Chemistry of Organo-hybrids: Synthesis and Characterization of Functional Nano-Objects

By Bernadette Charleux, Christophe Coperet and Emmanuel Lacote

This book provides readers with a one-stop entry into the chemistry of varied hybrids and applications, from a molecular synthetic standpoint

• Describes introduction and effect of organic structures on specific support components (carbon-based materials, proteins, metals, and polymers).
• Chapters cover hot topics including nanodiamonds, nanocrystals, metal-organic frameworks, peptide bioconjugates, and chemoselective protein modification
• Describes analytical techniques, with pros and cons, to validate synthetic strategies
• Edited by internationally-recognized chemists from different backgrounds (synthetic polymer chemistry, inorganic surfaces and particles, and synthetic organic chemistry) to pull together diverse perspectives and approaches

• Language: English
• Publisher: Wiley
• Release date: Jan 23, 2015
• ISBN: 9781118870006

Book preview

PREFACE

Science has moved and will increasingly move toward the molecular echelon. From biology and medicine to materials—and even the very French specialty, gastronomy!—many disciplines are claiming the molecular label. It features an understanding of events on a molecular level and the ability to control them through design and synthesis. The latter are thus keys to the future of many fields.

As practitioners of the field, we believe that tremendous opportunities are being created through the ability to design and synthesize novel functional tools—catalysts, devices from probes to batteries, photovoltaics, diagnostic and therapeutic agents, etc.—involving hybrid materials, made of ever more complex organic, organometallic and inorganic molecules linked to active nano-scaled supports (nanoparticles, clusters, polymers and biopolymers, etc.). In other words, molecular architectures certainly stand at the crossroad of most of the major present fields of research, provided their synthetic requirements are adapted to those of objects outside their traditional boundaries.

We define an organic hybrid as a material made by linking polymeric, carbon-rich or inorganic material to organics (small molecules or macromolecules). Our aim is to provide readers—organic synthetic chemists interested in applying their skills to function-oriented synthesis, non-organic chemists wishing to introduce molecular complexity to their field, students trying to make sense of objects that span over several fields of the curriculum—with a general overview of the diversity of solutions that organic, inorganic, and polymer synthetic chemists, chemical biologists, and materials scientists have come up with to control functions through the covalent attachment of specific supports to endlessly variable molecular architectures.

Each new material has specific requirements (availability, stability, surface functionality, etc.) that limit the options available for chemists to graft the desired function-laden molecules. It is also a challenge to determine whether a determined bonding between the entities has been achieved. This in turns generates new questions and opportunities for research and applications. Thus, this book emphasizes two main topics: synthesis and characterization.

The science of hybrids is growing at a galloping speed, so it is impossible, frustrating and in the end futile to pretend to be exhaustive. Rather, we have selected a few key items across the hybrids' family to illustrate the concept and approaches that can be replicated for other classes of supports. This is also why we have minimized the illustration of the properties of the materials as this is depicted in scores of reviews, generally at the expense of the presentation of the nuts and bolts of the synthesis, which the present book seeks to redress. We however explain what property(ies) the molecular template introduces or modifies as this drives the design of the functionalization itself and the methods used to achieve the goal.

In a nutshell, the authors have answered in the most concise way possible the following questions: Why and how to do it? And how to prove that you did it? That is, after a brief introduction on why go for a strategy involving hybrids—how this affects properties, or generates new ones—two issues are addressed, namely (i) what are the synthetic strategies and reactions required and (ii) how can one tell the reactions worked? Are there specific analytical techniques to support the claims? It is very interesting and enriching to see how each of them has interpreted these very simple constraints!

We have subdivided the book into 13 chapters, by classes of supports grouped in small clusters. The first cluster of chapters focuses on carbon-based materials, illustrating three different and complementary angles. Cécilia Ménard-Moyon has illustrated the functionalization of carbon nanotubes; Iban Azcarate, David Lachkar, Emmanuel Lacôte, Jennifer Lesage de la Haye, and Anne-Laure Vallet focused on the chemistry of graphenes; Maria Gunawan, Didier Poinsot, Bruno Domenichini, Peter Schreiner, Andrey Fokin, and Jean-Cyrille Hierso on the contrary focused on the chemistry of the very compact nanodiamonds.

After this carbon-rich foray, the second cluster deals with inorganic materials. In the first two chapters of this part, the functionalizations of titania (Laurence Rozes, Loïc D'Arras, Chloé Hoffman, François Potier, Niki Halttunen, and Lionel Nicole) and zirconia (Marc Petit and Julien Monot) are a perfect illustration of the different options a single change in chemical composition of the support can bring. The two other chapters are devoted to large surface area materials with two different perspectives: Flavien Morel, Xiaoying Xu, Marco Ranocchiari, and Jeroen van Bokhoven examine the highly porous MOFs, while Richard Brutchey, Zeger Hens, and Maksym Kovalenko discuss semiconducting nanoparticles.

The third cluster is devoted to three classes of biopolymers, with a representative example of each main natural building block. Michel Arthur and Mélanie Etheve-Quelquejeu show the challenges of nucleic acids modification; Divya Agrawal and Christian Hackenberger those of modified protein synthesis; and Maxime Guitet, Mickaël Ménand, and Matthieu Sollogoub those of the selective transformation of cyclodextrins.

The final cluster examines polymers: how one can selectively functionalize artificial polymers (Anja Goldmann, Mathias Glassner, Andrew Inglis, and Christopher Barner-Kowollik), as well as how one can couple artificial polymers with biopolymers (Paul Wilson, Julien Nicolas, and David Haddleton). Finally, Anne-Marie Caminade, Béatrice Delavaux-Nicot, and Jean-Pierre Majoral present the specific reactivity challenges of dendrimers, which are from two worlds, macromolecules and molecules… hybrids within hybrids.

We thank all the authors for their contribution and we hope the reader will enjoy the various and varied contributions as much as we did.

Bernadette Charleux

Christophe Copéret

Emmanuel Lacôte

CONTRIBUTORS

Divya Agrawal, Leibniz-Institut für Molekulare Pharmakologie, Berlin, Germany

Loïc D'Arras, Chimie de la Matière Condensée de Paris, UPMC Univ Paris 06-Collège de France, Paris, France

Michel Arthur, Laboratoire de Recherche Moléculaire sur les Antibiotiques, Université Pierre et Marie Curie, Paris, France

Iban Azcarate, Institut Parisien de Chimie Moléculaire, Université Pierre et Marie Curie, Paris, France

Christopher Barner-Kowollik, Preparative Macromolecular Chemistry, Institute for Chemical Technology and Polymer Chemistry, Karlsruhe Institute of Technology, Karlsruhe, Germany

Richard Brutchey, Department of Chemistry, University of Southern California, Los Angeles, CA, USA

Anne-Marie Caminade, Laboratoire de Chimie de Coordination du CNRS, Toulouse Cedex 4, France

Bernadette Charleux, Laboratoire de Chimie, Catalyse, Polymères et Procédés, Université Claude Bernard Lyon 1, Villeurbanne Cedex, France

Christophe Copéret, Department of Chemistry, ETH Zürich, Zürich, Switzerland

Jennifer Lesage de la Haye, Institut Parisien de Chimie Moléculaire, Université Pierre et Marie Curie, Paris, France

Béatrice Delavaux-Nicot, Laboratoire de Chimie de Coordination du CNRS, Toulouse Cedex 4, France

Bruno Domenichini, Laboratoire Interdisciplinaire Carnot de Bourgogne, Université de Bourgogne, Dijon Cedex, France

Mélanie Etheve-Quelquejeu, Laboratoire de Chimie et de Biochimie Pharmacologiques et Toxicologiques, Université Paris Descartes, Paris, France

Andrey A. Fokin, Institut für Organische Chemie, Justus-Liebig-Universität, Giessen, Germany; Department of Organic Chemistry, Kiev Polytechnic Institute, Kiev, Ukraine

Mathias Glassner, Preparative Macromolecular Chemistry, Institute for Chemical Technology and Polymer Chemistry, Karlsruhe Institute of Technology, Karlsruhe, Germany

Anja S. Goldmann, Preparative Macromolecular Chemistry, Institute for Chemical Technology and Polymer Chemistry, Karlsruhe Institute of Technology, Karlsruhe, Germany

Maxime Guitet, Institut Parisien de Chimie Moléculaire, Sorbonne Universités, UPMC Univ Paris 06, Paris, France

Maria A. Gunawan, Institut de Chimie Moléculaire de l'Université de Bourgogne, Dijon Cedex, France; Institut für Organische Chemie, Justus-Liebig-Universität, Giessen, Germany

Christian P. R. Hackenberger, Leibniz-Institut für Molekulare Pharmakologie, Berlin, Germany; Department Chemie, Humboldt-Universität zu Berlin, Berlin, Germany

David M. Haddleton, Department of Chemistry, University of Warwick, Coventry, UK

Niki Halttunen, Chimie de la Matière Condensée de Paris, UPMC Univ Paris 06-Collège de France, Paris, France

Zeger Hens, Department of Inorganic and Physical Chemistry, Ghent University, Ghent, Belgium

Jean-Cyrille Hierso, Institut de Chimie Moléculaire de l'Université de Bourgogne, Dijon Cedex, France and Institut Universitaire de France (IUF)

Chloé Hoffman, Chimie de la Matière Condensée de Paris, UPMC Univ Paris 06-Collège de France, Paris, France

Andrew J. Inglis, Preparative Macromolecular Chemistry, Institute for Chemical Technology and Polymer Chemistry, Karlsruhe Institute of Technology, Karlsruhe, Germany

Maksym V. Kovalenko, Department of Chemistry and Applied Biosciences, ETH Zürich, Zürich, Switzerland; EMPA-Swiss Federal Laboratories for Materials Science and Technology, Dübendorf, Switzerland

David Lachkar, CNRS - Institut de Chimie des Substances Naturelles, Gif s/Yvette, France

Emmanuel Lacôte, Laboratoire C2P2, CNRS - CPE Lyon - Université de Lyon, Villeurbanne Cedex, France

Jean-Pierre Majoral, Laboratoire de Chimie de Coordination du CNRS, Toulouse Cedex 4, France

Cécilia Ménard-Moyon, Laboratoire d'Immunopathologie et Chimie Thérapeutique, CNRS - Institut de Biologie Moléculaire et Cellulaire, Strasbourg, France

Mickaël Ménand, Institut Parisien de Chimie Moléculaire, Sorbonne Universités, UPMC Univ Paris 06, Paris, France

Julien Monot, Laboratoire Hétérochimie Fondamentale et Appliquée, Université Paul Sabatier, Toulouse Cedex 9, France

Flavien L. Morel, Laboratory of Catalysis and Sustainable Chemistry, Paul Scherrer Institute, Villigen PSI, Switzerland; Institute for Chemical and Bioengineering, ETH Zürich, Zürich, Switzerland

Julien Nicolas, Institut Galien Paris-Sud, UMR CNRS 8612, Université Paris-Sud, Faculté de Pharmacie, Châtenay-Malabry, France

Lionel Nicole, Chimie de la Matière Condensée de Paris, UPMC Univ Paris 06-Collège de France, Paris, France

Marc Petit, Institut Parisien de Chimie Moléculaire, Université Pierre et Marie Curie, Paris, France

Didier Poinsot, Institut de Chimie Moléculaire de l'Université de Bourgogne, Dijon Cedex, France

François Potier, Chimie de la Matière Condensée de Paris, UPMC Univ Paris 06-Collège de France, Paris, France

Marco Ranocchiari, Laboratory of Catalysis and Sustainable Chemistry, Paul Scherrer Institute, Villigen PSI, Switzerland

Laurence Rozes, Chimie de la Matière Condensée de Paris, UPMC Univ Paris 06-Collège de France, Paris, France

Peter R. Schreiner, Institut für Organische Chemie, Justus-Liebig-Universität, Giessen, Germany

Matthieu Sollogoub, Institut Parisien de Chimie Moléculaire, Sorbonne Universités, UPMC Univ Paris 06, Paris, France

Anne-Laure Vallet, CNRS - Institut de Chimie des Substances Naturelles, Gif s/Yvette, France

Jeroen A. van Bokhoven, Laboratory of Catalysis and Sustainable Chemistry, Paul Scherrer Institute, Villigen PSI, Switzerland; Institute for Chemical and Bioengineering, ETH Zürich, Zürich, Switzerland

Paul Wilson, Department of Chemistry, University of Warwick, Coventry, UK

Xiaoying Xu, Laboratory of Catalysis and Sustainable Chemistry, Paul Scherrer Institute, Villigen PSI, Switzerland; Institute for Chemical and Bioengineering, ETH Zürich, Zürich, Switzerland

1

COVALENT ORGANIC FUNCTIONALIZATION AND CHARACTERIZATION OF CARBON NANOTUBES

Cécilia Ménard-Moyon

Laboratoire d'Immunopathologie et Chimie Thérapeutique, CNRS – Institut de Biologie Moléculaire et Cellulaire, Strasbourg, France

1.1 INTRODUCTION

More than 20 years ago, Iijima reported the structural morphology of carbon nanotubes (CNTs) by use of high-resolution transmission electron microscopy (HRTEM) and electron diffraction [1]. A CNT can be defined as a graphene sheet rolled up to form a cylinder (Fig. 1.1a). CNTs can be classified into different types: single-wall CNTs (SWCNTs), double-wall CNTs (DWCNTs), and multi-wall CNTs (MWCNTs) depending on the number of layers. SWCNTs have diameters ranging from 0.7 to 2 nm and lengths up to several micrometers, while MWCNTs have diameters from a few to tens of nanometers and lengths up to a few micrometers. Hence, the structure of CNTs is characterized by a high aspect ratio (i.e., ratio between length and diameter). Approximately two-thirds of as-produced SWCNTs are semiconducting, whereas one-third is metallic. CNTs contain defects in their structure, such as vacancies, and five- or seven-membered rings that induce curvature, as illustrated in the transmission electron microscopy (TEM) image in Figure 1.1b.

Figure 1.1 Schematic representation of a SWCNT (a) and TEM image of MWCNTs (b).

The breadth and range of research involving CNTs has expanded greatly over the past years. Indeed, CNTs possess unique electronic, mechanical, and thermal properties that can be exploited for potential applications in a variety of fields from materials science [2], molecular electronics [3], photovoltaic devices [4] to nanomedicine [5]. However, CNTs have poor solubility in all solvents due to strong intermolecular cohesive forces among the nanotubes that form bundles, thus hampering full exploitation of their properties and presenting obstacles to their practical applications. Therefore, functionalization is required for manipulating and processing CNTs by inducing exfoliation, increasing dispersibility, and giving the possibility to associate molecules with specific properties to nanotubes.

Functionalization can be classified into two categories: covalent and noncovalent derivatization, the latter relying on hydrophobic, π–π, and/or electrostatic interactions [6–8]. Covalent functionalization can be achieved by oxidation of defect sites of CNTs and subsequent derivatization of the generated carboxylic acid groups. Other methods are based on halogenation, cycloaddition reactions, or direct additions of highly reactive species on the nanotube sidewall. Grafting functional groups on the nanotube surface in a covalent manner allows to obtain stable conjugates with desired properties by tailoring the physicochemical properties of the CNTs. Depending on the level of functionalization, the electrical conductivity of the CNTs can be significantly altered. It is of high importance to rigorously characterize functionalized CNTs. For this purpose, different spectroscopic, microscopic, and thermal techniques can be used for morphological, structural, and elemental analysis of functionalized CNTs.

This chapter is focused on covalent methodologies for nanotube functionalization. Section 1.2 is dedicated to the different chemical strategies used for covalent functionalization of CNTs, while Section 1.3 describes the analytical techniques for characterization of functionalized CNTs. Finally, Section 1.4 contains some concluding remarks.

1.2 COVALENT FUNCTIONALIZATION OF CARBON NANOTUBES WITH ORGANIC MOLECULES

1.2.1 Defect-Site Chemistry

Among various surface functionalization techniques, amidation or esterification of oxidized CNTs is probably the most extensively used to prepare soluble materials either in organic solvents or in water and for linking a wide range of molecules [9]. Generally, oxidation of CNTs is performed by treatment with strong acids such as nitric acid [10,11], sulfuric/nitric acid mixture [10], or with other strong oxidizing agents (H2SO4/KMnO4 [12] or OsO4 [13]). The oxidative treatment, in particular when assisted by sonication, usually induces shortening of the CNTs [10], but also frequently causes nanotube damage, limiting their use as mechanical and electrical reinforcements. Among treatments using strong acids, low-power sonication of MWCNTs in nitric acid followed by treatment with hydrogen peroxide was found to minimize nanotube damage [14]. Many research groups have studied the chemical nature of the oxygenated moieties (e.g., carboxylic acids, carbonyls, hydroxyls) [15] introduced on the nanotube surface by different techniques such as infrared (IR) spectroscopy [16] and thermogravimetry [15]. Oxidized CNTs are mainly decorated with carboxylic groups, as suggested by the pioneering work of the group of Smalley who derivatized the carboxyl functions with thiolalkylamines by amidation [10]. The CNTs bearing thiol moieties were labeled with gold nanoparticles and visualized by atomic force microscopy (AFM). Gold nanoparticles were found mainly at the nanotube ends. By using scanning tunneling microscopy (STM), Prato and coworkers visualized alkyl chains introduced by amidation of carboxylic acid functions, confirming that oxidation of CNTs occurs mainly at the nanotube tips [17].

Oxidized CNTs have been widely used as precursors for further covalent derivatization via amidation or esterification reactions, with amine or alcohol derivatives, respectively (Fig. 1.2). The carboxylic acid functions have to be pre-activated via the formation of acyl chlorides using oxalyl or thionyl chloride, followed by the addition of the appropriate amine or alcohol. Alternatively, the amidation can be performed by using the carbodiimide coupling chemistry. In this case, the carboxyl groups are treated with N-hydroxy succinimide (NHS) or 1-hydroxybenzotriazole (HOBt) in the presence of a carbodiimide, usually N,N-dicyclohexylcarbodiimide (DCC) or 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC). The corresponding esters are then displaced by amine or hydroxyl functions to form the amide or ester bonds, respectively.

Figure 1.2 Oxidation of CNTs (a) and, amidation (b) and esterification of oxidized CNTs (c). For clarity, a SWCNT segment is shown with only a single added functional group.

A range of molecules with various properties have been linked on the CNTs by this method, including several organic molecules [18,19], chromophores with optoelectronic properties [20], bioactive molecules [21–23], or polymers [24,25]. To provide evidence of the ester bond formation in soluble CNTs functionalized with lipophilic or hydrophilic chains, defunctionalization was performed by acid- or base-catalyzed hydrolysis, leading to recovery of the starting insoluble CNTs [26]. In another study, esterification between oxygenated functions at the tips of single oxidized SWCNTs has been exploited to form rings of nanotubes with a narrow size distribution according to AFM [27].

1.2.2 Halogenation

1.2.2.1 Fluorination.

Fluorination has been one of the first chemical methods developed to functionalize CNTs [28]. Most strategies involve elemental fluorine at high temperatures (up to 600°C) (Fig. 1.3a) [29–33]. The best temperature conditions are between 150°C and 400°C. The highest degree of functionalization was found to be one fluorine atom for every two carbon atoms according to elemental analysis [34].

Figure 1.3 Fluorination of CNTs (a). Substitution with Grignard reagents or alkyllithium derivatives (b). Substitution with amino compounds (c) or diols (d).

Alternative conditions implying CF4 plasma treatment have also been developed [35,36]. Due to rehybridation of a high number of sp² carbon atoms to sp³, the resulting fluoronanotubes are insulating.

Fluorination drastically enhances the reactivity of the nanotube sidewalls. Therefore, derivatization of fluoronanotubes by nucleophilic substitution reactions is possible (Figs. 1.3b–1.3d) [37]. Indeed, a variety of nucleophilic reagents has been used such as alkyl magnesium bromides (Grignard reagents) [38] and alkyllithium derivatives [39]. Fluoronanotubes have also been reacted with several amines [40], diamines [41], diols [42], or amino alcohols [42].

1.2.2.2 Bromination.

A few methods for bromination of CNTs have been recently reported using various conditions. DWCNTs have been brominated by elemental bromine using microwaves, leading to a mild alteration of the π-conjugated sidewall of the nanotubes according to Raman spectroscopy (5–8 wt% of Br) (Fig. 1.4a) [43]. Alternatively, bromination of DWCNTs using Br2 vapor at room temperature results in 5–6 at% bromine concentration [44]. Plasma using gaseous bromine has been applied for the functionalization of SWCNTs. The treatment is very efficient as one bromine atom per two carbon atoms is introduced on the nanotube surface in these conditions [45]. The bromo-functionalized nanotubes have been further derivatized by nucleophilic substitution with amine derivatives. Elemental bromine in the presence of a Lewis acid or dibenzoylperoxide as radical initiator allows to functionalize MWCNTs with 10–22 at% Br [46]. As alternative, SWCNTs have been brominated via a mild reaction using N-bromosuccinimide (NBS) (Fig. 1.4b) [47]. NBS readily decomposes into bromine and succinimide radicals with the help of light, heat, and ultrasound. The bromine radicals have higher reactivity toward metallic SWCNTs over semiconducting SWCNTs, allowing the separation of both types of SWCNTs by density gradient ultracentrifugation.

Figure 1.4 Bromination of CNTs with elemental bromine under various conditions (a). Bromination of CNTs with NBS (b). Iodination of CNTs by Hunsdiecker reaction (c).

1.2.2.3 Iodination.

Iodination of SWCNTs has been performed by oxidation of the nanotubes followed by a modified Hunsdiecker reaction in the presence of iodosobenzene diacetate (IBDA) and elemental iodine under UV irradiation via extrusion of CO2 (Fig. 1.4c) [48]. The attachment of iodine atoms on the nanotube sidewall offers the opportunity to apply metal-mediated C–C coupling reactions with aryl halides.

1.2.3 Arylation

Functionalization of CNTs with aryl diazonium salts is one of the most investigated addition reactions on the nanotube surface. The method, initially reported by the group of Tour, has been first performed via electrochemical reduction, resulting in the direct addition of aryl derivatives on the sidewall of SWCNTs (Fig. 1.5a) [49,50]. Later, the same group reported that arylation of SWCNTs could also be achieved by reaction with the aniline precursor, which was transformed in situ to the corresponding diazonium salt by action of isoamyl nitrite (Fig. 1.5b) [51]. The use of anilines has the advantage to avoid isolation of unstable and explosive diazonium salts. It has to be noted that alkyl diazonium salts cannot be used for functionalization of CNTs as alkyl diazonium salts are too unstable. The reaction between aryl diazonium salts and CNTs probably proceeds via highly reactive aryl radical intermediates, generated by electron donation from SWCNTs to the diazonium salts and the subsequent release of N2 gas [52]. Several aniline precursors have been tested, including anilines substituted with halogen atoms (e.g., F, Cl, I), as well as carboxylic acid, amine, or nitro groups [51, 53]. The use of anilines bearing an alkyl chain drastically enhances the solubility of SWCNTs [54]. Besides, iodophenyl-functionalized SWCNTs can be further derivatized by cross-coupling reactions, such as Suzuki [55] or Heck [56] coupling, with various aromatic molecules, like porphyrins, thiophene, and fluorene, or acrylates bearing photosensitive moieties, respectively. The hybrid CNT conjugates could find potential applications as photo- and electroactive materials.

Figure 1.5 Arylation of CNTs with diazonium salts (a), which can be generated in situ from anilines (b) or triazene derivatives (c).

Many experimental variations of the process have been proposed by Tour and coworkers. The reaction is possible in solvent-free conditions, opening possibilities for large-scale functionalization of CNTs [57]. The procedure has also been optimized to perform the reaction in water in the presence of sodium dodecyl sulfate (SDS) as surfactant, leading to unbundled and highly functionalized SWCNTs [58]. Alternatively, to avoid surfactant wrapping and sonication, arylation can be carried out in superacid solvent (oleum: H2SO4, 20% free SO3) with a radical initiator (azobisisobutyronitrile [AIBN] or di-tert-butylperoxide), affording aryl sulfonic acid-functionalized SWCNTs [59,60]. The group of Tour reported an environmentally friendly alternative method based on the use of ionic liquids instead of organic solvents [61]. In this case, grinding SWCNTs with aryl diazonium salts in the presence of imidazolium-based ionic liquids and potassium carbonate leads to rapid and efficient arylation of SWCNTs. Another green strategy has been developed based on the efficient reaction of SWCNTs on water in the presence of a substituted aniline and isoamyl nitrite under vigorous stirring [62]. Tour and coworkers also used triazene derivatives as stable diazonium precursors (Fig. 1.5c) [63]. The reaction can be performed at room temperature in water in the presence of SDS. This approach can be very useful when using molecules with functionalities that do not tolerate diazotization conditions. In addition to SWCNTs, functionalization using aryl diazonium salts has also been applied to MWCNTs [64] and DWCNTs [65]. By using a mixture of three different anilines, it is possible to prepare tri-functionalized CNTs in one step by arylation via in situ generation of the corresponding diazonium salts in the presence of isoamyl nitrite [66]. Using this approach, CNTs have been functionalized with benzylamine moieties blocked with three different protecting groups. The selective removal of these groups under specific conditions allows to control further derivatization of the nanotubes. Multifunctionalization of CNTs is of particular interest to impart multimodalities to the nanotubes in order to open novel opportunities for some applications, in particular in nanomedicine [67].

Smalley and coworkers demonstrated that arylation is selective to metallic SWCNTs under controlled conditions [68,69]. Several other studies have been conducted to explore both the mechanism and the selectivity of the arylation reaction, mainly by the groups of Strano and Tour [70–73]. Strano and coworkers also investigated the dependence of the arylation reaction on the nanotube diameter and the functional groups available on the aryl ring [74]. The selectivity toward metallic SWCNTs has been exploited to prepare field-effect transistors (FET) by reacting all the metallic SWCNTs in the device with diazonium salts in a controlled manner, keeping the semiconducting SWCNTs intact [75]. Separation of semiconducting and metallic SWCNTs can be possible after arylation with diazonium salts by simple filtration of the functionalized SWCNTs through silica gel [76]. Enriched fractions of metallic and semiconducting SWCNTs have been obtained by functionalization of SWCNTs with p-hydroxybenzene diazonium salts, followed by deprotonation in alkaline solution and free solution electrophoresis [77]. The nonfunctionalized semiconducting SWCNTs and the negatively charged metallic ones are found in the nonmobile and negative electrophoretic mobility fractions, respectively.

Arylation of CNTs allows to increase solubility and to further derivatize the functional groups on the aniline rings for specific applications. Direct functionalization of CNTs with diazonium precursors bearing molecules of interest is also possible. For instance, addition of in situ-generated porphyrin diazonium salts to SWCNTs imparts superior optical limiting properties to the nanotubes [78]. Arylation can also find interesting applications in nanoelectronics as it allows self-assembly of functionalized SWCNTs on silicon [79] and on other surfaces [80], which can be useful for the preparation of high-performance FET. Arylation of SWCNTs by using methylenedianiline as precursor gives the possibility to cross-link the nanotubes and to provide three-dimensional networks in the form of microfibers that can be used as scaffolds for hydrogen storage [81].

1.2.4 Cycloaddition Reactions

Functionalization of CNTs using cycloaddition reactions is a powerful strategy that offers advantages over other functionalization methods, such as acid treatment, where the nanotube structure is damaged and the oxygen-containing moieties are mainly localized at the tips and defect sites of the nanotubes. Cycloaddition reactions allow to enhance dispersibility of CNTs and introduce novel functionalities without significant disruption of the π-conjugated network of the nanotube surface, therefore preserving their electronic properties. Various cycloaddition reactions involving strained olefinic bonds of CNTs have been reported, such as [3+2], [2+1], [4+2], and [2+2] cycloadditions [82].

1.2.4.1 [3+2] Cycloadditions.

CNTs have been functionalized by [3+2] cycloaddition with different 1,3-dipoles, including azomethine ylides for introducing pyrrolidine rings on the nanotube surface and by Cu(I)-catalyzed Huisgen click chemistry, leading to the formation of 1,2,3-triazoles.

1.2.4.1.1   1,3-Dipolar Cycloaddition of Azomethine Ylides.

Functionalization of CNTs by 1,3-dipolar cycloaddition of azomethine ylides has been extensively used to increase dispersibility of the nanotubes in organic solvents and in aqueous media [83–85]. The highly reactive 1,3-dipole is generated in situ by condensation of an α-amino acid and an aldehyde derivative, leading to the formation of an iminium salt, followed by thermally induced decarboxylation (Fig. 1.6a). The addition of azomethine ylides on the nanotube surface introduces pyrrolidine rings. The use of paraformaldehyde and an N-functionalized glycine, bearing a triethylene glycol chain and a Boc (tert-butyloxycarbonyl)-protected amino end group, has allowed to introduce ammonium functionalities on the nanotubes and to increase drastically dispersibility in water [86]. The level of functionalization can be determined by using the colorimetric quantitative Kaiser test that gives the amount of primary amine functions. Further derivatization of the ammonium-terminated triethylene glycol-modified CNTs has been exploited, in particular by the groups of Bianco and Prato [87], for coupling of biologically active molecules such as peptides [88,89], antibodies [90,91], amphotericin B, which is an antifungal molecule [92,93], and the anticancer drug methotrexate [94]. The ammonium-functionalized CNTs have also been derivatized with diethylene triamine pentaacetic acid (DTPA), which is a chelating agent of radionuclides (e.g., ¹¹¹In, ⁹⁹mTc) for imaging purposes like the assessment of in vivo biodistribution of functionalized CNTs [95].

Figure 1.6 Functionalization of CNTs by 1,3-dipolar cycloaddition of azomethine ylides generated in situ by condensation of an α-amino acid and an aldehyde derivative (a), or generated from aziridines under microwave irradiation (b), or by double deprotonation of trialkyl-N-oxides (c). Derivatization of CNTs by click chemistry starting from CNTs functionalized with alkyne (d) or azide groups (e), with azido- or acetylene compounds, respectively.

Different functionalities can be directly introduced on the pyrrolidine rings by using a variety of α-amino acids and aldehyde derivatives [96,97]. The use of commercially available N-methylglycine and 3,4-dihydroxybenzaldehyde leads to hydroxyl-functionalized CNTs, which have been further derivatized by silylation and esterification [98]. Polyamidoamine dendrimers have been constructed from ammonium-modified SWCNTs, prepared by 1,3-dipolar cycloaddition of azomethine ylides, by repeating sequentially a two-step reaction using methyl acrylate and ethylenediamine [99]. Tetraphenylporphyrin moieties have been linked to the peripheries of these dendrimers, allowing to increase the number of porphyrins on the nanotubes without causing significant damage to their electronic properties [99]. Some porphyrin moieties do not interact with the SWCNTs, thus exhibiting a fluorescence lifetime similar to the free porphyrin, which could play the role of an antenna with photoinduced electron transfer. Other chromophores with interesting photophysical properties, phthalocyanines, have been covalently grafted by 1,3-dipolar cycloaddition of azomethine ylides [100]. The hybrid conjugate is promising as chemical and light-driven system for the development of light harvesting architectures since SWCNTs serve as electron acceptor and phthalocyanines as electron donor. Functionalization of CNTs by 1,3-dipolar cycloaddition of azomethine ylides has also been exploited to improve dispersibility of the nanotubes into polymer matrices [98] or to induce polymerization of acrylates via a grafting from approach for the preparation of nanocomposites [101].

However, the main limitations of the method using α-amino acids and aldehyde derivatives to generate the azomethine ylides are the long reaction times required. To overcome this issue, cycloaddition has been performed under microwave-assisted heating, reducing drastically the reaction time [102,103]. The use of microwave irradiation involves less severe reaction conditions and is in general more efficient in comparison with conventional techniques [104,105]. By functionalizing CNTs with thiol end groups, the location of the pyrrolidine units was assessed by TEM via specific interactions between gold nanoparticles and the thiol moieties. The functional groups were found evenly distributed on the nanotube surface. The methodology using azomethine ylides has been further improved to shorten reaction times and avoid the use of large amounts of organic solvents by developing continuous-flow [106] and solvent-free processes [107], opening opportunities for potential industrial scale-up. Prato and coworkers have proposed an alternative approach based on the generation of azomethine ylides by using aziridines as precursors under solvent-free microwave-assisted conditions (Fig. 1.6b) [108].

Another improvement of the cycloaddition of azomethine ylides has been developed to avoid the high temperatures that are required to induce decarboxylation of the iminium salts derived from the condensation of α-amino acids and aldehydes. In this study, the azomethine ylides are generated by double deprotonation of trialkylamine-N-oxide derivatives at moderate temperature (Fig. 1.6c) [109]. The functionalization is selective to semiconducting SWCNTs when the starting N-oxide bears a pyrene moiety, which acts as templating agent on the nanotube surface via π-stacking interactions. Solubilization of semiconducting SWCNTs in the presence of lignoceric acid through the formation of alkyl ammonium functionalities on the pyrrolidine rings allows to separate nonfunctionalized metallic SWCNTs from functionalized semiconducting SWCNTs.

1.2.4.1.2   Click Chemistry on CNTs.

Many functionalization reactions can be performed on the π-conjugated surface of CNTs. Most methods require experimental conditions, which may not be compatible with some organic or biological compounds. The Cu(I)-catalyzed azide–alkyne [3+2] cycloaddition (CuAAC) [110,111], which is a catalyzed variation of the Huisgen 1,3-dipolar cycloaddition, is a mild and efficient approach to conjugate CNTs with a variety of organic molecules and nanoparticles [112,113]. The CuAAC reaction, known under the term click chemistry, is highly regioselective and takes place between alkyne and azide functions, leading to the formation of 1,2,3-triazoles. The reaction cannot occur directly on the nanotube surface; it is necessary to first functionalize CNTs with the reactive groups, namely alkyne or azide groups. This is a limitation of this method. However, the click chemistry has several advantages: it can be performed in water, at room temperature in air, and the conditions are compatible with a wide range of functional groups. By exploiting the CuAAC reaction, various molecules have been coupled to CNTs pre-functionalized with alkyne groups in a controlled manner such as polymers (using grafting to or grafting from approaches) [114–117], chromophores [118–120], and nanoparticles [121–124] (Fig. 1.6d). These novel hybrid CNT conjugates can find potential applications in different fields, including nanocomposites and nanoelectronics. Giambastiani and coworkers recently reported the derivatization of phenylazido-functionalized SWCNTs by click chemistry with a variety of terminal alkynes bearing different functionalities (Fig. 1.6e) [125].

1.2.4.1.3   Other [3+2] Cycloaddition Reactions.

Other types of 1,3-dipolar cycloaddition have been investigated by using zwitterions resulting from the addition of pyridine and acetylene derivatives [126], nitrile oxides [127], nitrones [128,129], nitrile imines [130], pyridinium ylides [131], carbonyl ylides from oxiranes [132], and oxazolones via CO2 release [133] for the synthesis of cyclopentenone-, isoxazoline-, isoxazolidine-, pyrazoline-, indolizine-, tetrahydrofurane-, and pyrroline-grafted CNTs, respectively (Figs. 1.7 and 1.8).

Figure 1.7 1,3-Dipolar cycloaddition of CNTs using zwitterions resulting from the addition of pyridine and acetylene derivatives (a), nitrile oxides (b), and nitrones (c). Ar stands for aryl.

Figure 1.8 1,3-Dipolar cycloaddition of CNTs using nitrile imines (a), pyridinium ylides (b), carbonyl ylides (c), and oxazolones (d). Ar stands for aryl.

The group of Swager developed a strategy relying on the use of zwitterions as dipolar species, formed by addition of nucleophiles to activated electrophiles. Functionalization of SWCNTs was achieved with a zwitterion generated by reaction between 4-dimethylaminopyridine (DMAP) and dimethyl acetylenedicarboxylate (Fig. 1.7a) [126]. The positively charged DMAP on the cyclopentenone ring can be replaced by a methoxy or dodecanol moiety. The reaction has been extended to MWCNTs and different electrophiles have been tested, in particular acetylenedicarboxylate esters with chloroethyl, allyl, or propargyl groups, which can be further derivatized by SN2 substitution, thiol addition, or click chemistry, respectively. [134]. In addition, the charged DMAP in the intermediate cyclopentenone can be trapped by nucleophiles bearing a variety of functional groups. CNTs functionalized with this strategy have been exploited as chemiresistor sensor [135] and catalyst [136].

The 1,3-dipolar cycloaddition of nitrile oxides has been explored by a few research groups (Fig. 1.7b) [127, 137]. In particular, the reaction has been applied on easy dispersible SWCNTs pre-functionalized at the tips with pentyl esters to lead to the formation of isoxazolines bearing a pyridine ring [127]. Complexation of the pyridyl groups with porphyrins imparts interesting photophysical properties to the SWCNTs.

1.2.4.2 [2+1] Cycloadditions.

In addition to [3+2] cycloadditions, CNTs can be functionalized by [2+1] cycloadditions with highly reactive species such as nitrenes or carbenes.

1.2.4.2.1   Addition of Nitrenes.

Nitrenes contain a nitrogen atom with only six valence electrons. They can be either singlet-state nitrenes (with two filled p-orbitals, each filled with two electrons) or triplet-state nitrenes (with one filled p-orbital and two p-orbitals having unpaired electrons) [138]. Both states can add on the nanotube surface via [2+1] cycloaddition and via a reaction between biradicals and the π-conjugated sidewall of CNTs, respectively. Nitrenes are highly reactive electrophiles, therefore they cannot be isolated. They are generated in situ as reactive intermediates from azide derivatives by thermolysis or photolysis, with concomitant release of nitrogen gas.

The first cycloaddition of nitrenes on CNTs to form alkoxycarbonylaziridine rings has been reported by the group of Hirsch [139]. SWCNTs were functionalized by nitrenes using different azidocarbonates as precursors (Fig. 1.9a). Using this strategy, long alkyl chains, aromatic compounds, crown ethers, dendrimers, and oligoethylene glycol units were directly introduced on the nanotube surface [140]. By using di-functional molecules bearing two azidocarbonate moieties, Holzinger et al. obtained cross-linked SWCNTs [138].

Figure 1.9 [2+1] Cycloaddition of CNTs with nitrenes generated by thermal or photo-decomposition of azidocarbonates (a) or azides (b). [2+1] Cyclopropanation of CNTs via Bingel reaction (c), or with carbenes in situ generated from diaryldiazomethane (d) or diazirine derivatives (e). Ar stands for aryl.

Simple azide derivatives can be used as precursors of nitrenes. For instance, C2B10 carborane cages substituted with an azide function were in situ thermally decomposed to nitrene (Fig. 1.9b) [141]. The resulting SWCNTs functionalized with carborane were administered to mice and were found to concentrate in tumors cells, which make them promising for boron neutron capture therapy in the treatment of cancer. [2+1] Cycloaddition of nitrenes has also been extended to DWCNTs for grafting of different polyethylene glycol (PEG) chains to increase water dispersibility for potential biomedical applications [142]. An improvement of the method was brought by using microwave irradiation for thermal decomposition of azide derivatives in the presence of SWCNTs in order to decrease reaction times [143]. A green, low-cost, and super-gram-scale process has been developed to functionalize MWCNTs with nitrenes bearing bromine, carboxylic acid, hydroxyl, and amine moieties [144]. The functionalization degree can be controlled by adjusting the ratio of azides versus CNTs. The reactivity of the functional groups immobilized on the nanotube sidewall was demonstrated by inducing surface-initiated polymerization, amidation, and reduction of metal ions.

In addition to thermolysis of azide derivatives, photolysis of azides can lead to the formation of nitrenes. For instance, vertically aligned MWCNTs on a solid support have been functionalized with azidothymidine under UV irradiation [145]. The functional groups on the nanotube sidewall were visualized by TEM thanks to gold nanoparticles modified with complementary DNA. Photochemistry of azides has also been exploited for double functionalization of MWCNTs. Combination of UV irradiation at one side of a MWCNT film with azidothymidine and radical addition at the opposite side of the nanotube film affords dissymmetric end-functionalized MWCNTs [146]. Several perfluoroarylazides have also been used as nitrene precursors for functionalization of vertically aligned MWCNT forests [147]. The perfluoroarylazides contain a carboxyl moiety that can be derivatized by amidation or esterification.

1.2.4.2.2   Cyclopropanation.

CNTs have been functionalized by [2+1] cycloaddition via the Bingel reaction in the presence of a bromomalonate derivative and a base, which abstracts the acidic malonate proton. The resulting carbanion or enolate reacts with the double bonds of CNTs via a nucleophilic addition. This generates a carbanion on the nanotube surface that displaces bromine by a nucleophilic aliphatic substitution. This reaction leads to the formation of a cyclopropane ring on the nanotube sidewall [148]. Transesterification with an alcohol bearing thiol end groups allows to complex gold nanoparticles for direct visualization of the functional groups by AFM. Debundling and cyclopropanation have been accomplished on individual ultra-short (20–80 nm long) SWCNTs by in situ Bingel reaction under strongly reducing conditions (K°/THF) in the presence of a malonate derivative, a base, and CBr4 (Fig. 1.9c) [149]. The Bingel reaction has also been performed on dispersed SWCNTs, pre-functionalized by oxidation and amidation, under microwave irradiation to reduce reaction time and increase the amount of covalently linked substituents [150].

As alternative cyclopropanation strategy, CNTs have been functionalized by [2+1] cycloaddition of carbenes, which are the carbon analog of nitrenes. Oxidized SWCNTs, first functionalized by amidation, have been derivatized by dichlorocarbenes generated by the decomposition of PhHgCCl2Br [151]. Other sources of carbenes have been explored such as diaryldiazomethane [152] and diazirine derivatives [153] (under UV irradiation) with concomitant release of nitrogen gas (Figs. 1.9d and 1.9e).

1.2.4.3 [4+2] Diels–Alder Reaction and [2+2] Cycloadditions.

Diels–Alder reaction proceeds through [4+2] cycloaddition between dienes and dienophiles. The first Diels–Alder reaction was performed on oxidized SWCNTs, previously esterified with pentanol, and o-quinodimethane generated in situ from the corresponding 4,5-benzo-1,2-oxathiin-2-oxide under microwave irradiation (Fig. 1.10a) [154]. Clear indication of functionalization was given by scanning electron microscopy (SEM) showing at some points an increase in height of about 1–1.5 nm in the form of bumps on the surface of 2-nm diameter SWCNTs. Fluorinated SWCNTs have also been functionalized by Diels–Alder cycloaddition with a wide range of dienes, such as 2,3-dimethyl-1,3-butadiene, anthracene, and 2-trimethylsiloxyl-1,3-butadiene [155]. The high reactivity of fluoronanotubes can be explained by the bond strain of the remaining double bonds which is increased due to the fluorinated tetrahedral carbon atoms. In addition, the electron-withdrawing character of the fluorine atoms activates the double bonds. Another approach to favor Diels–Alder reaction on SWCNTs with electron-rich dienes relies on simultaneous activation by using high pressure (1.3 GPa) in the presence of a transition metal complex (chromium hexacarbonyl) (Fig. 1.10b) [156]. Complexation of SWCNTs to chromium, analogous to arene-chromium-tricarbonyl complexes, enhances the electrophilicity of nanotubes. In another study, cycloaddition of o-benzyne on SWCNTs has been performed via in situ thermal decomposition of benzenediazonium 2-carboxylate or by fluoride-induced elimination of trimethylsilyl and triflate groups from 2-(trimethylsilyl)aryltriflates [157].

Figure 1.10 Diels–Alder reaction on CNTs with o-quinodimethane (a) and 2,3-dimethoxy-1,3-butadiene in the presence of chromium hexacarbonyl under 1.3 GPa (b).

Reactivity of CNTs toward Diels–Alder cycloaddition is broad as the nanotubes can behave as either dienophile or diene depending on the reaction partner. Indeed, CNTs can react with furan derivatives as dienes or with maleimide compounds as dienophiles [158,159]. Semiconducting SWCNTs were shown to exhibit higher reactivity for Diels–Alder reaction with 1-aminoanthracene in comparison to metallic SWCNTs [160]. Both semiconducting and metallic SWCNTs with larger diameter display higher reactivity. In this case, the curvature-induced strain of SWCNTs is not the driving force, contrary to other functionalization reactions where the smaller diameter nanotubes have increased reactivity due to their increased curvature strain compared to the larger nanotubes.

Functionalization of SWCNTs by [2+2] cycloaddition of fluorinated olefins allows to convert pristine SWCNTs into high-mobility semiconducting nanotubes [161]. This approach is efficient and original as it does not require the separation of metallic and semiconducting SWCNTs.

1.2.5 Radical Addition

The group of Hirsch reported the first photoinduced addition of perfluorinated alkyl radicals onto SWCNTs [139]. Acyl peroxides, RC(O)O-O-(O)CR, can be used as radical precursors as they readily decompose to form free R• radicals upon mild heating with release of carbon dioxide. Using this strategy, SWCNTs have been reacted with benzoyl or lauroyl peroxide, as well as succinic or glutaric acid peroxide to produce phenyl-, undecyl-, 2-carboxyethyl-, or 3-carboxypropyl-functionalized SWCNTs, respectively (Fig. 1.11a) [162,163]. An ultra-fast (3–4 min), efficient, and solvent-free variation of this method has been developed for functionalization of SWCNTs in the presence of succinic acyl peroxide [164]. Thermal decomposition of benzoyl peroxide has also been exploited for alkylation of SWCNTs in the presence of alkyl iodides [165]. Besides, peroxides can serve as free radical initiators to induce in situ polymerization of polypropylene on SWCNTs [166]. The free radicals arising from thermal decomposition of benzoyl peroxide scavenge protons from polypropylene matrix, which provides reactive sites for covalent attachment on the nanotube surface. Alternatively, azo compounds, such as azonitriles, azoamides, and azoamidines can be employed to functionalize MWCNTs upon thermolysis by a free radical addition mechanism [167].

Figure 1.11 Functionalization of CNTs by radicals arising from thermal decomposition of glutaric acid acyl peroxide (a) and by reductive alkylation with 1-iodobutane (b).

Billups and coworkers have extensively studied reductive lithiation and alkylation of SWCNTs in liquid ammonia. Treatment of CNTs with lithium metal (or sodium, potassium) in liquid ammonia, followed by addition of either alkyl iodides [168], aryl iodides [169], or aryl/alkyl sulfides [170], yields partially debundled SWCNTs functionalized by alkyl or aryl moieties (Fig. 1.11b). The fast electron transfer from lithium to the CNTs induces debundling, thus allowing to perform chemistry at the single-tube level. In addition, the reduced CNTs have a higher nucleophilic character, which facilitates further reactions. This more pronounced reactivity has been exploited for alkylation of reduced SWCNTs in the presence of diacyl peroxides within a few minutes [171]. The reduced SWCNTs are highly reactive toward peroxides, which are strong oxidizing agents, thereby allowing the reaction to occur at room temperature. This is not the case for neutral SWCNTs as the functionalization requires thermal decomposition of the peroxides and long reaction time. Reductive lithiation of SWCNTs in the presence of N-halosuccinimide and dicumyl peroxide (isopropyl benzene peroxide) gives rise to succinimidyl- and alkoxy-derivatized SWCNTs, respectively [172]. Hydrolysis with hydrazine or with sulfuric acid yields amino- and hydroxyl-functionalized SWCNTs. The group of Hirsch has demonstrated that reductive alkylation is selective for metallic and small-diameter SWCNTs [173]. In another study, they investigated thoroughly carboxylation of SWCNTs by reductive lithiation under CO2 atmosphere [174]. In this case, the small-diameter semiconducting SWCNTs were derivatized first, followed by both larger diameter semiconducting and small-diameter metallic SWCNTs. The selectivity of the reaction is different in comparison with the previous study [173]. Finally, reductive functionalization of SWCNTs with lithium metal in the presence of acrylic ester monomers and catalytic amounts of di-tert-butylbiphenyl as electron carrier can induce polymerization on the nanotube surface, leading to polyacrylate-SWCNTs [175].

1.2.6 Nucleophilic and Electrophilic Additions

Hirsch and coworkers reported the addition of a nucleophilic carbene, dipyridyl imidazolidene, formed in situ by deprotonation of dipyridyl imidazolium [139]. This compound reacts with the nanotube surface to give a zwitterionic 1:1 adduct rather than a cyclopropane due to the stability of the resultant aromatic 14π perimeter. The same group has developed a method for alkylation of SWCNTs based on nucleophilic addition of organometallic reagents (carbon- and nitrogen-based carbanions) on the nanotube surface. This reaction is accompanied by the introduction of one negative charge per addition to the nanotube sidewall, leading to highly exfoliated negatively charged SWCNTs due to electrostatic repulsion. Subsequent reoxidation of the anionic intermediates with air oxygen gives neutral unbundled alkylated SWCNTs [176]. This method has been used with a variety of organolithium and organomagnesium compounds (nBuLi, tBuLi, EtLi, nHexLi, nBuMgCl, tBuMgCl) (Fig. 1.12a) [177]. This reaction displays a pronounced selectivity for metallic SWCNTs. Moreover, the reactivity toward the addition of organometallic compounds was found to be inversely proportional to the diameter of the nanotubes and dependent on the steric hindrance of the nucleophiles. The reaction has been extended to amine-based nucleophiles, in situ-generated lithium amides [178], and lithium acetylides [179] for direct covalent attachment of alkylamines or alkynes, respectively.

Figure 1.12 Functionalization of CNTs with n-hexyllithium (a) and with 4-ethoxybenzoic acid via a Friedel–Crafts reaction (b).

Functionalization of CNTs through organometallic reduction and electrophilic attack has been achieved by using alkyllithium compounds as nucleophilic reactive species, generating negative charges on the nanotubes with concomitant addition of alkyl moieties, and subsequent electrophilic attack with a series of halogenated electrophiles [180].

Tagmatarchis et al. have described the first electrophilic addition on CNTs via reaction with chloroform in the presence of aluminum chloride, followed by hydrolysis, resulting in the addition of hydroxyl functions on the sidewall of SWCNTs, which could be further derivatized by esterification [181]. Other research groups have exploited the Friedel–Crafts reaction to functionalize CNTs. In particular, the group of Baek has used this strategy to functionalize MWCNTs with various 4-substituted benzoic acids [182] and hyperbranched poly(ether-ketone) [183] in polyphosphoric acid (PPA)/phosphorus pentoxide medium (Fig. 1.12b). Alternative conditions for Friedel–Crafts acylation of SWCNTs with different acyl chlorides, including lauroyl chloride, involve the use of either aluminum chloride in nitrobenzene or an AlCl3/NaCl melt at 180°C [184]. MWCNTs functionalized with 1-butyric acid moieties have been prepared by Friedel–Crafts reaction using succinic anhydride in aluminum chloride [185]. Under microwave irradiation, the reaction of 3-chloropropene with AlCl3 gives rise to the formation of a cationic species that adds on the surface of SWCNTs in the form of polymer chains [186].

1.2.7 Plasma Functionalization and Mechanochemical Treatment

An alternative method to solution-phase reactions for functionalization of CNTs relies on plasma activation. Plasma is extensively used for surface modification of various materials. Plasma treatment presents many advantages: it can be scaled-up and reaction times are generally lower compared to conventional solution-phase reactions. In addition, when gases are used there is no generation of solvent waste. Plasma activation has been used initially by Chen et al. to treat aligned CNTs with acetaldehyde or ethylenediamine for the formation of Schiff bases to graft polysaccharide chains in the presence of sodium cyanoborohydride [187]. Various plasma gases can be used, such as oxygen to introduce hydroxyl, carbonyl, and carboxyl groups on the surface of MWCNTs; NH3 for the formation of amines, amides, and nitriles; and CF4 for the grafting of fluorine atoms [188]. Different plasma conditions, such as the power, type of gas, treatment time, pressure, and position of the sample inside the chamber, have been investigated to assess the influence on the functionalization of CNTs. Plasma treatment of MWCNTs with oxygen results mainly in the formation of carbonyl species, while the main functional groups formed by treatment with nitric acid are carboxylic and phenolic moieties [189]. Alternatively, water can be used to functionalize CNTs with oxygen-containing species [190]. The group of Meyyappan has investigated functionalization of SWCNTs with nitrogen-containing species by using either ammonia [191] or nitrogen [192] glow discharge.

A few mechanochemical treatments have been applied to functionalize CNTs. During mechanochemical processes, mechanical motions and energy control the chemical reactions. Local high-pressure spots are generated, which bring the reacting species into the closest contact to the nanotube surface and favor chemical reactions. The treatment also cuts drastically the nanotubes, inducing cleavage of C–C bonds of CNTs. The resulting active sites can react with molecules in gas phase or adsorbed on the nanotube surface. Ball milling of MWCNTs in reactive atmospheres such as H2S, NH3, and Cl2 gas allows to introduce a variety of functional groups such as amines, amides, thiols, and mercaptans under mild conditions at room temperature [193]. SWCNTs can also be functionalized by ball milling in the presence of various reagents, such as vapors of alkyl halides [194], potassium hydroxide to cover the nanotube sidewall with hydroxyl groups [195], C60 fullerene [196], or ammonium bicarbonate for the introduction of amines and amides [197]. As alternative to reductive alkylation in solution, SWCNTs have been alkylated and arylated via ball milling in the presence of alkyl or aryl chlorides and potassium, which transfers electrons to the nanotubes to make negatively charged CNTs that can react with the chloride derivatives [198]. Finally, Vázquez and coworkers have cut and functionalized pristine SWCNTs by [2+1] cycloaddition of nitrenes by ball milling in the presence of 2-azidoacetate at room temperature for 30 min [199].

1.3 CHARACTERIZATION OF FUNCTIONALIZED CARBON NANOTUBES

Several techniques have been used to characterize the structure and morphology of functionalized CNTs, including spectroscopic, microscopic, and thermal methods [200–202]. The most extensively used analytical techniques are Raman and UV/Vis/NIR spectroscopies that allow to determine whether the functional groups are covalently bonded or simply adsorbed on SWCNTs. Thermogravimetric analysis (TGA) provides quantitative information on the degree of functionalization. Microscopy techniques such as TEM, SEM, and AFM are used predominantly to qualitatively establish the morphology of CNTs. X-ray photoelectron spectroscopy (XPS) and IR spectroscopy give information on the nature of the functional groups grafted on the nanotube surface. No single measurement technique provides a complete characterization of functionalized CNTs. Therefore, only the combination of a variety of analytical techniques can give sufficient data for the characterization of functionalized CNTs.

1.3.1 Spectroscopic Techniques

1.3.1.1 Raman Spectroscopy.

Raman spectroscopy is extensively used to probe the structure, diameter, and electronic properties of SWCNTs [203]. It is one of the most important techniques for the characterization of functionalized SWCNTs [204]. Due to the one-dimensional structure of SWCNTs, the π-electronic density of states of SWCNTs form sharp singularities, namely van Hove singularities. The resonance Raman signals from SWCNTs can be obtained when the laser excitation energy is equal to the energy separation between the van Hove singularities in the valence and conduction bands. But, it is restricted to the selection rules for optically allowed electronic transitions according to the so-called Kataura plot that gives information on which nanotubes are resonant for a given excitation wavelength as a function of the nanotube diameter [205].

Raman spectra of SWCNTs contain three main peak areas (Fig. 1.13). The first feature between 150 and 350 cm−1 originates from the radial breathing mode (RBM) where all carbon atoms vibrate radially in phase. It is absent in graphite and is not detected for diameters above 2 nm, therefore this peak is also absent for MWCNTs. The RBM frequency is inversely proportional to the RBM position, thus it is possible to determine the diameter of SWCNTs present in the sample. As Raman scattering in CNTs is a resonant process, the RBM frequency is dependent on the excitation energy. To assess the diameter distribution, it is required to use different excitation energies. The second peak zone is the D-band (diamond band or disorder band) at 1300–1400 cm−1 and is assigned to defects in the nanotube structure, including sp³-hybridized carbon atoms and other carbon species present in the carbonaceous impurities. The intensity of the D-band is low for SWCNTs, but it is high for MWCNTs due to the presence of a large amount of defects. The third region located between 1500 and 1600 cm−1 is related to the C=C stretching tangential mode. The most intense band at ∼1590 cm−1 is named the graphitic (G) band. The shape of the G-band can be used to distinguish between metallic and semiconducting SWCNTs.

Figure 1.13 Raman spectra of pristine SWCNTs (solid line) and of covalently functionalized SWCNTs (dashed line).

Covalent functionalization induces rehybridation of the derivatized carbon atoms from sp² to sp³, resulting in an increase of the relative intensity of the D-band (ID) by comparison with the G-band (IG) (Fig. 1.13). A quantitative relationship between the relative intensity of the D-band and the degree of functionalization is not possible. However, it can give a rough estimate of the extent of functionalization if a significant amount of sp²-hybridized carbon atoms have been converted to sp³-hybridized carbons. In addition, the intensities of the RBM and G-band are decreased after functionalization. The RBM can completely disappear in the case of high functionalization degrees, while the G-band can be broader. Thus, the increase of the ID/IG