Synthesis and characterization of multi-walled carbon nanotubes modified with octadecylamine and polyethylene glycol

1 Introduction

Scientific and industrial efforts have been recently focused on the field of nanotechnology and new emerging nanomaterials such as carbon nanotubes which exhibit superior properties compared to their micro- or macro-size counterparts. Since the discovery of carbon nanotubes (CNTs) in 1991 (Iijima, 1991), CNTs have become the focus of scientists and researchers all over the globe. A momentous research work has been done to discover the unique properties and characteristics of CNTs and to explore their key applications. CNTs demonstrate the exceptional properties such as unique size distributions, novel hollow-tube structures, high specific surface areas and electrical semi-conductivity and conductivity (Aqel et al., 2012). These characteristics allow applications of CNTs in many fields such as photocatalysis (Zhang et al., 2011), medicine (Tran et al., 2009), nanoscale electronics (Hermann et al., 2010), hydrogen storage (Reyhani et al., 2012), analytical chemistry (Latorre et al., 2012), mechanical systems (Otero et al., 2012), SEM probes (de Jonge, 2009), electron field emission tips (Musatov et al., 2005), corrosion protection (Abdel Salam et al., 2011a; Hermas et al., 2012), and adsorbents for a wide range of chemicals such as aniline (Al-Johani and Abdel Salam, 2011), palladium (Afzali et al., 2012a), pentachlorophenol (Abdel Salam and Burk, 2010; Abdel Salam, 2012), heavy metals (Abdel Salam et al., 2011b), and manganese ions (Afzali et al., 2012b). The unusual interest in carbon nanotubes (CNTs) resides in their possible technological applications in various fields of science. Recent reports on the chemical compatibility and dissolution properties of CNTs show a great deal of interest in developing modification or functionalization of their surfaces (Baskaran et al., 2004). The modification procedure may be divided into two categories: non-covalent and covalent functionalizations. Noncovalent functionalization of nanotubes is of particular interest as it preserves the integrity and consequently the physical properties of CNTs meanwhile it improves their solubility and processability. This type of functionalization could be achieved by wrapping CNTs by polymers (Ntim et al., 2011) or surfactants (Gao et al., 2008). The main potential disadvantage of noncovalent attachment is that the forces between the wrapping molecule and the nanotube might be weak, thus as a filler in a composite the efficiency of the load transfer might be low. On the other hand, the covalent functionalization can improve both solubility and dispersion of CNTs in different solvents and polymers. This allowed more interactive forces between the CNTs and other materials. For example, the functional group at the surface of CNTs, can increase the selectivity of CNTs as an adsorbent toward certain pollutants (Abdel Salam and Burk, 2008; Abdel Salam and Burk, 2009; Abdel Salam, 2012). Also, the functional groups at the surface of CNTs make the strongest type of interfacial bonding with the polymer matrix. The covalent functionalization can be accomplished by either modification of surface bound carboxylic acid groups on the nanotubes or direct reagents attached to the side walls of nanotubes. The formation of the carboxylic group on the surface of CNTs is the first step in most of the covalent modifications of CNTs. Covalent functionalization has been studied previously using different functional groups (Darabi et al., 2012; Blondeau et al., 2011; Shi and Xing, 2009; Wang, 2009; Kakade and Pillai, 2008; Xu et al., 2007). The functionalized nanotubes usually have electrical, mechanical, or optical properties, which make them different from their original nanotubes. Therefore, it is an interesting subject to functionalize CNTs for various kinds of applications, such as sensors and gas storage devices (Bangerjee et al., 2003), molecular level electronic (Singh et al., 2006), and advanced technology (Meng et al., 2009).

In the present paper, the chemical modification and characterization of MWCNTs via side wall carboxylic acid functionalization and further modification using ODA (octadecylamine, CH₃–(CH₂)₁₇–NH₂) and PEG (polyethylene glycol, HO–(CH₂–CH₂–O)_n–OH), were studied. The percentage of functionalization of MWCNTs with both ODA and PEG was estimated using thermal analysis.

2 Materials and methods

2.2

2.2 Procedures

MWCNT surfaces were chemically functionalized in different steps. The functionalization scheme of MWCNTs and formation of MWCNTs–PEG (multi-walled carbon nanotubes with side wall functionalization with polyethylene glycol groups) and MWCNTs–ODA (multi-walled carbon nanotubes with side wall functionalization with octadecylamine groups) are shown in scheme 1.

Close

Scheme 1

Surface functionalization of MWCNT surface and the formation of MWCNTs–PEG and MWCNTs–ODA.

Export to PPT

2.2.1

2.2.1 Oxidation of MWCNTs

MWCNTs were purchased from Sun Nanotech (China), with average diameters of 100–200 nm, and were used as received. The MWCNTs (10.0 g) were added to 400 mL solution of 8 M HNO₃. The suspension was refluxed at 140 °C for 4 h, then washed with distilled water until the wash water was pH neutral and centrifugation (3000 rpm) was used for solid–liquid separation. The solid products were dried in air at 50 °C.

2.2.2

2.2.2 Formation of carbonyl chloride groups on MWCNTs

After oxidation of the MWCNTs with the nitric acid and the introduction of carboxylic groups (MWCNTs–COOH), the MWCNTs with carbonyl chloride groups (MWCNTs–COCl) were prepared as follows:

Oxidized MWCNTs (MWCNTs–COOH), 10 g, were stirred in 200 ml of SOCl₂ (thionyl chloride) in the presence of 5 mL of dimethylformamide (DMF) at 70 °C for 24 h. After centrifugation of the reaction solution, the brown–black supernatant was decanted and the remaining solid was washed with anhydrous tetrahydrofuran (THF). After centrifugation, the pale yellow-colored solution was decanted and discarded. The remaining solid was dried at room temperature under vacuum.

2.2.3

2.2.3 Formation of MWCNTs–PEG

After MWCNTs–COCl had been prepared, the dried solid was reacted with polyethylene glycol (PEG) as follows:

MWCNTs–COCl, 5 g, was mixed with 10 g of polyethylene glycol (PEG) in 100 ml benzene/THF solvent mixture (v/v = 3/1) and stirred for 40 h at 80 °C. After centrifugation, the black solid was washed with deionized water until the wash-water was clear. The remaining black solid was retained after drying at vacuum.

2.2.4

2.2.4 Formation of MWCNTs–ODA

MWCNTs–COCl, 5 g, was mixed with 20 g of octadecylamine (ODA) at 100 °C for 96 h. After cooling to room temperature, the excess ODA was removed by washing with ethanol with sonication for 10 min. This washing process was repeated eight times, until the ethanol was clear. The remaining black solid was dissolved in dichloromethane, and after centrifugation, the remaining black-colored solid was taken to dryness under vacuum.

3 Results and discussion

3.1

3.1 Thermal gravimetric analysis (TGA)

TGA was used to explore the degree of functionalization of MWCNTs with PEG and ODA. TGA experiment was performed on MWCNTs–PEG and MWCNTs–ODA and the results were compared with those of both the pristine MWCNTs and PEG and ODA pure materials, as a function of temperature at a heating rate of 10 °C/min. As presented in Fig. 1, the pristine MWCNTs weight decreased with increasing temperature, but this weight loss was insignificant (less than 2%), compared with pure PEG and MWCNTs–PEG. Pure PEG weight decreased sharply with increasing temperature and reached 98% weight loss near 400 °C. On the other hand, MWCNTs–PEG was more stable within the temperature range and its weight loss reached 16% at 800 °C. This weight loss may be due to mostly the decomposition of the PEG chains. Fig. 2 shows the TGA plots of pristine MWCNTs, MWCNTs–ODA and pure ODA. It is clear from the graph that pure ODA weight decreased sharply with raising the temperature, above its boiling point (55–57 °C) and reached 98% weight loss near 300 °C, whereas MWCNTs–ODA weight decreased steadily with increasing temperature and its weight loss reached 39% at 800 °C. This weight loss may be due to the decomposition of the ODA chains. The surface area of MWCNTs covered with PEG or ODA can be estimated from the % of MWCNTs functionalized according to the TGA analysis. Mainly, both compounds, ODA and PEG, when they attached to MWCNTs, form a long chain out of MWCNT surfaces. By imagining that this chain consists mainly of H–C–H groups connected together from the carbon atom the length and the width of these can be calculated from the C–H bond length (109 pm). So, the surface area occupied by each molecule would equal to 4.75 × 10⁻²⁰ m²/molecule. TGA analysis shows MWCNTs–ODA contains 39% by weight ODA. Calculating the total surface area covered by ODA chains indicated that almost 46.3% of the MWCNT surfaces is covered by ODA chains. Taking into consideration that the ODA chains are not attached to each other may lead to the approximation that more than 46.3% of the MWCNT surfaces is covered with ODA chains. The same calculation was done for PEG chains, and the result showed that 0.38% of the MWCNT surfaces was covered with the PEG chains. Taking into consideration the fact that PEG chains are long polymeric chains with an average molecular weight of 10,000/mole, may cover a larger surface of the MWCNTs, as these long chains can be tangled, folded and unfolded, which increase the surface of the MWCNTs covered. The total concentration of the oxygen containing active sites (carboxylic, phenolic, lactonic groups) on the surface of the oxidized MWCNTs was determined using acid–base titration, and it was found to be 1.27 mmole/g of MWCNTs. This concentration is almost close to the number of moles of ODA that had been attached to the MWCNTs; 1.46 mmole ODA/g of MWCNTs. This approximation could indicate that most of the acidic functional groups that were created on the surface of the MWCNTs via oxidation had been attached to ODA groups. Meanwhile, the number of PEG moles was estimated to be 0.016 mmole PEG/ g MWCNTs. This number was much lower than the number of acidic functional groups on the MWCNT surface available for binding.

Close

Figure 1

TGA plots of pristine MWCNTs, MWCNTs–PEG, and pure PEG.

Export to PPT

Close

Figure 2

TGA plots of pristine MWCNTs, MWCNTs–ODA, and pure ODA.

Export to PPT

3.2

3.2 Infrared spectroscopy

IR measurements were performed to provide more evidence of attachments of PEG and ODA to MWCNTs. Fig. 3 shows the IR spectra of the pristine and oxidized MWCNTs, and MWCNTs–PEG and pure PEG. It is clear from the Figure that introducing PEG groups on the surface of the oxidized MWCNTs changed its IR spectra slightly. The peak at approximately 3500 cm⁻¹, characteristic of an H bonded O–H stretch, became more pronounced in MWCNTs–PEG than the oxidized MWCNTs. This is may be due to the introduction of more OH groups with the PEG on the surface of MWCNTs. A sharp peak at 1730 cm⁻¹ corresponding to C⚌O stretching vibration of the carboxylic acid became less intense with the MWCNTs–PEG. The MWCNTS–PEG spectrum shows a peak at 1100 cm⁻¹ corresponding to the C–O stretch vibration of the ether group of PEG (same peak appears in pure PEG). Two more strong peaks appear more between 2850 and 3000 cm⁻¹ due to the C–H stretching in the PEG chain. The same peaks were stronger on the pure PEG spectrum. The intensities of the above mentioned peaks were not strong compared with the pure PEG compounds due to the fact that nearly 16% of the MWCNTs surface was functionalized with the PEG groups, as was concluded from the TGA study. This finding was confirmed by another study done to study the attachment of PEG to single-walled carbon nanotubes (Jung et al., 2004). Fig. 4 shows the IR spectra of the pristine MWCNTs, oxidized MWCNTs, and MWCNTs–ODA, as well as pure ODA. Sidewall functionalization of MWCNTs changed their spectrum significantly. The peak at approximately 3500 cm⁻¹; characteristic of an O–H stretch, became broader with MWCNTs–ODA compared with the oxidized MWCNTs, likely due to decreasing the number of the OH groups by reaction with ODA. Further evidence of the functionalization of the MWCNTs with ODA is the appearance of two strong peaks between 2850 and 3000 cm⁻¹ assigned to the C–H stretching of the ODA chains, similar to pure ODA.

Close

Figure 3

IR spectra of pristine MWCNTs, oxidized MWCNTs, MWCNTs–PEG and Pure PEG.

Export to PPT

Close

Figure 4

IR spectra of pristine MWCNTs, oxidized MWCNTs, MWCNTs–ODA and Pure ODA.

Export to PPT

3.3

3.3 Modified MWCNTs solubility

Solubilities of the modified MWCNTs were studied using different solvents; deionized water, dichloromethane, methanol, and hexane. Fig. 5 shows the photos of pristine MWCNTs, oxidized MWCNTs, MWCNTs–ODA, and MWCNTs–PEG, suspensions in different solvents. It is clear from the Figure that pristine MWCNTs were completely insoluble (unsuspendable) in both water and methanol, but significantly soluble, forming a dark black solution, in dichloromethane, a relatively non polar solvent. This is due to the hydrophobicity of the MWCNTs, which allows them to be suspended in non-polar solvents. Although hexane is a non-polar solvent, the pristine MWCNTs was not significantly soluble, this may be due to the different polarities between MWCNTs and hexane. Substance will dissolve in a solvent of similar polarity. If the solvent is too polar such as water or methanol, the relatively non-polar pristine MWCNTs cannot “separate” the solvent molecules from one another, and hence the solubility is low. On the other hand, if the solvent is too non-polar, such as hexane, then it cannot “separate” the MWCNTs from one another, and again solubility is low. If the polarity of both the solute (MWCNTs) and the solvent is just matching, such as dichloromethane, then the solute and solvent can separate on another, resulting in solubility. So, the solubility is due to both solvent and solute polarity. On the other hand, oxidation of the MWCNTs and the introduction of carboxylic groups increased the hydrophilic character of the MWCNTs. That was clear from the Figure as oxidized MWCNTs were significantly soluble in water and methanol and poorly soluble in dichloromethane and hexane. Further modification of the oxidized MWCNTs with PEG groups did not affect the solubility, as the MWCNTs–PEG was significantly soluble in water and methanol and poorly soluble in dichloromethane and hexane. On the other hand, further reaction of the oxidized MWCNTs with ODA, decreased the hydrophilic character of the CNTs, due to the formation of MWCNTs–ODA tubes. The solubility in polar solvents; water and methanol, was greatly inhibited whereas the solubility in dichloromethane and hexane relatively non-polar solvents was greatly enhanced. It is noteworthy to mention that introduction of the ODA groups along the surface of the MWCNTs only decreased the hydrophilicity of the tubes, but also it decreased their density; compared with water (MWCNTs–ODA floated on water).

Close

Figure 5

Solubilities of the modified MWCNTs in different solvents (from left to right, deionized water, dichloromethane, methanol, and hexane).

Export to PPT

3.4

3.4 X-ray diffraction measurements

X-ray diffraction was applied in order to investigate the change in the crystalline structure of the pristine MWCNTs upon the side-wall functionalization with PEG and ODA. Fig. 6 compares the XRD patterns for the pristine MWCNTs, MWCNTs–PEG, and MWCNTs–ODA. Generally, it is clear from the Figure that the pristine MWCNTs, MWCNTs–PEG, and MWCNTs–ODA, exhibit two peaks; one at 26.14° (0 0 2 plane) and the other at 44.22° (1 0 0 plane), corresponding to an interplanar space of 3.41 Å and 2.05 Å, respectively. The Figure shows that functionalization of the MWCNTs with PEG groups did not alter the crystallographic character of the MWCNTs as it possessed only two prominent peaks, whereas the MWCNTs–ODA was characterized by a slightly different pattern which could be due to the high degree of functionalization of ODA compared with PEG on the surface of the MWCNTs; nearly 39% (by weight) of MWCNTs were functionalized with ODA, and 16% (by weight) were functionalized with PEG (TGA data). MWCNTs–ODA sample contains three different peaks at 31.72°, 45.4°, and 56.56°, with interplanar d-spacing 2.8185 Å, 1.9960 Å, and 1.6258 Å, respectively. The XRD of the pure ODA was performed and its pattern was compared with that of the MWCNTs–ODA. There were no matches with previously mentioned peaks present on MWCNTs–ODA with those of the pure ODA. The pure ODA XRD pattern shows many uniform peaks, which are considered to result from the crystalline structure of the pure ODA. The new peaks present on the XRD pattern of the MWCNTs–ODA may be due to the breakdown of the ODA crystalline structure.

Close

Figure 6

XRD patterns of pristine MWCNTs, MWCNTs–PEG, MWCNTs–ODA, and pure ODA.

Export to PPT

3.5

3.5 Scanning electron microscopy (SEM)

The surface morphological structure of the modified MWCNTs was studied using SEM imaging. Fig. 7 shows representative images for the pristine MWCNTs, MWCNTs–PEG, and MWCNTs–ODA. It is clear from the Figure that the pristine MWCNTs are highly tangled tubes with diameters of 100–200 nm. Although, side-walled functionalization with PEG and ODA did not alter the morphological structure of MWCNT surfaces significantly, as observed from their SEM images, the diameter of the CNTs upon modification increased slightly.

Close

Figure 7

SEM images of pristine MWCNTs, MWCNTs–PEG, and MWCNTs–ODA.

Export to PPT

4 Conclusions

Side-walled functionalization of MWCNTs was achieved using polyethylene glycol and octadecylamine. Different characterization techniques were used to investigate the formation of MWCNTs–PEG and MWCNTs–ODA. TGA showed that almost 16% (by weight) of the MWCNTs were functionalized with PEG and 39% (by weight) were functionalized with ODA. The percent surface coverage was estimated for both MWCNTs–ODA and MWCNTs–PEG, and calculations showed that total surface areas covered were 46.3% and 0.38%, respectively. The IR spectra for MWCNTs–PEG and MWCNTs–ODA identified the characteristic stretching peaks for the successful linkage between MWCNTs and PEG and ODA. MWCNTs–PEG contains many of the hydrophilic groups and was suspendable in aqueous solution, whereas MWCNTs–ODA contains more hydrophobic groups and was not suspendable in aqueous solution, but suspendable in relatively non-polar organic solvents, such as dichloromethane. XRD and SEM measurements showed that there was not any significant change of either the crystalline structure or the surface morphology of the MWCNTs after the functionalization with PEG and ODA.