Defect functionalization is based on the attachment of the desired functional group to the existing or intentionally created defects on the CNT. In reality, CNTs are not ideal nanostructures but rather contain 1-3 % of defects naturally occur during the synthesis. Another type of defect can be invariably introduced during the preparation and purification stages. Ultrasonic treatment and shear mixing, widely used to disperse CNTs in solution, can introduce the defects. The effect of sonication is based on the cavitation effect that is able to break the CNT aggregates (bundles) in addition to the shear force effect. This causes the formation of the structural defects, eventually their breakage. After defect introduction the sidewalls are expected to be chemically active improving the reactivity of CNT.
Oxygenated sites can also to be considered as defects. Chemical oxidation of nanotubes is mainly achieved by using either wet chemistry or gaseous oxidants. One way of creating defects and functional groups is an oxidative damage to the nanotube framework by strong acids or oxidants, such as HNO3, H2SO4, KMnO4 which leave holes functionalized with oxygenated functional groups as well as open CNT’s caps caps [12][13][14]. The oxidation can be also archived by using hydrochloric acid [15][16], hydrogen peroxide [17], potassium permanganate [18], oxidative persulfates [18], hypochlorites [19]. In Ref. [53] it was shown that the oxidation of CNTs with concentrated HNO3 results in the defect formation in addition to the opening the ends of nanotubes. Besides that, it stimulates the formation of the aromatic polycyclic compounds, fulvic acids on the surface, which can be removed by alkali and then reversibly reabsorbed. This consequently results in the improving dispersion of CNTs in the aqueous solution.
The gas-phase functionalization of CNTs, including oxidation, has been carried out. Here, carbon dioxide [20][21], water vapor [22], air, oxygen [23][24][25], ozone [26][27][28], used as oxidants. Water vapor is effective as an oxidizing reagent only when it is introduced into the reaction during CVD synthesis of CNTs [22]. It was shown that in this method nanotubes are produced with a sufficiently large number of defects and oxygen content up to 42 wt. % in different functional groups. So far, the oxidation with oxygen in air is the most effective way of oxidation as it proceeds with the simultaneous CNT’s cleaning, shorting and opening. In general oxidation treatment leads the p-type doping of CNTs shifting the Fermi level towards valence band.
Moreover, some functionalization methods are combined together. For example, in [29], the effectiveness of the same carboxylation of CNTs in a mixture of nitrogen-nitric acid (3:1) is accompanied by cavitation caused by the ultrasonic action at room temperature. In addition, the hydrogen peroxide effectiveness in oxidation of CNTs increases with UV irradiation [30][31]. Air can be used for additional oxidization of CNTs pretreated with acids [32].
It was also shown that the oxidation processes strongly depends on the curvature of the CNT’s wall and consequently to the diameter dependence [33][34][35]. Upon oxidative functionalization of CNTs, the affinity of their surfaces to various solvents and polymer matrices increases. The defects on CNTs created by oxidants are stabilized by bonding with carboxylic acid (–COOH) or hydroxyl (–OH) groups, ketone, alcohol and ester groups formation. These groups can be used for precursors for further functionalization: silanization [36][37][38], esterification [39][40][41], thiolation, alkylation and arylation, biomolecules and polymer attachment.
CNT silanization is the interaction of oxidized CNTs with organosilanes. The silanization after oxidation usually takes place in a few steps. Firstly, the hydrolysis of silanes into silanols is performed. The creation of hydrogen bonds between oligomers and the surface hydroxyl groups of carbon nanofillers as well as the condensation of silanols into oligomers occur in the next phase. Finally, Si-O-C bonds are created by the interaction of Si-OH groups with OH groups on CNTs. The 3-aminopropyltriethoxysi-lane has been widely used as a silane agent [42][43][44][45][46]. Lin et al. used comparable techniques to attach both polyethylene glycol and iron oxide in-silica nanoparticles to the surface of CNTs [47].
Access to numerous CNT surface functionalizations is made possible by adding carboxylic and hydroxyl groups to CNTs. CNTs can be functionalized via esterification or amidation. Typically, carboxylic groups are activated to acyl chlorides by thionyl chloride, and then amidation and esterification reactions convert them to carboxamides (or carboxyesters) to introduce a variety of functional groups as nucleophiles, such as aliphatic or aryl amines, amino acid derivatives, ensims, peptides, or amino-group substituted dendrimers. Thiolation with sodium sulfite, sodium hydroxide, and carboxylation with thionyl chloride are used to attach a similar thiol group [48]. The presence of carboxil groups themselves leads to the reduction of van der Waals interactions between CNT which results in effective debundeling and separation into individual but shortened nanotubes.
The chemical manipulation of carbon nanotubes sidewall by fluorination and then defluorination is well developed [49][50][51]. It can be carried out efficiently in the presence of F2 at temperatures between 150 and 400 °C [52]. The interest is that the fluorine atoms can then be substituted by alkyl groups or by amines. Fluorinated CNTs are widely used for a wide range of chemical functionalization. For example, side-wall alkylation and arynation of fluorinated SWCNTs was performed using alkyllitium species, including methyl, ethyl, butyl, hexyl, octyl, dodecyl. Amino groups can be also attached to the sidewalls of fluorinated CNTs [53]. These materials exhibit an excellent reinforcement effect in epoxy-based systems which results in improved properties as mechanical robustness. The chlorination and bromination are another two methods that has been used for SWCNT functionalization. Chlorine applied in the functionalization of CNTs can be liquid precursors as thionyl chloride, chloroform, trichloromethane, tetrachloroethylene, or hydrochloric acid or solid precursors as dichlorocarbene or chlorine gaseous precursors [54][55][56][57]. Purified CNT sample attached 1.8 at.% of bromine after two days in saturated bromine vapors at room temperature [58][59].
The most common approach for hydrogenation is a wet chemistry which includes the dispersion of a mixture of lithium and CNT in diaminoethane [60]. The protonation of reduced lithium also reduce CNT with methanol leading to the formation of hydrogenated SWCNTs [61]. As the result the covalently bonded CH derivatives are formed. High-boiling polyomides also was used as hydrogenation reagent for CNTs [62]. At room temperature, it has been demonstrated that hydrogenated SWNTs with covalent C-H bonds constitute a stable hydrogen storage [63][64][65].
CNTs functionalized with fluorescein or a fluorescent bioactive peptide can also effectively cross the cell membrane without the significant damage of the membrane [66][67]. For therapeutic biomolecules, such as proteins, peptides, deoxyribonucleic acid (DNA), and siRNA, covalent functionalization has been widely reported [68]. For example, CNTs were used to deliver streptavidin into the cell [69]. It has been demonstrated that SWNTs may be covalently functionalized to peptide nucleic acid (PNA), an analog of uncharged DNA, and then hybridized to molecular wires containing complementary DNA [70]. There have been several reports of bovine serum albumin protein molecules adhering to f-MWCNTs in an aqueous buffer solution [71]. CNTs can introduce a DNA molecule into the cell nucleus which opens an opportunity to use CNTs in gen-delivery [72]. CNT, after covalent functionalization, can be used to carry peptide antigens [73]. Recently, the translocation of the RNA using non-functionalized SWCNT into breast cancer cells has also been reported [74].
In general, the CNTs covalently-functionalized are soluble in many organic solvents due to the attachment of polar or non-polar groups. The main difficulty in developing covalent func[71]tionalization [71] arise from the conversion sp2 into sp3 hybridization. This consequently lead to changes in mechanical, optical and electrical properties. From the other hand, the CNTs functionalized covalently open the road to many application which require the change in the optical and electrical properties [75][76].
The first commonly used non-covalent functionalization is surfactant based. These types of molecules results in the dispersion of the CNTs via physical adsorption. Both ionic and non-ionic surfactants have been used. The known surface charge of CNTs can be used to predict colloidal stability of CNT solutions and interaction (adsorption) mechanism with ionic surfactants.. A larger negative surface charge and steric repulsion, for instance, are produced when CNTs and SDS interact through the hydrophobic components, which increases the stability of the dispersion. Therefore, the purification procedure and tube wall functionalization, which influence its surface charge, may be used to regulate the adsorption mechanism of ionic surfactants [77]. While the polar portion of surfactants interacts with solvent molecules, the hydrophobic (non-polar) portion is often oriented toward the surface of CNTs. The most popular surfactants for dispersing CNTs in an aqueous media are sodium dodecyl sulphate (SDS), sodium dodecyl benzene sulfonate (SDBS), cetyltrimethyl ammonium bromide, Tween, Triton X, and a siloxane polyether copolymer [78][79][80][81][82][83][84]. One of the important parameters for the dispersion of CNTs is CNT to surfactant concentration. For example, a homogeneous dispersion of CNTs was found at the optimum composition of 0.5 wt% CNTs and 2 wt% SDS relative to water. In Ref. [85] found that the surfactant dispersing power follows the order of SDS < Tween 20 < Tween 80 < TX-100. Meanwhile, it was shown that SDBS exhibited superior capability than TX100 and SDS in terms of its capacity to disperse SWCNTs [10]. However, it was reported that the dispersion of MWCNTs in three surfactants followed the order of SDS < CTAB < TX100 [86]. While surfactants are often necessary to utilize individualized SWCNTs in the dispersion, improved surfactant-removing protocols to recover the initial SWCNT properties are needed to ensure optimal performance. The washing with organic solvents [87], acidic oxidation [88], and annealing in inert [87] or oxygen atmospheres are used for this purpose [89].
Through π-π interfaces, aromatic compounds like anthracene, phenanthrene, pentacene, porphyrin, pyrene, and fullerene interact with one another. Under mild sonication, the functional molecule's and large system's strong bonding allows for de-bundling and produces hydrophilicity. CNT@Pyrene hybrid was developed, aiming at improvement in heat transport by the effect of anion addition [90]. Biotin−streptavidin recognition method described in [91]. This complex showed stability in 18 days, and excellent loading capacity (about 1 streptavidin tetramer per 20 nm of SWNT). The anthracene groups were covalently added to the hydroxylated CNTs to aid in orienting the laccase active sites for direct electron transfer between enzymes and electrodes [92]. The excellent electrochemical catalytic property and stability of the functionalized with ferrocene SWCNTs was reported by Yang et al [93]. Huang et al. have used the ferrocene for non-covalent functionalization of SWCNT for L-glutamate detection [94].
The interaction of polymers with CNT surface via wrapping was used to improve the dispersion of CNT in water and organic solvent, and feather enable separation of CNTs by metallicity and chirality, and CNTs from carbonaceous and metal impurities [95][96][97]. The polyvinyl pyrrolidone, polyvinylidene fluoride, polystyrene sulfonate, cellulose derivatives, polypyrroles, glycolipids, and redox polymers was successfully used for CNT dispersion [98].
Dispersions of MWCNT with the pluronic surfactant F-127 were made with the intention of examining the interactions of CNT with biological tissues [99]. A water-soluble globular protein called bovine serum albumin adheres to the surface of CNTs and improves dispersibility [100]. Similar results were shown for MWNT-PEG composites [101]. It was concluded that the functionalizing CNT with PEG enhanced the dispersion of the CNTs and increased their heat capacities [102][103]. On act as a shRNA delivery vehicle, a modified branched polyethylenimine was grafted to carboxylated single-walled carbon nanotubes (SWCNT) using a polyethylene glycol (PEG) linker [104]. CNTs functionalized with DNA were shown to enhance stability used in different applications [105][106][107][108][109]. Similarly to proteins, DNA form tight helices around CNTs, or noncovalent conjugates. DNA-functionalized CNTs were more effective than plain MWCNTs against malignant tissues [109]. Despite the popularity of research on SWCNT-DNA conjugates, only a small number of works has studied the removal of adsorbed DNA on SWCNTs by biological, physical and chemical treatments. Numerous biological and pharmacological uses of chitosan, including drug delivery, cancer treatment, and biosensors, have been researched [110][111][112]. Chitosan could effectively disperse the SWCNTs and provide a suitable biological interface for immobilization of biomolecules [113]. The functionalized CNTs complexed siRNA demonstrated 10–30% silencing activity and 10–60% cytotoxicity [114]. siRNA delivery using CNTs were successfully utilized [115].
Finally, wresearchers would like to discuss the mechanical interlocking of SWCNT as a possible route for non-covalent functionalization. In particular, in [116] the macrocycle precursors that were terminated with bisalkenes that were wrapped around the NT by ring-closing metathesis were furnished with two-extended tetrathiafulvalene SWNT recognition units.
Nitrogen has one electron more than C, and when it more easily substitutes for C atoms. There are two types of C–N bonds occur: pyridine-type in which each N atom is bonded to two carbon atoms and a type in which N atom is bonded to three carbon atoms. The first type result in either a p- or n-type doping depending on the concentration of N atoms, meanwhile the second type to n-type conducting behavior. In situ doping techniques include high-temperature arc-discharge [117][118], chemical vapor deposition(CVD) [119][120][121][122],chemically solvothermal procedures [123], and laser ablation methods [124][125]. Also N type doping can be archived by various post-treatment methods [126][127]. The post-thermal treating of oxidized CNTs with ammonia was performed [128]. Doping of isolated SWNTs with nitrogen was achieved by mixing appropriate amounts of acetonitrile with xylene. The 2–6 at.% of nitrogen in the SWNTs was observed in controllable manner [129].
The halogens such as fluorine, chloride and bromide can be also attached to the sidewall of CNTs. As it was mentioned in section 1 the fluorination can be effectively used for feather covalent functionalization. The fluorination of CNTs, the formation of so-called fluorotubes, is usually performed by temperature treatment in molecular fluorine environment at 150-600 °C. The precise temperature control results in the control the activity of the substitution, and therefore, in the degree of the fluorination. The heavily doped CNTs was produced and showed increased solubility in alcohols [130]. The comparison between graphene and CNTs fluorination suggested that the curvature resulted in a weakening of the C-F bonding covalence [131]. By diluting fluorine with nitrogen in a ratio of 4:1 and treating SWCNTs with such a mixture at 250°C for 10 h the fluorinaized tubes with the stoichiometry of CF0.43. The heating resulted NTs under He with temperatures ranging from 100 to 400 °C desorb the fluorine[132]. Hydrazine defunctionalizes fluorinated nanotubes, allowing for the recovery of intact CNTs after treatment. Interestingly fluorinated single-walled nanotubes were highly insulating. The fluorination using various synthesis parameters has been studied [133][134].
The alternative commonly used method for CNTs functionalization via substitution is chlorination. Usually, the source of chlorine came from liquid precursors as thionyl chloride, chloroform, trichloromethane, tetrachloroethylene, or hydrochloric acid or solid as dichlorocarbene or gaseous precursors as Cl2 . The removal of metal catalyst and amorphous carbon residues from CNT has been reported to be accomplished using Cl2 gas and Cl2 aqueous solution treatments. Simultaneously, this protocol introduce chlorine but with a relatively low amount [135]. The chlorine substitution with the concentration of 2.5 atom % on SWCNTs was shown by using electrochemical functionalization in HCl and KCl aqueous solution [136]. A similar concentration (up to 2.9 atom %) of chlorine was introduced by treatment with SOCl2 under high temperature and pressure with prolonged reaction times of up to 14 days [137]. Recently, carbon tetrachloride cold plasma was shown to produce chlorinated CNTs [138]. Chlorine is fixed to the nanotubes' edges and flaws as a result of the operation.
A method of purification for CNT polluted by carbon impurities was used because of the low sensitivity of CNTs to bromination. However, the C-Br bond is more reactive than the comparable iodinated fluoro- or chloro-derivatives. CNTs bromination needed relatively difficult conditions as compared to other carbon nanostructures. With bromine in the liquid, gas, or plasma phases, it can only be done by electrochemical, electrothermal, or microwave processes at a higher temperature. Despite the improvement that has been shown, bromination causes a small amount of CNT degradation and fragmentation [139]. Bromine-containing SWCNT was demonstrated to have metallic properties [140]. In [141], the CNTs were combined with liquid bromine for 10 days at 55°C and then repeatedly washed with CCl4, carbon tetrachloride.The perpendicular arrangement of bromine atoms in relation to the CNT surface was readily seen using TEM. Br atom conjugation was made possible in the graphite structure of CNTs (2.8 percent weight) by electrochemical oxidation of inorganic salts such KBr [142]. Additionally, CNTs that had already been treated in aqueous solution were brominated using microwaves. A high Br atom concentration of 5-8 percent by weight was archived [143]. Elemental bromine and a group of Lewis acids (BBr3, BF3-Et2O, AlBr3, FeBr3, ZnBr2, etc.) may also be used to chemically bromine CNT, producing materials with 10–20% wt Br content. According to Bulusheva, the suction of CNT in liquid Br yields the C17Br stoichiometry. The covalent bromination process with a total weight of 4.9 % in CCl4 solution at microwave irradiation, and a cool CH3Br plasma were described in [144]. N-bromosuccinimide, NH4NO3/NBS, and Br2 were used in an electrophilic addition and radical reaction with UV light or conventional heating to archive the effective bromination [145]. In comparison to other methods, this one was proven to be the most mild, delicate, ecologically friendly, and quick. The effective bromine doping of single-walled carbon nanotube films under ultrasound was recently published [146].
The literature has covered a lot of various substitution methods, including the substitution by I, S, Si, Se, P-N, and Ar. An other technique for changing the graphitic structure of CNTs is iodination. First, a modified Hunsdiecker reaction was used to carry it out [147]. It was found that Cl, Br, and I easily produced halogenated CNTs at high temperatures and pressures. The CNTs have also been doped with iodine atoms using the straightforward solvent-thermal reaction process and their characteristics have been examined [148]. In order to illustrate the replacement of impurities in both materials, mono- and divacancies were produced in graphene and SWCNTs using the Ar plasma treatment [149]. Using STEM and other techniques, the Ar ions were plainly visible on the surface of CNTs.