesis is a multifunctional molecule (core) that covalently connects multiple polymer chains (arms) by either initiating polymer growth or linking preformed polymer chains through the multifunctional groups of the molecule [24]. The coupling of two (or more) chemically different polymers results in heteroarm star copolymers, which are often referred to as miktoarm star copolymers (Scheme 4.1). The synthesis of star copolymer particles is readily available by adopting multifunctional initiators for the various CRP techniques (Scheme 4.5). The R-group approach for RAFT-mediated polymerization results in the formation of homopolymer impurity and star-star coupling [25–27]. A whole range of products is obtained due to either premature termination of growing polymer chains or a mixture of branching point molecules from which a different number of polymer chains emanate. Star copolymers are also obtained via the synthesis of linear polymer chains and subsequent addition of (divinyl) cross-linker at high(er) conversion, as reported for ATRP [2]. 4.2.4 Gradient Copolymer Topology The synthesis of gradient copolymers can generally be carried out via two methods, that is, spontaneous (batch) and forced (M2 feed) gradient copolymerization. The second method is advantageous for monomer pairs with unfavorable reactivity ratios. The formation of a gradient copolymer requires that the concentration ratio of the monomer pair changes during the course of the polymerization. A forced gradient copolymer could be obtained when a second monomer is continuously added to the polymerization mixture, which contains only the irst monomer.

104

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Br S Z

S

Z

S S

R

S

S S

Z

O

S S

Z R

S multiple-arm RAFT agent (R-group approach)

Br

Br

S

O

S

CH3

CH3 O

1,1,1-tris(4-(2-bromoisobutyryloxy)-phenyl)ethane (TBiBPE)

Ph

O O

P O

N O

O

O O N

O

O

O O N

O

O

multiple-arm ATRP initiator

O O

O Ph

CH3 Br CH3

O

Br

O

Br

O

CH3 O

O

O

CH3

R

O

O O

O S S

N O

O

O

R

multiple-arm RAFT agent (Z-group approach)

O

Br

S

S

H3C

P

O

N

O Ph

n

multiple-arm SG1 initiator multiple-arm TEMPO initiator

SCHEME 4.5 Multifunctional initiators for the synthesis of star copolymers via various CRP techniques.

4.3 Polymer Nanoparticles in Heterogeneous Polymerization Polymerizations in dispersed systems show a higher rate of polymerization and a lower viscosity at high conversion than polymerizations carried out in solution or bulk. Due to the coninement of a radical within one particle (compartmentalization), radical-radical terminations are minimized and high conversions and high molecular weights are obtained in relatively short times. These are advantages that are of particular interest for industrial processes. Emulsion polymerization is one of the most important industrial processes via which stable polymer latexes are prepared. In most cases, the monomer swollen micelles are transformed into polymer particles via a free radical polymerization process. Emulsion polymerization requires transport of monomer from the droplets to the polymer particles. The fact that the primary nucleation mechanism in emulsion polymerization is either via micellar nucleation or via homogeneous nucleation (aggregation of insoluble chains in the continuous phase that are stabilized by surfactant molecules) complicates the synthesis of complex structures. These include core-shell particles that show homogeneity in composition throughout the particle size distribution and molecular weight distribution. The main mechanism of nucleation can be shifted to droplet nucleation if the size of the monomer droplets is small enough (large total surface area);

Block, Graft, Star, and Gradient Copolymer Particles

105

thus, monomer droplets could compete with the micelles for an incoming radical [28]. This technique is referred to as miniemulsion polymerization, and simpliies the process considerably as no monomer transport through the aqueous phase is required. For an ideal miniemulsion polymerization process, all kinetically stabilized monomer droplets are nucleated and are converted into polymer particles [29,30]. The disadvantages of miniemulsion are that the oil-in-water emulsion is subjected to sonication and that a (volatile) hydrophobe (e.g., hexadecane) is added to the formulation. This makes the realization of industrial implementation more cumbersome, although it has been shown that very hydrophobic (co)monomers (such as lauryl methacrylate) [31] or polymers (e.g., poly[methyl methacrylate], PMMA) [32] also can be used as effective hydrophobes. This eliminates the necessity of removing the hydrophobe after polymerization. Microemulsion is a heterogeneous polymerization technique that uses a very high surfactant load for the stabilization of monomer swollen micelles [33]. Typical weight ratios of surfactant to monomer range between 1 and 2.5. The surfactant concentration is well above the critical micelle concentration (CMC), and all monomer resides within these monomer swollen micelles. This eliminates the need for monomer transport through the aqueous phase, which makes latexes obtained from microemulsions (and miniemulsions) particularly suitable for use as seed. The seed latex can be swollen with monomer, and polymerization is then carried out via a batch emulsion polymerization process. Microemulsions are spontaneously formed and are therefore thermodynamically stable. Micelles that are not nucleated will disintegrate and the surfactant molecules will adsorb onto the surface of the growing latex particles to provide stabilization. Particle sizes obtained in microemulsion typically range between 10 and 50 nanometers. 4.3.1 Approaches for Successful Application of CRP in Heterogeneous Media The introduction of various CRP techniques (NMP, ATRP, and RAFTmediated polymerization, Scheme 4.2) opened up the way for the synthesis of polymer particles with complex architecture via a radical poly merization technique [1,2]. Application of these CRP techniques in heterogeneous media would combine the high rates of polymerization obtained in heterogeneous polymerization with the control over the polymer architecture offered with CRP [34–38]. Several approaches have been reported for successful application of CRP in heterogeneous media: • Microemulsion and miniemulsion polymerization • Preparation of a seed latex via micro- or miniemulsion polymerization and swelling of the particles with monomer (polymerization takes place within these seed latex particles via a batch emulsion polymerization process)

106

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

• In situ formation of polymer particles via nanoprecipitation • Self-assembly of block copolymers synthesized via CRP in bulk or solution Stable latexes can be synthesized via CRP in microemulsion [33] and miniemulsion [39–42] polymerization. Recently successful ab initio batch or semicontinuous (starved–feed) emulsion polymerization mediated via RAFT [43–48] and NMP [49–53] has been reported. In most cases, surfactant-free conditions were employed. Latex products were obtained without the formation of coagulum and particles that consist of polymer chains with complex topology of controlled molecular weight and low PDI. Additionally, these stable latex products could be used as a seed for (continued) polymerization via a batch emulsion process. The advantage then is that the polymerization starts in either Interval 2 or Interval 3, dependent on the presence of monomer droplets, and thus the process of particle nucleation is circumvented. This avoids the transport of the (hydrophobic) mediating agent to the locus of polymerization. The seed latexes are often synthesized under dilute conditions, and for this approach low molecular weight oligomers are targeted. Swelling of the polymer particles with monomer increases the solid content of the latex, and subsequent polymerization is carried out via a batch emulsion process. Nanoprecipitation involves the synthesis of an oligomer via CRP in solution or bulk. The resulting oligomer is dissolved in acetone, for example, and added dropwise to the aqueous surfactant solution. A stable latex is formed and the particles can be swollen with monomer, after which the polymerization can proceed via a batch emulsion polymerization process. The formation of monomer droplets during swelling of the polymer particles should be avoided, as polymerization within the monomer droplets is one of the main causes for signiicant latex destabilization. An advantage is that the polymer particles are formed in situ and that no monomer transport through the aqueous phase is required. This signiicantly improves the stability of the latex and the control over the molecular weight and PDI, as the mediating agent is present at the locus of polymerization. Prescott et al. reported on the successful application of RAFT-mediated polymerization using a similar approach [54]. They prepared a polystyrene (PS) emulsion and transported the RAFT agent to the locus of polymerization via the acetone transport technique, after which they successfully conducted RAFT-mediated poly merization. The drawback of this technique in their case is that the seed was prepared via conventional radical polymerization, and hence the latex product also contained uncontrolled polymer that could not be chain extended. Similarly, synthesis of amphiphilic block copolymers via CRP in solution or bulk allows for self-assembly of these block copolymers in the aqueous phase [47]. The amphiphilic block copolymers simultaneously act as mediating agent and stabilizing agent, which eliminates the requirement for a

Block, Graft, Star, and Gradient Copolymer Particles

107

low molecular weight surfactant. The self-assembly process can be induced either via the use of a selective solvent (e.g., water) or upon a change in environment, such as pH, temperature, or concentration of electrolyte. Use of a surface active mediating agent also allows for self-assembly in the aqueous phase [43,44]. The resulting micelles can be swollen with monomer via a controlled feed, and the polymerization can take place via a semicontinuous emulsion process. The use of a water-soluble monomer, such as (meth) acrylic acid ([M]AA) or 4-vinyl pyridine (4VP), allows for direct synthesis in the aqueous phase, and the resulting oligomer can be used as stabilizer in seeded semicontinuous emulsion for the polymerization (chain extension) of a water insoluble monomer [45–51,53]. However, due to interference of the acid group with the catalyst complex, this approach is not suitable for use in ATRP. Protected group chemistry is often used in this case to protect the acid functionality. 4.3.2 Challenges with the Application of CRP in Heterogeneous Media A common problem that had to be dealt with was the exit of small, mediating agent-derived radicals from the monomer droplets into the aqueous phase. An inherent characteristic of CRP is simultaneous growth of the polymer chains throughout the polymerization, and therefore slow formation of high molecular weight polymer. This could enhance exit of the radicals from the monomer droplets. Additionally, the resulting high concentration of oligomers at the early stage of the polymerization could lead to instability of the latex product because oligomers are known to be very effective swelling agents (superswelling) [55]. It should be noted that chain-transfer reactions in conventional free radical polymerization in (mini)emulsion also leads to the formation of small radicals, which can either terminate a long polymer chain (via short-long termination events) or exit the polymer particle. The absence of radicals in the polymer particle leads to rate retardation. In addition, partitioning of the mediating agent between the oil and aqueous phases determines the equilibrium concentrations in both phases and can be adequately controlled by using a suficiently hydrophobic mediating agent. Initial attempts to perform CRP in a (mini)emulsion process were met with dificulties such as colloidal instability and poor control over the molecular weight and PDI [34,35]. The inability to carry out CRP directly in batch emulsion was attributed to negligible transport of the hydrophobic mediating agent from the large monomer droplets to the polymerization loci. The resulting latex was characterized by phase separation and coagulation of small droplets, which lead to instability of the inal latex product. Therefore, CRP was applied to miniemulsion polymerization, which does not require transport of the (hydrophobic) mediating agent. When a rather hydrophilic mediating agent was used, the early stage of the polymerization was characterized by an inhibition period, due to the slow consumption of the

108

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

mediating agent and hence slow formation of surface active z-mers (oligomers with chain length that gives suficient hydrophobicity). 4.3.3 Characteristics of the Individual CRP Techniques in Miniemulsion Polymerization A large number of monomers have been used for RAFT-mediated polymerization in miniemulsion, such as styrene (S), methyl methacrylate (MMA), (2-ethyl)hexyl methacrylate (EHMA), methacrylic acid (MAA), n-butyl acrylate (n-BA), etc. Chain extension of the resulting homopolymers to form block copolymers (via seeded emulsion polymerization) has been carried out [40,41,56]. Anionic and cationic low molecular weight surfactants and nonionic polymeric surfactants have been investigated for their inluence on the stability of the latex particles. Provided that the concentration of surfactant (stabilization against coalescence/coagulation) and hydrophobe (stabilization against monomer diffusion [Ostwald ripening]) are suficiently high, stable latex particles with a minimum of coagulation can be synthesized via RAFT-mediated polymerization in miniemulsion [57]. Care has to be taken that the pH of the system is properly buffered because RAFT agents are sensitive toward hydrolysis at elevated pH [58,59]. Commonly observed phenomena in RAFT-mediated polymerization are an inhibition period and a retardation of the rate of polymerization compared to conventional radical polymerization [60]. These are disadvantages that particularly limit industrial application. It has been suggested that rate retardation in RAFT-mediated (mini)emulsion polymerization could be caused by exit of the leaving group radical into the aqueous phase [61] or by using a block copolymer stabilizer that contains a poly(acrylic acid) (PAA) block [34]. PAA is susceptible to proton abstraction, which results in formation of midchain radicals (MCRs) that propagate slowly and terminate quickly. ATRP in miniemulsion is commonly carried out via “activators generated by electron transfer (AGET) ATRP” [62,63]. Cu(II) is added as the catalyst precursor, which is more stable to oxygen than Cu(I). This circumvents the use of Cu(I) species that is very sensitive to oxidation, which would occur during the formation of a miniemulsion via sonication. AGET ATRP uses a reducing agent such as ascorbic acid, instead of a conventional radical initiator, to create the activator [62,63]. A substoichiometric ratio of reducing agent is used to leave some excess of Cu(II) species to regulate ATRP. This approach also increases the tolerance of the system to air, as the reducing agent can be used to scavenge oxygen from the air. The presence of air would, however, require an increased amount of reducing agent. The optimal amount of the reducing agent depends on the partitioning of the Cu(II) species between the aqueous phase and the monomer droplets, ATRP equilibrium constant, target molecular weight, etc. Too small an amount of reducing agent would lead to very slow polymerization, whereas too high an amount of reducing agent would deteriorate the control over the polymerization. Partitioning of

Block, Graft, Star, and Gradient Copolymer Particles

109

the more hydrophilic Cu(II) species between the oil and aqueous phases disturbs the Cu(I)-to-Cu(II) ratio, which is undesirable as this ratio regulates the control over the polymerization. Another advantage of AGET ATRP is the ability to synthesize pure block copolymers without the formation of homopolymer, as no free radicals are present that initiate new chains during the formation of the second block. The earlier developments of reverse ATRP and simultaneous reverse and normal initiation (SR&NI) ATRP also allowed for application in miniemulsion. However, reverse ATRP employs an external initiator (such as AIBN) to reduce the Cu(II) to Cu(I) and therefore this technique does not allow for the formation of block copolymers. SR&NI ATRP employs a normal alkyl halide initiator in combination with a conventional radical initiator and thus allows for the synthesis of block copolymers and structures of more complex architecture. In addition, a signiicantly decreased concentration of catalyst is required. However, as a conventional radical initiator is used for the generation of radicals, the synthesis of a block copolymer also leads to the formation of homopolymer impurity from polymerization of the second monomer, which requires an additional puriication step. Initial attempts to adopt “normal” ATRP in miniemulsion provided moderate success [64,65]. Only if a stable initial emulsion was obtained did the polymerization proceed in a controlled manner, that is, linear increase of molecular weight with conversion and low PDI throughout the reaction. An inherent disadvantage to ATRP is the use of a transition metal catalyst (mostly copper), which may be undesirable as it is retained as an impurity after the polymerization. The development of “activators re-generated by electron transfer (ARGET) ATRP” [68], which is similar in concept to AGET ATRP, allowed for the use of only parts per million (ppm) amounts of copper catalyst in the polymerization medium, while control over the polymerization was retained [66–68]. This development is a big step forward to application in industrial processes. Interaction of anionic surfactants with the copper catalyst results in poor control over molecular weight and molecular weight distribution. Nonionic surfactants can be used in ATRP instead, but they are generally not as preferable as they give bigger particle sizes, result in less stable latexes, and become less effective at elevated temperatures [37]. The development of the “second-generation” nitroxides in NMP provided a solution to the high temperatures required for activation of the TEMPOderived alkoxyamine (Scheme 4.3) [69–71]. The equilibrium constant of these second-generation nitroxides is much higher at lower temperatures and therefore an acceptable rate of polymerization is obtained. In addition, the second-generation nitroxides allow for controlled polymerization of (meth)acrylates. An example is the commercially available nitroxide SG1 (DEPN, Scheme 4.3), which can be used at temperatures of 90–120°C. Due to the formation of a large amount of dead chains at high conversion during the synthesis of homopolymer via NMP in miniemulsion, it is important to keep the conversion low to render the majority of chains active for

110

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

the synthesis of block copolymers. Alternatively, nitroxide end-capped oligomers can be synthesized in bulk or solution and applied as both hydrophobe and mediating agent to miniemulsion polymerization [69]. 4.3.4 Nanoparticles with Complex Architecture Synthesized via CRP in (Mini)emulsion The successful application of CRP in heterogeneous media allows for the synthesis of nanoparticles with complex architectures such as block [40,41,44–51,53,54], graft [10,20], star, and gradient [72] copolymers in (mini)emulsion. The synthesis of these structures does not cause additional complexity to the system compared to the synthesis of homopolymers in (mini)emulsion. RAFT-mediated polymerization H Z S S H

H Miniemulsion polymerization of n-BA

H S

T = 75°C, AIBN, Brij 98, hexadecane Z

S

Z

Z

S

S

H PBA

H

S

S

H

H

PBA

Z=CH3CH2CH2CH2S H

Z

Z

S

S

H PBA

H

PS

S PS

PBA

Sty, AIBN

S

T = 75°C

H

ATRP H2C

CuBr, PMDETA, EBiB

O O

T = 50°C, bulk CH3

Br PMA

CuBr2, BPMODA, Sty, Brij 98, hexadecane, ascorbic acid

Br

T = 80°C, miniemulsion

PMA

PS

NMP H2C O O

T = 115°C 2.5 mol% free SG1

C4H9 +

O N

O

N

O P

OCH3

Conversion to ~80%

O H3CO

PBA O

OC2H5

OC2H5

P

H2C T = 115°C

O OC2H5

OC2H5 CH3-O(CO)-CH(CH3)-SG1

PBA

PS O

O OCH3

SCHEME 4.6 Block copolymer synthesis via CRP in miniemulsion.

N

O P

OC2H5

OC2H5

111

Block, Graft, Star, and Gradient Copolymer Particles

(a)

(b)

250 nm

250 nm (c)

(d)

250 nm (e)

250 nm (f )

500 nm

500 nm

FIGURE 4.1 Height (a, c, and e) and phase (b, d, and f) images obtained with AFM from poly mer brushes synthesized via AGET ATRP in miniemulsion, using various ratios of ascorbic acid to Cu(II): 0.20 (a and b) and 0.35 (c–f). Monomer conversion is ~21% for images a and b, ~25% for images c and d, and ~80% for images e and f. (Reprinted from K. Min, S. Yu, H.-I. Lee, L. Mueller, S. S. Sheiko, and K. Matyjaszewski, Macromolecules 40: 6557–63, 2007. With permission.)

Figures 4.1 and 4.2 show examples of atomic force microscopy (AFM) and transmission electron microscopy (TEM) images of polymers with complex architecture obtained in mini(emulsion). Min et al. synthesized polymer brushes by growing PBA chains from a polymer backbone via AGET ATRP directly in miniemulsion [10]. Figure 4.1 shows the corresponding height and phase images obtained with AFM from these polymer brushes for various ratios of ascorbic acid to Cu(II) used, at a given conversion for n-BA. Dire et al. reported on the surfactant-free ab initio batch emulsion polymerization of MMA using a P(MMA-co-S)-SG1 macroalkoxyamine that simultaneously acted as initiator and stabilizing agent [49]. The TEM images they

112

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

500 nm

FIGURE 4.2 TEM images obtained from surfactant-free ab initio batch emulsion poly merization of MMA using a P(MMA-co-S)-SG1 macroalkoxyamine initiator. (Reprinted from C. Dire, S. Magnet, L. Couvreur, and B. Charleux, Macromolecules 42: 95–103, 2009. With permission.)

obtained from their stable latexes show a very narrow particle size distribution, as can be seen in Figure 4.2. Provided that stable latex particles are obtained, employment of the appropriate mediating agent during the polymerization (e.g., Scheme 4.7) will result in the formation of graft or star copolymers. The synthetic route is similar to that in homogeneous media, with the additional beneit in heterogeneous media of less termination/coupling reactions. Although coupling reactions can still occur within the latex particles, macroscopic gel formation will not occur as long as the (mini)emulsion system remains stable [10]. Cross-linking will not affect the luidity of the system and therefore high conversions can be achieved. Although the synthetic route is not signiicantly more complex compared to the synthesis of block copolymers, few reports have appeared on the synthesis of graft and star copolymer nanoparticles directly in (mini)emulsion. H2C

O O

1. BPMODA, Cu(II)Br2, EBiB, hexadecane, T = 80°C Aqueous solution of Brij 98 2. AscA, t-BA (0.01 mL/min for 200 min) gradient n-BA - t-BA copolymer

H3C

SCHEME 4.7 AGET ATRP of n-BA with a constant feed of tert-butyl acrylate (t-BA) in miniemulsion for the synthesis of gradient copolymers with controlled molecular weight and low polydispersity. (Reprinted from K. Min, J. K. Oh, and K. Matyjaszewski, Journal of Polymer Science, Part A: Polymer Chemistry 45: 1413–23, 2007. With permission.)

The synthesis of gradient copolymers has been reported for ATRP, where AGET ATRP was used for the synthesis of forced gradient copolymers [72]. The absence of a conventional radical initiator for the reduction of the precursor catalyst Cu(II) provides that the polymers exclusively be initiated via the

Block, Graft, Star, and Gradient Copolymer Particles

113

ATRP initiator, and therefore does not result in the formation of homopolymer. Factors that inluence the gradient are, for example, the molar feed ratio of the monomers, the feeding rate, the hydrophobicity of the monomers, and the reactivity ratios of the monomers. SR&NI ATRP is not suitable for the forced gradient method because M2 is fed to the reaction mixture. Radicals derived from a conventional radical source could escape monomer droplets where the polymerization takes place, and initiate polymerization of M2 in the aqueous phase to form homopolymer of M2. This not only broadens the molecular weight distribution, but also leads to the formation of unstable latexes.

4.4 Polymer Nanoparticles via Self-Assembly 4.4.1 Synthesis of Amphiphilic Copolymers Most polymerization techniques can be used to synthesize amphiphilic copolymers. Several polymerization techniques can be used in tandem for the synthesis of amphiphilic copolymers with complex architectures. This requires that several functional groups or initiating/mediating agents are present either at the chain end (block copolymer synthesis) or along the polymer backbone (for grafting reactions) or at the junction (for star copolymer synthesis). The amphiphilic nature is a result of having different poly mer blocks in the same molecule that have different properties. Synthesis of complex macromolecular architectures by CRP has been highlighted already in the previous section. A combination of ATRP and “click chemistry” is widely used for the synthesis of amphiphilic copolymers. ATRP and click chemistry were used, for example, to synthesize a multiresponsive double hydrophilic ABC miktoarm star terpolymer [73] and two oppositely charged graft ionomers consisting of poly(methacrylic acid-co-azidopropyl methacrylate)-g-poly(N-isopropyl acrylamide) (P[MAA-co-AzPMA]-g-P[NIPAM]) and methyl iodide quaternized poly(2-[dimethylamino]ethyl methacrylate-coazidopropyl methacrylate)-g-poly(N-isopropyl acrylamide) (P[DMAEMAco-AzPMA]-g-P[NIPAM]) [74]. ATRP was used to graft tert-butyl acrylate (t-BA) onto polystyrene-b-ethylene-co-butylene-b-polystrene (PSEBS) backbone [75]. The ATRP macroinitiator was initially prepared by hydrogenation and chloromethylation of PSEBS. Free radical polymerization in conjunction with coupling chemistry can also be used to synthesize amphiphilic copolymers. A cross-linkable amphiphilic graft copolymer was synthesized via free radical polymerization of N-acryloxysuccinimide (NAS) as the backbone, and then HEMA, amino-functionalized polyethylene glycol (PEG), and amino-functionalized PNIPAM were attached to the backbone [76]. Other techniques such as ROP [77,78] and ROMP [79] can also be used to synthesize amphiphilic copolymers. Anionic polymerization was used to synthesize

114

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

triblock terpolymers that self-assemble in organic media [80]. A combination of ATRP and anionic polymerization was used to synthesize star-like amphiphilic copolymers [81]. 4.4.2 Nanostructures via Self-Assembly of Copolymers Self-assembly of amphiphilic copolymers such as block [82,83], star [84–87], and graft copolymers [74,76,88–90] into nanostructures have been reported by various authors. Amphiphilic copolymers have a tendency of selfassembling into morphologies such as micelles, vesicles, lamellae, etc. when a selective solvent for one of the blocks is added. The driving force for self-assembly is the reduction in surface energy. The commonly used self-assembly procedure is to synthesize the copolymer (or dissolving the copolymer) in its common solvent (solvent for all the blocks) followed by solvent exchange with a selective solvent via dialysis. The self-assembly of these copolymers is reversible upon addition of a common solvent for the copolymer and upon heating. It is important to note that the self-assembled copolymers are thermodynamically and kinetically more stable than self-assembly structures from low molecular weight molecules [91]. The process of self-assembly of amphiphilic copolymers is affected by a number of factors, which include their chemistry, topology, molecular weight, (co)solvent, etc. Self-assembly of copolymers is also induced by incorporating thermo- or pH-responsive polymer chains in the copolymer, and thus nanostructures are formed upon a variation in the pH or temperature [92–97]. A double hydrophilic ABC miktoarm star terpolymer made up of PEG, PNIPAM, and poly(2-[diethylamino]ethyl methacrylate) (PDEAEMA) blocks was shown to self-assemble at 25°C and pH 10 and at 50°C and pH 4 but remained in solution at 25°C and pH 4 [73]. Narain and Armes used an aldehyde-functionalized ATRP initiator to synthesize various sugar containing methacrylate block copolymers (glycopolymers) that showed either pHor thermo-responsive behavior [98]. These copolymers formed well-deined micelles upon self-assembly, and the aldehyde-functional groups were found on the periphery of the micelles. Inverse polymeric micelles comprising a hydrophilic core and a hydrophobic shell can also be prepared. For example, micelles of an AB copolymer in a selective solvent for A can be reversed by adding a selective solvent for B while removing the selective solvent for A via dialysis. Cheng et al. prepared inverse polymeric micelles in chloroform using polystyrene-b-poly(4vinylpyridine) block copolymer by in situ quaternization of the pyridine moiety using HCl, and Figure 4.3 shows the latex particles obtained by varying the transfer time [99]. Self-assembly of macromolecules is useful for commercial applications such as coatings, detergents, cosmetics, electronics, thin ilms, drug delivery, and biocompatibility. The self-assembly structures are usually studied using (cryo) TEM, light scattering, and small-angle X-ray scattering. Freeze

Block, Graft, Star, and Gradient Copolymer Particles

(a)

(b)

(c)

(d)

(e)

(f)

115

FIGURE 4.3 TEM images of PS-b-P4VP inverse micelles at different transfer times (t): (a) 3 min, (b) 15 min, (c) 10 h, (d) 24 h, (e) 120 h, and (f) is the TEM image of PMACs (t = 120 h) noncovalently crosslinked by bivalent S2O82– ions. All the scale bars are 200 nm in length. (Reprinted from F. Cheng, X. Yang, H. Peng, D. Chen, and M. Jiang, Macromolecules 40: 8007–14, 2007. With permission.)

drying can be used to remove water without disrupting the micellar structure, allowing techniques such as AFM to be used. 4.4.3 Nanoparticles Synthesis by Cross-Linking of Self-Assembled Structures Synthesis of polymer nanoparticles is done via cross-linking of self-assembly structures of block, star, and graft copolymers. The synthesis of cross-linked nanoparticles requires cross-linkable groups to be part of the self-assembled structure. Scheme 4.8 illustrates different domains in the self-assembled structures that can be selectively cross-linked. Cross-linking of self-assembled structures will result in polymer nanoparticles that are resistant to changes in solvent, temperature, pH, or ionic strength. The advantage of cross-linking of self-assembly structures over synthesis in heterogeneous media for nanoparticle synthesis is that amphiphilic polymer nanoparticles of complex copolymer architectures can be easily

116

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Copolymer Self assembly

Cross-linking

Shell cross-linking at the hydrophilic end

Cross-linked shell

Cross-linked core

Core cross-linking at the hydrophobic end

SCHEME 4.8 Cross-linking of micelles at various possible strategic positions.

synthesized whereas their synthesis in emulsion may be more complex and induce latex stability problems. Depending on the position of the cross-linkable moieties in the copolymer, cross-linking can be carried out in the shell at the hydrophilic chain end, in the shell along the hydrophilic polymer backbone, at the hydrophilichydrophobic interface of the block copolymer, in the core along the hydrophobic polymer backbone, and in the core at the hydrophobic chain end [100]. The cross-linking density and the position of cross-links in the micellar structure hugely affect properties such as permeability, swellability, thermal stability, reactivity, etc. of the polymer nanoparticles. It is important to note that the self-assembled structures are preserved after adequate cross-linking at various strategic positions using various chemistries. Most research in this ield focused on cross-linked, self-assembled structures of amphiphilic block copolymers. O’Reilly et al. wrote a brief review on block copolymers that self-assemble into polymer micelles in a selective solvent and that possess functional groups throughout the core and/or shell for cross-linking purposes [100]. Guo et al. synthesized cross-linked diblock copolymer micelles of polystyrene-b-poly(cinnamoylethyl methacrylate) (PS-b-PCEMA) via photo-cross-linking of the PCEMA block [101]. Figure 4.4 shows TEM images of core photo-cross-linked diblock copolymer micelles illustrating that increasing the molecular weight of both blocks of the diblock copolymer resulted in an increase in the particle size.

117

Block, Graft, Star, and Gradient Copolymer Particles

(a)

(b)

250 nm

100 nm

FIGURE 4.4 Transmission electron micrographs of core cross-linked diblock copolymer micelles; (a) PS276-bPCEMA90 and (b) PS302-b-PCEMA570. (Reprinted from A. Guo, G. Liu, and J. Tao, Macromolecules 29: 2487–93, 1996. With permission.)

A number of cross-linking chemistries can be employed for cross-linking of self-assembled structures. End-functionalized amphiphilic copolymers are required for shell and core cross-linking at hydrophilic or hydrophobic polymer chain end, respectively. Scheme 4.9a shows examples of functional groups that can be used for cross-linking of copolymers at the chain end. Self-assembly of end-functionalized copolymers presents the functional groups at either the exterior or interior of the self-assembled structures. The (a) Examples of functional groups for cross-linking or modification 1

alkynyl

N3 azide

5

(b) Examples of vinyl cross-linkers

3

1,4-divinylbenzene 6 H2N amine 7 O

O O acrylate

4

OH alcohol

N3 12 N3 1,4-diazidobutene

9

O 2 H aldehyde

(c) Examples of chemical cross-linkers

O

glutaraldehyde

O

OH

carboxylic acid

O

CH3 11

O

O

13 O

NH

O O N,N’-methylenebisacrylamide

ketone 8

NH 10

O

NH2 14 H2N ethylenediamine 15

CH3

O CH dipropagyl ether

ethylene glycol dimethacrylate

(d) Examples of (co)monomers that impart functional groups for cross-linking or modification CH3 CH3 N O N H 3C O O O O 19 O 16 O 17 O 18 2-vinylpyridine (2-VP) 21 N O O O OH N-acryloxysuccinimide (NAS) 20 OH N3 2-hydroxy ethyl methacrylate acrylic acid (AA) (HEMA) 3-azidopropyl methacrylate (AzPMA)

O

CH3 O

7-(2-methacryloxyethoxy)4-methylcoumarin

SCHEME 4.9 Examples of (a) end-functional groups, (b) vinyl cross-linkers, (c) chemical cross-linkers, and (d) vinyl (co)monomers that can be used for cross-linking and modifying cross-linked nanoparticles.

118

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

end-functional groups may also include RAFT moieties, ATRP initiators, and vinyl groups (i.e., macromonomer) that allow polymer chains to be crosslinked using a difunctional vinyl monomer (Scheme 4.9b). The copolymer chain end can also be functionalized with functional molecules and then cross-linked after self-assembly using multifunctional complementary molecules. Scheme 4.9c shows examples of chemical cross-linkers. Graft copolymers (a) P(AA-co-MEA)-g-PNIPAM and (b) P(AA-co-MEA)-g(PNIPAM and PEG) were synthesized by grafting amine-terminated polymers PNIPAM and PEG onto P(NAS-co-MEA) backbone [76]. The unreacted NAS units in the copolymer backbone were then hydrolyzed forming acrylic acid residues. Cross-linking of the MEA residues after thermally induced self-assembly of the graft copolymer was performed using ammonium persulfate, and Figure 4.5 shows the TEM images of the graft copolymers (a) P(AA-co-MEA)-g-PNIPAM and (b) P(AA-co-MEA)-g-(PNIPAM and PEG). Graft copolymer (a) resulted in bulk cross-linking while graft copolymer (b) resulted in exclusively shell cross-linking. Cross-linking of the shell or the core is done through cross-linkable groups in the shell or core of the self-assembled structures. Some functional monomers (Scheme 4.9d) can be used in copolymerization or in modiication of the copolymer to provide cross-linkable groups. Bütün et al. reported on the synthesis of shell cross-linked block copolymer micelles with tunable hydrophilic/hydrophobic cores [102]. The core consisted of a thermo-responsive poly(N-[morpholino]ethyl methacrylate) (PMEMA) block that is hydrophilic at room temperature and hydrophobic above its lower critical solution temperature (LCST). Cross-linking of the PDMAEMA shell

(a)

(b)

100 nm

100 nm

FIGURE 4.5 TEM images of cross-linked micelles (a) P(AA-co-MEA)-g-PNIPAM and (b) P(AA-co-MEA)-g(PNIPAM and PEG). (Reprinted from W.-H. Chiang, Y.-H. Hsu, T.-W. Lou, C.-H. Chern and H.-C. Chiu, Macromolecules 42: 3611–19, 2009. With permission.)

119

Block, Graft, Star, and Gradient Copolymer Particles

above the LCST of PMEMA resulted in preserved micellar structure upon cooling to room temperature. Copolymers with carboxylic acid or amine functional groups in the hydrophilic part allow cross-linking to be done in the shell (or the core for inverse micelles). Scheme 4.9c shows examples of chemical cross-linkers. Difunctional amines such as 2,2’-(ethylenedioxy)bis(ethylamine) and aldehydes such as glutaraldehyde are used to cross-link the carboxylic acid and amino functional groups, respectively. Multifunctional (>2) molecules are statistically better cross-linkers. Monomers such as 2-vinyl pyridine (2VP) and NAS can be used in the copolymer and then cross-linked using difunctional halogenated molecules such as 1,4-dibromobutane and difunctional alcohols, respectively. Azide functionalized methacrylate can be used in copolymerization to allow for cross-linking reactions with difunctional alkynyls such as dipropargyl ether. Monomers such as 2-hydroxyethyl methacrylate (HEMA) can be used to react with PNAS units in a copolymer to allow for free radical cross-linking. It is also possible to cross-link micelles at the hydrophilic-hydrophobic interface. Scheme 4.10 shows how ATRP and click chemistry can be used to speciically functionalize a copolymer at the hydrophilic-hydrophobic interface. Self-assembly of the amphiphilic copolymer will result in the monomer functionality at the hydrophilic-hydrophobic interface. Cross-linking of the micelles is then done using a difunctional or multifunctional free radical monomer. Alternatively the hydrophilic-hydrophobic interface can be modiied with chemical groups, and the micelles are then stabilized by cross-linking using multifunctional cross-linkers. O

O

OH

Br O

CH O

O

O

O OH

OH

O

N N

CH

(eg PEG-OH)

Br

N

N3

HC DCC DMAP CH2Cl2 HO O

O

H3C

OH

O

CH O

Br

ATRP monomer (eg. MMA)

O O O

O

O CH3 O ‘click chemistry’

O O

O 1) self-assemble 2) cross-linking CH3 O O O CH3 EGDMA O

cross-linking at the hydrophilic-hydrophobic interface of micelles

SCHEME 4.10 An example of cross-linking chemistry at the hydrophilic-hydrophobic interface.

120

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Another method that can be used to cross-link micelles at the hydrophobic-hydrophilic interface is to use an ABC amphiphilic miktoarm star copolymer, for example. One of the blocks will be hydrophobic and the other two hydrophilic. One of the hydrophilic blocks should be pH- or thermoresponsive and bear cross-linkable groups. The micellization is carried out at normal conditions; then pH or temperature is changed and this results in the collapse of the pH- or thermo-responsive block. Cross-linking of this block is then carried out under these conditions to form micelles cross-linked at hydrophobic-hydrophilic interface, and a change in the pH or temperature will have little or no effect on the cross-linked block. Alternatively, an ABC triblock copolymer can be synthesized with crosslinkable B-block, which can be varied in length and solubility in selective solvents for both A and C. The A and C blocks are designed to be hydrophilic and hydrophobic, respectively. Self-assembly of the triblock copolymer will position block B at the hydrophilic-hydrophobic interface and cross-linking of B can be carried out. Liu et al. synthesized a thermo-responsive triblock copolymer that formed micelles, and cross-linking of the micelles was performed on the central block to avoid intermicellar cross-linking [103]. Intermicellar cross-linking during shell cross-linking is also avoided or minimized using dilute conditions. Figure 4.6 shows a TEM image of a PEO-b-PDMAEMA-b-PMEMA (PMEMA = poly[N-morpholino]ethyl methacrylate) micelles cross-linked at the PDMAEMA block [104].

250 nm

FIGURE 4.6 Transmission electron micrograph of a dilute suspension of cross-linked micelles prepared using the PEO-PDMAEMA-PMEMA triblock copolymer. (Reprinted from V. Bütün, X.-S. Wang, M. V. d. P. Banez, K. L. Robinson, N. C. Bilingham, and S. P. Armes, Macromolecules 33: 1–3, 2000. With permission.)

Block, Graft, Star, and Gradient Copolymer Particles

121

4.4.4 Functionalization of Cross-Linked Micelles Nanoparticles can be designed to have a particular functionality on the surface or in its interior to allow further modiication. Amphiphilic copolymers have long polymer chain segments in both the hydrophilic and hydrophobic parts in which speciic functional groups can be incorporated. After selfassembly and cross-linking, some of the functional groups may reside on the surface (for functional groups in the hydrophilic part) or interior (if functional groups are in the hydrophobic part) of the nanoparticles. The residual functionality is then used for modiication of the polymer nanoparticles with (bio)macromolecules or luorescent labels, for example. The use of functional initiators and CRP mediators (ATRP initiator, RAFT agent, alkoxyamine [NMP]) will result in cross-linked micelles with functional groups on the surface or in the micellar interior, which can be utilized for further CRP. Scheme 4.11 shows chemistries that have been used to modify the surface of cross-linked copolymer micelles. Alternatively, the copolymers can be functionalized irst with organic, (bio)molecules, dyes, etc., followed by self-assembly and then cross-linking. A biotin-functionalized ATRP-initiator was used to synthesize a poly(acrylic acid)-b-poly(methyl acrylate) (PAA-b-PMA) block copolymer [105]. After selfassembly of the block copolymer and cross-linking of the shell, the biotin functionality was retained at the surface of the nanoparticles. 4.4.5 Potential Applications of Cross-Linked Nanoparticles Polymeric micelles have great potential application in the drug delivery ield as nanocarriers. These self-assembled micelles have been widely studied in the ield of drug delivery because hydrophobic drugs can be retained within the hydrophobic inner core of micelles. Chan et al. synthesized reversible core cross-linked micelles of an amphiphilic block copolymer for drug release studies [106]. The drug loading in the core cross-linked micelles was high (60 wt%) and the hydrolysis of the cross-links due to acidic pH resulted in faster drug release at low pH than at neutral pH. Liu et al. investigated the potential use of double hydrophilic block copolymers as carrier [83]. An aldehyde-modiied ATRP initiator was used to prepare a triblock copolymer consisting of poly(oligo[ethylene glycol]methyl ether methacrylate((POEGMA), PDMAEMA, and PDEAEMA blocks (POEGMA-b-PDMAEMA-b-PDEAEMA) that self-assembled into micelles at alkaline pH and that molecularly dissolved at acidic pH. The middle block was cross-linked at alkaline pH to obtain shell cross-linked micelles that showed reversible pH-responsive swelling/deswelling behavior and formed bioconjugates by covalently attaching lysozyme to the aldehyde functionality of the triblock copolymer. Stenzel wrote a feature article showing how RAFT can be used to synthesize micelle-forming block copolymers for drug encapsulation and release [107].

122

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

NH2

H2N

O NH2

H2N

NH2

H2N

HN

O OH H N—R 2

O

O

Other coupling functional groups O

NH

NH

O NH2

H aldehyde

O

H

O

R— NH

NH—R O

O R

HO carboxylic acid

R = polymer chain, peptide, protein, dye etc. R=

N3 Further functionalization via ‘click chemistry’

O

HN

N

H

H

O

O

H2N—

H

N

Other coupling functional groups HO

H

H3C ketone

O

R= HO

O

O

NH

R

OH

O NH2

H2N O

NH2

NH2

HO

CH3-O

NH

H N O 2

BrMg

O

O

N

Br

Br

Br

Br

CRP grafting Br Br

Br

Br

Br Br Br

N3

Br Br

1) NaN3 2)

N N N

Br N N

N

Br

N3

: polymer chain, protein, enzyme, dye, drug etc.

SCHEME 4.11 Examples of surface modiication reactions for cross-linked poly mer nanoparticles.

Inorganic reactions can also be carried out in so-called poly mer micelles with aqueous core (PMACs). Inorganic compounds such as silica, AgCl, and Fe(SCN)2+ were synthesized in the aqueous core of a polystyrene-b-poly(4vinylpyridine) (PS-b-P4VP) micellar structure [99]. Shell cross-linked micelles of poly([ethylene oxide]-b-glycerol monomethacrylate-b-[diethylamino]ethyl methacrylate) (PEO-b-PGlMA-b-PDEAEMA) were used as pH-responsive particulate emulsifers [108].

Block, Graft, Star, and Gradient Copolymer Particles

123

Reversible cross-linking has been employed in cross-linking of micelles. Controlled release of bioactive agents can be achieved by regulating the reversibility of the cross-linked micelles. Reversible cross-linking of micelles provides stability to the micelles due to cross-links and facilitates faster release of the contents (drugs, polymers, or inorganic products) due to cleavage of the cross-links. Several authors used disulide links as degradable cross-links [80,109–111]. For example, Li et al. synthesized an ABC triblock copolymer that was thermo-responsive and underwent reversible cross-linking [111]. The triblock copolymer PEO-b-P(DMAEMA-stat-NAS)b-PNIPAM was cross-linked through the NAS units using cystamine. The cross-linked micelles were reductively cleaved using dithiothreitol or tris(2-carboxyethyl)phosphine and re-cross-linked by adding cystamine. Molecules that dimerize and decouple due to changes in external stimuli can also be used as reversible cross-links. Babin et al. synthesized a PDMAEMA-b-P(MMA-co-coumarin methacrylate [CMA]) (PDMAEMA54b-P[MMA55 -co-CMA19]) copolymer using ATRP and prepared reversible shell cross-linked poly mer micelles that were stimuli responsive [110]. Thus, poly mer chains in the copolymer were modiied with coumarin molecules which cross-link at UV > 310 nm and de-cross-link at UV < 260 nm.

4.5 Conclusions Nanoparticles of block, graft, star, and gradient copolymers can be synthesized in dispersed systems and via cross-linking of self-assembled copolymers. The two routes to copolymer nanoparticles can be ine-tuned to produce particles of the required size and properties (both chemical and physical). The method for synthesis of copolymer nanoparticles depends entirely on the intended application. Polymerization in dispersed systems can be carried out in either oil-in-water or water-in-oil (inverse dispersion), and large-scale synthesis of copolymer nanoparticles is possible. CRP has been successfully applied to polymerization in dispersed systems to synthesize well-deined copolymer topologies. There is, however, limited literature on the synthesis of complex copolymer architectures such as star and graft copolymers in dispersed systems, although the synthetic tools are at hand. Copolymer nanoparticles synthesized via polymerization in dispersed systems are widely used in industry for various applications. The self-assembly route of making copolymer nanoparticles is relatively simple since the copolymer can be synthesized in any suitable media. Generally, all polymerization techniques can be employed for the synthesis of copolymers that have the ability to self-assemble. The copolymer can be puriied

124

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

before self-assembly and cross-linking. This route, however, applies only to copolymer architectures that are capable of self-assembly. Copolymer nanoparticle modiication is easier on self-assembled nanoparticles, and both surface (shell) and core functionalization can be carried out selectively. Applications of cross-linked self-assembled copolymers are very broad and are not limited to the drug delivery and stimuli- (pH, temperature, ionic strength) responsive ields.

References 1. Shipp, D. A. 2005. Living radical polymerization: Controlling molecular size and chemical functionality in vinyl polymers. Journal of Macromolecular Science, Part C: Polymer Reviews 45:171–94. 2. Gao, H., and K. Matyjaszewski. 2009. Synthesis of functional polymers with controlled architecture by CRP of monomers in the presence of cross-linkers: From stars to gels. Progress in Polymer Science 34:317–50. 3. Hadjichristidis, N., Iatrou, H., Pitsikalis, M., and J. Mays. 2006. Macromolecular acrchitectures by living and controlled/living polymerizations. Progress in Polymer Science 31:1068–1132. 4. Wang, J.-S., and K. Matyjaszewski. 1995. Controlled/“living” radical polymerization. Atom transfer radical polymerization in the presence of transition-metal complexes. Journal of the American Chemical Society 117:5614–15. 5. Matyjaszewski, K., Patten, T. E., and J. Xia. 1997. Controlled/“living” radical polymerization. Kinetics of the homogeneous atom transfer radical polymerization of styrene. Journal of the American Chemical Society 119:674–80. 6. Chiefari, J., Chong, B. Y. K., Ercole, F., Krstina, J., Jeffery, J., Le, T. P. T., Mayadunne, R. T. A., Meijs, G. F., Moad, C. L., Rizzardo, E., and S. H. Thang. 1998. Living free-radical polymerization by reversible addition-fragmentation chain transfer: The RAFT process. Macromolecules 31:5559–62. 7. Perrier, S., and P. Takolpuckdee. 2005. Macromolecular design via reversible addition-fragmentation chain transfer (RAFT)/Xanthates (MADIX) polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 43:5347–93. 8. Lu, J., Liang, H., Li, A., and Q. Cheng. 2004. Synthesis of block and graft copolymers of β-pinene and styrene by transformation of living cationic polymerization to atom transfer radical polymerization. European Polymer Journal 40:397–402. 9. Quinn, J. F., Chaplin, R. P., and T. P. Davis. 2002. Facile synthesis of comb, star, and graft polymers via reversible addition-fragmentation chain transfer (RAFT) polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 40:2956–66. 10. Min, K., Yu, S., Lee, H.-I., Mueller, L., Sheiko, S. S., and K. Matyjaszewski. 2007. High yield synthesis of molecular brushes via ATRP in miniemulsion. Macromolecules 40:6557–63.

Block, Graft, Star, and Gradient Copolymer Particles

125

11. Tsarevsky, N. V., Bencherif, S. A., and K. Matyjaszewski. 2007. Graft copolymers by a combination of ATRP and two different consecutive click reactions. Macromolecules 40:4439–45. 12. Tsoukatos, T., Pispas, S., and N. Hadjichristidis. 2000. Complex macromolecular architectures by combining TEMPO living free radical and anionic polymerization. Macromolecules 33:9504–11. 13. Liu, P. 2005. Modiication of carbon nanotubes with polymers. European Polymer Journal 41:2693–2703. 14. Matyjaszewski, K., Dong, H., Jakubowski, W., Pietrasik, J., and A. Kusumo. 2007. Grafting from surfaces for “everyone”: ARGET ATRP in the presence of air. Langmuir 23:4528–31. 15. Stenzel, M. H., Zhang, L., and W. T. S. Huck. 2006. Temperature-responsive glycopolymer brushes synthesized via RAFT polymerization using the Z-group approach. Macromolecular Rapid Communications 27:1121–26. 16. Li, C., Han, J., Ryu, C. Y., and B. C. Benicewicz. 2006. A versatile method to prepare RAFT agent anchored substrates and the preparation of PMMA grafted nanoparticles. Macromolecules 39:3175–83. 17. Roy, D., Guthrie, J. T., and S. Perrier. 2005. Graft polymerization: Grafting poly(styrene) from cellulose via reversible addition-fragmentation chain transfer (RAFT) polymerization. Macromolecules 38:10363–72. 18. Tessier, L., Chancolon, J., Alet, P.-J., Trenggono, A., Mayne-l’Hermite, M., Deniau, G., Jégou, P., and S. Palacin. 2008. Grafting organic polymer ilms on surfaces electroinitiated emulsion polymerization. Physica Status Solidi a 205:1412–18. 19. Barsbay, M., Güven, O., Davis, T. P., Barner-Kowollik, C., and L. Barner. 2009. RAFT-mediated polymerization and grafting of sodium 4-styrenesulfonate from cellulose initiated via γ-radiation. Polymer 50:973–82. 20. Esteves, A. C. C., Bombalski, L., Trindade, T., and K. Matyjaszewski. 2007. Polymer grafting from CdS quantum dots via AGET ATRP in miniemulsion. Small 3:1230–36. 21. Fleet, R., McLeary, J. B., Grumel, V., Weber, W. G., Matahwa, H., and R. D. Sanderson. 2008. RAFT mediated polysaccharide copolymers. European Polymer Journal 44:2899–2911. 22. Riva, R., Rieger, J., Jérôme, R., and P. Lecomte. 2006. Heterograft copolymers of poly(ε-caprolactone) prepared by combination of ATRA “grafting onto” and ATRP “grafting from” processes. Journal of Polymer Science, Part A: Polymer Chemistry 44:6015–24. 23. Spiniello, M., Blencowe, A., and G. G. Qiao. 2008. Synthesis and characterization of luorescently labeled core cross-linked star polymers. Journal of Polymer Science, Part A: Polymer Chemistry 46:2422–32. 24. Chan, J. W., Yu, B., Hoyle, C. E., and A. B. Lowe. 2008. Convergent synthesis of 3-arm star polymers from RAFT-prepared poly(N,N-diethylacrylamide) via a thiol-ene click reaction. Chemical Communications 4959–61. 25. Miura, Y., Narumi, A., Matsuya, S., Satoh, T., Duan, Q., Kaga, H., and T. Kakuchi. 2005. Synthesis of well-deined AB20-type star polymers with cyclodextrin-core by combination of NMP and ATRP. Journal of Polymer Science, Part A: Polymer Chemistry 43:4271–79. 26. Chaffey-Millar, H., Stenzel, M. H., Davis, T. P., Coote, M. L., and C. BarnerKowollik. 2006. Design criteria for star polymer formation processes via living free radical polymerization. Macromolecules 39:6406–19.

126

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

27. Bernard, J., Favier, A., Zhang, L., Nilasaroya, A., Davis, T. P., Barner-Kowollik, C., and M. H. Stenzel. 2005. Poly(vinyl ester) star polymers via xanthate-mediated living radical polymerization: From poly(vinyl alcohol) to glycopolymer stars. Macromolecules 38:5475–84. 28. Ugelstad, J., El-Aasser, M. S., and J. W. Vanderhoff. 1973. Emulsion polymerization: Initiation of polymerization in monomer droplets. Polymer Letters Edition 11:503–13. 29. Asua, J. M. 2002. Miniemulsion polymerization. Progress in Polymer Science 27:1283–1346. 30. Antonietti, M., and K. Landfester. 2002. Polyreactions in miniemulsions. Progress in Polymer Science 27:689–757. 31. Bardajee, G. R., Vancaeyzeele, C., Haley, J. C., Li, A. Y., and M. A. Winnik. 2007. Synthesis, characterization, and energy transfer studies of dye-labeled poly(butyl methacrylate) latex particles prepared by miniemulsion polymerization. Polymer 48:5839–49. 32. Reimers, J., and F. J. Schork. 1996. Robust nucleation in polymer-stabilized miniemulsion polymerization. Journal of Applied Polymer Science 59:1833–41. 33. Liu, S., Hermanson, K. D., and E. W. Kaler. 2006. Reversible additionfragmentation chain transfer polymerization in microemulsion. Macromolecules 39:4345–50. 34. Save, M., Guillaneuf, Y., and R. G. Gilbert. 2006. Controlled radical polymerization in aqueous dispersed media. Australian Journal of Chemistry 59:693–711. 35. Prescott, S. W., Ballard, M. J., Rizzardo, E., and R. G. Gilbert. 2002. RAFT in emulsion polymerization: What makes it different? Australian Journal of Chemistry 55:415–24. 36. Oh, J. K. 2008. Recent advances in controlled/living radical polymerization in emulsion and dispersion. Journal of Polymer Science, Part A: Polymer Chemistry 46:6983–7001. 37. Cunningham, M. F. 2008. Controlled/living radical polymerization in aqueous dispersed systems. Progress in Polymer Science 33:365–98. 38. McLeary, J. B., and B. Klumperman. 2006. RAFT mediated polymerisation in heterogeneous media. Soft Matter 2:45–53. 39. Matahwa, H., McLeary, J. B., and R. D. Sanderson. 2006. Comparative study of classical surfactants and polymerizable surfactants (surfmers) in the reversible addition-fragmentation chain transfer mediated miniemulsion polymerization of styrene and methyl methacrylate. Journal of Polymer Science, Part A: Polymer Chemistry 44:427–42. 40. McLeary, J. B., Tonge, M. P., De Wet-Roos, D., Sanderson, R. D., and B. Klumperman. 2004. Controlled, radical reversible addition-fragmentation chain-transfer polymerization in high-surfactant-concentration ionic miniemulsions. Journal of Polymer Science, Part A: Polymer Chemistry 42:960–74. 41. De Brouwer, H., Tsavalas, J. G., Schork, F. J., and M. J. Monteiro. 2000. Living radical polymerization in miniemulsion using reversible addition-fragmentation chain transfer. Macromolecules 33:9239–46. 42. Nicolas, J., Charleux, B., Guerret, O., and S. Magnet. 2004. Novel SG1-based water-soluble alkoxyamine for nitroxide-mediated controlled free-radical polymerization of styrene and n-butyl acrylate in miniemulsion. Macromolecules 37:4453–63.

Block, Graft, Star, and Gradient Copolymer Particles

127

43. Stoffelbach, F., Tibiletti, L., Rieger, J., and B. Charleux. 2008. Surfactant-free, controlled/living radical emulsion polymerization in batch conditions using a low molar mass, surface-active reversible addition-fragmentation chaintransfer (RAFT) agent. Macromolecules 41:7850–56. 44. Rieger, J., Stoffelbach, F., Bui, C., Alaimo, D., Jérôme, C., and B. Charleux. 2008. Amphiphilic poly(ethylene oxide) macromolecular RAFT agent as a stabilizer and control agent in ab initio batch emulsion polymerization. Macromolecules 41:4065–68. 45. Ferguson, C. J., Hughes, R. J., Pham, B. T. T., Hawkett, B. S., Gilbert, R. G., Serelis, A. K., and C. H. Such. 2002. Effective ab initio emulsion polymerization under RAFT control. Macromolecules 35:9243–45. 46. Ferguson, C. J., Hughes, R. J., Nguyen, D., Pham, B. T. T., Gilbert, R. G., Serelis, A. K., Such, C. H., and B. S. Hawkett. 2005. Ab initio emulsion polymerization by RAFT-controlled self-assembly. Macromolecules 38:2191–2204. 47. Ganeva, D. E., Sprong, E., De Bruyn, H., Warr, G. G., Such, C. H., and B. S. Hawkett. 2007. Particle formation in ab initio RAFT mediated emulsion polymerization systems. Macromolecules 40:6181–89. 48. Bozovic-Vukic, J., Mañon, H. T., Meuldijk, J., Koning, C., and B. Klumperman. 2007. SAN-b-P4VP block copolymer synthesis by chain extension from RAFTfunctional poly(4-vinylpyridine) in solution and in emulsion. Macromolecules 40:7132–39. 49. Dire, C., Magnet, S., Couvreur, L., and B. Charleux. 2009. Nitroxide-mediated controlled/living free-radical surfactant-free emulsion polymerization of methyl methacrylate using a poly(methacrylic acid)-based macroalkoxyamine initiator. Macromolecules 42:95–103. 50. Delaittre, G., and B. Charleux. 2008. Kinetics of in-situ formation of poly(acrylic acid)-b-polystyrene amphiphilic block copolymers via nitroxidemediated controlled free-radical emulsion polymerization. Discussion on the effect of compartmentalization on the polymerization rate. Macromolecules 41:2361–67. 51. Delaittre, G., Nicolas, J., Lefay, C., Save, M., and B. Charleux. 2005. Surfactantfree synthesis of amphiphilic diblock copolymer nanoparticles via nitroxidemediated emulsion polymerization. Chemical Communications 614–16. 52. Nicolas, J., Charleux, B., Guerret, O., and S. Magnet. 2004. Nitroxide-mediated controlled free-radical emulsion polymerization of styrene and n-butyl acrylate with a water-soluble alkoxyamine as initiator. Angewandte Chemie, International Edition English 43:6186–89. 53. Delaittre, G., Nicolas, J., Lefay, C., Save, M., and B. Charleux. 2006. Aqueous suspension of amphiphilic diblock copolymer nanoparticles prepared in situ from a water-soluble poly(sodium acrylate) alkoxyamine macroinitiator. Soft Matter 2:223–31. 54. Prescott, S. W., Ballard, M. J., Rizzardo, E., and R. G. Gilbert. 2002. Successful use of RAFT techniques in seeded emulsion polymerization of styrene: Living character, RAFT agent RAFT agent transport, and rate of polymerization. Macromolecules 35:5417–25. 55. Luo, Y., Tsavalas, J., and F. J. Schork. 2001. Theoretical aspects of particle swelling in living free radical miniemulsion polymerization. Macromolecules 34:5501–7.

128

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

56. Bowes, A., McLeary, J. B., and R. D. Sanderson. 2007. AB and ABA type butyl acrylate and styrene block copolymers via RAFT-mediated miniemulsion polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 45:588–604. 57. Huang, X., Sudol, E. D., Dimonie, V. L., Anderson, C. D., and M. S. El-Aasser. 2006. Stability in styrene/HD miniemulsions containing a RAFT agent. Macromolecules 39:6944–50. 58. Thomas, D. B., Convertine, A. J., Hester, R. D., Lowe, A. B., and C. L. McCormick. 2004. Hydrolytic susceptibility of dithioester chain transfer agents and implications in aqueous RAFT polymerizations. Macromolecules 37:1735–41. 59. McCormick, C. L., and A. B. Lowe. 2004. Aqueous RAFT polymerization: Recent developments in synthesis of functional water-soluble (co)polymers with controlled structures. Accounts of Chemical Research 37:312–25. 60. Barner-Kowollik, C., Buback, M., Charleux, B., Coote, M. L., Drache, M., Fukuda, T., Goto, A., Klumperman, B., Lowe, A. B., McLeary, J. B., Moad, G., Monteiro, M. J., Sanderson, R. D., Tonge, M. P., and P. Vana. 2006. Mechanism and kinetics of dithiobenzoate-mediated RAFT polymerization. 1. The current situation. Journal of Polymer Science, Part A: Polymer Chemistry 44:5809–31. 61. Lansalot, M., Davis, T. P., and J. P. A. Heuts. 2002. RAFT miniemulsion polymerization: Inluence of the structure of the RAFT agent. Macromolecules 35:7582–91. 62. Min, K., Gao, H., and K. Matyjaszewski. 2005. Preparation of homopolymers and block copolymers in miniemulsion by ATRP using activators generated by electron transfer (AGET). Journal of the American Chemical Society 127:3825–30. 63. Oh, J. K., Perineau, F., and K. Matyjaszewski. 2006. Preparation of nanoparticles of well-controlled water-soluble homopolymers and block copolymers using an inverse miniemulsion ATRP. Macromolecules 39:8003–10. 64. Matyjaszewski, K., Shipp, D. A., Qiu, J., and S. G. Gaynor. 2000. Water-borne block and statistical copolymers synthesized using atom transfer radical polymerization. Macromolecules 33:2296–98. 65. Kagawa, Y., Minami, H., Okubo, M., and J. Zhou. 2005. Preparation of block copolymer particles by two-step atom transfer radical polymerization in aqueous media and its unique morphology. Polymer 46:1045–49. 66. Stoffelbach, F., Griffete, N., Bui, C., and B. Charleux. 2008. Use of a simple surface-active initiator in controlled/living free-radical miniemulsion polymerization under AGET and ARGET ATRP conditions. Chemical Communications 4807–9. 67. Min, K., Gao, H., and K. Matyjaszewski. 2007. Use of ascorbic acid as reducing agent for synthsis of well-deined polymers by ARGET ATRP. Macromolecules 40:1789–91. 68. Jakubowski, W., Min, K., and K. Matyjaszewski. 2006. Activators regenerated by electron transfer for atom transfer radical polymerization of styrene. Macromolecules 39:39–45. 69. Keoshkerian, B., MacLeod, P. J., and M. K. Georges. 2001. Block copolymer synthesis by a miniemulsion stable free radical polymerization process. Macromolecules 34:3594–99. 70. Farcet, C., Charleux, B., and R. Pirri. 2001. Poly(n-butyl acrylate) homopolymer and poly[n-butyl acrylate-b-(n-butyl acrylate-co-styrene)] block copolymer prepared via nitroxide-mediated living/controlled radical polymerization in miniemulsion. Macromolecules 34:3823–26.

Block, Graft, Star, and Gradient Copolymer Particles

129

71. Farcet, C., Lansalot, M., Charleux, B., Pirri, R., and J. P. Vairon. 2000. Mechanistic aspects of nitroxide-mediated controlled radical polymerization of styrene in miniemulsion, using a water-soluble radical initiator. Macromolecules 33:8559–70. 72. Min, K., Oh, J. K., and K. Matyjaszewski. 2007. Preparation of gradient copolymers via ATRP in miniemulsion. II. Forced gradient. Journal of Polymer Science, Part A: Polymer Chemistry 45:1413–23. 73. Zhang, Y., Liu, H., Hu, H., Hu, J., Li, C., and S. Liu. 2009. Synthesis and aggregation behavior of multi-responsive double hydrophilic ABC miktoarm star terpolymer. Macromolecular Rapid Communications 30:941–47. 74. Zhang, J., Zhou, Y., Zhu, Z., Ge, Z., and S. Liu. 2008. Polyion complex micelles possessing thermoresponsive coronas and their covalent core stabilization via “click” chemistry. Macromolecules 41:1444–54. 75. Ning, F., Jiang, M., Mu, M., Duan, H., and J. Xie. 2002. Synthesis of amphiphilic block-graft copolymers [poly(styrene-b-etylene-co-butylene-b-styrene)-gpoly(acrylic acid)] and their aggregation in water. Journal of Polymer Science, Part A: Polymer Chemistry 40:1253–66. 76. Chiang, W.-H., Hsu, Y.-H., Lou, T.-W., Chern, C.-H., and H.-C. Chiu 2009. Effects of mPEG grafts on morphology and cross-linking of thermally induced micellar assemblies from PAAc based graft copolymers in aqueous phase. Macromolecules 42:3611–19. 77. Halles, M., Barner-Kowollik, C., Davis, T. P., and M. H. Stenzel. 2004. Shell-crosslinked vesicles synthesized from block copolymers of poly(D,L-lactide) and poly(N-isopropyl acrylamide) as thermoresponsive nanocontainers. Langmuir 20:10809–17. 78. Shuai, X., Merdan, T., Schaper, A. K., Xi, F., and T. Kissel. 2004. Core-cross-linked polymeric micelles as paclitaxel carriers. Bioconjugate Chemistry 15:441–8. 79. Murphy, J. J., Kawasaki, T., Fujiki, M., and K. Nomura. 2005. Precise synthesis of amphiphilic polymeric architectures by grafting poly(ethylene glycol) to endfunctionalized block ROMP copolymers. Macromolecules 38:1075–83. 80. Schacher, F., Walther, A., Ruppel, M., Drechsler, M., and A. H. E. Müller. 2009. Multicompartment core micelles of triblock terpolymers in organic media. Macromolecules 42:3540–48. 81. Yu, X., Shi, T., An, L., Zhang, G., and P. K. Dutta. 2007. Synthesis of a H-shaped amphiphilic block copolymer by the combination of atom transfer radical polymerization and living anionic polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 45:147–56. 82. Chen, D., and M. Jiang. 2005. Strategies for constructing polymeric micelles and hollow spheres in solution via speciic intermolecular interactions. Accounts of Chemical Research 38:494–502. 83. Liu, H., Jiang, X., Fan, J., Wang, G., and S. Liu. 2007. Aldehyde surfacefunctionalized shell cross-linked micelles with pH-tunable core swellability and their bioconjugation with lysozyme. Macromolecules 40:9074–83. 84. Lorge, T. P., Rasdal, A., Li, Z., and M. A. Hilmer. 2005. Simultaneous, segregated storage of two agents in a multicompartment micelle. Journal of the American Chemical Society 127:17608–9. 85. Mao, J., Ni, P., Mai, Y., and D. Yan. 2007. Multicompartment micelles from hyperbranched star-block copolymers containing polycations and luoropolymer segment. Langmuir 23:5127–34.

130

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

86. Peng, D., Zhang, X., and X. Huang. 2006. Novel starlike amphiphilic graft copolymers with hydrophilic poly(acrylic acid) backbone and hydrophobic poly(methyl methacrylate) side chains. Macromolecules 39:4945–47. 87. Zhang, W., Zhang, W., Zhou, N., Cheng, Z., Zhu, J., and X. Zhu. 2008. Synthesis and self-assembly behaviors of three-armed amphiphilic block copolymers via RAFT polymerization. Polymer 49:4569–75. 88. Qian, J., and F. Wu. 2009. Preparation of supramolecular graft copolymers and the subsequent formation of pH-sensitive vesicles. Chemistry of Materials 21:758–62. 89. Chiang, W.-H., Hsu, Y.-H., Chern, C.-H., and H.-C. Chiu. 2009. Thermally induced polymeric assemblies from the PAAc-based copolymer containing both PNIPAAm and mPEG grafts in water. Journal of Physical Chemistry B 113:4187–96. 90. Gu, L., Shen, Z., Zhang, S., Lu, G., Zhang, X., and X. Huang. 2007. Novel amphiphilic centipede-like copolymer bearing polyacrylate backbone and poly(ethylene glycol) and polystyrene side chains. Macromolecules 40:4486–93. 91. Astaieva, I., Zhong, X., and A. Eisenberg. 1993. Critical micellization phenomena in block polyelectrolyte solutions. Macromolecules 26:7339–52. 92. Förster, S., Abetz, V., and A. H. E. Müller. 2004. Polyelectrolyte block copolymer micelles. Advances in Polymer Science 166:173–210. 93. Sunintaboon, P., Ho, K. M., Li, P., Cheng, S. Z. D., and F. W. Harris. 2006. Formation of nanostructured materials via coalescence of amphiphilic hollow particles. Journal of the American Chemical Society 128:2168–69. 94. Stubenrauch, K., Moitzi, C., Fritz, G., Glatter, O., Trimmel, G., and F. Stelzer. 2006. Precise tuning of micelle, core, and shell size by the composition of amphiphilic block copolymers derived from ROMP investigated by DLS and SAXS. Macromolecules 39:5865–74. 95. Houillot, L. Bui, C., Save, M., Charleux, B., Farcet, C., Moire, C., Raust, J.-A., and I. Rodriguez. 2007. Synthesis of well-deined polyacrylate particle dispersions in organic medium using simultaneous RAFT polymerization and self-assembly of block copolymers. A strong inluence of the selected thiocarbonylthio chain transfer agent. Macromolecules 40:6500–9. 96. Zhang, Y., Wang, Z., Zhao, J., and C. Wu. 2007. Facile preparation of pHresponsive gelatin-based core-shell polymeric nanoparticles at high concentrations via template polymerization, Polymer 48:5639–45. 97. Germack, D. S., Harrison, S., Brown, G. O. and K. L. Wooley. 2006. Inluence of the structure of nanoscopic building blocks on the assembly of micropatterned surfaces. Journal of Polymer Science, Part A: Polymer Chemistry 44:5218–28. 98. Narain, R., and S. P. Armes. 2003. Synthesis and aqueous solution properties of novel sugar methacrylate-based homopolymers and block copolymers. Biomacromolecules 4:1746–58. 99. Cheng, F., Yang, X., Peng, H., Chen, D., and M. Jiang. 2007. Well-controlled formation of polymeric micelles with a nanosized aqueous core and their applications as nanoreactors. Macromolecules 40:8007–14. 100. O’Reilly, R. K., Hawker, C. J., and K. L. Wooley. 2006. Cross-linked block copolymer micelles: Functional nanostructures of great potential and versatility. Chemical Society Reviews 35:1068–83. 101. Guo, A., Liu, G., and J. Tao. 1996. Star polymers and nanospheres from crosslinkable diblock copolymers. Macromolecules 29:2487–93.

Block, Graft, Star, and Gradient Copolymer Particles

131

102. Bütün, V., Bilingham, N. C., and S. P. Armes. 1998. Synthesis of shell cross-linked micelles with tunable hydrophilic/hydrophobic cores. Journal of the American Chemical Society 120:12135–36. 103. Liu, S., and S. P. Armes. 2001. The facile one-pot synthesis of shell cross-linked micelles in aqueous solution at high solids. Journal of the American Chemical Society 123:9910–11. 104. Bütün, V., Wang, X.-S., de Paz Báñez, M. V., Robinson, K. L., Bilingham, N. C., and S. P. Armes. 2000. Synthesis of shell cross-linked micelles at high solids in aqueous media. Macromolecules 33:1–3. 105. Qi, K., Ma, Q., Remsen, E. E., Clark, J. C. G., and K. L. Wooley. 2004. Determination of the bioavailability of biotin conjugated onto shell cross-linked (SCK) nanoparticles. Journal of the American Chemical Society 126:6599–6607. 106. Chan, Y., Wong, T., Byrne, F., Kavallaris, M., and V. Bulmus. 2008. Acidlabile core cross-linked micelles for pH-triggered release of antitumor drugs. Biomacromolecules 9:1826–36. 107. Stenzel, M. H. 2008. RAFT polymerization: An avenue to functional polymeric micelles for drug delivery. Chemical Communications 3486–3503. 108. Fujii, S., Cai, Y., Weaver, J. V. M., and S. P. Armes. 2005. Syntheses of shell crosslinked micelles using acidic ABC triblock copolymers and their application as pH-responsive particulate emulsiiers. Journal of the American Chemical Society 127:7304–5. 109. Zhang, Y., Liu, W., Lin, L., Chen, D., and M. H. Stenzel. 2008. Degradable disulide core-cross-linked micelles as a drug delivery system prepared from vinyl functionalized nucleosides via the RAFT process. Biomacromolecules 9:3321–31. 110. Babin, J., Lepage, M., and Y. Zhao. 2008. “Decoration” of shell cross-linked reverse polymer micelles using ATRP: A new route to stimuli-responsive nanoparticles. Macromolecules 41:1246–53. 111. Li, Y., Lokitz, B. S., Armes, S. P., and C. L. McCormick. 2006. Synthesis of reversible shell cross-linked micelles for controlled release of bioactive agents. Macromolecules 39:2726–28.

5 Polymer Nanoparticles by Reversible Addition-Fragmentation Chain Transfer Microemulsion Polymerization J. O’Donnell and E. Kaler CONTENTS 5.1 5.2

Introduction ................................................................................................ 134 Uncontrolled Microemulsion Polymerization ....................................... 135 5.2.1 Impact of Monomer Solubility ..................................................... 136 5.2.2 Microemulsion Copolymerization .............................................. 138 5.2.3 Multiple Addition and Semicontinuous Microemulsion Polymerization ............................................................................... 139 5.3 Reversible Addition-Fragmentation Chain Transfer Polymerization ........................................................................................... 140 5.4 Reversible Addition-Fragmentation Chain Transfer Microemulsion Polymerization ............................................................... 143 5.4.1 Microemulsion Polymerization Kinetics .................................... 144 5.4.1.1 Uncontrolled Microemulsion Polymerization Kinetics ............................................................................. 145 5.4.1.2 RAFT Microemulsion Polymerization Kinetics ......... 147 5.4.1.3 Model Results .................................................................. 150 5.4.1.4 Predicted RAFT Microemulsion Polymerization Kinetics with Negligible Biradical Termination ......... 150 5.4.1.5 Predicted RAFT Microemulsion Polymerization Kinetics with Biradical Termination ............................ 152 5.4.2 Molecular Weight and Polydispersity......................................... 153 5.4.2.1 Effect of Chain Transfer Agent per Micelle Ratio ...... 155 5.4.2.2 Effect of Monomer Solubility ........................................ 155 5.4.2.3 Effect of Chain Transfer Agent Solubility ................... 158 5.4.3 Latex Particle Size .......................................................................... 161 References............................................................................................................. 163

133

134

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

5.1 Introduction Microemulsion polymerization produces small latex nanoparticles (D < 50 nm) of high molecular weight (MN = 106 to 107 g/mol) polymer, and provides several advantages relative to other types of heterogeneous polymerizations, such as rapid reaction times and a product that is colloidally stable. Since the introduction of microemulsion polymerization by Stoffer and Bone [1] and Atik and Thomas [2] in the early 1980s, this technique has been widely studied because of the dramatic beneits the high surface area-to-volume ratio of the polymer particles provides for many applications including sensors [3], luorescence markers [4], and conductive ilms [5]. Control of the microstructural properties (such as molecular weight, polydispersity, monomer sequences, chain ends, and degree of branching) in a microemulsion polymerization could lead to enhanced chemical and mechanical properties and an even broader range of applications. Of the many controlled free radical polymerization techniques that have been developed to control these microstructural properties, including nitroxide-mediated polymerization (NMP) [6–9], atom transfer radical polymerization (ATRP) [10–13], and degenerative transfer (DT) [14,15], reversible addition-fragmentation chain transfer (RAFT) polymerization has emerged as the most promising due to its versatility and simplicity [16,17]. Also, in a RAFT polymerization the resulting polymer is free from the contamination of metal catalysts. Although the RAFT process has shown the ability to control a broad array of monomers over a wide range of polymerization conditions in numerous bulk and solution polymerizations, very few RAFT polymerizations have been successfully conducted in a heterogeneous polymerization environment [18–20]. Incorporating RAFT into heterogeneous polymerizations (such as emulsion or miniemulsion) frequently leads to poor control of molecular weight, high polydispersity, and loss of colloidal stability [18,21]. Additionally, rate retardation is often more signiicant in heterogeneous polymerizations than in homogeneous polymerizations [21]. The increase in rate retardation in miniemulsions relative to bulk has been shown to result from compartmentalization effects [22] and increased radical exit from the polymerizing particles [23]. Microemulsion polymerizations are advantageous relative to emulsion or miniemulsion polymerizations due to the absence of large monomer droplets. Liu et al. were the irst to demonstrate the feasibility of incorporating RAFT in oil-in-water microemulsion polymerizations while maintaining both control of the polymerization and colloidal stability of the polymerizing latex [24]. Subsequently, several series of RAFT microemulsion poly merizations of butyl acrylate and 2-ethylhexyl acrylate with the chain transfer agents methyl-2-(O-ethylxanthyl) propionate and methyl-2(O-dodecylxanthyl) propionate were performed to elucidate the effects of the chain transfer agent

Polymer Nanoparticles by RAFT Microemulsion Polymerization

135

molecule per micelle ratio, and the monomer and chain transfer agent aqueous solubilities on the control of the polymerization, and the inal polymer and latex characteristics [25,26]. To date, RAFT has not been incorporated into inverse water-in-oil microemulsion polymerizations. This chapter will irst review the key structure and transport properties that inluence the inal polymer and latex characteristics in an uncontrolled oil-in-water microemulsion polymerization. Then, the RAFT reaction mechanism will be reviewed, and the incorporation of RAFT into microemulsion polymerization will be described. Next, the development of a kinetic model to describe RAFT microemulsion polymerization will be summarized because it provides insight into the key parameters that affect the inal polymer and latex properties. Finally, the characteristics of the polymers produced by RAFT microemulsion polymerization to date will be summarized with respect to the key parameters of chain transfer agent molecule per micelle ratio, and aqueous solubility of the monomer and the chain transfer agent.

5.2 Uncontrolled Microemulsion Polymerization Polymerization of one-phase, thermodynamically stable oil-in-water microemulsions produces colloidally stable latex nanoparticles containing polymers of high molecular weight and low polydispersity [27–31]. Microemulsion poly mer i za tion begins when initiator-derived radicals (I•) react with the small amount of hydrophobic monomer (M) in the aqueous domain to form propagating polymers (P•) (Figure 5.1(1.)). The polymers propagate in the aqueous domain until a critical degree of polymerization is reached. At the critical degree of polymerization, the polymer is no longer soluble in the aqueous 1. Initiation

2. Propagation Pn–1

M

P

Polymer Particle

M Pn M

I

M

M

M

M

3. Final Latex Dispersion

M Empty Micelles

FIGURE 5.1 Typical oil-in-water microemulsion poly merization initiated with a water-soluble initiator. (1.) Initiator-derived radical (I•) reacts with monomer (M) in the aqueous domain to form a propagating poly mer (P•), which partitions into a monomer-swollen micelle to form a particle. (2.) Propagation of a poly mer in a particle with monomer diffusing from uninitiated micelles to the locus of poly merization. (3.) Final latex dispersion of surfactant-stabilized poly mer particles and empty micelles.

136

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

domain, and the polymer partitions into a monomer-swollen micelle, thus forming a surfactant-stabilized polymer particle. In a typical microemulsion polymerization, the concentration of micelles is approximately 1000 times greater than the concentration of polymer particles, so the probability of a propagating polymer entering a particle as opposed to a micelle and causing a termination reaction is negligible. Propagation continues in the polymer particles as monomer diffuses from surrounding uninitiated micelles to the locus of polymerization (Figure 5.1(2.)). The inal polymer latex consists of surfactant-stabilized particles containing a single poly mer and empty micelles (Figure 5.1(3.)). The transport and partitioning of monomer during a microemulsion polymerization depends on the aqueous solubility of the monomer. As such, the aqueous solubility of the monomer has a signiicant effect on the inal polymer and latex characteristics, as well as the ability to perform semicontinuous or multiple addition microemulsion (co)polymerizations. The following sections demonstrate the complications induced by monomer transport and partitioning. 5.2.1 Impact of Monomer Solubility The water solubility of the monomer dictates both the residence time of an oligomeric radical in a monomer-swollen micelle (τres) and the critical degree of polymerization (zcrit) that a propagating chain must reach before being segregated into a surfactant-stabilized particle. Decreasing the water solubility of the monomer decreases zcrit and increases τres, so an oligomeric radical explores fewer micelles and the probability of biradical termination is greatly reduced. The probability of an oligomeric radical entering a polymer particle and terminating the propagating polymer can be quantiied by the ratio of the characteristic propagation time (τprop) to τres. τprop is the average time required for a propagating polymer to add a single monomer unit, which depends on the propagation rate constant (kp) and the concentration of monomer in the micelle (Cmic), τ prop ≈

1 k p Cmic

(5.1)

τres can be approximated as the characteristic time for diffusion of a monomeric radical from the surface of a sphere immersed in an ininite medium [32], which is τ res ≈ q

2 R mic 3Daq mon

(5.2)

Polymer Nanoparticles by RAFT Microemulsion Polymerization

137

where q is the partition coeficient of the monomer between the micelles and the aqueous domain, Rmic is the radius of the micelle, and D aq mon is the aqueous diffusion coeficient of the monomer. The τprop /τres ratio is the number of micelles that a radical explores before adding one monomer unit. This ratio must be multiplied by zcrit to estimate the total number of micelles that an oligomeric radical enters and exits before being segregated into a surfactantstabilized polymer particle. If zcrit τprop /τres is greater than the ratio of micelles to polymer particles, then biradical termination is statistically possible. Bleger et al. demonstrated the effect of the aqueous solubility of the monomer on particle nucleation [33]. These authors investigated microemulsion polymerizations of methyl methacrylate, which has high aqueous solubility, and found an initial period of slow polymerization during which the concentration of particles did not signiicantly increase. After this initial period of slow polymerization, the rate and the particle concentration increased rapidly, indicating fast particle nucleation. The transition from slow to fast polymerization was hypothesized to signify a transition from homogeneous nucleation to nucleation by micelle entry. This hypothesis was supported by the observation that styrene, which has a much lower water solubility than methyl methacrylate, and therefore partitions into the micelles at a lower degree of polymerization, experienced a shorter period of slow poly merization. The aqueous solubility of the monomer also inluences the phase behavior, and thus the microstructure, of a microemulsion [34]. The link between phase behavior and microstructure can be understood qualitatively in terms of the curvature elastic energy of the surfactant-rich ilm that separates the water and oil (or polymer) domains [35]. On one hand, the swelling of polymer particles by monomer is promoted by the decrease in free energy of the bulk polymer as it is diluted, but on the other hand movement of the monomer to the polymer is opposed by the free energy penalty paid by changing the curvature elastic energy of the surfactant micelles that are swollen with monomer. Thus, there is a delicate balance between the curvature energy of the surfactant ilm and the solubility of the monomer in the polymer, and the result may lead to complex variations in particle diameter, molecular weight, and the ability to form copolymers as a function of monomer solubility. Capek et al. [30,36] have investigated series of alkyl acrylates and alkyl methacrylates to determine the effect of monomer solubility on the particle diameter, colloidal stability, and polymerization kinetics. Alkyl acrylates with more than four methyl groups are able to act as coemulsiiers, which increases the stability of the polymerizing particles and leads to the formation of smaller particles. Smaller alkyl acrylates are unable to serve as coemulsiiers, so agglomeration of the polymerizing particles occurs during these polymerizations and dramatically increases the particle diameter. The particle diameter of the more hydrophobic methacrylates does not depend on the length of the alkyl chain.

138

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

5.2.2 Microemulsion Copoly merization The synthesis of copolymers by microemulsion polymerization is complicated because of the partitioning of monomer between the micelles, the polymer particles, and the aqueous domain. Gan et al. have shown that the reactivity ratios measured in microemulsion copolymerizations of styrene and methyl methacrylate deviate signiicantly from those measured in bulk polymerizations [37] and attribute this to uneven partitioning. Bahwal et al. have investigated the impact of monomer partitioning on microemulsion copolymerizations by systematically varying the relative aqueous solubilities and polarities of the monomers [38]. As anticipated, the copoly merization of monomers with similar aqueous solubility and polarity, such as methyl methacrylate and ethyl acrylate, results in copolymers with a composition identical to that produced by bulk copolymerization. The combination of monomers with different aqueous solubilities and polarities produces unexpected results for microemulsion copolymerization [38]. When styrene (low aqueous solubility, nonpolar) is copolymerized with methyl acrylate (high aqueous solubility, polar), the ratio of styrene to methyl acrylate in the copolymer is expected to far exceed the monomer feed ratio because a signiicant amount of methyl acrylate will remain in the aqueous domain. However, when the monomer ratios were corrected for the aqueous solubility of the monomers, the authors found that the ratio of methyl acrylate in the copolymer was greater than anticipated from the monomer feed. The increased presence of methyl acrylate in the copolymer relects the partitioning of the monomers within the polymer particles and the signiicance of the locus of polymerization. The more polar methyl acrylate preferentially partitions into the surfactant tails, while the nonpolar styrene partitions into the growing particle core. Therefore, because polymerization is occurring in the corona of the polymer particle, a greater degree of methyl acrylate than styrene is incorporated into the copolymers than would be predicted based on the aqueous solubilities. The effect of monomer partitioning within the polymer particles was conirmed by polymerizing two monomers with similar aqueous solubilities but different polarities (styrene and butyl acrylate). The copolymer showed a greater composition of the more polar butyl acrylate than predicted from the feed composition. Several groups have studied microemulsion copolymerizations of a hydrophobic and a hydrophilic monomer [39–42]. Capek and Juranicova studied the microemulsion copolymerization of butyl acrylate and acrylonitrile in a three-component system, so as to simplify the kinetics [42]. Increasing the concentration of the more hydrophilic monomer decreases the rate of polymerization and the calculated concentration of radicals per particle, as well as the number average molecular weight. Therefore, the authors conclude that the hydrophilic acrylonitrile monomer regulates the concentration of radicals in the particles, and the polymerization kinetics, by controlling the entry and exit events. Pokhriyal and Devi have also studied the microemulsion

Polymer Nanoparticles by RAFT Microemulsion Polymerization

139

copolymerization of butyl acrylate and acrylonitrile and concluded that the concentration of the hydrophilic monomer determines the concentration of radicals in the particles [40]. 5.2.3 Multiple Addition and Semicontinuous Microemulsion Polymerization A signiicant disadvantage to the commercial implementation of microemulsion polymerization is the high ratio of surfactant to monomer required to form the initial, thermodynamically stable microemulsions. Two solutions to this problem have been proposed: multiple-addition and semicontinuous microemulsion polymerization. However, both of these methods suffer from complications arising from the partitioning of monomer between the micelles and polymer particles, which is addressed in this section. A more complete review of these methods is provided by Xu and Gan [43]. Hermanson and Kaler have investigated the multiple addition microemulsion polymerization of n-hexyl methacrylate [44]. The rate of polymerization decreased with each subsequent addition of monomer. Therefore, either the number of propagating radicals must decrease or the concentration of monomer at the locus of polymerization must decrease. The multiple additions of monomer did not result in signiicant growth of the polymer particles, which indicates that new particles were initiated with each addition of monomer. In addition, examination of the polymerization kinetics conirmed that none of the radicals initiated prior to the addition of monomer contributed to the polymerization. Therefore, the additional monomer swelled the empty micelles and the corona of the poly mer particles instead of the core of the polymer particles, and the concentration of radicals decreased with each addition as a result of the decrease in initiator decomposition rate with time. These results are consistent with the small-angle neutron-scattering studies of n-hexyl methacrylate microemulsion polymerizations that demonstrated partitioning of the monomer between the micelles and the polymer particles throughout the polymerization [35]. Ramirez et al. showed that monomer accumulates in the polymer particle when butyl acrylate is added to a microemulsion polymerization after the initial microemulsion polymerization reaches 95% conversion [45]. The instantaneous conversion decreases to 60% upon the introduction of neat monomer and, after the monomer feed is complete, the instantaneous conversion returns to 90–95%. The particle size increases throughout the course of monomer feeding and during the post-addition period, while the number density of particles decreases slightly during the post-addition period. The increase in the number of particles during the feed results from the competition between micelles and polymer particles for the monomer. A fraction of the monomer partitions into the micelles to form new polymer particles, while the remaining monomer swells the original latex particles. Most of the

140

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

increase in solids content results from the formation of new particles rather than the growth of the original particles. Xu et al. investigated semicontinuous microemulsion polymerization both by dropwise addition and by feeding monomer via a hollow tube [46]. In both polymerization methods, the feed of monomer must be slower than the rate of consumption of monomer. If the feed of monomer is greater than the rate of consumption of monomer, then the micelles swell with monomer and facilitate nucleation of new particles. In this case, the particle size remains approximately constant and the molecular weight of the polymers does not increase.

5.3 Reversible Addition-Fragmentation Chain Transfer Polymerization Reversible addition-fragmentation chain transfer (RAFT) is a robust method for controlling the free radical polymerization of a wide range of monomers in both homogeneous [16,47–51] and heterogeneous [24,52–55] polymerizations. The original RAFT reaction mechanism proposed by Chiefari et al. [16] augments the standard free radical polymerization reactions of initiation, propagation, and termination with two transfer reactions: activation of the chain transfer agent and equilibrium between active and dormant polymer chains (Figure 5.2). The chain transfer agent is activated by reaction with a polymeric radical to form an intermediate RAFT radical that fragments to produce a dormant polymer chain and a radical R group. The radical R group then reacts with monomer to initiate a new propagating polymer chain. The dormant and propagating polymers react to form an intermediate macroRAFT radical that fragments to release either polymer for further propagation. The transfer of radical activity between propagating and dormant polymers is the key to controlling the polymerization because termination reactions are nearly eliminated and all of the polymers experience the same reaction conditions. One of the primary advantages of RAFT is the signiicant range of functionalities that can be incorporated into the polymer end groups through variations in the R and Z groups of the chain transfer agents. Figure 5.3 shows several examples of chain transfer agents, and the review by Moad et al. provides a more comprehensive list [17]. In an ideal RAFT polymerization the exchange of radical activity between propagating and dormant polymers is rapid relative to the rate of propagation, so the polymerization kinetics should be unaffected. However, inhibition and rate retardation are frequently observed in RAFT polymerizations. Inhibition is commonly attributed to slow fragmentation of the RAFT radical and/or slow initiation by the R group radical [47]. The use of highly stable, electrophilic R groups eliminates inhibition and allows the chain transfer agent activation reaction to be simpliied in kinetic investigations of the RAFT reaction

141

Polymer Nanoparticles by RAFT Microemulsion Polymerization

Activation of the transfer agent:

P



S

+

S—R

P—S

S

P—S

+

R M



Z dormant radical polymer chain R group

Z RAFT radical

Z RAFT agent

propagating polymer chain

S—R



Simpliied activation of the transfer agent: P



+

propagating polymer chain

S

S—R

ktr

Z RAFT agent

P—S

S

+

Z dormant polymer chain

R M



radical R group

Equilibrium between active and dormant chains:

Pm



+

S

S—Pn

Z propagating dormant polymer chain polymer chain

ka kf

Pm—S



S—Pn

Z macroRAFT radical

kf ka

Pm—S

S +

Pn



Z propagating dormant polymer chain polymer chain

FIGURE 5.2 Proposed reversible addition-fragmentation chain transfer mechanism. (Reprinted from J. Chiefari, Y. K. Chong, F. Ercole, J. Krstina, J. Jeffery, T. P. T. Le, R. T. A. Mayadunne, G. F. Meijs, C. L. Moad, G. Moad, E. Rizzardo, and S. H. Thang. Macromolecules 31: 5559–62, 1998. With permission.)

mechanism. The rapid activation of the chain transfer agent also provides improved control over the molecular weight of the resulting polymers [50]. Proponents of the original RAFT reaction mechanism proposed by Chiefari et al. attribute rate retardation to slow fragmentation of the macroRAFT radical, as opposed to termination. Experiments [48,50,51] and ab initio molecular orbital calculations [56,57] have shown that the structure of the Z group determines the stability of the macroRAFT radical, and therefore, the rate of fragmentation. Phenyl and O-alkyl Z groups cause the greatest degree of rate retardation in RAFT polymerizations of styrene [48,50] and alkyl acrylates [50] because of the resonance structures available for radical stabilization. RAFT polymerizations of vinyl acetate have shown that the stability of the macroRAFT radical caused by the lone electron pair on an oxygen atom can be reduced by replacing the alkyl chains with more eficient electronwithdrawing groups [51]. Moad et al. have compiled the available experimental data for RAFT polymerizations and developed the guidelines shown in Figure 5.4 to assist with the selection of R and Z groups that will provide good control [17].

142

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

S

S

S

S

S

C12H25S

N

O

CO2H

S

S

S OR

CN

O

S S

O

S

S OCH3

O

S

S

S

S

RS S

CO2H

S O

OC2H5

S

S S

F3C

O

S

S

S

S

S

S

S

N+ – Br

CO2H

S

S S

S

S

O

S

Si OCH2

O

CO2H

HO2C

FIGURE 5.3 Examples of RAFT chain transfer agents. (Reprinted from G. Moad, E. Rizzardo, and S. H. Thang, Australian Journal of Chemistry 59: 669–92, 2006. With permission.)

Ã

=

00$

5

Ã

1

à 1

Ã&+

Ã6

+2

2

2

Ã

2

2

Ã

1

Ã2

Ã2

9$

60$$0$1 Ã

Ã

Ã1

1

Ã

Ã

00$ 9$193

60$$0$1

FIGURE 5.4 Guidelines for selection of RAFT agents for various poly merizations. For Z, addition rates decrease and fragmentation rates increase from left to right. For R, fragmentation rates decrease from left to right. The dashed line indicates partial control (i.e., control of molecular weight but poor polydispersity or substantial retardation in the case of VAc or NVP). Monomer abbreviations: Methyl methacrylate (MMA), vinyl acetate (VA), styrene (S), methyl acrylate (MA), acrylamide (AM), acrylonitrile (AN), n-vinyl pyrrolidone (NVP). (Reprinted from G. Moad, E. Rizzardo, and S. H. Thang, Australian Journal of Chemistry 59: 669–92, 2006. With permission.)

143

Polymer Nanoparticles by RAFT Microemulsion Polymerization

5.4 Reversible Addition-Fragmentation Chain Transfer Microemulsion Polymerization The proposed mechanism of RAFT microemulsion polymerization is shown in Figure 5.5. Monomer (M) and chain transfer agent (XR) in excess of their respective solubility thresholds partition into micelles (Figure 5.5(1.)). Polymerization commences when a water-soluble initiator (I) decomposes and reacts with the monomer solubilized in the aqueous domain to form a propagating polymer (Figure 5.5(2.)). Upon reaching a critical degree of polymerization, the propagating radical is no longer soluble in the aqueous domain and enters a monomer-swollen micelle, thus initiating a polymer particle. Propagation continues in the polymer particle with monomer and chain transfer agent diffusing from surrounding uninitiated micelles to the locus of polymerization (Figure 5.5(3.)). The ratio of the number of chain 1. Monomer and RAFT Agent Swollen Micelles M XR

2. Initiation M XR

M XR M XR

3. Propagation • Pn–1

M XR M XR

M

P•

M

I• 4. Reaction of Propagating Chain with RAFT Agent XR Pn• M M XR

M

5. Radical Initiation and Propagation R• XP

PnX•R

M

XPn + R•

P n• M

M XR

6. Equilibrium Between Active and Dormant Chains • XPn + Pm M

P•

M

M M XR

M

PmX•Pn XPm + Pn•

M XR

FIGURE 5.5 Proposed reversible addition-fragmentation chain transfer microemulsion poly merization mechanism. (1.) Initial microemulsion with monomer (M) and chain transfer agent (XR) partitioned into the micelles. (2.) Initiation and propagation in the aqueous domain followed by entry of the propagating poly mer (P•) into an uninitiated micelle. (3.) Propagation in a surfactant stabilized poly mer particle with monomer diffusing from surrounding micelles to the locus of poly merization. (4.) Activation of the chain transfer agent to form a dormant poly mer (XP) and a radical R group. (5.) Initiation of a new propagating poly mer in the same poly mer particle. (6.) Equilibrium between active and dormant polymers.

144

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

transfer agents (XR) to the number of micelles is critical in setting the degree of polymerization that occurs in the monomer-rich particle before the propagating polymer (Pn•) reacts with a chain transfer agent (XR) to form a RAFT radical (PnX•R). Subsequently, the RAFT radical cleaves to give a dormant polymer chain (XPn) and a new radical (R•) (Figure 5.5[4.]). The new R• radical reacts with monomer to initiate a new poly mer (P•) within the same particle (Figure 5.5(5.)). Once an active poly mer (P•) and one or more dormant polymers (XPn) are present within one particle, an equilibrium is formed in which the polymers alternate between the dormant and active states. The inal latex dispersion consists of empty micelles and surfactant-stabilized polymer particles in which the number of polymers per particle depends on the concentration of chain transfer agent (Figure 5.5[6.]). Table 5.1 summarizes the RAFT microemulsion polymerizations that have been performed to date. The results of these polymerizations have identiied several key parameters for achieving controlled polymerization, including the chain transfer agent molecule per micelle ratio, and the aqueous solubility of the monomer and the chain transfer agent. Kinetic models of both uncontrolled and RAFT microemulsion polymerization have proven useful in understanding the impact of these key parameters. Therefore, the kinetic models will be introduced prior to discussing the effects of these parameters on the molecular weight, polydispersity, and particle size. 5.4.1 Microemulsion Polymerization Kinetics An understanding of RAFT microemulsion polymerization kinetics and the development of a model are critical for gaining information on the impact of several key parameters. Hermanson and Kaler have developed a kinetic model to describe the rate of RAFT microemulsion polymerizations and to track the growth of the number average molecular weight as a function of conversion. However, this model requires the iterative solution of the rate equations for each species involved in the polymerization, which is computationally intensive and does not allow for an investigation of the effects of critical parameters. Alternatively, the Morgan model for uncontrolled microemulsion polymerization kinetics provides a solid framework for the development of a kinetic model of RAFT microemulsion polymerization that requires at most the solution of four coupled differential equations [25, 59]. This model allows the effects of the chain transfer agent per micelle ratio, and the monomer and chain transfer agent aqueous solubilities to be investigated. The Morgan model is described next, and then the development of the RAFT microemulsion polymerization model and the results of this model, both with and without biradical termination effects, are summarized.

145

Polymer Nanoparticles by RAFT Microemulsion Polymerization

TABLE 5.1 Summary of RAFT Microemulsion Polymerizations Performed to Date

Monomer (Solubility)

RAFT Agent (Solubility)

n-hexyl methacrylate [24] (0.4 mM)

2-cyanoprop-2-yl dithiobenzoate (0.12 mM)

n-butyl acrylate [58] (10.9 mM)

methyl-2-(Oethylxanthyl) propionate (1.0 mM)

methyl-2-(Ododecylxanthyl) propionate (1*10



45.0

47

5.06

21.0

34 3.85 28 1.80 23 1.43 17 1.24 12 1.20 8.7 1.25 >1*104 — 29.1 1.16 18.0 1.38 10.1 1.43 5.3 1.38 3.6 1.37 2.8 1.25 multimodal multimodal multimodal multimodal multimodal >1*104 — 23.6 1.77 23.7 1.60 13.4 1.73 6.0 1.88 4.7 1.66 — —

18.5 20.5 20.7 22.6 29.4 30.0 39.8 17.8 18.8 19.5 21.1 24.6 30.5 18.1 20.2 18.6 18.0 19.2 45.0 6.7 7.6 9.2 7.0 11.0 12.5

0.35

75

0.50 0.75 1.12 1.50 2.25 3.00 0 0.3 0.6 1.2 2.4 3.7 4.9 0.3 0.6 1.2 2.4 4.9 0 0.3 0.6 1.2 2.4 3.7 4.9

53 35 24 18 12 8.8 — 39.8 20.4 10.7 5.5 3.5 2.6 — — — — — — 35.8 19.4 9.6 5.1 3.3 —

4

5.4.1.1 Uncontrolled Microemulsion Polymerization Kinetics The Morgan model for the rate of microemulsion polymerization provides a single rate equation that is able to predict the poly merization kinetics a priori [32,59]. Valuable information about the impact of monomer partitioning and biradical termination, which are directly related to monomer solubility and microemulsion composition, can be obtained from the comparison of microemulsion polymerization kinetics with the Morgan model predictions [32].

146

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

The fundamental rate equation for the rate of fractional monomer conversion (df/dt) in a microemulsion polymerization is irst order in both the concentration of monomer and propagating radicals, part * df k p Cmon N = dt Mo

(5.3)

part where kp is the propagation constant, Cmon is the concentration of monomer at the locus of polymerization, N * is the concentration of propagating radicals, and Mo is the initial concentration of monomer in the microemulsion. Assuming negligible biradical termination, as a result of the low probability of a radical entering a particle as opposed to a micelle, N * increases linearly throughout the conversion as

N * = 2k d γ [I]t

(5.4)

where each initiator (at concentration [I]) is assumed to decompose into two radicals, kd is the dissociation rate constant, γ is the eficiency, and t is the reaction time. Equation (5.4) is often further simpliied by noting that [I] is approximately constant for initiators with a long half-life relative to the total polymerization time, so this term can be replaced by the initial initiator concentration (Io). part The concentration of monomer at the locus of polymerization (Cmon ) depends on the partitioning of monomer between the micelles and the polymer particles. Assuming that monomer transport is more rapid than monomer consumption, equilibrium thermodynamics determine the partitioning of monomer [35]. Therefore, the relative solubility of the polymer in the monomer and the monomer in the surfactant tails determines the locus of polymerization. If the polymer is suficiently soluble in the monomer, then the monomer swells the core of the polymer particles. Guo et al. have simulated the kinetics of styrene microemulsion polymerizations and shown that in this case the monomer is depleted from the core of the micelles before the polymerization reaches 5% conversion [60]. If the monomer thermodynamically favors solubilizing the surfactant tails as opposed to the polymer, then the monomer remains partitioned between the micelles and the corona of the polymer particles throughout the polymerization. In this case, the polymerization occurs in the corona of the polymer particles rather than the core. part Although Cmon depends on the thermodynamics of monomer partitioning, the concentration proile is independent of the locus of poly merization. part In both cases, Cmon decreases linearly as a function of conversion (f ):

(

part part Cmon = Cmon ,o 1 − f

)

(5.5)

Polymer Nanoparticles by RAFT Microemulsion Polymerization

147

part Substituting the expressions for N* and Cmon into Equation (5.3) and integrating results in the Morgan model for conversion as a function of time, viz.  k C part k γ I  f = 1 − exp − p mon ,o d o t 2  = 1 − exp −At 2 (5.6) Mo  

(

)

where all of the rate constants and concentrations are grouped into the single parameter A. A notable property of the Morgan model for the rate of microemulsion polymerization is that it predicts a rate maximum at a conversion equal to 1 – e–0.5 = 0.39, independent of the rate constants or initial composition of the microemulsion. Several monomers demonstrate rate maxima at conversions less than the 39% conversion predicted by the Morgan model, most notably styrene [32,61–66] and vinyl acetate [67–71]. deVries et al. have studied the deviations of the rate maximum from the Morgan model and predicted rate maximum in terms of nonlinear monomer partitioning, biradical termination, and diffusion limitations to propagation [32]. Both nonlinear monomer partitioning and biradical termination can be related to the monomer solubility; however, the deviations of low-water-solubility styrene and high-water-solubility vinyl acetate from the Morgan model indicate that another parameter is also important. The rate of transfer of radical activity relative to the rate of propagation is also an important parameter for minimizing biradical termination reactions. When biradical termination is not negligible, the deVries et al. model describes N * as  pterm  pterm    1 −  p p prop  dN * prop  part *  − 2 k mono = 2 γkd I o  Cmon (5.7) ,o 1 − f N tr   1 + pterm  1 + pterm  dt     p prop  p prop   

(

)

where pterm N* = part p prop τ res k p Cmon ,o 1 − f N mic

(

)

(5.8)

ktrmono is the rate constant for radical transfer from the polymer to the monomer, τres is the residence time of the radical in a micelle, and Nmic is the concentration of micelles in the microemulsion [32]. 5.4.1.2 RAFT Microemulsion Polymerization Kinetics [72] In an uncontrolled microemulsion polymerization, biradical termination is unlikely, so all of the radicals that enter a micelle to form a polymer particle

148

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

propagate throughout the entire polymerization. A RAFT microemulsion polymerization is fundamentally different because at any given time a fraction of the radicals exist as stable macroRAFT radicals that do not propagate. Therefore, a term must be introduced to the fundamental microemulsion polymerization rate equation [Equation (5.3)] to describe the fraction of active radicals (xact) in a RAFT microemulsion polymerization. xact is expected to depend on the polymerization time (t) and the concentration of the chain transfer agent ([XR]). Introducing xact to the fundamental rate equation with linear monomer partitioning gives

(

)

part * df k p Cmon ,o 1 − f x act N = dt Mo

(5.9)

In the absence of biradical termination, N * increases linearly as a function of time according to Equation (5.4). However, if biradical termination is not negligible, then the deVries et al. model for biradical termination [Equation (5.7)] must be adjusted to consider transfer and termination reactions involving only active radical species. Also, in addition to termination reactions caused by the exit of monomer radicals, the radical R group may exit the particle and cause termination. xact depends on the concentrations of all four different radical species that are present in RAFT microemulsion polymerizations, shown in Figure 5.5. Initiator (I•), R group (R•), and poly mer (P•) radicals are active radical species that react with monomer to form polymer and thus contribute to the rate of conversion. The macroRAFT radical (PX•P) is inactive and does not contribute to the rate of conversion. Therefore, the fraction of active radicals in a RAFT microemulsion polymerization is x act =

[I•] + [R•] + [P•] [I•] + [R•] + [P•] + [PX•P]

(5.10)

In principle, each time-dependent radical concentration in Equation (5.10) must be calculated by solving the coupled rate equations presented in the Hermanson model [73]. However, a simpler expression for xact can be obtained by modeling the kinetics in a single particle and calculating the fraction of time that the particle is in the active state as opposed to the dormant state. In this case, x act =

t act t act + t dorm

(5.11)

where tact is the characteristic time that a particle is active and tdorm is the characteristic time that a particle remains dormant. The conversion from

149

Polymer Nanoparticles by RAFT Microemulsion Polymerization

Equation (5.10) to (5.11) inherently assumes that all of the particles in the polymerizing microemulsion are identical, but the particles in a RAFT microemulsion polymerization experience different polymerization kinetics because of the distribution of the chain transfer agent between the particles [25]. Therefore, Equation (5.11) must be averaged over the particles in the polymerizing microemulsion to give x act =

t act

(5.12)

t act + t dorm

where ⟨tact ⟩ is the average time that a particle is in that active state and ⟨tdorm ⟩ is the average time that a particle remains in the dormant state. Modeling the kinetics in a single particle as opposed to in the entire microemulsion greatly simpliies the rate equations because each particle can contain only one radical, either P• in an active particle or PX•P in a dormant particle. The probability of a second radical entering a poly mer particle as opposed to a micelle is typically negligible in a microemulsion poly merization, as discussed previously, because the concentration of micelles is approximately 1000 times greater than the concentration of polymer particles. Additionally, the small particle size means that if a second radical enters a polymer particle, then rapid contact between the radicals occurs, and the radicals immediately terminate. The expressions for tact and tdorm in a single particle are determined from the rate equation for the concentration of propagating polymer molecules per particle [P•]part based on the mechanism shown in Figure 5.5. Integrating the rate equation for [P•]part for the conversion from a dormant particle ([P•]part = 0 and [PX•P]part = 1) to an active particle ([P•]part = 1 and [PX•P]part = 0), and from an active particle to a dormant particle, gives the expressions for tdorm and tact, respectively. In obtaining a closed expression for tact, it is necessary to assume that the concentration of dormant polymer molecules in the particle [XP]part and the volume of the polymer particle Vpart are constant over the time that a particle is active. These assumptions are supported by small-angle neutron-scattering experiments [26] and measured polymerization kinetics [25]. The inal closed expression for ⟨xact⟩ is then

()

x act t =

t act t act + t dorm

1

≈ 1+

ka

()

N A Vpart t k f

()

XP   part t

(5.13)

where ⟨[XP]part(t)⟩ and ⟨Vpart(t)⟩ can be approximated from the bulk RAFT microemulsion polymerization kinetics [72].

150

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 5.2 Parameters for Modeling the Rate of Conversion of Butyl Acrylate in RAFT Microemulsion Polymerizations with the Chain Transfer Agent MOEP Rate Constants

Microemulsion Conditions

Monomer and Surfactant Properties

kd72 kpa

1.65 × 10–5 s–1 1.66 × 103 M–1s–1

Mo Io

0.389 M 0.003 M

MWmono ρmono

128.17 g/mol 890 g/L

ka73

4.00 × 105 M–1s–1

C(part) mon,o

2.9 M

τres

1.2 × 10–6 s

ktr73

4.00 × 105 M–1s–1

Nmic

3.2*10–3 M

ltail

1.67 nm

γ

1.00

a

From itting the uncontrolled butyl acrylate microemulsion polymerization data to the Morgan model.

5.4.1.3 Model Results The dependence of the RAFT microemulsion polymerization rate on both the rate of macroRAFT radical fragmentation and the rate of diffusion of the chain transfer agent to the locus of polymerization can be investigated using the model described in the previous section. The reported values of kf in the literature range from 10 –2 s–1 when rate retardation is assumed to arise solely from slow fragmentation of the macroRAFT radical, [74] to 106 s–1 when rate retardation is assumed to arise solely from termination of the macroRAFT radical [75]. The characteristic diffusion time of a chain transfer agent from a micelle to a polymer particle ranges from approximately 10 –4 s to 10 s, while the characteristic activation time is approximately 10 –4 s. Therefore, the rate of chain transfer agent activation is likely diffusion limited rather than reaction limited. The rate constants, microemulsion characteristics, and monomer and surfactant properties used to calculate the rate of RAFT microemulsion polymerization for butyl acrylate with the chain transfer agent methyl 2-(O-ethylxanthyl) propionate (MOEP) are summarized in Table 5.2. The kinetic trends predicted by the model using these reaction parameters are irst examined in the absence of biradical termination. However, biradical termination is not negligible in butyl acrylate microemulsion polymerizations, so the effects of biradical termination are subsequently considered [72]. 5.4.1.4 Predicted RAFT Microemulsion Polymerization Kinetics with Negligible Biradical Termination The rate of RAFT microemulsion polymerization depends on both the rate of macroRAFT radical fragmentation and the rate of diffusion of the chain transfer agent to the locus of polymerization. Decreasing the rate of macroRAFT radical fragmentation increases tdorm and, therefore, is expected

151

Polymer Nanoparticles by RAFT Microemulsion Polymerization

to decrease the rate of polymerization, as shown by Equations (5.9) and (5.13). In fact, the model shows that decreasing kf from 106 s–1 to 10 –2 s–1 decreases the rate of polymerization by two orders of magnitude when the activation of the chain transfer agent is reaction limited. For diffusion-limited activation of the chain transfer agent, the polymerization rate decreases by only one order of magnitude when kf decreases from 106 s–1 to 10 –2 s–1. Diffusion-limited activation of the chain transfer agent results in a lower value of ⟨[XP]part⟩ relative to polymerizations with reaction-limited chain transfer agent activation. Therefore, ⟨xact⟩ is greater for polymerizations with diffusion-limited activation than reaction-limited activation [Equation (5.13)]. For RAFT microemulsion polymerizations with rapid macroRAFT radical fragmentation and diffusion-limited activation, the rate of polymerization is identical to the rate of uncontrolled microemulsion polymerization, as anticipated. The rate of chain transfer agent diffusion to the locus of poly merization has a signiicant effect on the location of the rate maximum when fragmentation of the macroRAFT radical is slow. The Morgan model predicts that the maximum rate of an uncontrolled microemulsion polymerization occurs at 39% conversion irrespective of the rate constants and initial microemulsion composition [59]. The model for RAFT microemulsion poly merization presented in the previous section predicts that the maximum rate of RAFT microemulsion polymerization occurs at conversions greater than 39% when the activation of the chain transfer agent is reaction limited and fragmentation of the macroRAFT radical is slow (Figure 5.6a). This shift of the rate maximum corresponds to the trend observed by Liu et al. for RAFT microemulsion polymerizations of hexyl methacrylate with the chain transfer 1.5

×10–4

1 Increasing MOEP/Micelle

(b)

0.8

1

Rate (s–1)

Rate (s–1)

(a)

×10–3 Increasing MOEP/Micelle

0.5

0.6 0.4 0.2

0

0

0.2

0.4

0.6

Conversion

0.8

1

0

0

0.2

0.4 0.6 Conversion

0.8

1

FIGURE 5.6 Predicted rate of conversion as a function of conversion of butyl acrylate and increasing MOEP:micelle ratios for slow fragmentation (k f = 10 –2 s–1) and (a) reaction-limited activation, (b) diffusion-limited activation. In both panels MOEP:micelle = 0.1, 0.3, 0.6, 1.2, 2.4, 3.7, and 4.9. Note that the scale of the y-axis varies. (Redrawn from J. O’Donnell and E.W. Kaler, Journal of Polymer Science Part A: Polymer Chemistry, 48(3): 604-613, 2010.)

152

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

agent 2-cyanoprop-2-yl dithiobenzoate (CPDB) at CPDB:initiator ratios less than 3.0 [24]. Liu et al. suggest that the shift of the rate maximum to conversions greater than 39% is caused by the complete consumption of the chain transfer agent early in the polymerization so that particles initiated late-experience rapid uncontrolled polymerization. Solving the RAFT microemulsion polymerization model for the concentration of the chain transfer agent does in fact show that when the activation of the chain transfer agent is reaction–limited, the chain transfer agent is consumed before the polymerization reaches 40% conversion. When activation of the chain transfer agent is diffusion-limited, the model predicts that the location of the rate maximum occurs at conversions less than 39% conversion (Figure 5.6b), in accordance with the measured kinetic trends for butyl acrylate and ethylhexyl acrylate RAFT microemulsion polymerizations with MOEP and methyl 2-(O-dodecylxanthyl) propionate (MODP). However, the location of the rate maximum does not shift as a function of the MOEP:micelle ratio as observed experimentally for these RAFT microemulsion polymerizations [58]. In addition, the maximum rate of conversion is experimentally observed to occur at conversions as low as 14% at an MOEP:micelle ratio of 4.9, and the lowest predicted conversion for the rate maximum using this model is 30%. Therefore, biradical termination must be considered as the source of the shift in the location of the rate maximum as a function of the MOEP:micelle ratio. 5.4.1.5 Predicted RAFT Microemulsion Polymerization Kinetics with Biradical Termination Both nonlinear monomer partitioning and biradical termination are known to shift the location of the rate maximum to conversions less than the 39% conversion predicted by the Morgan model [32]. However, small-angle neutron-scattering studies have shown that the concentration of butyl acrylate at the locus of polymerization decreases nearly linearly as a function of conversion during RAFT microemulsion polymerizations with MOEP [26]. Therefore, nonlinear monomer partitioning does not cause the shift of the rate maximum and only biradical termination must be considered. The impact of biradical termination reactions on the rate of conversion is most evident when fragmentation of the macroRAFT radical is fast (kf = 106 s–1) (Figure 5.7). Fast fragmentation of the macroRAFT radical means that the concentration of active radicals remains high throughout the polymerization, which facilitates biradical termination. When fragmentation of the macroRAFT radical is slow (kf = 10 –2 s–1) the concentration of active radicals is low, so a radical is likely to encounter a dormant particle as opposed to an active particle. Therefore, biradical termination is minimized and the kinetic trends are similar to the trends predicted without the inclusion of biradical termination reactions (Figure 5.6).

153

Polymer Nanoparticles by RAFT Microemulsion Polymerization

0.02

(a)

0.018

0.03

Increasing MOEP/Micelle

0.025

0.016 0.014

Increasing MOEP/Micelle

0.02

0.012

Rate (s–1)

Rate (s–1)

(b)

0.01 0.008

0.015 0.01

0.006 0.004

0.005

0.002 0

0

0.2

0.4 0.6 Conversion

0.8

1

0

0

0.2

0.4 0.6 Conversion

0.8

1

FIGURE 5.7 Predicted rate of conversion with biradical termination reactions as a function of conversion of butyl acrylate and increasing MOEP:micelle ratios for fast fragmentation (k f = 106 s–1) and (a) reaction-limited activation, and (b) diffusion-limited activation. MOEP:micelle = 0.1, 0.3, 0.6, 1.2, 2.4, 3.7, and 4.9. Note that the scale of the y-axis varies. (Redrawn from J. O’Donnell and E. W. Kaler, Journal of Polymer Science, Part A: Polymer Chemistry, 48(3), 604–613, 2010.)

The experimental shift of the maximum rate of polymerization for butyl acrylate with MOEP is quantitatively captured by the model when biradical termination is included, kf = 1585 s–1, and the activation rate coeficient is approximately one-tenth of the activation rate constant; that is, the activation of the chain transfer agent is diffusion-limited. Comparison of model its with and without biradical termination reactions shows that slow fragmentation of the macroRAFT radical is responsible for the decrease in the rate of polymerization. In addition to directly decreasing the rate of polymerization by decreasing the concentration of active radicals, slow fragmentation also increases the likelihood of termination reactions. 5.4.2 Molecular Weight and Polydispersity In a controlled polymerization when the concentration of the chain transfer agent is much greater than the initiator concentration, as is true for all of the chain transfer agent per micelle ratios that have been investigated, the number average molecular weight (MN) is expected to increase linearly as a function of conversion, that is, MN =

MWmon Mo f XR o

(5.14)

where MWmonomer is the molecular weight of the monomer, and Mo and XRo are the initial concentrations of monomer and chain transfer agent in the microemulsion. Experimentally MN does grow linearly for the polymerizations of hexyl methacrylate with CPDB and butyl acrylate with MOEP (Figure 5.8).

154

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

40000

MN (Da)

30000

20000

10000

0 0

20

40 60 80 Percent Conversion (f*100)

100

FIGURE 5.8 Number average molecular weight of poly(n-butyl acrylate) at MOEP:micelle ratios of (●) 0.3, (□) 0.6, (▲) 1.2, (◇) 2.4, (⬢) 4.9. The lines are least-squares linear its to the data.

However, the linear its to the growth of MN do not pass through the origin, which indicates there must be an initial period of rapid uncontrolled polymerization or slow activation of the chain transfer agent. The polymerizations of butyl acrylate with MODP exhibit multimodal molecular weight populations, so the growth of MN could not be traced as a function of conversion. An initial period of rapid uncontrolled growth is an unlikely source of the deviation of the MN from the origin because all of the polymers are signiicantly smaller than would be expected for an uncontrolled microemulsion polymerization. In an uncontrolled microemulsion polymerization, the very low concentration of polymers formed by decomposition of the initiator produces polymers with MN of ~106 Da. A more likely source of this deviation is slow activation of the MOEP due to the partitioning of MOEP between micelles and polymer particles. This partitioning means that at low conversions the concentration of polymers is signiicantly less than the initial concentration of chain transfer agent in the microemulsion. Therefore, the fraction of the chain transfer that has been activated as a function of converCTA sion (x act ) should be included in Equation (5.14), so MN =

MWmon Mo f CTA x act XR o

(5.15)

Equation (5.15) can be used to approximate the fraction of the chain transfer agent that has been activated at a given conversion from the MN

Polymer Nanoparticles by RAFT Microemulsion Polymerization

155

calculated from the least-squares linear its to the molecular weight data (Figure 5.8). For example, at 5% conversion of butyl acrylate the concentration of polymers calculated from the least-squares linear its to MN for each of the MOEP:micelle ratios corresponds to only 7–10% of the initial MOEP concentration. As the polymerization proceeds, the MOEP diffuses from the uninitiated micelles into the polymer particles, eventually resulting in activation of the entire chain transfer agent, a constant concentration of polymers, and linear growth of MN as a function of conversion. 5.4.2.1 Effect of Chain Transfer Agent per Micelle Ratio The deviation of MN from the predicted values that occurs early on in the RAFT microemulsion polymerizations of butyl acrylate decreases as the MOEP:micelle ratio increases, as anticipated. Increasing the chain transfer agent-to-micelle ratio reduces the probability of a propagating oligomer entering a micelle that does not contain a chain transfer agent, and therefore reduces the extent of uncontrolled polymerization that can occur in the polymer particles before MOEP diffuses to the locus of polymerization. However, increasing the MOEP:micelle ratio even as high as 4.9 does not completely eliminate uncontrolled polymerization because of the way that the MOEP is distributed among the micelles. The distribution of MOEP into the micelles may be approximated as a Poisson distribution about the mean MOEP:micelle ratio, and the Poisson distribution shows that 1% of the micelles do not contain any MOEP when the mean MOEP:micelle ratio is 5. The agreement between the measured and predicted inal MN [Figure 5.9, Equation (5.14)] shows that the concentration of chains does in fact correspond to the initial concentration of MOEP in the microemulsion, which indicates that all of the MOEP is activated during the polymerization. Because the number of micelles is ~1000 times greater than the number of polymer particles, activation of all of the MOEP signiies that the MOEP is diffusing into the polymer particles. The polydispersity (MW/MN) of the polymers (Figure 5.9 inset) increases as the average MOEP:micelle ratio increases because the diffusion of MOEP into propagating particles results in a distribution of the MOEP:particle ratio. The MOEP:particle ratio distribution means that all of the propagating polymer chains are not experiencing the same polymerization kinetics. 5.4.2.2 Effect of Monomer Solubility The aqueous solubility of the monomer is known to affect uncontrolled microemulsion polymerization kinetics, as discussed in Section 5.2.1. In RAFT microemulsion polymerizations, the aqueous solubility of the monomer may also affect the control of the polymerization. The probability of a chain transfer agent and a propagating polymer radical reacting in the aqueous domain is negligible. Therefore, zcrit is the minimum degree of uncontrolled

156

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

5

2.0 1.8

MW/MN

MN/104 (Da)

4

3

1.6 1.4 1.2

2 1.0 0

1

1

2 3 MOEP/Micelle

4

5

0 0

1

2 3 MOEP/Micelle

4

5

FIGURE 5.9 MN and MW/MN (inset) of poly(butyl acrylate) as a function of the MOEP:micelle ratio. (—) predicted MN from Equation (5.14).

polymerization that a propagating oligomer reaches in a RAFT microemulsion polymerization. The impact of monomer solubility has been examined by substituting butyl acrylate with 2-ethylhexyl acrylate (EHA). The τprop/τres ratios for butyl acrylate and 2-ethylhexyl acrylate are similar (Table 5.3), so the effect of the aqueous solubility of the monomer on biradical termination is negligible and any effects of the aqueous solubility of the monomer on the control of the polymerization are elucidated. The 2-ethylhexyl acrylate polymerizations are expected to experience less uncontrolled polymerization at low conversions than the butyl acrylate polymerizations because the 2-ethylhexyl acrylate chains are segregated into surfactant stabilized polymer particles at a lower zcrit. However, the kinetics and polymer characteristics of the EHA polymerizations indicate less control than the BA polymerizations. The butyl acrylate RAFT microemulsion polymerizations with MOEP demonstrate controlled, linear growth at all chain transfer agent:micelle ratios after the initial period of uncontrolled polymerization or slow chain transfer agent activation (Figure 5.8). In contrast, the 2-ethylhexyl acrylate RAFT microemulsion polymerizations with MOEP do not demonstrate controlled growth at an MOEP:micelle ratio of 0.3 (Figure 5.10). Controlled, linear growth of the EHA polymer chains is observed at MOEP:micelle ratios of 0.6 and 1.2; and at MOEP:micelle ratios greater than 1.2, MN decreases slightly, which indicates an increase in the concentration of small, controlled poly mer chains as the polymerization proceeds. At all chain transfer agent:micelle ratios investigated, the inal MN of the butyl acrylate chains is consistent with the activation of all of the MOEP (Figure 5.9). However, the inal MN of the 2-ethylhexyl acrylate chains is

157

Polymer Nanoparticles by RAFT Microemulsion Polymerization

TABLE 5.3 Monomer and Chain Transfer Agent Aqueous Solubilities (Caq), Monomer Concentration in the Micelles (Cmic), Radical Residence Time in a MonomerSwollen Micelle (tres), and Characteristic Time for Propagation of a Monomer Radical (t prop) or Reaction of a Transfer Agent with a Propagating Polymer (tact) Caq (mM)

Cmic (mM)

τres (s)

τprop/act (s)

τprop /τres

Monomers BA EHA

10.9a 0.5a

4200c 4200c

1.2*10–6 d 2.5*10–5 d

2.5*10–4 4.0*10–3

2.1*102 1.6*102

Chain Transfer Agents MOEP MODP

1.0b 10–4 – 10–3 e

~10–4 f ~10–4 f

a b c

d e f

Material safety data sheet values. Measured by ultraviolet absorption. Calculated from the concentration of monomer in a microemulsion at the phase boundary at 45°C. Calculated from Equation (5.2) with Rmic = 3.0 nm and Daqmon = 10–9 m2/s. Assumed based on values for monomers and relative solubilities. Concentration of a propagating polymer approximated as one propagating polymer in a spherical particle with a 3-nm radius.

40000

MN (Da)

30000

20000

10000

0 0

20 40 60 Percent Conversion (f*100)

80

100

FIGURE 5.10 Number average molecular weight of poly(ethylhexyl acrylate) at MOEP:micelle ratios of (○) 0.3, (■) 0.6, (△) 1.2, (◆) 2.4, and (▽) 3.7. The lines are least-squares linear its to the data.

158

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

5

2.0

1.8

MW/MN

MN/104 (Da)

4

3

1.6

1.4

1.2

2 1.0 0

1

2

3

4

5

MOEP/Micelle

1

0 0

1

2 3 MOEP/Micelle

4

5

FIGURE 5.11 MN and MW/MN (inset) polydispersity of poly(butyl acrylate) with MOEP (▲) and poly(ethylhexyl acrylate) with MOEP (□). The lines are predicted from Equation (5.14) with conversion equal to one.

greater than anticipated (Figure 5.11), which indicates that all of the MOEP is not activated. In addition, the polydispersity (MW/MN) of the 2-ethylhexyl acrylate chains is greater than the MW/MN of the butyl acrylate chains (Figure 5.11 inset). As discussed later with respect to the latex particle size, the fact that the butyl acrylate polymerizations are better controlled than the 2-ethylhexyl acrylate polymerizations is likely caused by coalescence of the poly(butyl acrylate) particles. 2-ethylhexyl acrylate polymer chains that partition into a surfactant stabilized polymer particle that does not contain a chain transfer agent must wait until a chain transfer agent diffuses into the particle, while the chain transfer agent is rapidly transported into the poly(butyl acrylate) particles by coalescence. 5.4.2.3 Effect of Chain Transfer Agent Solubility The aqueous solubility of the chain transfer agent is expected to impact the RAFT microemulsion polymerization kinetics and the characteristics of the polymer chains. The residence time of the chain transfer agent in the micelles (τres,CTA) is approximated in the same manner as τres [Equation (5.2)]. Decreasing the aqueous solubility of the chain transfer agent increases the residence time in the micelles and slows the transport of the chain transfer agent to the locus of polymerization. When the time necessary for a chain

Polymer Nanoparticles by RAFT Microemulsion Polymerization

159

transfer agent to diffuse to the locus of polymerization exceeds the time required for activation of the chain transfer agent by a propagating polymer (τact), the activation of the chain transfer agent is diffusion limited. The time necessary for a chain transfer agent to diffuse to the locus of polymerization is approximately τres,CTA multiplied by the micelle:polymer particle ratio, and τact is a function of the transfer rate constant (ktr) and the concentration of part radicals in the polymer particle (Crad ): τ act ≈

1 part k tr Crad

(5.16)

Slow diffusion of the chain transfer agent to the locus of polymerization decreases the average and the standard deviation of the distribution of the chain transfer agent:particle ratio. Decreasing the standard deviation of the chain transfer agent:particle ratio is expected to decrease the polydispersity of the polymer chains because all of the polymer particles experience similar polymerization conditions. Decreasing the average chain transfer agent:particle ratio is expected to alleviate the rate retardation caused by slow fragmentation of the macroRAFT radical. The effect of the chain transfer agent aqueous solubility is investigated by replacing MOEP with MODP, which differs from MOEP only in the length of the alkyl chain in the Z group so that the functionality of the chain transfer agent is unaltered. Given that the ratio of micelles to polymer particles is 100–1000, the activation of both of the chain transfer agents is diffusion limited. Therefore, the average and the standard deviation of the chain transfer agent:particle ratio distribution are expected to be lower in the MODP polymerizations than the MOEP polymerizations. The RAFT microemulsion poly meri zations of BA with MODP, the low aqueous solubility chain transfer agent, demonstrate several distinct molecular weight populations (Figure 5.12). However, all of the polymers are smaller than those produced by uncontrolled poly meri zation (MN ~ 103 kDa), which indicates that the poly meri zation is controlled to some extent. At MODP:micelle ratios of 0.3 and 0.6, the gel permeation chromatography (GPC) traces show predominantly 40–60 kDa polymers with a low molecular weight tail. At an MODP:micelle ratio of 1.2, the concentration of low molecular weight polymers increases and two distinct peaks are observed in the GPC trace. The high molecular weight polymers are now 30–50 kDa and the low molecular weight polymers are less than 10 kDa. As the MODP:micelle ratio increases from 1.2 to 4.8, the ratio of the high molecular weight poly mer to low molecular weight poly mer decreases. Additionally, a large concentration of oligomers is apparent in the GPC traces. An examination of the experimentally observed molecular weight proiles as a function of the chain transfer agent per particle distribution using the kinetic model described in Section 5.4.1.2 has demonstrated that

160

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Refractive Index Response (a.u.)

4

Conversion = 90%

3

84% 73%

2

47% 37% 1

26% 9% 3%

0 0

5

10 Time (min) (a)

15

20

4

Refractive Index Response (a.u.)

Conversion = 92% 77%

3

67% 51%

2

36% 24% 1

18% 6% 2%

0 0

5

10 Time (min) (b)

15

20

Refractive Index Response (a.u.)

5 Conversion = 87%

4

79% 69%

3

59% 48%

2

36% 1

24% 11%

0 0

5

10 Time (min) (c)

15

20

FIGURE 5.12 Gel permeation chromatography traces of poly(butyl acrylate) with, from top to bottom, MODP/micelle = 0.3, 1.2, and 4.8.

Polymer Nanoparticles by RAFT Microemulsion Polymerization

161

the distinct molecular weight populations in the RAFT microemulsion poly meri zations with the low water solubility chain transfer agent are a direct result of the slow diffusion of the chain transfer agent to the locus of poly meri zation. 5.4.3 Latex Particle Size The poly(butyl acrylate) latex particles produced by RAFT microemulsion polymerization remain colloidally stable for at least three years. The latex particle size distribution depends on the distribution of MOEP in the polymer particles (Figure 5.13). The uncontrolled microemulsion polymerization of butyl acrylate produces latex particles with a very narrow distribution of diameters between 35 and 45 nm. The addition of the chain transfer agent 0.5

0.5 MOEP/Micelle = 0

0.3 0.2

0

0

10 20 30 40 50 Particle Diameter (nm)

0.5 MOEP/Micelle = 1.2

0.3 0.2 0.1

MOEP/Micelle = 4.9

0.4 Volume Fraction

0.4 Volume Fraction

0.2

0.0

10 20 30 50 40 Particle Diameter (nm)

0.5

0.0

0.3

0.1

0.1 0.0

MOEP/Micelle = 0.3

0.4 Volume Fraction

Volume Fraction

0.4

0.3 0.2 0.1

0

40 50 10 20 30 Particle Diameter (nm)

0.0

0

40 50 10 20 30 Particle Diameter (nm)

FIGURE 5.13 Average of four CONTIN its to quasi-elastic light scattering data for poly(butyl acrylate) latex particles initiated with 3.0 mM VA044. (From J. O’Donnell, 2007. Reversible addition-fragmentation chain transfer polymerizationin microemulsions. PhD diss., Chemical Engineering, University of Delaware.)

162

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

at an MOEP:micelle ratio of 0.3 drastically reduces the latex particle diameter, but the distribution is still narrow. As the MOEP/micelle ratio increases beyond 0.3, the particle size distribution broadens, and the average latex particle diameter increases. The initial decrease and subsequent increase in the average latex particle diameter as the MOEP:micelle ratio increases is directly related to the observed rate retardation. The increased poly merization time both increases the number of polymer particles formed by thermal decomposition of the initiator, resulting in smaller particles, and facilitates coalescence of the monomer-swollen polymer particles, thereby increasing the size of some of the particles. Both events contribute to the broader distribution. In contrast to the highly stable poly(butyl acrylate particles), the poly(ethylhexyl acrylate) particles are colloidally stable for only one month, and the poly(butyl acrylate) particles polymerized with MODP are stable for several months. The reason for the decrease in stability with the lowsolubility monomer and chain transfer agent is unknown. The RAFT microemulsion polymerizations of butyl acrylate with MODP and 2-ethylhexyl acrylate with MOEP also produce smaller latex particles than the corresponding uncontrolled microemulsion polymerizations (Figure 5.14). The poly(butyl acrylate) particles polymerized with MOEP and MODP increase in diameter as the chain transfer agent:micelle ratio increases. As the chain transfer agent:micelle ratio increases the polymerization time 50

Particle Diameter (nm)

40

30

20

10

0 0

1

2 3 Chain Transfer Agent/Micelle

4

5

FIGURE 5.14 Average latex particle diameter from quadratic cumulants its to four autocorrelation functions measured by quasi-elastic light scattering for poly(butyl acrylate) with MOEP (▲), poly(butyl acrylate) with MODP (●), and poly(ethylhexyl acrylate) with MOEP (□). Error bars are smaller than the size of the symbols. (From J. O’Donnell, 2007. Reversible addition-fragmentation chain transfer polymerizationin microemulsions. PhD diss., Chemical Engineering, University of Delaware.)

Polymer Nanoparticles by RAFT Microemulsion Polymerization

163

also increases, which provides an extended time where the polymer particles can coalesce with the high concentration of monomer-swollen micelles. The coalescence of the polymer particles and monomer-swollen micelles can impact the polymerization kinetics and polymer characteristics by increasing the chain transfer agent:particle ratio more rapidly than diffusion of the chain transfer agent alone. The poly(ethylhexyl acrylate) particles polymerized with MOEP do not increase in size as the chain transfer agent:micelle ratio increases. The lower water solubility of 2-ethylhexyl acrylate relative to butyl acrylate likely limits coalescence because the 2-ethylhexyl acrylate is partitioned closer to the core of the polymer particles while the butyl acrylate may partition closer to the surface.

References 1. Stoffer, J. O., and T. Bone. 1980. Polymerization in water-in-oil micro-emulsion systems 1. Journal of Polymer Science, Part A: Polymer Chemistry 18:2641–48. 2. Atik, S. S., and J. K. Thomas. 1981. Polymerized microemulsions. Journal of the American Chemical Society 103:4279–80. 3. Clark, H. A., Hoyer, M., Philbert, M. A., and R. Kopelman. 1999. Optical nanosensors for chemical analysis inside single living cells. 1. Fabrication, characterization, and methods for intracellular delivery of PEBBLE sensors. Analytical Chemistry 71:4831–36. 4. Gao, H. F., Zhao, Y. Q., Fu, S. K., Li, B., and M. Q. Li. 2002. Preparation of a novel polymeric luorescent nanoparticle. Colloid and Polymer Science 280:653–60. 5. Jang, J., and J. H. Oh. 2005. Fabrication of a highly transpartent conductive thin ilm from polypyrrole/poly(methyl methacrylate) core/shell nanospheres. Advanced Functional Materials 15:494–502. 6. Fukuda, T., Terauchi, T., Goto, A., Ohno, K., Tsujii, Y., Miyamoto, T., Kobatake, S., and B. Yamada. 1996. Mechanisms and kinetics of nitroxide-controlled free radical polymerization. Macromolecules 29:6393–98. 7. Georges, M. K., Kee, R. A., Veregin, R. P. N., Hamer, G. K., and P. M. Kazmaier. 1995. Nitroxide mediated free-radical polymerization process—autopolymerization. Journal of Physical Organic Chemistry 8:301–5. 8. Hawker, C. J., Bosman, A. W., and E. Harth. 2001. New polymer synthesis by nitroxide mediated living radical polymerizations. Chemical Reviews 101:3661–88. 9. Sciannamea, V., Jerome, R., and C. Detrembleur. 2008. In-situ nitroxide-mediated radical polymerization (NMP) processes: Their understanding and optimization. Chemical Reviews 108:1104–26. 10. Coessens, V., Pintauer, T., and K. Matyjaszewski. 2001. Functional polymers by atom transfer radical polymerization. Progress in Polymer Science 26:337–77. 11. Kamigaito, M., Ando, T., and M. Sawamoto. 2001. Metal-catalyzed living radical polymerization. Chemical Reviews 101:3689–3745.

164

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

12. Matyjaszewski, K., and J. H. Xia. 2001. Atom transfer radical polymerization. Chemical Reviews 101:2921–90. 13. Wang, J. S., and K. Matyjaszewski. 1995. Controlled living radical polymerization—atom-transfer radical polymerization in the presence of transition-metal complexes. Journal of the American Chemical Society 117:5614–15. 14. Gaynor, S. G., Wang, J. S., and K. Matyjaszewski. 1995. Controlled radical polymerization by degenerative transfer—effect of the structure of the transfer agent. Macromolecules 28:8051–56. 15. Goto, A., Ohno, K., and T. Fukuda. 1998. Mechanism and kinetics of iodidemediated polymerization of styrene. Macromolecules 31:2809–14. 16. Chiefari, J., Chong, Y. K., Ercole, F., Krstina, J., Jeffery, J., Le, T. P. T., Mayadunne, R. T. A., Meijs, G. F., Moad, C. L., Moad, G., Rizzardo, E., and S. H. Thang. 1998. Living free-radical polymerization by reversible addition-fragmentation chain transfer: The RAFT process. Macromolecules 31:5559–62. 17. Moad, G., Rizzardo, E., and S. H. Thang. 2006. Living radical polymerization by the RAFT process—A irst update. Australian Journal of Chemistry 59:669–92. 18. Prescott, S. W., Ballard, M. J., Rizzardo, E., and R. G. Gilbert. 2002. RAFT in emulsion polymerization: What makes it different? Australian Journal of Chemistry 55:415–24. 19. Ferguson, C. J., Hughes, R. J., Nguyen, D., Pham, B. T. T., Gilbert, R. G., Serelis, A. K., Such, C. H., and B. S. Hawkett. 2005. Ab initio emulsion polymerization by RAFT-controlled self-assembly. Macromolecules 38:2191–2204. 20. de Brouwer, H., Tsavalas, J. G., Schork, F. J., and M. J. Monteiro. 2000. Living radical polymerization in miniemulsion using reversible addition-fragmentation chain transfer. Macromolecules 33:9239–46. 21. Cunningham, M. F. 2008. Controlled/living radical polymerization in aqueous dispersed systems. Progress in Polymer Science 33:365–98. 22. Luo, Y. W., Wang, R., Yang, L., Yu, B., Li, B. G., and S. P. Zhu. 2006. Effect of reversible addition-fragmentation transfer (RAFT) reactions on (mini)emulsion polymerization kinetics and estimate of RAFT equilibrium constant. Macromolecules 39:1328–37. 23. Lansalot, M., Davis, T. P., and J. P. A. Heuts. 2002. RAFT miniemulsion polymerization: Inluence of the structure of the RAFT agent. Macromolecules 35:7582–91. 24. Liu, S., Hermanson, K. D., and E. W. Kaler. 2006. Reversible additionfragmentation chain transfer polymerization in microemulsion. Macromolecules 39:4345–50. 25. O’Donnell, J. 2007. Reversible addition-fragmentation chain transfer polymerization in microemulsions. PhD diss., Chemical Engineering, University of Delaware. 26. O’Donnell, J., and E. W. Kaler. 2008. Microstructure evolution and monomer partitioning in reversible addition-fragmentation chain transfer microemulsion polymerization. Macromolecules 41:6094–99. 27. Lusvardi, K. M., Schubert, K. V., and E. W. Kaler. 1996. Microemulsions: A medium for polymerization reactions. Berichte Der Bunsen-Gesellschaft-Physical Chemistry Chemical Physics 100:373–79. 28. Antonietti, M., Basten, R., and S. Lohmann. 1995. Polymerization in microemulsions—a new approach to ultraine, highly functionalized polymer dispersions. Macromolecular Chemistry and Physics 196:441–66. 29. Candau, F. 1992. Polymerization in microemulsions. In Polymerization in organized media, ed. C. M. Paleos, 215. Philadelphia: Gordon and Breach Science Publishers.

Polymer Nanoparticles by RAFT Microemulsion Polymerization

165

30. Capek, I., Juranicova, V., Barton, J., Asua, J. M., and K. Ito. 1997. Microemulsion radical polymerization of alkyl acrylates. Polymer International 43:1–7. 31. Guo, J. S., Sudol, E. D., Vanderhoff, J. W., and M. S. Elaasser. 1992. Kinetics and mechanism of styrene microemulsion polymerization. ACS Symposium Series 492:99–113. 32. de Vries, R., Co, C. C., and E. W. Kaler. 2001. Microemulsion polymerization. 2. Inluence of monomer partitioning, termination, and diffusion limitations on polymerization kinetics. Macromolecules 34:3233–44. 33. Bleger, F., Murthy, A. K., Pla, F., and E. W. Kaler. 1994. Particle nucleation during microemulsion polymerization of methyl-methacrylate. Macromolecules 27:2559–65. 34. Safran, S. A. 1991. Saddle-Splay modulus and the stability of spherical microemulsions. Physical Review A 43:2903–4. 35. Co, C. C., de Vries, R., and E. W. Kaler. 2001. Microemulsion polymerization. 1. Small-angle neutron scattering study of monomer partitioning. Macromolecules 34:3224–32. 36. Capek, I., and V. Juranicova. 1998. On the free-radical microemulsion polymerization of alkyl methacrylates. European Polymer Journal 34:783–88. 37. Gan, L. M., Lee, K. C., Chew, C. H., Hg, S. C., and L. H. Gan. 1994. Copolymerization of styrene and methyl methacrylate in ternary oil-in-water microemulsions: Monomer reactivity ratios and microstructures by 1HNMR and 13CNMR. Macromolecules 27:6335–40. 38. Bhawal, S., Sanghvi, P. G., and S. Devi. 2003. Recalculation of monomer reactivity considering the effect of monomer partitioning in microemulsion. European Polymer Journal 39:389–96. 39. Shi, Y. C., Wu, Y. S., Hao, J. C., and G. Z. Li. 2005. Kinetics of microemulsion copolymerization of styrene and acrylonitrile in the presence of cosurfactant. Colloids and Surfaces A: Physicochemical and Engineering Aspects 262:191–97. 40. Pokhriyal, N. K., and S. Devi. 2000. Effect of water solubility of monomer on reaction kinetics oil-water microemulsion copolymerization. European Polymer Journal 36:333–43. 41. Zheng, C., He, W.-D., Liu, W.-J., Li, J., and J.-F. Li. 2006. Novel one-step route for preparing amphiphilic block copolymers of styrene and N-isopropylacrylamide in a microemulsion. Macromolecular Rapid Communications 27:1229–32. 42. Capek, I., and V. Juranicova. 1996. On kinetics of microemulsion copolymerization of butyl acrylate and acrylonitrile. Journal of Polymer Science, Part A: Polymer Chemistry 34:575–85. 43. Xu, X. J., and L. M. Gan. 2005. Recent advances in the synthesis of nanoparticles of polymer latexes with high polymer-to-surfactant ratios by microemulsion polymerization. Current Opinion in Colloid & Interface Science 10:239–44. 44. Hermanson, K. D., and E. W. Kaler. 2003. Kinetics and mechanism of the multiple addition microemulsion polymerization of hexyl methacrylate. Macromolecules 36:1836–42. 45. Ramirez, A. G., Lopez, R. G., and K. Tauer. 2004. Studies on semibatch microemulsion polymerization of butyl acrylate: Inluence of the potassium peroxodisulfate concentration. Macromolecules 37:2738–47. 46. Xu, X. J., Chew, C. H., Siow, K. S., Wong, M. K., and L. M. Gan. 1999. Microemulsion polymerization of styrene for obtaining high ratios of polystyrene/surfactant. Langmuir 15:8067–71.

166

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

47. Chong, Y. K., Krstina, J., Le, T. P. T., Moad, G., Postma, A., Rizzardo, E., and S. H. Thang. 2003. Thiocarbonylthio compounds S=C(Ph)S-R in free radical polymerization with reversible addition-fragmentation chain transfer (RAFT polymerization). Role of the free-radical leaving group (R). Macromolecules 36:2256–72. 48. Chiefari, J., Mayadunne, R. T. A., Moad, C. L., Moad, G., Rizzardo, E., Postma, A., Skidmore, M. A., and S. H. Thang. 2003. Thiocarbonylthio compounds (S=C(Z)S-R) in free radical polymerization with reversible addition-fragmentation chain transfer (RAFT polymerization). Effect of the activating group Z. Macromolecules 236:2273–83. 49. Davis, T. P., Barner-Kowollik, C., Nguyen, T. L. U., Stenzel, M. H., Quinn, J. F., and P. Vana. 2003. Inluences of the structural design of RAFT agents on living radical polymerization kinetics. Advances in Controlled/Living Radical Polymerization 854:551–69. 50. Destarac, M., Taton, D., Zard, S. Z., Saleh, T., and Y. Six. 2003. On the importance of xanthate substituents in the MADIX process. Advances in Controlled/Living Radical Polymerization 854:536–50. 51. Stenzel, M. H., Cummins, L., Roberts, G. E., Davis, T. R., Vana, P., and C. BarnerKowollik. 2003. Xanthate mediated living polymerization of vinyl acetate: A systematic variation in MADIX/RAFT agent structure. Macromolecular Chemistry and Physics 204:1160–68. 52. Butte, A., Storti, G., and M. Morbidelli. 2001. Miniemulsion living free radical polymerization by RAFT. Macromolecules 34:5885–96. 53. Simms, R. W., Davis, T. P., and M. F. Cunningham. 2005. Xanthate-mediated living radical polymerization of vinyl acetate in miniemulsion. Macromolecular Rapid Communications 26:592–96. 54. Ferguson, C. J., Hughes, R. J., Pham, B. T. T., Hawkett, B. S., Gilbert, R. G., Serelis, A. K., and C. H. Such. 2002. Effective ab initio emulsion polymerization under RAFT control. Macromolecules 35:9243–45. 55. McLeary, J. B., Tonge, M. P., Roos, D. D., Sanderson, R. D., and B. Klumperman. 2004. Controlled, radical reversible addition-fragmentation chain-transfer polymerization in high-surfactant concentration ionic miniemulsions. Journal of Polymer Science, Part A: Polymer Chemistry 42:960–74. 56. Coote, M. L., and L. Radom. 2003. Ab initio evidence for slow fragmentation in RAFT polymerization. Journal of the American Chemical Society 125:1490–91. 57. Coote, M. L. 2004. Ab initio study of the addition-fragmentation equilibrium in RAFT polymerization: When is polymerization retarded? Macromolecules 37:5023–31. 58. O’Donnell, J. and E. W. Kaler. 2010. Reversible addition-fragmentation chain transfer in microemulsion polymerizations: Effect of chain transfer agent aqueous solubility, Macromolecules, 43 (4): 1730–1738. 59. Morgan, J. D., Lusvardi, K. M., and E. W. Kaler. 1997. Kinetics and mechanism of microemulsion polymerization of hexyl methacrylate. Macromolecules 30:1897–1905. 60. Guo, J. S., Sudol, E. D., Vanderhoff, J. W., and M. S. Elaasser. 1992. Particle nucleation and monomer partitioning in styrene O/W microemulsion polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 30:691–702. 61. Chern, C. S., and L. J. Wu. 2001. Microemulsion polymerization of styrene stabilized by sodium dodecyl sulfate and short-chain alcohols. Journal of Polymer Science, Part A: Polymer Chemistry 39:3199–3210.

Polymer Nanoparticles by RAFT Microemulsion Polymerization

167

62. Chern, C. S., and C. W. Liu. 2000. Effect of short-chain alcohols on the oil-inwater microemulsion polymerization of styrene. Colloid and Polymer Science 278:329–36. 63. Puig, J. E., Perezluna, V. H., Perezgonzalez, M., Macias, E. R., Rodriguez, B. E., and E. W. Kaler. 1993. Comparison of oil-soluble and water-soluble initiation of styrene polymerization in a 3-component microemulsion. Colloid and Polymer Science 271:114–23. 64. Gan, L. M., Chew, C. H., and I. Lye. 1992. Styrene polymerization in oil-inwater microemulsions: Kinetics of polymerization. Makromolekulare ChemieMacromolecular Chemistry and Physics 193:1249–60. 65. Gan, L. M., Chew, C. H., Lye, I., and T. Imae. 1991. Microemulsion polymerization of styrene. Polymer Bulletin 25:193–98. 66. Guo, J. S., Elaasser, M. S., and J. W. Vanderhoff. 1989. Microemulsion polymerization of styrene. Journal of Polymer Science, Part A: Polymer Chemistry 27:691–710. 67. Trevino, M. E., Lopez, R. G., Peralta, R. D., Becerra, F., Mendizabal, E., and J. E. Puig. 1999. Polymerization of vinyl acetate in microemulsions stabilized with a mixture of dodecyltrimethylammonium bromide and didodecyldimethylammonium bromide. Polymer Bulletin 42:411–17. 68. Donescu, D., Fusulan, L., Petcu, C., Boborodea, A.-G., and D.-S. Vasilescu. 2000. Polymerization of vinyl acetate in microemulsions with methanol. Colloid and Polymer Science 278:927–35. 69. Gomez-Cisneros, M., Trevino, M. E., Peralta, R. D., Rabelero, M., Mendizabal, E., Puig, J. E., Cesteros, C., and R. G. Lopez. 2005. Surfactant concentration effects on the microemulsion polymerization of vinyl acetate. Polymer 46:2900–7. 70. Gomez-Cisneros, M., Lopez, R. G., Peralta, R. D., Cesteros, L. C., Katime, I., Mendizabal, E., and J. Puig. 2002. Polymerization of vinyl acetate in microemulsions stabilized with dodecyltrimethylammonium bromide and cetyltrimethylammonium bromide. Polymer 43:2993–99. 71. Sosa, N., Lopez, R. G., Peralta, R. D., Katime, I., Becerra, F., Mendizabal, E., and J. Puig. 1999. Electrostatic effects on the polymerization of vinyl acetate in three-component anionic microemulsions. Macromolecular Chemistry and Physics 200:2416–20. 72. O’Donnell, J., and E. W. Kaler. 2010. Kinetic model of reversible addition-fragmentation chain transfer polymerization in microemulsions. Journal of Polymer Science, Part A: Polymer Chemistry, 48(3):604–613. 73. Hermanson, K. D., Liu, S., and E. W. Kaler. 2006. Kinetic modeling of controlled living microemulsion polymerizations that use reversible additionfragmentation chain transfer. Journal of Polymer Science, Part A: Polymer Chemistry 44:6055–70. 74. Barner-Kowollik, C., Quinn, J. F., Morsley, D. R., and T. P. Davis. 2001. Modeling the reversible addition-fragmentation chain transfer process in cumyl dithiobenzoate-mediated styrene homopolymerizations: Assessing rate coeficients for the addition-fragmentation equilibrium. Journal of Polymer Science, Part A: Polymer Chemistry 39:1353–65. 75. Wang, A. R., and S. P. Zhu. 2003. Modeling the reversible addition-fragmentation transfer polymerization process. Journal of Polymer Science, Part A: Polymer Chemistry 41:1553–66.

6 pH-Responsive Polymer Nanoparticles Jonathan V. M. Weaver CONTENTS 6.1 Introduction ................................................................................................ 169 6.2 pH-Responsive Polymer Micelle Particles.............................................. 172 6.3 pH-Responsive Cross-Linked Micelle Particles .................................... 177 6.4 pH-Responsive Microgel Particles .......................................................... 179 6.5 pH-Responsive Branched Copolymer Particles..................................... 182 6.6 Polymer Nanoparticles with pH-Responsive Surfaces......................... 183 6.7 Applications of pH-Responsive Polymer Particles ............................... 185 6.8 Outlook ........................................................................................................ 189 References............................................................................................................. 190

6.1 Introduction The ield of pH-responsive polymer nanoparticles has stimulated growing academic and commercial interest in recent years due to signiicant developments in the synthetic methodologies used to make these materials and their broadening potential applicability in various applications. In this chapter, we highlight some of the most common pH-responsive polymeric particles currently accessible and highlight how advances in control over these systems has presented various new opportunities for this important class of material. The rapid development of pH-responsive polymer nanoparticles (PRPNs) over the past 10 years has been facilitated by the broad range of synthetic methodologies available to make these materials, such as emulsion and dispersion heterogeneous polymerizations and controlled and conventional radical and ionic homogeneous polymerizations. A variety of common PRPN structures are shown schematically in Figure 6.1. We restrict discussion to nanoparticles in water with submicron dimensions here although many of the principles discussed can be translated to responsive macroscopic materials at larger, micron-sized, length scales. In the course of this chapter, we will focus on the most common classes of PRPNs with leeting mention of other less common, and in some instances less well-understood, responsive nanoparticle systems. This necessary restriction, however, emphasizes the 169

170

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

+\GURSKLOLF K\GUDWHGIRUP

+\GURSKRELF GHK\GUDWHGIRUP pH change

pH-responsive block copolymer micelles: Section 6.2

pH change

pH-responsive cross-linked micelles: Section 6.3

pH change

pH-responsive micro(nano)gels: Section 6.4

pH change

pH-responsive branched copolymers: Section 6.5

FIGURE 6.1 Schematic of PRPN structures shown as idealized two-dimensional images. White components represent permanently hydrophilic domains and gray components represent pH-responsive domains.

magnitude of the rate of development of new synthetic capabilities to produce ever more elaborate organic nanoparticles. This is encouraging for the ield and, of course, augurs well for stimulating further interest and broadening the potential applicability of PRPNs over the coming decade or so. Synthetic PRPNs are typically either weakly acidic or weakly basic polyelectrolytes. In aqueous solution and in their ionic form these polymers are hydrated and swollen, but dehydrate and become deswollen and compact in their neutral form. This swelling and deswelling process induces a size and volume change (Figure 6.1) and is usually reversible up to the point at

pH-Responsive Polymer Nanoparticles

171

which the buildup of background salt affects aqueous solubility. The reversible swelling process can be conveniently monitored using various common laboratory techniques such as nuclear magnetic resonance (NMR) to monitor the extent of hydration, light scattering to observe in situ size variations, and rheometry to measure changes in the low of these materials. PRPNs are typically multicomponent in nature—that is, there are multiple domains within the structure. In the simplest sense PRPNs adopt “core-shell” morphologies where the core material is pH-responsive, and this domain is stabilized in aqueous solution by a permanently hydrophilic, nonresponsive, domain. The hydrophilic domain provides either steric or electrostatic stabilization when the responsive polymer is in its hydrophobic form similarly to the stabilization of colloidal particles. A range of highly complex morphologies are possible, some of which are shown in Figure 6.1. The precise morphologies are deined by the PRPN type and synthetic method and will be discussed throughout the chapter in the different sections. A key property of pH-responsive nanoparticles is that their response to solution pH occurs in a distinctly nonlinear fashion. Thus, PRPNs change between their swollen and deswollen states over a relatively small pH-range. This transitional window is deined by the apparent pKa (or pKb) of the acidic or basic residues of the polyelectrolyte and is governed by Equation (6.1), which is described for poly(methacrylic acid) (PMAA). Above the pKa, PMAA is in its deprotonated anionic form, and below the pKa it exists in its protonated neutral form. − +  CH 2C(CH 3 )COOH + H 2O   CH 2C(CH 3 )COO + H 3O

Ka =

[CH 2C(CH 3 )COO − ] + [H 3O+ ] [CH 2C(CH 3 )COOH]

(6.1)

The pKa of the monomer unit in the PRNP deines the transition between the swollen and deswollen states. The apparent pKa of PRPNs is also strongly inluenced by polymer architecture. The Ka [Equation (6.1)] can be affected by the proximity of multiple acid groups in a polymer chain—this explains differences in pKa of monomers relative to the analogous polymers and is often referred to as the polyelectrolyte effect [1]. This effect is exacerbated in cross-linked or branched PRPN systems where high local concentrations of acid group are trapped by cross-linked or branched points. Thus, in these systems careful consideration of monomer choice and architecture can allow the swelling and deswelling transition to be adjusted over relatively large pH ranges. This presents interesting opportunities in terms of designing complex controlled delivery systems. The multiple cooperative interactions observed in responsive polymeric materials serve to amplify the effect of their response and induce dramatic and useful property changes.

172

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

6.2 pH-Responsive Polymer Micelle Particles Conventional, “nonresponsive” polymer micelles are prepared from amphiphilic copolymers, usually diblock copolymers, and are typically assembled using a cosolvent approach. This is a multistep process that involves dissolving the block copolymer in a thermodynamically good solvent for each component of the block. Providing the solvent is miscible with water (i.e., THF, low alcohols, etc.) slow addition of water causes the hydrophobic domains within the block copolymers to assemble via the hydrophobic effect. The structure of a model block copolymer micelle is shown in Figure 6.1 and involves a hydrophobic polymer core domain stabilized by a hydrophilic polymer corona or shell. The hydrodynamic diameter of the poly mer micelle is deined by the relative block lengths, aggregation number, and the degree of solvation of the micelles. A variety of other self-assembled structures such as worm-like micelles [2], vesicles, and torroids can be assembled by simply altering the polymer architecture and solvent composition; however, here we will focus primarily on simple spherical micelles. In contrast to the permanently amphiphilic block copolymers, so-called responsive double hydrophilic block copolymers offer a direct assembly strategy. Double hydrophilic block copolymers typically comprise two hydrophilic polymer blocks where one of the blocks is permanently hydrophilic (i.e., poly[ethylene glycol], PEG) and the other is responsive or “smart.” When the hydrophilicity of the responsive block responds to changes in solution pH, these block copolymers are referred to as pH-responsive block copolymers. Using a pH-responsive diblock copolymer, comprising a weakly polybasic and permanently hydrophilic block as an example, the block copolymer can initially be dissolved in aqueous solution at acidic pH when the amine is in its protonated, cationic, and hydrophilic form. Slow addition of base to this solution deprotonates the amine group, and the block becomes increasingly hydrophobic. The hydrophobic blocks self-assemble into micellar structures, which are stabilized by the permanently hydrophilic block. Thus, the addition of base has a similar effect to water addition in cosolvent-induced block copolymer assembly—that of slowly and selectively desolvating one block. The synthesis and pH-induced assembly of pH-responsive block copolymers is shown schematically in Figure 6.2. In 1996, Webber and co-workers reported the pH-induced micellization of a poly(2-vinylpyridine)-block-poly(ethyleneglycol) copolymer (P2VP-bPEG) synthesized by stepwise anionic polymerization [3]. This AB diblock copolymer was soluble in aqueous acidic solution as discrete unimers when the P2VP block was protonated and cationic, however it spontaneously selfassembled on raising the solution pH to neutral or basic conditions. The driving force for self-assembly was the deprotonation of the P2VP block, which became increasingly hydrophobic and phase separated from solution above pH 4.8. The covalently attached PEG block provided suficient steric

173

pH-Responsive Polymer Nanoparticles

(a) pH-responsive monomer

(b) pH-responsive homopolymer

(d)

block copolymer micelle

(c) pH-responsive block copolymer

block copolymer micelle

(e)

shell cross-linked micelle (amphiphilic)

shell cross-linked micelle (hydrophilic)

FIGURE 6.2 Schematic of the step-wise synthesis and self-assembly of core-shell poly mer micelles. (a) Controlled poly merization of pH-responsive monomer. (b) Controlled chain-extension polymerization of permanently hydrophilic poly mer block. (c) Solution pH change to induce block copolymer self-assembly (and disassembly). (d) Selective cross-linking of the shell domain of the block copolymer micelle. (e) Solution pH change to induce shell cross-linked micelle swelling (and deswelling).

stabilization under these conditions to produce localized, as opposed to macroscopic, phase separation of the hydrophilic P2VP domains. The resulting assembled polymer structures comprised a dehydrated PVP “core” with a hydrophilic PEG “shell.” This process was reversible and the micelles could be disassembled back into individual unimers by addition of acid. Since this seminal work, an enormous number of advanced pH-responsive copolymers have been synthesized and shown to self-assemble into higher-order structures in aqueous solution. In the rest of this section, we emphasize some of the key developments in this area. Driven by the emergence of various controlled free radical polymerization techniques over the last two decades, the synthesis of pH-responsive block copolymers has become signiicantly more accessible. Techniques such as reversible addition fragmentation chain transfer (RAFT) polymerization [4,5], atom transfer radical polymerization (ATRP) [6,7], and nitroxide-mediated polymerization (NMP) [8,9] permit the facile preparation of controlledstructure block copolymers with increasing architectural, compositional, and functional diversity. Matyjaszewski and co-workers demonstrated the irst controlled ATRP of a pH-responsive monomer, 2-(dimethylamino)ethyl methacrylate (DMAEMA), in a variety of organic solvents [10]. Shortly after, Pelton and co-workers demonstrated that this monomer could be poly merized using aqueous ATRP [11]—that is, ATRP conducted in a more convenient

174

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

aqueous-based solution [12]. Following this the Armes group has demonstrated the synthesis of a range of weakly basic pH-responsive copoly mers using protic ATRP—that is, solution ATRP performed in low alcohol and water solvents—and their pH-driven self-assembly in aqueous solution. In related work, Wooley and co-workers used NMP techniques to prepare a poly(acrylic acid)-block-poly(p-hydroxystyrene) (PAA-b-PpHS) diblock copolymer [13]. In contrast to the tertiary amine-based systems where the pKa of the conjugate acid forms of the amines occur over relatively narrow pH ranges, this acid-based copolymer functional polymer blocks with comprised vastly differing pKa values, which allowed the formation of PpHS-core micelles in aqueous solution from pH 4 to 10. At higher pH values (i.e., above pH 10) the entire polymer was water soluble, and at lower pH values (i.e., below pH 4) the copolymer was entirely hydrophobic and precipitated from solution. An intriguing extension of pH-responsive block copolymer assembly was reported by Armes and Bütün in 2000. It was shown that judicious choice of each block in a pH-responsive diblock copolymer could allow the synthesis of so-called “schizophrenic” diblock copolymers. These new materials required the synthesis of block copolymers with independent responsivity and were deined by their ability to form two distinct types of micelle, that is, micelles and reverse micelles, in aqueous solution by merely changing an external stimulus [14]. The irst example of a purely pH-responsive schizophrenic copolymer concerned a zwitterionic block copolymer comprising a weak polyacid block and a weak polybases block [15] and is shown in Figure 6.3. The copolymer of 2-(diethylamino)ethyl methacrylate (DEAEMA) CH3

(

CH CH2 )60

CH2

C

68

C

O

O CH2 C

O

CH2

OH

N CH3CH2

PVBA

PVBA-core micelles at pH 2

CH2CH3

PDEA

base

base

acid

acid

PVBA-block-PDEA zwitterionic diblock copolymer insoluble at isoelectric point

PDEA-core micelles at pH 10

FIGURE 6.3 Schematic representation of the aqueous solution behavior of the irst example of a pH-responsive “schizophrenic” diblock copolymer. (Reprinted from S. Liu and S. P. Armes, Angewandte Chemie, International Edition 41: 1413–16, 2002. With permission.)

pH-Responsive Polymer Nanoparticles

175

and 4-vinylbenzoic acid (VBA) was synthesized by a two-step ATRP process and formed PVBA-core micelles (Dh = 37 nm) with cationic PDEA shells at pH 2 and PDEAEMA-core micelles (Dh = 35 nm) with anionic PVBA shells in aqueous solution at pH 10. This PDEAEMA-b-PVBA copolymer underwent macroscopic phase separation from solution at around neutral pH due to the presence of an isoelectric point. While no realistic applications of these fascinating copolymers have currently been identiied, they have undoubtedly served to stimulate substantial activity and interest in the ield of responsive polymeric materials. A conceptually different approach to the formation of core-shell PRPNs is to use pH-driven reversible hydrophobic inter- or intrapolymer interactions to form the hydrophobic domains. Gohy et al. have shown that PEGblock-PMAA copolymers are molecularly soluble as discrete unimers at neutral/basic solution pH; however, at low-pH hydrogen bonds form between the EG and MAA units [16]. Providing the block copolymers were asymmetric (in favor of the PEG block) an inter- and intrahydrogenbonded core was formed, which was stabilized by a shell of the excess PEG residues. Oppositely charged polyelectrolytes are known to form strong, multiple cooperative bonds and can facilitate the assembly of various highly complex polymeric structures [17,18]. In 1999, Harada and Kataoka demonstrated chain-length-dependent selectivity over polyelectrolyte interactions using oppositely charged pairs of poly(ethylene glycol)-block-poly(α,β-aspartic acid) and poly(ethylene glycol)-block-poly(L-lysine) [19]. The authors showed that mixtures of the oppositely charged polyions formed weak and illdeined assemblies when the charged blocks were asymmetric; however, when closely matched polyion block lengths were employed, highly monodisperse polyion complex-core micelles assembled, selectively. In an extension of this work, Weaver et al. have also shown that a binary mixture of PMAA with a PEG-b-PDEAEMA block copolymer can form three different types of micelle in aqueous solution by changing solution pH alone. The driving force for micellization is different in each case [20]. At basic pH, PDEAEMA-core micelles stabilized by the PEG block are formed and the anionic PMAA homopolymer is merely a noninteracting “spectator” polymer in solution. On lowering the solution pH below neutral pH, the PDEAEMA residues start to protonate and form electrostatic bonds with the anionic PMAA polymer. Thus, electrostatically core-complexed micelles are formed that remain stabilized in solution by hydrophilic PEG shell. On further reducing the solution pH below around pH 4, the PMAA protonates and loses its anionic character. However, the neutral PMAA homopolymer hydrogen bonds with the PEG block thus forming a reverse micelle comprising a PEG/PMAA hydrogen-bonded core and a cationic PDEAEMA hydrophilic shell. An alternative strategy to reversible pH-induced polymer assembly is to incorporate a hydrophobic core-forming polymer block that can undergo

176

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

acid-triggered degradation. This approach is particularly attractive if the hydrophobic component is a molecule which it would be desirable to deliver in a triggered manner, such as a drug. Park [21] and Kataoka [22] have independently attached doxorubicin, an anticancer drug, to micelle cores using labile hydrazone bonds, which break under acidic conditions. Heller and co-workers have used acid-labile poly(ortho esters) [23] while Fréchet and Gillies favor acetal groups [24]. In principle, covalent attachment of the drug within the PRPN is an attractive strategy since this ensures controllable encapsulation eficiencies while variation of the linking chemistry can provide optimized, or triggered, release rates that can be tuned to occur around physiological pH. In their initial studies, Fréchet and Gillies conjugated a hydrophobic model dye, Nile Red, via a sensitive cyclic acetal linker to a preformed poly(ethylene glycol)-block-poly(aspartic acid) double-hydrophilic block copolymer (Figure 6.4). The inal dye-modiied block copolymer was amphiphilic and thus assembled into core-shell micelles in aqueous solution. Hydrolysis of the acetal group located in the core of the micelle was shown to be pH-sensitive (faster at acidic pH), and since this bond-cleavage process produced more polar diol functionalities, the micelle simultaneously disassembled and released the encapsulated material. The use of pH-controlled hydrolytically labile bonds has more recently been extended by the same group to dendrimers [25]; the interested reader is pointed toward seminal articles for further information on dendrimers [26–28]. OH O O

H N

O x

O OMe O

N H O

H N

O

y

OH NH2

MeO O OMe

EDC, NEt3, H2O, DMF OH O O

H N

O x

O

N H O

H N

O

y

O

Acid hydrolysis

OH O O

H N

O x

O NH

N H O

H N y

NH

O

MeO ‘Protected’ amphiphilic block copolymer

Hydrophilic block copolymer

OH OH OMe

OMe

Hydrophilic block copolymer and release of 2,4,6-trimethoxybenzaldehyde

FIGURE 6.4 Schematic to show triggered release from poly mer particles using acid-labile bonds.

O

pH-Responsive Polymer Nanoparticles

177

Block copolymer micelles with nonglassy, low-Tg cores are typically in dynamic equilibrium with free polymer, or unimers. Similarly to conventional low molar mass surfactants (i.e., sodium dodecylsulphate) the assembled micellar structures exist only above their critical micelle concentration, which can be determined by various techniques including luorescence spectroscopy and surface tensiometry. While polymer nanoparticles are routinely touted as potential drug delivery vehicles, this inherent dynamic unimer-micelle transition unfortunately means that the particles are unstable to dilution and are thus unsuitable for these high-dilution applications.

6.3 pH-Responsive Cross-Linked Micelle Particles In 1996, Thurmond, Kowalewski, and Wooley reported the irst example of covalently-stabilized block copolymer micelles or shell cross-linked micelles (SCMs) [29]. This strategy involved chemically cross-linking self-assembled block copolymer micelles thereby preventing their dilution-induced dissociation. In this irst example, polystyrene-block-poly(4-vinyl pyridine) block copolymers quaternized with 4-(chloromethyl)styrene were used as building blocks to prepare conventional core-shell micelles using a THF/water cosolvent approach. Selective oligomerization of the pendant styrenic residues located in the shell domain at high dilution produced well-deined and chemically stabilized PS-core micelles that were single molecules and thus remained completely stable to dilution. The ield of SCMs has now been developed into an extremely robust platform for the preparation of welldeined spherical polymer nanoparticles in the 10- to 100-nm diameter size range. Some key developments in the synthesis of SCMs have included the use of triblock copolymers (as opposed to diblocks) to allow cross-linking to be performed at high concentrations [30,31], the use of nontoxic physical cross-linking strategies [32] (as opposed to covalent), the formation of reversible cross-linking strategies to allow disassembly of the SCL micelles [33], and the ability to remove the core domain to produce hollow nanocages [34]. In 1999, the Armes group used pH-responsive block copolymers to prepare SCMs with tunable core hydrophilicities in aqueous solution and without recourse to cosolvents [35]. These micelles were assembled from a schizophrenic (see Section 6.2) poly[2-(dimethylamino)ethyl methacrylateblock-methacrylic acid] (PDMAEMA-b-PMAA) diblock copolymer. This copolymer was synthesized by group transfer polymerization and included a postpolymerization deprotection of poly(2-tetrahydropyranyl methacrylate) (THPMA) residues. This block copolymer could form either PMAAcore or PDMAEMA-core micelles depending on the precise sequence and

178

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

conditions of the assembly steps. Importantly, these responsive SCMs possessed cores that could be reversibly hydrated and dehydrated primarily in response to changes in the solution pH. These covalently stabilized block copolymer micelles expanded and contracted as the cores became hydrophilic and hydrophobic, respectively. Various analogous systems have since been developed [36] that have opened up potential applications as drug delivery vehicles that can target speciic regions of the body and release their drug payload in response to subtle changes in the local pH [37] (see Section 6.7 for further applications). A generalized schematic of the cross-linking of block copolymer micelles is also shown in Figure 6.2. An array of shell cross-linking strategies now exist, such as carbodiimide-mediated strategies [38], oligomerization of pendant vinyl groups [29], quaternization of tertiary amines [39], Michael additions [40], inorganic condensation reactions [41], photo-cross-linking of cinnamoyl groups [42], and so-called “click” reactions such as Huisgen 1,3-dipolar cycloadditions between azides and terminal alkynes [43]. Unfortunately, the majority of cross-linking chemistries used to stabilize SCL micelles are invariably toxic and this limits their potential application, especially in vivo. A recent development has shown that polyion complexation can be used to electrostatically cross-link pH-responsive triblock copolymer micelles [32]. In this system, shown schematically in Figure 6.5, a cationic PEG113-block-[QDMAEMA33/ DMAEMA5]-block-DEAEMA50 triblock copolymer (where QDMAEMA O O

O

113

5

33

PEO stabilising corona

pH-tunable DEA core

54

O O O NaOH O O O PEO113-[QDMA33/DMA5]-DEA54 molecularly dissolved at pH 2 – – Cl– H N+ Cl H N+ Cl N+ Cationic QDMA inner-shell

+

+

+

+

DEA-core micelles at pH>7.5 27 nm diameter

+ +

+

+

O O

O

113

PEO113-NaSts34 addition at NaStS/QDMA >1.0 at 20°C Shell and core cross-linked micelles at pH 7.5 35–50 nm diameter

±

FIGURE 6.5 Schematic of the formation of shell cross-linked micelles using the ionic cross-linking strategy. (Reprinted from J. V. M. Weaver, Y. Tang, S. Liu, P. D. Iddon, R. Grigg, N. C. Billingham, S. P. Armes, R. Hunter, and S. P. Rannard, Angewandte Chemie, International Edition 43: 1389–92, 2004. With permission.)

pH-Responsive Polymer Nanoparticles

179

represents methyl-quaternized DMAEMA residues) are self-assembled into three-layer core-shell-corona micelles on increasing the solution pH from acidic to basic pH. This assembly process is entirely reversible in the absence of added polyelectrolyte; however, on addition of a PEG113-block-poly(sodium 4-styrenesulphonate)34 (PEO113-NaStS34) anionic diblock copolymer to the micelle at pH 10, polyion complexation in the central micelle domain occurred that effectively cross-linked the cationic micelles. Interestingly, the PDEAEMA cores remained responsive, and on reducing the solution pH of the electrostatically cross-linked micelles below the pKa of the conjugate acid form of the PDEAEMA residues, additional core cross-linking occurred as the PDEAEMA protonated, providing an excess of the PEO113NaStS34 was employed. These ionically stabilized SCMs were stable to relatively high electrolyte concentration, which, in conjunction with the benign cross-linking, implies that they could be suitable candidates for drug delivery applications.

6.4 pH-Responsive Microgel Particles Microgel particles [44] are cross-linked colloidal polymeric (nano)particles that swell in thermodynamically good solvents [45]. As such, microgels exist in solution in two states: as “hard,” desolvated and compact latexes, or as “soft,” solvated and swollen microgels as depicted in Figure 6.1. A proportionately large subset of these materials are pH-responsive microgels that exhibit a latex-to-microgel transition that is controlled by solution pH. The pH-induced transition between latex and microgel can be conveniently followed by standard techniques such as light scattering as a function of solution pH. A representative transition is shown in Figure 6.6c for a range of PEGMA-stabilized poly(2-vinylpyridine) microgels of different sizes [46]. Transitions such as these are often accompanied by a change in appearance from turbid, desolvated latexes to opaque, solvated microgels. The extent of swelling is deined by factors including the cross-linking density, nature of the cross-linking, presence of comonomers, and ionic strength. The overall degree of swelling is dictated by mobile counterions within the microgel that inluence the internal osmotic pressure, and this is balanced by the internal electrostatic repulsion [47]. The kinetics of pH-responsive microgel swelling has been studied [46,48] following Tanaka’s work on macroscopic gels [49,50]. The precise nature of the swelling is an important component in deining the properties and potential application of these pH-responsive microgels but is relatively poorly understood. Most theories of microgel swelling assume a uniform cross-linking density; however, in reality this is unlikely to be the

Differential Weight Distribution

Entry 6

Entry 1

0.9

0.7

2 µm

1.1

Particle Diameter (µm) (a)

Hydrodynamic Diameter (nm)

4500 4000 3500 3000

2 µm

Entry 3

Entry 4

6

5

2 µm

2 µm

2500 2000

Entry 2

Entry 5

4

Entry 6

1500 1–3 1000 500 2

3

4

5

6

pH (c)

7

8

9

10

11

2 µm

2 µm (b)

FIGURE 6.6 (a) Size distribution and (b) scanning electron microscopy images of a series of PEGMA-stabilized poly(2-vinylpyridine) particles in their latex form. (c) Mean hydrodynamic diameters as a function of solution pH to show the latex-to-microgel transition. (Reprinted from D. Dupin, S. Fujii, S. P. Armes, S. Reeve, and S. M. Baxter, Langmuir 22: 3381–87, 2006. With permission.)

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

0.5

0.3

Entry 5

180

Entry 1 Entry 2 Entry 3 Entry 4

pH-Responsive Polymer Nanoparticles

181

case for most systems. A more likely scenario is that cross-linking density decreases from the core of the microgel particles, and this hypothesis has been supported for poly(methyl methacrylate) microgel particles dispersed in benzene that demonstrated a distinctly inhomogeneous cross-linking density [51]. Flory’s theories concerning network swelling in organic solvents [52] are unfortunately inadequate for aqueous pH-responsive microgel systems with heterogeneous internal structures. However, Hoare and Mclean have recently developed a more suitable kinetic model to semiquantitatively predict the spatial distribution of functionality in microgel systems [53]. pH-responsive microgels are typically comprised of weakly ionic monomers where the swollen microgel phase is induced by electrostatic repulsion, and inherent hydrophilicity increases in the monomer units in their ionic state. The main classes include alkali-swellable microgels based on (meth)acrylic acid [54,55], acid-swellable microgels based on basic monomers such as vinyl pyridines [48], and tertiary amine methacrylates and N-isopropylacrylamide-based microgels with acidic or basic comonomers. They are commonly synthesized by precipitation [56] or inverse emulsion polymerization [57]. The microgel diameter is strongly inluenced by the synthetic approach employed, and microgels have been reported with dimensions ranging from around 10 nm to several hundreds of microns. pH-responsive nanogels share many of the properties of microgels, however they have signiicantly smaller hydrodynamic diameters. They can be synthesized via emulsion polymerization, miniemulsion polymerization, or postmodiication of self-assembled copolymers. Their smaller dimensions mean their surface-to-volume ratios are signiicantly higher than microgels, and these sizes are potentially more viable for biotechnology applications. Nagasaki’s group has prepared pH-responsive nanogels based on DEAEMA in the 50- to 350-nm size range [58]. These nanogels were synthesized via emulsion polymerization using a specially designed heterotelechelic PEGbased macromonomer as steric stabilizer. These nanogels exhibited a sharp volume transition around neutral pH, which was deined by the apparent pKa of the PDEAEMA residues. Such nanogels are postulated to have useful properties for bio-delivery such as the enhancement of the endosomal escape. In fact, it has been suggested that the buildup of osmotic pressure and endosomal protonation from polyelectrolyte nanogels may destabilize and disrupt endosomes [59,60]. Yusa’s group has developed a self-assembly based strategy to form welldeined pH-responsive nanogels based on DEAEMA [61]. A poly(ethylene glycol)-block-poly(DEAEMA-co-2-cinnamoyloxyethyl acrylate) statistical diblock copolymer synthesized by RAFT polymerization was self-assembled into PDEAEMA-core micelles using a solution pH-switch followed by core cross-linking of the cinnamoyl groups by photo-dimerization yielding a nanogel. These particles had smaller hydrodynamic diameters of around 25–30 nm at pH 10.

182

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

6.5 pH-Responsive Branched Copolymer Particles The preparation of architecturally and compositionally controlled branched copolymers has been a perennial synthetic polymer challenge. They remain an active research area since branched polymers offer numerous different properties to linear polymers. For example, branched polymers have lower viscosities, higher solubilities, and higher concentrations of end groups compared to analogous linear polymers—each of which can be desirable for speciic applications. The most common synthetic approach to branched copolymers has been step-growth polymerization [62]; however, there is a strong commercial drive to produce branching technologies that can be applied to vinyl monomers, and in pursuit of this goal Fréchet and co-workers developed self-condensing vinyl polymerization [63–65]. This strategy uses a multifunctional vinyl monomer that can undergo both conventional addition polymerization while also initiating new “branched” chains. While this technique is relevant to large classes of vinyl monomers, it remains restrictive by requiring specially designed branching monomers. In 2000, Sherrington’s group reported a generic, eficient, and scaleable route to the preparation of branched copolymers of essentially any vinyl monomer [66]. Their strategy, often referred to as the Strathclyde approach to branched polymers, uses a conventional divinyl cross-linking monomer to introduce branching points during a standard free radical polymerization. Providing suficient chain transfer agent is employed, relative to the concentration of the multifunctional “branching monomer,” completely soluble branched polymer is produced in high yield and without recourse to tailor-made reactants. More recently it has been shown that controlled polymerization processes can be used to prevent gelation by targeting relatively low primary chain lengths [69]. While the mechanism of branching appears to alter slightly depending on the polymerization technique employed [67,68], in general, providing less than one branching monomer is employed, highly branched, soluble polymer with controlled primary chain lengths is possible using this technique [69]. It should be emphasized that above this critical branching level, these structures transform from soluble branched polymers to “gel-like” materials, which will not be discussed in any detail here [70,71]. The Strathclyde approach to branched polymers has recently been exploited by groups at Liverpool as a platform to synthesize novel core-shell PRPNs, similar to micelles and shell cross-linked micelles; however, unlike these materials, they are prepared in one pot and do not require time-consuming assembly and toxic cross-linking steps. One example of their synthesis involves the statistical conventional free radical branching polymerization of the common pH-responsive tertiary amine methacrylate monomer, DEAEMA, with a permanently hydrophilic macromonomer, PEG22MA, in the presence of the branching monomer ethyleneglycol dimethacrylate, EGDMA, and a chain transfer agent that prevents gelation. A schematic of

183

pH-Responsive Polymer Nanoparticles

TG

DDT O

R

OH

O

O O

or HO 5 SH AIBN, ethanol, 40h

SH

S

O

O O

O

O O

O

N

O 22

DEA

PEGMA

O

O O

EGDMA

N

O

O polymer

O

22

polymer

Branched copolymer

FIGURE 6.7 Schematic of the one-step synthesis of PRPNs via the Strathclyde approach. (Reprinted from J. V. M. Weaver, R. T. Williams, B. J. L. Royles, P. H. Findlay, A. I. Cooper, and S. P. Rannard, Soft Matter 4: 985–92, 2008. With permission.).

the synthesis is shown in Figure 6.7. There is a range of different polymer architectures with controllable compositions and chain ends that are deined by the choice of chain transfer agent. These branched copolymers can be dissolved molecularly in acidic aqueous solution to give cationic and highly hydrated structures; however, on addition of base, they contract to give core-shell structures similar to pH-responsive micelles or SCMs [95] (see Sections 6.2 and 6.3, respectively). Interestingly, although the free radical polymerization process is “uncontrolled” in terms of the molecular weight and polydispersity of primary chains, the resulting branched copolymer nanoparticles that exist in basic solution have relatively narrow particle size distributions as determined by dynamic light scattering and transmission electron microscopy (TEM). In principle, the core, shell, and chain end functionality can all be readily varied by judicious choice of monofunctional monomers, branching monomers, and chain transfer agent, and thus various applications are predicted from this new class of viable particles. These pH-responsive branched copolymer particles have some structural similarity to star polymers—that is, polymers synthesized by propagating several linear chains from a single, multifunctional core initiator [72,73]—however, discussion of this class of material is beyond the scope of this chapter.

6.6 Polymer Nanoparticles with pH-Responsive Surfaces The majority of PRPNs described so far comprise cores that respond to solution pH; however, surface or shell responsivity can also be highly desirable for certain systems. In these examples, however, the shell responsivity tends to produce a disperse-to-agglomerated nanoparticle transition as the shell functionality reversibly switches between being capable of stabilizing the core to having insuficient hydrophilicity to allow solubilization. A prevalent example includes surface-functionalized polymer latexes—that is, insoluble

184

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(or hydrophobic) polymer particles with surfaces that enable their eficient dispersion and stabilization in water. Conventional latexes are typically synthesized by techniques including emulsion or dispersion poly merization and use conventional surfactants to provide either electrostatic or steric stability. An alternative synthetic approach involves substituting the surfactant with a steric stabilizer such as a hydrophilic polymer. In this instance, the polymer is chemically or physically incorporated onto the latex surface during the polymerization via either the pendant reactive functionality or presence of hydrophobic domains, respectively. If the polymer stabilizer comprises weak polyacid or polybase, then the surfaces of the latex are effectively functionalized with a responsive polymer that can undergo a transition from charged and hydrophilic to neutral and hydrophobic. For example, well-deined polystyrene latexes can be prepared via alcoholic dispersion polymerization using a PDMAEMA-block-poly(alykyl methacrylate) copolymer as a physically adsorbed steric stabilizer [74]. Under optimized conditions, the latex size can be controlled by varying the concentration and molar mass of the hydrophilic stabilizer. Some applications of surface-responsive latexes are described in Section 6.7. Vincent and co-workers have exploited the pH-responsive surfaces of near-monodisperse polymer latexes to assemble these colloids into onedimensional “microrods” using temporary electric dipoles induced by an applied ac electric ield [75]. The particles are prevented from coming into direct contact by extensive electrostatic repulsion; however, the authors demonstrated that these transient structures could be “glued” together using a smaller, oppositely charged microgel particle thus maintaining the string structures after removing the external ield. The observed periodicity in the particle separations of these assemblies in combination with the reversibility of the system suggests that these materials could be used as sensing devices where diffraction of white light by the ordered spacings could produce different colors in response to control of surface charges using solution pH. Wanless and co-workers investigated the interfacial adsorption behavior of pH-responsive diblock copolymers with responsive cores and shells on mica: a model lat, solid, and charged surface [76]. Weakly charged pH-responsive (PDMAEMA-stat-PQDMAEMA)-block-PDEAEMA diblock copolymers were shown not only to controllably self-assemble from solvated unimers to core-shell micelles in aqueous solution in response to increasing the solution pH from acidic to basic, but also to self-organize into ordered monolayertype ilms as judged by wet atomic force microscopy (AFM). Interestingly, micelle dissociation and desorption did not occur to any appreciable extent on reducing the solution pH, presumably due to the low levels of permanent cationic functionality present on the micelle surfaces. These results augur well for the organization of various polymeric species and encapsulants on surfaces in multiple dimensions.

pH-Responsive Polymer Nanoparticles

185

6.7 Applications of pH-Responsive Polymer Particles The self-assembly of pH-responsive block copolymers offers a number of opportunities because of their controllable and reversible nanoparticleunimer transition. Indeed this nonlinear transition from highly soluble and chain expanded to collapsed core is ideal for the controlled uptake and release of actives, and it is not surprising that perhaps the most widely mooted application of block copolymer micelles is as drug delivery vehicles [77]. For example, PRPNs capable of release of actives at acidic pH could have application in environments where local pH variances could be used as convenient triggers such as in tumor tissue or endosomic or lysosomic cellular compartments. Moreover their sizes, which are typically around 20–100 nm, are suficiently small to avoid uptake by the reticuloendothelial system and also prevent rapid renal exclusion [78]. Of course, a prerequisite for these systems is the use of biocompatible and approved polymers and starting materials—concepts that are often ignored in favor of demonstrating viability. In terms of intrinsic physiological release of encapsulated drugs from PRPNs within the body, an important consideration is the presence and nature of inherent pH variation. The most widely suggested potential application of pH-responsive SCMs in the literature is as biological encapsulation and delivery vehicles for in vivo administration of drug payloads. Wooley and co-workers have investigated in detail the cross-linking of acidic shell functionality using diamines, thus producing SCMs with pH-responsive surfaces. The same group has used the responsive and functional surfaces of these SCMs as chemical handles onto which various biologically relevant motifs can be attached [79]. In an alternative approach, they prepared biotin-functionalized SCMs from biotin-labeled poly(acrylic acid)-block-poly(methyl acrylate) copolymers and also studied the competitive binding of these PRPNs with avidin [80]. Liu and co-workers have prepared SCMs from α-aldehyde-functionalized poly(DMAEMAblock-DEAEMA) copolymers with both pH-responsive cores and shells [81]. The free surface aldehyde groups permitted their subsequent bioconjugation with lysozyme via the formation of Schiff bases. The McCormick group has shown that cystamine cross-linked poly(ethylene glycol)-block-poly(N,Ndimethylacrylamide/N-acryloxysuccinimide)-block-poly(N-isopropylacrylamide) triblock copolymers can retard the release of hydrophobic drugs from the core compared to un-cross-linked micelles [33]. While these SCL micelles are not truly pH-responsive, but rather thermally sensitive, it demonstrates the added beneit of cross-linking on release. Moreover, since these SCL micelles were cross-linked using cystamine, they could be readily un-cross-linked by cleavage of the disulide bond using either tris(2-carboxyethyl)phosphine or dithiothreitol. This further supports their proposed applicability in terms

186

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

of allowing their disassembly and thus eficient renal exclusion after use. Wooley and co-workers have prepared pH-responsive SCL micelles with hydrolytically labile cross-linkers that also bear UV-vis active components to detect SCL cleavage events during pH-induced destabilization [82]. The authors observed accelerated degradation of the cross-links at around lysosomal pH, which provides further corroboration that these diverse systems can, in principle, be optimized for speciic bio-applications. An alternative potential application of PRPNs is as nanoreactors for chemical transformations and catalysis. O’Reilly points out in a recent review article on this subject [37] that micellar catalysis is already well explored, and provides examples of reaction rate enhancement and retardation; however, these systems typically use unstabilized, that is, un-cross-linked, conventional surfactant micelles and are thus inherently less robust and optimized than many PRPN systems highlighted in this chapter. To demonstrate this point, three-layer pH-responsive SCL micelles assembled and cross-linked from a poly(ethylene oxide)-block-poly(glycerol monomethacrylate)-blockpoly(2-diethylaminoethyl methacrylate) triblock copolymer have been used as nanoreactors for the preparation of gold nanoparticles [40]. Having prepared the SCM with hydrophobic PDEAEMA cores at basic solution pH, addition of an HAuCl4 solution protonated the tertiary amine groups thereby electrostatically binding the AuCl4-counterion selectively within the SCL core domain. Reduction of the gold salt by NaBH4 produced gold nanoparticles, presumably templated by the core, that exhibited excellent long-term colloid stability. Shell cross-linking is vital to this application as protonation of the PDEAEMA residues in the analogous un-cross-linked polymer micelle resulted in dissociation of the micelle back into individual unimeric chains. Several potential applications of responsive microgels have been postulated and are currently being investigated such as drug encapsulation and delivery, regenerative medicine, and templates for chemical syntheses [83,84]. Drug delivery is particularly well suited to microgel systems since they can be readily prepared to have suficient “stealth” properties to successfully evade the immune system. Fréchet and co-workers have used labile acetalbased cross-linkers (see Section 6.2) as a pH-triggered release mechanism [85]. This approach has been further developed to produce biocompatible poly(2-hydroxyethyl methacrylate) (PHEMA) microgels that could release bovine serum albumin payloads over several hours. More importantly, the encapsulation and release process did not denature the protein. The mechanical strength of partially pH-responsive microgels has recently been exploited by Saunders and co-workers to aid degenerated intervertebral disks—a common global cause of back pain [86,87]. In this process, poly(methacrylic acid)-based microgel dispersions are delivered to affected areas by injection. A luid-to-gel transition occurs in vivo as the pH increases above ~pH 6. The use of ionic pH-responsive microgels is crucial for this application since the charge concentration generated when the microgels swell results in signiicant swelling pressures.

pH-Responsive Polymer Nanoparticles

187

Surface-responsive polymer latexes have recently been exploited in the context of responsive or reversible interfacial stabilization of biphasic materials. Indeed, various PRPNs have been investigated because of their ability to controllably and reversibly adsorb at interfaces. Polystyrene latexes with pHresponsive PDMAEMA surfaces have been used as particulate emulsiiers for the preparation of oil-in-water emulsions when the PDMAEMA surface is in its neutral, and more hydrophobic, form [88]. Addition of acid to the emulsion causes the PDMAEMA residues to protonate and become cationic and hydrophilic, which induces desorption from the oil/water droplet interface. Consequently this pH switch behaves as a suitable trigger for rapid demulsiication, or phase separation, which has various potential applications, not least in tertiary oil recovery where demulsiication can be a key step in the recovery of the oil. More recently pH-responsive branched copolymers (see Section 6.5) based on PEGMA and DEAEMA have also been shown to eficiently stabilize relatively small emulsion droplets at basic solution pH and at relatively low polymer concentrations [89]. Importantly, judicious control of the degree of branching and the nature of the branching monomer in the PRPN allowed unprecedented control over the pH-responsive properties of the resulting emulsion droplets. Consequently responsive emulsions were prepared that could either remain completely stable (in which case the surface charge changes from neutral to cationic), partially demulsify, or completely demulsify by subtle variation of the branches’ length and chain end properties on reduction of the solution pH. This implies that these responsive emulsion systems could be used as controllable release devices for the delivery of substantial hydrophobic payloads. In addition to the ability to “make or break” emulsions, carefully designed functional branched copolymer particles have also been used to assemble and disassemble emulsion droplets in a process referred to as emulsion engineering [90,96]. Branched copolymers of PMAA and PEGMA with hydrophobic dodecane chain ends have been synthesized using the Strathclyde branching technique (see Section 6.5). The architecture and composition of these copolymers were speciically chosen so as to provide: (i) strong and multiple points of adhesion of the polymer onto the droplet surface via the dodecane chain ends, and (ii) simultaneous steric and electrostatic stabilization of the droplets at basic solution. On lowering the solution pH of the emulsion droplets below ~pH 4, the PMAA residues protonated and lost their ability to charge-stabilize the emulsion droplets. In addition, the PMAA residues in their neutral state formed hydrogen bonds with the ethylene glycol units of the PEGMA and thus secondary bonding cross-links were established between functionality on the emulsion droplet surfaces. The reversible transition between free-lowing emulsion dispersion at basic pH and the assembled emulsion droplets at acidic pH is shown in Figure 6.8a–b. The extensive hydrogen-bonded networks formed between surface-functionalized droplets at acidic pH can be used to kinetically trap the droplets in precise morphologies and with signiicant degrees of complexity, as shown

188

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

a

b base acid

Emulsion droplet attraction ENGINEERED EMULSIONS d

c

2.5 mm

Emulsion droplet repulsion CONVENTIONAL DISPERSION e

f

2.5 mm

g

2.5 mm

700 µm

h

4 mm

1.5 mm

250 µm

100 µm

0.8 mm

1 mm

250 µm

250 µm

1 mm

1 mm

1 mm

1 mm

i

j

FIGURE 6.8 (a–b) Schematic representation of the reversible assembly and disassembly of emulsion droplets—the emulsion engineering process [90]. (c–g) Images of various engineered emulsion structures. (h–i) Images of morphologically controlled engineered emulsions at increasing resolution. (j) Base-triggered disassembly of engineered emulsion ibers. (Reprinted from J. V. M. Weaver, S. P. Rannard, and A. I Cooper, Angewandte Chemie International Edition 48: 2131–34, 2009. With permission.)

in Figure 6.8c–i. However, since the emulsion droplet surface functionality is responsive to solution pH, the engineered emulsions can be readily disassembled into conventional, disperse, and free-lowing emulsions by raising the solution pH (Figure 6.8j). The ability to reversibly assemble emulsion droplets into controlled structure arrays without destabilizing the droplets has potential in advanced encapsulation and delivery applications where the controlled release and isolation of actives could be advantageous. In a process somewhat analogous to the stabilization of liquid-liquid interfaces in emulsions, under certain conditions solid colloidal particles can

pH-Responsive Polymer Nanoparticles

189

stabilize biphasic mixtures of air and water [91]. Binks and Murakami have shown that either air-in-water foams or water-in-air materials, dubbed “dry water,” can be stabilized by judicious control of the surface hydrophobicity of silica colloids [92]. The different phases can be accessed by varying the nature of the liquid/air phases and the contact angle of the particle at the interface. Foams and dry water have both existing and potential applications, and thus their extension to more complex, responsive systems stands to make substantial impact in areas such as home and personal care products, foods, separations, and the synthesis of advanced materials. The groups of Binks and Armes have shown that polymer latexes with poly(acrylic acid) surfaces can stabilize foams eficiently at low pH when the latex is in its less hydrophilic, neutral form [93]. Defoaming could be triggered by simply increasing the pH of the foam. This pH increase deprotonated the PAA residues and produced anionic, colloidally stable latexes that desorbed from the air-water interface. A similar concept is described in a recent manuscript by Dupin and Armes, who show that “liquid marbles”—millimeter-sized water droplets stabilized by hydrophobic particulates—can be formed using polymer latexes with pH-responsive PDEAEMA-surfaces at high pH when the PDEAEMA is in its neutral, hydrophobic form [94]. The liquid marbles remain intact on being placed on water at basic pH or solid surfaces for several hours at 20ºC. The authors postulate that instabilities over longer time frames are caused by water evaporation. In the case of the liquid marbles on water, reduction of the solution pH causes the structures to rapidly dissociate as the PDEAEMA residues are protonated and become cationic and hydrophilic. The authors demonstrate the encapsulation and triggered release of hydrophilic dyes within the liquid marbles, which suggests their potential application in microencapsulation and biotechnology processes.

6.8 Outlook In this chapter it has been possible to cover only a few of the most pertinent concepts in the rapidly progressing, expanding, and diversifying ield of pH-responsive polymer nanoparticles. A number of excellent reviews have been cited throughout, which the interested reader is directed to for further information. The ield of pH-responsive polymer particles is continuing to expand rapidly and broaden its applicability in ever more diverse areas of materials science and beyond. However, continued development is strongly reliant on two key factors: irst, the development of new and more advanced synthetic routes to the preparation of responsive materials, and second, the identiication and optimization of applications for these complex systems. These two factors should be inherently linked, especially where the synthesis stimulates further applications research. In this chapter, a number

190

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

of direct and speciic applications of PRPNs have been described, such as the use of pH-responsive particles for drug delivery applications and the use of pH-responsive microgels to address degenerated intervertebral disks (see Section 6.7); however, their broader scope should not be underestimated and may prove to be a fruitful application of these materials over the next few decades. PRPNs are incredibly versatile materials and their synthesis is fundamentally well understood and can be trivial. There are a number of examples of using PRPNs to effectively impart their responsivity onto other materials, such as in the case of emulsion engineering (see Section 6.7), and this is a clear opportunity to maximize the potential exploitation of these fascinating classes of materials.

References 1. Weaver, J. V. M., Williams, R. T., Royles, B. J. L., Findlay, P. H., Cooper, A. I., and S. P. Rannard. 2008. pH-responsive branched polymer nanoparticles. Soft Matter 4:985–92. 2. Wang, X-S., Winnik, M. A., and I. Manners. 2002. Synthesis and aqueous selfassembly of a polyferrocenylsilane-block-poly(aminoalkyl methacrylate) diblock copolymer. Macromolecular Rapid Communications 23:210–13. 3. Martin, T. J., Prochazka, K., Munk, P., and S. E. Webber. 1996. pH-dependent micellization of poly(2-vinylpyridine)-block-poly(ethyleneoxide). Macromolecules 29:6071–73. 4. Chiefari, J., Chong, Y. K., Ercole, F., Krstina, J., Jeffrey, J., Le, T. P. T., Mayadunne, R. T. A., Mejis, G. F., Moad, C. L., Moad, G., Rizzardo, E., and S. H. Thang. 1998. Living free-radical polymerization by reversible addition-fragmentation chain transfer: The RAFT process. Macromolecules 3:5559–62. 5. Moad, G., Rizzardo, E., and S. H. Thang. 2005. Living radical polymerization by the RAFT process. Australian Journal of Chemistry 58:379–410. 6. Kato, M., Kanigaito, M., Sawamoto, M., and T. Higashimura. 1995. Polymerization of methyl methacrylate with the carbon tetrachloride dichlorotris(triphenylph osphine) ruthenium(II) methylaluminium bis(2,6-di-tert-butylphenoxide) initiating system—possibility of living radical polymerization. Macromolecules 28:1721–23. 7. Wang, J-S., and K. Matyjaszewski. 1995. Controlled living radical polymerization—atom transfer radical polymerization in the presence of transition metal catalysts. Journal of the American Chemical Society 117:5614–15. 8. Georges, M. K., Veregin, R. P. N., Kazmaier, P. M., and G. K. Hamer. 1993. Narrow molecular weight resins by a free radical polymerization process. Macromolecules 26:2987. 9. Hawker, C. J., Bosman, A. W., and E. Harth. 2001. New polymer synthesis by nitroxide mediated living radical polymerizations. Chemical Reviews 101:3661–88.

pH-Responsive Polymer Nanoparticles

191

10. Zhang, X., Xia, J., and K. Matyjaszewki. 1998. Controlled/“living” radical polymerization of 2-(dimethylamino)ethyl methacrylate. Macromolecules 31:5167–79. 11. Zheng, F., Shen, Y., Zhu, S., and R. Pelton. 2000. Atom transfer radical polymerization of 2-(dimethylamino)ethyl methacrylate in aqueous solution. Journal of Polymer Science, Part A: Polymer Chemistry 38:3821–27. 12. Wang, X-S., Jackson, R. A., and S. P. Armes. 2000. Facile synthesis of acidic copolymers via atom transfer radical polymerization in aqueous media at ambient temperature. Macromolecules 33:255–57. 13. Lee, N. S., Li, Y., Marcus Ruda, C. M., and K. L. Wooley. 2008. Aqueous-only, pH-induced nanoassembly of dual pKa-driven contraphilic block copolymers. Chemical Communications 5339–41. 14. Butun, V., Liu, S., Weaver, J. V. M., Bories-Azeau, X., Cai, Y., and S. P. Armes. 2006. A brief review of “schizophrenic” block copolymers. Reactive and Functional Polymers 66:157–65. 15. Liu, S., and S. P. Armes. 2002. Polymeric surfactants for the new millennium: A pH-responsive, zwitterionic, schizophrenic diblock copolymer. Angewandte Chemie, International Edition 41:1413–16. 16. Gohy, J. F., Varshney, S. K., and R. Jerome. 2001. Water-soluble complexes formed by poly(4-vinylpydrinium)-block-poly(ethyleneoxide) and poly(sodium methacrylate)-block-poly(ethyleneoxide) copolymers. Macromolecules 34:3361–66. 17. Decher, G. 1997. Fussy nanoassemblies: Towards layered polymeric multicomposities. Science 277:1232–37. 18. Caruso, F., Caruso, R. A., and H. Mohwald. 1998. Nanoengineering of inorganic and hybrid hollow spheres by colloidal templating. Science 282:1111–14. 19. Harada, A., and K. Kataoka.1999. Chain length recognition: Core-shell supramolecular assembly from oppositely charged block copolymers. Science 283:65–67. 20. Weaver, J. V. M., Armes, S. P., and S. Liu. 2003. A “holy trinity” of micellar aggregates in aqueous solution at ambient temperature: Unprecedented selfassembly behavior from a binary mixture of a neutral-cationic diblock copolymer and an anionic polyelectrolyte. Macromolecules 36:9994–98. 21. Yoo, H. S., Lee, E. A., and T. G. Park. 2002. Doxorubicin-conjugated biodegradable polymeric micelles having acid-cleavable linkages. Journal of Controlled Release 82:17–27. 22. Bae, Y., Fukushima, S., Harada, A., and K. Kataoka. 2003. Design of environmentsensitive supramolecular assemblies for intracellular drug delivery: Polymeric micelles that are responsive to intracellular pH change. Angewandte Chemie, International Edition 42:4640–43. 23. Heller, J., Barr, J., Ng, S. Y., Abdellauoi, K. S., and R. Gurny. 2002. Poly(ortho esters): Synthesis, characterisation, properties and uses. Advanced Drug Delivery Reviews 54:1015–39. 24. Gillies, E. R., and J. M. J. Fréchet. 2003. A new approach towards acid sensitive copolymer micelles for drug delivery. Chemical Communications 1640–11. 25. Gillies, E. R., and J. M. J. Fréchet. 2005. pH-responsive copolymer assemblies for controlled release of doxorubicin. Bioconjugate Chemistry 16:361–68. 26. Tomalia, D. A., Baker, H., Dewald, J., Hall, M., Kallos, G., Martin, S., Roeck, J., Ryder, J., and P. Smith. 1985. A new class of polymers—starburst-dendritic macromolecules. Polymer Journal 17:117–32.

192

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

27. Newkome, G. R., Yao, Z. Q., Baker, G. R., and V. K. Gupta. 1985. Micelles. 1. Cascade molecules—a new approach to micelles. Journal of Organic Chemistry 50:2003–4. 28. Hawker, C. J., and J. M. J. Frechet. 1990. Preparation of polymers with controlled molecular architecture—A new convergent approach to dendritic macromolecules. Journal of the American Chemical Society 112:7638–47. 29. Thurmond, K. B., Kowalewski, T., and K. L. Wooley. 1996. Water-soluble knedellike structures: The preparation of shell-cross-linked small particles. Journal of the American Chemical Society 118:7239–40. 30. Butun, V., Wang, X-S., de Paz-Banez, M. V., Robinson, K. L., Billingham, N. C., Armes, S. P., and Z. Tuzar. 2000. Synthesis of shell cross-linked micelles at high solids in aqueous media. Macromolecules 33:1–3. 31. Liu, S., and S. P. Armes. 2001. The facile one-pot synthesis of shell cross-linked micelles in aqueous solution at high solids. Journal of the American Chemical Society 123:9910–11. 32. Weaver, J. V. M., Tang, Y., Liu, S., Iddon, P. D., Grigg, R., Billingham, N. C., Armes, S. P., Hunter, R., and S. P. Rannard. 2004. Preparation of shell cross-linked micelles by polyelectrolyte complexation. Angewandte Chemie, International Edition 43:1389–92. 33. Li, Y., Lokitz, B. S., Armes, S. P., and C. L. McCormick. 2006. Synthesis of reversible shell cross-linked micelles for controlled release of bioactive agents. Macromolecules 39:2726–28. 34. Huang, H., Remsen, E. E., Kowalewski, T., and K. L. Wooley. 1999. Nanocages derived from shell cross-linked micelle templates. Journal of the American Chemical Society 121:3805–6. 35. Butun, V., Lowe, A. B., Billingham, N. C., and S. P. Armes. 1999. Synthesis of zwitterionic shell cross-linked micelles. Journal of the American Chemical Society 121:4288–89. 36. Zhang, Z., Liu, G., and S. Bell. 2000. Synthesis of poly(solketal methacrylate)block-poly(2-(dimethylamino)ethyl methacrylate) and preparation of nanospheres with cross-linked shells. Macromolecules 33:7877–83. 37. O’Reilly, R. K. 2007. Spherical polymer micelles: Nanosized reaction vessels? Philosophical Transactions of Royal Society A 365:2863–78. 38. Huang, H., Kowalewski, T., Remsen, E. E., Gertzmann, R., and K. L. Wooley. 1997. Hydrogel-coated glassy nanospheres: A novel method for the synthesis of shell cross-linked knedels. Journal of the American Chemical Society 119:11653–59. 39. Butun, V., Billingham, N. C., and S. P. Armes. 1998. Synthesis of shell crosslinked micelles with tunable hydrophilic/hydrophobic cores. Journal of the American Chemical Society 120:12135–36. 40. Liu, S., Weaver, J. V. M., Save, M., and S. P. Armes. 2002. Synthesis of pHresponsive shell cross-linked micelles and their use as nanoreactors for the preparation of gold nanoparticles Langmuir 18:8350–57. 41. Yuan, J. J., Mykhaylyk, O. O., Ryan, A. J., and S. P. Armes. 2007. Cross-linking of cationic block copolymer micelles by silica deposition. Journal of the American Chemical Society 129:1717–23. 42. Underhill, R. S., and G. Liu. 2000. Triblock nanospheres and their use as templates for inorganic nanoparticle preparation. Chemistry of Materials 12:2082–91.

pH-Responsive Polymer Nanoparticles

193

43. Joralemon, M. J., O’Reilly, R. K., Hawker, C. J., and K. L. Wooley. 2005. Shell clickcrosslinked (SCC) nanoparticles: A new methodology for synthesis and orthogonal functionalization. Journal of the American Chemical Society 127:16892–199. 44. Baker, W. O. 1949. Microgel, a new macromolecule—relation to sol and gel as structural elements of synthetic rubber. Industrial Engineering and Chemistry 41:511–20. 45. Staudinger, H., and E. Huseman. 1935. On the swellable polystyrene. Chemische Berichte 68:1618–34. 46. Dupin, D., Fujii, S., Armes, S. P., Reeve, S., and S. M. Baxter. 2006. Eficient synthesis of sterically stabilized pH-responsive microgels of controllable particle diameter by emulsion polymerization. Langmuir 22:3381–87. 47. Saunders, B. R., and B. Vincent. 1999. Microgel particles as model colloids: Theory, properties and applications. Advances in Colloid and Interface Science 80:1–25. 48. Loxley, A., and B. Vincent. 1997. Equilibrium and kinetic aspects of the pHdependent swelling of poly(2-vinylpyridine-co-styrene) microgels. Colloid and Polymer Science 275:1108–14. 49. Tanaka, T., Hocker, L. O., and G. B. Benedek. 1973. Spectrum of light scattered from a viscoelastic microgel. Journal of Chemical Physics 59:5151–59. 50. Tanaka, T., and D. J. Fillmore. 1979. Kinetics of swelling of gels. Journal of Chemical Physics 70:1214–18. 51. Nieuwenhuis, E. A., and A. Vrij. 1979. Light scattering of PMMA latex-particles in benzene—structural effects. Journal of Colloid and Interface Science 72:321–41. 52. Flory, P. J. 1953. Principles of polymer chemistry. Ithaca, NY: Cornell University Press. 53. Hoare, T., and D. Mclean. 2006. Kinetic prediction of functional group distributions in thermosensitive microgels. Journal of Physical Chemistry B 41:20327–36. 54. Rodriguez, B. E., Wolfe, M. S., and M. Fryd. 1994. Nonuniform swelling of alkali swellable microgels. Macromolecules 27:6642–47. 55. Saunders, B. R., Crowther, H. M., and B. Vincent. 1997. Poly[(methyl methacrylate)-co-(methacrylic acid)] microgel particles: Swelling control using pH, co-nonsolvency, and osmotic deswelling. Macromolecules 30:482–87. 56. Pelton, R. 2000. Temperature-sensitive aqueous microgels. Advances in Colloid and Interface Science 85:1–33. 57. Landfester, K. 2006. Synthesis of colloidal particles in miniemulsions. Annual Review of Materials Research 36:231–79. 58. Hayashi, H., Iijima, M., Kataoka, K., and Y. Nagasaki. 2004. pH-sensitive nanogel possessing reactive PEG tethered chains on the surface. Macromolecules 37:5389–96. 59. Hu, Y., Litwin, T., Nagaraja, A. R., Kwong, B., Katz, J., Watson, N., and D. J. Irvine. 2007. Cytosolic delivery of membrane-impermeable molecules in dendritic cells using pH-responsive core-shell nanoparticles. Nano Letters 7:3056–64. 60. Soppimath, K. S., Liu, L. H., Seow, Y. W., Liu, S. Q., Powell, R., Chan, P., and Y. Y. Yang. 2007. Multifunctional core/shell nanoparticles self-assembled from pHinduced thermosensitive polymers for targeted intracellular anticancer drug delivery. Advanced Functional Materials 17:355–62. 61. Yusa, S., Sugashara, M., Endo, T., and Y. Morishima. 2009. Preparation and characterization of a pH-responsive nanogel based on a photo-cross-linked micelle formed from block copolymers with controlled structure. Langmuir 25:5258–65.

194

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

62. Kim, Y. H., and O. W. Webster. 1990. Water-soluble hyperbranched polyphenylene—a unimolecular micelle. Journal of the American Chemical Society 112:4592–93. 63. Fréchet, J. M. J., Henmi, M., Gitsov, I., Aoshima, S., Leduc, M. R., and R. B. Grubbs. 1995. Self-condensing vinyl polymerization—An approach to dendritic materials. Science 269:1080–83. 64. Hawker, C. J., Frechet, J. M. J., Grubbs, R. B., and J. Dao 1995. Preparation of hyperbranched and star polymers by a living, self-condensing free radical polymerization. Journal of the American Chemical Society 117:10763–64. 65. Weimar, M. W., Frechet, J. M. J., and I. Gitsov. 1998. Importance of active-site reactivity and reaction conditions in the preparation of hyperbranched polymers by self-condensing vinyl polymerization: Highly branched vs. linear poly[4-(chloromethyl)styrene] by metal-catalyzed “living” radical polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 36:955–70. 66. O’Brien, N., McKee, A., and D. C. Sherrington. 2000. Facile, versatile and cost effective route to branched vinyl polymers. Polymer 41:6027–31. 67. Gao, H., Min, K., and K. Matyjaszewski. 2007. Determination of gel point during atom transfer radical copolymerization with cross-linker. Macromolecules 40:7763–70. 68. Bannister, I., Billingham, N. C., Armes, S. P., Rannard, S. P., and P. H. Findlay. Development of branching in living radical copolymerization of vinyl and divinyl monomers. Macromolecules 39:7483–92. 69. Isaure, F., Cormack, P. A. G., Graham, S., Sherrington, D. C., Armes, S. P., and V. Butun. 2004. Synthesis of branched poly(methyl methacrylate)s via controlled/living polymerisations exploiting ethylene glycol dimethacrylate as branching agent. Chemical Communications 9:1138–39. 70. Gao, H. F., and K. Matyjaszewski. 2009. Synthesis of functional polymers with controlled architecture by CRP of monomers in the presence of cross-linkers: From stars to gels. Progress in Polymer Science 34:317–50. 71. Blencowe, A., Tan, J. F., Goh, T. K., and G. G. Qiao. 2009. Core cross-linked star polymers via controlled radical polymerisation. Polymer 50:5–32. 72. Hedrick, J. L., Trollsas, M., Hawker, C. J., Atthoff, B., Claesson, H., Heise, A., Muller, R. D., Mecerreyes, D., Jerome, R., and P. Dubois. 1998. Dendrimer-like star, block and amphiphilic copolymers by combination of ring opening and atom transfer radical polymerization. Macromolecules 31:8691–8705. 73. Jones, M-C., Ranger, M., and J-C. Leroux. 2003. pH-sensitive unimolecular polymeric micelles: Synthesis of a novel drug carrier. Bioconjugate Chemistry 14:774–81. 74. Baines, F. L., Dionisio, S., Billingham, N. C., and S. P. Armes. 1996. Use of block copolymer stabilizers for the dispersion polymerization of styrene in alcoholic media. Macromolecules 29:3096–3102. 75. Snoswell, D. R. E., Brill, R. K., and B. Vincent. 2007. pH-responsive microrods produced by electric-ield-induced aggregation of colloidal particles. Advanced Materials 19:1523–27. 76. Webber, G. B., Wanless, E. J., Butun, V., Armes, S. P., and S. Biggs. 2002. Selforganized monolayer ilms of stimulus-responsive micelles. Nano Letters 2:1307–13. 77. Bader, H., Ringsdorf, H., and B. Schmidt. 1984. Water soluble polymers in medicine. Angewandte Makromolekulare Chemie 123:457–85.

pH-Responsive Polymer Nanoparticles

195

78. Kwon, G. S., and T. Okano. 1996. Polymeric micelles as new drug delivery vehicles. Advanced Drug Delivery Reviews 21:107–16. 79. Nystrom, A. M., and K. L. Wooley. 2008. Thiol-functionalized shell crosslinked knedel-like (SCK) nanoparticles: A versatile entry for their conjugation with biomacromolecules. Tetrahedron 64:8543–52. 80. Qi, K., Ma, Q. G., Remsen, E. E., Clark, C. G., and K. L. Wooley. 2004. Determination of the bioavailability of biotin conjugated onto shell cross-linked (SCK) nanoparticles. Journal of the American Chemical Society 126:6599–6607. 81. Liu, H., Jiang, X., Fan, J., Wang, G., and S. Liu. 2007. Aldehyde surface-functionalized shell cross-linked micelles with pH-tunable core swellability and their bioconjugation with lysozyme. Macromolecules 40:9074–83. 82. Li, Y., Du, W., Sun, G., and K. L. Wooley. 2008. pH-responsive shell cross-linked nanoparticles with hydrolytically labile cross-links. Macromolecules 41:6605–07. 83. Saunders, B. R., Laajam, N., Daly, E., Teow, S., Hu, X., and R. Stepto. 2009. Microgels: From responsive polymer colloids to biomaterials. Advances in Colloid and Interface Science 147:251–62. 84. Tan, B. H., and K. C. Tam. 2008. Review on the dynamics and micro-structure of pH-responsive nano-colloidal systems. Advances in Colloid and Interface Science 136:25–44. 85. Murthy, N., Thng, Y. X., Schuck, S., Xu, M. C., and J. M. J. Fréchet. 2002. A novel strategy for encapsulation and release of proteins: Hydrogels and microgels with acid-labile acetal cross-linkers. Journal of the American Chemical Society 124:12398–99. 86. Freemont, T. J., and B. R. Saunders. 2008. pH-responsive microgel dispersions for repairing damaged load-bearing soft tissue. Soft Matter 4:919–24. 87. Lally, S., Mackenzie, P., Le Maitre, C. L., Freemont, T. J., and B. R. Saunders. 2007. Microgel particles containing methacrylic acid: pH-triggered swelling behaviour and potential for biomaterial application. Journal of Colloid and Interface Science 316:367–75. 88. Amalvy, J. I., Armes, S. P., Binks, B. P., Rodrigues, J. A., and G.-F. Unali. 2003. Use of sterically-stabilised polystyrene latex particles as a pH-responsive particulate emulsiier to prepare surfactant-free oil-in-water emulsions. Chemical Communications 1826–27. 89. Woodward, R. T., Slater, R. A., Higgins, S., Rannard, S. P., Cooper, A. I., Royles, B. J. L., Findlay, P. H., and J. V. M. Weaver. 2009. Controlling responsive emulsion properties via polymer design. Chemical Communications 3554–56. 90. Weaver, J. V. M., Rannard, S. P., and A. I Cooper. 2009. Polymer-mediated hierarchical and reversible emulsion droplet assembly. Angewandte Chemie International Edition 48:2131–34. 91. Binks, B. P., and T. S. Horozov. 2006. Colloidal particles at liquid interfaces. Cambridge: Cambridge University Press. 92. Binks, B. P., and R. Murakami. 2006. Phase inversion of particle-stabilized materials from foams to dry water. Nature Materials 5:865–69. 93. Binks, B. P., Murakami, R., Armes, S. P., Fujii, S., and A. Schmid. 2007. pHresponsive aqueous foams stabilized by ionizable latex particles. Langmuir 23:8691–94. 94. Dupin, D., Armes, S. P., and S. Fujii. 2009. Stimulus-responsive liquid marbles. Journal of the American Chemical Society 131:5386–87.

196

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

95. Weaver, J. V. M. and D. J. Adams. 2010. Synthesis and application of pH-responsive branched copolymer nanoparticles (PRBNs): A comparison with pHresponsive shell cross-linked micelles. Soft Matter. Manuscript in press. 96. Woodward, R. T., Chen, L., Adams, D. J., and J. V. M. Weaver. 2010. Journal of Materials Chemistry. Manuscript in press.

7 Smart Thermo-Responsive Nanoparticles Peng Tian and Qinglin Wu CONTENTS 7.1 Introduction to Thermo-Responsive Polymer and Nanoparticles ..... 197 7.1.1 Thermo-Responsive Polymer ....................................................... 197 7.1.2 Phase Transition of Thermo-Responsive Polymer .................... 198 7.1.3 Thermo-Responsive PNIPAAm Nanoparticles ......................... 199 7.2 Thermo-Responsive Nanoparticle Synthesis and Characterization ......................................................................................... 200 7.3 Thermo-Responsive Nanoparticle Size Control .................................... 205 7.4 Thermo-Responsive Nanoparticle Applications ................................... 208 7.5 Conclusions and Future Development ................................................... 214 References............................................................................................................. 215

7.1 Introduction to Thermo-Responsive Polymer and Nanoparticles 7.1.1 Thermo-Responsive Polymer Stimuli-responsive polymers, also called smart materials or intellectual materials, show an interesting property change behavior to external stimulus. These external stimuli, including pH [1,2], temperature [3–6], light [7,8], magnetic and electric ield [9,10], and solvent/salt [11,12], result in a change of the polymers in conformation, solubility, and hydrophilic/hydrophobic balance. The property change of the polymer to external environmental condition change can be linear or nonlinear. The nonlinear change attracts more scientiic interest since a small stimulus change may result in a large response. Based on their response to different stimuli, stimuli-responsive polymers can be classiied into several categories: thermo-responsive poly mer, pH-responsive polymer, magnetic-responsive polymer, and double-/multiresponsive polymer due to their response to a combination of two or more stimuli. Furthermore, based on different structures of stimuli-responsive polymers, there are linear polymers, cross-linked polymers (hydrogel), micelles, core-shell particles, and polymer-protein conjugates. Stimuli-responsive 197

198

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

polymers have shown applications in many ields, such as drug encapsulation and delivery [13–17], diagnostics [18], sensing [19,20], fabrication of photonic crystals [21,22], nanoparticle templates [23,24], and separation and puriication technologies [25,26]. Among the stimuli-responsive polymers, thermo-responsive polymers are one of the most widely studied polymers because of their temperature-sensitive properties [27,28]. 7.1.2 Phase Transition of Thermo-Responsive Polymer Thermo-responsive polymers undergo phase transition in a suitable solvent at certain temperature. The volume phase transition behavior of thermoresponsive polymer is an imbalance between repulsive and attractive forces acting in the polymer [29], such as hydrogen bonding, Van der Waals interaction, attractive ionic interaction, and hydrophobic interaction. This volume phase transition results in a sudden change of solubility of the polymer in the solvent. When the temperature is increased above the lower critical solution temperature (LCST) or decreased below the upper critical solution temperature (UCST), a phase separation or the coil-to-globule transition behavior is observed. When the polymer is immersed in a solvent like aqueous solution, negative entropy of mixing is formed. As temperature changes, the forces between the hydrophilic and hydrophobic parts of the polymer chain and water molecule, such as hydrogen bonding, come to a balance, and the polymer changes from transparent to opaque [30–32]. Compared to the polymer-polymer and water-water interactions, the hydrogen bonding between the polymer and water becomes weak and unfavorable at LCST/UCST; thus, a transition occurs because the hydrated hydrophilic macromolecule dehydrates quickly [33]. Polymers that undergo LCST behavior include poly(N-isopropylacrylamide) (PNIPAAm or PNIPAM) [31,32,34], polymers of N-substituted acrylamides [35,36], poly(N,N-diethylacrylamide) (PDEAM) [37], poly(ethylene oxide)(PEO) [38], poly(methyl vinyl ether) (PMVE) [39,40], poly(N-vinylcaprolactam) (PNVCL) [41,42], and copolymers based on these polymers. Polymers that undergo UCST behavior include copolymers of acrylamide (PAAm) and acrylic acid (PAAc) interpenetrating polymer network (IPN) [43–46], etc. The LCST/UCST of polymers can be controlled by polymerizing copolymers of the aforementioned polymers with other hydrophilic or hydrophobic polymers. Furthermore, copolymers of the previously mentioned polymers and polymers having other stimuli-responsive behavior, such as pH or magnetic, have the capacity of double or multiresponse to environmental change [44]. For example, core/shell poly(N-isopropylacrylamide-co-glycidyl methacrylate) (PNIPAAm-GMA) particles can be synthesized with emulsion copolymerization [45]. Thiol compounds are then used to modify the template particles for incorporating ionic groups such as amino, suffonate, or carboxyl groups and in situ synthesis of magnetic nanoparticles. Another example is

Smart Thermo-Responsive Nanoparticles

199

Bhattacharya’s work on ferromagnetic nanoparticles, which show responsive ability to temperature, pH, and magnetic ield [46]. Among thermo-responsive polymers, PNIPAAm- and PNIPAAm-based copolymers are well known and attract the most signiicant scientiic research attention because they show a low critical phase separation from water at an LCST of approximately 30–34°C [35]. Unlike PDEAM, which has an LCST around 32–34°C as well [37], the LCST of PNIPAAm is irrelevant to the tacticity of the polymer and independent of molecule weight and concentration [47]. The close-to-human-body-temperature LCST and the biocompatibility property of PNIPAAm make it an ideal polymer for controlled drug delivery. We will mainly focus our discussion of thermo-responsive poly mer on PNIPAAm and its copolymers. 7.1.3 Thermo-Responsive PNIPAAm Nanoparticles Although extensive work on thermo-responsive polymers based on PNIPAAm has been done in the past 20 years, and some promising results of the polymeric self-aggregates as drug carriers have been reported, study of thermo-responsive polymeric nanoparticles and nanogels is still a new ield that attracts research and development interest just in recent years. The synthesis and characterization of the PNIPAAm-based copolymer nanoparticles still need much research before they can be used for drug delivery, and how the nanoscale behavior of the particles affects the delivery eficiency needs to be thoroughly understood. Beside the thermo-responsive behavior as their macro counterpart, the micro- or nanoscopic dimension of thermo-responsive nanoparticles introduces more advantages [48–56]. For example, with their much smaller size, it is easier for thermo-responsive nanoparticles to access areas of the human body where macroscopic networks cannot reach. It is of signiicant importance for therapeutic drugs to be delivered to targeted cells. Besides, the larger surface area of nanoparticles has more beneit compared with macroscopic ones when conjugating with siRNAs and peptides, thus resulting in better in vivo treatment performance. Furthermore, thermo-responsive nanoparticles have the capacity of much rapid response to temperature change, which provides a broader choice for many applications. Some of the PNIPAAm-based thermo-responsive nanoparticles reported include poly(acrylonitrile-co-N-isopropylacrylamide) (p[AN-co-NIPAAM]), poly(N-isopropylacrylamide)-b-poly(ε-caprolactone), poly(NIPAAm-co-N-[2-hydroxypropyl] methacrylamide-dilactate) (poly[NIPAAm-co-HPMAm-dilactate]), poly(ethylene oxide)-b-poly(Nisopropylacrylamide) (PEO-b-PNIPAAm), and poly(N-isopropylacrylamide-co-N,N-dimethylacrylamide-co-10-undecenoic acid) (P[NIPAAm-co-DMAAm-co-UA]) [57–61]. For thermo-responsive polymer nanoparticles and hydrogel, it is dificult to control the stability and size of the particles. In the following sections we summarize the research work

200

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

on the PNIPAAm-based block copolymers as temperature-sensitive polymeric nanoparticles. It is noted that the “nanoparticles” here do not strictly limit to particles with diameter less than 100 nm.

7.2 Thermo-Responsive Nanoparticle Synthesis and Characterization By introducing hydrophilic or hydrophobic groups into PNIPAAm polymer, the LCST of the thermo-responsive nanoparticles can be controlled to some extent, as well as introducing other stimuli-responsive properties. Moreover, chemical or physical cross-linkers can help the polymer to form a network structure (i.e., the hydrogel), which increases the structure stability and hinders polymer chains from dissolving into aqueous solution, thus changing the LCST of the nanoparticles. For example, by adding a small amount of cross-linking agents with the polymerization, such as the methyleneBisacrylamide (BisAAm), cross-linked PNIPAAm hydrogels could be obtained. These hydrogels also show a volume phase transition at the LCST. Similar to the linear ones, when the temperature is increased above the LCST, the cross-linked PNIPAAm collapses substantially [62–64]. The different structures of the linear and cross-linked PNIPAAm give different levels of hydrogen bonds between the polymer chains and water, and thus lead to different phase transition behaviors, although the process is similar. At a temperature above the LCST, the hydrogel gives out water in the pore and becomes stiff and opaque. As the temperature is decreased below the LCST, the PNIPAAm hydrogel reswells in water, but at a slower rate than the initial deswelling process [62]. The hydrogel tends to shrink rapidly at irst and then deswell slowly, as shown by Okano and co-workers [65,66]. Approaches for developing polymer nanoparticles include, but are not limited to, cross-linking of block copolymer micelles in solution, surface-initiated polymerization of monomers, layer-by-layer adsorption of electrolyte polymers on templates, etc. [67–71]. The achievement of thermo-responsive nanoparticles and nanogels is often fulilled by emulsion polymerization and precipitation polymerization methods [72,73], or by physical-chemical technology such as particle replication in nonwetting templates (PRINT) [74,75]. In both cases, the control of particle size, shape, particle size distribution (PSD), particle stability, the responsive time for temperature change, and polymer composition is vitally important, as it affects the LCST behavior of the nanoparticles, thus affecting their inal applications. Fernandez et al. studied thermo-responsive nanostructured PNIPAAm hydrogel synthesized by a two-stage process [76]. First, slightly cross-linked PNIPAAm nanoparticles were fabricated via microemulsion polymerization.

Smart Thermo-Responsive Nanoparticles

201

Second, the developed particles were cross-linked with N,N-methylene-bisacrylamide (BisAAm) in aqueous solution after drying, cleaning, and dispersing to form nanostructured hydrogel. It is shown that z-average diameter was 29.5 nm for the nanoparticles at irst stage. Larger equilibrium water uptake, faster responsive behavior, and larger Young moduli were found for the nanogels compared with their macro counterparts, and a similar LCST of 34°C was shown. Sahiner and co-workers reported poly(acrylonitrile-co-N-isopropylacrylamide) (p[AN-co-NIPAAm]) hydrogel nanoparticles synthesized by microemulsion polymerization [77]. The nanoparticle has a core-shell structure with poly(N-isopropylacrylamide) as the shell and poly(acrylonitrile) (p(AN)) as the hydrophobic core. By adjusting the reaction conditions, highly monodispersed nanoparticles with particle size of range 50–150 nm can be obtained. The synthesized nanoparticles were characterized with transmission electron microscopy (TEM), scanning electron microscopy (SEM), and dynamic light scattering (DLS) for morphology and particle size, and tested with release of a model drug, propranolol (PPL). The amidoximation of the hydrophobic core material helped increase the loading/release capacity by two times due to tunability in the hydrophobicity/hydrophilicity balance of the composite nanoparticle. Thermo-sensitive poly(N-isopropylacrylamide)-b-poly(ε-caprolactone) (PNIPAAm-CL) with different PCL block lengths was synthesized by hydroxyterminated PNIPAAm-initiated ring opening polymerization of ε-caprolactone by Choi et al. [78]. In their study, self-assembled polymeric nanoparticles were formed in aqueous solution with PNIPAAm as the shell. They showed that PCL block length had great effect on particle sizes as well as critical aggregation concentrations. Drug release tests with a model drug of clonazepam (CNZ) demonstrated that the thermo-sensitive hydrogel layer worked as an additional diffusion barrier and enhanced sustained drug release patterns. Another example is the core-shell nanoparticles of poly(ethylene oxide)-bpoly(Nisopropylacrylamide) (PEO-b-PNIPAAm) with N,N-bis(acryloylcystamine) (BAC) slightly cross-linked, reported by Zeng and co-workers [79]. DLS and luorescence measurements indicated the nanoparticle size was less than 150 nm with BAC content ranging from 0.75 wt% to 0.2 wt% of the mass of NIPAAm. Stability study showed that with 0.5 wt% BAC, the nanoparticles were stable up to two weeks at room temperature. Atom transfer radical polymerization (ATRP) was used to synthesize PEG-bPNIPAAm block copolymer nanoparticles with N,N’-ethylenebisacrylamide as cross-linker in different solvent [80]. It is shown that compositions of the mixed solvent have an effect on the inal particle size. For example, hydrogel particles prepared in H2O and H2O/THF solvents had different particle sizes (i.e., 67.5 and 503.3 nm, respectively). Besides, photo-cross-link reactions can also be utilized for synthesizing nanogels, by which surfactant may or may not be used and nanogel with more

202

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

homogeneous cross-linking density can be prepared [81,82]. Kuckling and co-workers developed poly(N-isopropylacrylamide-co-2-dimethylmaleinimido ethylacrylamide) (PNIPAAm-DMIAAm) nanogels by UV-irradiated solutions of thermo-sensitive polymers in water at 45°C and studied the factors that affect the particle size, such as the concentration of photopolymer solutions, the amount of DMIAAm in the photopolymer chains, and surfactant sodium dodecyl sulfate (SDS) concentrations [81]. Signiicant scientiic effort has been made on the hybrid nanoparticles with PNIPAAm shell with a metal or inorganic/organic core [83–89]. The advantage of this method is that the particle size can be controlled by the mental or inorganic template, and speciic optical, electronic, and magnetic properties can be introduced by the template materials. Furthermore, hollow nanoparticles can be obtained if needed. Chemical properties of the hybrid nanoparticles are determined by the polymer shell. To synthesize this kind of nanoparticle, physical or chemical surface modiication of the metal or inorganic core particles is necessary for the polymers to be adsorbed or covalently grafted. Hollow nanoparticles can be achieved by etching of the metal or inorganic core from the polymer-coated nanoparticles. Hollow PNIPAAm nanoparticles with sub 50-nm dimension were synthesized by precipitation polymerization process onto Au nanoparticles, followed by etching of the Au core with KCN [83]. Similar work using isothiocyanate luorescein (FITCentrapped SiO2) as the core template to synthesize novel thermo-sensitive PNIPAAm nanocapsules was reported by Gao et al. [88]. In their work, SiO2 core was etched by hydroluoric acid, and the pretrapped FITC molecules left were used to study the permeability of the PNIPAAm nanogels. It was shown that the FITC could permeate the PNIPAAm shell around its LCST at 32.8°C. The preceding examples represent only a small number of PNIPAAmbased nanoparticles reported to show the synthesis and characterization of the thermo-responsive nanoparticles and nanogels. More discussion can be seen in published papers and review articles [90–94]. We use poly(Nisopropylacrylamided-co-methacrylic acid) (PNIPAAm-MAA) nanoparticles as an example for some detailed discussion on synthesis of nanoparticles and how the monomer ratio and the amount of cross-linker and surfactant affect the particle size and the phase transition behavior of the nanoparticles. The MAA group introduces additional ionized groups to the copolymer and results in pH responsive behavior, but this property is not the main focus in this chapter. The structure of PNIPAAm-MMA is shown in Figure 7.1. Temperature- and pH-responsive polymeric composite membranes prepared from nanoparticles of PNIPAAm-MAA, and their permeability to proteins and peptides in response to environmental stimuli were investigated by Zhang et al. [95]. They showed that permeability of the solutes across the membranes increased with increased temperature or particle concentration, and decreased with increased pH and molecular size of the solutes. However, there is a lack of detailed investigation on factors that affect the particle size and phase transition behavior. Huang et al. studied volume phase transition

203

Smart Thermo-Responsive Nanoparticles

x

y

O HN

O

HO

FIGURE 7.1 Chemical structure of PNIPAAm-MMA.

of PNIPAAm-MAA copolymer nanoparticles in buffer solutions at various pH levels and in aqueous solutions of KCl or ionic surfactants using the DLS technique [96]. It was found that the swelling behavior of the nanoparticles had a close relationship with pKa of copolymer, and there was a sharp volume phase transition for these nanoparticles. PNIPAAm-MAA particles can be synthesized by precipitation polymerization [97]. In preparing the particles, the target amount of NIPAAm, MAA, BisAAm, and SDS was added to deionized water, where NIPAAm and MAA served as the monomers, BisAAm as the cross-linking agent, and SDS as surfactant. To study how these components affect the particle size, different molar ratios of MAA to NIPAAm, and various levels of BisAAm and SDS were used (Table 7.1). The solution was stirred for 1 h followed by deoxygenating by bubbling with nitrogen for 40 min, and then heated to 70°C in a nitrogen environment. After that, 0.083 g of potassium persulfate (KPS) in a 10-ml water solution was added to the reactor to initiate poly merization. Reaction was kept at 70°C under nitrogen atmosphere for 4 h with constant stirring. The formed particle solution was then puriied by membrane dialysis against distilled deionized (DDI) water using Spectra/Pros membrane (Fisher Scientiic Inc., Pittsburgh, Pennsylvania) with a molecular weight cutoff of 12,000–14,000. LCSTs of the PNIPAAm-MAA particles can be measured by turbidity test or differential scanning calorimetry (DSC). The DSC allows the measurement of the heat of phase transition over a wide range of temperatures and pH-values. The onset of the thermogram corresponds to the temperature of collapse of the polymer and thus was treated as the LCST of PNIPAAm-MAA particles. Particle size and size distribution of the copolymers can be determined using DLS technology. Measurements were performed over a temperature range of 20–60°C for diluted PNIPAAm-MAA particles in 0.1M PBS solution. Z-average diameter was used as the hydrodynamic size because it is more reproducible than volume and number weighted mean diameters. Figure 7.2 shows the effect of monomer ratio, amount of surfactant, and cross-linker on phase transition temperature of pNIPAAm-MAA particles. Notice in the igure that phase transition temperatures (i.e., LCST) of PNIPAAm-MAA particles were all higher than that of PNIPAAm itself.

204

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 7.1 Summary of Experimental Design and Selected Results of Synthesized PNIPAM-co-MAA Particles (1.9 g N-isopropylacrylamide [NIPAAm] and 120 ml Water for Every Polymerization Process). Methacrylic Acid (MAA) (g)

N,N’methylenebisacrylamide (BisAAm) (g)

Sodium dodecyl sulfate (SDS) (g)

Potassium persulfate (KPS) (g)

Low Critical Solution Temperature (LCST) (°C)

Particle Diameter at 20°C (nm)

0 0.036 0.072 0.108 0.144 0.576 1.008

0.033 0.033 0.033 0.033 0.033 0.033 0.033

0.075 0.075 0.075 0.075 0.075 0.075 0.075

0.083 0.083 0.083 0.083 0.083 0.083 0.083

32.06 32.47 32.93 35.45 32.13 33.16 51.31

300.1(4.482) 311.3(2.454) 358.6(3.251) 425.4(5.347) 396.7(5.261) 444.5(5.828) 554.5(1.671)

0.036 0.036 0.036 0.036

0.033 0.033 0.033 0.033

0.100 0.150 0.225 0.300

0.083 0.083 0.083 0.083

35.23 40.02 40.02 45.20

274.4(1.054) 140.1(2.548) 50.5(0.228) 39.9(2.427)

0.036 0.036 0.036

0.017 0.500 0.067

0.075 0.075 0.075

0.083 0.083 0.083

32.76 34.35 37.42

341.3(1.577) 300.3(3.003) 292.0(4.055)

Source: Data from P. Tian, Q. Wu, and K. Lian, Journal of Applied Polymer Science 108: 2226–32, 2008. Note: Reaction time was four hours for every poly merization process. The num-

bers in parentheses for diameter (D) at 20°C represent standard derivation of three measurements.

However, an obvious increase can be detected for MAA:NIPAAm ratios of only 0.075 and 0.7 (Figure 7.2). Apparently PNIPAAm-MAA particles inherited the thermo-responsive behavior of PNIPAAm. DSC results of particles under various pH levels suggested that LCSTs of the particles were also sensitive to pH changes. PMAA undergoes a marked pH-induced conformation transition. At low pH levels, PMAA chains are highly compacted to minimize the hydrophobic interaction, while at high pH levels, PMAA chains show expanded coil. The swelling/collapsing of PMAA gels are also highly pH dependent [98]. The polyelectrolyte behavior of the MAA unit introduced various coexistent intra- and intermolecular forces, thus complicating the phase transition process of the hydrogel particles. The MAA group introduced more hydrophilic group (COO-) to the copolymer, which decreased the hydrophobic interaction that determines the phase transition of the hydrogel. Besides, with highly ionized MAA groups, the electrostatic force also helped balance the hydrophobic interaction, thus leading to an increase of the LCST. This effect is especially obvious at MAA content level

205

Smart Thermo-Responsive Nanoparticles

55 50

LCST (degree C)

45 40 35 MAA

30

SDS Cross-linker

25 20 0

0.08

0.16 0.24 0.32 0.4 0.48 0.56 0.64 MAA/NIPAM Ratio/SDS(g)/Cross-linker(g)

0.72

0.8

FIGURE 7.2 Effect of monomer ratio, surfactant content, and cross-linker content to phase transition temperature, LCST, of PNIPAAm-MAA copolymer particles. (Reprinted from P. Tian, Q. Wu, and K. Lian, Journal of Applied Polymer Science 108: 2226–32, 2008. With permission.)

of 0.7, where an LCST of about 51.31°C was observed. Acrylic acid groups of MAA in PNIPAAm-MAA copolymer made the particle gel undergo a discontinuous transition, a higher LCST, and a larger volume change, which is consistent with results reported by other researchers [99,100]. Figure 7.2 also shows the surfactant effect on phase transition temperature of the PNIPAAm-MAA particles with MAA/NIPAAm = 0.025. It is clear that as the surfactant content increased, there was an apparent increase of the LCST. This was due to the fact that amphiphilic structure of the surfactant SDS helped solubilize the polymer in aqueous solutions, thus isolating the hydrophobic polymer segments from the aqueous environment. The contents of the cross-linker also played an important role in the LCST of the copolymer particles, as depicted in Figure 7.2. Although not as obvious as the effect of surfactants, LCST of the copolymer particles increased with higher cross-linker contents. This is possibly because more cross-linker gave more compact size of the copolymer particles, thus increasing the electrostatic repulsion forces between the polymer chains and preventing the particles from shrinking.

7.3 Thermo-Responsive Nanoparticle Size Control For the nanoparticles, stability, mean particle size, and PSD are the main focus in design and synthesis of thermo-responsive polymers/copolymers.

206

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Besides, much research effort needs to be made on determining how particle size inluences the phase transition behavior, thus the releasing mechanism of the nanoparticles or nanogels. The shape of the PSD, which includes the broadness of the PSD, the modality, and the speciic particle size covered by the PSD, has a signiicant effect on the properties of nanoparticles and the performance of the products made with nano-sized precursors. The controlled adjustment of particle size is of great interest due to size-dependent physical and chemical properties of nanoparticles [101]. In the pharmaceutical industry, for example, PSD of the active pharmaceutical ingredients is one of the most important aspects of the dosage form that affect the effectiveness of drug release [102]. Besides, PSD may also affect the stability of the dosage and absorption rate of active intergradients. As a result, it is essential to control PSD in the development of pharmaceutical dosages [103]. Factors that inluence the particle size of the thermo-responsive nanoparticles are complicated. It is strongly dependent on the ratio and block length of polymer/copolymer components, the amount of cross-linker and surfactant used, the reaction conditions, and the diameter of the template particles. When referring to “particle size” or “particle diameter,” the condition has to be deined since the particles are sensitive to temperature and environmental change. Neradovic and co-workers studied the dependence of nanoparticle size on the processing and formulation parameters of poly(N-isopropylacrylamide-co-ethylene glycol) (PNIPAAm-PEG), where PEG is hydrophilic block and PNIPAAm or PNIPAAm-co-HPMAm-dilactate act as a thermoresponsive block with DLS [104]. They found that the formation of small and stable nanoparticles strongly depends on PEG molecule weight. Small particle size is more easily obtained with high molecular weight PNIPAAm and lower polymer concentration in water instead of PBS buffer as solvent. Moreover, the heating rate of the polymer solutions plays a dominant role for nanoparticle size control, and a fast heating rate is more favorable for smaller nanoparticle formation. Continuing with our discussion of PNIPAAm-MAA particles introduced earlier, hydrodynamic diameters of PNIPAAm-MAA particles with different MAA:NIPAAm ratios as a function of temperature are shown in Figure 7.3. It is clear that the Z-average diameter of particles was composition dependent. As the MAA ratio increased, the diameter of the copolymer particles increased because more MAA units resulted in longer chains. For each set of samples, the diameter of the particles tended to decrease as temperature increased because hydrogel shrunk as temperature increased, especially around its phase transition temperature. At a certain temperature level, PNIPAAm-MAA particles tended to aggregate. For example, for PNIPAAm particles without MAA, particles tended to aggregate at above 30°C (around its LCST). This made it dificult to measure the particle size, leading to incomplete data points over the temperature range of 20–60°C as shown in Figure 7.3. Increasing temperature weakened

207

Smart Thermo-Responsive Nanoparticles

600 MAA/NIPAAm = 0 MAA/NIPAAm = 0.025 MAA/NIPAAm = 0.05 MAA/NIPAAm = 0.075 MAA/NIPAAm = 0.1 MAA/NIPAAm = 0.4 MAA/NIPAAm = 0.7

Diameter (nm)

500 400 300 200 100 0 15

25

35

45

55

65

LCST (°C) FIGURE 7.3 The effect of temperature on the size of the PNIPAAm-MAA particles with different MAA:NIPAAm ratio. (Reprinted from P. Tian, Q. Wu, and K. Lian, Journal of Applied Polymer Science 108: 2226–32, 2008. With permission.)

the hydrogen bonding between copolymer and water, thus polymer chains tended to collapse with each other. As there were more MAAs, the particles were more stable in aqueous solution as temperature changed. This should ascribe to the fact that the MAA groups introduced electrolyte behavior, and the electrostatic repulsion force of the particles helped stabilize the copolymer particles in solution. The ratio of the particle diameter at a given temperature (D) to that at 20°C (D20), D/ D20, was used to indicate the changing tendency of particles. The ratio was reduced as the MAA ratio increased, showing that hydrophobic interaction was balanced by the effect of electrostatic repulsion from the charged MAA groups. With higher MAA ratios, the electrostatic force in particles became stronger. The temperature dependence of copolymer particles’ diameter with different surfactant contents is shown in Figure 7.4. Obviously, at the same temperature, the diameter of the copolymer particles decreased as the surfactant content increased. The amphiphilic structure of SDS surfactants helped solubilize polymer particles in aqueous solutions, thus making it possible for polymer chains to form smaller particles. The effect of surfactant on the diameter of the particles was signiicant. For example, as SDS changed from 0.10 g to 0.3 g, the particle size changed from 274.4 nm to 39.9 nm at 20°C. Since PNIPAAm-MAA copolymer was negatively charged because of ionization of COOH groups, and SDS was an anionic surfactant, the contents of the SDS helped disperse the copolymer particles in water. This could be beneicial for controlling the size of the particles for a speciic purpose, such as drug delivery. The sharpness of the D:D20 ratio curve change was reduced as SDS increased. Higher surfactant contents gave stronger repulsion in particles and

208

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

350 SDS = 0.075g 300

SDS = 0.15g SDS = 0.1g

Diameter (nm)

250

SDS = 0.225g 200

SDS = 0.3g

150 100 50 0 15

25

35

45

55

65

LCST (°C) FIGURE 7.4 The effect of temperature on the diameter of the PNIPAAm-MAA particles with different surfactant contents. (Reprinted from P. Tian, Q. Wu, and K. Lian, Journal of Applied Polymer Science 108: 2226–32, 2008. With permission.)

smaller volume reduction as temperature increased. At temperatures higher than 30°C, the particles tended to aggregate, since higher temperature broke the balance of hydrogen bonds between polymer and water, and neighboring polymer chains had increased possibility to collapse with each other. Figure 7.5 shows the temperature dependence of the diameter of the copolymer particles with different cross-linking agent contents. It can be seen from the igure that although a decrease in diameter with increased cross-linker content was observed, the effect of cross-linker on the particle size was not as obvious as that of the monomer ratio and surfactant. At 20°C, the diameter of the particle decreased from 341.3 nm to 292.0 nm as the cross-linker content increased from 0.017 g to 0.068 g. The sharpness of the D/D20 curve transition was reduced upon the increase of the cross-linker. This was possibly because the increase of the cross-linking of the polymer helped increase the electrostatic repulsion within particles and thus reduced the volume change. This result suggested that by controlling the amount of surfactant and monomer ratio, PNIPAAm-MAA nanoparticles with desired size for speciic applications could be developed.

7.4 Thermo-Responsive Nanoparticle Applications It was discussed earlier that thermo-responsive polymers show potential applications in many ields including targeted drug delivery, sensing,

209

Smart Thermo-Responsive Nanoparticles

400 BIS = 0.017g BIS = 0.033g BIS = 0.05g BIS = 0.068g

350

Diameter (nm)

300 250 200 150 100 50 0 15

25

35

45

55

65

LCST (°C) FIGURE 7.5 The effect of temperature on the diameter of the PNIPAAm-MAA particles with different cross-linker contents. (Reprinted from P. Tian, Q. Wu, and K. Lian, Journal of Applied Polymer Science 108: 2226–32, 2008. With permission.)

nanoparticle templates, and separation and puriication technologies. Current research on thermo-responsive nanoparticles mainly focuses on utilizing them as drug carriers. For this purpose, the encapsulation of drugs onto the nanoparticles and the releasing behavior need to be investigated. Core/shell thermo-responsive nanoparticles with pH-sensitive ability of poly(N-isopropylacrylamide-co-N,N-dimethylacrylamide-co-10-undecenoic acid) (PNIPAAm-co-DMAAm-co-UA) were self-assembled from the amphiphilic tercopolymer, studied by Soppimath et al. [105]. Doxorubicin (DOX), an anticancer drug, was encapsulated into the nanoparticles and its release behavior was investigated. It is shown that multifunctional polymer core/shell nanoparticles provide a promising carrier for the releasing of DOX to cancer cells. Cuie and co-workers studied driving forces in the loading and release of protein onto PNIPAAm-based nanoparticles [106]. In their study, bovine serum albumin (BSA) and γ-globulins (γG) were loaded onto various cationically or anionically charged PNIPAAm-based nanoparticles and their release behavior was investigated. It was shown that the loading and release strongly depended on temperature, pH, and salinity. As temperature increased, the loaded amount of both proteins increased. Loaded proteins can be completely released after the temperature is above the LCST and the process is reversible. It is well-known that some ionized polymers can capture multivalent ions in cooperation with several ionized side chains (i.e., chelation) [107–109]. The chelation behavior of gel is very useful in detecting and/or capturing toxic multivalent ions in waste solutions and can be used to design ionized-gel

210

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

sensor, which is sensitive to multivalent metal ions. Furthermore, the ion-capture property of the ionized gel is very useful in environmental puriication technology. Several studies have been reported with stimuli-responsive polymer gels that adsorb metal ions by forming an ionic bond [110–115]. To design and develop a polymer gel that can selectively absorb and release metal ions, the polymer should consist of two monomer groups, each having a different role. One group forms a complex with the target (adsorption part) and the other allows the polymers to stretch and shrink reversibly in response to environmental change (the responsive part). Generally, for hydrogel to adsorb metal ions, the monomer with inherent properties that incorporates onto gelatin acts as a determinant of the water-absorption part, and metal ions are absorbed by effectively partitioning between hydrogels and solution phase. Apart from the nature of metal ions, the structural aspects of hydrogels also determine the quantum of metal ion uptake. The rates of adsorption and desorption were dominated by diffusion within the gel. For the synthesized PNIPAAm-MAA particles, beside their application as a drug delivery vehicle, one of the other applications is for heavy metal adsorption. It has been shown that bivalent ions, such as Cu, Cd, and Zn, can interact with methacrylic acid and form metal-polyacid complexes involving two carboxylate groups for each ion in a large domain of pH values [116,117]. Furthermore, stimuli-responsive polymer with nanostructure is expected to be more eficient in capturing or detecting metal ions because of their much greater contact areas. Another advantage with stimuli-responsive copolymers for target metal adsorption is that potentially it can be a temperature swing process. To our knowledge, there have been very few investigations on the ion-capturing feature of the stimuli-responsive polymer with nanostructure. Such investigations are very important because these systems are stabilized with a balance of the interactions in a nanoscopic scale as mentioned earlier. It is of great practical interest to show how the particle size affects the metal ion adsorption process. The mechanism of the Cu2+ adsorption with PNIPAAm-MAA is shown in Figure 7.6. Cu2+ can chelate with -COOH groups in PNIPAAm-MAA. Generally, a copper cation can chelate with two PNIPAAm-MAA molecules C O

O

Cu O

O

O

O

O C

C O

Cu O

O C

O

O

Cu

Cu O

C

C

C O

O

O C

FIGURE 7.6 Mechanism of the Cu2+ adsorption with PNIPAAm-MAA. (Reprinted from Q. Wu and P. Tian, Journal of Applied Polymer Science 109: 3740–46, 2008. With permission.)

211

Smart Thermo-Responsive Nanoparticles

and two copper cations can further link four or more PNIPAAm-MAA molecules to form an interlocking structure. To test adsorption capacity of PNIPAAm-MAA particles to Cu2+ ion, a sample of 2 ml of polymerized PNIPAAm-MAA gel solution was put into 20 ml of 5mM CuSO4 aqueous solution in a test tube. The adsorption experiments were carried out at two temperatures, one at 5°C (below LCST) and one at 50°C (above LCST). Samples in test tubes were placed for two days for Cu2+ adsorption, which was suficient time for reaching equilibrium. After adsorption, samples were centrifuged to separate the particle and the solution. The pH of the CuSO4 solution was adjusted with Clark-Lubs buffer solution. The initial and equilibrium concentration of Cu2+ ions in solution were measured by an inductively coupled plasma-atomic emission spectrometry (ICP-AES) system [118]. The equilibrium adsorption amount of Cu adsorbed by the particles, which was the net amount of Cu2+ interacted with copolymer, was determined from the mass balance of the initial and equilibrium Cu2+ ion contents in solutions, assuming that the concentrations of the nonreacted Cu2+ ion was retained in the equilibrium solution. The equilibrium adsorption amount, qe (mmol/g-drygel), was determined as follows: qe =

(

Vl Co − Ce Wg

)

(7.1)

where Vl (ml) is the volume of the adsorption solution and Wg (g) is the weight of the corresponding dry gel adsorbent, and Co and Ce (mol/L) are the initial and equilibrium Cu2+ ion contents, respectively. The adsorption eficiency, η, which is based on the actual adsorption amount qe over the theoretical adsorption amount qt, was calculated as: η=

qe × 100% qt

(7.2)

The theoretical adsorption amount, qt, was calculated based on 2 mol of Cu2+ chelating with 4 mol of -COOH group of PNIPAAm-MAA. MAA:NIPAAm ratio dependence for the amount of Cu2+ adsorbed from CuSO4 solution to the PNIPAAm-MAA copolymer particles at pH = 5 is shown in Figure 7.7. It is obvious that the adsorbed amount increased in a stepwise manner with the increase of MAA:NIPAAm ratio. Furthermore, the adsorbed amount at 5°C was slightly higher than that at 50°C. This could be due to the fact that at temperatures higher than its LCST, the copolymer shrunk, decreased the eficient contact area of the MAA unit with Cu2+ ions, and decreased part of the carboxyl groups of MAA in adsorption reaction. For pH change from 5 to 6, Cu2+ adsorption amount increased slightly at

212

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Cu Adsorption Amount (mmol/g-drygel)

0.7 0.6 0.5 0.4 0.3 0.2

5C 50C

0.1 0 0

0.1

0.2

0.3 0.4 0.5 MAA/NIPAAM Ratio

0.6

0.7

0.8

FIGURE 7.7 PNIPAAm-MAA particle Cu2+ adsorption with MAA:NIPAAm ratio at pH = 5. (Reprinted from Q. Wu and P. Tian, Journal of Applied Polymer Science 109: 3740–46, 2008. With permission.).

low MAA ratio. However, for higher MAA ratios (MAA/NIPAAm = 0.1 – 0.7), no apparent trend of Cu2+ adsorption amount with pH was detected. The equilibrium adsorption amount, qe, ranged from 0.1 to 0.7 mmol/g-dry gel, depending on different MAA contents. This is much higher than the adsorption eficiency reported in the literature. For example, the adsorption amount with PNIPAAm-VBEDA ranged from 0.001 to 0.012 mmol/g-dry gel [113], and that of PNIPAAm-MEP ranged from 0.015 to 0.035 mmol/g-drygel [119]. The magniicent increase of adsorption should ascribe to the nanoscale particle size. The PNIPAAm-MAA particle size ranged from 30 to 500 nm, which greatly increased the effective contact area of the MAA and Cu2+ ions. This is a beneicial and important factor of PNIPAAm-MAA particles for its potential application in puriication processes. The particle size effect on the amount of Cu2+ adsorbed to the PNIPAAmMAA copolymer particles from CuSO4 solution at pH = 5 is shown in Figure 7.8. The adsorbed amount at temperature below LCST was higher than that above LCST, as discussed before. For particles with size in the range of 30–300 nm, the amount of Cu2+ adsorbed onto particles showed no signiicant change; the equilibrium adsorption amount was about 0.1 mmol/ g-drygel. For these samples, the MAA:NIPAAm ratio was the same (i.e., MAA/NIPAAm = 0.025). As the particle size increased from 300 to 500 nm, the total adsorbed amount increased. However, this does not necessarily suggest that increasing particle size increased the adsorption amount since the MAA contents in the copolymer particles also increased signiicantly. The adsorption of Cu ions onto PNIPAAm-MAA particles is a balance between the particle size and MAA contents.

213

Smart Thermo-Responsive Nanoparticles

Cu Adsorption Amount (mmol/g-drygel)

0.7 0.6 0.5 0.4 0.3 0.2

5C 50C

0.1 0 250

300

350 400 450 Particle Size at 25°C (nm)

500

550

FIGURE 7.8 PNIPAAm-MAA particle Cu2+ adsorption with particle size at pH = 5. (Reprinted from Q. Wu and P. Tian, Journal of Applied Polymer Science 109: 3740–46, 2008. With permission.)

As mentioned before, particles within 20- to 300-nm size range were synthesized with the same amount of MAA:NIPAAm ratio, but different cross-linker contents or surfactant contents, leading to different particle size. Particles with size range of 300–500 nm represented increasing MAA:NIPAAm ratio with the same cross-linker and SDS contents. As shown in Figure 7.9, with small particle size, the adsorption eficiency was pretty 1 Cu Adsorption Efficiency

0.9

5C 50C

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

250

300

350

400

450

500

550

Particle size at 25ºC (nm) FIGURE 7.9 Adsorption eficiency of Cu2+ with PNIPAAm-MAA with particle size at pH = 5. (Reprinted from Q. Wu and P. Tian., Journal of Applied Polymer Science 109: 3740–46, 2008. With permission.)

214

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

high, most at 80–98% range. With increased particle size, although the absolute adsorption amount per g-dry gel increased, the adsorption eficiency decreased dramatically. This strongly suggested that particle size played an important role in the adsorption of metal ions with the copolymer particles. By controlling surfactant contents, PNIPAAm-MAA particles can be synthesized with higher MAA ratio but smaller particle size, and can be used to help the adsorption eficiency of metal ion adsorption with the copolymer.

7.5 Conclusions and Future Development Thermo-responsive copolymer is a very promising material for varieties of applications based on their response to temperature change. Extensive work on thermoresponsive polymers based on PNIPAAm has been done in the past 20 years. However, only recently has the investigation of the studies moved into micro- and/or nanoscale for this kind of polymer. A number of novel, thermo-sensitive nanoparticles and their synthesis route and property characterizations has been described in the literature. Each system has its own advantages and drawbacks. The choice of a particular nanoparticle and nanogel depends on its intrinsic properties and target application. The main focus of research in thermo-responsive nanoparticles lies in developing speciic nanoparticles with designed properties for proper loading and release of speciic drugs to targeted cells. The synthesis and characterization of the PNIPAAm-based copolymer nanoparticles still need much research effort before they can be widely used as drug delivery vehicles, and how the nanoscale behavior of the particles affects the delivery eficiency also requires much research attention. It is worth noting that current research still mainly focuses on design of the thermo-responsive nanoparticles. For these nanoparticles to be used in actual clinical applications, a cooperative effort needs to be carried out for studying the load and release eficiency of speciic drugs, the toxicity of the polymer nanoparticles, the immunogenicity of the human body to the encapsulated particle-drug, and its release kinetics among chemist, pathologist, and therapist. Besides the intense interest in thermo-responsive nanoparticles’ biomedical application, other applications such as separation and puriication are well worth scientiic study as well since they may supply a swing process for adsorption and desorption.

Smart Thermo-Responsive Nanoparticles

215

References 1. Tanaka, T., Nishio, I., Sun, S.-T., and S. Ueno-Nishio. 1982. Collapse of gels under an electric ield. Science 218:467–69. 2. Osada, Y., Okuzaki, H., and H. Hori. 1992. A polymer gel with electrically driven motility. Nature 355:242–44. 3. Tanaka, T. 1978. Collapse of gels and the critical endpoint. Physical Review Letters 40:820–23. 4. Hirokawa, Y., and T. Tanaka. 1984. Volume phase transition in a nonionic gel. Journal of Chemical Physics 81:6379–80. 5. Amiya, T., Hirokawa, Y., Hirose, Y., Li, Y., and T. Tanaka. 1987. Reentrant phase transition of N-isopropyl acrylamide gels in mixedsolvents. Journal of Chemical Physics 86:2375–79. 6. Chen, G., and A. S. Hoffman. 1995. Graft copolymers that exhibit temperatureinduced phase transition over a wide range of pH. Nature 373:49–52. 7. Irie, M. 1993. Stimuli-responsive poly(N-isopropylacryl-amide). Photo- and chemical-induced phase transitions. Advances in Polymer Science 110:49–65. 8. Suzuki, A., and T. Tanaka. 1990. Phase transition in polymer gels induced by visible light. Nature 346:345–47. 9. Szabo, G., Szeghy, G., and M. Zrinyi. 1998. Shape transition of magnetic ield sensitive polymer gels. Macromolecules 3:6541–48. 10. Tanaka, T., Fillmore, D., Sun, S. T., Nishio, I., Swislow, G., and A. Shah. 1980. Phase transitions in ionic gels. Physical Review Letters 45:1636–39. 11. Higa, M., and T. Yamakawa. 2004. Design and preparation of a novel temperature-responsive ionic gel. 1. A fast and reversible temperature response in the charge density. Journal of Physical Chemistry B 108:16703–7. 12. Maeda, T., Kanda, T., Yonekura, Y., Yamamoto, K., and T. Aoyagi. 2006. Hydroxylated poly(N-isopropylacrylamide) as functional thermo-responsive materials. Biomacromolecules 7:545–49. 13. Murthy, N., Xu, M., Schuck, S., Kunisawa, J., Shastri, N., and M. Frechet. 2003. Macromolecular delivery vehicle for protein-based vaccines: Acid-degradable protein-loaded microgels. J Proc Natl Acad Sci USA 100:4995–5000. 14. Lopez, V. C., and M. J. Snowden. 2003. The role of colloidal microgels in drug delivery. Drug Delivery Systems Science 3:19–23. 15. Cammas, S., Suzuki, K., Sone, S., Sakurai, Y., Kataoka, K., and T. Okano. 1997. Thermo-responsive polymer nanoparticles with a core-shell micelle structure as site-speciic drug carriers. Journal of Controlled Release 48:157–64. 16. Lin, C-L., Chiu, W-Y., and C.-F. Lee. 2005. Preparation of thermoresponsive coreshell copolymer latex with potential use in drug targeting. Journal of Colloid and Interface Science 290:397–405. 17. Ottenbrite, R. M., and S. W. Kim. 2001. Polymeric drugs and drug delivery systems. Boca Raton, FL: CRC Press. 18. Tian, P., Newman L., and R. Toomey. 2006. Peptide-modiied responsive networks with built in logic. Paper presented at the annual meeting of AIChE.

216

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

19. Retama, J. R., Lopez-Ruiz, B., and E. Lopez-Cabarcos. 2003. Microstructural modiications induced by the entrapped glucose oxidase in cross-linked polyacrylamide microgels used as glucose sensors. Annual Reviews of Materials Research 24:2965–73. 20. Guo, Z., Sautereau, H., and D. E. Kranbuehl. 2005. Structural evolution and heterogeneities studied by frequency-dependent dielectric sensing in a styrene/ dimethacrylate network. Macromolecules 38:7992–99. 21. Xu, S., Zhang, J., Paquet, C., Lin, Y., and E. Kumacheva. 2003. From hybrid microgels to photonic crystals. Advanced Functional Materials 13:468–72. 22. Lyon, L. A., Debord, J. D., Debord, S. B., Jones, C. D., McGrath, J. G., and M. J. Serpe. 2004. Microgel colloidal crystals. Journal of Physical Chemistry B 108:19099–19118. 23. Zhang, J., Xu, S., and E. Kumacheva. 2004. Polymer microgels: Reactor for semiconductor, metal, and magnetic nanoparticles. Journal of the American Chemical Society 126:7908–14. 24. Bromberg, L., Temchenko, M., and T. A. Hatton. 2003. Dually responsive microgels from polyether-modiied poly(acrylic acid): Swelling and drug loading. Langmuir 19:8675–84. 25. Tanaka, T., Wang, C., Pande, V., Grosberg, Y., English, A., Masamune, S., Gold, H., Levy, R., and K. King. 1996. Polymer gels that can recognize and recover molecules. Faraday Discussions 102:201–6. 26. Chauhan, G., Kumar, S., Kumari, A., and R. Sharma. 2003. Study on the synthesis, characterization, and sorption of some metal ions on gelatin- and acrylamide-based hydrogels. Journal of Applied Polymer Science 90:3856–71. 27. Mrkic, J., and B. R. Saunders. 2000. Microgel particles as a matrix for polymerization: A study of poly(N-isopropylacrylamide)-poly(N-methylpyrrole) dispersions. Journal of Colloid and Interface Science 222:75–82. 28. Murray, M. J., and M. Snowden. 1995. The preparation, characterisation and application of colloidal microgels. Advances in Colloid and Interface Science 54:73–91. 29. Ilmain, F., Tanaka, T., and E. Kokufuta. 1991. Volume transition in a gel driven by hydrogen bonding. Nature 349:400–1. 30. Taylor, L. D., and L. D. Cerankowski. 1975. Preparation of ilms exhibiting a balanced temperature dependence to permeation by aqueous solution: A study of lower consolute behavior. Journal of Polymer Science, Polymer Chemistry Edition 13:2551–70. 31. Heskins, M., and J. E. Guillet. 1968. Solution properties of poly (N-isopropyl acrylamide). Journal of Macromolecular Science, Chemistry Edition A2:1441–55. 32. Schild, H. G. 1992. Poly(N-isopropylacrylamide): Experiment, theory and application. Progress in Polymer Science 17:163–249. 33. Bae, Y. H., Okano, T., and S. W. Kim. 1990. Temperature dependence of swelling of cross-linked poly(N,N’-alkyl substituted acrylamide) in water. Journal of Polymer Science, Polymer Physics Edition 28:923–36. 34. Shibayama, M., Norisuye, T., and S. Nomura. 1996. Cross-link density dependence of spatial inhomogeneities and dynamic fluctuations of poly(N-isopropylacrylamide) gels. Macromolecules 29:8746–50. 35. Liu H. Y., and X. X. Zhu. 1999. Lower critical solution temperature of N-substituted acrylamide copolymers in aqueous solutions. Polymer 40:6985–90.

Smart Thermo-Responsive Nanoparticles

217

36. Platé, N. A., Lebedeva T. L., and L. I. Valuev. 1999. Lower critical solution temperature in aqueous solutions of N-alkylsubstituted polyacrylamides. Polymer Journal 31:21–27. 37. Idziak, I., Avoce, D., Lessard, D., Gravel, D., and X. X. Zhu. 1999. Thermosensitivity of aqueous solutions of poly(N,N-diethylacrylamide). Macromolecules 32:1260–63. 38. Ataman, M. A. 1987. Properties of aqueous salt solutions of poly(ethyleneoxide), cloud point, θ temperature. Colloid and Polymer Science 265:19–25. 39. Horne, R., Almeida, J. P., Day, A. F., and N. Yu. 1971. Macromolecule hydration and the effect of solutes on the cloud point of aqueous solutions of polyvinyl methyl ether: A possible model for protein denaturation and temperature control in homeothermic animals. Journal of Colloid and Interface Science 35:77–84. 40. Mikheeva, L. M., Grinberg, N. V., Mashkevich, A. Y., Grinberg, V. Y., Thanh, L. T. M., Makhaeva, E. E., and A. R. Khokhlov. 1997. Microcalorimetric study of thermal cooperative transitions in poly(N-vinylcaprolactam) hydrogels. Macromolecules 30:2693–99. 41. van Durme, K., Verbrugghe, S., DuPrez, F. E., and B. van Mele. 2004. Inluence of poly(ethylene oxide) grafts on kinetics of LCST behavior in aqueous poly(Nvinylcaprolactam) solutions and networks studied by modulated temperature DSC. Macromolecules 37:1054–61. 42. Makhaeva, E. E., Tenhu, H., and A. R. Khokhlov. 1998. Conformational changes of poly(vinylcaprolactam) macromolecules and their complexes with ionic surfactants in aqueous solution. Macromolecules 31:6112–18. 43. Zentner, G. M., Rathi, R., Shih, C., McRea, J. C., Seo, M.-H., Oh, H., Rhee, B. G., Mestecky, J., Moldoveanu, Z., Morgan, M., and S. Weitman. 2001. Biodegradable block copolymers for delivery of proteins and water-insoluble drugs. Journal of Controlled Release 72:203–15. 44. Wakamatsu, H., Yamamoto, K., Nakao, A., and T. Aoyagi. 2006. Preparation and characterization of temperature-responsive magnetite nanoparticles conjugated with N-isopropylacrylamide-based functional copolymer. Journal of Magnetism and Magnetic Materials 302:327–33. 45. Suzuki, D., and H. Kawaguchi. 2006. Stimuli-sensitive core/shell template particles for immobilizing inorganic nanoparticles in the core. Colloid and Polymer Science 284:1443–51. 46. Bhattacharya, S., Eckert, F., Boykoand V., and A. Pich. 2007. Temperature-, pH-, and magnetic-ield-sensitive hybrid microgels. Small 3:650–57. 47. Fujishige, S., Kubota, K., and I. Ando. 1989. Phase transition of aqueous solutions of poly2 (N2isopropylacrylamide) and poly-(N-isopropylmethacrylamide). Journal of Physical Chemistry 93:3311–13. 48. Peppas, N. A., Bures, P., Leobandung, W., and H. Ichikawa. 2000. Hydrogels in pharmaceutical formulations. European Journal of Pharmaceutics and Biopharmaceutics 50:27–46. 49. Gupta, P., Vermani, K., and S. Garg. 2002. Hydrogels: From controlled release to pH-responsive drug delivery. Drug Discovery Today 7:569–79. 50. Peppas, N. A., Hilt, J. Z., Khademhosseini, A., and R. Langer. 2006. Hydrogels in biology and medicine: From molecular principles to bionanotechnology. Advanced Materials 18:1345–60. 51. Nayak, S., and L. A. Lyon. 2005. Soft nanotechnology with soft nanoparticles. Angewandte Chemie, International Edition 44:7686–7708.

218

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

52. Oh, J. K., Drumright, R., Siegwart D. J., and K. Matyjaszewski. 2008. The development of microgels/nanogels for drug delivery applications. Progress in Polymer Science 33:448–77. 53. Malmsten, M. 2006. Soft drug delivery systems. Soft Matter 2:760. 54. Yallapu, M. M., Reddy M. K., and V. Labhasetwar. 2007. Biomedical applications of nanotechnology. New York: John Wiley & Sons. 55. Ulanski, P., and J. M. Rosiak. 2004. Polymeric nano/microgels. In Encyclopedia of nanoscience and nanotechnology, ed. H. S. Nalwa, 845–71. Stevenson Ranch, CA: American Scientiic Publishers. 56. Raemdonck, K., Demeester, J., and S. De Smedt. 2009. Advanced nanogel engineering for drug delivery. Soft Matter 5:707–15. 57. Choi, C., Chae, S., and J. Nah. 2006. Thermosensitive poly(Nisopropylacrylamide)-b-poly(3-caprolactone) nanoparticles for eficient drug delivery system. Polymer 47:4571–80. 58. Neradovic, D., Soga, O., Van Nostrum, C. F., and W. E. Hennink. 2004. The effect of the processing and formulation parameters on the size of nanoparticles based on block copolymers of poly(ethylene glycol) and poly(N-isopropylacrylamide) with and without hydrolytically sensitive groups. Biomaterials 25:2409–18. 59. Zeng, Y., and W. Pitt. 2005. A polymeric micelle system with a hydrolysable segment for drug delivery. Journal of Biomaterials Science, Polymer Edition 16:371–80. 60. Kim, K., Kim, J., and W. Jo. 2005. Preparation of hydrogel nanoparticles by atom transfer radical polymerization of N-isopropylacrylamide in aqueous media using PEG macro-initiator. Polymer 46:2836–40. 61. Weng, H., Zhou, J., Tang, L, and Z. Hu. 2004. Tissue responses to thermallyresponsive hydrogel nanoparticles. Journal of Biomaterials Science, Polymer Edition 15:1167–80. 62. Hoffman, A. S. 1991. Environmentally sensitive polymers and hydrogels. MRS Bulletin 42–46. 63. Hoffman, A. S., Afrassiabi, A., and L. C. Dong. 1986. Thermally reversible hydrogels: II. Delivery and selective removal of substances from aqueous solutions. Journal of Controlled Release 4:213–22. 64. Hirokawa, Y., and T. Tanaka. 1984. Volume phase transitions in nonionic gel. Journal of Chemical Physics 81:6379–80. 65. Yoshida, R., Sakai, K., Okano, T., and Y. Sakurai. 1994. Modulating the phase transition temperature and thermosensitivity in N-isopropylacrylamide copolymer gels. Journal of Biomaterials Science, Polymer Edition 6:585–98. 66. Yoshida, R., Uchida, K., Kaneko, Y., Sakai, K., Kikuchi, A., Sakurai, Y., and T. Okano. 1995. Comb-type grafted hydrogels with rapid de-swelling response to temperature changes. Nature 374:240–42. 67. Quaroni, L., and G. Chumanov. 1999. Preparation of polymer-coated functionalized silver nanoparticles. Journal of the American Chemical Society 121:10642–43. 68. Li, D., Jones, G. L., Dunlap, J. R., Hua, F., and B. Zhao. 2006. Thermosensitive hairy hybrid nanoparticles synthesized by surface-initiated atom transfer radical polymerisation. Langmuir 22:3344–51. 69. Huang, H., Remsen, E. E., Kowalewski, T., and K. L. Wooley. 1999. Nanocages derived from shell cross-linked micelle templates. Journal of the American Chemical Society 121:3805–6.

Smart Thermo-Responsive Nanoparticles

219

70. Watson, K. J., Zhu, J., Nguyen, S. T., and C. A. Mirkin. 1999. Hybrid nanoparticles with block copolymer shell structures. Journal of the American Chemical Society 121:462–63. 71. Schneider, G., and G. Decher. 2004. From functional core/shell nanoparticles prepared via layer-by-layer deposition to empty nanospheres. Nano Letters 4:1833–39. 72. Landfester, K. 2006. Synthesis of colloidal particles in miniemulsions. Annual Reviews of Materials Research 36:231–79. 73. Pelton, R. 2000. Temperature-sensitive aqueous microgels. Advances in Colloid and Interface Science 85:1–33. 74. Napier M. E., and J. M. Desimone. 2007. Nanoparticle drug delivery platform. Polymer Reviews 47:321–27. 75. Rolland, J. P., Maynor, B. W., Euliss, L. E., Exner, A. E., Denison, G. M., and J. M. Desimone. 2005. Direct fabrication and harvesting of monodisperse, shape speciic nano-biomaterials. Journal of the American Chemical Society 127:10096–100. 76. Fernandez, V. V. A., Tepale, N., Sanchez-Dıaz, J. C., Mendizabal, E., Puig, J. E., and J. F. A. Soltero. 2006. Thermoresponsive nanostructured poly(N-isopropylacrylamide) hydrogels made via inverse microemulsion polymerization. Colloid and Polymer Science 284:387–95. 77. Sahiner, N., Alb, A., Graves, R., Mandal, T., McPherson, G., Reed, W., and V. John. 2007. Core–shell nanohydrogel structures as tunable delivery systems. Polymer 48:704–11. 78. Choi, C., Chae, S., and J. Nah. 2006. Thermosensitive poly(N-isopropylacrylamide)-b-poly(ε-caprolactone) nanoparticles for eficient drug delivery system. Polymer 47:4571–80. 79. Zeng, Y., and W. Pitt. 2005. Poly(ethylene oxide)-b-poly(N-isopropylacrylamide) nanoparticles with cross-linked cores as drug carriers. Journal of Biomaterials Science, Polymer Edition 16:371–80. 80. Kim, K., Kim, J., and W. Jo. 2005. Preparation of hydrogel nanoparticles by atom transfer radical polymerization of N-isopropylacrylamide in aqueous media using PEG macro-initiator. Polymer 46:2836–40. 81. Kuckling, D., Vo, C. D., Adler, H.-J. P., Voelkel, A., and H. Coelfen H. 2006. Preparation and characterization of photo-cross-linked thermosensitive PNIPAAm nanogels. Macromolecules 39:1585–91. 82. Kuckling, D., Vo, C. D., and S. Wohlrab. 2002. Preparation of nanogels with temperature-responsive core and pH-responsive arms by photo-cross-linking. Langmuir 18:4263–69. 83. Singh, N., and A. Lyon. 2007. Au nanoparticle templated synthesis of pNIPAm nanogels. Chemistry of Materials 19:719–26. 84. Ash, B. J., Siegel, R. W., and L. S. Schadler. 2004. Glass transition temperature behavior of alumina/PMMA nanocomposites. Macromolecules 37:1358–69. 85. Zhang, Y., Luo, S., and S. Liu. 2005. Fabrication of hybrid nanoparticles with thermoresponsive coronas via a self-assembling approach. Macromolecules 38:9813–20. 86. Astruc, D., and M.-C. Daniel. 2004. Gold nanoparticles: Assembly, supramolecular chemistry, quantum-size-related properties, and applications toward biology, catalysis, and nanotechnology. Chemical Reviews 104:293–346. 87. Alivisatos, A. P. 1996. Semiconductor clusters, nanocrystals, and quantum dots. Science 271:933–37.

220

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

88. Gao, H., Yang, W., Min, K., Zha, L., Wang, C., and S. Fu. 2005. Thermosensitive poly(N-isopropylacrylamide) nanocapsules with controlled permeability. Polymer 46:1087–93. 89. Kim, J.-H., and T. R. Lee. 2004. Thermo- and pH-responsive hydrogel-coated gold nanoparticles. Chemistry of Materials 16:3647–51. 90. Qian, F., Cui, F., and C. Yin. 2006. Preparation, characterization and enzyme inhibition of methylmethacrylate copolymer nanoparticles with different hydrophilic polymeric chains. European Polymer Journal 42:1653–61. 91. Chilkotia, A., Drehera, M. R., Meyera, D. E., and D. Raucherb. 2002. Targeted drug delivery by thermally responsive polymers. Advanced Drug Delivery Reviews 54:613–30. 92. Das, M., Zhang, H., and E. Kumacheva. 2006. Microgels: Old materials with new applications. Annual Reviews of Materials Research 36:117–42. 93. Raemdonck, K., Demeester J., and S. DeSmedt. 2009. Advanced nanogel engineering for drug delivery. Soft Matter 5:707–15. 94. Schmaljohann, D. 2006. Thermo- and pH-responsive polymers in drug delivery. Advanced Drug Delivery Reviews 58:1655–70. 95. Zhang, K., and X. Wu. 2004. Temperature and pH-responsive polymeric composite membranes for controlled delivery of proteins and peptides. Biomaterials 25:5281–91. 96. Huang, J., and X. Y. Wu. 1999. Effects of pH, salt, surfactant and composition on phase transition of poly(NIPAm/MAA) nanoparticles. Journal of Polymer Science, Part A: Polymer Chemistry 37:2667–76. 97. Tian, P., Wu, Q., and K. Lian. 2008. Preparation of temperature- and pH-sensitive, stimuli-responsive poly(N-isopropylacrylamide-co-methacrylic acid) nanoparticles. Journal of Applied Polymer Science 108:2226–32. 98. Seno, M., Lin, M. L., and K. Iwamoto. 1991. Swelling of poly(methacrylic acid) gels and adsorption of L-lysine and its polymer on the gels. Colloid and Polymer Science 269:873–79. 99. Shibayama, M., Mizutani, S., and S. Nomura. 1996. Thermal properties of copolymer gels containing N-isopropylacrylamide. Macromolecules 29:2019–24. 100. Schild, H. G. 1992. Poly(N-isopropylacrylamide): Experiment, theory and application. Progress in Polymer Science 17:163–249. 101. Cao, Q., Tian, P., and Q. Wu. 2009. Modeling diameter distributions of poly(Nisopropylacrylamide-co-methacrylic acid) nanoparticles. Journal of Applied Polymer Science 111:2584–89. 102. Jounela, A. J., Pentikäinen, P. J., and A. Sothmann. 1975. Effect of particle size on the bioavailability of digoxin. European Journal of Clinical Pharmacology 8:365–70. 103. Tinke, A. P., Vanhoutte, K., De Maesschalck, R., Verheyen, S., and H. De Winter. 2005. A new approach in the prediction of the dissolution behavior of suspended particles by means of their particle size distribution. Journal of Pharmaceutical and Biomedical Analysis 39:900–7. 104. Neradovic, D., Soga, O., Van Nostrum, C. F., and W. E. Hennink. 2004. The effect of the processing and formulation parameters on the size of nanoparticles based on block copolymers of poly(ethylene glycol) and poly(N-isopropylacrylamide) with and without hydrolytically sensitive groups. Biomaterials 25:2409–18.

Smart Thermo-Responsive Nanoparticles

221

105. Soppimath, K. S., Liu, L.-H., Seow, W. Y., Liu, S.-Q., Powell, R., Chan, P., and Y. Y. Yang. 2007. Multifunctional core/shell nanoparticles self-assembled from pHinduced thermosensitive polymers for targeted intracellular anticancer drug delivery. Advanced Functional Materials 17:355–62. 106. Cuie, Y., Abdelhamid, E., and P. Christian. 2006. Loading and release studies of proteins using poly(N-isopropylacrylamide) based nanogels. Journal of Biomedical Nanotechnology 2:208–16. 107. Haner, M., Henzl, M. T., Raissouni, B., and E. R. Birnbaum. 1984. Synthesis of a new chelating gel: Removal of Ca2+ ions from parvalbumin. Analytical Biochemistry 138:229–34. 108. Eskeland, T. 1977. The effect of various metal ions and chelating agents on the formation of noncovalently and covalently linked IgM polymers. Scandinavian Journal of Immunology 6:87–95. 109. Degeiso, R. C., Donaruma, L. G., and E. A. Tomic. 1965. Preparation and chelating properties of 8-hydroxy-quinoline-formaldehyde polymers. Journal of Applied Polymer Science 9:411–19. 110. Tanaka, T., Wang, C., Pande, V., Grosberg,, Y., English, A., Masamune, S., Gold, H., Levy, R., and K. King. 1996. Faraday Discussions 102:201–6. 111. Chauhan, G., Kumar, S., Kumari, A., and R. Sharma. 2003. Study on the synthesis, characterization, and sorption of some metal ions on gelatin- and acrylamide-based hydrogels. Journal of Applied Polymer Science 90:3856–62. 112. Saitoh, T., Satoh, F., and M. Hiraide. 2003. Concentration of heavy metal ions in water using thermoresponsive chelating polymer. Talanta 61:811–17. 113. Tokuyama, H., Kanazawa, R., and S. Sakohara. 2005. Equilibrium and kinetics for temperature swing adsorption of a target metal on molecular imprinted thermosensitive gel adsorbents. Separation and Puriication Technology 44:152–59. 114. Hara, K., Sugiyama, M., Annaka, M., and Y. Soejima. 2004. Nanostructural characterization of the dehydrated (NIPA/SA+additive ion) gels. Colloidas and Surfaces B: Biointerfaces 38:197–200. 115. Yamashita, K., and T. Nishimura. 2003. Preparation of IPN-type stimuliresponsive heavy-metal-ion adsorbent gel. Polymers for Advanced Technologies 14:189–94. 116. Gregor, H. P., Luttinger, L. B., and E. M. Loebl. 1955. Metal–polyelectrolyte complexes. II. Complexes of copper with cross-linked polyacrylic and polymethacrylic acids. Journal of Physical Chemistry 59:366–68. 117. Geuskens, G., Pauw, A., and C. De David. 1968. Formation de chelates de l’acide polyméthacrylique avec les ions Cu2+ provenant d’une résine échangeuse d’ions—I Etude du mécanisme. European Polymer Journal 4:365–72. 118. Wu, Q., and P. Tian. 2008. Adsorption of Cu2+ ions with poly(N-isopropylacrylamide-co-methacrylic acid) micro/nanoparticles. Journal of Applied Polymer Science 109:3740–46. 119. Tokyyama H., Yanagawa K., and S. Sakohara. 2006. Temperature swing adsorption of heavy metals on novel phosphate-type adsorbents using thermosensitive gels and/or polymers. Separation and Puriication Technology 50:8–14.

8 Surface Tailoring of Polymer Nanoparticles with Living Polymerization Methods Koji Ishizu and Dong Hoon Lee CONTENTS 8.1 Introduction ................................................................................................223 8.2 Architecture of Polymer Particles Possessing Radical Initiating Sites on the Surface by Emulsion Copolymerization and Synthesis of Core-Brush Structures by Photoinduced ATRP Approach .....................................................................................................225 8.3 Synthesis of Silica Particles Coated with High-Density Polymer Brushes ........................................................................................................ 228 8.4 Synthesis of Polymer Brushes Encapsulated Silica Particles by DC-Mediated Living Radical Polymerization ....................................... 230 8.5 Synthesis of Silica Hybrid Nanoparticles Modiied with Photofunctional Polymers and Construction of Colloidal Crystals ... 235 8.6 Architecture of Polymer Particles Composed of Brush Structure at Surface and Application for Structural Color Materials .................. 239 8.7 Surface Modiication of Polymer Particles via RAFT Polymerization ........................................................................................... 249 8.8 Surface Modiication of Silica Nanoparticles via NitroxideMediated Radical Polymerization ........................................................... 253 8.9 Conclusions................................................................................................. 255 References............................................................................................................. 255

8.1 Introduction Latex particles have exceptional uniform shape and wide diameter range, which is of interest because their combination with other inorganic (or metallic) colloid particles should extend their application in the ield of academic research as well as industrial material development. This is especially true of the synthesis of composite particles comprised of organic and inorganic substances, in which the adoption of monodisperse polymer lattices as their one component largely expands the variety of composite particles because many 223

224

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

techniques to prepare latex samples with different sizes and structures have been developed [1]. Control of polymer particle size and its uniformity has been a major area of interest especially for particles in the micron size range. Micron-size monodisperse latex particles have been prepared by seeded emulsion poly merization, by either the successive seeding method [2] or the two-stage swelling method [3,4]. They have also been prepared by dispersion polymerization in organic media [5–8]. The dispersion polymerization method has recently received great attention because of the simplicity of the process and the wide variety of monomers that can be successfully polymerized. On the other hand, core-brush-type nanoparticles were prepared by cross-linking of core domains in block copolymers [9]. In general, block and graft copolymers with incompatible sequences exhibit characteristic morphological behavior and interesting properties, owing to micelle formation in a selective solvent and microdomain in the solid state. Many types of micelles formed by diblock copolymers in solution were reported. Spherical self-assembly formed in the solvent for one sequence and not for the other forms a so-called polymer micelle with core-corona morphology. As a result, the core-brush-type polymer particles were prepared easily by cross-linking of the spherical parts (the spherical microdomains in the solid state and the core in solution). The microphase-separated structures in bulk ilm are more stable in thermal equilibrium than the micelle in solution. So the cross-linking of spherical microdomains in the ilm is a superior preparative method for core-brush type polymer particles than cross-linking of micellar cores. Recently, there has been rapid growth in the number of techniques employed in the area of controlled/living radical polymerization, such as atom transfer radical polymerization (ATRP) [10,11], nitroxide-mediated radical [12], and reversible addition-fragmentation chain transfer (RAFT) polymerization [13]. Until recently, ionic polymerizations (anionic or cationic) were the only living techniques available that eficiently controlled the structure and architecture of vinyl polymers. Although these techniques ensure low-polydispersity materials, controlled molecular weight, and deined chain ends, they are not useful for the polymerization and copolymerization of a wide range of functionalized vinyl monomers. On the other hand, the photochemical reactions and initiation polymerization of N,N-diethyldithiocarbamate (DC) derivatives have been extensively studied by Otsu [14]. Generally, this type is used as an inifeter. More recently, we demonstrated that density functional theory calculations provide a prediction of the trend in C-S bond dissociation energies and atom spin densities for radicals using model compounds as DC-mediated inifeters [15–17]. Such photopolymerizations proceeded with a living fashion. We also presented a novel route to hyperbranched polystyrene from 4-vinylbenzyl N,N-diethyldithiocarbamate (VBDC) as an inimer by one-pot photopolymerization [18,19]. Photoinduced ATRP of multifunctional polymers having DC pendant groups with vinyl monomers (“grafting from” approach) formed nanocylinders consisting of graft-block

Surface Tailoring of Polymer Nanoparticles

225

copolymers [20]. More recently, Yamago et al. [21] reported several new organotellurium-based initiators for controlled/living radical polymerization of vinyl monomers that allowed accurate molecular weight control with deined end groups. These living radical polymerizations can be applied for the architecture of various types of nanostructured polymers. This chapter deals mainly with surface modiications of inorganic and organic particles with living radical polymerization methods.

8.2 Architecture of Polymer Particles Possessing Radical Initiating Sites on the Surface by Emulsion Copoly mer i zation and Synthesis of Core-Brush Structures by Photoinduced ATRP Approach One [22–24] of the authors and Dai et al. [25] developed a convenient method for preparing cross-linked core-brush particles by free radical dispersion copolymerization with hydrophobic macromonomers in nonaqueous media. On the other hand, there are several reports concerning the synthesis of corebrush (or shell) particles [26–33]. In particular, Rühe et al. could demonstrate that photoinitiators chemically attached to the surface allow us to obtain planar brushes with deined structure [28]. As mentioned in the introduction, we presented a new method for hyperbranched polymer synthesis by DC-mediated living radical polymerization [18]. These hyperbranched polymers exhibited photofunctional DC groups on their surface. On the basis of these studies, cross-linked polystyrene (PS) particles possessing photofunctional DC groups on their surface were synthesized by free radical emulsion copolymerization of a mixture of styrene (St), divinylbenzene (DVB), and VBDC with redox system as an initiator under ultraviolet (UV) irradiation [34]. In short, the monomer droplets (VBDC, DVB, and St) were dispersed in the aqueous phase by means of surfactants (dodecylbenzenesulfonic acid sodium salt [SDBS] and 1-hexadecanol [HAD]), and a stable emulsion was produced. A water-soluble initiator potassium persulfate (K2S2O8), and one droplet of N,N-dimethylaniline (redox system) was used so that the initial locus of the polymerization was in the aqueous phase. This radical mechanism proceeded with a quasi-living radical nature because the initiating sites were very few. At the same time, the photolysis of inimer VBDC within monomer droplets led to the initiating benzyl radical with the inactive DC radical. This benzyl radical could add to the vinyl group of St, DVB, or VBDC. By repeating these elementary reactions, this polymerization system proceeded to form the hyperbranched structures. Then, this reaction system provided highly localized DC groups attached chemically to the surface. The copolymerization conditions and results for PS-MI1 to PS-MI4

226

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 8.1 Emulsion Copoly merization Conditions and Results for PS-MI1 to PS-MI4a Surfactant (wt% for the Total Monomer) Experiment PS-MI1 PS-MI2 PS-MI3 PS-MI4

SDBS

HAD

Dn (nm)b

Dw/Dnb

1 3 6 12

2 6 12 24

523 420 230 214

1.05 1.10 1.10 1.08

Note: From K. Ishizu, N. Kobayakawa, S. Takano, Y. Tokuno, and M. Ozawa, Journal of Polymer Science, Part A: Polymer Chemistry 45: 1771–77, 2007. With permission. a The emulsion copolymerizations were carried out at room temperature for 8 h of UV irradiation in feeds of [St] = 16 × 10–3 mol/L, [DVB] = 7.7 × 10–2 mol/L, [VBDC] = 6.3 × 10–2 mol/L, and [K2S2O8] = 3 wt% for the total monomer (one droplet of N,N-dimethylaniline). b Determined by a survey of 300 samples packed from SEM photographs.

are listed in Table 8.1. Typical scanning electron microscopy (SEM) photographs of the copolymerization products PS-MI1 and PS-MI4 are shown in Figure 8.1a and b, respectively. The products provided spherical particles (particle diameter Dn = 523 nm for PS-MI1 and Dn = 214 nm for PS-MI4) and relatively narrow size distributions (Dw/Dn = 1.05 for PS-MI1 and Dw/Dn = 1.08 for PS-MI4). Table 8.1 shows that the particle diameter decreased gradually with a decreasing feed ratio of the surfactants to the total monomer. This was a reasonable result for an emulsion polymerization system. The graft polymerization of monomer methyl methacrylate (MMA) initiated by the PS-MI macroinitiator under UV irradiation often led to macrogelation due to intermolecular radical couplings. The main reasons are (1) a highly localized radical concentration on the PS particle surface and (2) a high propagation rate of PMMA radicals. Then, we employed photoinduced ATRP for such a grafting process because the propagation rate, especially of MMA with copper(I) chloride (CuCl)/2,2’-bipyridine (bpy), was very slow compared with that without CuCl/bpy. The core-brush particles were synthesized by the photoinduced ATRP of MMA monomer initiated by PS-MI2 as a macroinitiator in 8/2 (v/v) tetrahydrofuran (THF)/methanol (MeOH) for 8 h of UV irradiation under the following conditions: a monomer concentration of 66.7 wt% and a [DC]/[MMA]:[CuCl]/[bpy] ratio of 1:300:2:6, where [DC] indicates the DC concentration of the PS-MI2 macroinitiator. The reaction scheme for core-brush particle synthesis is shown in Scheme 8.1. In the size distributions for dynamic light scattering (DLS) data of PS-MI2 and CS-2 core-brush particles in THF, both proiles showed unimodal distributions, and after the grafting of the MMA monomer, the hydrodynamic

227

Surface Tailoring of Polymer Nanoparticles

(a)

(b)

1.0 µm

2.5 µm

(c)

1.5 µm

FIGURE 8.1 SEM photographs of (a) PS-MI1 cross-linked PS particles, (b) PS-MI4 cross-linked PS particles, and (c) CS-2 core-brush particles. (Reprinted from K. Ishizu, N. Kobayakawa, S. Takano, Y. Tokuno, and M. Ozawa, Journal of Polymer Science, Part A: Polymer Chemistry 45: 1771–77, 2007. With permission.)

diameter (Dh) increased (to 477 nm) compared with that (Dh = 411 nm) of the PS-MI2 macroinitiator. On the other hand, a typical SEM photograph of the CS-2 core-brush particles is shown in Figure 8.1c. Spherical particles (Dn = 450 nm) are clearly visible for the sample. This result supports the solution data on DLS. The synthesis of the core-brush particles would result in the continuing growth of PMMA chains within PS beads. This is a possibility because THF and MMA monomer swell the beads extensively, and consequent PMMA chain on growth is highly probable and also may cause an increase in the particle size.

228

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

diethyldithiocarbamate group Et CH2

S

C S

UV

N Et

CH2•

Et

•S

C

N Et

S

CuCl

bpy

CH3 H C C H C OCH3 O

MMA

CH2

Et S

C S

N Et

SCHEME 8.1 Schematic illustration for the synthesis of core-shell particles by grafting from the approach of photo-initiated ATRP. (Reprinted from K. Ishizu, N. Kobayakawa, S. Takano, Y. Tokuno, and M. Ozawa. 2007. Journal of Polymer Science, Part A: Polymer Chemistry 45: 1771–77, 2007. With permission).

8.3 Synthesis of Silica Particles Coated with High-Density Polymer Brushes Silica (SiO2) nanoparticles received recent attention because of their superior properties over the microsize particles. The silica surface consists of two types of functional groups, siloxane (Si-O-Si) and silanol (Si-OH) [35,36], which provide functionalization with different functional groups. Thus, silica modiication can occur via the reaction of particular molecules with either the siloxane or the silanol. Modiication of the surface is mostly done by using an appropriate molecule designated as a precursor silane coupling agent. A covalent bond can be formed between the silica surface and the silane coupling agent to give a new modiied silica surface with an anchored functionality. Strategies have been developed to tailor silica particle surfaces with polymers by surface-initiated living radical polymerization techniques, in particular, ATRP [37–50]. Patten et al. irst succeeded in surface-initiated ATRP of styrene and MMA on silica particles with an average diameter of

229

Surface Tailoring of Polymer Nanoparticles

CH2=CH(CH2)4OH

+

BrOC(CH2)2Br

TEA

CH2=CH(CH2)4OOC(CH3)2Br

In THF

BPH (CH3CH2O)3SiH Karstedt’s cat. in toluene

SiO2

(CH3CH2O)3Si(CH2)6OOC(CH3)2)Br BHE NH4OH cat. in EtOH

SiO2

MMA CuCl/Nbipy

PMMA SiO2

70°C O O Si(CH2)6OOC(CH3)2Br O

PMMA

SCHEME 8.2 Schematic representation for the synthesis of poly mer-coated silica particle by surface-initiated ATRP. (Reprinted from K. Ohno, T. Morinaga, K. Koh, Y. Tsujii, and T. Fukuda, Macromolecules 38: 2137–42, 2005. With permission.)

75 and 300 nm [37,38]. Matyjaszewski et al. synthesized an initiator-functionalized silica particle with a diameter of about 20 nm via the sol-gel chemistry and grafted homo and block polymers on the silica surface by ATRP [40]. Carrot et al. used a commercially available silica particle with an even smaller average diameter of 12 nm and carried out ATRP of n-butyl acrylate from its surface [42]. Müller et al. synthesized a hybrid particle with a silica core and the shell of a hyperbranched polymer by surface-initiated ATRP initiating group-holding acrylic monomer [43]. In search for a better route to the modiication of silica particle by surfaceinitiated ATRP, Fukuda et al. synthesized a new triethoxysilane derivative to introduce ATRP initiation sites onto silica surfaces without causing any aggregation of the particles [51]. Scheme 8.2 shows the schematic representation for the synthesis of polymer-coated silica particle by surface-initiated ATRP. The surface initiator, (2-bromo-2-methyl)pro-pionyloxyhexyltriethoxysilane (BHE), was synthesized via a two-step reaction. The irst step was the reaction of 5-hexene-1-ol with 2-bromoisobutyryl bromide provided 1-(2-bromo-2-methyl)propionyloxy-5-hexene (BPH). The second step was the reaction of BPH with triethoxysilane provided the initiator BHE using Karstedt’s catalyst. Next, the reaction of silica particle suspension in ethanol (EtOH) with BHE using NH4OH catalyst provided the initiator-coated SiO2 particle. The surface-initiated ATRP of MMA mediated by CuCl/4,4’-dionyl2,2’-bipyridine (dNbipy) complex was carried out with the initiator-ixed SiO2 particle in the presence of a “sacriicial” (free) initiator. The polymerization proceeded in a living manner in all examined cases, producing SiO2 particle coated with well-deined PMMA of a target molecular weight up to 480K with a graft density as high as 0.65 chains/nm2. These hybrid particles had an exceptionally good dispersibility in organic solvents.

230

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

On the same analogy of the aforementioned works, gold nanoparticles coated with well-deined high-density polymer brushes were synthesized by the surface-initiated ATRP approach [52,53]. For example, the gold nanoparticle (AuNP) coated with an initiator group for living radical polymerization was prepared by the simple one-pot reduction of tetrachloroaurate with sodium borohydrate in the presence of an initiator group-holding disulide. The surface-initiated ATRP of MMA mediated by a copper complex was carried out with the initiator-coated AuNP [53]. Living polymerization proceeded exhibiting a irst-order kinetics of monomer consumption and an evolution of molecular weight of the graft polymer proportional to monomer conversion, thus providing well-deined, low-polydispersity graft polymers with an approximate graft density of 0.3 chain/nm2.

8.4 Synthesis of Polymer Brushes Encapsulated Silica Particles by DC-Mediated Living Radical Polymerization As mentioned in Section 8.2, we established a new synthetic method for cross-linked PS particles possessing photofunctional DC groups on their surface by the free radical emulsion copolymerization. We expected that this method would be applied for the synthesis of PMMA brush encapsulated silica (SPM) particles [54]. The reaction scheme for the synthesis of the SPM particles is shown in Scheme 8.3. The synthesis of well-deined core-shell structures by the method devised here required the architecture of the crosslinked PS shell parts to possess as many photofunctional DC groups as possible on the surface. In this study, the monomer droplets (St, DVB, VBDC, and 2-hydroxyethyl methacrylate [HEMA])-encapsulated SiO2 particles (diameter Dn = 20 nm) were dispersed in the aqueous phases by means of a surfactant (SDBS), and a stable emulsion was produced. A small amount of HEMA that was miscible to the SiO2 surface was also used to achieve the initial formation of the random copolymer composed of rich poly(2-hydroxyethyl methacrylate) (PHEMA) sequences because of the rapid propagation rate constant. An organic-soluble initiator [2,2’-azobis(4-methoxy-2,4-dimethylvaleronitrile): V-70] was used so that effective initiation occurred in the monomer droplets. At the same time, the photolysis of the inimer VBDC within the monomer droplets led to the initiating benzyl radical with the inactive DC radical. This benzyl radical could add to the vinyl groups of St, DVB, HEMA, or VBDC. By repeating these elementary reactions, this polymerization system proceeded to form the hyperbranched structures. The copolymerization conditions and results for the SPS macroinitiator particles are listed in Table 8.2. The SPS1 and SPS2 particles were polymerized in the water phase only. A typical SEM photograph of the copolymerization product SPS2 is shown in Figure 8.2a.

SiO2

SDBS +

H2O/EtOH

+

+

Styreme

VBDS

CH2

C

O

V-70 hv

OCH2CH2OH

CH2DC Silica particle

C

DVB

HEMA CH3

DC

DC

DC DC

SiO2

DC

DC

MMA CuCl, bpy THF, hv

DC

DC DC

SiO2

DC

DC

DC

DC

SiO2-cross-linked PS particle (SPS)

PMMA brushes encapsulated SiO2 particle (SPM)

(CH2 CH)a (CH2 CH)b (CH2 CH)c (CH2 C)d DC

Surface Tailoring of Polymer Nanoparticles

DC

DC DC

CH3

C O OCH2CH2OH CH2DC

(CH2

CH)a

DC

SCHEME 8.3 Reaction scheme for the synthesis of SPM particles; VBDC is 4-vinylbenzyl N,N-diethyldithiocarbamate (Reprinted from K. Ishizu, D. H. Lee, Y. Tokuno, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 109: 3968–74, 2008. With permission.).

231

232

Emulsion Copoly merization Conditions and Results for SiO2-Cross-Linked PS Particles (SPS)a Feed of Monomer Code

SiO2b (mg)

St (mL)

VBDC (mg)

DVB (mL)

HEMA (mL)

SDBS (mg)

V-70 (mg)

Solvent (mL/mL)

Total Momonerc Conversion (%)

Dnd (nm)

Dw/Dnd

SPS1 SPS2 SPS3 SPS4

140 140 140 140

1.2 1.2 0.3 0.6

400 400 100 300

0.2 0.2 0.05 0.1

0.2 0.2 0.05 0.1

200 300 278 278

50 50 50 50

H2O (30) H2O (30) H2O (45)/EtOH (5) H2O (45)/EtOH (5)

56 55 45 52

55 50 35 40

1.20 1.16 1.04 1.05

Note: From K. Ishizu, D. H. Lee, Y. Tokuno, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 109: 3968–74, 2008. With permission. a Polymerized under UV irradiation at room temperature for 4 h in nitrogen atmosphere. b Particle diameter D = 20 nm. n c Determined by gravimetric method. d Determined by a survey of 300 samples picked from SEM photographs.

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 8.2

233

Surface Tailoring of Polymer Nanoparticles

(a)

(b)

100 nm

100 nm (c)

100 nm FIGURE 8.2 SEM photographs of the (a) SPS2, (b) SPS3, and (c) SPM1 particles. (Reprinted from K. Ishizu, D. H. Lee, Y. Tokuno, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 109: 3968–74, 2008. With permission.)

The products provided large spherical particles (Dn = 50 nm) and a broad size distribution (weight-average particle diameter/number-average particle diameter [Dw/Dn = 1.16]). The result for SPS1 showed a similar trend. It was also found from both experiments that the particle diameter decreased gradually with increasing feed ratio of the surfactant SDBS to the total monomer. The thickness of cross-linked PS shell parts was too large because the particle diameter of the SiO2 core used in this study was 20 nm. Therefore, the SPS3 and SPS4 experiments were carried out with a small feed amount of total monomer in the water/EtOH phase. We expected that the addition of EtOH would promote uniform emulsion droplets in miscible water/monomer and the effect of the phase-transfer catalyst. A typical SEM photograph of SPS3 is shown in Figure 8.2b. From this texture, small spherical particles (Dn = 35 nm) were clearly visible and exhibited a narrow size distribution (Dw/Dn = 1.04). Figure 8.3 shows the size distribution for the DLS data of SPS3 in an emulsion aqueous solution. The proile showed unimodal distribution, and the hydrodynamic diameter (Dh) was 38 nm. This value was well in agreement with that observed from SEM. The result for the SPS4 particles showed almost the same trend. The content of DC groups for the SPS particles was determined with the radical transfer reaction. For example, SPS3 was reduced with a lower excess

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Size Distribution (LS%)

234

SPS3 SPM1

1

10

100 Dh (nm)

1000

104

FIGURE 8.3 Size distributions of the DLS data for the SPS3 particles in an emulsion aqueous and the SPM1 particles in THF. (Reprinted from K. Ishizu, D. H. Lee, Y. Tokuno, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 109: 3968–74, 2008. With permission.)

of tributyltin hydride (Bu3SnH) under 0.5–1.0 h of UV irradiation in THF. The consumed amount of Bu3SnH was constant for the previous irradiation times from gas chromatography with decahydronaphthalene as an internal standard sample. The DC groups were estimated to be 1785 number/particle. The formation mechanism of the SPS macroinitiator particles can be speculated as follows. PHEMA-co-PS (PVBDC or PDVB) copolymer was formed in the initial stage because the propagation rate constant of HEMA was very rapid. These copolymers covered the surface of the SiO2 particles because of good afinity between the PHEMA units and silanol (Si-OH) groups on the SiO2 particles. Subsequently, the network and hyperbranched structures at the shell part were generated during polymerization. The grafting from photoinduced ATRP conditions and results for the SPM particles are listed in Table 8.3. The monomer conversion for SPM1 was 24% after 8 h of UV irradiation. A typical SEM photograph of the SPM1 corebrush particles is shown in Figure 8.2c. Spherical particles (Dn = 65 nm) are clearly visible for the sample. This result supports the solution data from DLS. That is, the proile shows unimodal distribution, and Dh was 95 nm (see SPM1 in Figure 8.3). This means that the PMMA brush chains expanded in the THF solution because THF is a good solvent for PMMA. To determine the composition of core, cross-linked PS shell, and PMMA brushes, we

235

Surface Tailoring of Polymer Nanoparticles

TABLE 8.3 Grafting from Photoinduced ATRP Conditions and Results for PMMA Brushes Encapsulated SiO2 Particles (SPM)a Code

SPS3 (mg)

[DC]:[CuCl]:[bpy]:[MMA]

Time (h)

Conversionb (%)

Dnc (nm)

Dw /Dnc

Dhd (nm)

SPM1 SPM2

215 80

1:1.1:2.1:40 1:1:2:80

8 4

24 18

65 56

1.03 1.04

95 80

Note: From K. Ishizu, D. H. Lee, Y. Tokuno, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 109: 3968–74, 2008. With permission. a Polymerized in THF (60 vol% monomer concentration) under UV irradiation at room temperature in high vacuum. b Determined by gravimetric method. c Determined by a survey of 300 samples picked from SEM photographs. d Hydrodynamic diameter was determined by DLS in THF at 25°C.

also performed thermogravimetric analysis (TGA) measurements for SPS3 and SPM1. As a result, the volume ratio SiO2/PS shell:PMMA was estimated to be 1:2.43:11.04. The value of the radius of SPM1 (r3) was determined to be 24.9 nm. This value was in relative agreement with that observed from the SEM image (SPM1, Dn = 65 nm). The thickness of the PMMA brush phase was estimated to be 9.8 nm, in the solid state. Then, the molecular weight of the PMMA brush chain was estimated to be 1.85 × 104, with the assumption that all of the DC initiation sites on the macroinitiator led to the propagation of MMA.

8.5 Synthesis of Silica Hybrid Nanoparticles Modified with Photofunctional Polymers and Construction of Colloidal Crystals More recently, Park and co-workers [55] have prepared PMMA-SiO2 core-shell nanocomposite particles from the dispersion polymerization in supercritical carbon dioxide. In their approach, 3-(trimethoxysilyl)propyl methacrylatefunctionalized SiO2 were irst dispersed in the reaction medium followed by the polymerization with MMA. The polymer-grafted SiO2 was also prepared as reported previously [56,57]. For example, the PMMA-grafted SiO2 particles were synthesized as follows [57]: (1) radical polymerization of MMA was carried out in the presence of (3-mercaptopropyl)-trimethoxysilane (MPMS) as chain transfer agent to obtain PMMA-Si(OMe)3; (2) the PMMA-grafted SiO2 was prepared by coupling of SiO2 particles with PMMA-Si(OMe)3. Moreover, colloidal crystals formed by PMMA-grafted SiO2 particles were immobilized by a procedure consisting of gelation by radical copolymerization. We also synthesized the silica hybrid nanoparticles (diameter Dn = 192 nm) modiied with photofunctional polymer or monofunctional silane coupling

236

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

O SH O Si O (MPMS)

V-70 O

+ OMe

SCNEt2

O O Si O

S

in THF, 30°C, 24h, under dark room

(

( (

O

OMe

(H

SCNEt2

S MMA

VBDC

S

PFD

SCHEME 8.4 Polymerization scheme for the synthesis of photofunctional polymers (PFD) as silane coupling agent. (Reprinted from K. Ishizu, Y. Tokuno, D. H. Lee, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 112: 2434–40, 2009. With permission.)

agents [58]. Scheme 8.4 shows the reaction scheme for the synthesis of photofunctional polymer as silane coupling agent (PFD). Such PFD was prepared by free radical copolymerization of VBDC and MMA in the presence of MPMS as chain transfer agent. Next, SiO2 nanoparticles were surface-modiied with PFD and 3-(trimethoxysilyl)propyl methacrylate (γ-MPS) by covalent bond formed between silanol groups and silane coupling agents (see Figure 8.4). The PFD and γ-MPS functionalizations changed the silica surface into hydrophobic nature and provided grafting initiation sites and methacrylate terminal groups, respectively (SiO2 hybrid nanoparticles ([H-SiO2]). A typical SEM photograph of H-SiO2 hybrid particles is shown in Figure 8.5b, where Figure 8.5a shows an SEM photograph of starting SiO2 particles (Dn = 192 nm, Dw/Dn = 1.01 ± 0.05). The product H-SiO2 provided somewhat large spherical particles (Dn = 198 nm) and kept narrow size distribution (Dw/Dn = 1.01 ± 0.06). Figure 8.6 shows size distribution (Dh, hydrodynamic diameter) on DLS data of H-SiO2 in THF and starting SiO2 particles in water. Both proiles show unimodal distribution. Because H-SiO2 particles were stabilized sterically with both PMMA grafted chains and γ-MPS fragments, these hybrid particles formed a single molecule in THF. Then, the hydrodynamic diameter of H-SiO2 showed a larger value (Dh = 243 nm) than that of starting SiO2 particles (Dh = 209 nm) in solution. DC

DC DC SiO2

+

O O O

O

Si

O

+

O

DC

DC

DC Si

O

in Toluene, 30°C, 200h, under dark room

SiO2

O

O O

γ-MPS

PFD

H-SiO2

FIGURE 8.4 Schematic illustration for the synthesis of SiO2 hybrid nanoparticles (H-SiO2). (Reprinted from K. Ishizu, Y. Tokuno, D. H. Lee, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 112: 2434–40, 2009. With permission.)

Surface Tailoring of Polymer Nanoparticles

237

(a)

400 nm

(b)

400 nm

FIGURE 8.5 SEM photographs of (a) SiO2 and (b) SiO2 hybrid nanoparticles (H-SiO2). (Reprinted from K. Ishizu, Y. Tokuno, D. H. Lee, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 112: 2434–40, 2009. With permission.)

To determine the composition of H-SiO2 particles, we performed TGA measurements for H-SiO2 particles. A 6.6 wt% mass increase was observed after silane coupling reaction. Assuming that this organic layer is composed of PFD, the number (NPFD) of grafted PFD on SiO2 surface is calculated to be 6.09 × 104 mol/one H-SiO2 particle, that is, (4/3)πr13 dSiO2 Nav(6.6/93.4Mn,PFD), where r1 (96 nm), dSiO2 (1.74 g/cm3), Nav (6.025 × 1023 mol–1), and Mn,PFD (4500) are radius of SiO2, density of SiO2, Avogadro’s number, and molecular weight of PFD silane coupling agent, respectively. Then, the number of DC groups on H-SiO2 particle surface was estimated to be 3.2 mol/nm2, that is, 6.4N/4πr22, where r2 (99 nm) is the radius of H-SiO2 hybrid particles. The content of DC groups for H-SiO2 hybrid particles was determined by using the radical transfer reaction. H-SiO2 (22.2 mg) was reduced with a less

238

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

SiO2 in Water (Dn = 209 nm)

Intensity (LS%)

H-SiO2 in THF (Dn = 243 nm)

10

102

103

104

Hydrodynamic Diameter (nm) FIGURE 8.6 Size distribution of DLS data for SiO2 in water and H-SiO2 in THF. (Reprinted from K. Ishizu, Y. Tokuno, D. H. Lee, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 112: 2434–40, 2009. With permission.)

excess of Bu3SnH under 30 min of UV irradiation in THF. The consumed amount of Bu3SnH (9.85 × 10 –7 mol) was constant for the aforementioned irradiation times from gas chromatography using decahydronaphthalene as the internal standard sample. The particle number of feed H-SiO2 is calculated to be 5.339 × 10 –12 using the following equation: (22.2 × 10 –3 × 0.934) ÷ [(4/3) πr13 dSiO2 Nav], where r1 = 96 nm, dSiO2 = 1.74 g/cm3, and Nav = 6.025 × 1023 mol–1. Then, the number of DC groups per particle can be estimated to be 1.845 × 105 mol. Finally, the number of DC groups on H-SiO2 particle surface was estimated to be 1.50 mol/nm2, that is, 1.845 × 105/4πr22, where r2 = 99 nm. This value is smaller than that (3.2 mol/nm2) calculated from TGA data because the organic shell consisted of PFD and γ−MPS silane coupling fragments. The number of grafted γ−MPS fragments (Nγ-MPS) can be estimated as follows. The molecular weight (MH-SiO2 ) of H-SiO2 hybrid particle is given by the relation: MH-SiO2 = MSiO2 + Mshell, where MSiO2 and Mshell are the molecular weights of SiO2 core and organic shell part, respectively. From TGA data, the ratio of MSiO2 to Mshell is given by MSiO2/Mshell = 93.4/6.6. Then, Mshell is estimated to be 2.744 × 108 using the relation: MSiO2 = (4/3)πr13 dSiO2 Nav. This organic shell is composed of a mixture of PFD and γ−MPS fragments. Then,

Surface Tailoring of Polymer Nanoparticles

239

the mass of PFD grafted chains can be estimated to be 1.296 × 106. So, the mass of grafted γ−MPS fragments is estimated to be 1.448 × 108. Therefore, the number of grafted γ−MPS fragments was calculated to be 5.83 × 105 mol/ one particle. We performed the construction of hybrid nanocomposites by using these modiied SiO2 nanoparticles [58]. The MMA solution of H-SiO2 (H-SiO2; 84 wt%) was poured into a petri dish. This weight percent was determined from calculation to give a body-centered cubic (BCC) packing structure. The H-SiO2 was dispersed in MMA under ultrasonic irradiation before poly merization. Radical photopolymerization was carried out in nitrogen atmosphere at 30°C under UV irradiation. This DC-mediated radical poly merization proceeded with living radical mechanism as proven previously by kinitic approaches [20,59]. All the polymerization products provided transparent ilms and exhibited opal-like color. Such prepared samples were then immersed in hydroluoric acid (HF) solution to etch away SiO2 particles, yielding nanoporous PMMA ilms. Figure 8.7a shows an SEM photograph of a vertical section of composite ilm after etching SiO2 particles. The SEM image shows a matrix with quite uniform nanopore voids. On the other hand, Figure 8.7b shows an enlarged image of nanoporous ilms. The average pore size was about 190 nm, which agrees well with the size of colloidal SiO2 particles. The nearest-neighbor distance Ds of SiO2 particles was about 204 nm, which closes relatively to the calculated value (243 nm) within experimental errors considering the effects of the volume shrinkage of PMMA matrix and the cutting angle of a vertical section of composite ilm. This texture indicates that SiO2 particles are locked in a state of molecular dispersion in a PMMA matrix, but that the two-dimensional long-range order such as colloidal crystals is not perfectly maintained. The photofunctional SiO2 particles in this system undergo graft polymerization radially to form hybrid materials.

8.6 Architecture of Polymer Particles Composed of Brush Structure at Surface and Application for Structural Color Materials The unique ability of photonic crystals to manipulate the transmission of light may lead to potential applications ranging from simple optical switches to an optical computer. Opals are naturally occurring three-dimensional (3D) photonic crystals: their microstructure consists of SiO2 spheres of about 150–300 nm in diameter, which are tightly packed into repeating hexagonal or cubic arrangements [60]. Synthetic opals (colloidal crystals) use this same pattern, although they can be made from different materials. In the former section, we mentioned the construction method of colloidal crystals

240

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(a)

1 µm

(b)

500 nm

FIGURE 8.7 SEM photographs of (a) vertical section of composite ilm after etching SiO2 particles and (b) enlarged image. (Reprinted from K. Ishizu, Y. Tokuno, D. H. Lee, S. Uchida, and M. Ozawa, Journal of Applied Polymer Science 112: 2434–40, 2009. With permission.).

by DC-mediated radical polymerization of MMA initiated by DC groups on SiO2 hybrid nanoparticles modiied with PFD [58]. In Section 8.4, we also mentioned the architecture of polymer brushes encapsulated SiO2 particles by DC-mediated living radical polymerization. This technique can be applied not only for SiO2 but also for the surface modiication of various polymer particles. Scheme8.5 shows the schematic illustration for the synthesis of polymer particles composed of brush structure at surface [61]. First, the cross-linked poly(t-butyl methacrylate) (PBMA) core particles (CCP) were prepared by surfactant-free emulsion copolymerization of t-butyl methacrylate (BMA) and ethylene glycol dimethacrylate (EGDM). Table 8.4 lists the polymerization

Surface Tailoring of Polymer Nanoparticles

CH3 H2C

C

CH3 H 2C

C

O

O

C CH2

C

CH2

+

CH2

O

90°C, H2O

MMA, VBDC

MMA, CuCl, bpy

45°C, H2O

in THF, UV, ct

O

O

C

C(CH3) H2C

O

O CH3

tBMA

K2S2O8

EGDMA

Seed Particle (CCP) Emulsion Polymerization

Core-Shell Particle (CSP) Living Radical Photopolymerization

Brush-Core Particle (BCP) Photo-induced ATRP

SCHEME 8.5 Schematic illustration for the synthesis of poly mer particles composed of brush structure at surface. (Reprinted from D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission.)

241

242

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 8.4 Surfactant-Free Emulsion Polymerization Conditions and Results for PBMA Cross-Linked Core Particles (CCPs)a Code

Temperature (°C)

Total Momonerb Conversion (%)

Dnc (nm)

Dhd (nm)

Dw /Dnc

CCP1 CCP2 CCP3 CCP4

93 90 83 82

83 80 88 87

194 229 251 282

197 233 252 288

1.0002 1.0002 1.0001 1.0003

Note: From D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission. a The emulsion copolymerizations were carried out in feeds of BMA (5.8 mL, 3.62 × 102 mol/L), EGDM (1.7 mL, 0.014 × 10–4 mol/L), and K2S2O8 (0.206 g, 7.65 × 10–4 mol/L) under stirring speed condition for 800 rpm in nitrogen atmosphere for 2 h. b Determined by gravimetric method. c Determined by a survey of 300 samples picked from SEM photographs. d Hydrodynamic diameter was determined by DLS in H O at 25°C. 2

conditions and results for CCP series. The effect of polymerization temperature on particle morphology and particle size was studied under the constant condition of the feed monomer ratio of BMA to EGDM and total monomer concentration. Figure 8.8 shows the SEM photographs of CCP1–CCP4. In all the photographs, the spherical particles are clearly visible and each particle size distribution is extremely narrow (Dw/Dn = 1.001 – 1.003). The particle diameters of CCPs increase with decreasing the polymerization temperature. At high temperature, CCP particles formed in emulsion seem to be stabilized sterically in the scale of small particle size with not only initiator fragments (persulfate anion) but also solubility of PBMA units in water. Next, we synthesized photofunctional core-shell particles (CSPs) possessing DC groups on the surface by encapsulation of seed particles. The poly meri zation conditions and results are listed in Table 8.5. The core particles encapsulated in monomer droplets (MMA, EGDM, and VBDC) start the radical photopoly meri zation and form the thin layer of the shell part. The reaction was run at 45°C, which is much higher than room temperature. An organic-soluble initiator (V-70) was used so that effective initiation occurred in the monomer droplets. This reaction system formed a macroinitiator with highly localized DC groups attached chemically to the surfaces of the colloidal particles. The poly mer i zation rate increased during the time in which the conversion reached 72% after 4 h. A typical SEM photograph of the copoly mer i zation product CSP1 (CCP1 as seed particle) is displayed in Figure 8.9a. The texture shows that all particles have very smooth surfaces and remain unagglomerated (Dn = 204 nm; Dw /Dn = 1.002). The particle size distribution obtained from analysis of the DLS data is

243

Surface Tailoring of Polymer Nanoparticles

(a)

(b)

500 nm (c)

500 nm (d)

500 nm

500 nm

FIGURE 8.8 SEM photographs of cross-linked PBMA core particles (CCP): (a) CCP1, (b) CCP2, (c) CCP3, and (d) CCP4. (Reprinted from D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission.)

presented in Figure 8.9b. The texture shows spherical particles larger (D h = 209 nm) than the SEM image, and narrow size distribution. The content of DC groups for the CSP1 particles was also determined with the radical transfer reaction. As a result, the DC groups were estimated to be 7.59 × 105 number/particle. Brush-core particles (BCPs) were prepared by grafting from photoinduced ATRP. Table 8.6 lists the polymerization conditions and results for BCP1 initiated by CSP1 as macroinitiator. After 8 h of polymerization, the monomer conversion reached 15%. A typical SEM photograph of BCP1 is shown in Figure 8.10a. Spherical particles (Dn = 240 nm) are clearly visible for the sample and particle size distribution was very narrow (Dw /Dn = 1.007). The thickness of the grafted PMMA brushes was observed to be 18 nm. Figure 8.10b shows particle size distribution of BCP1 on DLS data. The distribution is unimodal, and hydrodynamic diameter Dh was 240 nm. This means that PMMA brush chains expanded in THF solution because THF is a good solvent for PMMA. The mass of the PMMA brush phase (MB) and molecular weight

244

Living Radical Photopoly mer i zation Conditions and Results for Core-Shell Particles (CSP)a Feed of Monomer Code

CCP1 (mL)

CSP1

30

b

MMA (mL)

EGDM (mL)

VBDC (mg)

V-70 (mg)

Temperature (°C)

Total Momonerc Conversion (%)

Dnd (nm)

Dhe (nm)

0.2

0.05

0.05

0.05

45

72

204

209

Dw /Dnd 1.0002

Note: From D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission. a Polymerized under UV irradiation for 3 h. b D = 194 nm. n c Determined by gravimetric method. d Determined by a survey of 300 samples picked from SEM photographs. e Hydrodynamic diameter was determined by DLS in H O at 25°C. 2

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 8.5

245

Surface Tailoring of Polymer Nanoparticles

(a)

500 nm 35 (b)

Size Distribution (LS%)

30 25 CSP Dh = 204 nm

20 15 10 5 0

1

10

100

1000

104

Dh (nm) FIGURE 8.9 SEM photograph (a) and size distribution (b) on DLS data for core-shell particles (CSP1) encapsulated CCP seed. (Reprinted from D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission.)

246

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 8.6 Grafting from Photo-Induced ATRP Conditions and Results for PMMA Brush-Core Particles (BCP)a Code

[CSP1]:[CuBr]:[bpy]:[MMA]

Time (h)

Conversionb (%)

Dnc (nm)

Dhd (nm)

Dw /Dnc

BCP1

1:2:4:70

8

15

240

244

1.0007

Note: From D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission. a Polymerized in THF/EtOH(3:1v/v) under stirring speed condition for 800 rpm in UV irradiation. b Determined by gravimetric method. c Determined by a survey of 300 samples picked from SEM photographs. d Hydrodynamic diameter was determined by DLS in THF at 25°C.

of PMMA brush were calculated to be 1.999 × 109 and 1.625 × 103, respectively, with the assumption that all of the DC initiation sites on the microinitiator led to the propagation of MMA. The BCP particles exhibited one DC group at each PMMA brush end. We performed the construction of colloidal crystals and studied their optical properties. The brilliant monochromatic colors covering the visible parts were prepared by controlling the diameter of the monodisperse particles. Figure 8.11a displays color photographs of the prepared ilms CCP1~CCP4 deposited on the glass substrate. The CCP4, CCP3, CCP2, and CCP1 ilms show structural colors, that is, red, yellow, green, and blue, respectively. This means that each ilm forms a colloidal crystal (CC). Figure 8.11b shows the relection spectra of CC-CCP1~CC-CCP4 under the condition of the incident angle θ = 90°. Table 8.7 lists the peak wavelength for colloidal crystals CC-CCP1~CC-CCP4. The Bragg equation is given by

(

λ max = 2Ds ( 2/3) neff 2 − cos 2 θ

)

1/2

(8.1)

which describes the bandgap position for light incident on the (111) face of a face-centered cubic (FCC) lattice, where λmax, Ds, θ, and neff are the peak wavelength, diameter of spheres, angle between the incident light and the normal to the diffraction planes, and the mean effective refractive index of this crystalline array, respectively. The neff can be deined by neff = nP2ϕ P + nM2ϕ M

(8.2)

where nP, nM, ϕP, and ϕM are the refractive index of particles, refractive index of matrix, volume fraction of particles, and volume fraction of matrix, respectively. The CC-CCP ilm is an array composed of PBMA particles and air matrix. Then, neff is estimated to be 1.36 using nP (1.464)62, nM (1.0), ϕP (0.74), and ϕM (0.26). The calculated wavelength (λcald) from Equation (8.1) is 430 nm

247

Surface Tailoring of Polymer Nanoparticles

(a)

600 nm 35 (b)

Size Distribution (LS%)

30 25 BCP Dh = 240 nm

20 15 10 5 0

1

10

100 Dh (nm)

1000

104

FIGURE 8.10 SEM photograph (a) and size distribution (b) on DLS data for brush-core particle (BCP1). (Reprinted from D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission.)

(b)

90

CC-CCP4

CC-CCP3

CC-CCP2

CC-CCP1

Reflectance (%)

75 60 45 30 15 0

20°

200

200

400 600 800 Wavelength (nm)

1000

1200

CC-CCP1

Reflection (%) 0

400 600 800 Wavelength (nm)

(d)

30° 50°

Reflectance (%)

(c)

0

1000

1200

CC-CSP CC-BCP

0

200

400 600 800 Wavelength (nm)

1000

1200

FIGURE 8.11 Photographs and relection spectra of colloidal crystals (CC): (a) color photographs of CC-CCP1 (blue), CC-CCP2 (green), CC-CCP3 (yellow), and CC-CCP4 (red), (b) relection spectra of CC-CCP1~CC-CCP4 (θ = 90°), (c) angle dependence of the relection for CC-CCP4, and (d) relection spectra of CC-CCP1, CC-CSP1, and CC-BCP1.

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

CC-CCP1 CC-CCP2 CC-CCP3 CC-CCP4

248

(a)

249

Surface Tailoring of Polymer Nanoparticles

TABLE 8.7 Optical Properties of Colloidal Crystals Code CC-CCP1 CC-CCP2 CC-CCP3 CC-CCP4

λmaxa (nm)

λcaldb (nm)

Dsc (nm)

Color

450 506 525 626

430 508 557 626

194 229 251 282

Blue Green Yellow Red

Note: From D. H. Lee, Y. Tokuno, S. Uchida, M. Ozawa, and K. Ishizu, Journal of Colloid and Interface Science, 340: 27–34, 2009. With permission. a λ max was determined from the wavelength of relection. b λ cald was calculated by Bragg’s law. c D was determined by the SEM. s

using Ds = 194 nm. This value was well in agreement with the observed one (λmax = 450 nm). Figure 8.11c shows the relection spectra of colloidal crystal CC-CCP4, varying the incident angle θ. The λmax at θ = 50°, 30°, and 20° were 581 (red), 558 (yellow), and 491 nm (green), respectively. The λmax led to blue-shift with decreasing the incident angle. On the other hand, Figure 8.11d shows the relection spectra of colloidal crystals CC-CCP1, CC-CSP1, and CC-BCP1. The observed λmax at θ = 90° is 477 (sky blue) and 516 nm (green) for colloidal crystals CC-CSP1 and CC-BCP1, respectively. The wavelengths shift to the long wavelength side with the increment of particle diameters, and the relection intensities for CC-CSP1 and CC-BCP1 somewhat decrease compared to that of CC-CCP1 due to weak FCC packing. Both colloidal particles had photofunctional DC groups on their surfaces. Therefore, polymeric superstructure ilms (mesoscopically ordered cubic lattices) will be able to construct by living radical graft copolymerization, initiated by using photofunctionalized polymer particles as macroinitiator. Future work will involve controlling the interplanar spacings of structures by changing monomer concentration and type of cubic lattices. These systems will have numerous applications in nanotechnology, such as optical and electronic devices.

8.7 Surface Modification of Polymer Particles via RAFT Polymerization RAFT is arguably the most applicable technique, since most of the monomers that can be polymerized by conventional radical polymerization can also be polymerized through RAFT, which is not possible with other living radical polymerization methods. The conditions for RAFT polymerization

250

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

are similar to that of a conventional radical poly merization except for the addition of a RAFT agent [63–67]. The combination of surface-initiated poly merization and RAFT techniques has been widely explored as a route to design the surface properties and functionality of various substrates. Previous work has used the RAFT polymerization technique to graft polymers to silica particles either by using “grafting to” [68,69] or “grafting from” methods [70–73]. Li and Benicewicz [70] have reported the synthesis method for well-deined polymer brushes grafted onto silica nanoparticles via surface RAFT poly merization. Scheme 8.6 shows the synthetic procedures for attaching RAFT agent onto silica nanoparticles. Novel RAFT-silane agents were prepared that contained both an active RAFT moiety and a silane coupling agent. RAFT agents were anchored to silica nanoparticles by the functionalization of colloidal silica with the RAFT-silane agents. RAFT polymerizations were then conducted from the particle surface to graft homopolymer and block copolymer brushes to the particle. The kinetics of styrene and n-butyl acrylate surface RAFT poly merizations were investigated and compared with model polymerizations mediated by free RAFT agent. The molecular weights of grafted polymers increased linearly with conversions, and irst-order kinetics were observed in the conversion range studied, indicating that the surface graft poly merization proceeded in a controlled manner. More recently, Liu and Pan [73] have also reported the preparation method for surface modiication of silica nanoparticles via “grafting onto” RAFT polymerization. Scheme 8.7 shows a synthetic scheme for RAFT agent onto silica nanoparticles. The RAFT agent, 2-butyric acid dithiobenzoate (BDB), was prepared by substitution of dithiobenzoate magnesium bromide with sodium 2-bromobutyrate under alkali conditions in aqueous solution. Epoxy groups were covalently attached to silica nanoparticles by condensation reaction of 3-glycidyloxypropyltrimethoxysilane (GPS) with the hydroxyl on the silica particle surface. RAFT agent–functionalized nanoparticles were produced by the ring-opening reaction of the epoxy group with carbonyl group of BDB. Then, PS chains with controlled molecular weights and narrow polydispersities (less than 1.1) were grown from the RAFT agent– anchored nanoparticle surface. On the other hand, PMMA-silica nanocomposites were produced by “grafting through” using RAFT polymerization [74]. Scheme 8.8 shows the synthesis procedures for methacrylate-functionalized silica nanoparticles. The surface of silica nanoparticle was modiied covalently by attaching methacryl group to the surface using 3-methacryloxypropyldimethylchlorosilane. Polymerization of MMA using the 4-cyano-4-(dodecylsulfanylthiocarbonyl) sulfanyl pentanoic acid RAFT agent produced the PMMA-SiO2 nanocomposites. It is not different from the studies concerning the surface modiication of polymer particles via RAFT polymerization, but there are several reports that the core-shell polymer particles were prepared by cross-linking of core

O H

Karstedt’s catalyst dimethylchlorosilane

O Br

O H

1

Si Cl

CH3OH/pyridine

O Br

O H

2

Si OCH3 3

S SMgBr stir at rt.

S

O S

O H 4

Si O

CH3 nanoparticles

S

O S

MIBK, 85°C

O

Si O

H 5 S

O S

Surface Tailoring of Polymer Nanoparticles

O Br

O H

6 SCHEME 8.6 Synthesis procedures for attaching RAFT agent onto silica nanoparticles. (Reprinted from C. Li and B. C. Benicewicz, Macromolecules 38: 5929–36, 2005. With permission.)

251

252

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Br

1. H2O

Mg/EtO2 (C2H5)2O

S

CS2

MgBr

C

(C2H5)2O

Br

SMgBr

S

2. CH3CH2CHCOONa

CH3CH2CH S C

3. HCl

COOH

OH +

(CH3O)3Si(CH2)3OCH2CH CH2 O COOH CH3CH2CH

S

S C

CH2

OSi(CH2)3OCH2CH O

O OSi(CH2)3OCH2CH CH2OC OH

S CHSC C2H5

SCHEME 8.7 Synthesis scheme for RAFT agent onto silica nanoparticles. (Reprinted from C.-H. Liu and C.-Y. Pan, Polymer 48: 3679–85, 2007. With permission.) O

O Cl

Si

O

+

SiO2

SiO2

~20 nm methacryloxypropyldimethylchlorosilane (MCPDCS)

O SiO2-MMA

RAFT Polymerization MMA + AIBN S CN S S CO2H

11

PMMA + SiO2 nanocomposite

SiO2

SiO2

SCHEME 8.8 Synthesis procedures for methacrylate-functionalized silica nanoparticles. (Reprinted from P. S. Chinthamanipeta, S. Kobukata, H. Nakata, and D. A. Shipp, Polymer 49: 5636–42, 2008. With permission.)

Surface Tailoring of Polymer Nanoparticles

253

or corona parts in micelles formed by block copolymers via RAFT polymerizations [75–77].

8.8 Surface Modification of Silica Nanoparticles via Nitroxide-Mediated Radical Polymerization Actually, the controlled living polymerization technique nitroxide-mediated radical polymerization (NMRP) in conjunction with surface-initiated polymerization is among the most useful routes to precisely design and functionalize the surfaces of various solid materials by well-deined polymers. However, because the preparation of functional alkoxyamine is a complicated, multistep process, there have been few explorations about synthesis of polymer/silica hybrid composites by NMRP in recent years [78–81]. 2,2,6,6-Tetramethylpiperidine-N-oxyl (TEMPO) bromide salt was used to functionalize a silica surface with nitroxyl moieties [81]. Scheme 8.9 shows schematic representation of the functionalization of silica mediated by oxoaminium bromide salt and the graft polymerization of poly[styrene-co(maleic anhydride)]. The functionalization reaction took place in 48 h under mild conditions. In a second step, grafts of styrene-maleic anhydride copolymer were grown from the functionalized silica surface by heating it in the presence of the monomers. FT-IR and TGA analysis showed that the silica surface irst functionalized with nitroxide moieties, and then that grafts of styrene-maleic anhydride grew from the functionalized silica surface. The results suggest that the oxoaminium salts were good candidates for the functionalization and grafting of surfaces that contain hydroxyl groups and for the generation of hybrid materials with improved properties. More recently, Wang et al. [82] have also reported the synthesis of poly(styrene-co-maleic anhydride) (PSMA)/SiO2 hybrid composites via “grafting onto” strategy based on NMRP. Scheme 8.10 shows the grafting polymerization of styrene from vinyl trimethoxysilane (A-171) modiied nanosilica surface. Two steps were used to graft styrene/maleic anhydride copolymer chains to the surface of nanosilica: irst, anchoring A-171 onto the surface of nanosilica, and then, using TEMPO as a radical trap, trapping the radicals produced by reaction of benzoyl peroxide (BPO) with styrene, maleic anhydride, and the vinyl group in grafted A-171 molecules. Finally, well-controlled molecular weight and narrow molecular weight distribution of PSMA chains were grown from the surface of nanosilica. PSMA is a polymeric material available in the market that has some good properties such as solubility, ilming, and miscibility when mixed with styrene polymers [83,84]. Meanwhile, the anhydride groups can easily react with low molecular

254

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(1) HO

OH

HO

(2)

– Br + N O

OH

OH HO

OH

HO

OH

O

O

N

+

Et3N, CH2Cl2 25°C, 48h HO

HO

OH

Et3NHBr

OH OH

OH Unmodified Silica

O O 126°C

O

2h OH

(3) N O

O

H2C

O

HO

O CH2

HC

OH

O

OH +

“Free” polymer

HC HO m

OH

n

OH

SCHEME 8.9 Schematic representation of the functionalization of silica mediated by oxiaminium bromine salt and the graft poly merization of poly[styrene-co-(maleic anhydride)]. (Reprinted from B. C. José, L. C. Tania, S. G. Enrique, and J. R. Enrique, Macromolecular Rapid Communications 28: 1397–1403, 2007. With permission.)

OH

HO

Si

HO

OH

HO

OH OH

A-171 Ultrasound dispersed

HO

O

O

Si

OH

HO HO

O

O Si

HO HO

O

Si

Si OH HO

Styrene/Maleic anhydride TEMPO/BPO/130°C

SCHEME 8.10 The grafting poly merization of styrene from the A-171 modiied nanosilica surface. (Reprinted from Y. Wang, Y. Shen, X. Pei, S. Zhang, H. Liu, and J. Ren, Reactive and Functional Polymers 68: 1225–30, 2008. With permission.)

Surface Tailoring of Polymer Nanoparticles

255

weight compounds such as water, alcohols, and amines. So, PSMA can be readily subjected to modiication and has been widely used in industry.

8.9 Conclusions We demonstrated various strategies for the surface modiication of polymeric/inorganic particles via living radical mechanisms such as ATRP, NMRP, DC-mediated, and RAFT poly mer i zations. Previous work has used these poly mer i zation techniques to graft polymers to poly mer or silica particles by using the “grafting to,” “grafting from,” or “grafting through” method. We also established a new synthetic method for crosslinked shell possessing poly mer or silica particles encapsulated in photofunctional DC groups by the free radical emulsion copoly mer i zation using inimer VBDC as one component monomer. In this copoly mer i zation, the inimer VBDC had the formation of hyperbranched structures at its surface by living radical photopoly mer i zation. Subsequently, core-brush particles were synthesized by the photoinduced ATRP of vinyl monomers initiated by photofunctional cross-linked particles as a macroinitiator. Because of the promising combination of a nanostructured poly mer with the properties of reactive particles, these nanoparticles can be applied in the ields of hybrid nanocomposites and optic materials (such as structural color materials).

References 1. Furusawa, K., Norde, W., and J. Lyklema. 1972. Method for preparing surfactantfree polystyrene lattices of high surface change. J Kolloid Z Z Polym 250:908–9. 2. Vanderhoff, J. W., El-Aasser, M. S., Micale, F. J., Sudol, E. D., Tseng, C. M., Kornfeld, D. M., and F. A. Vincente. 1984. Preparation of large particle-size monodisperse latexes in space-polymerization kinetics and process-development. Journal of Dispersion Science and Technology 5:231–46. 3. Ugelstad, J., Mork, P. C., Kaggerud, K. G., Ellingsen, T., and A. Berge. 1980. Swelling of oligomer-polymer particles. New methods of preparation of emulsions and polymer dispersions. Advances in Colloid and Interface Science 13:101–4. 4. Ugelstad, J., Mfutakamba, H. R., and P. C. Mork. 1985. Preparation and application of monodisperse polymer particles. Journal of Polymer Science, Polymer Symposia 72:225–40.

256

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

5. Antl, L., Goodwin, J. W., Hill, R. D., Ottewill, R. H., Owens, S. M., and S. Papworth. 1986. The preparation of poly(methyl methacrylate) lattices in nonaqueous media. Colloids and Surfaces 17:67–78. 6. Williamson, B., Lukas, R., Winnik, M. A., and M. D. Croucher. 1987. The preparation of micron-size polymer particles in nonpolar media. Journal of Colloid and Interface Science 119:559–64. 7. Ober, C. K., and K. P. Lok. 1987. Formation of large monodisperse copolymer particles by dispersion polymerization. Macromolecules 20:268–73. 8. Tseng, C. M., Lu, Y. Y., El-Aasser, M. S., and J. W. Vanderhoff. 1986. Uniform polymer particles by dispersion polymerization in alcohol. Journal of Polymer Science, Part A: Polymer Chemistry 24:2995–3007. 9. Ishizu, K. 1998. Synthesis and structural ordering of core-shell polymer microspheres. Progress in Polymer Science 23:1383–1408. 10. Matyjaszewski, K., and J. Xia. 2001. Atom transfer radical polymerization. Chemical Reviews 101:2921–90. 11. Kamigaito, M., Ando, T., and M. Sawamoto. 2001. Metal-catalyzed living radical polymerization. Chemical Reviews 101:3689–3746. 12. Hawker, C. J., Bosman, A. W., and E. Harth. 2001. New polymer synthesis by nitroxide mediated living radical polymerizations. Chemical Reviews 101:3661–88. 13. Chiefari, J., and E. Rizzardo. 2002. In Handbook of radical polymerization, ed. K. Matyjaszewski and T. P. Davis, 629. Hoboken, NJ: Wiley-Interscience. 14. Otsu, T. 2000. Inifeter concept and living radical polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 38:2121–36. 15. Ishizu, K., Khan, R. A., Ohta, Y., and M. Furo. 2004. Controlled radical polymerization of 2-hydroxyethyl methacrylate initiated by photofunctional 2-(N,N- diethyldithiocarbamate)isobutyric acid. Journal of Polymer Science, Part A: Polymer Chemistry 42:76–82. 16. Ishizu, K., Katsuhara, H., and K. Itoya. 2005. Controlled radical polymerization of methacrylic acid initiated by diethyldithiocarbamate-mediated inifeter. Journal of Polymer Science, Part A: Polymer Chemistry 43:230–33. 17. Ishizu, K., Katsuhara, H., Kawaguchi, S., and M. Furo. 2005. Controlled radical polymerization of styrene initiated by diethyldithiocarbamate-mediated inifeters. Journal of Applied Polymer Science 95:413–18. 18. Ishizu, K., and A. Mori. 2000. Synthesis of hyperbranched polymers by selfaddition free radical vinyl polymerization of photo-functional styrene. Macromolecular Rapid Communications 21:665–68. 19. Ishizu, K., Tsubaki, K., Mori, A., and S. Uchida. 2003. Architecture of nanostructured polymers. Progress in Polymer Science 28:27–54. 20. Ishizu, K., and H. Kakinuma. 2005. Synthesis of nanocylinders consisting of graft block copolymers by photo-induced ATRP technique. Journal of Polymer Science, Part A: Polymer Chemistry 43:63–70. 21. Yamago, S., Iida, K., and J. Yoshida. 2002. Organotellurium compounds as novel initiators for controlled/living radical polymerizations. Synthesis of functionalized polystyrenes and end-group modiications. Journal of the American Chemical Society 124:2874–75. 22. Ishizu, K., and S. Shiratori. 2001. Microsphere synthesis by dispersion copolymerization using poly(t-butyl methacrylate) macromonomers in nonaqueous media. Journal of Colloid and Interface Science 236:266–69.

Surface Tailoring of Polymer Nanoparticles

257

23. Ishizu, K., Shiratori, S., and A. Abdel-Naby. 2001. Microsphere synthesis by dispersion copolymerizations using polyisoprene macromonomers in non-aqueous media. Journal of Materials Science Letters 20:1675–77. 24. Ishizu, K., Yasuda, M., and T. Tamura. 2003. Synthesis of core-crosslinked core-shell polymer particles by free-radical dispersion copolymerization of 4-vinylpyridine with polystyrene macromonomers in nonaqueous media. Journal of Colloid and Interface Science 267:320–25. 25. Dai, Q., Wu, D. Z., Zhang, Z. C., and Q. Ye. 2003. Preparation of monodisperse poly(methyl methacrylate) particles by radiation-induced dispersion polymerization using vinyl terminus polysiloxane macromonomer as polymerizable stability. Polymer 44:73–77. 26. Prucker, O., and J. Rühe. 1998. Synthesis of poly(styrene) monolayers attached high surface area silica gels through self-assembled monolayers of azo initiators. Macromolecules 31:592–601. 27. Prucker, O., and J. Rühe. 1998. Mechanism of radical chain polymerizations initiated by azo compounds covalently bound to the surface of spherical particles. Macromolecules 31:602–13. 28. Prucker, O., Schimmel, M., Tovar, G., Knoll, W., and J. Rühe. 1998. Microstructuring of molecularly thin polymer layers by photolithography. Advanced Materials 10:1073–77. 29. Lin, C.-L., Chiu, W.-Y., and C.-F. Lee. 2006. Preparation, morphology, and thermoresponsive properties of poly(N-isopropylacrylamide)-based copolymer microgels. Journal of Polymer Science, Part A: Polymer Chemistry 44:356–70. 30. Albertsson, A.-C., David, G., Stranderg, C., Bilba, D., and C. Paduraru. 2005. Synthesis of core-shell structured carboxylated microparticles with a straightforward procedure and their evaluation as a polymer support. Journal of Polymer Science, Part A: Polymer Chemistry 43:5889–98. 31. Wan, D., Li, Z., and J. Huang. 2005. Synthesis of a new type of core-shell particles from hyperbranched polyglycerol. Journal of Polymer Science, Part A: Polymer Chemistry 43:5458–64. 32. Wan, D., Fu, Q., and J. Huang. 2005. Synthesis of a thermoresponsive shellcrosslinked 3-layer onion-like polymer particle with a hyperbranched polyglycerol core. Journal of Polymer Science, Part A: Polymer Chemistry 43:5652–60. 33. Lee, C. F. 2005. Effects of surfactants on the morphology of composite polymer particles produced by two-stage seeded emulsion polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 43:2224–36. 34. Ishizu, K., Kobayakawa, N., Takano, S., Tokuno, Y., and M. Ozawa. 2007. Synthesis of polymer particles possessing radical initiating sites on the surface by emulsion copolymerization and construction of core-shell structures by a photoinduced atom transfer radical polymerization approach. Journal of Polymer Science, Part A: Polymer Chemistry 45:1771–77. 35. Akl, M. A. A., Kenawy, I. M. M., and R. R. Lasheen. 2004. Organically modiied silica gel and lame atomic absorption spectrometry: Employment for separation and preconcentration of nine trace heavy metals for their determination in natural aqueous systems. Microchemical Journal 78:143–56. 36. Filha, M. G., Wanderley, A. F., Sousa, K. S., Espinola, J. G. P., Fonseca, M. G., Arakaki, T., and L. N. H. Arakaki. 2006. Thermodynamic properties of divalent cations complexed by ethylenesulide immobilized on silica gel. Colloids and Surfaces A 279:64–68.

258

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

37. von Werne, T. and T. E. Patten. 1999. Preparation of structurally well-deined polymer-nanoparticle hybrids with controlled/living radical polymerizations. Journal of the American Chemical Society 121:7409–10. 38. von Werne, T. and T. E. Patten. 2001. Atom transfer radical polymerization from nanoparticles: A tool for the preparation of well-deined hybrid nanostructures and understanding the chemistry of controlled/”living” radical polymerizations from surfaces. Journal of the American Chemical Society 123:7497–7505. 39. Pyun, J., Matyjaszewski, K., Kowalewski, T., Savin, D. A., Patterson, G. D., Kickelbick, G., and N. Huesing. 2001. Synthesis of well-deined block copolymers tethered to polysilsesquioxane nanoparticles and their nanoscale morphology on surfaces. Journal of the American Chemical Society 123:9445–46. 40. Pyun, J., Jia, S., Kowalewski, T., Patterson, G. D., and K. Matyjaszewski. 2003. Synthesis and characterization of organic/inorganic hybrid nanoparticles: Kinetics of surface-initiated atom transfer radical polymerization and morphology of hybrid nanoparticle ultrathin ilms. Macromolecules 36:5094–5104. 41. Savin, D. A., Pyun, J., Patterson, G. D., Kowalewski, T., and K. Matyjaszewski. 2002. Synthesis and characterization of silica-grafted-polystyrene hybrid nanoparticles: Effect of constraint on the glass-transition temperature of spherical polymer brushes. Journal of Polymer Science, Part B: Polymer Physics 40:2667–76. 42. Carrot, G., Diamanti, S., Manuszak, M., Charleux, B., and J.-P. Vairon. 2001. Atom transfer radical polymerization of n-butyl acrylate from silica nanoparticles. Journal of Polymer Science, Part A: Polymer Chemistry 39:4294–4301. 43. Mori, H., Seng, D. C., Zhang, M., and A. H. E. Müller. 2002. Hybrid nanoparticles with hyperbranched polymer shells via self-condensing atom transfer radical polymerization from silica surfaces. Langmuir 18:3682–93. 44. Wang, Y.-P., Pei, X.-W., and K. Yuan. 2005. Reverse ATRP grafting from silica surface to prepare well-deined organic/inorganic hybrid nanocomposites. Materials Letters 59:520–23. 45. El Harrak, A., Carrot, G., Oberdisse, J., Eychenne-Baron, C., and F. Boué. 2004. Surface-atom transfer radical polymerization from silica nanoparticles with controlled colloidal stability. Macromolecules 37:6376–84. 46. Liu, P., Tian, J., Liu, W., and Q. Xue. 2004. Surface-initiated atom transfer radical polymerization (ATRP) of styrene from silica nanoparticles under UV irradiation. Polymer International 53:127–30. 47. Parvole, J., Laruelle, G., Guimon, C., Francois, J., and L. Billon. 2003. Initiatorgrafted silica particles for controlled free radical polymerization: Inluence of the initiator structure on the grafting density. Macromolecular Rapid Communications 24:1074–78. 48. Perruchot, C., Khan, M. A., Kamitsi, A., Armes, S. P., von Werne, T., and T. E. Patten. 2001. Synthesis of well-deined, polymer-grafted silica particles by aqueous ATRP. Langmuir 17:4479–81. 49. Chen, X. Y. and S. P. Armes. 2003. Surface polymerization of hydrophilic methacrylates from ultraine silica sols in protic media at ambient temperature: A novel approach to surface functionalization using a polyelectrolytic macroinitiator. Advanced Materials 15:1558–62. 50. Chen, X., Armes, S. P., Greaves, S. J., and J. Watts. 2004. Synthesis of hydrophilic polymer-grafted ultraine inorganic oxide particles in protic media at ambient temperature via atom transfer radical polymerization: Use of an electrostatically adsorbed polyelectrolytic macroinitiator. Langmuir 20:587–95.

Surface Tailoring of Polymer Nanoparticles

259

51. Ohno, K., Morinaga, T., Koh, K., Tsujii, Y., and T. Fukuda. 2005. Synthesis of monodisperse silica particles coated with well-deined high-density polymer brushes by surface-initiated atom transfer radical polymerization. Macromolecules 38:2137–42. 52. Kotal, A., Mandal, T. K., and D. R. Walt. 2005. Synthesis of gold-poly(methyl methacrylate) core-shell nanoparticles by surface-conined atom transfer radical polymerization at elevated temperature. Journal of Polymer Science, Part A: Polymer Chemistry 43:3631–42. 53. Ohno, K., Koh, K., Tsujii, Y., and T. Fukuda. 2002. Synthesis of gold nanoparticles coated with well-deined, high-density polymer brushes by surface-initiated living radical polymerization. Macromolecules 35:8989–93. 54. Ishizu, K., Lee, D. H., Tokuno, Y., Uchida, S., and M. Ozawa. 2008. Novel synthesis of poly(methyl methacrylate) brushes encapsulated silica particles. Journal of Applied Polymer Science 109:3968–74. 55. Park, E. J., Kim, W. S., Hwang, H. S., Park, C., and K. T. Lim. 2007. PMMA encapsulated silica particles by dispersion polymerization in supercritical carbon dioxide. Macromolecular Symposia 249/250:196–201. 56. Takeoka, Y., and M. Watanabe. 2002. Polymer gels that memorize structures of mesoscopically sized templates. Dynamic and optical nature of periodic ordered mesoporous chemical gels. Langmuir 18:5977–80. 57. Yoshinaga, K., Fujiwara, K., Mouri, E., Ishii, M., and H. Nakamura. 2005. Stepwise controlled immobilization of colloidal crystals formed by polymergrafted silica particles. Langmuir 21:4471–77. 58. Ishizu, K., Tokuno, Y., Lee, D. H., Uchida, S., and M. Ozawa. 2009. Synthesis of silica hybrid nanoparticles modiied with photofunctional polymers and construction of colloidal crystals. Journal of Applied Polymer Science 112:2434–40. 59. Ishizu, K., Ohta, Y., and S. Kawauchi. 2002. Kinetics of hyperbranched polystyrenes by free radical polymerization of photofunctional inimer. Macromolecules 35:3781–84. 60. Darragh, P. J., Gaskin, A. J., Terrell, B. C., and J. V. Sanders. 1996. Origin of precious opal. Nature 209:13–16. 61. Lee, D. H., Tokuno, Y., Uchida, S., Ozawa, M., and K. Ishizu. 2009. Architecture of polymer particles composed of brush structure at surface and construction of colloidal crystals. Journal of Colloid and Interface Science, 340:27–34. 62. Uozu, Y., Hirota, N., and K. Horie. 2004. Improvement of chromatic aberration of the plastic rod-lens array, 2-A rod-lens array consisting of novel transparent polymer blends composed of monomer units with different refractive indices and almost equal Abbe numbers. Macromolecular Materials and Engineering 289:56–62. 63. Shipp, D. A. 2005. Living radical polymerization: Controlling molecular size and chemical functionality in vinyl monomers. Journal of Macromolecular Science, Part C: Polymer Reviews 45:171–94. 64. Chiefari, J., Chong, Y. K. B., Ercole, F., Krstina, J., Jeffery, J., Le, T. P. T., Mayadunne, R. T. A., Meijs, G. F., Moad, C. L., Moad, G., Rizzardo, E., and S. H. Thang. 1998. Living free-radical polymerization by reversible additionfragmentation chain transfer: The RAFT process. Macromolecules 31:5559–62. 65. Schilli, C., Lanzendorfer, M., and A. H. E. Muller. 2002. Benzyl and cumyl dithiocarbamates as chain transfer agent in the RAFT polymerization of N-isopropylamide. In situ FT-NIR and MALDI-TOF MS investigation. Macromolecules 35:6819–27.

260

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

66. Moad, G., Rizzardo, E., and S. H. Thang. 2005. Living radical polymerization by the RAFT process. Australian Journal of Chemistry 58:379–410. 67. Moad, G., Chong, Y. K., Postma, A., Rizzardo, E., and S. H. Thang. 2005. Advances in RAFT polymerization: The synthesis of polymers with deined end-groups. Polymer 46:8458–68. 68. Bridger, K., Fairhurst, D., and B. Vincent. 1979. Nonaqueous silica dispersions stabilized by terminally-grafted polystyrene chains. Journal of Colloid and Interface Science 68:190–95. 69. Vincent, B. 1993. The preparation of colloidal particles having (post-grafted) terminally-attached polymer chains. Chemical Engineering and Science 48:429–36. 70. Li, C., and B. C. Benicewicz. 2005. Synthesis of well-deined polymer brushes grafted onto silica nanoparticles via surface reversible addition-fragmentation chain transfer polymerization. Macromolecules 38:5929–36. 71. Li, C., Han, J., Ryu, C. Y., and B. C. Benicewicz. 2006. A versatile method to prepare RAFT agent anchored substrates and the preparation of PMMA grafted nanoparticles. Macromolecules 39:3175–83. 72. Zhao, Y., and S. Perrier. 2006. Synthesis of well-deined homopolymer and diblock copolymer grafted onto silica particles by Z-supported RAFT polymerization. Macromolecules 39:8603–8. 73. Liu, C.-H., and C.-Y. Pan. 2007. Grafting polystyrene onto silica nanoparticles via RAFT polymerization. Polymer 48:3679–85. 74. Chinthamanipeta, P. S., Kobukata, S., Nakata, H., and D. A. Shipp. 2008. Synthesis of poly(methyl methacrylate)-silica nanocomposites using methacrylatefunctionalized silica nanoparticles and RAFT polymerization. Polymer 49:5636–42. 75. Kakwere, H., and S. Perrier. 2009. Orthogonal “relay” reactions for designing functionalized soft nanoparticles. Journal of the American Chemical Society 131:1889–95. 76. Pearson, S., Allen, N., and M. H. Stenzel. 2009. Core-shell particles with glycopolymer shell and polynucleoside core via RAFT: From micelles to rods. Journal of Polymer Science, Part A: Polymer Chemistry 47:1706–23. 77. Yao, X., Yao, H. W., Li, Y., and G. Chen. 2009. Preparation of honeycomb scaffold with hierarchical porous structures by core-crosslinked core-corona nanoparticles. Journal of Colloid and Interface Science 332:165–72. 78. Husseman, M., Malmstro, E. E., McNamara, M., Mate, M., Mecerreyes, D., Benoit, D. G., Hedrick, J. L., Mansky, P., Huang, E., Russell, T. P., and C. J. Hawker. 1999. Controlled synthesis of polymer brushes by “living” free radical polymerization techniques. Macromolecules 32:1424–31. 79. Kasseh, A., Ait-Kadi, A., Riedl, B., and J. F. Pierson. 2003. Organic/inorganic hybrid composites prepared by polymerization compounding and controlled free radical polymerization. Polymer 44:1367–75. 80. Andruzzi, L., Hexemer, A., Li, X. F., Ober, C. K., Kramer, E. J., Galli, G., Chiellini, E., and D. A. Fischer. 2004. Control of surface properties using luorinated polymer brushes produced by surface-initiated controlled radical polymerization. Langmuir 20:10498–506.

Surface Tailoring of Polymer Nanoparticles

261

81. José, B. C., Tania, L. C., Enrique, S. G., and J. R. Enrique. 2007. Towards controlled graft polymerization of poly[swtyrene-co-(maleic anhydride)] on functionalized silica mediated by oxoaminium bromide salt. Facile synthetic pathway using nitroxide chemistry. Macromolecular Rapid Communications 28:1397–1403. 82. Wang, Y., Shen, Y., Pei, X., Zhang, S., Liu, H., and J. Ren. 2008. In situ synthesis of poly(styrene-co-maleic anhydride)/SiO2 hybrid composites via “grafting onto” strategy based on nitroxide-mediated radical polymerization. Reactive and Functional Polymers 68:1225–30. 83. Shiomi, T., Karaz, F. E., and W. J. Mcknight. 1986. Compatibility in blends of two random copolymers having a common monomer segment. Macromolecules 19:2274–80. 84. Aoki, Y. 1988. Miscibility of poly(acrylonitrile-co-styrene) with poly[styrene-co(maleic anhydride)] and poly[styrene-co-(N-phenylmaleimide)]. Macromolecules 21:1277–82.

9 Effects of Nano-Sized Polymerization Locus on the Kinetics of Controlled/ Living Radical Polymerization Hidetaka Tobita CONTENTS 9.1 9.2

Introduction ................................................................................................ 264 Origin of Livingness in Controlled/Living Radical Polymerization (CLRP).............................................................................. 265 9.3 Polymerization Rate .................................................................................. 267 9.3.1 Basic Concept.................................................................................. 267 9.3.2 SFRP and ATRP (Radical Long Life-ization Process) ............... 268 9.3.2.1 SFRP and ATRP in Bulk ................................................. 268 9.3.2.2 SFRP and ATRP in Dispersed Systems ........................ 271 9.3.3 RAFT and DT ................................................................................. 282 9.3.3.1 RAFT and DT in Bulk ..................................................... 282 9.3.3.2 RAFT and DT in Dispersed Systems............................ 285 9.4 Molecular Weight Distribution (MWD) ................................................. 293 9.4.1 Fundamental Distribution in CLRP ............................................ 293 9.4.1.1 Chain Length Distribution During a Single Active Period: Most Probable Distribution .............................. 293 9.4.1.2 Fundamental Distribution: Broader than the Poisson Distribution ....................................................... 295 9.4.2 SFRP and ATRP in Dispersed Systems ...................................... 297 9.4.3 RAFT Polymerization in Dispersed Systems (Normal Lifetime Process) ............................................................................ 298 9.4.3.1 Monomer-Concentration-Variation (MCV) Effect ...... 299 9.4.3.2 Effect of Particle Size on the MWD .............................. 299 9.5 Summary..................................................................................................... 302 Acknowledgment ................................................................................................ 303 References............................................................................................................. 303 Appendix: Formulation of Various Average Times........................................305

263

264

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

9.1 Introduction This chapter is aimed at clarifying the effects of small reaction locus on the kinetics of controlled/living radical polymerization (CLRP). The polymerization kinetics in a dispersed system is complex partly because of the transport phenomena of various species, which could be utilized to prepare functional nanoparticles. To clarify the very basic characteristics of polymer formation through controlled/living radical polymerization, the transport phenomena among particles are neglected in this chapter. Theoretical investigation is conducted on how the small reaction locus changes the kinetics of CLRP, assuming the polymer particles as ideally isolated microreactors. When the initiator in continuous phase is used, only the radical entry into the particles is accounted for and the exit of species from a particle is neglected. To further simplify the discussion, the particles are assumed to be monodisperse. The present polymerization systems correspond to the ideal miniemulsion polymerization. The experimental and theoretical investigations conducted so far for various types of CLRPs in bulk (Goto and Fukuda 2004; Braunecker and Matyjaszewski 2007) and also in a dispersed medium (Butte, Storti, and Morbidelli 2001; Cunningham 2008; Zetterlund, Kagawa, and Okubo 2008) have already been summarized. In this chapter, I try to add a fresh dimension to the kinetics of CLRP. In the present chapter, the mathematical tool used for the analysis of polymerization kinetics inside nano-sized particles is the Monte Carlo (MC) simulation method reported earlier (Tobita and Yanase 2007; Tobita 2009a, 2009b). The MC simulation method is a powerful tool to analyze the kinetics of emulsion polymerization (Tobita 1995), and it enables investigation of what is happening inside each particle. Very detailed information can be extracted from this numerical technique. The effects of statistical variation among particles on the kinetics of CLRP in nano-sized particles, introduced in this chapter by the name of the luctuation effect (for stable free-radical mediated polymerization [SFRP] and atom transfer radical polymerization [ATRP]) and monomer-concentration-variation effect (for reversible additionfragmentation chain transfer [RAFT] polymerization, degenerative transfer [DT] polymerization, and conventional free radical polymerization), are fruitful discoveries made possible through this powerful technique. The effects of particle size on the polymerization rate and the formed molecular weight distribution (MWD) proiles are discussed for various types of CLRPs in a uniied manner.

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

265

9.2 Origin of Livingness in Controlled/Living Radical Polymerization (CLRP) In free radical polymerization, the bimolecular termination reactions are inevitable. Therefore, the living polymerization in a strict sense in which the chain termination reactions are totally absent is impossible. However, if a large percentage of polymer chains are dormant and can potentially grow further, such free radical polymerization systems can be regarded as pseudo-living polymerization. Because the lifetime of a generated radical is short, normally ~100 s, a basic strategy to keep the chain potentially active is to distribute very short active periods throughout the whole reaction time. Figure 9.1 shows a schematic representation of a pseudo-living polymer formation. The thickness of each vertical line shows the time length of an active period, which is typically 10 –4–10 –3 s. On the other hand, it takes hours for the whole polymerization time. From the point of view of the formation history of a chain, most of the time is spent as a dormant form. Two types of methods have been proposed to make the active periods distributed throughout the whole polymerization time. The irst method is to protect a radical from termination reactions by reversible capping with a trapping agent. The trapping reaction provides a temporal protection of active radicals, and the pseudo-living condition is assured by reducing the frequency of bimolecular termination. SFRP and ATRP fall into this category (see Figure 9.2 for various reversible reaction schemes). The history of chain formation and that of a radical are the same and are both represented schematically by Figure 9.1. The time fraction of the active period in terms of growing chain, ϕA,Chain, and in terms of radical species, ϕA, is the same, and for both is given by tR•/(tR• + tPX), where tZ is the average lifetime spent in the chemical form, Z. Both ϕA,Chain and ϕA must be very much smaller than unity, essentially very close to zero. The fundamental procedure to explicitly represent various average lifetimes can be found in the Appendix to this chapter. In this polymerization process, the apparent lifetime of a radical is extended signiicantly, and such polymerization process is called the radical long lifeReaction time, tP Active period

Dormant period FIGURE 9.1 Schematic representation of the formation history of a poly mer chain in CLRP.

266

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

SFRP

Pi X

k1 k2

ATRP Pi X + Y RAFT Ri• + XPj DT

Ri• + XPj

Ri• + X k1 k2 k2 k1 kex kex

Ri• + XY PiXPj

k1 k2

Pi X + Rj•

Pi X + Rj•

FIGURE 9.2 Reversible reaction scheme in each type of CLRP. In the igure, Pi is the poly mer with chain length i, and R•i is the active poly mer radical with chain length i.

ization process in this chapter. Note, however, the actual total time length of all growth periods of a chain is still smaller than a few seconds and is very short. The lifetime is extended simply because of very long dormant periods in which an active radical is capped by a trapping agent. In addition, the radical long life-ization should not be confused with the persistent radical or stable radical. The persistent radical and/or stable radical in SFRP refer to the trapping agent, such as TEMPO and verdazyls, which cannot propagate. On the other hand, the long-life radical named here can propagate during the active period. The second method for distributing the chain growth period is to utilize the chain transfer reactions, such as in RAFT polymerization and DT polymerization. In these cases, the pseudo-livingness does not necessarily come from a reduced frequency of bimolecular termination, but is provided by relaying an active radical to a large number of chains before inally being stopped by bimolecular termination. In this type of process, the initiation reaction is needed to start polymerization. The history of a radical lifetime is shown in Figure 9.3. All of the chains except for the inally terminated chains can grow further when the next chain transfer reaction occurs on the chain end. In this process, the termination reactions are anticipated and do not have to be prevented. The radical long life-ization is not necessarily required. The total number of relayed chains is important to keep the pseudo-livingness. The time fraction of the active period in terms of growing chain, ϕA,Chain, is given by tR•/(tR• + tXP) for RAFT and DT, and is much smaller than unity, as in the cases of SFRP and ATRP. This is why the theoretical MWD function is essentially the same for all CLRPs. On the other hand, when discussing the polymerization rate, what is important is the time fraction of the active radical period in terms of radical species, ϕA, which is given by tR•/(tR• + tPXP) for RAFT and is ϕA = 1 for DT. For RAFT, ϕA does not have to be much smaller than unity. Note that ϕA = 1 in the conventional, nonliving free radical polymerization.

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

267

Termination

Initiating chain end FIGURE 9.3 Schematic representation of the chain formation in RAFT and DT.

9.3 Polymerization Rate 9.3.1 Basic Concept The rate of free radical polymerization is most popularly represented by: Rp = k p [M][R• ]

(9.1)

where kp is the propagation rate constant, [M] is the monomer concentration, and [R•] is the active radical concentration. It might be less popular, but the following expression can also be used for free radical polymerization, as long as the initiation is slow and the lifetime of a radical is short: R p = RI ν

(9.2)

where R I is the initiation rate, ν is the kinetic chain length, represented by ν = kp [M]/kt [R•], and kt is the termination rate constant. By using the relationship, R I = Rt = kt [R•] 2, it is straightforward to show that Equations (9.1) and (9.2) are equivalent. Note that the termination rate constant deined by Rt = kt [R•] 2 is used in this chapter, not Rt = 2kt [R•] 2. To facilitate the discussion on the CLRP, Equation (9.2) is generalized. Let RRG be the radical generation rate in the reaction locus (mol/[L•s]), and Lν be the average number of monomeric units added to the generated radical until the chain stops growing. Then, the polymerization rate, Rp, is represented by: Rp = RRG Lν

(9.3)

268

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

TABLE 9.1 Concentration of a Single Molecule Assuming the Density of a Particle Is 1 g/cm3 Diameter, Dp [nm]

Concentration of a Single Molecule in mol•L–1

200 100 50 30 10

3.98 × 10–7 3.18 × 10–6 2.54 × 10–5 1.18 × 10–4 3.18 × 10–3

In the conventional free radical polymerization in bulk, RRG = R I and Lν = ν. The theoretical usefulness of Equation (9.3) will be understood when the kinetics of SFRP and ATRP are considered in Section 9.3.2. In conventional, nonliving free radical polymerization, the polymerization rate increases as the particle diameter, Dp, decreases, especially for Dp < 100 nm, due to so-called compartmentalization (Gilbert 1995). In terms of Equation (9.1), the rate increase can be explained by the increase in the radical concentration, [R•]. By separating radicals into different particles, the termination reaction between the radicals located in different particles is prohibited and the apparent (overall) termination rate constant decreases, leading to a higher radical concentration. A signiicant rate increase could be observed for the zero-one systems (Gilbert 1995) where the time fraction in which more than one radical exists in a particle can be neglected. When the exit of a radical from a particle can be neglected, the average number of radicals per particle is nR• = 0.5. In a nano-sized particle, the concentration of a single molecule becomes rather high, as shown in Table 9.1. Note that in usual free radical polymerization in bulk, [R•] is smaller than ~10 –6 mol•L–1, and a single radical concentration for Dp < 50 nm is extraordinarily high, leading to a very high polymerization rate. On the basis of Equation (9.2), Rp = R Iν, the polymerization rate increase in the conventional free radical polymerization in dispersed systems can be rationalized from the increase in the kinetic chain length ν. The bimolecular termination frequency in a particle is reduced because of the compartmentalization, which makes ν larger. In a typical zero-one system where radicals enter a particle one by one, the irst radical entry generates a poly mer radical that keeps on growing until the second radical entry. The ν-value can be made extraordinarily large in an emulsiied system. 9.3.2 SFRP and ATRP (Radical Long Life-ization Process) 9.3.2.1 SFRP and ATRP in Bulk To simplify the discussion, irst consider SFRP as an example of the radical long life-ization process, although the discussion also can be extended for

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

269

ATRP in a straightforward manner. The reversible reaction of SFRP is shown in Figure 9.2, and other elementary reactions are the same as conventional free radical polymerization. Figure 9.1 shows the history of chain growth, but it can also represent the history of a radical for SFRP and ATRP. To further simplify the discussion, let us neglect the initiation reaction. After a long dormant period, a radical is generated by the activation reaction whose rate constant is k1. The average time required to generate a radical from a PX is tPX = 1/k1, which is constant during polymerization in SFRP, assuming a constant reaction temperature. For a given polymerization time, tP, the average number of active periods is given by nA = tp/tPX , which is also a constant. The total number of monomeric units added during the polymerization time tp is equal to nA × Pn,SA, where Pn,SA is the average number of monomeric units added during a single active period. Therefore, the polymerization rate for a given initial dormant concentration [PX]0 is determined solely by Pn,SA, which is represented by: Rp ∝ Pn,SA = k p [M]tR• =

k p [ M] k 2 [ X]

(9.4)

Equation (9.4) shows that the polymerization rate is inversely proportional to the trapping agent concentration [X]. Let us generalize the discussion in the context given by Equation (9.3) proposed earlier. Imagine an active radical as a lashlight. When a radical is generated by the activation reaction, the light is on for about tR• = 10 –4 – 10 –3 s. After that there is a very long dark period for tens (or hundreds) of seconds (=tPX). During this long time interval, lots of lashlights are on in various locations. When the observed radical is active again and lashes, it would be very dificult for the observer to recognize that this is the very radical that the observer was watching. If the dark period (tPX = 1/k1) is much larger than the time required to add a monomeric unit (tM = 1/(kp[M])), the memory is practically lost and it would be reasonable to consider that the activation reaction is essentially the radical generation rate R RG in Equation (9.3). The activation reaction can be considered as the pseudo-initiation, and the deactivation reaction as the pseudo-termination, as shown in Table 9.2. TABLE 9.2 Pseudo-Initiation and -Termination in the CLRP Based on the Radical Long Life-ization Process SFRP k

Pseudo-initiation

1 Pi X  → R•i + X

Pseudo-termination

2 R•i + X  → Pi X

k

ATRP k

1 Pi X + Y  → R•i + XY

k

2 R•i + XY  → Pi X + Y

270

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

The livingness in the long life-ization process results from the fact that the rate of pseudo-initiation and -termination are much more signiicant than the genuine initiation and termination, respectively. For SFRPs, the radical generation rate in the reaction locus, RRG, is given by: RRG = RI + k1[PX] ≅ k1[PX]

(9.5)

and the average number of monomeric units added to the generated radical until the chain stops growing, Lν, is given by: Lν =

k p [ M] kt [R • ] + k 2 [ X]



k p [ M] = Pn,SA k 2 [ X]

(9.6)

Therefore, the polymerization rate, Rp, is given by: Rp = RRG Lν = k1[PX]

k p [ M] = k1[PX]Pn,SA k 2 [ X]

(9.7a)

or Rp = k p [M]K

[PX] [ X]

(9.7b)

where K ≡ k1/k2. Equation (9.7a) agrees with Equation (9.4) developed earlier. Normally, [X] 1, the difference in the number of monomer molecules among particles nM,i can be neglected, at least for low conversion regions. What is important here is the variation of nX,Act. The rate increase because of the number luctuation of trapping agents in a particle was named the luctuation effect (Tobita 2007). The rate increase due to the luctuation effect can be assessed by considering the following ratio:

Ω=

1 Nt

Nt

∑1 n

X , Act , i

i=1

1 nX , Act

(9.16)

Mathematically, the ratio Ω is always larger than unity, or equal to unity for a uniform distribution. Let us consider a simple example. Suppose we have two systems, both with nX,Act = 3. All particles contain three trapping agents in system 1, while 50% of particles contain one trapping agent and 50% contain ive trapping agents in system 2. In this case, the Ω-value is given by: Ω=

0.5 × 1 1 + 0.5 × 1 5 0.6 = = 1.8 13 13

(9.17)

Equation (9.17) illustrates that system 2 shows an 80% faster poly meri zation rate than system 1. This simple example shows that the luctuation in nX,Act can accelerate the poly mer i zation rate. Note, however, that “acceleration” in the present context is that Rp is larger than the system having the same number of nX,Act, not compared with the corresponding bulk poly mer i zation.

281

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

2.2 = 0.047



2.0 1.8

1 0.5 0.1

= 0.211

1.6 1.4 1.2 Ω = 1.1 1.0

0

5

10 – nX,Act

15

20

FIGURE 9.13 Estimated rate increase due to the luctuation in the number of trapping agents among particles, as a function of the average number of trapping agents during the active period nX,Act. In the igure, ξ = k2/kt.

To quantify the acceleration due to the luctuation effect, one needs to determine the distribution function of the trapping agents among particles. Of course, the MC simulation can provide the distribution function. However, the MC method may not be a convenient tool for everybody. To roughly assess the degree of acceleration, a simple model for the number distribution was proposed (Tobita 2007). Note that the model is not exact but based on rather simpliied assumptions. However, it could be used to roughly estimate the condition where the luctuation effect becomes important. In addition, the exit of the trapping agent from the particle did not change the distribution proile signiicantly in a preliminary investigation (Tobita 2007). Figure 9.13 shows the relationship between Ω and the average number of free trapping agents during the active period nX,Act . The luctuation effect can be neglected for large values of nX,Act , and it becomes signiicant as the particle size is reduced. In the igure, ξ = k2/kt , and ξ is usually smaller than unity. As ξ becomes smaller, there is a tendency to cause bimolecular termination more frequently, which makes the distribution of nX,Act broader, leading to a more signiicant acceleration. For 0.5 < ξ < 1, the luctuation effect becomes signiicant when nX,Act is smaller than about 10. In the case of Figure 9.10, nX,Act/ (vpNA) ≌ [X]bulk. Let the threshold nX,Act-value below which the acceleration is signiicant be nf . The particle diameter below which the luctuation effect becomes important, Dp,Fluct , can be estimated from the following equation:   6n f  Dp , Fluct =   πN A [ X]bulk 

1 3

= n f Dp ,SMC 13

(9.18)

For Dp,SMC < Dp < Dp,Fluct , an acceleration window (Tobita 2007), a particle diameter range for which the polymerization rate is larger than that for bulk polymerization, may exist.

282

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Dp,SMC

Dp,Fluct

2 1 Rp/Rp,bulk

6 4

Slope = 3

2

Fluctuation effect

Single molecule conc. effect

0.1 6 4

Acceleration window

8

9

10

3

2

4

5

6

7

Dp (nm) FIGURE 9.14 Estimated values of Dp,SMC and Dp,Fluct for the SFRP system shown in Figure 9.6 with [PX]0 = 0.2. (Modiied from H. Tobita, Macromolecular Theory and Simulation 16: 810–23, 2007.)

Under the calculation condition shown in Figure 9.6, ξ = 0.211. From Figure 9.13, nf is about 13 for ξ = 0.211. Figure 9.14 shows the calculated acceleration window for [PX]0 = 0.2 shown in Figure 9.6. The acceleration window reasonably applies for the present reaction condition. Note that Ω shows the relative rate increase compared with the cases without the statistical variation, and not necessarily compared with the bulk polymerization. As discussed earlier (Tobita 2007), there are cases in which the [X]Act-value in a particle is larger than [X]bulk . In such cases, the polymerization rate certainly becomes larger compared with the system without the statistical variation; however, Rp could be smaller than the corresponding bulk polymerization. Figure 9.15 shows the comparison for the cases with ATRP, shown in Figure 9.7. In this case, ξ = 0.047, and nf is about 20. The acceleration window agrees reasonably well for all [PX]0 conditions. Figure 9.13 shows that the maximum rate increase is predicted to occur at nX,Act = 5.3 for ξ = 0.047. (However, in ATRP shown in Figure 9.2, X corresponds to XY.) Zetterlund, Kagawa, and Okubo (2009) reported that the number of trapping agents at the maximum polymerization rate is about ive, which also agrees satisfactorily with the present luctuation theory. 9.3.3 RAFT and DT 9.3.3.1 RAFT and DT in Bulk The chain formation process in RAFT and DT was schematically shown in Figure 9.3. In this mechanism, the radical long life-ization is not required. The generated radicals are terminated after causing a large number of chain

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

283

10

Rp/Rp,bulk

1

0.1

Acceleration window

0.001

[PX]0 mol/L

Slope = 3

0.01

89

10

2

3

0.0435 0.00435 0.00087

4

5 6 7 89 100 Dp(nm)

2

FIGURE 9.15 Estimated acceleration window for the ATRP system shown in Figure 9.7.

transfer reactions. The time fraction of the active radical period is ϕA = 1 for DT, and does not have to be very small for RAFT. Note that the condition ϕA > tM. Similarly to the SFRP and ATRP process discussed in Section 9.3.2.1, the memory is practically lost, and pseudo-initiation and pseudo-termination dominate the kinetics. This is the case of another radical long life-ization process. Therefore, the Lν-value is represented by: Lν = Pn,SA =

k p [ M] (for the SF model in RAFT) k 2 [ XP]

(9.22)

The SF model in RAFT belongs to the radical long life-ization process. Equations (9.21) and (9.22) are the key equations to clarify the differences in the kinetics of two types of models in dispersed media. 9.3.3.2 RAFT and DT in Dispersed Systems For DT and the IT model for RAFT, the kinetic chain length ν becomes larger as the particle size decreases, and therefore the polymerization rate Rp is expected to become larger as the particle size is made smaller. Figure 9.16 shows that the polymerization rate increases signiicantly by reducing the particle size (Tobita and Yanase 2007; Tobita 2008b, 2009a). In Figure 9.16, the monomer:initiator ratio is kept constant in order to focus attention on the effect of particle size. The radical generation rate is chosen 1.0 Dp = 50 nm 75 nm

Conversion; x

0.8

100 nm

0.6 0.4

150 nm

0.2

300 nm Bulk

0.0 0

2000

4000

6000

8000

10000

Time (s) FIGURE 9.16 Conversion development calculated for the IT model in RAFT poly merization. The kinetic parameters used are shown in Table 9.3. (Reprinted from H. Tobita and F. Yanase, Macromolecular Theory and Simulation 16: 476–88, 2007. With permission.)

286

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

2.0×10–2

Bulk 50 nm 75 nm 100 nm 150 nm

Conversion; x

1.5

1.0

0.5

0.0

0

2000

4000

6000

8000

10000

Time (s) FIGURE 9.17 MC simulation results showing the effect of particle size on the conversion development for the SF model in RAFT poly merization. The kinetic parameters used are shown in Table 9.3. (Reprinted from H. Tobita and F. Yanase, Macromolecular Theory and Simulation 16: 476–88, 2007. With permission.)

to make the average time interval between radical entry te = 50 s for Dp = 100 nm. On the other hand, the SF model in RAFT belongs to the radical long lifeization process, similarly to SFRP and ATRP. In RAFT, the RAFT concentration [XP] is large compared with the trapping agent concentration in SFRP and ATRP, and the number of the RAFT molecules in a particle is rather large. Therefore, the SMC effect, as well as the luctuation effect, is insigniicant. The Lν-value inside a nano-sized reaction locus represented by Equation (9.19) is essentially the same as for the bulk polymerization, and the polymerization rate is expected to be the same as for the bulk polymerization. The MC simulation results for the SF model are shown in Figure 9.17 (Tobita and Yanase 2007). The conversion development is essentially the same as bulk polymerization irrespective of the particle size, within the statistical errors. The present simulation results show that the miniemulsion polymerization experiment is a convenient method to discriminate the IT and SF models. If the polymerization rate increases signiicantly by reducing the particle size, the k1-value must be large enough, and the IT model might be an appropriate model to choose. 9.3.2.2.1 Monomer-Concentration-Variation (MCV) Effect in Normal Lifetime Process As clariied in the previous section, an increase of the polymerization rate by reducing the particle size is an important feature of the normal lifetime

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

287

TABLE 9.4 Average Time Interval between Radical Entry to a Particle Dp [nm] te [s]

30 1852

50 400

100 50

150 14.8

300 1.85

process. On the other hand, it will be clariied in this section that the polymerization rate may be reduced for very small particles because of the statistical variation of the monomer concentration among particles (Tobita 2009a). This type of rate reduction applies to the IT model in RAFT, DT, and the conventional free radical polymerization in miniemulsion. In the present MC simulation condition, the monomer:initiator ratio is kept constant, and the average time interval between radical entry te used for the simulation is shown in Table 9.4. When the particle size is small, the radical entry interval is large, and the transient behavior of the radical concentration development in the earlier stage of polymerization cannot be neglected. For a RAFT polymerization, a growing radical experiences two different states, R• and PXP. The sum of the numbers of R• and PXP is represented by n (= nR• + nPXP). Assuming a zeroone system, the time development of the average number of n per particle, n, is given by (Tobita 2009a): nR• = e −t

te

( )

sinh t te ⇒ n = e −t

te

( )

sinh t te

(9.23)

Figure 9.18 shows a comparison with the MC simulation results for Dp = 50 nm, which shows an excellent agreement. Note that the n-development is not affected by the initial RAFT concentration because both R• and PXP can 0.5 0.4 0.3 – n

– – n– = e –t/te sinh(t/te)

0.2

MC Simulation (5000 particles) [XP]0 = 0

0.1 0.0

[XP]0 = 4×10–2 0

500

1000 Time (s)

1500

2000

FIGURE 9.18 Time development of n for Dp = 50 nm. The theoretical calculation is based on Equation (9.23).

288

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

be terminated by an entered radical. The case without RAFT agent, [XP]0 = 0, corresponds to the DT polymerization and also to the conventional free radical polymerization. The conversion development of polymerization in the emulsiied system is normally given by: R p = k p [ M]

nR• φ A nφ A ⇒ R p = k p [ M] Vp N A Vp N A

(9.24)

In Equation (9.24), the time fraction of the active radical period ϕA is needed because the monomer addition occurs during that fraction of time. For DT and the conventional free radical polymerization, ϕA = 1. By substituting Equation (9.23) into Equation (9.24), the conversion development would be obtained. Figure 9.19 compares Equation (9.24) and the MC simulation results for Dp = 50 nm. The polymerization rate is clearly slower than that calculated from Equation (9.24). For the particles with Dp > 100 nm, such rate reduction was not observed. Figure 9.20 shows the time development of conversion, x in each polymer particle for Dp = 50 and 100 nm, for [XP]0 = 0, that is, the conventional miniemulsion polymerization without using RAFT agent or DT polymerization. The smooth curves are the average of all particles simulated, which corresponds to the conversion-time curve that can be obtained in an experiment. [XP]0 = 0

1.0

[XP]0 = 4×10–2

Conversion

0.8 0.6 0.4 MC simulation

0.2 0.0

Rp = kp[M]nR•φA/(VpNA) 0

500

1000

1500 Time (s)

2000

2500

3000

FIGURE 9.19 Conversion development for Dp = 50 nm. The poly merization rate is slower than the calculation based on Equation (9.24), assuming the same monomer concentration for all particles.

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

289

1.0

Conversion

0.8

Dp = 50 nm

0.6

Average

[XP]0 = 0

0.4 0.2 0.0

0

500

1000 Time (s)

1500

2000

1.0

Conversion

0.8

Dp = 100 nm

0.6

Average

[XP]0 = 0

0.4 0.2 0.0

0

1000

2000 3000 Time (s)

4000

5000

FIGURE 9.20 Conversion development in each poly mer particle obtained in the MC simulation for Dp = 50 and 100 nm, with [XP]0 = 0.

When the particle size is small, the statistical variation in the conversion development among particles is signiicant. Equation (9.24) tacitly assumes that the monomer concentration is the same for all particles. Such an assumption might be reasonable for large Dps, but not so for smaller particles, as shown in Figure 9.20. Note that the monomer transfer among particles may not be neglected in real systems, depending on the types of monomer used. When the monomer is transferred from one particle to another, the monomer concentration would be equalized, and therefore the present simulation results show the extreme cases where the monomer transfer is totally absent. As shown in Figure 9.20, some particles are almost dried out while the other particles have not started polymerization. In the radicals that exist in the particles with high conversion levels, the polymerization rate is much slower than that expected from the average monomer concentration, which may slow down the overall polymerization rate.

290

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

For the zero-one system without monomer and radical exit, the analytical solution for the conversion development that accounts for the monomerconcentration-variation (MCV) effect is given by Tobita (2009a): ∞

x(t) = 1 −

∑ y (t)

(9.25)

i

i=0

λ y 2 i (t) =    η

2i

(−1)

i

i!

{ (

)

(

e − λt U −i, 1 − 2i, −ηt − ie − ηtU 1 − i, 1 − 2i, ηt

)}

(9.26)

(for i = 0, 1, 2, …) λ y 2 i+1 (t) =    η

2 i+1

(−1) ⋅ (ηt) i

i!

i+(1/2 )

π

e

−λ+( η/2 ) t

{(−1)

i+(1/2 )

(

)

(

K i+(1/2) −ηt/2 − K i+(1/2) ηt/2

(for i = 0, 1, 2, …)

)}

(9.27)

where U(a,b,z) is the conluent hypergeometric function of the second kind, Kν(z) is the modiied Bessel function of the second kind, λ ≡ 1/te, and η ≡ k p ϕA/(v p NA). Figure 9.21 compares the conversion development x from Equations (9.25)–(9.27) and the MC simulation results. The theoretical calculation results [XP]0 = 0

1.0

Conversion

0.8

[XP]0 = 4×10–2

0.6 0.4 MC simulation MCV model

0.2 0.0

0

500

1000

1500

2000

2500

3000

Time (s) FIGURE 9.21 Conversion development for Dp = 50 nm. The dotted curves are calculated from Equations (9.25)–(9.27). The statistical variation of monomer concentration needs to be accounted for in order to rationalize the MC simulation results.

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

291

1.0 Dp = 50 nm

75 nm

0.8 Conversion

30 nm 100 nm

0.6 0.4

150 nm

0.2

300 nm Bulk

0.0

0

2000

4000

6000

8000

10000

Time (s) FIGURE 9.22 Effect of particle size on the poly merization rate for the RAFT poly merization with [XP]0 = 4 × 10 –2 mol/L. (Reprinted from H. Tobita, Macromolecular Theory and Simulation 18: 108–19, 2009. With permission.)

agree perfectly with the MC simulation. The MCV effect also may need to be considered for the conventional miniemulsion polymerization with smaller particle sizes. The rate reduction due to the MCV effect is expected to be signiicant for the smaller particles. Figure 9.22 shows the conversion development for the RAFT polymerization with the IT model, which is the same as Figure 9.16 except that the MC simulation results for Dp = 30 nm are added. Because the rate reduction by the MCV effect is very signiicant for Dp = 30 nm, the polymerization rate is smaller than that for Dp = 50 nm in the present calculation condition. The MCV effect may play a role also in microemulsion poly merization, whose particle size is even smaller. 9.3.2.2.2 Retardation of RAFT in the Intermediate Termination Model In this section, the RAFT polymerization whose t PXP is small (t PXP 1, which means that CLRP always leads to a broader distribution than the Poisson distribution. Equation (9.45) shows how to produce narrow MWD through CLRP. The most dominant factor is nA . By increasing the number of active periods, the MWD becomes narrower, and approaches unity irrespective of the p-value. On the other hand, for a given nA -value, the MWD is narrower for a larger p-value. It is sometimes stated that a higher number of monomer units added per activation-deactivation cycle leads to a broader distribution, but such a statement is wrong, as clearly shown in Equation (9.45). A higher number of monomer units added per activation-deactivation cycle is preferable to obtain a narrower MWD for a ixed value of nA as long as the bimolecular termination reaction and other chain transfer reactions are suppressed effectively. For a RAFT polymerization, the possible repetition of the active periods makes the P-value larger than p, while the nAoverall -value is about one-half of the cases without repetition nA , such as DT polymerization. Figure 9.25 shows a comparison with the cases without the repetition (Tobita 2008a). The repetition of the active periods forms clusters of active periods, and such clustering makes the MWD slightly broader than the cases without the repetition. The fraction of dead polymer chain must be made small in CLRP. Other than the bimolecular termination, chain transfer reactions lead to the formation of dead polymer chains. Within various types of chain transfer reactions, the monomer transfer reactions can never be totally suppressed. The maximum number-average chain length attainable in free radical poly merization is limited by the monomer transfer reaction, and Pn, Max = 1/Cm , where

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

50

3

40 30

2.5 W(log10r)

297

RAFT Process

20

No repetition of active radical periods

– nAOA = 10

2 1.5 1 0.5 0.25

0.5

0.75

1

1.25

1.5

log10 r FIGURE 9.25 Example of the calculated weight fraction distribution development of RAFT polymers (solid curves). Comparison is made with the cases without the repetition of the active radical periods. (Reprinted from H. Tobita, Macromolecular Theory and Simulation 18: 120–26, 2009. With permission.)

Cm is the monomer transfer constant. A simple calculation method to account for the effect of monomer transfer reaction on the MWD proile was proposed in Tobita (2006a). To obtain a narrow distribution with high livingness, it is better to design Pn smaller than about 5% of Pn, Max (Tobita 2006a). The fundamental MWD function represented by Equation (9.41) is useful to consider with regard to the basic characteristics of the MWD formed in CLRP. On the other hand, the three assumptions discussed earlier need to be modiied to calculate the MWD formed in more realistic conditions. The full MWD proiles were obtained by numerically solving an ininite set of differential equations using a commercial package (Predici®), and also by applying an MC simulation method. For the CLRP in dispersed systems, the MC method is the most powerful tool for the simulation. 9.4.2 SFRP and ATRP in Dispersed Systems Figure 9.26 shows the calculated MWD proiles for the SFRP whose conversion development was shown in Figure 9.8. Because the polymerization rate is the largest at Dp = 50 nm within the particle sizes shown in Figure 9.8, the MW is the largest at Dp = 50 nm. Figure 9.27 shows Dp polydispersity index (Pw /Pn ) of the MWD shown in Figure 9.26. The polydispersity index is the largest at which the polymerization rate is the largest, that is, at Dp = 50 nm. The broadening of the MWD can be rationalized on the basis of the luctuation theory. The main reason for the rate increase is the number luctuation of trapping agents, and therefore, the conversion development is different among particles, as was shown in

298

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

2.5

W(log10r)

2.0 1.5 1.0

t = 10,000 [s]

Dp[nm] 50 75 100 150

0.5 0.0 0.5

1.0

1.5

2.0

2.5

log10r (a) 2.5

W(log10r)

2.0 1.5 1.0

t = 5,000 [s]

Dp[nm] 50 75 100 150

0.5 0.0 0.5

1.0

1.5

2.0

2.5

log r 10

(b) FIGURE 9.26 Calculated weight fraction distribution of the SFRP polymers formed in miniemulsion poly merization at t = 5000 s (a) and 10,000 s (b), whose conversion development is shown in Figure 9.8. (Reprinted from H. Tobita and F. Yanase, Macromolecular Theory and Simulation 16: 476–88, 2007. With permission.)

Figure 9.11. The MWD proiles formed in each particle are different signiicantly when the luctuation effect is important. Therefore, the polydispersity index is the largest at Dp = 50 nm where the luctuation effect accelerates the polymerization rate most signiicantly. The acceleration window could be used to increase the productivity; however, the MWD becomes broader because of the luctuation effect at that particle size region. 9.4.3 RAFT Polymerization in Dispersed Systems (Normal Lifetime Process) In this section, the MWD formed in the IT model whose kinetic parameters are shown in Table 9.3 is considered.

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

299

1.5 Bulk

PDI

1.4

t = 5,000 [s]

1.3 1.2 t = 10,000 [s]

1.1 50 nm 1.0

40

60

80

100

120

140

160

Dp (nm) FIGURE 9.27 The polydispersity index (=Pw/Pn) of the MWDs shown in Figure 9.26. (Reprinted from H. Tobita and F. Yanase, Macromolecular Theory and Simulation 16: 476–88, 2007. With permission.)

9.4.3.1 Monomer-Concentration-Variation (MCV) Effect In Section 9.3.3.2.1, it was shown that the MCV effect may slow down the polymerization rate for small particle sizes because the time interval between radical entry tends to become larger for small Dps. When the MCV effect is signiicant, the MWD formed in different particles may differ signiicantly, leading to a broad distribution. One of the methods to prevent the broadening is to increase the radical entry frequency, which reduces the necessary transient time to reach the steady state radical concentration in the particle. (See Figure 1 of Tobita 2009b.) The problem with this hasty method is that the effect of dead polymer chains becomes signiicant especially at higher conversions. Another simpler way to obtain a narrow MWD is a wait-and-see method. The MWD becomes narrower as the number of active periods increases, as shown in Figure 9.28. 9.4.3.2 Effect of Particle Size on the MWD Figure 9.29 shows the effect of particle size on the MWD at the inal conversion level (x > 0.99). A signiicant amount (weight) of dead poly mer chains is formed in larger particles. These dead polymer chains are large in weight, but small in the number, and therefore these peaks cannot be found if the distribution is plotted on the number basis (Tobita 2009b). Mainly because of a signiicant dead polymer chain formation in the larger particles, the MWD is narrower for smaller particles. The chain lengths of dead polymers are about twice and three times as large as the main peak polymers at the inal conversion level. Because the main peak keeps on moving to larger chain lengths as the polymerization

300

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

12

W(log10r)

10

Dp = 50 nm t[s] 1000 5000

8 6 4

er

w e ro m ar ti N ith w

Broadened by the MCV Effect PDI = 1.39

2 0

PDI = 1.02

Conv. 55.2% 99.1%

1.6

1.8

2.0

2.2

2.4

log10r FIGURE 9.28 Calculated MWD development for the RAFT miniemulsion poly merization with the IT model for Dp = 50 nm. The MWD broadening due to the MCV effect weakens as the time goes by.

12

W(log10r)

10

Dp[nm] 50

8

100 6

150

4

300

Dead polymers

2 0

2.2

2.4

2.6

2.8

log10r 2-armed

3-armed

FIGURE 9.29 Simulated MWD for the RAFT miniemulsion poly merization with the IT model at the inal conversion levels, x > 0.99. (Reprinted from H. Tobita, Macromolecular Theory and Simulation 18: 120–26, 2009. With permission.)

proceeds, this fact shows that the dead polymer chains are formed mainly at the inal stage of polymerization. In fact, the dead polymer peaks were not observed at lower conversion levels. The reason for forming dead polymer chains in a inal stage of polymerization can be rationalized as follows. Figure 9.30 shows the conversion developments for various Dps. Because the polymerization rate is slower for larger particle cases, it takes a very long time to reach high conversion levels for larger Dps. In addition, the polymerization rate becomes very slow at higher conversion levels. Figure 9.31 shows how the number of dead polymer chains increases with time. The initiation and termination rates are balanced for both bulk and

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

1.0

301

50 100

Conversion

0.8

Very long time is needed.

150

0.6

Dp = 300 nm

0.4 0.2 0.0 0.0

0.5

1.0

1.5

2.0

2.5

3.0×105

Time (s) FIGURE 9.30 Conversion development for the RAFT miniemulsion poly merization with Dp = 50, 100, 150 and 300 nm. (Reprinted from H. Tobita, Macromolecular Theory and Simulation 18: 120–26, 2009. With permission.)

1.0×10–3

Dead Polym. Conc. [mol/L]

0.8

0.6

0.4

Bulk Polym. Dp = 30 nm

0.2

50 nm 100 nm 150 nm 300 nm

0.0 0.0

0.5

1.0

1.5

2.0

2.5

3.0×104

Time (s) FIGURE 9.31 Time development of the dead poly mer chain concentration for the RAFT poly merization with various particle sizes. (Reprinted from H. Tobita, Macromolecular Theory and Simulation 18: 120–26, 2009. With permission.)

miniemulsion polymerization. In the present series of calculations, a constant initiation rate RI is used, and the number of dead polymer chains increases linearly with time. In terms of the required reaction time, the longest time is required at the inal stage of polymerization as shown in Figure 9.30. This is why a signiicant number of dead polymer chains are formed at that stage.

302

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

A larger number of dead polymer chains are formed for larger Dps, simply because it takes a longer time to reach a high conversion level.

9.5 Summary The effect of nano-sized polymerization locus on the polymerization rate of CLRP can be rationalized conveniently by using the formula, Rp = R RGLν, where R RG is the radical generation rate in the reaction locus, and Lν is the average number of monomeric units added to the generated radical until the chain practically stops growing. SFRP and ATRP belong to the radical long life-ization process, identiied in this chapter, R RG is equal to the radical generation rate from the dormant species, and Lν is given by the average number of monomeric units added during a single active period, Pn,SA ; that is, Rp = k1[PX]Pn,SA ∝ 1/[ X] for SFRP. An important conclusion is that (1) the polymerization rate is in inverse proportion to the trapping agent concentration [X], and (2) the bimolecular termination rate does not directly change the polymerization rate. The bimolecular termination affects the polymerization rate only through the production of free trapping agents through bimolecular termination. The polymerization rate cannot be increased signiicantly through the compartmentalization of radicals, as in the case of the conventional, nonliving radical polymerization. Two important factors control the polymerization rate of SFRP and ATRP inside nano-sized polymerization locus, namely, (1) the single-moleculeconcentration (SMC) effect, and (2) the statistical luctuation in the number of trapping agents in a particle. With the SMC effect, the polymerization rate 3 RP decreases signiicantly by making the particle size smaller; that is, Rp ∝ Dp . Below the diameter Dp,SMC, the polymerization rate must be smaller than that in bulk polymerization, and the Dp,SMC-value can be determined by equating the single molecule concentration in a particle and the trapping agent concentration in bulk polymerization, [X]bulk. On the other hand, when the average number of trapping agents in a particle is smaller than about 10–20, the statistical variation among the particles becomes signiicant, which makes the polymerization rate faster, compared with the cases without luctuation. Because of the effects of SMC and luctuation, Rp may show an acceleration window, Dp,SMC < Dp < Dp,Fluct where Rp is slightly larger than that in bulk polymerization. The acceleration occurs because of the luctuation, and therefore the MWD obtained becomes slightly broader in this region. For the cases of DT and RAFT conforming to the IT model, the lifetime of a radical is in the same order of magnitude as conventional free radical polymerization, and the termination reaction does not have to be avoided. For the normal lifetime process, Lν = ν, where ν is the kinetic chain length. Because ν can be made larger by separating the radicals into different particles through

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

303

compartmentalization, the polymerization rate can be increased signiicantly by reducing the particle size, as in the case of the conventional free radical polymerization. On the other hand, the RAFT polymerization system conforming to the SF model belongs to the radical long life-ization process, and the polymerization rate is essentially the same as that in bulk. The normal radical lifetime process, including conventional free radical polymerization, may show the reduction in the polymerization rate for small particles due to the MCV effect. For RAFT, the rate retardation by increasing the RAFT concentration occurs with or without IT in a zero-one system. For RAFT following the normal radical lifetime process, smaller particles are advantageous in implementing a faster polymerization rate, narrower MWD, and a smaller number of dead polymer chains. The present theoretical model might be too simple to apply to real systems. However, it is hoped that the present theoretical investigation will provide greater insight into the kinetics of CLRP.

Acknowledgment This work was supported by a Grant-in-Aid for Scientiic Research, the Ministry of Education, Culture, Sports, Science and Technology, Japan (Grant-in-Aid 21560790).

References Barner-Kowollik, C., et al. 2001. Kinetic investigations of reversible addition fragmentation chain transfer polymerizations: Cumyl phenyldithioacetate mediated homopolymerizations of styrene and methyl methacrylate. Macromolecules 34:7849–57. ———. 2003. The reversible addition-fragmentation chain transfer process and the strength and limitations of modeling: Comment on “The Magnitude of the Fragmentation Rate Coeficient.” Journal of Polymer Science, Part A: Polymer Chemistry 41:2828–32. ———. 2006. Mechanism and kinetics of dithiobenzoate-mediated RAFT polymerization. 1. The current situation. Journal of Polymer Science, Part A: Polymer Chemistry 44:5809–31. Braunecker, W. A., and K. Matyjaszewski. 2007. Controlled/living radical polymerization: Features, developments, and perspectives. Progress in Polymer Science 32:93–146. Butte, A., Storti, G., and M. Morbidelli. 2001. Miniemulsion living free radical polymerization by RAFT. Macromolecules 34:5885–96. Cunningham, M. F. 2008. Controlled/living radical polymerization in aqueous dispersed systems. Progress in Polymer Science 33:365–98.

304

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Fischer, H. 1999. The persistent radical effect in controlled radical polymerization. Journal of Polymer Science, Part A: Polymer Chemistry 37:1885–1901. Fukuda, T. et al. 1996. Mechanisms and kinetics of nitroxide-controlled free radical polymerization. Macromolecules 29:6393–98. Gilbert, R. G. 1995. Emulsion polymerization. London: Academic Press. Goto, A., and T. Fukuda. 1997. Effects of radical initiator on polymerization rate and polydispersity in nitroxide-controlled free radical polymerization. Macromolecules 30:4272–77. ———. 2004. Kinetics of living radical polymerization. Progress in Polymer Science 29:329–85. Konkolewicz, D. et al. 2008. RAFT polymerization kinetics: Combination of apparently conlicting models. Macromolecules 41:6400–12. Kwak, Y., Goto, A., and T. Fukuda. 2004. Rate retardation in reversible additionfragmentation chain transfer (RAFT) polymerization: Further evidence for cross-termination producing 3-arm star chain. Macromolecules 37:1219–25. Monteiro, M. J., and H. de Brouwer. 2001. Intermediate radical termination as the mechanism for retardation in reversible addition-fragmentation chain transfer polymerization. Macromolecules 34:349–52. Tobita, H. 1995. Monte Carlo simulation of emulsion polymerization—linear, branched, and crosslinked polymers. Acta Polymerica 46:185–203. ———. 2006a. Molecular weight distribution of living radical polymers. 1. Fundamental distribution. Macromolecular Theory and Simulation 15:12–22. ———. 2006b. Molecular weight distribution of living radical polymers. 2. Monte Carlo simulation. Macromolecular Theory and Simulation 15:23–31. ———. 2007. Kinetics of stable free radical mediated polymerization inside submicron particles. Macromolecular Theory and Simulation 16:810–23. ———. 2008a. Fundamental molecular weight distribution of RAFT polymers. Macromolecular Reaction Engineering 2:371–81. ———. 2008b. Kinetics of controlled/living radical polymerization in emulsiied systems. Macromolecular Symposia 261:36–45. ———. 2009a. RAFT miniemulsion polymerization kinetics. 1. Polymerization rate. Macromolecular Theory and Simulation 18:108–19. ———. 2009b. RAFT miniemulsion polymerization kinetics. 2. Molecular weight distribution. Macromolecular Theory and Simulation 18:120–26. Tobita, H., and F. Yanase. 2007. Monte Carlo simulation of controlled/living radical polymerization in emulsiied systems. Macromolecular Theory and Simulation 16:476–88. Wang, A. R. et al. 2003. A difference of six orders of magnitude: A reply to “The Magnitude of the Fragmentation Rate Coeficient.” Journal of Polymer Science, Part A: Polymer Chemistry 41:2833–39. Zetterlund, P. B., Kagawa, Y., and M. Okubo. 2008. Controlled/living radical polymerization in dispersed systems. Chemical Reviews 108:3747–94. ———. 2009. Compartmentalization in atom transfer radical polymerization in dispersed systems: Effects of target molecular weight and halide end group. Macromolecules 42:2488–96. Zetterlund, P. B., and M. Okubo. 2007. Compartmentalization in TEMPO-mediated radical polymerization in dispersed systems: Effects of macroinitiator concentration. Macromolecular Theory and Simulation 16:221–26.

Effects of Nano-Sized Polymerization Locus on the Kinetics of CLRP

305

Appendix Formulation of Various Average Times First, consider the average time required for an active radical to add one monomer molecule, tM . The average number of times for the monomer addition to the active radicals in a unit volume per second is the polymerization rate, Rp mol•L–1•s–1. Therefore, the average number of times that a single active radical adds monomer molecules in a second is Rp/[R•] = kp[M] s–1. The average time required to add one monomer molecule is the inverse of this value, and is given by: tM =

1 k p [ M]

(9.A1)

Next, consider the average lifetime of an active radical, tR• , in the reversible reaction in SFRP. The average number of times the active radical is deactivated by the coupling reaction with X in a unit volume per second is k2[X][R•] mol•L–1•s–1. Therefore, the average number of times that a single radical forms PX from R• in a second is k2[X] s–1. The average lifetime of an active radical, tR• is given by: tR• =

1 k 2 [ X]

(9.A2)

Similarly, various lifetimes can be obtained. In the cases of SFRP and ATRP, the dormant period of a radical tDR and the dormant period of a chain tDC are the same. On the other hand, tDR and tDC are different for RAFT and DT. The results are summarized in Table A9.1. TABLE A9.1 Various Average Lifetimes tDR

tR• ( tM k1

1 ≈ tM k 2 [ X]

tPX =

1 >> tM k1

ATRP

tPX =

1 >> tM k1 [Y]

1 ≈ tM k 2 [XY]

tPX =

1 >> tM k1 [Y]

RAFT

tPXP =

1 ≈ tM k 2 [XP]

tXP + tPXP =

tDC

0

1 ≈ tM kex [XP]

tXP =

1 ≈ tM or >> tM k1

1 1 + >> tM k 2 [R• ] k1

1 >> tM kex [R• ]

10 Functional Polymer Particles by Emulsifier-Free Polymerization V. Mittal CONTENTS 10.1 Introduction ................................................................................................ 307 10.2 Particle Nucleation.....................................................................................308 10.3 Functional Particles by Surfactant-Free Polymerization ......................309 Acknowledgments .............................................................................................. 324 References............................................................................................................. 325

10.1 Introduction The presence of emulsiier during emulsion polymerization helps the particles to retain colloidal stability. It forms the micelles in the beginning of the polymerization, which subsequently become the loci of polymerization due to the entry of the radicals in them [1,2]. The amount of the surfactant can also be varied to achieve different rates of polymerization or different sizes of the particles, so the presence of surfactant is important in more than one way to tune the properties of the polymer particles. Therefore, the majority of the emulsion polymerization reactions are carried out in the presence of the surfactants. However, there are also some cases where the poly merization is achieved in the absence of any surfactant. Such a polymerization process is then termed surfactant-free polymerization [1,3]. This kind of polymerization is carried out when the polymer particles are required for speciic applications (like calibration of particle characterization methods or use of the particles as standards) that cannot tolerate the presence of even minor amounts of impurities on the surface of the particles or in the aqueous phase due to the desorption of the surfactant. Surfactant-free polymerization is also required when the surface charges are needed to be accurately known. Apart from that, this mode of polymerization is required when the particles are subjected to subsequent surface functionalization processes that are not compatible with the presence of the emulsiier molecules on the surface. One example is the poisoning of the copper catalyst by reaction with emulsiier 307

308

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

when carrying out particle surface functionalization by atom transfer radical polymerization (ATRP) [4,5].

10.2 Particle Nucleation The mechanism of the particle nucleation in this mode of poly merization is completely different as compared to emulsiied polymerization methods. In the case of polymerization in the presence of surfactant, the micelles nucleate the polymer particles as these provide ideal conditions for the radical to enter and subsequently initiate the polymerization. When no surfactant is present in the system, the generated radicals in the aqueous phase in the absence of micelles start reacting with the monomer dissolved in the aqueous phase. The chains so formed are colloidally unstable and keep on collapsing with each other to attain stability [1,3,6]. This mass of collapsed chains gets swollen with the monomer and continues to polymerize monomer in these particles. The negative charges from the potassium persulphate initiator moieties present at the end of the polymer chains make these ends hydrophilic, and as a result these groups are present on the surface of the particles. The presence of the charges on the surfaces helps to give colloidal stability to these particles. Figure 10.1 shows the representation of such particles stabilized by the negative charges from the initiator. Because the surfactant is absent, the number of particles reaching the stable state is generally two orders of magnitude less than the number of particles generated in conventional emulsion polymerization. The smaller particles keep on collapsing with each other owing to the colloidal instability, thereby generating particles much bigger in size as compared to the size in the emulsiied polymerizations. Also, because the number of particles is less, the time required to achieve the full conversion or higher extents of conversion is high. Figure 10.2a also shows the effect of different amounts of surfactant on the conversion of the monomer as a function of time [3]. As mentioned earlier, the conversion is much faster when the amount of surfactant is increased due to the nucleation of more particles. The effect of changing the amount of surfactant on the mean particle size is also depicted in Figure 10.2b [3]. The particle size reached near 900 nm in the absence of surfactant as compared to roughly 100 nm in the presence of 3% surfactant (based on the amount of monomer) when all other reaction components were kept the same. Figure 10.3 also demonstrates the scanning electron microscopy (SEM) micrographs of the particles generated without and with surfactant. The differences in particle sizes owing to the surfactant are clearly visible [6,7]. The reaction outcome, that is, colloidal stability and surface morphology of the particles in the case of surfactant-free poly merization, is very sensitive to the changes in the reaction parameters or components; therefore, it is

309

Functional Polymer Particles by Emulsifier-Free Polymerization



SO4



SO4



SO4



SO4 SO4–



SO4 –

SO4



SO4 –



SO4

SO4



SO4



SO4

– SO4



SO4

SO4–

SO4– SO4–

SO4–

FIGURE 10.1 Representation of poly mer particles stabilized by the sulphate ions from the initiator.

important to study these changes and their respective effects on the particles in order to tune the particles according to the requirement. It is particularly important to understand the behavior of the system for the successful functionalization of the particle surfaces by surfactant-free polymerization to generate functional latexes.

10.3 Functional Particles by Surfactant-Free Polymerization ATRP is a versatile method to achieve controlled morphologies in the polymer particles. It is also used to control the molecular weight as well as molecular weight distribution of the polymer chains in the particles [8–17]. ATRP can also be used to functionalize the polymer particles by grafting the polymers of interest from the surface of the particles. One example is the grafting of poly(N-isopropylacrylamide) from the surface of the particles to achieve thermally responsive behavior in the particles. Similarly, other polymers can be grafted on the surface to tune the properties of the polymers. To achieve such polymer grafting, it is important to functionalize the surface of the particles with an ATRP initiator irst, which can subsequently be used to

310

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

(a) 100 3% Surfactant 1% Surfactant

Conversion (%)

80

60 No surfactant 40

20

0 0

100

200

300 400 Time (min)

500

600

(b) 900

Particle Size (nm)

750 No surfactant 600 450 1% Surfactant

300 150 3% Surfactant 0

100

200

300 400 Time (min)

500

600

FIGURE 10.2 Monomer conversion and mean particle size as a function of amount of surfactant used in the poly merization. (Reprinted from V. Mittal, Advances in Polymer Latex Technology. New York: Nova Science Publishers, 2009. With permission.)

graft the chains of various polymers of interest. Because the ATRP catalyst gets poisoned by the ionic emulsiier used in the emulsion poly merization, the polymerization is required to be carried out in the absence of surfactant. There is also a possibility of using the nonionic surfactants when using ATRP method, but even with such surfactants, it is not an optimal process and colloidal stability can be a problem.

Functional Polymer Particles by Emulsifier-Free Polymerization

300 nm

311

100 nm (a)

(b)

FIGURE 10.3 SEM images of the particles generated (a) without surfactant and (b) with surfactant.

The most promising way for the controlled functionalization of the particles’ surface with ATRP initiator to achieve functional latexes is to polymerize a functional monomer (i.e., ATRP initiator carrying a terminal acrylic or methacrylic reactive group) onto the seed polymer particles in the form of a thin shell. One example of such a functional monomer is 2-(2-bromopropionyloxy) ethyl acrylate (BPOEA) reported in the literature [6,18]. The process can be achieved in two ways: In the irst, a two-step method that is based on batch polymerization, the seed is polymerized to 100% conversion of the monomer followed by polymerization of the functional monomer on the surface of the seed particles either alone or together with styrene. However, the process can also be achieved in a single step by using the semibatch conditions, where the seed is not allowed to fully polymerize and the shell forming functional monomer is added when the seed is polymerized to 60–70%. Because the surfactant-free polymerization is sensitive to the changes in reaction conditions, the different methodologies lead to different surface morphologies depending on the kinetic and thermodynamic factors [19–22]. The factors whose effect on the particles’ functionalization need to be studied are homopolymerization or copolymerization of the functional monomer, preswelling of the seed particles before functionalization, mode of addition of the shell-forming monomers, cross-linker, etc. One must be careful that the aforementioned parameters are optimized so as to achieve a uniform layer of the ATRP initiator on the surface of the particles, indicating uniform functionalization. It is also required that the amount of initiator on the surface of the particles be controlled, which can subsequently help to control the amount of polymer grafting.

312

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

300 nm

250 nm

(a)

(b)

300 nm

(c) FIGURE 10.4 (a) Fully poly merized polystyrene particles when modiied with (b) styrene and (c) functionalizing monomer on the surface.

Figure 10.4 demonstrates the irst trials to generate functional particles [6]. Fully polymerized polystyrene seed particles (Figure 10.4a) were used for these trials, and the batch polymerization route was taken for the particle functionalization. As a reference, the polystyrene seed particles were also functionalized on the surface with a thin layer of polystyrene (Figure 10.4b). First, functionalization reactions were carried out by polymerizing a thin layer of BPOEA on the surface of the particles as shown in Figure 10.4c after the surface functionalization reactions. This way, the behavior of the homopolymers toward the shell formation on the surface formation could be understood. Also, the surface morphologies obtained in both cases are quite different from each other. The particles after functionalization with styrene were smooth, and the increase in the size of the particles after the functionalization reactions also indicated corresponding diffusion of the monomer

Functional Polymer Particles by Emulsifier-Free Polymerization

313

and subsequent polymerization on the surface of the particles. However, the morphology in the case of BPOEA functionalization of the particles is more like orange peel or patchy. It was opined that the observed behavior is a result of the mismatch between the compatibility of the fully polymerized seed particles and BPOEA. It also indicated that the BPOEA chains may not be colloidally stable on their own, thus leading to the kinetic adsorption of these chains on the surface of the particles during stirring. Therefore, the diffusion of BPOEA on the surface of the particles may not be thermodynamic but more a kinetic effect. This process proceeds as a precipitation polymerization where the precipitation of the BPOEA polymer nuclei takes place on the surface of the seed particles. BPOEA chains then continue to polymerize on the surface and lead to the patchy morphology after the completion of the polymerization. Some coagulum was also observed in the aqueous phase, indicating that a part of functional monomer may also be present away from the particles, which was conirmed by the nuclear magnetic resonance investigation of the dried particles. The amount of functionalizing monomer groups present in the particles was observed to be less than originally added during polymerization. However, the uniform coverage of the particles with even a thin layer of the ATRP initiator can be enough to achieve the surface functionalization as it is only the surface initiation sites that are usable in the subsequent polymerization reactions. As another functionalization trial, styrene and BPOEA were both used to generate the functioning shell on the surface of the particles. It is also important to vary the amount of styrene as it helps to control the amount of ATRP initiator on the surface and hence subsequently the density of the grafted brushes. Figure 10.5 shows the results of the functionalization trials when both styrene and BPOEA are added during the functionalization step. Two modes of addition of shell-forming monomer can be tried: In the irst mode, styrene can be added irst to swell the preheated seed particles to some extent followed by the addition of functional monomer. In the second mode, both the monomers can be added together. It is obvious from Figure 10.5 that the morphologies of the resulting particles in both cases are quite different from each other, indicating the impact of monomer addition protocols on the inal outcome. In both cases, secondary nucleation of the particles was observed. The morphology of the particles in the irst case (as shown in Figures 10.5a and b) is more like a strawberry type, whereas the particles in the second case are smoother at the surface. To further understand the polymerization behavior of BPOEA in the presence of styrene, polymerization of particles with 1:1 styrene:BPOEA and 3:1 styrene:BPOEA molar ratios were also separately carried out. The system with equimolar ratio of the monomers was slow to polymerize and generated only very small particles (Figure 10.6a) and coagulum, whereas the system with higher amounts of styrene led to the fast generation of larger and stable particles (Figure 10.6b), indicating that monomer composition has signiicant effects on polymerization rate as well as formation of particles. It can

314

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

250 nm

300 nm (a)

(b)

250 nm

500 nm (c)

(d)

FIGURE 10.5 (a) SEM and (b) TEM (transmission electron microscopy) images of the functionalized particles where styrene was added before the functionalizing monomer in the shell-forming batch; (c) SEM and (d) TEM images of the functionalized particles where styrene and the functionalizing monomer in the shell-forming batch were added together.

also be concluded from Figure 10.6 that the presence of styrene promotes the formation of stable particles by homogenous nucleation. Thus, the results in Figure 10.5 can be explained in light of these indings. The differences in the morphology of the particles probably originate from the initial stages of the functionalization reaction. When styrene is added irst to the system, it can be expected to swell the polystyrene seed; thus, the overall styrene:BPOEA mole ratio is lower than the originally added amount. This explains the colloidal instability of the copolymer particles formed by the reaction of styrene with BPOEA, leading to their sticking or coagulation with the seed latex and thus generating the strawberry morphology. On the other hand, when both the monomers are added together, the formed copolymer chains owing to the higher amount of styrene present in the system can quickly form stable

Functional Polymer Particles by Emulsifier-Free Polymerization

250 nm (a)

315

300 nm (b)

FIGURE 10.6 SEM images of the particles demonstrating the copoly merization of styrene and BPOEA in the absence of any seed particles using styrene to BPOEA mole ratios of (a) 1:1 and (b) 3:1.

secondary particles. On the other hand, the higher amount of styrene present also allows more compatibility of these copolymer chains with the seed particles. This results in one part of the monomer mixture forming the secondary particles and the other part forming a uniform layer on the surface of the seed particles. This is also conirmed by the increase in the size of the seed particles after the functionalization reaction. Generation of secondary nucleation is not optimal as it leads to nonuniform distribution of the ATRP initiator in the particles, which then subsequently affects the quality of the polymer grafts from the surface of the particles. It should also be noted that the time provided to the shell-forming monomers to equilibrate with the seed particles may also have an effect on the inal particle morphology. In the aforementioned trials, the time provided for these monomers to swell the seed was 30 minutes, which may not be suficient owing to the very hydrophobic nature of BPOEA, thus requiring a greater amount of time for swelling the seed particles. To analyze such effects, the seed particles were added with either styrene or the functional monomer BPOEA, or a mixture of styrene and BPOEA, and were allowed to stir at room temperature overnight in order to better swell the seed particles with the monomers. Figure 10.7 shows the results of such trials. The resulting particle morphologies were similar to the earlier case, indicating that the amount of time provided to the system did not impact the resulting morphology in this case. Functionalization reactions in the semibatch mode of monomer addition can also be carried out. In this process, which consists of a single step, the seed particles are not allowed to polymerize fully, and the shell-forming monomers are added to the system at certain conversion of the seed monomers (e.g., 70%). Two different possible types of the addition of shell-forming

316

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

300 nm

300 nm

(a)

(b)

300 nm

500 nm (c)

(d)

FIGURE 10.7 (a) SEM image of the particles obtained after overnight swelling of the seed particles with styrene and subsequent poly merization, (b) SEM image of the particles obtained after overnight swelling of the seed particles with BPOEA followed by polymerization, (c) TEM image of the particles obtained after overnight swelling of the seed particles with BPOEA and styrene in the mole ratio of 1:3 followed by poly merization, and (d) SEM image of the particles shown in (c).

monomers can be analyzed: shot addition or delayed addition. In the shot addition mode, the whole amount of BPOEA with or without styrene was added to the seed as a shot, whereas in the delayed-addition mode, the monomers forming the shell are added by a syringe in a delayed fashion. Shot addition adds the monomer at a rate that is higher than the polymerization rate, whereas delayed addition adds the monomers at a rate that may be comparable to or smaller than the polymerization rate. However, in the case of emulsiier-free polymerization, even delayed addition may have some buildup of the monomer in the system owing to the very slow polymerization in the absence of emulsiier. Figures 10.8 and 10.9 represent these results of shot and delayed addition of the monomers to the partially polymerized

317

Functional Polymer Particles by Emulsifier-Free Polymerization

200 nm

500 nm (a)

(b)

500 nm

300 nm (c)

(d)

FIGURE 10.8 (a) and (d) TEM images of the partially polymerized seed particles; (b,c) TEM and SEM images of functionalized particles obtained by the shot addition of BPOEA to the seed particles and subsequent polymerization; and (e,f) TEM and SEM images of functionalized particles obtained by the shot addition of a mixture of styrene and functionalizing monomer followed by polymerization.

seed particles. The shot addition of the shell-forming monomers to the seed polymerized to 70% led to the generation of smooth and fairly monodisperse particles, and there was no secondary nucleation of the shell-forming monomers. A smooth surface also indicates a uniform distribution of the ATRP initiator on the surface of the particles. The increase in the size of the particles as compared to the partially polymerized seed particles also indicates the incorporation of the shell-forming monomers on the surface of the particles. The behavior is completely different from the case when fully polymerized seed was used, indicating that the compatibility of the seed particles toward the copolymer chains can also affect the inal morphology of the particles. The fully polymerized seed may be considered to be more compact than the

318

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

500 nm

300 nm (e)

(f )

FIGURE 10.8 (continued).

partially polymerized seed swollen or enriched with the remaining monomer. The delayed addition of the shell-forming monomers, on the other hand, led to the generation of varying amounts of secondary particles depending on whether the functional monomer was added alone or was added in the presence of styrene. This behavior is signiicantly different from the shot addition, conirming the earlier notion that the monomer addition methodology has a signiicant effect, especially on the polymerization reactions carried out in the absence of surfactant. The behavior of pure BPOEA is also different from the case when the fully polymerized seed was used. The generated particle morphologies in the case of shot addition of the monomers to the seed have the following factors affecting them. First, the presence of unreacted styrene owing to the incomplete polymerization of seed particles causes the effective ratio of styrene as compared to BPOEA to increase. Also, because the particles are already swollen with styrene, monomer from the second batch can partition at the interphase much faster. As mentioned previously, the partially polymerized seed particles can also be considered to be “softer” as compared to the fully polymerized seed particles, thus improving their compatibility with the polymer chains formed by the second batch of monomers. In the case of shot addition of BPOEA to the system, because the formed polymer chains from BPOEA are colloidally unstable, they keep on collapsing with the seed particles. Because the seed particles, owing to their softer surface, favor the coalescence or adsorption of the nuclei produced by homogenous nucleation, the seed surface is more compatible with the collapsing BPOEA chains, and thus smoother particle surfaces are formed in the end. Similarly, when BPOEA is added together with styrene, similar monomer partitioning at the interphase happens, which

319

Functional Polymer Particles by Emulsifier-Free Polymerization

500 nm

300 nm (a)

(b)

1000 nm

250 nm (c)

(d)

FIGURE 10.9 (a) SEM and (b) TEM images of the functionalized particles obtained by the delayed addition of BPOEA to the partially poly merized seed particles and subsequent poly merization; and (c) SEM and (d) TEM images of the functionalized particles obtained by the delayed addition of a mixture of styrene and functionalizing monomer followed by poly merization.

occurs at a much faster rate than the monomer polymerization. Thus, most of the styrene added in the secondary batch of monomers is used to swell the seed particles, causing BPOEA to initiate polymerization alone; and owing to the similar reasons mentioned earlier, these chains intermix with the seed particles, generating the smooth particles. In the case of delayed addition of monomers to the seed particles, the morphology is affected by the effective mole ratio of the monomers at a particular time during the course of

320

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

polymerization. Slow addition of monomers to the polymerization medium leads to the ratio of styrene to BPOEA being higher than in the case of shot addition, as all of the added styrene is not quickly taken by the seed particles to swell them. Therefore, this increased amount of styrene in the system leads to the quick generation of the stable secondary particles. Thus, the difference in the case of both shot and delayed addition is the competition between the rate of coalescence of the copolymer chains or nuclei to the seed and the rate of growth of the secondary particles to stability. In order to again underline the sensitivity to the reaction conditions in the absence of the surfactant, particles synthesized in the emulsiied conditions were compared with the particles generated by emulsiier-free polymerization. The homopolymer particles have already been compared in Figure 10.3, but the comparison of the functionalized particles generated with or without the surfactant is also required to gain further insights into the particles’ morphology changes. Figures 10.10a and b show the particles generated without the addition of surfactant for comparison with the particles synthesized in the presence of surfactant shown in Figures 10.10c and d. The particles in the nonemulsiied conditions are the same as shown in Figure 10.5 and were synthesized by the addition of a batch of styrene and BPOEA to the fully polymerized seed particles. The differences between the particle morphologies in the two systems are quite visible. The seed particles in the emulsiied polymerization were smooth (Figure 10.10c), similar to the seed particles in emulsiier-free polymerization conditions. However, the functional particles modiied by a copolymer layer of styrene and BPOEA in the emulsiied polymerization conditions are also smooth in nature in contrast to the strawberry morphology in the case of emulsiier-free polymerization. Also, owing to the presence of the surfactant, the generated particles are much smaller in size. The particles in Figure 10.10d were functionalized by the addition of the shell-forming monomers to the fully polymerized seed particles. Other trials to achieve the functionalization of the particles by other modes like semibatch polymerization, shot addition, delayed addition, addition of BPOEA either alone or with styrene, etc., in the presence of emulsiied conditions also led to particles similar to those in Figure 10.10d. These particles are demonstrated in Figure 10.11 [7], clearly indicating that the presence of the emulsiier eliminates most of the dificulties associated with the emulsiier-free polymerization reactions. Though the functional polymer particles with the thin shell of polymer containing ATRP initiator could easily be formed in the emulsiied polymerization conditions, it should be noted that the presence of surfactant adsorbed on the particles would interfere with the functioning of the ATRP imitator when used later onto graft-controlled polymer brushes from the surface. One possibility is to desorb the polymer particles by successive washing, but this may generate colloidal instability in the particles. Thus, it is justiied to approach such functionalization by following the surfactant-free polymerization conditions.

Functional Polymer Particles by Emulsifier-Free Polymerization

250 nm

300 nm (a)

200 nm

321

(b)

100 nm (c)

(d)

FIGURE 10.10 Comparison of particles generated without and with surfactant: (a,b) particles generated without surfactant and functionalized with a copolymer of styrene and functionalizing monomer, (c) pure seed polystyrene particles generated by the use of surfactant, and (d) particles of (c) functionalized with a copolymer of styrene and functionalizing monomer.

Additional functionalization reactions can be carried out in the presence of cross-linker, as a certain extent of cross-linking is required in many applications of the polymer particles. The cross-linker was added to both the seed and the shell. Similar permutations of polymerization reactions with shot or delayed addition of the secondary batch of monomers, batch or semibatch functionalization, polymerization of only BPOEA or in copolymerization with styrene, etc., in the shell were carried out. These particles are shown in Figure 10.12. There were signiicant differences in the morphology of the polymer particles in the presence of cross-linker as compared to un-crosslinked particles. The cross-linked particles were smaller in size than their un-cross-linked counterparts. Also, the surface of the cross-linked particles

322

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

200 nm

(a)

250 nm

(b)

100 nm

100 nm (c)

(d)

FIGURE 10.11 SEM images of the functionalized particles using emulsiied poly merization conditions. Functionalization methodologies, that is, shot addition, starved addition, batch or semibatch generation of particles, etc., were varied.

was rougher in nature. No secondary nucleation of the particles was observed in any of the functionalization reactions in the presence of cross-linker. The surface morphologies were also different if only BPOEA was used to form the shell or if the copolymer of styrene and BPOEA were used to functionalize the particles. BPOEA functionalization without the addition of styrene was observed to generate somewhat smoother particles. This is quite surprising as only a small amount of cross-linker led to signiicant changes in the morphology of the resulting particles. The molar ratio of styrene to divinylbenzene cross-linker in the seed latex was approximately 50, whereas a ratio of 40 was used for the shell-forming monomers. Most importantly, the secondary nucleation was completely eliminated. Divinylbenzene is generally

Functional Polymer Particles by Emulsifier-Free Polymerization

323

250 nm

300 nm (a)

(b)

250 nm

250 nm

(c)

(d)

FIGURE 10.12 SEM images of the functionalized latex particles obtained by the use of cross-linker in core and shell: (a) when BPOEA and divinylbenzene were added as a shot to 70% poly merized seed latex followed by poly merization, (b) when styrene, BPOEA, and divinylbenzene were added as a shot to 70% poly merized seed latex succeeded by poly merization, (c,d) when delayed addition of the monomers of cases (a) and (b), respectively, was carried out, and (e) when styrene, BPOEA, and divinylbenzene were added in a delayed mode to fully poly merized seed latex and subsequent poly merization.

observed to enhance the colloidal instability of the forming particles as well as to reduce the swelling of the seed with the functionalizing monomers. This generates a large number of nuclei, which keep on coalescing with the seed particles owing to their colloidal instability. Because the presence of styrene leads to the fast generation of the particles, their coalescence on the surface of the particles is more clearly visible than the case of BPOEA functionalization only, owing to the colloidal instability of the formed BPOEA chains.

324

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

300 nm (e) FIGURE 10.12 (continued).

The amount of the functionalizing monomer BPOEA used during the shell-forming reaction also affects the resulting morphology of the particles. These trials are shown in Figure 10.13. In the irst case (Figure 10.13a), 15 g of fully poly merized seed with a solid content of 4.4% was added with 0.21 g of the functionalizing monomer. In the second and third case, the amount of the functionalizing monomer was increased by two and three times, respectively. Increasing the amount of the functionalizing monomer led to the generation of different morphologies. The morphology in the irst case was more like an “orange peel” type, whereas it changes to “moon crater” type morphology in the second case; in the third case, these craters were absent probably owing to a large amount of shell-forming monomer. However, the surface of the particles was still rough as a result of the collapse of the BPOEA chains on the surface purely kinetically and not thermodynamically because of the incompatibility between the BPOEA and seed particles.

Acknowledgments The author would like to express heartfelt gratitude to Professor M. Morbidelli at Institute of Chemical and Bioengineering at Swiss Federal Institute of Technology, Zurich. The technology described in this chapter has also been reported in Polymer, 2007, Volume 48, pp. 2806–2817.

325

Functional Polymer Particles by Emulsifier-Free Polymerization

250 nm

300 nm (a)

(b)

300 nm (c) FIGURE 10.13 Change in the surface morphology of the surface of the particles when the amount of functionalizing monomer is increased in the shell-forming monomer batch.

References 1. Odian, G. 2004. Principles of polymerization. Hoboken, NJ: John Wiley & Sons. 2. Gowariker, V. R., Viswanathan, N. V., and J. Sreedhar. 1986. Polymer science. New Delhi, India: John Wiley & Sons. 3. Mittal, V. 2009. Advances in polymer latex technology. New York: Nova Science Publishers. 4. Matyjaszewski, K., and T. P. Davis. 2002. Handbook of radical polymerization. Hoboken, NJ: John Wiley & Sons.

326

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

5. Cunningham, M. F. 2008. Controlled/living radical polymerization in aqueous dispersed systems. Progress in Polymer Science 33:365–98. 6. Mittal, V., Matsko, N. B., Butté, A., and M. Morbidelli. 2007. Functionalized polystyrene latex particles as substrates for ATRP: Surface and colloidal characterization. Polymer 48:2806–17. 7. Mittal, V., Matsko, N. B., Butté, A., and M. Morbidelli. 2008. PNIPAAM grafted polymeric monoliths synthesized by the reactive gelation process and their swelling/deswelling characteristics. Macromolecular Reaction Engineering 2:215–21. 8. Wu, T., Eimenko, K., and J. Genzer. 2001. Preparing high-density polymer brushes by mechanically assisted polymer assembly. Macromolecules 34:684–86. 9. Zhao, B., and W. J. Brittain. 2000. Synthesis, characterization, and properties of tethered polystyrene-b-polyacrylate brushes on lat silicate substrates. Macromolecules 33:8813–20. 10. Kim, J., Bruening, M. L., and G. L. Baker. 2000. Surface-initiated atom transfer radical polymerization on gold at ambient temperature. Journal of American Chemical Society 122:7616–17. 11. Sun, T., Wang, G., Feng, L., Liu, B., Ma, Y., Jiang, L., and D. Zhu. 2004. Reversible switching between superhydrophilicity and superhydrophobicity. Angewandte Chemie International Edition 43:357–60. 12. Coca, S., Jasieczek, C., Beers, K. L., and K. Matyajaszewski. 1998. Polymerization of acrylates by atom transfer radical polymerization. Homopolymerization of 2-hydroxyethyl acrylate. Journal of Polymer Science, Part A: Polymer Chemistry 36:1417–24. 13. Gaynor, S. G., Qiu, J., and K. Matyajaszewski. 1998. Controlled/“living” radical polymerization applied to water-borne systems. Macromolecules 31:5951–54. 14. Qiu, J., Gaynor, S. G., and K. Matyajaszewski. 1999. Emulsion polymerization of n-butyl methacrylate by reverse atom transfer radical polymerization. Macromolecules 32:2872–75. 15. Kizhakkedathu, J. N., Norris-Jones, R., and D. E. Brooks. 2004. Synthesis of well-deined environmentally responsive polymer brushes by aqueous ATRP. Macromolecules 37:734–43. 16. Jayachandran, K. N., Takacs-Cox, A., and D. E. Brooks. 2002. Synthesis and characterization of polymer brushes of poly(N,N-dimethylacrylamide) from polystyrene latex by aqueous atom transfer radical polymerization. Macromolecules 35:4247–57. 17. Kizhakkedathu, J. N., and D. E. Brooks. 2003. Synthesis of poly(N,N-dimethylacrylamide) brushes from charged polymeric surfaces by aqueous ATRP: Effect of surface initiator concentration. Macromolecules 36:591–98. 18. Matyajaszewski, K., Gaynor, S. G., Kulfan, A., and M. Podwika. 1997. Preparation of hyperbranched polyacrylates by atom transfer radical polymerization. 1. Acrylic AB* monomers in “living” radical polymerizations. Macromolecules 30:5192–94. 19. Sundberg, D. C., Casassa, A. P., Pantazopoulous, J., Muscato, M. R., Kronberg, B., and J. Berg. 1990. Morphology development of polymeric microparticles in aqueous dispersions. I. Thermodynamic considerations. Journal of Applied Polymer Science 41:1425–42.

Functional Polymer Particles by Emulsifier-Free Polymerization

327

20. Juang, M. S., and I. M. Krieger. 1976. Emulsiier-free emulsion polymerization with ionic comonomer. Journal of Polymer Science, Polymer Chemistry Edition 14:2089–2107. 21. Zukoski, C. F., and D. A. Saville. 1985. The formation of small scale granularities in latex particles. Journal of Colloid and Interface Science 104:583–86. 22. van Herk, A. M., and R. G. Gilbert. 2005. Chemistry and technology of emulsion polymerization. Oxford: Blackwell Publishing.

11 Polymer Nanoparticles with Surface Active Initiators and Polymer Initiators Klaus Tauer CONTENTS 11.1 General Introduction ................................................................................. 329 11.2 Historical Introduction ............................................................................. 330 11.3 Speciic Introduction ................................................................................. 333 11.4 Role of Initiators in Aqueous Emulsion Polymerization......................334 11.5 Initiation and Interfaces ............................................................................ 338 11.6 Surface Active Initiators—Inisurfs .......................................................... 341 11.7 Polymers as Initiators for Emulsion Polymerization ............................ 347 11.8 Summary..................................................................................................... 353 References............................................................................................................. 353

11.1 General Introduction Heterophase polymerization can be deined as polymerization reaction involving two or more different phases leading eventually to so-called polymer dispersions or latexes. Examples of such polymerization are dispersion polymerization, suspension polymerization, emulsion polymerization, miniemulsion polymerizations, and microemulsion polymerization. All these systems are characterized by a continuous phase containing segregated or dispersed phases. Two phases in direct contact with different compositions are characterized geometrically by an interface and energetically by an interfacial tension. In comparison to the bulk phases of the two major components (continuous phase and monomer/polymer), the dispersed state possesses an excess free energy that roughly corresponds to the product of interfacial area and interfacial tension. Due to this excess free energy, latex particles made of lyophobic polymers possess an inherent tendency to decrease the interfacial area and, hence, are prone to coagulation. In order to stabilize the colloidal state at a given interfacial area, certain measures are required to reduce the interfacial tension. The most effective way to do this is the decoration of the interface with lyophilic groups mediating between both phases. 329

330

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

Here, only water is considered as continuous phase, which means that the dispersed particles are hydrophobic and stabilized by hydrophilic groups attached to their surface via either adsorption or covalent binding. Although aqueous heterophase polymerization is fundamentally not restricted to a particular chain growth mechanism, this contribution is exclusively focused on classical free radical polymerization, which is under practical aspects nowadays still the most important method of synthetic latex production. Under these constraints, stabilization via adsorption requires emulsiiers with Krafft temperatures below the polymerization temperature [1], high HLB (hydrophylic-lipophilic balance) values, and phase inversion temperatures above the polymerization temperature [2]. Covalent attachment of hydrophilic groups is possible via any kinetic event of radical polymerization—that is, initiation, chain propagation, chain transfer, and chain termination. However, the title of this chapter requires focus on the initiation step of free radical polymerization. It deals with various possibilities to initiate the chain growth in aqueous heterophase polymerization that have been developed over the decades that this polymerization technique has been in practical use.

11.2 Historical Introduction Emulsion polymerization or more general heterophase polymerization was invented in the early twentieth century at Bayer in Germany [3]. The merit belongs to Kurt Gottlob, who polymerized isoprene in an aqueous system and made for the irst time ever artiicial caoutchouc. This was at a time before the application of radical initiators was common and also before the invention of synthetic emulsiiers or stabilizers. Consequently, the procedure was quite simple but long lasting [4]. A mixture of 500 g of water containing 7 g of natural stabilizer (egg albumin, starch, or gelatin) and 300 cm3 of isoprene was placed in a pressure vessel and kept under stirring or shaking at 60°C for several weeks. The polyisoprene separated spontaneously over time from the latex-like liquid that was obtained at the end of the reaction, due to clotting of the stabilizers under ambient conditions. The addition of alkali or organic bases led to an improvement of the latex stability, and coagulation could be deliberately induced by decreasing the pH. In these experiments the polymerization was started obviously either thermally or due to the action of oxygen more or less randomly depending on the particular amount of oxygen in the reaction vessel. It has been known for more than a century that oxygen plays a double role in radical poly merization as it can either initiate and accelerate or decelerate and inhibit the polymerization [5,6].

Surface Active Initiators

331

In the same years, the irst patents were iled on the usefulness of ozonides or organic peroxides to improve the properties of rubber-like polymers (butadiene and its homologues) [7] and to start the polymerization of vinylesters with inorganic or organic peroxides (peroxoborates, peroxycarbonates, benzoyl peroxide) [8]. However, these reactions were not carried out under aqueous heterophase conditions. In 1913, Ludwig Lautenschläger completed his doctoral thesis on the autoxidation and polymerization of unsaturated hydrocarbons under supervision of Hermann Staudiger [9]. This topic for his doctoral studies was chosen following a suggestion of C. Engler, who did a great deal of work on autoxidation in the second half of the nineteenth century [10]. With progressing studies and applications of radical initiators, polymerization eficiency was considerably enhanced and a much better control of the process was achieved. In the early 1930s, various types of compounds were known that were useful in inluencing the course of polymerizations (polymerization accelerators) such as halides, stannic luoride or boron triluoride, caustic alkyls, alkali peroxides [11,12], and sodium persulfate [13]. Note that at this time there was not yet a clear distinction between catalysts and initiators or radical and ionic polymerization (cf. the discussion in [14]). From today’s perspective, this is quite astonishing as it was already known at this time that free radicals (alkyl radicals generated during thermal decomposition of mercury dialkyls start polymerization in an ethylene gas phase) [15] and alkali metals (sodium for starting isoprene polymerization to get sodium rubber) [7] or alkali metal alkyls (such as butyllithium for butadiene polymerization) [16] can also be used to start polymerization reactions. It was, however, generally accepted that these substances accelerate the radical production rate (rate of initiation) and decrease the average chain length of the polymer molecules under homogeneous conditions. It was also known that under heterogeneous reaction conditions (emulsion polymerization) acceleration of the monomer conversion is possible without simultaneously decreasing the degree of polymerization. It is interesting to note that the initiation mechanism with persulfate, which nowadays can be considered as a kind of standard initiator for aqueous polymerizations, was not yet understood during the early 1940s [17]. The initiation was thought to take place via complex formation between Caro’s acid (formed via the reaction S2O82– + H2O → SO42– + H2SO5) and the monomer followed by splitting off of the labile peroxide oxygen, which then starts chain growth. Organic peroxides became popular initiators for polymerizations in organic solvent, including bulk polymerizations. During the 1930s the stateof-the-art initiator for aqueous emulsion polymerization was hydrogen peroxide, because the polymerization was started in the aqueous phase [18–20]. Later it was shown that aqueous emulsion poly merization could also be carried out with oil-soluble initiators such as benzoyl peroxide and azo-bis-

332

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

1-phenylethane [21,22] or 2,2-azobisisobutyronitrile (AIBN) and p-tert-butylisoproylbenzenehydroperoxide [23–25]. The latter authors assumed, based on their systematic studies on the inluence of potassium peroxodisulfate as water soluble and the mentioned oil-soluble initiators, that radical generation and polymerization takes place in the adsorbed surfactant layer around the latex particles. First reports on the application of azo-compounds as catalyst or initiator appeared in the late 1930s. The action of triphenylmethylazobenzene as accelerator of styrene polymerization (bulk and solution polymerization) applied in a molar ratio between 0.345 × 10 –5 and 6.67 × 10 –4 relative to the monomer at a temperature of 50–60°C was described in 1939 [26]. Possibly the irst example of the application of an azo-compound in radical heterophase polymerization was reported in 1942 when it was found that p-bromobenzene diazonium hydroxide can start styrene polymerization [27]. This was observed when “alkali was added to vigorously stirred suspension of 30 ml of styrene in an aqueous solution of 11 g of diazotized p-bromoaniline at 0°C” [27]. From the separated viscous organic layer, the authors isolated polystyrene with p-bromobenzene end groups and determined an average degree of polymerization of 22 [27]. Aliphatic azo-compounds have been in use at least since the late 1940s and quickly became quite popular sources of free radicals for polymerizations in both organic media (azo-bis-alkylnitriles) and water (4-azo-bis-4-cyanopentanoic acid) because of their particular properties [28]. These compounds possess fairly good chemical stability during storage, irst-order decomposition kinetics almost independent of reaction media, and the ability to act not only as thermal initiators but also as eficient photoinitiators (absorption in the near ultraviolet) [28–30]. Moreover, further chemical modiication is quite easily possible, which is of special importance with respect to the topic of this chapter. The studies of Fritz Haber and others on the catalytic decomposition of hydrogen peroxide by both ferrous and ferric salts formed the basis for the development of redox initiation for radical poly mer i zations [31]. The authors have shown that the reaction cycle combines a chain and a radical reaction, involving in its stages the radicals (HO· and HO2·) and the hydroxyl anion. Another important breakthrough was achieved at the ICI laboratories in Manchester in October 1940 with the extension of the principle of “reduction activation” to other peroxides [32]. It was found that polymerizations initiated with persulfate could be improved in combination with hydroquinone, copper, bisulite, or other reductants. Particularly for aqueous acrylonitrile polymerization, the observations were made that the induction period in nearly all cases was shortened or even vanished, that the rate of polymerization, at least initially, was enhanced, and that both effects occurred frequently in an almost proportional fashion. This method, which nowadays is called redoxinitiation, enables one to conduct polymerizations at low temperatures. In

Surface Active Initiators

333

the decades since it has largely been used in both industrial and lab-scale production of polymer dispersions.

11.3 Specific Introduction The colloidal nature of emulsion polymerizations and polymer dispersions is the origin for all pros and cons regarding their use and application. The basics of emulsion polymerization and polymer latexes are quite well understood [33–35] but still an object of ongoing reinements [36–38]. The huge number of small particles of an aqueous emulsion polymerization allows easy heat removal, the combination of hydrophilic and hydrophobic monomers in just a single polymerization step, and tailored modiication of particle properties by subsequent monomer addition leading to layer-by-layer buildup of the desired particle morphology. The main drawback of emulsion polymers is the contamination of the desired polymer by auxiliary materials, mainly emulsiiers. Depending on the average particle size, the solids content, and the polymerization procedure, the emulsiier content can be as high as 5 wt% relative to the amount of monomers. All emulsiiers for aqueous emulsion polymerizations are hydrophilic, and their presence in the inal polymer increases the hydrophilicity during application. Moreover, low molecular weight emulsiiers are easily able to desorb or to migrate upon changing conditions. In latexes frequently this leads to undesired foaming. Migration of surfactant molecules is also an issue in the inal application form of the polymer dispersion because they assemble in hydrophilic spots, forming entry sites for water molecules. Such effects contribute essentially to the failure of latex-based polymeric products such as protective coatings and are a strong driving force for the development of strategies to ix stabilizers in place [37,39]. Fixing the stabilizer in place is a compromise. The better alternative, the change in the properties of the surfactants from hydrophilic as needed during the polymerization to hydrophobic as desirable for a nonlatex application of the polymer during the service life, seems obviously not attainable. In summary, the actual problem of an emulsion poly mer is the stabilizing system. On the one hand, it should be optimized to preserve stability during the polymerization process. This requires optimized adsorption properties including a certain hydrophilicity and mobility. On the other hand, the inal destination of most of the polymer latexes is the solid state with more or less hydrophobic properties. Clearly, these demands are not supported by optimized emulsiiers for aqueous emulsion polymerizations. So, strategies have been developed during the last decades trying to improve the emulsiier eficiency regarding both contradicting requirements [39–45]. In order to keep the stabilizers in place, their chemical structure has been modiied in various ways, allowing them to participate in the polymerization reaction

334

Advanced Polymer Nanoparticles: Synthesis and Surface Modifications

either as monomer (so-called surfmers), as transfer agents (so-called transurfs), or as initiators (so-called inisurfs) [39]. These various functionalities allow covalent (permanent) binding to the polymer chains. Typically these reactive surfactants are based on low molecular weight emulsiiers. But polymeric reactive stabilizers also have been developed such as partly sulfonated polyoleins [45] belonging to the class of surfmers. Stabilizing and initiating properties can also be combined in polymers, offering a variety of clear advantages compared with only initiators or even surface active initiators. Surfactants for emulsion polymerization are “only” auxiliary materials facing strict cost requirements. Emulsiiers contribute between 5 and 10%, depending on the particular product, to the materials costs for latex production (cf. also the discussion in [37]). Reactive surfactants always have to compete with “classical” nonreactive standard emulsiiers. The improvements in the performance of the latex and the inal polymer product must justify possibly higher costs of the reactive surfactants. Hence, synthesis of reactive surfactants for potential industrial applications should always be based on readily available raw materials. Studying the application of reactive surfactants is a matter not only of scientiic curiosity but also of interest for application-oriented research, as the performance of poly mer dispersions can be improved compared to similar dispersion prepared with conventional surfactants. An example of the application is given in [46]. The author showed that the overall performance of binder dispersions for industrial coatings could be improved when the nonreactive surfactant was replaced by a surfmer based on “easily” obtainable maleic acid esters. This chapter is exclusively devoted to initiators or initiating systems for aqueous emulsion polymerizations that are able to act twofold in the course of the reaction. They must be able to eficiently start the polymerization and to stabilize a huge interfacial area; that is, they must allow the manufacture of polymer dispersions with high solids content and small particles. The chapter is organized in a way that irst the role of hydrophilic initiators in aqueous emulsion polymerization is discussed, followed by considerations regarding the features of inisurfs and polymers as initiating species.

11.4 Role of Initiators in Aqueous Emulsion Polymerization In most industrial emulsion polymerizations, water-soluble initiating systems are employed and radical generation takes place predominantly in the continuous aqueous phase via either thermal decomposition or redox activation [47]. Because the main monomers are hydrophobic, the polymer chain ends derived from primary radicals might be considered to be amphiphilic at least to a certain extent. The ability to stabilize latex particles depends

335

Surface Active Initiators

A:

R HO•