Engineered Carbohydrate-Based Materials for Biomedical Applications: Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels

ENGINEERED CARBOHYDRATE-BASED MATERIALS FOR BIOMEDICAL APPLICATIONS Polymers, Surfaces, Dendrimers, Nanoparticles, and ...

Author: Ravin Narain

211 downloads 846 Views 6MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

ENGINEERED CARBOHYDRATE-BASED MATERIALS FOR BIOMEDICAL APPLICATIONS Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels

Edited by

RAVIN NARAIN University of Alberta Edmonton, Alberta, Canada

A JOHN WILEY & SONS, INC., PUBLICATION

ENGINEERED CARBOHYDRATE-BASED MATERIALS FOR BIOMEDICAL APPLICATIONS

ENGINEERED CARBOHYDRATE-BASED MATERIALS FOR BIOMEDICAL APPLICATIONS Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels

Edited by

RAVIN NARAIN University of Alberta Edmonton, Alberta, Canada

A JOHN WILEY & SONS, INC., PUBLICATION

C 2011 by John Wiley & Sons, Inc. All rights reserved. Copyright

Published by John Wiley & Sons, Inc., Hoboken, New Jersey. Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com Library of Congress Cataloging-in-Publication Data: Engineered carbohydrate-based materials for biomedical applications : polymers, surfaces, dendrimers, nanoparticles, and hydrogels / edited by Ravin Narain. p. ; cm. Includes bibliographical references and index. ISBN 978-0-470-47235-4 (cloth) 1. Carbohydrates–Biotechnology. I. Narain, Ravin. [DNLM: 1. Biopolymers–physiology. 2. Biocompatible Materials. 3. Biomedical Engineering–methods. 4. Dendrimers. 5. Hydrogels. 6. Polysaccharides–chemistry. QT 37.5.P7] TP248.65.P64E54 2011 660.6–dc22 2010039787 Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

CONTENTS

PREFACE

vii

CONTRIBUTORS

xi

1 SYNTHESIS OF GLYCOPOLYMERS

1

Samuel Pearson, Gaojian Chen, and Martina H. Stenzel

2 BLOCK GLYCOPOLYMERS AND THEIR SELF-ASSEMBLY PROPERTIES

119

Qian Yang

3 CATIONIC GLYCOPOLYMERS

143

Marya Ahmed and Ravin Narain

4 GLYCOPOLYMER BIOCONJUGATES

167


5 GLYCOPOLYMER-FUNCTIONALIZED CARBON NANOTUBES

189


6 GLYCONANOPARTICLES: NEW NANOMATERIALS FOR BIOLOGICAL APPLICATIONS

213

Isabel Garc´ıa, Juan Gallo, Marco Marradi, and Soledad Penades ´ v

vi

CONTENTS

7 GLYCODENDRIMERS AND THEIR BIOLOGICAL APPLICATIONS

261

Elizabeth R. Gillies

8 GLYCOSURFACES

307

Anca Mateescu and Maria Vamvakaki

9 CARBOHYDRATE-DERIVED HYDROGELS AND MICROGELS

337

Mitsuhiro Ebara

10

MODIFIED NATURAL POLYSACCHARIDES AS NANOPARTICULATE DRUG DELIVERY DEVICES

355

Archana Bhaw-Luximon

INDEX

397

PREFACE

Carbohydrates are the most abundant, easily accessible and cheap biomolecules in nature. Besides their potential uses as key chemical raw materials and energy production, they have been recognized to play a key role in a wide variety of complex biological processes. They are involved to a large extent in mediating recognition processes through their interactions with proteins and other biological entities. They have been recognized to play a significant role in many important cellular recognition processes including cell growth regulation, differentiation, adhesion, cancer cell metastasis, cellular trafficking, inflammation by bacteria and viruses, and immune response. Individual carbohydrate–protein interactions are generally weak, and multivalent forms of carbohydrate ligands are usually involved in those biological processes. This book has been conceived in order to provide an up-to-date account of the major developments on the biomedical applications of synthetic carbohydrate-based materials. This book is organized into five main themes such as polymers, nanoparticles, surfaces, dendrimers, and hydrogels. Synthetic glycopolymers are essential macromolecules that display many structural and functional features. With functions similar to those of natural carbohydrates, synthetic glycopolymers with specific pendant saccharide moieties can play a significant role in pathological and biological processes via multivalent carbohydrate– protein interactions. With recent progress in organic and polymer chemistry, functional glycopolymers have been prepared with remarkable ease. Carbohydrate-based polymers with different properties were also synthesized, including biodegradable, thermosensitive, and acid-degradable core-crosslinked glyconanoparticles, with neuroactivity and with chiroptical properties. Chapter 1 provides a comprehensive review on the synthesis of glycomonomers and their corresponding glycopolymers via a wide range of organic and polymerization synthesis approach. Some biological vii

viii

PREFACE

interaction studies and applications of glycopolymers, such as in antivirus/bacteria and gene delivery, are also described. Chapter 2 discusses the solution properties of block glycopolymers and their biological relevance. The synthesis of smart block glycopolymers using various polymerization techniques has been discussed. The usage of these smart glycopolymers in tissue engineering, drug delivery, and pathogen interactions is discussed. One of the well-studied types of block glycopolymers is cationic glycopolymers. The use of cationic polymers for gene delivery purposes is a facile technique that is extensively studied as a possible source of noninvasive and efficient gene delivery. Chapter 3 discusses the role and importance of cationic glycopolymers for gene delivery purposes. The brief overview of synthesis of cationic glycopolymers by different polymerization techniques is provided. The detailed study of cationic glycopolymer for gene delivery purposes is specifically discussed. The incorporation of glycopolymers or their corresponding copolymers to macromolecules of choice can further enhance their physiological impact for biological applications. The major challenge in this regard is the synthesis of glycopolymer bioconjugates of controlled dimensions to explore their uses for biomedical applications. Chapter 4 describes the synthetic techniques used in the literature for the production glycopolymer bioconjugates and their importance in biological applications. The facile approaches to synthesize glycopolymer bioconjugates of controlled dimensions are highlighted and their role in biological assays, diagnostics, and in the study of carbohydrate- and protein-based interactions is elaborated. The synthesis of glycoclusters is an important aspect of synthetic carbohydratebased materials under study to understand their interactions with macromolecules such as pathogens and several proteins. These interactions of glycoclusters with living organisms or macromolecules make the basics of most biological phenomena, including invasion, metastasis, and infections. Nanotechnology is a rapidly growing field of materials science that has also extensively been explored in biological applications, owing to the facile introduction of functional groups on the surface of nanomaterials. The introduction of glycopolymer-based moieties on the surface of nanomaterials are found to produce glycoclusters with enhanced biological significance compared to glycopolymers alone, due to the multivalent effect of functional groups present on the surface of nanomaterials. These nanomaterials are largely studied in literature as a function of their structure, nature of materials, surface functionalization properties, and morphology-dependent interactions with living organisms. Chapter 5 discusses the various strategies to synthesize glycopolymer-functionalized carbon nanotubes and their interactions in vitro and in vivo. The inherent properties of carbon nanotubes toward cellular uptake and their toxicity issues are discussed. Moreover, the uses of glycopolymer-functionalized nanotubes for biomedical applications, including gene and drug delivery, and tissue engineering is described. Chapter 6 provides a brief overview about the synthesis and surface functionalization of another type of nanomaterial, namely metallic nanoparticles. The synthesis and surface functionalization of gold and magnetic nanoparticles and of quantum dots with biocompatible carbohydrate-based polymers has opened various possibilities for their uses in biotechnology and biomedicines. This chapter describes a review of few biomedical applications of these glyconanoparticles, including their use in pathogen

PREFACE

ix

inhibition, fluorescent probes, magnetic resonance imaging, and cancer metastasis. Another approach to obtain multivalency and to enhance the function of glycopolymers is the synthesis of glycodendrimers, which compared to their corresponding polymers are of controlled molecular weight and architecture. Chapter 7 describes the synthesis of glycodendrimers using various strategies and their interactions with proteins are studied. The interactions of glycodendrimers with various proteins at physiological and pathological levels are the discussed. In addition to the surface functionalization of nanoscaffolds in colloidal form, the synthesis of glycopolymer-coated macroscaffolds are found to be an attractive platform for the tissue engineering purposes. These glycopolymer-modified surfaces provide not only biocompatibility but are also shown to possess the potential to provide the selectivity in cellular growth and proliferation. Chapter 8 provides a detailed overview of the synthetic techniques involved in the functionalization of macroscaffolds with glycopolymers or their corresponding copolymers. Moreover, the characterization of these surfaces and their role in tissue engineering and as nonfouling surfaces for the inhibition of pathogens is discussed. Chapter 9 provides a different synthetic route to produce biomaterials for tissue engineering and gene delivery. The chapter focuses on the synthesis of glycopolymer-functionalized hydrogels by various techniques. The use of these hydrogels in tissue engineering and drug delivery is discussed. Chapter 10 mainly focuses on the modification of natural carbohydrate-based scaffolds for drug delivery purposes via various administration routes. These modifications are thought to increase the efficacy of drug delivery, in addition to eliminating the hypersensitivity reactions associated with various drug treatments. Ravin Narain

CONTRIBUTORS

Marya Ahmed, Department of Chemical and Materials Engineering and Alberta Ingenuity Centre for Carbohydrate Science, University of Alberta, Edmonton, AB, Canada Archana Bhaw-Luximon, Department of Chemistry, University of Mauritius, Réduit, Mauritius Gaojian Chen, Centre for Advanced Macromolecular Design, University of New South Wales, Sydney, Australia Mitsuhiro Ebara, Smart Biomaterials Group, Biomaterials Center, National Institute for Materials Science, Tsukuba, Japan Juan Gallo, Laboratory of Glyconanotechnology, Biofunctional Nanomaterials Unit, CIC biomaGUNE/CIBER-BBN, San Sebastian, Spain Isabel Garc´ıa, Laboratory of Glyconanotechnology, Biofunctional Nanomaterials Unit, CIC biomaGUNE/CIBER-BBN, San Sebastian, Spain Elizabeth R. Gillies, Department of Chemistry, Department of Chemical and Biochemical Engineering, The University of Western Ontario, London, Canada Marco Marradi, Laboratory of Glyconanotechnology, Biofunctional Nanomaterials Unit, CIC biomaGUNE/CIBER-BBN, San Sebastian, Spain Anca Mateescu, Institute of Electronic Structure and Laser, Foundation for Research and Technology—Hellas Heraklion, Crete, Greece, Department of Chemistry, University of Crete, Heraklion, Crete, Greece

xi

xii

CONTRIBUTORS

Ravin Narain, Department of Chemical and Materials Engineering and Alberta Ingenuity Centre for Carbohydrate Science, University of Alberta, Edmonton, AB, Canada Samuel Pearson, Centre for Advanced Macromolecular Design, University of New South Wales, Sydney, Australia Soledad Penadés, Laboratory of Glyconanotechnology, Biofunctional Nanomaterials Unit, CIC biomaGUNE/CIBER-BBN, San Sebastian, Spain Martina H. Stenzel, Centre for Advanced Macromolecular Design, University of New South Wales, Sydney, Australia Maria Vamvakaki, Institute of Electronic Structure and Laser, Foundation for Research and Technology—Hellas, Heraklion, Crete, Greece, Department of Materials Science and Technology, University of Crete, Heraklion, Crete, Greece Qian Yang, Lehrstuhl für Technische Chemie II, Universität Duisburg-Essen, Essen, Germany

CHAPTER 1

SYNTHESIS OF GLYCOPOLYMERS SAMUEL PEARSON, GAOJIAN CHEN, and MARTINA H. STENZEL Centre for Advanced Macromolecular Design, University of New South Wales, Sydney, Australia

1.1 Introduction 1.2 Synthesis of Vinyl-Containing Glycomonomers 1.2.1 Monomers from Protected Carbohydrates 1.2.2 Monomers from Unprotected Sugars 1.3 Conventional Free Radical Polymerization 1.3.1 Acrylamide Monomers 1.3.2 (Meth)acrylate Monomers 1.3.3 Styrene-Based Monomers 1.3.4 Other Vinyl-Containing Glycomonomers 1.4 Controlled/Living Radical Polymerization 1.4.1 Stable Free Radical Polymerization 1.4.2 Atom Transfer Radical Polymerization 1.4.3 Reversible Addition–Fragmentation Chain Transfer Polymerization 1.5 Ring-Opening Polymerization 1.6 Ionic Chain Polymerization 1.6.1 Anionic Chain Polymerization 1.6.2 Cationic Chain Polymerization 1.7 Ring-Opening Metathesis Polymerization (ROMP) 1.8 Postfunctionalization of Preformed Polymers Using Sugar Moieties 1.8.1 Amide Linkage 1.8.2 Click Approach 1.8.3 Other Nonclick Approaches 1.9 Conclusions References

2 2 2 4 5 6 6 18 19 20 20 33 50 71 74 74 80 82 90 98 101 103 104 104

Engineered Carbohydrate-Based Materials for Biomedical Applications: Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels, Edited by Ravin Narain C 2011 John Wiley & Sons, Inc. Copyright

1

2

SYNTHESIS OF GLYCOPOLYMERS

1.1 INTRODUCTION Glycopolymers—synthetic polymers with pendant carbohydrates—have received considerable attention in the fields of polymer chemistry, material science, and biomedicine due to their biocompatibility and their bioactivity. From humble beginnings where glycopolymers were synthesized from vinyl-functionalized sugars via free radical polymerization with little control over the resulting polymer characteristics, glycopolymer synthesis has now developed into a mature area where the control over molecular weight and polymer architecture is routinely sought and indeed achieved. Glycopolymer synthesis has now infiltrated most known techniques of polymer synthesis and is not restricted to controlled radical processes; ionic techniques also provide feasible means to polymerize glycomonomers in a controlled manner. The aim of this chapter is to provide a comprehensive review of all the techniques that have been utilized for the synthesis of glycopolymers; however, greater emphasis will be placed on the techniques that supercede free radical polymerization. After highlighting the important achievements in the synthesis of glycopolymers via free radical polymerization, focus will turn toward glycopolymer synthesis via the controlled/living radical polymerization processes known as nitroxide-mediated polymerization (NMP), cyanoxyl-mediated polymerization, atom transfer radical polymerization (ATRP), and reversible addition–fragmentation chain transfer (RAFT) polymerization. Glycopolymer synthesis by ring-opening polymerization (ROP), anionic polymerization and cationic polymerization will detail the progress made in the area of ionic polymerization, and a discussion of the work carried out using ring-opening metathesis polymerization (ROMP) will conclude the section on the synthesis of glycopolymers by polymerizing sugar-containing monomers. The functionalization of reactive polymer scaffolds with carbohydrate species will then be discussed as an alternative strategy for synthesizing glycopolymers.

1.2 SYNTHESIS OF VINYL-CONTAINING GLYCOMONOMERS 1.2.1 Monomers from Protected Carbohydrates The commercial availability of a range of carbohydrates provides access to a wide array of different glycomonomers, and significant efforts have been dedicated to the synthesis of polymerizable vinyl sugars. In an early feature article, Wulff et al. [1] highlighted possible avenues for generating glycomonomers in which an important distinction must be made between protected and unprotected carbohydrates. The choice of employing protected or unprotected sugars is dependent on the ease of stereospecific functionalization of the sugar, the solubility of the monomer and polymer, the potential incompleteness of the removal of the protective group, and the ease of purification. The most common synthetic approaches are outlined below, but they are discussed in more detail elsewhere [1].

SYNTHESIS OF VINYL-CONTAINING GLYCOMONOMERS

3

1.2.1.1 Reactions Using Isopropylidene-Protected Sugars Many sugars can easily be protected using acetone to form isopropylidene derivatives. This approach, which is suitable for a range of sugars including glucose, galactose, fructose, and sorbose, allows easy functionalization of the remaining hydroxyl functionality with acrylate [2], methacrylate [3], and 4-vinylbenzyl groups [4].

1.2.1.2 Glycosides from Halogeno Sugars Glycoside monomer synthesis via the reaction between halogeno sugars and hydroxyl groups of vinyl-containing species has been explored in detailed with varying degrees of success. The starting materials, typically acetylated 1-halogeno sugars, can be expensive or difficult to obtain, but the technique is especially useful for inserting longer spacers between the polymerizable moiety and the sugar. The cleavage of the acetyl protecting groups in alkaline media does not affect the glycoside bond [5].

1.2.1.3 Grignard Reactions The aldehyde functionality of a sugar molecule can be targeted by Grignard reagents [6]. Prior protection of the remaining hydroxyl groups is essential.

4


1.2.2 Monomers from Unprotected Sugars

1.2.2.1 Enzymatic trans-esterification Enzyme-catalysed transesterification reactions present a highly efficient and regioselective avenue for obtaining glycomonomers that would otherwise be inaccessible without utilizing protecting groups. The reaction of sugars with vinyl-containing esters in organic solvents is catalyzed by lipases such as Candida antarctica, usually yielding derivatives functionalized in 6-position [7], although efficient functionalization in the 1-position has also been reported [8].

1.2.2.2 Fischer Glycoside Synthesis Direct monosubstitution in the anomeric (C-1) position without the recourse to protective chemistry can be achieved by the reaction of an excess of hydroxyl groups, such as in hydroxyethyl acrylate, with the sugar in the presence of phosphomolybdic acid as catalyst [9].

β

1.2.2.3 Synthesis via Barbituratic Acid Barbituratic acid reacts readily with the C-1 position of the unprotected sugar to generate a reactive salt. Subsequent reaction with bromides such as 4-vinyl benzyl bromide leads to polymerizable monomers. Conversion of the barbituratic acid ring to a diamide further improves water solubility [10].

1.2.2.4 Conversion of Aminosugars A popular route to glycomonomers is the fast reaction between aminosugars and acyl halides or anhydrides. The high reactivity of the amine group ensures its preferential reaction even in the presence of unprotected hydroxyl groups. Reactions of acryloyl chloride and methacryloyl

CONVENTIONAL FREE RADICAL POLYMERIZATION

5

chloride [11] and also isocyanates [12] and epoxides with various aminosugars have been explored to confer the glycomonomers in high yields.

1.2.2.5 Reaction between Oxidized Sugars and Amines A range of amide-linked glycomonomers are accessible from sugars that have been oxidized to their corresponding lactones and can therefore be reacted with vinyl-functionalized amines [13].

While these are the most common strategies used for glycomonomer synthesis, other pathways have emerged in recent years such as Cu(I) click chemistry [14]. Some of these approaches are highlighted in this chapter.

1.3 CONVENTIONAL FREE RADICAL POLYMERIZATION Free radical polymerization is one of the most widely used techniques for making polymers. The polymerization reaction is initiated by free radical initiators and has been used to synthesize linear vinyl saccharide polymers since the 1960s. Despite its disadvantages, such as high polydispersities of the resulting polymers and difficulties in controlling terminal functionalities, the robustness of free radical polymerization has encouraged its widespread use. Indeed, a large number of reports have emerged on the synthesis of glycopolymers via free radical polymerization in both aqueous and nonaqueous media. Glycopolymers, polymers with pendant sugar groups, were first reported in 1961 when Kimura et al. [15] and Whistler et al. [11a, 11c] reported the free radical homo- and copolymerization of glycomonomers. Significant activity in the field of free radical polymerization of glycomonomers emerged in the following years, which only declined in the late 1990s with the birth of living free radical polymerization

6


techniques. The feature article by Wulff et al. [1] highlighted the body of work and the array of structures. Here, we only highlight some of the latest achievements in this area, mainly publications after 1990. 1.3.1 Acrylamide Monomers Roy et al. copolymerized 4-acrylamidophenyl-␤-lactoside (Table 1.1, entry 1) and acrylamide in water at 90◦ C in the presence of ammonium persulfate (APS) and tetramethylethylenediamine (TMEDA) [16]. The antigenicity of the resulting carbohydrate copolymer was then demonstrated by agar gel diffusion with peanut and castor bean lectins. Nishimura et al. outlined the synthesis of 3-(N-acryloylamino)propyl 2-acetamido-2-deoxy-␤-d-glucopyranoside [17] (Table 1.1, entry 2) and 3-(N-acryloylamino)propyl O-(␤-d-galactopyranosyl)-(l-4)-2acetamido-2-deoxy-␤-d-glucopyranoside [18] (Table 1.1, entry 3) and polymerized them in a similar manner to Roy et al. [16]. Methacrylamide-functionalized mannose monomers (Table 1.1, entries 4–5) were polymerized by Tagawa et al. using a lipophilic azo-initiator containing two long alkyl chains per initiating fragment [19]. Incorporation of the amphiphiles into liposomes generated structures that were able to recognize Concanavalin A (Con A) with little difference observed between the species with varying spacer lengths between the polymer backbone and the sugar residue. Also starting with protected sugars, Carpino et al. reported the synthesis of 2,3,4,6-tetra-O-acetyl-1-O-(4methacryloylaminophenyl)-␤-d-glucopyranoside (Table 1.1, entry 6) and 1-O-(4methacryloylaminophenyl)-␤-d-glucopyranoside (Table 1.1, entry 7), and homopolymerized them with 2,2 -azobisisobutyronitrile (AIBN) as initiator in dimethylformamide (DMF) to afford polymers that were then deprotected with sodium methoxide to give water-soluble glycopolymers [20]. 1.3.2 (Meth)acrylate Monomers Novel (meth)acrylic monomers (Table 1.1, entry 9) bearing a monosaccharide residue were developed by Kitazawa et al. by reacting methyl glycosides with 2-hydroxyethyl acrylate or methacrylate in the presence of heteropoly acid. The monomers were then polymerized in aqueous solution with potassium persulfate as initiator [9]. A galactose-based monomer containing a galactopyranose unit attached through an ester linkage to a vinyl group (Table 1.1, entry 10) was synthesized by Fortes and co-workers and was then copolymerized with ethyl acrylate in DMF under free radical conditions [22]. The protected monomers 2-(2 ,3 ,4 ,6 -tetra-O-acetyl-␤-d-glucosyloxy)ethyl methacrylate (Table 1.1, entry 11) and 2-(2 ,3 ,4 ,6 -tetra-O-acetyl-␤-dgalactosyloxy)ethyl methacrylate (Table 1.1, entry 12) were polymerized by Cameron et al. in chloroform, and the polymers deacetylated in a mixture of dichloromethane and methanol [23]. The alternative approach for obtaining deprotected polymers was also adopted; entries 11 and 12 were deprotected to give the corresponding monomers 2-(␤-d-glucosyloxy)ethyl methacrylate (GlcEMA; Table 1.1, entry 13) and

7

1

2

3

4

Glucosamine

Lactosamine

Mannose

HO

HO

O

OH OH

O

OH

4

OH

O

S

O

O

HO

NHAc

O

n = 3 or 6 HO OH

HO HO

OH

Entry Monomer

2

O

NHAc

O

H N O

n = 3 or 6

OH

H N n

O n

H N O

Glycomonomers Synthesized via Free Radical Polymerization

Lactose

Carbohydrate

TABLE 1.1

2:1 THF/MeOH

H2 O/DMSO

H2 O

H2 O

Solventa

70◦ C

DODA-ACPA

APS

APS

25◦ C

50◦ C

APS

90 C

◦

Temperature Initiatorb

(continued)

19

18

17

16

Reference

8 5

6

7

Glucose

Glucose

O

n = 4 or 7

HO

OH OH

O

OH

Entry Monomer

Mannose

Carbohydrate

n

H N O

TABLE 1.1 Glycomonomers Synthesized via Free Radical Polymerization (Continued)

DMF

DMF

n = 4: H2 O n = 7: 1:1 THF/MeOH

Solventa

AIBN

AIBN

60◦ C

V50 DODA-ACPA

60◦ C

70◦ C 70◦ C


20

20

19

Reference

9

8

9

10

11

Glucose

Glucose Galactose Mannose Xylose

Galactose

Glucose

Chloroform

DMF

H2 O

Benzene

AIBN

AIBN

80◦ C

65◦ C

No mention KPS in the paper

40 or 50◦ C AIBN

(continued)

23a

22

9

21

10 12

13

14

15

16

Glucose

Galactose

Glucose

Galactose AcO

AcO OAc

HO

HO O H

AcO

AcO OAc

Entry Monomer

Galactose

Carbohydrate

AcO

O

OH

O

AcO

O

O

O

O

O

CO2Et

O

O

O


Chlorobenzene

Chlorobenzene

H2 O/MeOH

H2 O/MeOH

Chloroform

Solventa

AIBN

AIBN

70◦ C

K2 S2 O8

65◦ C

70◦ C

K2 S2 O8

AIBN

65◦ C

65 C

◦


24

24

23a, 23b

23a

23a, 23b

Reference

11

17

18

19

Arabinose

Fructose

Galactose

HO

OH

OH

OH

OH

OH

OH,H

OH OH

O

OH,H

HO

OH

HO

HO

HO

OH, H

H2 O

H2 O

H2 O, DMF

70◦ C

70◦ C

70◦ C

AAPD

AAPD

(continued)

25

25

AAPD, AIBN 6b-d, 25, 26

12 20

21

22

Lactose

Maltose

HO

HO OH

Entry Monomer

Lactose

Carbohydrate

OH

O O HO

OH

N R

O

R = H or CH3

HO

OH N H


DMSO, H2 O

H2 O

H2 O

Solventa

60◦ C

25◦ C

RT

c

AIBN, KPS

KPS

KPS


13b, 30

28

27

Reference

13

23

24

25

26

Lactose

Maltotriose

Lactose

Lactosamine HO

HO OH

HO

HO OH

AcHN

O

OH

O

O

O

HO

HO

OH

OH

AcHN

O

OH

O

H N

H N

O

O

DMSO

DMSO

DMSO, H2 O

DMSO, H2 O

AIBN

60◦ C

AIBN

AIBN, KPS

60◦ C

60◦ C

AIBN, KPS

60◦ C

(continued)

29

29

13b

13b, 30

14 27

28

29

Glucuronamide

Galactose/ gluconamide HO

HO OH

HO

O

HO HO

Entry Monomer

Glucose

Carbohydrate

O

OH

OH O

NH


DMSO

DMSO

DMSO

Solventa AIBN

AIBN

AIBN

60◦ C

60◦ C

60◦ C


30

30

30

Reference

15

30

31

32

␣,␣-Galactotrehalose

␣,␤-Galactotrehalose

Glucosamine HO HO

HO HO

HO HO

O

O

H N

n

OH

NH

O

O

OH

NH

O

O

OH

O HN

OH

NHAc

O

O

HN

HN

n = 1 or 2 or 3 or 9

OH

HO O HO

O

OH

HO

HO O

O

OH

O

H2 O

DMSO, H2 O

DMSO, H2 O

RT

60◦ C

60◦ C

APS

AAPD

AAPD

(continued)

32

31

31

16 33

34

35

36

Chitobiose

Lactosamine

Galactose O O

HO

HO

Me

HO OH

HO HO

OH

Entry Monomer

Lactosamine

Carbohydrate

O

O O O

HO

OH

O

O

AcHN

O

O

OH

OH

HO

O

O

OH

NHAc

O

AcHN

O

O

O


Copolymerization in different solvents

H2 O

H2 O

H2 O

Solventa

65◦ C

RT

RT

RT

AIBN

APS

APS

APS


34

33

32

32

Reference

17

38

39

Lactitol

Glucose

HO

HO

HO HO

HO

O

OH O HO

HO O HO

O

OH

O

OH

O

OH OH

O

O

OH

OH OH

O

(CH2)8

O

O (CH2)8

O

H2 O or methanol ethanol 2-propanol

H2 O/DMSO

H2 O/DMSO

60◦ C

60◦ C or 35◦ C

60◦ C

ACPA

AAPD or H2 O2 with l-ascorbic acid

AAPD or H2 O2 with l-ascorbic acid

b APS

a DMSO

= dimethyl sulfoxide; THF = tetrahydrofuran; MeOH = methanol; DMF = N,N-dimethyl formamide. = ammonium persulfate; DODA-ACPA = dioctadecylamine-functionalised 4,4 -azobis(cyanopentanoic acid); V50 = 2,2 -azobis(2-methylpropionamidine) dihydrochloride; AIBN = 2,2 -azobisisobutyronitrile; KPS = potassium persulfate; AAPD = 2,2 -azobis(2-amidinopropane)dihydrochloride. c RT = room temperature.

37

Maltitol

36

35

35

18


2-(␤-d-galactosyloxy)ethyl methacrylate (GalEMA; Table 1.1, entry 14), which were then polymerized in a water–methanol mixture. Poly(GalEMA) synthesized via the second approach was tested for the binding with peanut agglutinin (PNA) and the thermodynamic binding parameters were calculated [23]. In another report, Cuervo-Rodriguez and co-workers synthesized methacrylate derivatives bearing acetylated glucopyranoside (Table 1.1, entry 15) and galactopyranoside (Table 1.1, entry 16) residues. Glycopolymers were then obtained by homopolymerization (and copolymerization with methyl methacrylate) in chlorobenzene, and their binding to Ricinus communis agglutinin (RCA120 ) was investigated after deprotection using methoxide [24]. 1.3.3 Styrene-Based Monomers Wulff et al. invested significant effort into the synthesis and polymerization of styrenic glycomonomers (Table 1.1, entries 17–19). The oxidation of sugars to aldehydes and a subsequent Grignard reaction using 4-vinyl-phenylmagnesium chloride was the preferred method for generating the glycomonomers [6b]. Deprotection after the free radical polymerization in water produced polymers with high molecular weights [25]. Narain et al. polymerized the same monomer 4vinylphenyl-d-gluco(d-manno)hexitol (Table 1.1, entry 17) using 2,2 -azobis-(2amidinopropan)dihydrochloride (AAPD) initiator in water to obtain copolymers with acrylamide [26]. Thermal properties of the polymers were studied by differential scanning calorimetry (DSC). In the same report, the monomer was copolymerized with styrene in DMF using AIBN as initiator. Kurth and co-workers prepared a new type of sugar monomer with an oxime linkage, d-lactose-O-(p-vinylbenzyl)oxime (Table 1.1, entry 20) and polymerized it using similar conditions as described above. High-molecular-weight glycopolymers were obtained that displayed narrow polydispersities, a feature attributed to the processing method: precipitation into methanol and two-stage thermal degradation [27]. Similar glycomonomers (Table 1.1, entry 21) containing a urea linkage were polymerized by the same researchers, with the resulting polymers displaying multimodal molecular weight distributions and high glass transition temperatures due to the presence of urea [28]. Kobayashi et al. prepared styrene derivatives with maltose, lactose, and maltotriose substituents on each benzene ring (Table 1.1, entries 22–24) by coupling the corresponding oligosaccharide lactones with p-vinylbenzylamine [13b]. The monomers were then polymerized in either dimethyl sulfoxide (DMSO) using AIBN or in water using potassium peroxydisulfate at 60◦ C. The maltose- and maltotriose-containing polymers interacted specifically with Con A. Kobayashi et al. reported the synthesis of other types of p-vinylbenzamide glycoside derivatives (Table 1.1, entries 25–26). They were homo- and copolymerized with acrylamide using 2,2 -azobisisobutyronitrile (AIBN) as initiator in DMSO at 60◦ C [29]. They also investigated the interaction of the glycopolymers with lectins by means of a two-dimensional immunodiffusion test in agar and inhibition of the

CONVENTIONAL FREE RADICAL POLYMERIZATION

19

hemagglutinating activity. It was found that the specificity of lectins with these glycopolymers was similar to that reported for naturally occurring glycoconjugates and binding between wheat germ agglutinin lectin (WGA) and poly[(p-vinylbenzamido)␤-diacetylchitobiose] was increased by 103 times compared with that of the oligosaccharide itself. Similar monomers (Table 1.1, entries 27–29) have also been synthesized by Akaike and co-workers and polymerized with AIBN in DMSO [30]. They investigated the specific binding of the glucose-derivatized polymers to the asialogylcoprotein receptor of mouse primary hepatocytes. More recently, Nishida and co-workers reported the synthesis of novel vinyl monomers (Table 1.1, entries 30–31) bearing galacto-trehalose (GT), a novel class of 1,1 -linked nonreducing disaccharide possessing an ␣-galactoside epitope. The monomers were copolymerized with acrylamide, and the resulting glycopolymers showed specific binding to ␣-galactoside-specific proteins (BSI-B4 lectin and Shiga toxin-1) [31]. 1.3.4 Other Vinyl-Containing Glycomonomers Monomers derived from N-acetyl-d-glucosamine (Table 1.1, entry 32) and N-acetyld-lactosamine (Table 1.1, entry 33) and chitobiose (Table 1.1, entry 34) were synthesized by Nishimura and co-workers [32]. They were polymerized with acrylamide in deionized water in the presence of ammonium peroxodisulfate (APS) and tetramethylethylenediamine (TMEDA) at room temperature. The synthetic glycopolymers exhibited good solubility in water and specific adhesion to rat hepatocytes. This research group later reported the preparation of a trisaccharide monomer with a Lex structure, n-pentenyl-O-(␤-d-galactopyranosyl)-(1-4)-[O-(␣-l-fucopyranosyl)(1-3)]-2-acetamido-2-deoxy-␤-d-glucopyranoside (Table 1.1, entry 35) [33]. The monomer was then polymerized under the same conditions as described earlier. A monomer based on vinyl ketone, 7,8-didesoxy-1,2:3,4-di-O-isopopylidene-␣d-galacto-oct-7-ene-1,5-pyranose-6-ulose (Table 1.1, entry 36) was copolymerized with a range of other monomers to generate hydrophilic surfaces. The reactivity ratios of different glycomonomers with methyl methacrylate (MMA), styrene, and acrylonitrile were determined [34]. Chemoenzymatically synthesized matitol- and lactitol-based glycomonomers (Table 1.1, entries 37–38) were polymerized to afford glucose-containing and galactosecontaining polymers. The glycopolymers showed specific biological activities toward Con A and RCA120 . Furthermore, positive adhesion to hepatocytes was also observed [35]. The solvent and oxygen effects on the free radical polymerization of 6-Ovinyladipoyl-d-glucopyranose (6-O-VAGlc; Table 1.1, entry 39) were recently investigated by Albertin and co-workers [36]. They polymerized 6-O-VAGlc in water and different alcohols at 60◦ C in the presence of 4,4 -azobis(cyanopentanoic acid) and found that in all cases long polymerization times (>24 h) were necessary to achieve reasonable conversions, and oxygen removal was critical for the success of the experiments.

20


1.4 CONTROLLED/LIVING RADICAL POLYMERIZATION Controlled/living radical polymerization techniques have received widespread interest in recent years due to their ability to produce polymers with precise architectures, predefined compositions, and narrow molecular weight distributions, all of which are inaccessible via conventional radical polymerization. The term living polymerization implies that irreversible chain transfer and termination events are absent, a condition that is not strictly upheld in controlled/living radical polymerization since termination events are unavoidable; however, many of the features commonly associated with living polymerization are still attained [37]. Living characteristics include: r The linear evolution of molecular weight with monomer conversion r A constant concentration of active species, which is indicated by a linear plot of ln([M]0 /[M]t ) vs. time r Narrow molecular weight distributions, with the polydispersity index (PDI = Mw /Mn ) remaining below 1.2; a conventional radical polymerization in which termination occurs exclusively by combination gives a theoretical minimum PDI of 1.5 r The ability to polymerize until all monomer is consumed and continue chain growth with the addition of more monomer due to the retention of active end groups Glycopolymers have been synthesized in a controlled/living fashion using the stable free radical polymerization techniques nitroxide-mediated polymerization (NMP) and cyanoxyl-mediated polymerization, the atom transfer radical polymerization (ATRP) technique, and the reversible addition–fragmentation chain transfer (RAFT) technique. 1.4.1 Stable Free Radical Polymerization

1.4.1.1 Nitroxide-Mediated Polymerization Nitroxide-mediated polymerization was developed as a controlled polymerization technique by Solomon et al. in 1985 [38], but it was not until the end of the 1990s that its potential for the synthesis of well-defined glycopolymers was finally realized. Nitroxide-mediated polymerization relies on the reversible capping of growing radical chains with nitroxide species, which are known as persistent radicals (2 in Scheme 1.1). The nitroxides themselves kt Pn

R O N R'

kact k deact

Pn

Dead polymer +

R O N R'

M k p 1

2

SCHEME 1.1 Mechanism of NMP.

21

CONTROLLED/LIVING RADICAL POLYMERIZATION

are not capable of initiating polymerization but instead serve to reduce the active radical concentration in the system and thereby limit bimolecular termination events. The initiating radicals can originate from the fragmentation of an alkoxyamine (as in Scheme 1.1) or can be provided by another source such as a conventional free radical initiator; in the latter case, addition of a nitroxide rather than the corresponding alkoxyamine is the most suitable strategy for controlling the polymerization. Nitroxides are often commercially available as stable radical species. Additives known as accelerators are also commonly employed in NMP and serve the role of regulating the concentration of free nitroxide in the system, which would otherwise build up and retard the polymerization as propagating radicals are inevitably lost through termination. Sulfonic acids, conventional radical initiators, and unstable nitroxides are often used for this purpose [39]. Controlling agents used for the NMP of glycomonomers are shown in Figure 1.1. N1 (TEMPO, a very common controlling agent), N2, N6, and N10 are nitroxides, whereas the remaining species are alkoxyamines, which fragment to generate their corresponding nitroxides. Detailed discussion on the affect of nitroxide structure on NMP kinetics and monomer choice is provided elsewhere [40]. The range of monomers that can be polymerized using NMP is more restricted than other controlled polymerization techniques, and glycopolymer syntheses have largely involved styrenic and acrylic monomers. The majority are protected monomers. Table 1.2 lists the glycomonomers polymerized by NMP. Ohno et al. were the first researchers to synthesize well-defined glycopolymers using NMP. N-(p-vinylbenzyl)-[O-␤-d-galactopyranosyl-(1-4)]-d-gluconamide (VLA; Table 1.2, entry 1) and its acetylated equivalent (AcVLA; Table 1.2, entry 2) were polymerized in N,N-dimethylformamide (DMF) at 90–105◦ C using N3 and a radical initiator as accelerator [41]. Initial attempts to polymerize VLA using N1 were unsuccessful due to partial decomposition of the monomer at the temperatures required for

N1

O

N3

N2

N

O

2,2,6,6-tetramethyl-1-piperidynyl-N-oxy (TEMPO)

O

N

O

O

N

O

O

N

N7

O

N

N

O

di-tert-butyl nitroxide (DBN) O

N6

N5

N4

N

N9

N8

N O

C18H37 N C18H37

C18H37 N C18H37

O

N O O

N

O P O OH O

N10

O

N O P O O

FIGURE 1.1 Nitroxide and alkoxyamine species used for NMP of glycomonomers.

22 1

2

3

Lactose

Glucose

OH

O O HO

OH

HO O

OH NH

O OAc AcO

O

OAc

O

OAc AcO

NH

O O O

O

3-O-acryloyl-1,2:5,6-di-O-isopropylidine-D-glucofuranose (AIpGlc)

O O

O

N-(p-vinylbenzyl)-2,3,5,6-tetra-O-acetyl-4-O-(2,3,4,6-tetraO-acetyl- -D-galactopyranosyl)-D-gluconamide (AcVLA)

AcO

AcOOAc

(1-4)]-D-gluconamide (VLA)

N-(p-vinylbenzyl)-[O- -D-galactopyranosyl-

HO

HOOH

Entry Monomer

Lactose

Sugar

TABLE 1.2 Glycomonomers Polymerized Using NMP

DMF p-xylene

DMF 1,2-dichloroethane

DMF

Solventa

105◦ C 100◦ C

105◦ C 90◦ C

105◦ C

N5, N6 N3

N3 N4

N3

Temperature Initiator

— DCP

DCP DCP

DCP

43 44

41 42

41

Acceleratorb Reference

23

4

5

Gluconolactone

Mannitol

O O O

CH OH

2,3-isopropylidine-1-(4-vinylphenyl) -D-threo(D-erythro)triol

CH,OH O O

2,3;4,5-diisopropylidine-1(4-vinylphenyl)-Dgluco(D-manno)pentitol

O

130◦ C

130◦ C

Bulk, diphenyl ether

Bulk

120◦ C

Bulk

N7

DCP

Poly(styrene)- — N1 N7 DCP

(continued)

46

46

45

24 6

7

8

Fructose

Galactose

O O

CH OH O O O

CH OH

OAc

OAc

O O

4-vinylbenzyl glucoside peracetate

AcO AcO

2,3;4,5-di-O-isopropylidine-1(4-vinylphenyl)-D-manno(D-gluco)hexulo-2,6-pyranose

O O

O O

1,2;3,4-di-O-isopropylidine-1-(4-vinylphenyl)D-glycero(L-glycero)- -D-galactopyranose

Entry Monomer

Galactose

Sugar

TABLE 1.2 Glycomonomers Polymerized Using NMP (Continued)

Chlorobenzene m-xylene

Bulk

Bulk

Solventa

125◦ C 138◦ C

130◦ C

130◦ C

DCP

DCP

47a 47b

46

46

Acceleratorb Reference

N8 CSA Poly(styrene)- — N1

N7

N7


25

b DCP

AcO

O O AcO

O

4

AcO

OAc

O AcO

OAc

AcO

O O O

O

OAc

O

2-(2′,3′,4′,6′-tetra-O-acetyl- -D-galactosyloxy) ethyl methacryl (AcGalEMA)

AcO

AcOOAc

4-vinylbenzyl maltohexaoside peracetate

AcO AcO

OAc

= N,N-dimethyl formamide. = dicumyl peroxide; CSA = (1S)-(+)-10-camphorsulfonic acid.

10

Galactose

a DMF

9

Maltohexaose

O

Dioxane

m-xylene

85◦ C

138◦ C

N9, N10

—

Poly(styrene)- — N1

48

47b

26


sufficiently rapid dissociation of TEMPO (> 120◦ C). In contrast, N3 proved effective at lower temperatures (90–105◦ C) due to its higher dissociation constant. Polymerization of VLA using various concentrations of N3 was reasonably well controlled at high N3 concentrations (targeting lower molecular weights); however, high conversions could not be reached, PDIs increased with conversion, and no polymerization occurred when targeting higher molecular weights (target DPn > 100). The addition of the radical initiator dicumyl peroxide (DCP) accelerated the polymerization rate (as expected) without significantly increasing the number of polymer chains in the system; PDIs were unaffected, but the same limiting conversions were observed. Substitution of VLA with its protected analog AcVLA transformed the system into a highly living one, with PDIs consistently around 1.10 and 90% conversion attainable in short polymerization time. Again, the radical initiator greatly enhanced the polymerization rate without affecting the degree of control. Interestingly, the undesirable features of the VLA polymerization are attributed to side reactions, in particular chain transfer to the unprotected hydroxyl groups of the monomer; however, numerous unprotected glycomonomers have since been successfully polymerized using other controlled radical polymerization techniques without such reactions occurring. Despite this discrepancy, almost all subsequent glycopolymer syntheses by NMP have been restricted to protected monomers. The same authors polymerized AcVLA using N4, an alkoxyamine with two long alkyl chains, to generate a lipophilic glycopolymer (Fig. 1.2) [42]. The polymerization was well controlled in 1,2-dichloroethane at 90◦ C, and after deprotection of the acetyl groups using hydrazine the resulting polymer formed a stable liposome when mixed with phospholipid. The liposomes specifically bound the galactose-specific lectin RCA120 . A similar approach was adopted by Götz et al., who polymerized 3-O-acryloyl-1,2:5,6-di-O-isopropylidine-d-glucofuranose (AIpGlc; Table 1.2, entry 3) in DMF at 105◦ C using N5 [43]. PDIs remained between 1.06 and 1.20, and the molecular weight increased linearly with conversion. Copolymerization of the glycomonomer with a lipophilic acrylamide monomer was also an effective strategy to generate amphiphilic lipo-glycopolymers. Polymerizations of AIpGlc in p-xylene at 100◦ C using both N3- and an N2-capped poly(styrene) macroinitiator were deemed “living” by Ohno et al., although the level

O

N

AcVLA 1,2-Dichloroethane Dicumyl peroxide 90°C

H C CH

O N

CH

Deprotection

H C CH

H C

CH

N O C H

C H

N O C H

HN OAc

N O C H

O OAc

AcO

O OH

O OH

O OAc O OAc AcO

HN OH

HO

O AcO AcO

O N n

n

C H C H

CH

OH

HO HO HO

FIGURE 1.2 Lipophilic glycopolymer synthesis by Ohno et al. [42].

27


of control more closely resembled that of the unprotected monomer VLA than the protected AcVLA; even in the best case the PDI climbed beyond 1.6 at only 25% conversion, and high molecular weights were unattainable [44]. Despite the questionable “livingness” of the process, deprotection of the isopropylidine groups using formic acid generated amphiphilic diblock copolymers that exhibited microphase separation when cast as thin films. Chen and Wulff produced diblock copolymers of styrene and 2,3;4,5-di-Oisopropylidine-1-(4-vinylphenyl)-d-gluco(d-manno)pentitol (Table 1.2, entry 4) and showed that either block could be polymerized first [45]. N7 (a TEMPO-derived alkoxyamine) was chosen as the mediating agent, and in contrast to the observations of Ohno et al. using VLA, no thermal decomposition of this monomer was noted even after polymerizing for 48 h at 130◦ C. The glycopolymer PDIs were around 1.33, and increased to 1.35–1.80 with the addition of different length poly(styrene) segments. After deprotection, films cast from the block copolymers showed surface properties that varied with the block lengths of the two components. A more extensive kinetic investigation into the N7-mediated polymerization of this and three other styrene-based glycomonomers 2,3-isopropylidine-1-(4-vinylphenyl)-d-threo(d-erythro)triol (Table 1.2, entry 5), 1,2;3,4-di-O-isopropylidine-1-(4-vinylphenyl)-d-glycero(l-glycero)␣-d-galactopyranose (Table 1.2, entry 6), and 2,3;4,5-di-O-isopropylidine-1-(4vinylphenyl)-d-manno(d-gluco)-hexulo-2,6-pyranose (Table 1.2, entry 7), was performed by the same authors [46]. Polymerizations were conducted at 130◦ C in bulk without any monomer degradation; thermogravimetric analysis confirmed that the monomers are thermally stable up to 150◦ C. The polymerization of all four monomers saw a linear increase in Mn with conversion, and PDIs were low (1.24–1.28) when targeting low molecular weights, except in the case of entry 7 whose PDI was inexplicably large (> 1.50). High conversions were attainable for all but monomer 6, which was attributed to higher steric hindrance around its vinyl group. It is interesting to note that the presence of a free hydroxyl group in each monomer does not appear to have compromised the degree of control. Narumi et al. extended the NMP process to synthesize triblock copolymers of 4-vinylbenzyl glucoside peracetate (Table 1.2, entry 8) and styrene using N8, a difunctional TEMPO-based mediator [47a]. Polymerization of the glycomonomer was performed in chlorobenzene at 125◦ C for 5 h using (1S)-(+)-10-camphorsulfonic acid (CSA) as an accelerator to give a well-defined polymer (Mn 8.5 kDa, PDI 1.09) whose two terminal TEMPO groups were used to chain extend with styrene (Fig. 1.3). Small low-molecular-weight shoulders were observed in the size exclusion chromatography (SEC) traces when the polymerization time extended beyond 3 h, but

N O

Ch orobenzene 10-Camphorsulfonic acid 125°C

OAc AcO AcO

O O OAc

N O

CH CH

CH

CH

OAc AcO AcO

O O OAc

CH CH

O N m/2

Styrene Ch orobenzene 125°C

CH

CH

CH m/2

OAc AcO AcO

O O OAc

FIGURE 1.3 Triblock copolymer synthesis by Narumi et al. [47].

CH

O N n/2

28


restricting the polymerization time allowed generation of well-defined triblocks (Mn 26 kDa, PDI 1.17). The same monomer was used in the synthesis of star poly(styrene) with a glycopolymer core [47b]. Chains of well-defined poly(styrene) with TEMPO end groups were linked together by chain extending using divinylbenzene with either 4-vinylbenzyl glucoside peracetate or 4-vinylbenzyl maltohexaoside peracetate (Table 1.2, entry 9), thereby incorporating sugar groups into the core. The final star polymers had on average 18 arms, PDIs around 1.35, and after alkaline deprotection of the sugar groups were capable of encapsulating water-soluble molecules in chloroform solution. Ting et al. published the most recent report on NMP of a sugar monomer, namely 2(2 ,3 ,4 ,6 -tetra-O-acetyl-␤-d-galactosyloxy)ethyl methacrylate (AcGalEMA; Table 1.2, entry 10), which is the first report detailing the polymerization of a methacrylic sugar monomer by NMP [48]. Difficulties in polymerizing methacrylates by NMP (such as enhanced disproportionation between growing radicals and nitroxide) were overcome by copolymerizing with a small proportion of styrene, which reduced the average activation–deactivation constant and thereby suppressed the irreversible termination events. The commercially available alkoxyamine N9 (which contains an N10 nitroxide end group) was used to mediate the polymerization at 85◦ C in dioxane to generate the random copolymer poly(AcGalEMA-co-styrene), which exhibited linear evolution of molecular weight with conversion and a final PDI of 1.26. However, chain extension of this block with styrene at 115◦ C resulted in some low-molecular-weight tailing, which was attributed to loss of alkoxyamine end groups. Alternatively, chain extension of a well-defined poly(styrene) homopolymer with glycomonomer/styrene furnished the block copolymer with a PDI of 1.15. The polymerization was commenced at 120◦ C to encourage cleavage of the poly(styrene)-N10 macroinitiator before reducing the temperature to 85◦ C for the remainder of the polymerization, since the activation–deactivation constant for the methacrylate-based alkoxyamine is higher than that for styrene. Deprotection of the galactose acetyl groups afforded amphiphilic block copolymers capable of forming bioactive micelles and porous films.

1.4.1.2 Cyanoxyl-Mediated Polymerization Cyanoxyl radicals, NCO• , were found by Druliner to act as persistent radicals capable of mediating the polymerization of acrylates, methacrylates, and methacrylamides [49]. The NCO• radicals are generated in situ by the one-electron reduction of cyanate anions using arenediazonium ions (e.g., p-chlorobenzenediazonium cations, Scheme 1.2). The aryl radicals produced simultaneously initiate polymerization, and their associated cyanoxyl radicals (which are not capable of initiating polymerization) act as reversible capping agents to form dormant species that can be reactivated by homolytic cleavage of the C–O bond. The C–O bonds in these systems are more easily cleaved than the equivalent bonds in alkoxyamines, which offers a distinct advantage over NMP in the polymerization of thermally sensitive glycomonomers; cyanoxyl-mediated polymerizations are typically conducted from 25 to 70◦ C. In addition, tolerance to a wide range of functional groups including hydroxyls, amines, and carboxylic acids broadens the scope of potential monomer–solvent combinations. It must be noted, however,

29


N N+

Cl

+

-

OC

N

N2

+

Cl

OC

N

Dead polymer

X kt kact Cl

CH2

CH CH2

CH OC N

X

X

n

kdeact

Cl

CH2

CH CH2

CH

X

X

n

+

OC

N

X kp

SCHEME 1.2 Mechanism of cyanoxyl-mediated polymerization.

that while cyanoxyl radicals introduce some degree of mediation, the inability to target predefined molecular weights by stopping the polymerization at a particular time precludes cyanoxyl-mediated polymerization systems from being considered truly living/controlled. Low initiator efficiency due to irreversible termination of aryl radicals in the initial stages of the polymerization is responsible for this lack of predictability in molecular weights [50]. Chaikof’s work group pioneered the application of cyanoxyl-mediated polymerization for the synthesis of glycopolymers from unprotected monomers. Cyanoxyl radicals were generated in situ by the reduction of cyanate using pchlorobenzenediazonium cations, simultaneously producing p-chlorobenzene radicals that initiated polymerization (Scheme 1.2). The glycoolymers were designed as heparin and heparin sulfate mimetics, with later publications focusing on their various biomimetic capabilities. In their early work, nonsulfated and sulfated N-acetyl-d-glucosamine-based monomers with two different spacer arms between the vinyl group and the sugar (Table 1.3, entries 1–2) were copolymerized with acrylamide to generate statistical copolymers with different proportions of sugar groups [51]. The molecular weights increased with conversion and the final statistical copolymers displayed PDIs ranging from 1.10 to 1.47. A higher glycomonomer feed ratio did compromise the level of control, presumably due to the lower reactivity of the unactivated vinyl group. Indeed, further investigation revealed that homopolymerization of these alkene-derivatized monomers was not possible, prompting the development of acrylate alternatives (Table 1.3, entries 3–4) that were capable of homopolymerization [50]. It is interesting to note, however, that copolymers of these nonsulfated and sulfated acrylates with acrylamide generally displayed higher PDIs than those synthesized using the alkene glycomonomers, potentially due to the higher frequency of termination events in acrylate systems. Extension of the same synthetic protocol to include alkene-derivatized nonsulfated and sulfated lactose monomers (Table 1.3, entries 5–6) broadened the library of copolymers to test for heparin-like abilities [52]. In vivo, heparin sulfate is known to bind to fibroblast growth factor 2 (FGF-2) and promote its binding to FGF receptor1 (FGFR-1). When tested in this capacity, polymers containing either sulfated

30 1

2

3

Glucosamine

Glucosamine

O NHAc

O n = 3, 9

O NHAc

O n = 3, 9

NHAc

O O O

O

(4,5-Dihydroxy-6-hydroxymethyl3-methyl-carboxamidotetrahydro-2H-2pyranyloxy)ethyl acrylate

HO HO

OH

n=3: n-Pentenyl 2-acetamido-2-deoxy3,4,6-trisulfoxy-D-glucopyranoside n=9: n-Undecenyl 2-acetamido-2-deoxy3,4,6-trisulfoxy-D-glucopyranoside

O3SO – O3SO

–

OSO3–

n=3: n-Pentenyl 2-acetamido-2deoxy-D-glucopyranoside n=9: n-Undecenyl 2-acetamido-2deoxy-D-gluco pyranoside

HO HO

OH

Entry Monomer

Glucosamine

Sugar

H2 O/THF

H2 O

H2 O

Solventa

TABLE 1.3 Glycomonomers Polymerized by Cyanoxyl-Mediated Polymerization

50◦ C

50◦ C

50◦ C

ClC6 H4 N≡N+ BF4 − /NaOCN



Temperature Initiating System

50, 54

50–52

50–52

Reference

31

4

5

6

Glucosamine

Lactose

Lactose

O3SO – O3SO O NHAc

O O

O

OH

O O HO

OH

HO

O O

O3SO –

–

O O3SO

O3SOOSO3 –

O O3SO –

O O3SO

OSO3– O

n-Pentenyl O-(2,3,4,6-tetra-O-sulfo- -D-galactopyranosyl)(1-4)-2,3,6-tri-O-sulfo- -D-glucopyranoside

–

–

n-Pentenyl O-( -D-galactopyranosyl)-(1-4)-D-glucopyranoside

HO

HOOH

(4,5-Disulfoxy-6-sulfoxymethyl-3-methylcarboxamidotetrahydro-2H-2-pyranyloxy) ethyl acrylate

–

OSO3–

H2 O/THF

H2 O/THF

H2 O/THF, H2 O

50◦ C

50◦ C

50◦ C



(continued)

52

52

ClC6 H4 N≡N+ BF4 − /NaOCN 50, 53, 54

32

= tetrahydrofuran.

8

Lactose

a THF

7

OH

O O HO

OH

HO

O O N H

O

O3SO

–

O O O3SO

– –

O O3SO

OSO3– O N H

O

2-N-Acryloyl-aminoethoxyl-4-O-(2,3,4,6-tetra-O-sulfo- -Dgalactopyranosyl)-(1-4)-2,3,6-tri-O-sulfo- -D-glucopyranoside

O3SO

–

–

O3SO OSO3

–

2-N-Acryloyl-aminoethoxyl-4-O-( -D-galactopyranosyl)-(1-4)-D-glucopyranoside

HO

HOOH

Entry Monomer

Lactose

Sugar

H2 O

H2 O

H2 O 1:1 H2 O/THF H2 O

Solventa

65◦ C

50◦ C

50◦ C 50◦ C 65◦ C

BiotinC6 H4 N≡N+ BF4 /NaOCN VariousC6 H4 N≡N+ BF4 /NaOCN

ClC6 H4 N≡N+ BF4 − /NaOCN BiotinC6 H4 N≡N+ BF4 /NaOCN VariousC6 H4 N≡N+ BF4 /NaOCN

Temperature Initiating System

TABLE 1.3 Glycomonomers Polymerized by Cyanoxyl-Mediated Polymerization (Continued)

56

53, 54

54 55 56

Reference

33


RO OR

OR O

RO OR

O

O

O RO

O RO + O

N H

Biotin Biotin C H N N BF /NaOCN 1:1 H O/THF, 50°C

CH

CH O

RO OR

co CH

CH O

NH

OR O

RO

H N

NH

OR

O RO

OC N n

O O RO

R = H / SO

FIGURE 1.4 Synthesis of biotinylated statistical copolymers by Hou et al. [56].

N-acetyl-d-glucosamine or sulfated lactose groups showed a pronounced enhancement in the binding of FGF-2 to FGFR-1, particularly for the lactose candidate, compared to their nonsulfated counterparts. Interestingly, the linker length had no influence. Further investigation into the chaperone ability of sulfated lactose/acrylamide copolymers found that a particular copolymer composition (Mn 9.3 kDa, 10 mol% glycomonomer) was as effective as heparin sulfate in protecting FGF-2 from enzymatic, acidic, and thermal degradation and for promoting binding to FGFR-1 [53]. Acrylamide-based lactose monomers (Table 1.3, entries 7–8) were copolymerized with acrylamide in an identical fashion to above, giving statistical copolymers with PDIs ranging from 1.19 to 1.47 [54]. Polymers containing sulfated lactose groups demonstrated a significant anticoagulation ability, whereas nonsulfated lactose polymers and neither sulfated nor nonsulfated N-acetyl-d-glucosamine-containing polymers showed any effect. Interestingly, sulfated lactose homopolymers were outperformed by their copolymers with acrylamide. End-functionalization of glycopolymers synthesised by cyanoxyl-mediated polymerization was also reported by Chaikof’s work group. Two biotin-derivatized arylamine initiators (each with different spacer lengths between the biotin and aryl groups) were used to generate biotin-terminated lactose/acrylamide copolymers that interacted strongly with streptavidin (irrespective of spacer length) [55]. Elaborating on this concept, a series of arylamine initiators derivatized with alkoxy, amino and Fmoc-protected amino, biotinyl, hydrazido, and carboxyl functionalities were effective in mediating the familiar glycomonomer/acrylamide copolymerizations (Fig. 1.4) [56]. The unprotected amino species and the hydrazine-functionalized species both gave lower yields, a feature that was attributed to quaternization of the amino group that suppressed radical formation. The series of end-functionalized glycopolymers were promising candidates for various bioconjugation reactions. 1.4.2 Atom Transfer Radical Polymerization Atom transfer radical polymerization (ATRP, Scheme 1.3) is a controlled radical polymerization technique that was developed independently by the research groups headed by Sawamoto [57] and Matyjaszewski [58] in 1995. An ATRP system contains a halogenated organic compound (initiator) R–X, a transition metal Mn , which can increase its oxidation number, a complexing ligand L to stabilize the metal, and monomer M. Initiation involves abstraction of the initiator’s halogen atom by the metal complex, which simultaneously undergoes a single electron oxidation.

34


R X

+

Mn / L

k act k de act

Mn+1X / L

+

R

kt

Dead polymer

M k p

SCHEME 1.3 ATRP mechanism.

This gives an organic radical R• and a new metal complex Mn+1 X/L in which the metal’s oxidation number has increased by 1. The radical can add monomer units, thereby initiating chain growth, before abstracting the hydrogen atom from the metal complex and restoring its dormant state Pn –X. This halogen-capped polymer chain assumes the same role that the initiator occupies in the initiation step; it remains in its dormant state until activated by the metal complex to reform the radical Pn • , which can then add a few more monomer units before being deactivated. These atom transfer equilibria lie heavily toward the dormant species, which reduces the effective radical concentration and thereby limits bimolecular termination. Termination still occurs but is greatly suppressed, imparting living characteristics on the polymerization process. Atom transfer radical polymerization has been successfully performed using a variety of transition metals, but copper (Cu) complexes have proven to be the most efficient and versatile. The choice of ligand influences the relative rates of activation and deactivation and, therefore, the degree of control over the polymerization. Nitrogen-containing multidentate ligands are commonly employed for Cu-mediated ATRP, and those relevant to glycopolymer synthesis are shown in Figure 1.5. Figure 1.6 shows the initiators relevant to the present review. For simplicity, the bold code will be used in text to refer to each ligand and initiator, rather than including its full or abbreviated name. In cases where the initiating group was attached to another structure, such as a premade polymer or a surface, the structure will be stated followed by the code of the initiating group. For example, a poly(ε-caprolactone) (PCL) species end-functionalized with an ethyl 2-bromoisobutyryl group A1 will be denoted PCL-A1. The glycomonomers polymerized using ATRP are summarized in order of appearance in Table 1.4, including their polymerization conditions.

1.4.2.1 Linear Polymers Ohno et al. from Kyoto University reported the first ATRP of a sugar monomer by polymerizing 3-O-methacryloyl-1,2:5,6-di-Oisopropylidine-d-glucofuranose (MAIpGlc; Table 1.4, entry 1) using a CuBr/L3 catalyst in veritrole containing A1 as initiator at 80◦ C [59]. The first-order kinetic plots for the polymerizations of MAIpGlc were approximately linear, which means that the system obeys first-order kinetics with respect to monomer concentration. However, a twofold increase in the initiator concentration (holding all else constant) had a negligible effect on the rate of polymerization rp , whereas a doubling of the activator concentration increased rp by a factor of 3. Previous studies using styrene indicated that rp is first order with respect to both the dormant alkyl halide concentration and


L1

N

L2

N

N

L3

N

N

4,4'-Dimethyl-2,2'-bipyridine (DMDP)

2,2'-Bipyridine (bipy)

C7H15

35

L4

C7H15

C10H21

N

C10H21

N

N

4,4'-di-n-Heptyl-2,2'-bipyridine (dHbipy)

4,4'-bis(1-Decyl)-2,2'-bipyridine (dDbipy)

L5

L6 N

N

N

N

N N

N

1,1,4,7,7-Pentamethyldiethylene triamine (PMDETA)

1,1,4,7,10,10-Hexamethyltriethylene tetramine (HMTETA)

L7

L8

N

N N

N

N C8H17

N N-n-octyl-2-pyridylmethanimine

tris(2-(Dimethylamino)ethyl)amine (Me6-TREN)

FIGURE 1.5 Ligands employed Cu(I) catalyst systems for glycopolymer synthesis using ATRP.

A1

A2

A3

O Br

O

Br

Br

Ethyl 2-bromoisobutyrate

A4

Br Dibromoxylene

1-Phenylethyl bromide

A5

O O

Br

Ethyl 2-bromoisopropionate

(MeO)3Si CH2 CH2

O S Cl O

2-(4-Chlorosulfonylphenyl)ethyltrimethoxysilane

FIGURE 1.6 Initiators employed for glycopolymer synthesis using ATRP.

36

1

2

Glucose

O O

O

O

OAc

O O O

O

glucopyranosyloxy)ethyl acrylate (AcGEA)

2-(2′,3′,4′,6′-tetra-O-acetyl- -D-

AcO AcO

OAc

3-O-methacryloyl-1,2:5,6-di-Oisopropylidine-D-glucofuranose (MAIpGlc)

O O

O

Entry Structure

Glucose

Sugar

TABLE 1.4 Glycomonomers Polymerized Using ATRP

Chlorobenzene Chlorobenzene

Anisole Ethyl acetate Veratrole Anisole, ethyl acetate Toluene

Ethyl acetate

Ethyl acetate

Veratrole Ethyl acetate

Solventa

80◦ C

80◦ C

70◦ C

90◦ C 100◦ C 80◦ C 60◦ C

60◦ C, 60◦ C, 100◦ C 60◦ C

60◦ C

80◦ C

A2 A2

N-succinimidyl-A4

25-arm star silsesquioxane-A1 A1-poly(sulfone)-A1 Silica-A1 Silica-A5 Carbon nanotube-A1

A1 poly(2-(A1)ethyl methacrylate) A1


CuBr/L1 CuBr/L5

CuBr/L8

CuBr/L2 Ni(PPh3 )2 Br2 CuBr/L3 CuBr/L6

CuBr/L6, CuBr/L5 Ni(PPh3 )2 Br2 CuBr/L6

CuBr/L3 CuBr/L6

Catalyst

61–63 64

92

82 85 86 87

81

80

59 79

Reference

37

3

4

Glucose

Lactose

OH

O O HO

OH OH H N HO O O

2-Lactobionamidoethyl methacrylate (LAMA)

HO HO

OH

O

O

methacrylate (GAMA)

OH H N HO O

D-gluconamidoethyl

HO HO

OH

O

25◦ C 35◦ C 30◦ C

NMP DMSO

MeOH/H2 O

NMP, 3:2 MeOH/H2 O NMP H2 O, MeOH/H2 O NMP NMP/H2 O

4-Arm PCL-A1 Gold-A1 Biotin-PEG-A1 A1-disulfide-A1

Aldehyde-A1 25◦ C 20◦ C 20◦ C 25◦ C

Titanium-A1 20◦ C

Poly(ethylene oxide)-A1 Poly(ethylene oxide)-A1 4-Arm star PCL-A1 4-Arm poly(peptide)-A1 12-Arm star PCL-A1 A4poly(pseudorotaxane)A4 Membranepoly(HEMA)-A4 Gold-A1 25◦ C

20◦ C

25◦ C 25◦ C

NMP NMP

H2 O, MeOH/H2 O 3:2 MeOH/H2 O

20◦ C

20◦ C

MeOH, MeOH/ H2 O, H2 O MeOH

CuBr/L1 CuBr/L1 CuBr/L1 CuBr/L1

CuBr/L1

CuBr/CuBr2 /L1

CuBr/L1

CuBr/CuBr2 /L1

CuBr/L1 CuBr/L5

CuBr/L1 CuBr/L1

CuBr/L1

CuBr/L1

(continued)

75 89 95 98b

65b, 66

90

89

88

77 78

74 76

66

65a

38

5

6

7

8

Lactose

Glucose

Glucose

O

O O

O OAc AcO

O

OAc

AcO

O O O

O

O O

O

O

OH

O O O

O

2-methacryloxyethyl glucoside

HO HO

OH

3-O-acryloyl-1,2:5,6-di-O-isopropylidine-Dglucofuranose (AIpGlc)

O O

O

2-O-acryloyloxyethyl-(2,3,4,6-tetra-O-acetyl-β-D-galactopyranosyl)(1-4)-2,3,6-tri-O-acetyl-β-D-glucopyranoside (AEL)

AcO

AcOOAc

6-O-methacryloyl-1,2;3,4-di-Oisopropylidine-D-galactopyranose (MAIpGal)

O

OO

O

Entry Structure

Galactose

Sugar

TABLE 1.4 Glycomonomers Polymerized Using ATRP (Continued)

3:2 MeOH/H2 O

Ethyl acetate

Chlorobenzene

THF Toluene

Anisole Chlorobenzene

Solventa

25◦ C

60◦ C

100◦ C

90◦ C 60◦ C, 60◦ C, 25◦ C 60◦ C 70◦ C

Azide-A1

A1

A3

A1-PCL-A1 A4, A1, A1 PCL-A1 N-succinimidyl-A4


CuBr/L1

CuBr/L5

CuBr/L1

CuBr/L4 CuCl/L1, CuBr/L1, CuBr/L5 CuBr/L6 CuBr/L8

Catalyst

91

83, 84

73

72 92

70 71

Reference

39

11

12

Glucosamine

Mannose

O

O O

OH HO O O

O NHAc

O

O NHAc

O

O

O

O

O

O

O

(MEMan)

2-methacryloxyethyl-D-mannopyranoside

HO HO

HO HO

OH

AcO AcO

OAc

6-O-(4-vinylbenzyl)-1,2:3,4-di-O-isopropylidineD-galactopyranose (VBIG)

O

OO

MeOH

85:15 MeOH/H2 O 3:1 MeOH/H2 O

DMSO, MeOH

DMSO, MeOH

Chlorobenzene

MeOH = methanol; NMP = N-methyl-2-pyrrolidone; DMSO = dimethyl sulfoxide; THF = tetrahydrofuran.

10

Glucosamine

a

9

Galactose

Pyrene-PCL-A1

Peptide-A1 Pyridyl disulfide-A1 A1-disulfide-A1

23◦ C 30◦ C 25◦ C

25◦ C, 30◦ C Biotin-A1

25◦ C, 30◦ C Biotin-A1

115◦ C

CuBr/L1

CuBr/CuBr2 /L1

CuBr/L7, CuBr/L1 CuBr/L1

CuBr/L7, CuBr/L1

CuBr/L1

98a

97

96

94

94

93

40


the activator concentration [60]. Despite these unexpected kinetic features, the polydispersity index (PDI) of the resulting glycopolymers was less than 1.30, indicating that the polymerizations were reasonably well controlled. Liang and co-workers were one of the first research groups to generate well-defined glycopolymers using ATRP [61]. The polymerization of 2-(2 ,3 ,4 ,6 -tetra-O-acetyl␤-d-glucopyranosyloxy) ethyl acrylate (AcGEA; Table 1.4, entry 2) was performed in chlorobenzene at 80◦ C using CuBr/L1 catalyst and A2 initiator. The polymerization obeyed first-order kinetics up to 70% conversion, after which the rising viscosity caused deviation from linearity. The measured molecular weights agreed with theoretical values and increased linearly with conversion, and the PDIs remained less than 1.4 throughout. Chain extension of a poly(styrene)-A2 macroinitiator under the same conditions generated well-defined block copolymers (PDI 1.31–1.37) as shown in Figure 1.7 [62]. Deprotection of the hydroxyl groups on the sugar units using sodium methoxide afforded amphiphilic block copolymers that self-assembled in water to form spheres, rods, vesicles, tubules, and finally large compound vesicles as the polymer concentration increased from 0.1 to 2.0 wt% [63]. The CuBr/L5-catalyzed ATRP of the same monomer using a poly(ethylene glycol)-A1 (PEG-A1) macroinitiator furnished a well-defined hydrophilic-hydrophilic block copolymer (after deprotection) that assembled into aggregates capable of binding concanavalin A (Con A) [64]. Narain and Armes synthesized the two unprotected monomers 2-gluconamidoethyl methacrylate (GAMA; Table 1.4, entry 3) and 2-lactobionamidoethyl methacrylate (LAMA; Table 1.4, entry 4) by reacting 2-aminoethyl methacrylate with d-gluconolactone [65]. Homopolymerizations were performed at 20◦ C using CuBr/L1 catalyst and a PEO-A1 macroinitiator in different methanol–water solvent combinations. A drastic increase in polymerization rate and a corresponding decline in control was observed when the polymerization medium was changed from pure methanol to 9:1 methanol–water, 3:2 methanol–water, and finally pure water (Fig. 1.8); full conversion was reached in 15 h using methanol (final PDI of 1.19) and only 30 min in water (final PDI of 1.82). [65a] The polymerization of LAMA proceeded with good control in 3:2 methanol–water (final PDI of 1.10) but not in pure water (final PDI of 1.78); its highly polar nature prohibited dissolution in pure methanol. Various GAMA- and LAMA-containing stimuli-responsive block copolymers were synthesized from macroinitiators or by the direct addition of a second methacrylic monomer to the homopolymerization system at complete conversion [66]. These results are consistent with numerous literature reports on the effect of solvent on the ATRP process. Polar solvents generally increase the rate of polymerization CH3

CH

CH2

CH

Br m

AcGEA CuBr/L1 Chlorobenzene 80°C

CH3

CH

CH2

CH

m

CH2

CH

n O

O OAc AcO AcO

O OAc

Br

O

Deprotection

CH3

CH

CH2

CH

m

CH2 OH

HO HO

CH

n O

O O

OH

Br

O

FIGURE 1.7 Synthesis of amphiphilic block copolymers of poly(styrene)-b-poly(2(2 ,3 ,4 ,6 -tetra-O-acetyl-␤-d-glucopyranosyloxy) ethyl acrylate) by Li et al. [63a].

4 Water

3.5

9:1 Methanol/water

2.5

1.8

14000

1.7

12000

1.6

10000

1.5

Mn

Ln([M ]0/[M ])

3

16000

2

8000

1.4

6000

1.3

4000

1.2

0.5

2000

1.1

0

0

1.5 Methanol

1

0

5

10 Time (h)

15

0

20

40

60

80

Mw /Mn

41


1 100

Conversion (%)

FIGURE 1.8 Pseudo-first-order kinetic plots (left) for the ATRP of 2-gluconamidoethyl methacrylate (GAMA) in different solvent systems, and number-average molecular weight and PDI as a function of conversion (right) for the methanol system. In both cases, [GAMA]0 = 1.675 mol L−1 , [GAMA]0 :[PEO-A1]0 :[CuBr] 0 :[L1] 0 = 50:1:1:2 and T = 20◦ C. Reproduced with permission from ref. [66].

by increasing the activation rate coefficient and decreasing the deactivation rate coefficient. Aqueous systems are especially prone to this effect [67]. Matyjaszewski and Xia state that in a bulk or apolar solvent the a copper catalyst complex is neutral [68] but in aqueous systems most likely forms the cationic [Cu(I)(L1)2 ]+ species with halide counterion [69]. The ionic catalyst is far more active than its neutral counterpart and as a result generates a high concentration of active radicals that increases the polymerization rate but also encourages termination events such as radical coupling and disproportionation reactions. The control over the polymerization is therefore compromised.

1.4.2.2 ABA and Star-Block Copolymers Chen and Wulff synthesized triblock copolymers in which the two terminal blocks are the same (denoted ABA triblock copolymers from here onwards) and star block copolymers by polymerizing 6-O-methacryloyl-1,2:3,4-di-O-isopropylidine-d-galactopyranose (MAIpGal; Table 1.4, entry 5) using linear and four-arm poly(ε-caprolactone)-A1 (PCL-A1) macroinitiators [70]. Linear and star PCL species are conveniently synthesized by ring-opening polymerization (ROP) of ε-caprolactone and can be easily functionalized with ATRP initiators. PCL-based macroinitiator species are therefore popular for the synthesis of block copolymers using ATRP. In Chen and Wulff’s work, a complete shift in the molecular weight distribution and PDIs less than 1.19 confirmed successful and controlled chain growth with no evidence of high-molecular-weight products that are sometimes formed in star synthesis by ATRP. A 5:1 Cu(0)/Cu(II) ratio with L4 ligand was used as the catalyst system. A detailed investigation into the effect of initiator, catalyst, ligand, and temperature on the ATRP of MAIpGal was presented by Meng et al. [71]. Due to its higher initiating efficiency, A1 was more effective than A4 in controlling the polymerization using CuBr/L1 catalyst at 60◦ C, although replacing

42


CuBr with CuCl did improve the control in the A4-initiated system. A1/CuBr/L5 at room temperature proved equally effective. Suriano et al. demonstrated that copolymerization of MAIpGal with ␣-methoxy, ␻-methacrylate poly(ethylene oxide) using a PCL-A1 macroinitiator is well controlled by CuBr/L6 in tetrahydrofuran (THF) at 60◦ C, generating amphiphilic block copolymers after deprotection [72]. Atom transfer radical polymerization was combined with ring-opening polymerization (ROP) by Dong et al. to generate ABA triblock copolymers of polypeptide-b-poly(2-acryloxyethyl-lactoside)-b-polypeptide [73]. ATRP of the acetyl-protected lactose monomer 2-O-acryloyloxyethyl-(2,3,4,6-tetra-O-acetyl-␤d-galactopyranosyl)-(1-4)-2,3,6-tri-O-acetyl-␤-d-glucopyranoside (AEL; Table 1.4, entry 6) was performed using A3 initiator and CuBr/L1 complex in chlorobenzene at 100◦ C (Fig. 1.9). The absolute molecular weights as determined by multiangle laser light scattering were consistent with predicted values and the PDIs remained between 1.19 and 1.35. The bromo end groups of the isolated polymers were converted to amino groups, and the resulting diamines were used to initiate the ROP of l-alanine N-carboxyanhydride or ␤-benzyl-l-glutamate N-carboxyanhydride. Removal of the acetyl groups on the sugar units using hydrazine generated well-defined triblocks with the central glycopolymer sandwiched between two polypeptide blocks. Dai and Dong synthesized 4-arm stars starting with a 4-arm star poly(εcaprolactone) (SPCL) produced by ROP [74]. The hydroxyl end groups were converted to A4 species and the star was used to polymerize GAMA in N-methyl-2pyrrolidone (NMP) at room temperature employing CuBr/L1 as catalyst. NMP is a dipolar aprotic solvent that is often chosen over water since it effectively dissolves many polar monomers but avoids the poor control that is often observed in aqueous ATRP systems. The PDIs of the final star copolymers were between 1.09 and 1.33 for different block lengths of glycopolymer. Differential scanning calorimetry (DSC) analysis of the block copolymers showed that the crystallinity of the PCL block was greatly reduced by the presence of the glycopolymer, dropping from almost 70% for the SPCL alone to less than 2% after grafting the GAMA. A transition from micelles to wormlike rods to vesicles was noticed in the self-assembly of the star copolymers as the glycopolymer block length decreased. The stars also specifically recognized Con A in solution. A very similar study using a lactose-based monomer was published by the same authors [75], and the protocol was extended to the synthesis of star glycopolymers from a 4-arm polypeptide macroinitiator [76] and a 12-arm PCL macroinitiator [77].

AcAEL CuBr/L1

Br Br

a) End-group conversion

Chlorobenzene 100°C

n/2

O

AcOOAc

O

OAc O

AcO

Br

O OAc AcO

n/2

O

HOOH

O

O N H

H N

H

m/2

O

OH O

O AcO

H N

b) ROP c) Deprotection

HO OH

O HO

O O HO

FIGURE 1.9 Synthesis of the ABA triblock copolymer poly(l-alanine)-b-poly(2acryloxyethyl lactoside)-b-poly(l-alanine) by Dong et al. [73b].


43

A later publication by the same group presented the synthesis of polypseudorotaxane/glycopolymer hybrids [78]. ␣-Cyclodextrin (␣-CD) was threaded onto a difunctional PCL-A4 macroinitiator via a supramolecular inclusion reaction to give a polypseudorotaxane with ATRP-initiating end groups. It is interesting to note that the more efficient and, therefore, preferred ATRP-initiating group A1 was too bulky to penetrate the ␣-CD cavities; therefore, A4 had to be employed. The polypseudorotaxane was used as initiator to polymerize GAMA in DMSO at 35◦ C using CuBr/L5 catalyst. Different [monomer]0 :[initiator]0 ratios were used, giving final polymers with Mn values of 33–39 kDa and PDIs of 1.26–1.56. Wide-angle X-ray diffraction (WAXD) and DSC analysis confirmed that the crystallization of the PCL was completely suppressed by its inclusion into the ␣-CD cavities and that none of the ␣-CD units unthreaded during polymerization. The biohybrids self-assembled in solution to form spherical micelles or vesicles that specifically recognized Con A and not bovine serum albumin (BSA). Molecular “sugar sticks” were synthesized by Muthukrishnan et al. from narrowly dispersed poly(2-hydroxyethyl methacrylate) of 67 and 418 kDa [79]. The hydroxyl groups of these two polymers were converted to A1 groups to give the polyinitiator species poly(2-(2-bromoisobutyryloxy)ethyl methacrylate). In a preliminary study, the effectiveness of three different catalysts in controlling the homopolymerization of MAIpGlc in ethyl acetate using A1 initiator was investigated, finding that (PPh3 )2 NiBr2 at 100◦ C and CuBr/L6 at 60◦ C both demonstrated good control (final polymer PDIs of 1.18 and 1.19, respectively), but CuBr/L5 resulted in broader molecular weight distributions (final polymer PDI of 1.51) [80]. The CuBr/L6 catalyst system was therefore chosen for the sugar stick synthesis. Conversions were limited to 10% in order to obtain well-defined polymer brushes with poly(MAIpGlc) side chains; the high-molecular-weight polyinitiator’s PDI of 1.08 remained unchanged after grafting. However, the cleaved grafted chains displayed PDIs of approximately 1.30, which suggests that initiation was slow and therefore the level of control was not optimal. In addition, the initiator efficiency was lower than expected (0.23 < f < 0.38). Imaging of the brushes using scanning force microscopy (SFM) and cryogenic transmission electron microscopy (cryo-TEM) revealed highly uniform wormlike structures that were not fully stretched due to the relatively low grafting density. Star inorganic–organic hybrids were synthesized in a similar fashion by ATRP of MAIpGlc from an A1-functionalized 25-arm silsesquioxane [81]. ABA triblock copolymers were synthesized by Wang et al. by polymerizing MAIpGlc from an A1-terminated difunctional polysulfone (Mn 5.7 kDa, PDI of 1.34) in anisole at 90◦ C using CuBr/L2 [82]. The final polymer (Mn 12.5 kDa, PDI of 1.18) self-assembled after deprotection of the sugar units to give spherical aggregates in mixed N,N-dimethyl formamide (DMF)–water.

1.4.2.3 Hyperbranched Glycopolymers Hyperbranched glycopolymers were synthesised by Muthukrishnan et al. by using a technique called self-condensing atom transfer radical copolymerization (Fig. 1.10) [83]. The acrylic glycomonomer 3-O-acryloyl-1,2:5,6-di-O-isopropylidine-d-glucofuranose (AIpGlc; Table 1.4, entry 7) was copolymerized with the acrylic AB∗ initiator–monomer (or inimer)

44


*M B*

m

*A m

O O O

O

Br

+

O O

b a

O

CuBr/L5 Ethyl Acetate 60°C

O

O

*M

a

m

m

m

M*

m *M

m

m m

a m

B*

a

b

O O

A

m

m

m

b

M

m

a *B

m

b

m

A* B*

m

FIGURE 1.10 Muthukrishnan et al.’s synthesis of hyperbranched structures by copolymerizing the monomer AIpGlc (denoted M in the scheme) with the “initiator–monomer” species 2-(2-bromopropionyloxy)ethyl acrylate (BPEA, denoted A B∗ ). A∗ , B∗ , and M∗ are active groups and a, b, and m are reacted ones [83].

2-(2-bromopropionyloxy)ethyl acrylate (BPEA) using ATRP with the CuBr/L5 catalyst. The copolymerization introduces both AIpGlc units and BPEA units into the polymer chains, and each of the BPEA units is capable of initiating ATRP via its bromo group, which allows the generation of hyperbranched structures without significant crosslinking since the ATRP conditions minimize chain transfer and combination events. Investigation of the homopolymerization of AIpGlc using A1 indicated that a lower reaction temperature of 60◦ C was preferable to avoid the bimodal molecular weight distributions observed at 80 and 100◦ C as a result of recombination events. At 60◦ C in ethyl acetate the homopolymer PDIs remained between 1.09 and 1.14; therefore, 60◦ C was preferred for the generation of the hyperbranched structures. Removal of the isoproylidine groups gave hydrophilic hyperbranched polymers, which were used to generate thin bioactive glycopolymer films on a hydrophobic surface using a low-pressure plasma immobilization technique [84]. Hyperbranched structures based on MAIpGlc were generated in a similar fashion using (PPh3 )2 NiBr2 at 100◦ C and CuBr/L6 at 60◦ C in ethyl acetate [80]; (PPh3 )2 NiBr2 was preferred because it facilitated a higher degree of branching. A silicon wafer functionalized with a bromoester initiator facilitated the grafting of hyperbranched poly(MAIpGlc) from the surface [85]. As expected, the monomer feed ratio ␥ = [monomer]0 /[inimer]0 had a significant effect on the properties and morphologies of the resulting films since the degree of branching declines as less inimer is incorporated. The surface attained an enormous swelling ability once the isopropylidine groups were removed, with the wet surface thickness jumping from 4.6 to 29.9 nm.

1.4.2.4 Surface Grafting Surface-initiated ATRP is an effective method for achieving a high density of polymer brushes covalently linked to a surface, but maintaining a sufficient deactivator concentration during polymerization is crucial in maintaining the balance between the dormant and active species. In a conventional solution or bulk ATRP, sufficient Cu(II) is generated by reaction of the Cu(I) catalyst with the initiator at the start of the polymerization. However, the concentration of initiator in a surface-initiated ATRP is generally insufficient to achieve this purpose,

45


so alternative means are required in order to boost the deactivator concentration. The first approach is to add sacrificial initiator, which means that polymerization proceeds both in solution (initiated by the sacrificial initiator) and from the surface (initiated by the surface-immobilized initiator). The advantage of this approach is the production of free polymer, which is more easily analyzed than the grafted polymer on the surface, and which can often be assumed to represent the characteristics of the grafted polymer. However, the free polymer chains can be adsorbed to the surface of the substrate and interfere with the grafting process. In addition, monomer is consumed not only at the surface but in solution, which may be undesirable if the monomer is precious. The alternative approach is to add Cu(II) to the initial polymerization mixture, which ensures sufficient deactivator is present at the commencement of polymerization. Ejaz et al. used ATRP to functionalize a solid surface with a densely grafted glycopolymer layer (Fig. 1.11) [86]. MAIpGlc was grafted from a silica surface affixed with a monolayer of A5, which was attached using the Langmuir–Blodgett technique. The chlorosulfonylphenyl group of A5 is a highly effective ATRP initiator, and is claimed by the authors to be more capable than A1 at controlling the polymerization of MAIpGlc. In this case, free initiator was added to the system. Despite the bulkiness of the monomer, the thickness of the polymer layer as measured by ellipsometry increased concurrently with the Mn of the free polymer in solution, strongly supporting the proposal that the chains grow from the surface in a controlled fashion. The PDI of the free chains remained less than 1.2 up to high conversion, at which the Mn reached 60 kDa. Quantitative deprotection of the isoproylidine groups gave a surface covered with a high density of well-defined glucose-containing polymer chains. Gao and co-workers grafted linear and hyperbranched MAIpGlc glycopolymers from the surface of multiwalled carbon nanotubes (MWNTs) functionalized with A1 groups using CuBr/L6 at 60◦ C in ethyl acetate [87]. Thermogravimetric analysis (TGA) was used to determine the molecular weight of the polymer on the MWNTs and revealed that Mn of the grafted polymer increased linearly with monomer conversion, although the Mn values were much lower than the expected values due to the low initiating efficiency. Interestingly, the grafted polymer synthesized without sacrificial

Si O Si CH2 CH2

O S Cl O

MAIpGlc CuBr/L3, A5 Veratrole 80°C

Si O Si CH2 CH2

O S O

CH2

CH3 C C O

O O

O

Cl n

O O

Formic Acid Deprotection

Si O Si CH2 CH2

O S O

CH2

O

CH3 C

Cl n

C O OH

O

OH

HO

O OH

FIGURE 1.11 Surface grafting of MAIpGlc from a silica substrate by Ejaz et al. [86].

46


initiator demonstrated almost identical kinetics to the system containing the sacrificial initiator. When sacrificial initiator was used, SEC analysis of the free polymer showed the emergence of high-molecular-weight coupling products beyond approximately 45% conversion. This indicates that the MWNTs are not simply an inert scaffold but are actually involved in the polymerization mechanism since a conventional ATRP of the same monomer in an equivalent homogeneous system showed no highmolecular-weight coupling even beyond 80% conversion. This effect was attributed to the “unique electronic property of carbon” but was not explained in further detail. Microscopy of the grafted MWNTs revealed the distinct core–shell structure of the polymer-coated nanotubes. In the same study, hyperbranched glycopolymers were also grafted from the MWNT–A1 using an identical protocol to the group’s previous hyperbranched glycopolymer grafting publications [85]. A polypropylene microporous membrane was grafted with glycopolymer brushes by Yang et al. using a combination of ultraviolet (UV)-induced graft polymerization and ATRP [88]. The UV-promoted polymerization of 2-hydroxyethyl methacrylate (HEMA) introduced a high density of hydroxyl groups to the surface, approximately 62% of which were converted to A4 groups. GAMA was grafted from the substrate using CuBr/L1 catalyst in two different water–methanol mixtures at 30◦ C without the use of sacrificial initiator. A rapid initial polymerization rate followed by a plateau in the number-average degree of polymerization (as determined gravimetrically) was observed in the case where the methanol–water ratio was 3:2. Lowering the water content to 4:1 reduced the polymerization rate and imparted greater control over the polymerization. The addition of CuBr2 , as expected, markedly improved the control. Mateescu and co-workers polymerized GAMA and LAMA from a gold surface functionalized with ␻-mercaptoundecyl bromoisobutyrate (i.e., gold–A1) [89]. Polymerizations were performed at 20◦ C using CuBr/L1 catalyst in water or methanol/water mixtures. Either sacrificial initiator or Cu(II) deactivator were used to provide a sufficient concentration of Cu(II) to control the polymerization, although PDIs of free polymer produced from sacrificial initiator ranged from 1.6–1.8. An increase in the maximum film thickness and a decrease in the polymerization rate were observed as the water content of the polymerization medium decreased, which is consistent with numerous reports on the destabilizing effect of water in the ATRP process. Despite the rather poor control, the poly(GAMA)- and poly(LAMA)-grafted surfaces showed strong binding affinities for Con A and RCA120 , respectively. Raynor et al. performed surface-initiated ATRP of GAMA from a titanium surface functionalized with A1 groups using CuBr/CuBr2 /L1 in methanol–water [90]. The brush thickness increased for the first 4 h but then plateaued due to a loss of ATRP end groups as observed by X-ray photoelectron spectroscopy (XPS). The living behavior of the polymerization was not verified. The glycopolymer-grafted surface resisted protein and cell adhesion.

1.4.2.5 End-Functionalization and Bioconjugation An interesting application of ATRP and click chemistry was reported by Gupta et al. (Fig. 1.12) [91]. An azide-terminated glycopolymer was synthesized by the ATRP of the unprotected monomer 2-methacryloxyethyl glucoside (Table 1.4, entry 8) using an

47

CONTROLLED/LIVING RADICAL POLYMERIZATION OH

O N3

O

O

+

Br

O

O

O

O

HO HO

O

CuBr/L1 3:2 MeOH/H O 25°C

O

OH

N

O

O

Br n

O O

O

OH HO HO O

O

O

N N N

O

excess HN

O O

O

O HN

O OH

Br n

O

O

O

OH

fluorescein

"Click" conditions

O

O

O

HO HO

O OH

COOH

HO

H N

H N

O

O

"Click" conditions

H N

H N

N

N N N

Br n

CPMV

CPMV O

O

150

O

O fluorescein

O R

O 125±12

FIGURE 1.12 Synthesis of virus–glycopolymer conjugates using ATRP and click chemistry by Gupta et al. [91].

azide-terminated A1 derivative with CuBr/L1 in 3:2 methanol–water at 25◦ C to give a relatively well-defined polymer (Mn = 13 kDa and PDI of 1.30). The azideterminated polymer was reacted with an excess of fluorescein dialkyne to give a fluorescently labeled glycopolymer with a single alkyne group at the terminus. Azide groups were then installed on the outer surface of cowpea mosaic virus (CPMV) particles at 150 of the available 240 lysine locations, and 125±12 polymer chains per particle were able to be clicked on via these azide functionalities. This method of attaching polymer chains to virus particles was more efficient than previously reported strategies, and the resulting glycopolymer–CPMV conjugates were investigated further as targeting agents for overexpressed carbohydrate receptors on cancer cells. Indeed, strong and specific binding to both immobilized and free Con A occurred almost instantaneously. An N-(hydroxy)succinimide-terminated A4 initiator was used by Ladmiral et al. to polymerize the protected monomers MAIpGlc and MAIpGal with CuBr/L8 catalyst in toluene at 70◦ C to give well-defined homopolymers [92]. Initiating efficiencies were less than 50% compared to almost 100% for the analogous initiator based on A1, but the less efficient species was preferred because its reactivity toward bioconjugation to peptides and proteins was found to be far superior. Copolymerization of each monomer with a fluorescent methacrylate generated fluorescent statistical copolymers capable of reacting with primary amines courtesy of the terminal N(hydroxy)succinimide groups. A PCL species with a pyrene end group was synthesized by Lu et al., converted to a pyrene–PCL–A1 macroinitiator, and used for the polymerization of 6-O(4-vinylbenzyl)-1,2:3,4-di-O-isopropylidene-d-galactose (VBIG; Table 1.4, entry 9)

48


using CuBr/L1 in chlorobenzene at 115◦ C [93]. The block copolymer (Mn 9.0 kDa, PDI 1.28) was deprotected, self-assembled in water, and the glycopolymer shell was crosslinked using a difunctional aldehyde. Hydrolysis of the PCL core under basic conditions was proven by the loss of fluorescence from the pyrene group. Vázquez-Dorbatt and Maynard polymerized both a protected and unprotected monomer (Table 1.4, entries 10–11) synthesized by functionalizing N-acetyl-dglucosamine with 2-hydroxyethyl methacrylate (HEMA). The ATRP was carried out using a biotinylated A1 derivative with CuBr/L7 as catalyst in DMSO at room temperature [94]. The polymerization of the protected monomer was very rapid, reaching >90% conversion in 15 min. Some low-molecular-weight tailing was observed in the SEC traces when the [monomer]0 :[initiator]0 ratio was 100 (rather than 50 or 10), which suggests some early termination events occurred, but the PDIs of the final polymers were quite narrow, ranging from 1.17 to 1.23. Deprotection of the glycopolymers using a catalytic amount of methoxide occurred within minutes at room temperature but did not cleave any of the sugar units from the backbone or affect the biotin end group. The alternative to deprotection was the direct polymerization of the unprotected monomer under identical conditions, which interestingly afforded glycopolymers with narrower polydispersities (between 1.07 and 1.16) compared to the protected monomer system. The researchers also polymerized both monomers in methanol using a CuBr/L1 catalyst at 30◦ C and found that under these conditions a conversion comparable to the DMSO system was reached after 90 min and the polymerization was better controlled, with linear evolution of molecular weight with conversion and PDIs less than 1.13. Again, the unprotected monomer performed slightly better in terms of control. The biotinylated glycopolymers displayed a high affinity for the protein streptavidin. Narain was also interested in the synthesis of biotin-functionalized glycopolymers [95], and synthesized an ATRP initiator (Mn 5.1 kDa, PDI 1.07) that contained a PEG block end-functionalized with biotin. This biotin–PEG–A1 initiator was used to polymerize LAMA at 20◦ C in NMP using a CuBr/L1 catalyst system. The obtained polymers displayed PDIs less than 1.35. Binding of the biotinylated glycopolymers was investigated using a mutant streptavidin protein, which is much more conducive to biotin exchange assays than wild-type streptavidin. Fluorescein-biotin was used to first occupy all binding sites of the mutant streptavidin. Then in the presence of excess biotin-functionalized glycopolymer the fluorescent species was displaced and provided a means to quantify the degree of binding between the polymer and the protein. Unsurprisingly, the high-molecular-weight polymer (24 kDa) was sterically prohibited from accessing all four binding sites on each streptavidin, whereas the lower molecular weight species (11 and 16 kDa) were able to occupy all four binding sites. Broyer et al. developed a new procedure to modify amino acids with ATRP initiators (Fig. 1.13) [96]. A serine–A1 species was synthesized in five steps and used to initiate the ATRP of the protected glucosamine monomer (Table 1.4, entry 10) used by Vázquez-Dorbatt and Maynard [94]. After proving the robustness of the initiator for the polymerization of HEMA in MeOH at 23◦ C using CuBr/L1 at various [monomer]0 :[initiator]0 ratios, the initiator was applied to the polymerization of a methacrylic glucosamine monomer in a 85:15 methanol–water mixture. The small

49


V

M

O

H N

OH V

O O

Br

V

Q

T

K

G

+

HO HO

O

O O NHAc

O

CuBr/L1 4:1 MeOH/H O 23°C

V

M

O

H N

V

V

Q

O

Fluorophore

Br n

O O

OAc AcO AcO

T

K

G

Fluorophore

O

O O NHAc

FIGURE 1.13 Fluorescent peptide–glycopolymer conjugate synthesised by ATRP from an A1-functionalized amino acid by Broyer et al. [96].

fraction of water was necessary to prevent precipitation of the polymers at higher molecular weights. Slight curvature was noted in the pseudo-first-order kinetic plot, but Mn increased linearly with conversion and the PDI of the final polymer (15.4 kDa, 84% conversion) was low at 1.19, indicating that the presence of water did not have a significant detrimental effect on the polymerization. Given these promising results, the serine-based initiator was incorporated into a model peptide containing nine amino acid residues using solid-phase peptide synthesis, and polymerization under similar conditions generated a well-defined conjugate (12.2 kDa, 93% conversion) with PDI of 1.14. Proof that the polymer was indeed conjugated to the peptide was provided by incorporating a fluorophore into the lysine (K) residue and measuring the fluorescence of the purified polymer. This important work provided a more efficient and selective method of conjugating proteins with (glyco)polymers at particular amino acids, compared to previous attempts to selectively modifying premade sequences. Vasquez-Dorbatt and Maynard performed ATRP of the same unprotected monomer (Table 1.4, entry 11) using a pyridyl disulfide-containing A1 derivative in 3:1 methanol–water at 30◦ C utilizing a CuBr/CuBr2 /L1 catalyst system [97]. Polymerization in DMSO was exceedingly rapid, reaching 90% conversion in less than 5 min, and methanol, as previously discovered [96], was not suitable due to its inability to solubilize the higher molecular weight chains. Optimization of the polymerization using A1 revealed that 3:1 methanol–water containing 1:1 CuBr/CuBr2 was most effective in controlling the polymerization. A kinetic study using the pyridyl disulfide–A1 initiator under these optimal conditions showed a linear first-order kinetic plot up to 80% conversion and PDIs less than 1.2 throughout. The polymer used for subsequent steps had Mn of 13 kDa and PDI of 1.12. Interestingly, this measured Mn was much higher than the theoretical Mn of only 3.2 kDa, a discrepancy that was also observed in the polymerization of this monomer using the biotinylated initiator [94] but not with an amino-acid-containing initiator [96]. The same initiating group is common to all three systems; therefore, the low initiator efficiency for the pyridyl disulfide species was undetermined. Some chain transfer to the pyridyl disulfide species was also observed. After purifying the polymer, the pyridyl disulfide was conjugated to thiol-terminated small inhibitory ribonucleic acid (siRNA). A dithiothreitol (DTT) solution quantitatively reversed the conjugation to release free siRNA, which is promising for the use of this conjugate system for therapeutic siRNA delivery.

50


Symmetrical disulfide ATRP initiators with A1 end groups were utilized by Mizukami et al. and Kitano et al. to polymerize LAMA in 4:1 NMP–water and 2-methacryloxyethyl-d-mannopyranoside (MEMan; Table 1.4, entry 12) in methanol using CuBr/L1 at room temperature [98]. The MEMan polymerization was reasonably well controlled (PDI less than 1.5) up to high conversion, but control was poorer in 7:3 NMP–water. The disulfide-containing polymers were accumulated on surfaces covered by a monolayer of colloidal gold, and the resulting galactose- and mannose-displaying surfaces were able to reversibly associate Con A and RCA120 , respectively.

1.4.3 Reversible Addition–Fragmentation Chain Transfer Polymerization Reversible addition–fragmentation chain transfer (RAFT) polymerization is a controlled polymerization technique developed by researchers at CSIRO in 1998 [99]. Around the same time, French researchers revealed a technique referred to as macromolecular design by interchange of xanthate (MADIX), which operates via the same mechanism and can be viewed as a subset of RAFT [100]. RAFT is the most recent of the controlled polymerization techniques to garner widespread interest, primarily due to its tolerance to a wide range of reaction conditions and monomers that are difficult to polymerize in a controlled fashion using other controlled radical polymerization techniques. Glycomonomers are an important class of monomers that can be polymerized in a relatively routine manner using RAFT. In contrast to NMP and ATRP, the RAFT process is more conducive to the controlled polymerization of unprotected glycomonomers. Control is attained in a RAFT polymerization by the involvement of a thiocarbonyl thiocompound called a RAFT agent (I, Scheme 1.4). The RAFT process consists of the familiar initiation, propagation, and termination steps encountered in conventional free radical polymerization but with two superimposed equilibria controlling the growth of the polymer chains. The first equilibrium, called the pre-equilibrium, involves the addition of a propagating radical Pn • to the RAFT agent I, generating the RAFT-centred radical II. This radical can fragment in either direction; that is, revert to its original structure (and release Pn • ) or undergo ␤-scission to generate the macro-RAFT agent III and release the new radical R• . Either of the two released radicals is capable of undergoing propagation by reacting with monomer units. The pre-equilibrium ceases once every RAFT agent molecule in the system has undergone this forward fragmentation process to release its R• group and form a new oligomeric RAFT species. The second equilibrium, known as the main equilibrium, closely mirrors the preequilibrium, but in this case it is the macro-RAFT species III, which acts as the chain transfer agent. Addition of a propagating radical Pm • to III gives the RAFT-centred radical IV, which can fragment in the forward direction to release Pn • and generate another macro-RAFT species essentially the same as III (ignoring differences in chain length).

51


Initiation Initiator

M ki ,1

I

I M

M kp

Pn

Pre-equilibrium and propagation

Pn

+

kadd

S R

S

Pn S

k-add

S R

kβ k-β

Pn S

Z

Z

Z

I

II

III

S

+

S

+

R

M k p

Re-initiation M k i ,2

R

M kp

R M

Pm

Main equilibrium and propagation

Pm +

S

kaddP

S Pn

k-addP

Pm S

S Pn

k-addP kaddP

Z

Z

Pm S

Pn

Z M k p

M k p III

IV

III

Termination Pm

+

Pn

kt

Dead polymer

SCHEME 1.4 RAFT mechanism.

The structure of the RAFT agent is crucial to its ability to control the polymerization of a particular monomer. RAFT agents generally fall into four classes of compounds according to the nature of their Z groups; dithioesters (Z = R ), xanthates (Z = OR ), dithiocarbamates (Z = NR R ) and trithiocarbonates (Z = SR ). The Z group determines the susceptibility of the C=S bond to radical addition, and also the lifetime of the intermediate RAFT-centred radical (II, Scheme 1.4). The choice of R group also has a pronounced effect on the performance of the RAFT agent and allows fine-tuning of its overall reactivity. The R group must be a good free radical leaving group, and the resulting radical must be capable of adding to the monomer. Detailed explanations into the appropriate choice of RAFT agent for the controlled polymerization of a particular monomer are provided elsewhere [101] and will therefore not be discussed further. Table 1.5 contains the glycomonomers in order of appearance in

52 Entry 1

2

3

Glucose

Glucose

Mannose

OH

O O O

O

O

OH OMe

O

HO O

OMe

Methyl 6-O-methacryloyl-α-Dmannoside (6-O-MAMMan)

HO HO

O

O

Methyl 6-O-methacryloyl-D-glucoside (6-O-MAMGlc)

HO HO

O

2-Methacryloxyethyl glucoside (2-MAOEGlc)

HO HO

OH

Monomer

Glycomonomers Polymerized Using RAFT

Sugar

TABLE 1.5

9:1 H2 O/EtOH

9:1 H2 O/EtOH, H2 O 86:14 D2 O/d6 -DMSO

9:1 H2 O/EtOH

H2 O 9:1 H2 O/EtOH

Solventa

70◦ C

60◦ C

70◦ C

70◦ C

70 C 70◦ C

◦

ACPA

ACPA

ACPA

ACPA

ACPA ACPA


poly(2MAOEGlc)-C1

C1

C1

C1

C1 C1

RAFT Agent

103

108

103, 105, 119, 130 104

102 103

Reference

53

4

5

6

Mannose

Glucose

Galactose

HO O

O

OH

OH

O OH

O

O

O

O O

6-O-methacryloyl-1,2;3,4-di-Oisopropylidine-D-galactopyranose (MAIpGal)

O

OO

O

6-O-vinyladipoyl-D-glucopyranose (6-O-VAGlc)

HO HO

O

O

6-O-methacryloyl mannose (MAM)

HO HO

O

DMF

H2 O, MeOH DMAc

9:1 H2 O/EtOH DMAc

60◦ C

60◦ C 70◦ C

70◦ C 70◦ C

AIBN

ACPA ACPA

ACPA AIBN

C4, C5

C2, C3 C8

C1 Acetylene-C1

(continued)

110

109 116

107 123

54 Entry 7

8

Glucose

Glucose

NH

O O

O

(CH2)n

O

n = 5 10

O O

O

O

3-O-methacryloyl-1,2:5,6-di-Oisopropylidine-D-glucofuranose (MAIpGlc)

O O

O

3'-(1',2':5',6'-di-O-isopropylidine-D-glucofuranosyl)-6-methacrylamido hexanoate (n = 5) 3'-(1',2':5',6'-di-O-isopropylidine-D-glucofuranosyl)-11-methacrylamido undecanoate (n = 10)

O O

O

Monomer

Glycomonomers Polymerized Using RAFT (Continued)

Sugar

TABLE 1.5

Hexadecane/H2 O emulsion Ethyl acetate

n = 5: Dioxane n = 10: Anisole

Solventa

AIBN AIBN

75◦ C

AIBN AIBN

70◦ C

70 C 80◦ C

◦


C11

C4, C5, C6

C5 C5

RAFT Agent

121

114

113

Reference

55

9

10

11

Fructose

Galactose

Glucosamine

O

O

O

O

O

O O

O

NH

O OH

N-acryloyl-D-glucosamine (AGA)

HO HO

OH

6-O-(4-vinylbenzyl)-1,2:3,4-di-Oisopropylidine-D-galactopyranose (VBIG)

O

OO

3-O-methacryloyl-1,2;5,6-di-Oisopropylidine- -D-fructopyranose (MAIpFrc)

O O

O

5:1 H2 O/EtOH H2 O 5:1 H2 O/EtOH 5:1 H2 O/EtOH

Toluene

Hexadecane/H2 O emulsion

60◦ C 65◦ C 60◦ C 60◦ C

90◦ C

70◦ C

ACPA ACPA ACPA ACPA

AIBN

AIBN

C9, C10 ␤-CD-C9 C9 C9, silica-C9

C7

C4, C5, C6

(continued)

117 118 125 127

115

114

56 12

13

14

Glucose

Glucose

Entry

OH

O O O

O

O

NH

O OH

OAc

O O O

O

2-(2',3',4',6'-tetra-O-acetyl-β-D-glucopyranosyloxy) ethyl methacrylate (AcGEMA)

AcO AcO

OAc

2-Methacrylamido glucopyranose (MAG)

HO HO

OH

2-( -D-galactosyloxy)ethyl methacrylate (GalEMA)

HO

HOOH

Monomer


Galactose

Sugar

TABLE 1.5

MeOH

9:1 DMAc/H2 O

9:1 H2 O/EtOH

Solventa

70◦ C

60◦ C

70 C

◦

AIBN

AIBN

ACPA


C5

C1

C1

RAFT Agent

122

120

119, 129, 130

Reference

57

15

16

17

Galactose

Galactose

Lactose

O

O O

CHO

O

O O

O OAc AcO

O

OAc

AcO

O O

2-O-methacryloxyethyl-(2,3,4,6-tetra-O-acetyl-D-galactopyranosyl)-(1-4)-2,3,6-tri-O-acetyl-D-glucopyranoside (MAEL)

AcO

AcO OAc

6-O-acryloyl-1,2;3,4-di-O-isopropylidineα-D-galactopyranose (AIpGal)

O

OO

O

1,2:3,4-di-O-isopropylidine6-O-(2′-formyl-4′-vinylphenyl)D-galactopyranose (IVDG)

O

OO

O

O

Chloroform

␣,␣,␣Trifluorotoluene

THF

70◦ C

70◦ C

60◦ C

AIBN

AIBN

AIBN

C4

C12, PCL-C13

C6

(continued)

128

126

124

58 Entry 18

19

20

Glucose

Lactose

Galactose

OH

O O HO

OH OH H N HO O O

O

O O

6-O-acrylamido-6-deoxy-1,2:3,4-diO-isopropylidine-α-D-galactopyranose (GalAm)

O

O NH

O

2-Lactobionamidoethyl methacrylate (LAMA)

HO HO

OH

O

O

methacrylate (GAMA)

OH H N HO O

D-gluconamidoethyl

HO HO

OH

Monomer

O


Sugar

TABLE 1.5

Dioxane

5:1 H2 O/DMF

3:2 H2 O/DMF, 4:1 H2 O/MeOH

Solventa

90◦ C

60◦ C, 65◦ C

AIBN

ACPA

C16

C14, C15

C14, C15

60◦ C, 65◦ C ACPA

RAFT Agent


132, 133

131

131

Reference

59

21

22

23

24

Mannose

Glucosamine

Glucosamine

Glucose

O

NH

O

NHAc

O O

NH

O

O

N H

NH

O OH O O

OH H N HO O N H

O

2-Gluconamidoethyl methacrylamide (GAEMA)

HO HO

OH

2-Deoxy-2-N-(2′-methacryloyloxyethyl) aminocarbamyl-D-glucose (GUMA)

HO HO

OH

p-Acrylamidophenyl N-acetyl-β-glucosamine

HO HO

OH

p-Acrylamidophenyl α-mannoside

HO HO

OH HO O

7:1 H2 O/DMF 7:1 H2 O/DMF 1:1 H2 O/Dioxane

3:1 H2 O/MeOH

H2 O/DMSO

H2 O/DMSO

70◦ C 70◦ C 70◦ C

70◦ C

60◦ C

60◦ C

ACPA ACPA ACPA

ACPA

AAPD

AAPD

C1 C1 C13

C1

C17

C17

(continued)

136 137 139

135

134

134

60 25

26

27

Lactose

Glucose

Mannose

OH

O O HO

OH OH H N HO O N H

O

OH H N HO O H N O

N N N

2′-(4-vinyl-[1,2,3]-triazol-1-yl)ethylO-α-D-mannopyranoside

HO HO

OH HO O

3-Gluconamidopropyl methacrylamide (GAPMA)

HO HO

OH

2-Lactobionamidoethyl methacrylamide (LAEMA)

HO

HOOH

Monomer

H2 O/MeOH

5:1 H2 O/DMF

7:1 H2 O/DMF

Solventa

60◦ C

70◦ C

70◦ C

ACPA

ACPA

ACPA


C9

C1

C1, C13

RAFT Agent

14

137, 138

136 138

Reference

a EtOH = ethanol; D O = deuterium oxide; DMSO = dimethyl sulfoxide; DMAc = N,N-dimethyl acetamide; MeOH = methanol; DMF = N,N-dimethyl formamide; 2 THF = tetrahydrofuran. b ACPA = 4,4 -azobis(cyanopentanoic acid); AIBN = 2,2 -azobisisobutyronitrile; AAPD = 2,2 -azobis(2-amidinopropane)dihydrochloride.

Entry


Sugar

TABLE 1.5

61


C1 S

C2 S

CN OH

S

C3 OH

S

N

C4

S

CN

O

S O

S

O

S

O

O

C8 C7

C6

C5 S

O

S

S

O

O O

S

S

O

S S

CN

S

S

S O

S O

S

O

O

C9

S

S

O S

S

S

OH

C11

S

O

O

S

S

C13

C14

S

S

S

S

S

OH

HOOC

S

S

C12 C12H25

S S

O

S

S

C15 S

S

COOH

HOOC

O

H N

S

O

O

O

S

S

C12H25

C17

C16 S

O

O

S

O

O

O

S

S

C10

S

O S

S

S

N H

OH

S O

NH

HN O

FIGURE 1.14 RAFT agents used for the synthesis of glycopolymers.

the text and their corresponding RAFT polymerization conditions. The RAFT agents and initiators used for glycopolymer synthesis are included in Figures 1.14 and 1.15, respectively. For simplicity, RAFT agents will generally be referred to by their label rather than their chemical name. Since the number of initiators is small, they will be referred to by their commonly accepted acronyms.

O

NH2

CN N

HO NC

OH

N O

4,4'Azobis(cyanopentanoic acid) (ACPA)

NC

N

N

CN

2,2'-Azoisobutyronitrile (AIBN)

HCl·HN

N

N

NH·HCl NH2

2,2'-Azobis(2-amidinopropane)dihydrochloride (AAPD)

FIGURE 1.15 Initiators used for the RAFT polymerization of glycomonomers.

62


1.4.3.1 Homopolymers and Copolymers The RAFT process is not restricted to the generation of homopolymers, although the first literature reports of glycopolymer synthesis naturally focused on single monomer systems. Well-defined random copolymers are also accessible by RAFT provided the comonomer reactivities are well matched to the RAFT agent. The “living” nature of RAFT polymerization also provides the possibility to chain extend homopolymers with another monomer to generate block copolymers. This section will focus on homopolymers and statistical/block copolymers generated without the intent of higher applications such as self-assembly or bioconjugation. The first polymerization of glycomonomers using RAFT was reported by Lowe et al. [102] in which 2-methacryloxyethyl glucoside (2-MAOEGlc; Table 1.5, entry 1) was polymerized in an aqueous system using the now commonly used RAFT agent (4-cyanopentanoic acid)-4-dithiobenzoate C1 and the water-soluble initiator 4,4 -azobis(4-cyanopentanoic acid) (ACPA) (Fig. 1.16). The solubility of the RAFT agent and the initiator were enhanced by choosing slightly basic polymerization conditions, which may have contributed to higher than expected Mn values beyond 40% conversion as a result of RAFT agent hydrolysis. Despite this, however, the low PDIs below 1.07 and the pseudo-first-order kinetics indicate that the polymerization was well controlled. No induction period was observed. Retention of the dithioester end groups was also supported by the chain extension of a poly(MAOEGlc)-C1 macro-RAFT agent (Mn = 14.2 kDa, PDI = 1.07) with 3-sulfopropyl methacrylate (SPMA), in which no low-molecular-weight tailing was observed to indicate a loss of active RAFT functionalities. However, broadening of the molecular weight (MW) distribution occurred, with the PDI rising to 1.63 accompanied by the appearance of high MW termination products in the SEC traces. A self-blocking experiment [chain extension of poly(MAOEGlc)-C1 with MAOEGlc] also resulted in a PDI over 1.5. In contrast to this, the chain extension of a poly(SPMA) macro-RAFT agent with 2-MAOEGlc gave a narrowly dispersed block copolymer with a PDI of 1.18. 1-O-methyl-␣-d-glucopyranose and 1-O-methyl-␣-d-mannopyranose were regioselectively acylated in the 6-position with vinyl methacrylate using an enzymatic approach by Albertin et al. to give the monomers 6-O-methyl-␣-d-glucopyranose (6-O-MAMGlc; Table 1.5, entry 2) and 6-O-methyl-␣-d-mannopyranose (6O-MAMMan; Table 1.5, entry 3) [103]. Stereoselective functionalization of

O O OH HO HO

ACPA, C1 H2 O, 70°C

S

NC HO

CH2

C

O

O

S n O

O

O

OH

OH HO HO

O O OH

FIGURE 1.16 Homopolymerization of 2-methacryloxyethyl glucoside (2-MAOEGlc) by Lowe et al. [102].


63

unprotected sugars has been employed by several researchers in good yields without the need for protection/deprotection steps. 6-O-MAMGlc and the commercially available monomer 2-methacryloxyethyl glucoside (2-MAOEGlc; Table 1.5, entry 1) were polymerized utilizing C1 as RAFT agent and ACPA as initiator, but the addition of 10% ethanol as cosolvent was preferred to aid the dissolution of these two species, rather than risking C1 hydrolysis by adopting the basic conditions chosen by Lowe et al. [102]. Indeed, a direct comparison between aqueous RAFT using carbonate–bicarbonate versus 10% ethanol to dissolve C1 found that the added base resulted in long inhibition periods, significant retardation, bimodal molecular weight distributions, and loss of control at moderate conversions, all of which are attributed to hydrolysis of the RAFT agent. In contrast, control was maintained in the ethanol system up to high conversion [104]. Pseudo-first-order polymerization kinetics for the polymerizations of 6-OMAMGlc and 2-MAOEGlc were observed after a short induction period, indicating that the radical concentration was constant throughout. In both cases, the molecular weight increased linearly with conversion and PDIs remained below 1.12 throughout. Two resulting macro-RAFT agents were chain extended with 2-MAOEGlc and 6-O-MAMMan, respectively, to give the block copolymers poly(6-O-MAMGlc)-bpoly(2-MAOEGlc) and poly(2-MAOEGlc)-b-poly(6-O-MAMMan). Increasing the C1 concentration in the polymerization of MAOEGlc resulted in a marked retardation in the polymerization rate [105], a phenomenon which is common in RAFT polymerization using dithioesters and has been tentatively, although not conclusively, attributed to slow reinitiation in the RAFT pre-equilibrium [106]. A similar effect was observed by the same researchers in the homopolymerization of 6-O-methacryloyl mannose (MAM; Table 1.5, entry 4) [107]. Chain extension of poly(2-MAOEGlc) with 2-hydroxyethyl methacrylate (HEMA) gave well-defined hydrophilic-hydrophilic block copolymers (PDI of 1.2), although the chain extension was not quantitative because the first block was polymerized to high conversion (98%), which encouraged loss of RAFT end groups through termination in the latter stages when the monomer was depleted [105]. A detailed investigation into the kinetics of 6-O-MAMGlc polymerization using nuclear magnetic resonance (NMR) revealed that the initial non-steady-state (induction) period previously observed [103] was inversely proportional to the [C1]0 :[initiator]0 ratio [108]; that is, a higher initiator concentration (keeping the RAFT agent concentration unaltered) reduces the induction period. Interestingly, replacing C1 with a poly(6-O-MAMGlc)-C1 macroRAFT agent did not completely eliminate this induction period. The vinyl adipoyl monomer 6-O-vinyladipoyl-d-glucopyranose (6-O-VAGlc; Table 1.5, entry 5) was synthesized in a similar enzymatic fashion by Albertin et al. [109] and was polymerized at 60◦ C using a dithiocarbamate RAFT agent C2 in water and a xanthate RAFT agent (or more specifically a MADIX agent) C3 in methanol. Vinyl esters are relatively unreactive monomers whose propagating radicals are, therefore, very reactive, and as a result, controlled polymerization of these monomers is difficult. The two aforementioned RAFT agents are suitable for this purpose because both of their Z groups are electron donating and therefore have a destabilizing effect on the intermediate RAFT radical in the pre-equilibrium (species II in Scheme 1.4).

64


The destabilizing nature of these Z groups encourages forward fragmentation of the RAFT-centered radical, a condition that is necessary for controlled polymerization. Excellent control was observed for both homopolymerizations, with the PDI of 1.19 attained using C2, and 1.10 using C3; however, the conversions were much lower than the control experiments without RAFT agent, particularly in the case of C2. No investigation into the kinetics was performed. Lowe and Wang polymerized the protected monomer 3-O-methacryloyl-1,2;3,4di-O-isopropylidine-d-galactopyranose (MAIpGal; Table 1.5, entry 6) using cumyl dithiobenzoate C4 and 1-cyano-1-methylethyl dithiobenzoate C5 as RAFT agents in DMF at 70◦ C [110]. The C4-mediated polymerization exhibited a 50-min induction period before proceeding with first-order kinetics. The induction period is common in C4 systems and is attributed to slow fragmentation of the intermediate RAFT-centered radical. Other explanations have also been proposed, such as an initialization period [111] and the influence of C4 impurities [112], but none have been unequivocally verified. Number-average molecular weights increased linearly with conversion, and although a discrepancy was noted between theoretical and SEC-determined Mn values, the polymerization was well controlled (PDIs below 1.20). A series of poly(MAIpGal) macro-RAFT agents were chain extended with (2-dimethylamino)ethyl methacrylate (DMAEMA) with good control, and deprotection of the isopropylidine groups using aqueous formic acid gave well-defined double-hydrophilic block copolymers. Methacrylamide variants of MAIpGlc with C5 and C10 spacers between the vinyl ¨ urek et al. using C5 and sugar groups (Table 1.5, entry 7) were polymerized by Ozy¨ [113]. Control was reasonable at best and declined beyond 50% conversion. Very broad molecular weight distributions (over 1.5) were obtained when a well-defined poly(NiPAAm) macro-RAFT agent was chain extended with the sugar monomers. Despite this lack of optimization, the thermoresponsive properties of a variety of random and block copolymers were investigated and were found to depend, among other parameters, on the spacer length of the sugar monomers. 3-O-methacryloyl-1,2:5,6-di-O-isopropylidine-d-glucofuranose (MAIpGlc; Table 1.5, entry 8) and the fructose-based monomer 3-O-methacryloyl-1,2;5,6-di-Oisopropylidine-␣-d-fructopyranose (MAIpFrc; Table 1.5, entry 9) were polymerized by Al-Bagoury et al. in a miniemulsion system using the three RAFT agents 1phenylethyl dithiobenzoate C6, C5, and C4 [114]. C6 was ineffective in controlling the polymerizations, giving significant retardation, much higher than expected Mn ’s, and broad PDIs due to its low chain transfer constant to methacrylates. Similar effects were encountered using C5, which were in this case partly attributed to easy escape of the small and highly polar 2-cyanoprop-2-yl free radical into the water continuous phase, which reduced the number of radicals in the system. In contrast, C4 allowed the generation of well-defined homopolymers (PDIs of 1.10–1.25) and the number of particles Np present throughout the polymerization followed a profile typical of a controlled miniemulsion process. Higher C4 concentrations favored better control. Interestingly, the polymerization of MAIpFrc was significantly slower than that of MAIpGlc, which was attributed to greater steric hindrance around the vinyl group in the pyranose ring of MAIpFrc compared to the furanose ring of MAIpGlc.


65

Wang et al. polymerized 6-O-(4-vinylbenzyl)-1,2:3,4-di-O-isopropylidine-dgalactopyranose (VBIG; Table 1.5, entry 10) using benzyl dithiobenzoate C7 in a well-controlled manner to give optically active homopolymers (courtesy of the chiral pendant sugar groups) [115]. The polymers were useful for chiral recognition of several racemic compounds.

1.4.3.2 Stars The scope of the RAFT technique was extended by the development of multifunctional RAFT agents that allow the synthesis of star polymers. Both the R group and the Z group approaches have been applied to star glycopolymer synthesis and are detailed below. Bernard et al. synthesized glycostars by polymerizing 6-O-VAGlc (Table 1.5, entry 5) in N,N-dimethylacetamide (DMAc) at 70◦ C using a tetrafunctional xanthate RAFT agent C8 [116]. The R-group approach was adopted in this case, which means that the star acts as the R group in the RAFT process, and consequently the polymer chains grow from the core outward with the dithioester group residing at the terminus of each arm. The alternative Z-group approach has each RAFT group attached via its Z group to the star. In this case, the radicals propagate as linear chains in solution and must return to the dithioester groups, which remain at the core, in order to undergo reversible chain transfer. Both approaches have their advantages and disadvantages, but in this case the R-group approach gave unexpectedly low conversions, which did not exceed a maximum of 50% despite a continuing radical flux. Linear increase in molecular weight was observed with conversion, but the molecular weights were systematically higher than expected, which was attributed to competing side reactions that were not specifically identified by the authors. N-acryloyl glucosamine (AGA; Table 1.5, entry 11) was polymerized by Bernard et al. using mono- and trifunctional trithiocarbonate RAFT agents C9 and C10 in 5:1 H2 O/EtOH [117] and the polymerizations in this case were well controlled and reached high conversions in less than 7 h. Synthesis of the linear block copolymer poly(AGA)-b-poly(NiPAAm) was successful but was accompanied by the presence of some low-molecular-weight material that suggested incomplete chain extension. The initial attempt at star glycopolymer synthesis was performed using the Z-group approach in polar aprotic solvents to ensure solubility of both monomer and RAFT agent, but despite achieving good solubility, the control was poor. The problem was overcome by first polymerizing 2-hydroxyethyl acrylate (HEA) to give a trifunctional macro-RAFT agent whose short poly(HEA) blocks permitted dissolution in the H2 O/EtOH cosolvent system used for the subsequent sugar polymerization. The molecular weight increased with conversion, but not linearly, and polydispersities were low (1.3–1.6) but not comparable to those of the linear copolymers. Steric congestion around the RAFT groups at higher conversion is blamed for this loss of control. Reasonably well-defined poly(AGA) glycostars were synthesized similarly by Zhang and Stenzel using ␤-cyclodextrin (␤-CD) with seven C9 functionalities, but chain extension using NiPAAm proved unsuccessful [118]. 1.4.3.3 Self-Assembly The ability of the RAFT technique to generate block copolymers presents the opportunity to synthesize copolymers with blocks of

66


disparate character, which can often self-assemble in an appropriate solvent to give higher order structures such as micelles or vesicles. In an aqueous system, the hydrophobic core of the micelles can be designed to encapsulate a therapeutic species such as hydrophobic drug or gene. Glycopolymers are interesting candidates for the hydrophilic shell because their biological activity can facilitate targeted delivery of the drug to particular cells. As an interesting anomaly, the final example in this section demonstrates the self-assembly of a sugar-containing homopolymer rather than a block copolymer. Cameron et al. synthesized block copolymers that self-assembled in aqueous solution into wormlike micelles capable of encapsulating a hydrophobic dye. The hydrophilic block was synthesized from 2-(␤-d-galactosyloxy)ethyl methacrylate (GalEMA; Table 1.5, entry 12) and the hydrophobic block from n-butyl acrylate (BA) with high blocking efficiency and PDIs less than 1.2 [119]. A block copolymer synthesized by Pearson et al. also exhibited a tendency to form rods rather than spherical micelles in solution as the length of the hydrophobic block was increased. Homopolymerization of 2-methacrylamido glucopyranose (MAG; Table 1.5, entry 13) to generate the hydrophilic block was well controlled by C1 [120]. The polymerization of 3-O-methacryloyl-1,2:5,6-di-O-isopropylidine-d-glucofuranose (MAIpGlc; Table 1.5, entry 8) using cumyl phenyl dithioacetate C11 RAFT agent by Ramiah et al. showed an apparent lack of control in the early stages of the polymerization, after which the PDI decreased from 1.5 down to 1.16 at higher conversion [121]. This feature of C1-mediated polymerization of (meth)acrylates is attributed to slow fragmentation of the RAFT-centered radicals in the pre-equilibrium (II, Scheme 1.4). A conventional free radical polymerization mechanism therefore dominates at low conversion until all of the RAFT agent species have undergone fragmentation, after which the polymerization proceeds in a controlled manner and the PDI declines. Chain extension with styrene or methyl acrylate generated amphiphilic block copolymers that formed spherical micelles in aqueous solution after deprotection of the isopropylidine groups. Paspakaris and Alexander polymerized acetyl-protected 2glucosyloxyethyl methacrylate (AcGEMA; Table 1.5, entry 14) using C5 and, after deprotection, chain extended with diethyleneglycol methacrylate (DEGMA) to give a thermosensitive block copolymer [122]. In aqueous solution the polymer chains formed vesicles whose diameters decreased from 500 to 300 nm when the temperature of the system exceeded 37◦ C. Above this temperature, the thermoresponsive poly(DEGMA) block becomes hydrophobic and therefore the PDEGMA core of the vesicles collapses. These thermoresponsive vesicles were used to investigate polyvalent binding events of Con A and mutant Escherichia coli, which expresses specific receptors for glucose and mannose. A block copolymer of vinyl acetate (VAc) and 6-O-methacryloyl mannose (MAM; Table 1.5, entry 4) was synthesized by Ting et al. [123], but a macromolecular click reaction was used rather than chain extension of a macro-RAFT agent because the large difference in monomer reactivities precludes their union in this manner; the success of the RAFT process relies on careful selection of R and Z groups according to the reactivity of the propagating species, and in this case the reactivities of the two monomers are too different to permit a chain extension approach. MAM was

67

CONTROLLED/LIVING RADICAL POLYMERIZATION S

S C6, AIBN

CH2

CH

THF, 60°C CHO OO

Deprotection

O

CH2

S n CHO

HO O O

O HO

O O

CH

CHO OO

O O

S n

O

O

OH OH

FIGURE 1.17 Polymerization of 1,2:3,4-di-O-isopropylidine-6-O-(2 -formyl-4 -vinylphenyl)-d-galactopyranose by Xiao et al. [124].

polymerized using a C1-based RAFT agent containing a silyl-protected alkyne to give a well-defined glycopolymer (Mn = 7.6 kDa, PDI = 1.11), which, after deprotection of the silyl group, was reacted with an azide-functionalized poly(VAc) for 48 h in DMAc using a 1,8-diaza[5,4,0]bicycloundec-7-ene and CuI catalyst system. SEC analysis did indicate that the desired block copolymer was formed, but the PDI was inexplicably broad (1.48) due to the presence of unreacted homopolymers. A new glycomonomer containing an aldehyde functionality was developed by Xiao et al. (Fig. 1.17) [124]. Polymerization of 1,2:3,4-di-O-isopropylidine-6-O(2 -formyl-4 -vinylphenyl)-d-galactopyranose (IVDG; Table 1.5, entry 15) was performed at 60◦ C in tetrahydrofuran (THF) using C6. The Mn increased linearly with conversion and PDIs remained less than 1.10 throughout, highlighting the versatility of the RAFT process in accommodating monomers with various functional groups. Removal of the isoproylidine groups using formic acid gave an amphiphilic homopolymer that self-assembled in aqueous medium to give discrete spherical micelles of uniform size. Aldehyde groups can react with primary amines to form a Schiff base linkage under mild conditions, and were thereby used in this system to conjugate bovine serum albumin (BSA), a model protein, to the micelles.

1.4.3.4 Crosslinked Micelles Researchers from the Centre for Advanced Macromolecular Design (CAMD) have published several reports of crosslinked micelles made from glycopolymer-containing block copolymers. Zhang et al. [125] synthesized poly(AGA)-b-poly(NiPAAm) block copolymers as reported previously [117] and crosslinked the core of the resulting micelles by polymerizing with 3,9-divinyl-2,3,8,10-tetraoxaspiro[5.5]undecane (an acid-degradable crosslinker). Crosslinking in this case involved chain extension of the assembled block copolymers with a hydrophobic difunctional species that was encapsulated in the micelle core (where the RAFT end groups also resided). The process was essentially an emulsion polymerization with radicals provided by a water-soluble initiator. No gelation occurred due to successful mediation of the polymerization by the RAFT agent, giving stable core–shell structures that readily decomposed under acidic conditions. A similar approach using 1,6-hexanediol diacrylate also gave stabilized micelles [118]. A macro-RAFT agent synthesized from a poly(lactide) (PLA) was used to chain extend 6-O-acryloyl-1,2;3,4-di-O-isopropylidine-␣-d-galactopyranose (AIpGal;

68


O O

O

O

O

O C13 SnOct2, 140°C

O

HO O

S

O m

S

AlpGal, AIBN, 70°C α,α,α-trifluorotoluene

S

O

HO O

S

O

S S

m

n

OO

O O

O a) Deprotection b) Self-assembly c) Crosslinking d) Core removal

O

O

Hollow sugar balls

FIGURE 1.18 Synthesis of hollow sugar balls by Ting et al. [126].

Table 1.5, entry 16) by Ting et al. (Fig. 1.18) [126]. Suitable conditions were found by polymerizing AIpGal with C12 in ␣,␣,␣-trifluorotoluene, after DMAc and DMSO proved ineffective. Polymerization using the PLA-C13 macro-RAFT agent proceeded at a higher rate than the preceding RAFT experiment under the same conditions. A higher macro-RAFT concentration reduced the polymerization rate, which is attributed to lower initiator efficiency due to the higher viscosity (rather than slow fragmentation of the intermediate RAFT radicals, since this is not usually observed with trithiocarbonates). After deprotection of the sugar, the amphiphilic block copolymer self-assembled into micelles, which were crosslinked with 1,6-hexandiol diacrylate at the interface between the two blocks (since the RAFT functionality resided there). Removal of the PLA core using hexylamine afforded hollow galactose-bearing nanocages.

1.4.3.5 Surface and Particle Modification Various strategies have been successfully employed to generate surfaces and particles functionalized with welldefined glycopolymers. The majority focus on attachment of premade glycopolymers via various reactive groups installed at the chain end or along the backbone, but the first report mentioned in this section details direct RAFT polymerization from a surface. Stenzel et al. synthesized glycopolymer brushes by attaching a trithiocarbonate RAFT agent to a silica substrate and polymerizing NiPAAm and N-acryloyl glucosamine (AGA; Table 1.5, entry 11) in a sequential fashion (Fig. 1.19) [127]. The trithiocarbonate RAFT agent C9 was affixed using the Z-group approach, which

Si O Si O Si Si O

O N H

S S

O

Si O Si O Si Si O

AGA, ACPA, C9 5:1 H2O:EtOH 60°C

S

N H

S S

S

C

CH2 m

O

NH

OH

O

HO HO OH NiPAAm, ACPA, C9 1:1 H2O:DMSO 60°C

Si O Si O Si Si O

O N H

S S

S

C

CH2

C

CH2

n HN

O

O

m NH

OH

O

HO HO OH

FIGURE 1.19 Grafting of glycopolymer from a silica surface by Stenzel et al. [127].


69

means the trithiocarbonate group remained between the growing polymer chain and the silica substrate. The polymerization was performed by submerging the RAFTfunctionalized silica in a 5:1 H2 O/EtOH solution of AGA, ACPA, and free C9. The use of free RAFT agent in solution suppresses termination since the propagating radicals are not attached to the surface in the Z-group approach and are more likely to undergo combination rather than reversible chain transfer if the only RAFT agent in the system is confined to the surface. Both the molecular weight of the free polymer and the corresponding brush thickness both increased with conversion, suggesting a controlled polymerization process. Surprisingly, chain extension using NiPAAm showed none of the steric hindrance problems that plagued previous chain extensions of star macro-RAFT agents using the Z-group approach. Success in this case is attributed to entrapment of the growing radicals close to the surface, which meant they were always in close proximity to the trithiocarbonate controlling agent (which resides between the surface and the growing chains). A radical approach was used by Guo et al. to attach a polymer synthesized from 2-O-methacryloyloxyethyl-(2,3,4,6-tetra-O-acetyl-␤-d-galactopyranosyl)-(14)-2,3,6-tri-O-acetyl-␤-d-glucopyranoside (MAEL; Table 1.5, entry 17) to a vinylfunctionalized silica surface [128]. Polymerization of the protected monomer was conducted using C4 in chloroform, which is an unusual choice of solvent for controlled polymerization because of its relatively high chain transfer constant. The final PDIs were still low (1.07–1.34), but the measured molecular weights were significantly higher than expected. More judicious solvent choice may have improved the level of control. Glycosylated nanoparticles were synthesized from GalEMA (Table 1.5, entry 12) by Spain et al. in an aqueous system containing 20% ethanol using C1 at 70◦ C [129]. Despite approaching complete conversion after 2 h, the polymerization showed all indications of well-controlled living polymerization. The dithiocarbonate end group of the final polymer (Mn = 24.1 kDa, PDI = 1.09) was reduced to a thiol by NaBH4 in the presence of HAu(III)Cl4 , which was simultaneously reduced to Au(0), giving glycopolymer-stabilized gold nanoparticles (AuNPs). Thiols are well known for their strong binding to gold. The biological activity of the galactose groups on the particles was demonstrated by their ability to agglomerate peanut agglutinincoated beads. As an aside, the same polymer was included in a small library of polymers containing peptide and vinyl-derived backbones that were investigated for their antifreeze abilities [130]. The glycopolymer was found to have a small but significant inhibiting effect on crystallization. Several other reports of glyconanoparticles prepared in a similar manner have emerged since that of Spain et al. [129] was published. Trithiocarbonate-mediated polymerization of d-gluconamidoethyl methacrylate (GAMA; Table 1.5, entry 18) and 2-lactobionamidoethyl methacrylate (LAMA; Table 1.5, entry 19) gave glycopolymers that were reduced in the presence of the thiol (denoted –SH) species biotin–PEG–SH and Au(III) to give AuNPs stabilized with glycopolymer and biotinPEG. The particles were bioconjugated to streptavidin [131]. Random copolymers of 6-O-acrylamido-6-deoxy-1,2:3,4-di-O-isoproylidine-␣-d-galactopyranose (GalAm; Table 1.5, entry 20) and N-acryloylmorpholine (NAM) made using the biotinlabeled RAFT agent C16 also gave glycoparticles that could bind streptavidin [132].

70


Compositional drift during the random copolymerization saw an increasing proportion of the galactose monomer incorporated in the latter stages of the polymerization [133]. AuNPs with random copolymers containing p-acrylamidophenyl ␣-mannoside (Table 1.5, entry 21) or p-acrylamidophenyl N-acetyl-␤-glucosamine (Table 1.5, entry 22) (copolymerized with acrylamide) showed strong, specific molecular recognition of lectins and bacterium [134]. A homopolymer of 2-deoxy-2-N(2 -methacryloyloxyethyl)aminocarbamyl-d-glucose (GUMA; Table 1.5, entry 23) was synthesized by RAFT using C1 after a failed attempt using ATRP due to the presence of the urea group. The resulting RAFT polymer (PDI of 1.17) was reduced and attached to a solid substrate coated in AuNPs to give a surface with excellent resistance to nonspecific protein adsorption [135]. Deng et al. polymerized the two monomers 2-gluconamidoethyl methacrylamide (GAEMA; Table 1.5, entry 24) and 2-lactobionamidoethyl methacrylamide (LAEMA; Table 1.5, entry 25) in 7:1 H2 O/DMF at 70◦ C using C1 and S,S -bis(␣,␣ -dimethyl-␣ -acetic acid)trithiocarbonate C14 RAFT agents. Surprisingly, the polymerization of the more bulky LAEMA using C1 was approximately twice as fast as that of GAEMA. A long inhibition period was observed for both. Much better control was attained using the dithioester C1 rather than the trithiocarbonate C14. Macro-RAFT agents synthesized from primary amine-containing monomers were chain extended with GAEMA, resulting in some broadening in the molecular weight distributions. Biotin was attached to the amine groups, and AuNPs stabilized with the resulting block copolymers specifically recognized both streptavidin and RCA120 lectin [136]. In a separate publication the same group found the polymerization of the slightly bulkier 3gluconamidopropyl methacylamide (GAPMA; Table 1.5, entry 26) to be slower than that of GAEMA. In this report a second primary amine-containing block was used to condense plasmid deoxyribonucleic acid (DNA) [137], and in a further publication the same poly(amine)-b-poly(GAPMA) and the lactose equivalent poly(amine)-bpoly(LAEMA) attached to single-wall carbon nanotubes (SWNTs) displayed high biocompatibility and transfection efficiency as potential gene delivery agents [138]. Jiang et al. attached a RAFT-synthesized random copolymer containing glucose (GAEMA; Table 1.5, entry 24), primary amine, and biotin pendant groups to quantum dots (Fig. 1.20) [139]. Quantum dots are semiconductor nanocrystals whose dimensions impart unique optical properties useful for biomedical imaging applications.

+

+ O O

O

O O

O

ACPA, C14 HO 70°C

HOOC

CH

C co CH

OH HO HO

O NH O

S

C n O

O

S

COOH

S

O

O

OH

OH HO

C co CH

O

NH

NH ·HCl

HO HO

O

OH NH HO

O

NH ·HCl

NH O

H N S

H N

O N H

S

O N H

FIGURE 1.20 Statistical copolymer used by Jiang et al. to functionalize quantum dots [139].

RING-OPENING POLYMERIZATION

71

The three monomers were copolymerized using C14 in an aqueous system to give a well-defined statistical copolymer with Mn = 12.9 kDa and PDI = 1.19. Reaction of the primary amine groups in the polymer with activated ester groups on the quantum dot surface had no effect on the physical properties of the quantum dots but gave water-soluble particles with the ability to recognize streptavidin and with improved biocompatibility compared to the naked particles.

1.5 RING-OPENING POLYMERIZATION Ring-opening polymerization (ROP) of cyclic monomers such as cyclic ethers, acetals, amides (lactams), esters (lactones), and siloxanes results in the formation of similar products as step polymerization products, but usually with better control over molecular weight and smaller molecular weight distributions. Ring sizes commonly used are 3-, 4-, 5-, 7-, and 8-membered rings, while the polymerization of 6-membered rings is thermodynamically unfavorable. Ring-opening polymerization can be initiated with anionic and cationic species including strong protic acids and Lewis acids in conjunction with water for cationic polymerization and alkali metals, metal alkoxides, metal complexes and others for the anionic process (Scheme 1.5). Many ROPs proceed as living polymerizations with the molecular weight increasing with conversion. However, significant chain transfer reaction can be present especially in cationic ROPs. The synthesis of complex architectures such as star and block copolymers is feasible [140]. Two reports on the synthesis of glycopolymers using ROP emerged simultaneously in 1985 when Good and Schuerch [141] and Uryu et al. [142] used the anhydride form of various sugars (Table 1.6, entries 1–3) to yield polymers of up to 11 kDa

Anionic

X R R

X

R

X * n

Cationic

X

R X

X

R R X

X

R

X * n

SCHEME 1.5 Mechanism of anionic and cationic ring-opening polymerization.

72 1

2

3

4

Galactose

Glucose

Glucose

Entry

OBz

OOBz

BzO

O

OR

OR

O OR R= p-Br C6H4CH2R= Bz R= p-CH3C6H4CH2-

OAc

O O

HN O

O

O

O-(tetra-O-acetyl-β-D-glucopyranosyl)L-serine N-carboxyanhydride

AcO AcO

OAc

1,3-anhydro-2,4,6-tri-O-alkylβ-D-glucopyranose

O

1,4-Anhydro-2,3,6-tri-O-benzylβ-D-galactopyranose

BzO

BzO O

1,4-Anhydro-2,3,6-tri-O-benzylα-D-glucopyranose

BzO

O

Monomer

Dioxane CH2 Cl2 Acetonitrile

CH2 Cl2 Toluene Benzene

CH2 Cl2

CH2 Cl2

Solvent

Glycomonomers Synthesized with Ring-Opening Polymerization

Glucose

Carbohydrate

TABLE 1.6

25◦ C

Et3 N n-HexNH2 t-BuNH2 Polyoxaline-NH2 p-Vinylbenzylamine

PF5 SbCl5 BF3 ·O(C2 H5 )2 (CF3 SO2 )2 O (C6 H5 )3 CCl

PF5 SbCl5 BF3 ·O(C2 H5 )2

−78◦ C – + 0◦ C

−78◦ C – + 50◦ C

PF5 SbCl5 BF3 ·O(C2 H5 )2 SnCl4 (CF3 SO2 )2 O (i-Bu)3 Al-H2 O

Initiator

−78◦ C – +30◦ C

Temperature

144–146, 149

141

142

142

Reference

RING-OPENING POLYMERIZATION

(a)

O

73

OH

O OR

OR

O

HO

OH O

OR

n

(b)

O

OAc AcO AcO

HN

O O OAc

O

O R-NH2

O

R

N H

H N

H n

O

O AcO

OAc

AcO AcO

FIGURE 1.21 Ring-opening polymerization of glycomonomer: (a) anhydride [141] and (b) N-carboxyanhydride [145].

(Fig. 1.21). Depending on the initiator, stereoregular glycopolymers were obtained as evidenced by optical rotation and NMR. Glycomonomers based on N-carboxyanhydrides were first reported by Rüde et al. [143] but not until the pioneering work of Okada and co-workers were these molecules employed to design complex glycopolymer architectures (Fig. 1.21) [144]. Welldefined homopolymers were obtained by initiating the polymerization with primary amines (PDI of 1.1) while tertiary amines led to broader molecular weight distributions. The amino end-functionality of the product was then utilized to generate block copolymers in a subsequent step [145]. The initiating species for the polymerization—primary amines—provide an avenue to broaden the array of available structures. Amino-terminated polyoxazolines were used as macroinitiators to initiate the polymerization of O-(tetra-O-acetyl-␤-dglucopyranosyl)-l-serine N-carboxyanhydride (Table 1.6, entry 4), resulting in block copolymers with molecular weights which were in good agreement with the theoretical values albeit with broader molecular weight distributions (PDIs of 1.2–1.6) [146]. A similar block copolymer was created by joining two functional homopolymers, polyoxazoline generated via cationic polymerization and poly(O-(tetra-O-acetyl-␤d-glucopyranosyl)-l-serine N-carboxyanhydride) with an amino end-functionality obtained via living anionic ring-opening polymerization. Polyoxazoline was added to the anionic ring-opening polymerization, terminating the polymerization, thus yielding block copolymers with PDIs less than 1.1 [147]. Two avenues were investigated to prepare macromonomers based on carbohydrates in combination with ring-opening polymerization. Comb polymers with glucose functionalities at the end of each branch were generated from N-acetyl-d-glucosamine. The sugar molecule was converted into oxazoline, which then acted as the initiated species in the ring-opening polymerization of various 2-oxazolines (Fig. 1.22). Termination with acrylate anions and the following radical copolymerization with styrene resulted in comb polymers of molecular weights of up to 45,000 g mol−1

74


OAc

OAc O

AcO AcO

N

N

MeOTf O

O

R

O-

O

R

O

O

O

AcO AcO Me

N

N O

n

O

Me

Me

FIGURE 1.22 Synthesis of macromonomers with glucose endfunctionality [148].

with a polystyrene backbone and poly(oxazoline) braches with molecular weights of 1850 g mol−1 [148]. An elegant and simple method for glycol macromonomers was developed by using p-vinylbenzylamine as the initiator for the ring-opening polymerization of O-(tetra-O-acetyl-␤-d-glucopyranosyl)-l-serine N-carboxyanhydride (Table 1.6, entry 4). Subsequent polymerization of the macromonomer with acrylamide lead to comb polymers with high activity when interacting with wheat germ agglutinin [149].

1.6 IONIC CHAIN POLYMERIZATION 1.6.1 Anionic Chain Polymerization Anionic polymerization is initiated with strong nucleophiles such as alkyl or naphthalenide anions. To date, anionic polymerization is the only polymerization method that is truly living, providing polymers with extremely narrow molecular weight distributions while termination and chain transfer reactions are usually absent. However, this technique requires stringent reaction conditions and the choice of monomers is limited since competing reaction with other functional groups need to be absent to ensure living behavior (Scheme 1.6) [150]. It is therefore not surprising that the reports on glycopolymers using living anionic polymerization are limited. Hirao and co-workers achieved well-defined glyco-homo and block copolymers with polydispersity indices as low as 1.04. Prerequisite, however, was the careful design of the structure of the glycomonomers, which were all based on styrene derivatives. Styrene with acetyl-protected ␣-d-galactopyranose, ␤-d-fructopyranose, and ␤-l-sorbofuranose in the meta position (Table 1.7, entries 1–6) underwent living anionic polymerization affording molecular weights close to predicted values. In contrast, para-substituted products resulted in no polymerization. This was explained by an 1,6-elimination step forming a biradical intermediate [151]. The hypothesis was confirmed by synthesizing a para-substituted glycomonomer with a hydrocarbon spacer between styrene and carbohydrate (Table 1.7, entry 7), which is n M+Alkyl -

Alkyl CH2 CH R

R

R

H+ Alkyl CH2 CH CH2 CH R R n

Alkyl CH2 CH H R n+1

SCHEME 1.6 Mechanism of anionic living polymerization.

75

1

2

Glucose

Entry

O

O O

O

O

O O

O

m-(1,2:5,6-Di-O-cyclohexyli dene-αD-glucofuranose-3-oxy-methyl)styrene

O O

m-(1,2:5,6-Di-O-isopropylidene-αD-glucofuranose-3-oxymethyl)styrene

O O

Monomer

Glycomonomers Synthesized via Living Anionic or Cationic Polymerization

Glucose

Carbohydrate

TABLE 1.7

THF

THF

Solvent

−78◦ C

−78 C

◦

Temperature

s-BuLi

s-BuLi

Initiator

(continued)

151

151

Reference

76 3

4

5

Fructose

Sorbose

Entry

O

O O

O

OO

O

O

O

O O

m-(2,3:4,6-Di-O-isopropylidene-αL-sorbofuranose-1-oxy-methyl)styrene

O

O

O

m-(1,2:4,5-Di-O-isopropylidene-αD-fructopyranose-3-oxymethyl)styrene

O

m-(1,2:3,4-Di-O-isopropylideneD-galactopyranose-6-oxy-methyl)styrene

O

OO

Monomer

THF

THF

THF

Solvent

Glycomonomers Synthesized via Living Anionic or Cationic Polymerization (Continued)

Galactose

Carbohydrate

TABLE 1.7

−78◦ C

s-BuLi

s-BuLi

s-BuLi

−78◦ C

−78◦ C

Initiator

Temperature

151

151

151

Reference

77

6

7

Glucose

Glucose

O

O O

O

O

OO n O

n= 3, 11

m-(1,2:5,6-Di-O-isopropylideneα-D-glucofuranose-3oxypropyl)styrene n=3

O O

m-(1,2:5,6-Di-O-isopropy lideneα-D-glucofuranose-3oxyundecyl)styrene n=11

p-(1,2:5,6-Di-O-isopropylidene-αD-glucofuranose-3-oxymethyl)styrene

O O

THF

THF

−78◦ C

−78◦ C

s-BuLi

s-BuLi potassium naphthalate

(continued)

152

151

78 8

9

Glucosamine

Entry

O O

O O

O

O

O O

O

O O

O

O

N

O O O

O

3,4,6-tri-O-acetyl-2-deoxy-2phthalimide-β-D-glucopyranoside

O

O

O

1-O-(vinyloxy)ethyl-2,3,4,6-tetra-Oacetyl-β-D-glucopyranoside

O

O

Monomer

Toluene

Toluene

Solvent

Glycomonomers Synthesized via Living Anionic or Cationic Polymerization (Continued)

Glucose

Carbohydrate

TABLE 1.7

−15◦ C or 0◦ C

−15 C

◦

Temperature

HCl/ZnI2 or TFA/EtAlCl2 (dioxane)

HCl/ZnI2

Initiator

154, 156

153

Reference

10

11

Glucose

Galactose

79

O O

O

O

O

O O

1,2:3,4-di-O-isopropylidene-6-O(2-vinyloxy ethyl)-D-galactopyranose

O

O

O

3-O-(vinyloxy)ethyl-1,2:5,6-di-Oisopropylidene-D-glucofuranose

O O

O

O

−20◦ C

2. −30◦ C

2. Benzene

Toluene

1. −40◦ C

1. Toluene

CH3 CHI(OEt)/ ZnCl2

1. CH3 CH (OiBu)Cl/ZnI2 2. PS-I/ZnCl2

159

158

156

80


incapable of elimination and therefore anionic polymerization proceeded according to the expected living behavior [152].

1.6.2 Cationic Chain Polymerization Cationic polymerization can be initiated using protic acids, Lewis acids, halonium ions, or oniom salts in combination with photoinitiation. While cationic polymerization is closely related to anionic chain polymerization, chain transfer and termination reactions are frequently observed. Side reactions such as ␤-proton transfer, combination with counterion, and chain transfer to polymer limit the growth of a terminating chain. Therefore, PDIs of up to 2 can be theoretically expected and are observed. However, fast initiation and the absence of chain transfer reactions can lead to narrow molecular weight distributions. Under specific circumstances, living cationic polymerization is operational with the molecular weight increasing with conversion (Scheme 1.7) [150]. Glycopolymers via living cationic polymerization are frequently reported. A HCl/ZnI2 initiating system was employed by Minoda et al. to achieve the living polymerization of 1-O-(vinyloxy)ethyl-2,3,4,6-tetra-O-acetyl-␤-d-glucopyranoside (Table 1.7, entry 8) [153]. The molecular weight increased linearly with conversion, although it leveled off at around 80% monomer conversion. A similar result was observed with a monomer based on glucosamine (Table 1.7, entry 9) [154]. This phenomenon was attributed to poor solubility of the resulting glycopolymers in the solvent. Attempts to improve the livingness by using different solvents at higher temperatures succumbed to broader molecular weight distributions. A solution was found by adopting a different initiation system, trifluoroacetic acid with ethylaluminum dichloride, which led to a living process up to 100% monomer conversion [154]. Subsequent work by Minoda and co-workers demonstrated that the catalyst system can moreover be employed to successfully synthesize block copolymers based on poly(isobutyl vinylether) and poly(3,4,6-tri-O-acetyl-2-deoxy-2-phthalimide-␤-dglucopyranoside) (Table 1.7, entry 9) (Fig. 1.23) [155]. The specific interaction of the glycopolymer obtained after deprotection and acetylation of the amino functionality with wheat germ agglutinin (WGA) was evaluated using fluorescence spectroscopy. The block copolymers showed significantly enhanced recognition abilities compared to the homopolymer, let alone the monovalent acetyl glucosamine. A similar block copolymer, based on poly(isobutyl vinylether), was obtained using a protected glucofuranose pendant group (Table 1.7, entry 10). Well-defined block copolymers with PDIs below 1.1 were created and the number of repeating units varied between 20 and 90 [156]. Upon deprotection, amphiphilic block copolymers

n H+ R

H CH2 CH R

R

H CH2 CH CH2 CH R R n

H2O

SCHEME 1.7 Mechanism of cationic polymerization.

H CH2 CH OH R

n +1

81

IONIC CHAIN POLYMERIZATION

O O

O O

O HO

O

O

O

CF3

O

EtAlCl2 1,4-Dioxane

O O N

CF3 O

O

AlEtCl2

O O

O

O O

O

O

O

R O

N

R= carbohydrate

OO O

O

O

n

m O

O

O m

n-1

δ− O

δ+ O

O

n

O

m

a) H2NNH2*H2O b) Ac2O/MeOH

O O

O O

O

O

HO

O

HOHO

OO

O

O

O

O NH

N

FIGURE 1.23 Synthesis of poly(3,4,6-tri-O-acetyl-2-deoxy-2-phthalimide-␤-d-glucopyranoside)-block-poly(isobutyl vinylether) via living cationic polymerization [155].

were formed while simultaneously converting glycofuranose into its pyranose form. Depending on the composition, these block copolymers were found to form various morphologies in solid state, including spheres, cylinders, and lamellas. The monomer reactivity ratios of the statistical copolymerization of these two monomers, isobutyl vinylether (IBVE) and 3-O-(vinyloxy)ethyl-1,2:5,6-di-O-isopropylidene-dglucofuranose (IpGlcVE; Table 1.7, entry 10), were calculated to be r1 (IBVE) = 1.65 and r2 (IpGlcVE) = 1.15. The solution behavior of the statistical copolymer differed substantially from that of the block copolymer. As evidenced by NMR spin-lattice relaxation times, the block copolymer was in an aggregated state in solution at low temperatures while the statistical copolymer was unrestricted in motion [157]. Combination of living anionic and living cationic polymerization resulted in amphiphilic block copolymers with polystyrene as the hydrophobic block. The anionic styrene polymerization was terminated with 3-chloropropionaldehyde diethyl acetal. Addition of trimethylsilyl iodide led to a macroinitiator for cationic polymerization (Fig. 1.24) [158].

O O OO O n -1

Cl

O O

n

I O

I-SiMe

n

O O

O

O Block copolymer

ZnCl

FIGURE 1.24 Polystyrene-block-poly(glucofuranose vinyl ether) by combination of living anionic and living cationic polymerization by Labeau et al. [158].

82


Glycopolymers carrying galactose moieties (Table 1.7, entry 11) were prepared for the immobilization of DNA probes. The aldehyde functionality of the acylic form of the galactose group underwent Schiff base formation with the amino-terminated oligonucleotide, which was followed by reductive amination. A more detailed investigation was dedicated to the end group analysis of the polymer obtained during the living cationic process. The polymerization was terminated in a highly alkaline solution of aqueous KOH. As a result, polymers with aldehydes as ␻-end groups were obtained [159]. Matrix-Assisted Laser Desorption Ionization-Time of Flight (MALDI-TOF) mass spectroscopy analysis indeed confirmed the presence of aldehyde end-functionalities but also the presence of unsaturated products as a result of ␤-elimination. 1.7 RING-OPENING METATHESIS POLYMERIZATION (ROMP) Transition-metal coordination initiators instigate the reaction of cycloalkenes to form polymers via a ring-opening process. Metal–carbene complexes such as the molybdenum- and tungsten-based Schrock initiators and the ruthenium-based Grubbs initiators allow control over the polymerization. Grubbs initiators are considered advantageous due to their low sensitivity against air, moisture, and many functional groups. Depending on the fine structure and further metal ligands, initiators with good polymerization rates in combination with limited side reactions can be designed. In general, ROMP proceeds in a living fashion with narrow molecular weight distributions and the ability to generate block copolymers (Scheme 1.8) [160]. Kiessling and co-workers pioneered the area of glycopolymers via ROMP. In their initial work, a 7-oxobornene derivative with glucose functionalities (Table 1.8, entry 1) was polymerized in water in the presence of RuCl3 at 55◦ C. The polymer obtained with a molecular weight of 106 g mol−1 showed an efficient inhibition of erythrocyte agglutination by Con A. Inhibition doses required were 2000 times lower compared to the monomeric unit, demonstrating its multivalent effect [161]. This study was extended to compare ␣-C-glucoside, ␣-O-glucoside, ␣-C-mannoside, and ␣-O-mannoside (Table 1.8, entries 1–4). The four monomers were obtained by polymerization in water, all resulting in molecular weights of 106 g mol−1 . The ␣-Cmannoside glycopolymer (Table 1.8, entry 3) was measured to have a most efficient inhibition effect of Con-A-induced hemagglutination, followed by the polymer with ␣-O-mannoside (Table 1.8, entry 4) [162]. Concerns were raised regarding the effect of ruthenium impurities in biological applications of these polymers. The use of recycled catalyst or preformed catalyst (obtained by heating RuCl3 with a small amount of monomer) not only overcame this problem, since these are more easily removed, but also enhanced the initiation rate and polymer yield [163].

RCH M

RCH M

RCH M

RCH

M

RCH=CHCH2CH2CH2CH=M n

SCHEME 1.8 Ring-opening metathesis polymerization (ROMP).

83

1

2

Glucose

Entry

HO HO

HO HO

OH

OH

Monomer

O O O

H O

O

O

O O OH HO

H

O

O

O

HO

O O

H O

OH

H O

O

OH OH

OH

OH OH

OH

Glycomonomers Polymerized via ROMP

Glucose

Carbohydrate

TABLE 1.8

Water

Water

Solvent

55◦ C

55◦ C

RuCl3

RuCl3


(continued)

162

161

Reference

84 3

4

Mannose

Entry

Mannose

Carbohydrate

HO HO

HO HO

OH

OH

Monomer

O OH

H O

O OH

H O

O O

O O

O

O O

O

O OH

H O

OH OH

OH OH

OH

OH

O OH

H O

TABLE 1.8 Glycomonomers Polymerized via ROMP (Continued)

Water

Water

Solvent

55◦ C

55 C

◦

RuCl3

RuCl3


162

162

Reference

85

5

6

7

Mannose

Glucose

Galactose

O

HO

H O

OH

O

O OH

H O

OH OR

O O

O

O O

MO3SO

H O

H O

O

O

HN

O

OH

OH OH

OH

OH OH

R= H or MO3S

O

Water/ ClCH2 CH2 Cl/ DTAB

Water

Water

65◦ C

55◦ C

55◦ C

P(Cy)3 Ph Ru H Cl P(Cy)3 Cl

RuCl3

RuCl3

(continued)

164

163, 164

163

86

Glucose

Galactose

Carbohydrate

10

9

8

Entry

O3SO

O

RO HN

-

O

OH

O

OR'

OR OR

O

OH

O

HO OH

H3C

OH OR6'

MO3SO

OH OR

Monomer

O

OR6 OH

O O

HN

R= H or MO 3S

O

R=R'=H I R=R'=Ac II R=R'=CH2Ph III R=R'=SiEt3 IV R=H; R'=CPh3 V

OH

O

O

HN


R6=SO3-, R6'=H

R6= H, R 6'= SO 3-

O

CH2 Cl2 Benzene and solvent mixtures



Solvent

25–50◦ C

60◦ C

65 C

◦

C

C

C

C

P(Cy)3

Ru

PPh3 P(Cy)3

Ru

PPh3

P(Cy)3 Ph Ru H C P(Cy)3 C

P(Cy)3 Ph Ru H Cl P(Cy)3

Cl


Ph Ph

Ph Ph

166

165

164

Reference

87

11

12

13

Mannose

Galactose

Galactose

O O

O

O

O

O

OO

O

OH

O

O

O

OH

O

O N

H

H

O O

O

O

OH OH

O

O O

O

O

O

O

Toluene

Toluene

MeOH/ H2 O/ (CH2 Cl2 )

Room temperature

Room temperature

Room temperature

O

iPr

O

iPr

O

iPr N Ph Me Me Mo

O

iPr N Ph Me Me Mo

P(Cy)3 Ph Ru H Cl P(Cy)3

Cl

(continued)

168

168

167

88 14

15

Ribonic acid

Entry

Mannose

Carbohydrate

O

O

O

O

O

O

O

O

O

Monomer

O

O

O


Toluene

Toluene

Solvent

Room temperature

Room temperature

O

iPr

O

iPr


O

iPr N Ph Me Me Mo

O

iPr N Ph Me Me Mo

168

168

Reference

89

RING-OPENING METATHESIS POLYMERIZATION (ROMP)

O H O

O

H O

O O H C R

RuCl

H O

O H H OO

H H OO

O O H C R

O

O O R

CH

O H H OO

O O H C R

O O R H C

H O

R=

O OH

OH OH OH

FIGURE 1.25 ROMP of an asymmetric glycomonomer for the synthesis of stereochemically diverse glycopolymers [163].

An interesting feature of ROMP-synthesized polymers for biological applications is the simultaneous existence of cis- and trans-isomers in the backbone, which is supposed to enhance the interaction with lectins. To test the hypothesis that variety in polymer geometries enhances the interaction with lectins, the isomer diversity was further increased by employing asymmetric glycomonomers (Table 1.8, entries 5 and 6). Indeed, stereochemically diverse glycopolymers based on 5 and 6 showed an increased activity in their binding to Con A (Fig. 1.25) [163]. To date, most glycomonomers based on 7-oxobornene have been polymerized in water with RuCl3 as initiator. The molecular weights obtained were typically around 106 g mol−1 . The polymerizations, however, were not living. To address the need to generate glycopolymers of different molecular weights via ROMP the Grubbs ruthenium alkylidene catalyst was employed; however, a challenge was presented by the contrasting solubilities of catalyst and monomer. Finally, emulsion conditions were employed with dodecyltrimethylammonium bromide (DTAB) in a 1,2-dichloroethane/water mixture. These conditions were employed using sulfated glycomonomers with 7-oxobornene (Table 1.8, entry 7) as the polymerizing moiety, but also norbonene (Table 1.8, entry 8), and the resulting polymers were found to be effective P-selectin inhibitors with the disulfated compounds showing higher activity [164]. The inhibition of L-selectin was achieved by the norbonene derivate with a disulfated trisaccharide side group (Table 1.8, entry 9). The polymerization was carried out at a low monomer to Grubbs initiator ratio to obtain oligomeric species. The narrow molecular weight distribution of PDI < 1.2 indicated a well-controlled process [165]. The control of molecular weight in conjunction with narrow molecular weight distributions is an attractive feature of living techniques such as ROMP. A detailed study into the livingness of the ruthenium carbene catalyzed polymerization of glycomonomers was first carried out by Fraser and Grubbs. A set of protected and unprotected norbornene derivatives (Table 1.8, entry 10) was polymerized using an active and a less active ruthenium carbene catalyst (Fig. 1.26) [166]. The less active catalyst A (Fig. 1.26) was only able to initiate the polymerization of the acetylated monomer II, albeit forming insoluble gel under certain circumstances. Better results were obtained using the more reactive catalyst B, but the outcome varied depending on the solvent. While heating was usually not required with catalyst B, lower PDIs (

Engineered Carbohydrate-Based Materials for Biomedical Applications: Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels

Engineered Carbohydrate-Based Materials for Biomedical Applications: Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels

Biomedical Applications of Hydrogels Handbook

Biomedical Applications of Hydrogels Handbook

Surfaces of nanoparticles and porous materials

Engineering Materials for Biomedical Applications

Engineering Materials for Biomedical Applications

Natural-Based Polymers for Biomedical Applications

Dendrimers and other dendritic polymers

Biomedical polymers

Hydrogels for medical and related applications

Hydrogels: Biological Properties and Applications

Plasma Technology for Hyperfunctional Surfaces: Food, Biomedical, and Textile Applications

Plasma Technology for Hyperfunctional Surfaces: Food, Biomedical, and Textile Applications

Dendrimers: Towards Catalytic, Material and Biomedical Uses

Biomedical Hydrogels: Biochemistry, Manufacture and Medical Applications (Woodhead Publishing in Materials)

Polymers at Surfaces and Interfaces

Biomedical materials

Biomedical Materials

Polyphosphazenes for Biomedical Applications

Dendrimers

Analytical Ultracentrifugation of Polymers and Nanoparticles

Dendrimers and Dendrons: Concepts, Syntheses, Applications

Macroporous Polymers: Production Properties and Biotechnological Biomedical Applications

coating materials for electronic applications polymers processes reliability testing

Mechanical Properties of Engineered Materials

Mechanical Properties Of Engineered Materials

Coating Materials for Electronic Applications: Polymers, Processing, Reliability, Testing

Biodegradable polymers for industrial applications

Biomedical Data and Applications

Engineered Carbohydrate-Based Materials for Biomedical Applications: Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels