Green tea-aided dispersion of single-walled carbon nanotubes in non-water media: Application for extraordinary reinforcement of nanocomposite fibers

Abstract

Green tea has recently been reported to be a good dispersant of single-walled carbon nanotubes (SWCNTs) in aqueous media. In the current work, green tea extract powder was examined for dispersion of SWCNTs in organic solvents. Dimethyl sulfoxide (DMSO) was found to be a good solvent of green tea extract for dispersing SWCNTs. A combination of green tea (dispersant)/DMSO (solvent)/polyvinyl alcohol (PVA) (nanotube wrap) was obtained that resulted in the dispersion of SWCNTs almost to individual nanotubes or to very thin nanotube bundles. The reinforcing ability of highly dispersed SWCNTs was explored in a PVA matrix producing PVA-SWCNT gel-spun fibers. Extraordinary reinforcing effects on the tensile and dynamic mechanical properties of the composites were observed by incorporating a minute amount of SWCNTs (∼1 wt% to PVA). The analysis revealed effective stress transfer through the PVA crystalline interfaces surrounding the nanotubes.

Keywords

Polymer-matrix composites green tea carbon nanotubes tensile strength toughness

Single-walled carbon nanotubes (SWCNTs) are known to have an extremely high Young’s modulus, close to 1 TPa, and tensile strength in the range of 50–150 GPa.¹ They have generated a great deal of interest as ideal candidates for mechanical reinforcement of polymers and polymeric fibers. In spite of their excellent mechanical properties, their use in effective reinforcement of SWCNTs in composites has not been fully realized due to their poor dispersion in most solvents; this limits the effective surface area of nanotubes in composites and hinders stress transfer.² To obtain the optimum reinforcing effect, the SWCNTs must be dispersed to the level of perfectly or nearly isolated tubes individually coated with polymer in polymer-based composites. The strength and toughness of the composites tend to be very sensitive to the quality of dispersion and adhesion of polymer/SWCNTs.³

A variety of methods have been attempted to improve the dispersion of nanotubes by non-covalent functionalization, including surfactants,⁴ polymers^5,6 and biomolecules such as DNA,⁷ peptides,⁸ polysaccharides⁹ and proteins.¹⁰ Recently, Nakamura et al.¹¹ discovered that green tea can solubilize SWCNTs in water. Green tea is composed of a group of water-soluble polyphenol compounds, epigallocatechin gallate (EGCg), epicatechin gallate (ECg), epigallocatechin (EGC) and epicatechin (EC).¹² Among these, EGCg is the most abundant component in green tea, and is acknowledged to have antioxidant, antitumor and anticancer properties.¹³

Although dissolution of SWCNTs in green tea has been reported in water media,^11,14 the utility of green tea in other organic solvents, that is, in non-water media, is of great interest. In this connection, we adopted green tea extract powder instead of aqueous green tea and found that the tea powder dissolves easily in dimethyl sulfoxide (DMSO). A combination of green tea as the dispersant, DMSO as the solvent and diluted polyvinyl alcohol (PVA) solution as the nanotube wrap appears to disperse SWCNTs to almost individual nanotubes or very narrow bundles of a few tubes. The potential reinforcement potential of dispersed SWCNTs was explored by applying it in a PVA matrix. PVA-SWCNT fibers have been fabricated by gel spinning, and the morphology, thermal and mechanical properties of the prepared fibers were investigated.

Experimental details

Materials

PVA chips (Poval-HC, Kuraray Co. Ltd) with degree of polymerization (DP) and saponification of ca. 1500 and 99.9%, respectively, were used. The SWCNTs were synthesized by a method called ‘enhanced-Direct Injection Pyrolytic Synthesis’ (e-DIPS) by our co-research group of the National Institute of Advanced Industrial Science and Technology (AIST), Japan.¹⁵ The SWCNTs had a purity of 97.5%, average tube diameter ca. 2 nm and average tube length of 10–20 µm. Green tea extract powder (brand name: Sunphenon DK) of Taiyo Kagaku Co. Ltd, Japan was used. Sunphenon DK contains over 80% polyphenols and EGCg is the major component.

PVA-SWCNT suspensions

PVA chips were dissolved in DMSO in a static rotary mixer (50 rpm) at 90°C for 4 h to obtain 3 wt% (as the nanotube wrap) and 25 wt% (as the matrix polymer) solutions. The green tea extract powder was dissolved in DMSO, and SWCNTs were added to it and sonicated (with 125 W on a Branson water-bath ultrasonic soniﬁer 2510 J-DTH) for 2 h. The green tea/DMSO/SWCNT suspension was transferred to a homogenizer (PH91, SMT Co. Ltd, Japan) and homogenized for 10 min at a rotational speed of 15,000 rpm. The 3 wt% PVA solution in DMSO was poured into the homogenizer and homogenized for 15 min. This suspension was then blended for 15 min with the required amount of the 25 wt% PVA/DMSO solution to produce blends of 0.3, 0.5, 0.7 and 1.0% SWCNT/PVA. The PVA concentration was 15 wt% in each suspension. The green tea extract was four times the nanotubes’ weight. The prepared spinning suspensions were de-aerated at 80°C for 4 h prior to spinning.

Gel spinning

The polymer solution was placed into a syringe and injected at a rate of 0.5 g min⁻¹ through a 0.57 mm diameter needle into a methanol coagulating bath by pressing the syringe using a pump. A heater surrounding the syringe was set to 80°C. The temperature of the coagulating bath was maintained at around −20°C; however, the composite fibers showed an inhomogeneous surface structure after drying (as detailed in the ‘Results and discussion’ section). The temperature of the spinning bath was optimized at 0°C where fibers with homogeneous morphology were obtained. During spinning, the fibers were continuously wound at 5.3 m min⁻¹ onto an 11-cm diameter polyvinyl chloride bobbin. The fiber bobbin was kept rotating in a pure methanol bath by changing the methanol several times until the green tea was removed from the fibers (traced by color change of methanol). The fibers were subsequently dried in air for 48 h.

Measurements

Transmission electron microscopy

A few drops of nanotube suspensions were deposited on carbon-coated electron microscope grids and allowed to dry. The sample grids were then observed in a JEOL JEM-2010 at an accelerating voltage of 200 kV.

Optical microscopy

The morphology of the suspensions was observed with an optical microscope (Olympus BH-2, Japan).

Fourier transform infrared spectroscopy

The fiber samples were dried under vacuum at 40°C for 12 h and then the Fourier transform infrared (FTIR) spectra were taken with an FTIR-8400 S instrument, Shimadzu Ltd, Japan. All spectra were collected with a 4 cm⁻¹ wavenumber resolution after 50 continuous scans.

Scanning electron microscopy

The morphology of the fibers was examined using a Hitachi S-2380 N scanning electron microscope (SEM) with an accelerating voltage of 25 kV after sputter-coating the samples with platinum.

Differential scanning calorimetry

Differential scanning calorimetry (DSC) analyses were performed on a Perkin-Elmer Pyris-1 analyzer. Both temperature and heat ﬂow were calibrated with a standard indium reference. Fiber weights were maintained within the range of 4–5 mg. All thermal analyses were carried out under a dry nitrogen atmosphere.

The degree of crystallinity (χ _c ) of the PVA component in the composite fibers was calculated based on the following equation:

(1)

where ΔH⁰ _m = 161 J g⁻¹ is the heat of fusion for 100% crystalline PVA¹⁶ and w is the weight fraction of the PVA matrix in the composite.

Before DSC scans, fibers were cut into tiny fragments and dried in an oven at 40°C overnight. For the non-isothermal crystallization study, the samples were first heated to 250°C and then maintained at this temperature for 10 min to ensure complete melting of the PVA. They were then cooled to room temperature at a cooling rate of 10°C min⁻¹.

Wide-angle x-ray diffraction

Wide-angle x-ray diffraction (WAXD) was carried out with a Rigaku Rotorflex RU200B X-ray generator operated at 40 kV and 150 mA. The radiation used was Ni-filtered Cu-K_α (wavelength 0.1542 nm).

Tensile measurements

Tensile properties were measured with a Tensilon Model RTC-1250 A using a 50 N load cell, 20 mm gauge length and 20 mm min⁻¹ crosshead speed. Prior to tensile testing, samples were conditioned with relative humidity (RH) of 65% at 20°C for at least 24 hrs.

Dynamic mechanical analysis

Dynamic mechanical analyses (DMAs) were performed in tensile mode using an ITK Co. DVA-225 instrument, at a frequency of 10 Hz, strain amplitude of 0.15% and a heating rate of 10°C min⁻¹ on fibers 20 mm in length. Fibers were dried at 40°C and conditioned with RH 65% at 20°C for 24 hrs before DMA testing.

Raman spectroscopy

Raman spectra were obtained using a Hololab 5000 instrument in conjunction with an attachment specially designed to take Raman spectra together with a gradual increase in the load on the fiber. The excitation source was an argon laser with a wavelength of 532 nm and 50 mW power. The laser power was adjusted by an attenuator to minimize the heating effects of the laser beam on the sample.

Results and discussion

The dissolution of green tea in DMSO is shown in Figure 1. The green tea extract powder instantaneously dissolved in DMSO (Figure 1(a)) and after stirring for 1 min yielded a clear reddish solution (Figure 1(b)). With the addition of SWCNTs and following sonication for 2 h, a homogeneous nanotube dispersion was obtained (Figure 1(c)). This dispersion was treated in a mechanical homogenizer, and a 3 wt% PVA solution in DMSO as the nanotube wrap was agitated through the homogenizer. Finally, a 25 wt% PVA/DMSO solution was added as the matrix polymer. When green tea was used, no macroscopic nanotube aggregates were observed in spinning suspensions up to 1 wt% nanotube loading. The optical micrograph of 0.5 wt% SWCNT loaded dispersion is shown in Figure 2(a). In order to substantiate the good SWCNT dispersion with the assistance of green tea, a separate spinning suspension without green tea with the same nanotube content (0.5 wt%) was prepared in a similar manner and is shown in Figure 2(b). A distinct improvement in the dispersion quality of nanotubes is seen when green tea was used as the dispersant.

Figure 1.

(a) Just after the addition of green tea extract powder into dimethyl sulfoxide (DMSO) showing instantaneous dissolution. (b) Green tea/DMSO solution. (c) Green tea/DMSO/single-walled carbon nanotube (SWCNT) suspension after sonication for 2 h. Inset: green tea extract powder.

Figure 2.

Optical micrographs of the polyvinyl alcohol-single-walled carbon nanotube (PVA-SWCNT) 0.5 wt% spinning suspensions prepared (a) with the assistance of green tea and (b) without green tea.

The dispersion states of SWCNTs were observed by transmission electron microscopy (TEM). Figure 3(a) shows pristine SWCNTs indicating their inherent tendency to form larger bundles. After dipping them into the green tea/DMSO solution and sonicated for 10 min, the green tea particles were found to be adhered to the nanotubes’ surface, as shown in Figure 3(b). After sonication for 2 h, the larger bundles of SWCNTs were converted into smaller bundles (Figure 3(c)). Since green tea contains numerous polyphenols,¹⁴ non-covalent interaction between SWCNTs and polyphenols by π-π dispersive interactions is believed to cause separation of the nanotube bundles through sonication.¹¹ After the addition of the diluted PVA solution as the nanotube wrap and mechanical homogenization for 15 min, as shown in Figure 3(d), it is clear that a major portion of the SWCNTs was converted into individual nanotubes (marked by arrows). We anticipated much better nanotube dispersion in the final spinning suspension after adding 25 wt% PVA solution into it; however, it was difficult to trace the nanotubes in the TEM image since the well-dispersed nanotubes were hidden within the surrounding highly concentrated PVA polymer.

Figure 3.

Transmission electron microscopy (TEM) images of (a) pristine single-walled carbon nanotubes (SWCNTs), (b) SWCNTs in green tea/dimethyl sulfoxide (DMSO) after 10 min sonication, (c) SWCNTs in green tea/DMSO after 2 h sonication and (d) SWCNTs dispersed in green tea/DMSO/diluted PVA (3 wt%) solution after homogenization.

The surface morphologies of the gel-spun neat PVA and composite fibers were studied by SEM images and are shown in Figure 4. For the neat PVA fibers, a homogeneous cross-section was obtained in a methanol coagulating bath at any temperature in the range from 0 to −20°C. The outer layer formed by the gelation is softer and thus the fiber has a smooth appearance when prepared even at low temperatures (–20°C). The neat PVA fibers prepared at 0°C are shown in Figure 4(a). Unlike neat PVA, the composite fibers spun at ∼−20°C exhibited an irregular morphology (Figure 4(b)). This is considered to be caused by the low temperature quenching, leading to faster gelation of the solution, along with formation of a hard outer layer. Consequently, the fibers shrink inhomogeneously with slow elusion of DMSO. However, the bath temperature was then further investigated, and homogeneous composite fibers were obtained at a bath temperature of 0°C (Figure 4(c)). In this study, the coagulation bath temperature appeared to be a significant factor dictating the morphology of the PVA-SWCNT composite fibers.

Figure 4.

Scanning electron microscope (SEM) images of (a) neat polyvinyl alcohol (PVA) fibers prepared at 0°C, (b) and (c) are PVA-single-walled carbon nanotube (SWCNT) 0.5 wt% fibers prepared respectively at ∼ −20°C and at 0°C methanol coagulating bath.

While surfactants are typically used to disperse nanotubes, any additional surfactant is usually left in the polymer matrix, which has a negative influence on the final properties of the composite.¹⁷ In this work, care was taken to avoid this, since the remaining green tea components surrounding the nanotubes, as seen in Figure 3(b) and (c), could hamper effective stress transfer from the matrix to the nanotubes. However, green tea was found to be spontaneously eliminated in the spinning bath and subsequent methanol wash bath, traced by the color of the methanol turning reddish. FTIR spectra taken for the neat PVA and composite fibers did not show any discernible difference, indicating the absence of untraceable remnant amounts of green tea (not shown here). For preparing nanotube-reinforced composites using a dispersant, wet spinning was found to be advantageous, since the dispersant has the scope to be removed with solvent from the spinning dope to the spinning bath while coagulating the fiber.

Tensile testing was carried out for the fibers, and representative stress–strain curves are shown in Figure 5(a). Figure 5(b) shows an enlarged view of the initial portion of these curves. A substantial improvement in the mechanical properties of the composites can be observed with increasing SWCNT content. For the neat PVA fiber, the curve shows an initial linear portion, a marked yield point and a natural necking plateau followed by strain-induced crystallization and subsequent stiffening. For the composite fibers, the yield point shifted steadily upward with rising initial slopes (modulus); the plateau region before strain hardening gradually decreased with increasing nanotube content and was barely visible for the fiber containing 1.0 wt% nanotubes. The tensile strength increased with increasing nanotube amount along with decreasing elongation at break.

Figure 5.

(a) Representative stress–strain curves of neat polyvinyl alcohol (PVA) and PVA-single-walled carbon nanotube (SWCNT) fibers, and (b) enlarged initial portion of stress–strain curves to show the modulus increase with the increase of SWCNT content.

The tensile properties, including yield strength, tensile strength, Young’s modulus, elongation at break and toughness (energy required to break), evaluated from the stress–strain curves, are summarized in Table 1. For the composite fibers, all the tensile properties showed remarkably higher values than those of the neat PVA. For the 1 wt% nanotube content sample, the yield stress, tensile strength, Young’s modulus and toughness increased by factors of 3.6 (from 37 to 135 MPa), 4.2 (from 42 to 177 MPa), 4.1 (from 1.199 to 4.93 GPa) and 2.9 (from 333 to 970 J g⁻¹), respectively.

Table 1.

Tensile and dynamic mechanical properties of neat polyvinyl alcohol (PVA) and PVA-single-walled carbon nanotube (SWCNT) composite fibers

Sample	Yield strength (MPa)	Breaking strength (MPa)	Young’s modulus (GPa)	Elongation at break%	Toughness (J g⁻¹)	Storage modulus at
Sample	Yield strength (MPa)	Breaking strength (MPa)	Young’s modulus (GPa)	Elongation at break%	Toughness (J g⁻¹)	(−100°C) GPa	(25°C) GPa	(100°C) GPa
Neat PVA	37	42	1.19	13.5	333	13	1.57	0.28
SWCNT 0.3%	86	105	2.97	12.4	818	19	3.86	0.48
SWCNT 0.5%	104	145	4.06	11.3	960	25	4.95	0.66
SWCNT 0.7%	119	165	4.52	10.4	1073	32	5.28	0.76
SWCNT 1.0%	135	177	4.93	8.5	970	35	5.62	0.98

With a view to observe the influence of SWCNT dispersion by green tea to the mechanical properties of fibers, PVA/SWCNT 0.5 wt% fibers without using green tea (an image of dispersion is shown in Figure 2(b)) were prepared and analyzed. The fiber properties were as follows: yield stress: 77 MPa, tensile strength: 107 MPa, Young’s modulus: 2.46 GPa, elongation at break: 10.45% and toughness: 695 J g⁻¹. Comparing these data with those provided for green tea/PVA/SWCNT 0.5 wt% fibers (Table 1), the merit of using green tea for dispersing SWCNTs in attaining remarkably higher mechanical properties is evident.

Figure 6 shows the dynamic storage modulus (E′) curves of the fibers over a wide temperature range. The E′ curve for the neat PVA fiber shows the typical behavior of a semicrystalline polymer. The drop in E′ in the temperature range of 0–70°C is associated with the glass-rubber transition of the PVA amorphous phase. At a temperature around 230°C, E′ dropped sharply due to melting of the PVA crystalline regions. Compared to neat PVA, all the PVA-SWCNT fibers displayed higher E′ from the glassy to the melting of PVA, and the E′ increased with increasing nanotube content. The values of E′ in the glassy (−100°C), room temperature (25°C) and rubbery (100°C) regions collected from the respective E′ curves are listed in Table 1. By adding 1 wt% SWCNTs, the composite fibers demonstrated an increase by factors of 2.7 (13 to 35 GPa), 3.6 (1.57 to 5.62 GPa) and 3.5 (0.28 to 0.98 GPa) at the glassy, room temperature and rubbery regions, respectively.

Figure 6.

Temperature dependence of storage modulus (E′) curves for neat polyvinyl alcohol (PVA) and PVA-single-walled carbon nanotube (SWCNT) fibers.

In addition to the large reinforcement effect, a thermal stabilization effect in the E′ curves of the composite fibers was also observed after PVA melting at ca. 230°C (Figure 6) for a SWCNT loading of 0.5 wt% and higher. This can be explained by the possible restriction of the PVA melt-flow supported by the nanotube network. Since nanotubes are not highly oriented in fibers, such networks can be formed by the available surfaces of dispersed nanotubes, especially when nanotube content is higher, such as 0.5 wt% or above in the present case. The increasing trend in E′ with increasing temperature may be due to stiffening of the fibers caused by the gradual dehydration of the side chains and the main chains of the PVA polymer itself:¹⁸

(2)

or by the formation of some kind of intermolecular crosslinking.

The reinforcing mechanism in the nanocomposites was elucidated by a combination of DSC and Raman spectroscopy measurements. The heat of fusion (ΔH_m), degree of crystallinity (χ _c ) and melting temperature (T_m) of the PVA components obtained from the DSC analysis are compiled in Table 2. The χ _c of neat PVA was found to be higher than that usually observed for as-spun PVA fiber.¹⁹ This might have resulted from the orientation-induced crystallization of PVA chains during continuously winding of fiber during spinning, and/or during the sample drying for moisture evaporation before the DSC experiment. However, WAXD patterns (not shown here) obtained for the neat PVA and composite fibers exhibited an even Debye–Scherrer ring, indicating no preferential orientation of PVA crystallites. Similar to the increase of nanotube content in the composites, the χ _c value increased linearly for composite fibers, although no change in T_m was observed. It is noteworthy that the neat PVA and all composite fibers were wound at the same winding speed. Thus, there was no opportunity to increase the PVA crystallinity in the composites by more orientation of the PVA chains. Therefore, the increase in χ _c could be ascribed to the formation of a crystalline layer of PVA surrounding the nanotubes’ surface. Such crystal growth can be expected to occur easily due to the interaction between PVA and SWCNTs by charge transfer.²⁰ The nucleation effect of the SWCNTs is also responsible for PVA crystal growth surrounding the nanotubes.^3,21

Table 2.

The heat of fusion (ΔH_m), degree of crystallinity (χ_c) and melting temperature (T_m) of polyvinyl alcohol (PVA) component in neat PVA and PVA-single-walled carbon nanotube (SWCNT) fibers

Sample	ΔH_m (J g⁻¹)	χ_c (%)	T_m (°C)
Neat PVA	92	57	232
SWCNT 0.3%	95	60	232
SWCNT 0.5%	102	64	232
SWCNT 0.7%	108	68	232
SWCNT 1.0%	116	73	232

Figure 7 shows DSC thermograms of dynamic cooling crystallization for the neat PVA and PVA-SWCNT fibers. Compared to the neat PVA, both the crystallization temperature (T_c) corresponding to the peak temperature of the crystallization exotherm and the temperature associated with the initiation of the crystallization process increased for the composite fibers. This phenomenon supports the possible interaction of PVA and SWCNTs by which nanotube-promoted PVA crystallinity was anticipated. Such crystalline interfacial regions of PVA surrounding the nanotubes are expected to play a significant role in reinforcing the composite properties in two ways: (i) the interfacial crystalline regions are inherently strong and stiff, thus acting as a reinforcing agent themselves; and (ii) since the crystalline interfacial regions result in better stress transfer to nanotubes compared to amorphous regions,^22,23 a strong load-bearing effect can be expected from this crystalline interface. Further evidences to confirm the interfacial stress transfer were obtained by Raman spectroscopy.

Figure 7.

Differential scanning calorimetry (DSC) cooling curves of neat polyvinyl alcohol (PVA) and PVA-single-walled carbon nanotube (SWCNT) fibers.

The D* Raman band of the carbon nanotubes appearing at ∼2600 cm⁻¹ is usually used to study load transfer from the matrix to the nanotubes.^24,25 In this connection, the PVA-SWCNT composite fibers were subjected to gradual loading and simultaneously the Raman shift corresponding to the D* band of SWCNTs at ∼2667 cm⁻¹ was monitored. For all the composite fibers, the shift in peak (band) position was observed. Figure 8 shows the relationship between the stress applied to the PVA-SWCNT 1.0 wt% fiber and the measured peak position of the Raman D* band. The shifting of the D* band position with applied stress indicates excellent stress transfer from the matrix to the filler, possibly generated from the strong interaction of the crystalline PVA-wrapped SWCNTs.

Figure 8.

Relation between the stress applied to the polyvinyl alcohol-single-walled carbon nanotube (PVA-SWCNT) 1 wt% fiber and the shift of Raman D* band. Inset: shifting of Raman D* band with applied stress.

Conclusions

SWCNTs were highly dispersed in green tea/DMSO/PVA solution, and the reinforcibility of dispersed nanotubes was examined by preparing nanocomposite fibers using PVA as a matrix. By incorporating a small amount of SWCNTs, outstanding improvements in tensile and dynamic mechanical properties of the PVA fibers were achieved. The study indicates that the applied load was effectively transferred from the matrix to the SWCNTs based on the strong interface between the nanotubes and PVA.

If nanotubes are incorporated in larger bundle forms, only an increase in the Young’s modulus of the composites is generally obtained. Two other essential tensile parameters, the tensile strength and toughness, are very sensitive to the quality of the nanotube dispersion.²⁶ In the current work, we obtained impressive enhancement in the strength, modulus and toughness of PVA fibers with a minute amount of SWCNTs. The possible reinforcing effects can be summarized as being due to the following reasons:

i) the good SWCNT dispersion that greatly enhances the available surfaces of nanotubes in the composites (judging from Figures 2 and 3),

ii) the formation of a large number of PVA crystallites surrounding nanotubes (judging from the increased crystallinity, see Table 2);

iii) the broad network of nanotubes in the whole composite (judging from thermal stabilization of the storage modulus curves, see Figure 6); and

iv) the strong interfacial interaction between PVA and the nanotubes (not only by van der Waals interaction but also due to charge transfer) that resulted in effective stress transfer to the reinforcing phase (judging from Figures 7 and 8).

Footnotes

Funding

This work was supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan through a Grant-in-Aid for the Global COE program.

References

Coleman

Khan

Blau

Gun’ko

. Small but strong: a review of the mechanical properties of carbon nanotube-polymer composites. Carbon 2006; 44: 1624–1652.

Cadet

Coleman

Ryan

Nicolosi

Bister

Fonseca

. Reinforcement of polymers with carbon nanotubes: the role of nanotube surface area. Nano Lett 2004; 4: 353–356.

Coleman

Cadek

Blake

Nicolsi

Ryan

Belton

. High performance nanotube-reinforced plastics: understanding the mechanism of strength increase. Adv Func Mat 2004; 14: 791–798.

Moore

Strano

Haroz

Hauge

Smalley

Schmidt

. Individually suspended single-walled carbon nanotubes in various surfactants. Nano Lett 2003; 3: 1379–1382.

O’Connell

Boul

Ericson

Huffman

Wang

Haroz

. Reversible water-solubilization of single-walled carbon nanotubes by polymer wrapping. Chem Phys Lett 2001; 342: 265–271.

Star

Stoddart

Steuerman

Diehl

Boukai

Wong

. Preparation and properties of polymer-wrapped single-walled carbon nanotubes. Angew Chem Int Ed 2001; 40: 1721–1725.

Zheng

Jagota

Semke

Diner

Mclean

Lustig

. DNA-assisted dispersion and separation of carbon nanotubes. Nat Mater 2003; 2: 338–342.

Xie

Ortiz-Acevedo

Zorbas

Baughman

Draper

Musselman

. Peptide-crosslinking modulated stability and assembly of peptide-wrapped single-walled carbon nanotubes. J Mater Chem 2005; 15: 1734–1741.

Star

Steuerman

Heath

Stoddart

. Starched carbon nanotubes. Angew Chem 2002; 41: 2508–2512.

10.

Karajanagi

Yang

Asuri

Sellitto

Dordick

Kane

. The synthesis of reactive poly(p-phenylene terephthalamide) s: a route towards compression stable aramid fibres. Langmuir 2006; 22: 1392–1395.

11.

Nakamura

Narimatsu

Niidome

Nakashima

. Green tea solution individually solubilizes single-walled carbon nanotubes. Chem Lett 2007; 36: 1140–1141.

12.

Chang

Chiu

K-L

Chen

Y-L

Chang

C-Y

. Separation of catechins from green tea using carbon dioxide extraction. Food Chem 2000; 68: 109–113.

13.

Kuzuhara

Suganuma

Fujiki

. Green tea catechin as a chemical chaperone in cancer prevention. Cancer Lett 2008; 261: 12–20.

14.

Chen

Lee

Vedala

Allen

Star

. Exploring the chemical sensitivity of a carbon nanotube/green tea composite. ACS Nano 2010; 4: 6854–6862.

15.

Saito

Ohshima

Okazaki

Ohmori

Yumura

Iijima

. Selective diameter control of single-walled carbon nanotubes in the gas-phase synthesis. J Nanosci Nanotech 2008; 8: 6153–6157.

16.

Kubo

Kadla

. The formation of strong intermolecular interactions in immiscible blends of poly(vinyl alcohol) (PVA) and lignin. Biomacromolecules 2003; 4: 561–567.

17.

Liu

Sun

Lin

Zhou

Wang

Fan

. Scratch-resistant, highly conductive, and high-strength carbon nanotube-based composite yarns. ACS Nano 2010; 4: 5827–5834.

18.

Yang

Shao

Liu

Guan

. Nanofibers of CeO₂ via an electrospinning technique. Thin Solid Films 2005; 478: 228–231.

19.

Hwang

K-S

Lin

C-A

Lin

C-H

. Preparation of high-strength and high-modulus poly(vinyl alcohol) fibers by cross-linking wet spinning/multistep drawing method. J Polym Sci Part B 1994; 52: 1181–1189.

20.

Uddin

Aoki

Gotoh

Saito

Yumura

. Fabrication of high strength PVA/SWCNT composite fibres by gel spinning. Carbon 2010; 48: 1977–1984.

21.

Minus

Chae

Kumar

. Interfacial crystallization in gel-spun poly(vinyl alcohol)/single-wall carbon nanotube composite fibers. Macromol Chem Phys 2009; 210: 1799–1808.

22.

Frankland

SJV

Caglar

Brenner

Griebel

. Molecular simulation of the influence of chemical cross-links on the shear strength of carbon nanotube-polymer interfaces. J Phys Chem B 2002; 106: 3046–3048.

23.

Cadet

Coleman

Barron

Hedicke

Blau

. Morphological and mechanical properties of carbon-nanotube-reinforced semicrystalline and amorphous polymer composites. Appl Phys Lett 2002; 81: 5123–5126.

24.

Zhang

Liu

Sreekumar

Kumar

Moore

Hauge

. Poly(vinyl alcohol)/SWCT composite film. Nano Lett 2003; 3: 1285–1288.

25.

Ruan

Gao

. Ultra-strong gel-spun UHMWPE fibers reinforced using multiwalled carbon nanotubes. Polymer 2006; 47: 1604–1611.

26.

Dalton

Collins

Munoz

Razal

Ebron

Feraris

. Super-tough carbon-nanotube fibres. Nature 2003; 423: 703–703.