Conductive properties and mechanisms of different polymers doped by carbon nanotube/polypyrrole 1D hybrid nanotubes

Abstract

We suggested a method to reduce the amount of multi-walled carbon nanotubes (MWNTs) used for fabricating polymer composites, and revealed the electron transport mechanism in polymer matrixes with different polarities. Specifically, one-dimensional (1D) hybrid nanotubes (MPPy) have been easily prepared by in situ polymerization with an appropriate ratio of pyrrole to MWNTs. Three matrixes including polyvinyl chloride (PVC), poly(methyl methacrylate) (PMMA) and polystyrene (PS) were separately blended mechanically with the MPPy nanotubes. Three families of composites (MPPy/PVC, MPPy/PMMA, MPPy/PS) achieved electrical conductivities higher than 10⁻⁵ S cm⁻¹ at 0.3 wt%, 0.8 wt% and 1.5 wt% of MWNTs. The amounts of MWNTs were an order of magnitude lower compared to bare MWNTs used as fillers. The highly ordered chain structure of PPy grown along the surface of the MWNTs might be responsible for the good performance of the MPPy nanotubes, as indicated by FESEM, X-ray photoelectron spectroscopy and Hall Effect Measurement System analysis. We combined the doping effect and tunneling distance to explore the conductive mechanism in the three matrixes. The more potent doping effect and longer tunneling distance in the MPPy/PVC families enabled a sharp improvement of carrier concentration and therefore a lower percolation threshold, compared to that of the MPPy/PMMA and MPPy/PS families.

---

1. Introduction

Polymers doped by carbon nanotubes (CNTs) with notably electrical, thermal, and mechanical properties have attracted tremendous attention in both academia and industry.^1–4 Many researchers have centered on improving the dispersion of CNTs, to produce CNTs/polymer composites with high conductivities and low CNT content in the past years.^5–7 Besides, finding new CNT-based multi-component fillers would be a new way to achieve this goal.

Up to now, chemical functionalization methods, such as covalent grafting,^8,9 doping,¹⁰ and ion beam treatment have been used to improve the intertube contact, but these methods inevitably disturbed the conjugated system of CNTs and thus impair the conductive properties of the resulting composites. Over the past few years, hybrid CNT fillers, with no destruction to the conjugated structure of the CNTs, have been synthesized to fabricate CNT/polymer composites with a reduced consumption amount of CNTs. Multi-branched polyaniline/multi-walled carbon nanotube (MWNTs) hybrids,¹¹ nano-hybrid dopants such as CNT/PPy nano-composites,¹² and CNTs coated with polyaniline¹³ have been employed to dope polymer matrixes to obtain high dielectric permittivity and low electrical percolation thresholds. Hybrid fillers of polyethylene containing 15 wt% of MWNTs have been melt mixed with the polyethylene matrix, and the composites obtained a low percolation threshold at 0.4 vol%.¹⁴ Nevertheless, these hybrid fillers are limited upon application because complex mixed processes (e.g. melt blending) were needed to improve the dispersion of CNTs. These hybrid fillers also exhibit good electrical conductivity at the cost of a large CNT content. Thus, it is necessary to synthesise new hybrid fillers with good conductivity and a low CNT content in a simple and low-cost way.

The reported electrical conductivities of assembled CNT/polymer products (e.g., fibers, films) are always much lower than that of pure CNTs, mainly due to the lack of effective electron transport pathways from tube to tube.^15–17 CNT thin films (assembled large area networks of CNTs) have been reviewed previously with most emphasis on the electrical properties of the network and their application in electronics.^18–20 It has been reported that the conformation and arrangement of polymer chains are critical to the electrical conductivity of CNT/polymer composites.²¹ An expanded chain conformation and an ordered chain arrangement would result in a reduced barrier of both interchain and intrachain hopping, and thus, enhanced electrical conductivity.^22,23 It is considered to be vital to construct an ordered polymer structure for realizing high electrical performance in conducting polymers.

CNTs are the best candidates for 1D templates for generating ordered polymer layers owing to their extremely stable well-defined 1D nanostructure and excellent electric and mechanical properties.^24,25 Some researchers have utilized self-assembly to obtain ordered polymer chains around the surface of CNTs.^26,27 CNT/poly(vinyl acetate),³ CNT/poly(3,4-ethlenedioxythiophene) poly(styrenesulfonate) composite films,²⁸ and CNT/polyaniline hybrid nanocomposites²¹ have achieved remarkably enhanced thermoelectric properties, benefiting from their ordered construction of polymers.

Among the best-known conducting polymers, PPy is considered as one of the most promising materials because of its relatively facile processability, good electrical conductivity, environment stability and low price. Moreover, MWNT/PPy hybrides with core–shell structure have been compounded and characterized by a large number of researchers. It has also been reported that π–π interactions existed between the CNT surface and PPy,^29–31 which would benefit the growth of a uniform layer along the surface of the CNTs. However, the configuration of PPy affected by the presence MWNTs, behind the phenomenon of ordered chain growth, has been largely neglected.

Our work presents the preparation of highly conductive MPPy 1D hybrid nanotubes by in situ polymerization and three families of composites by mechanical blending with PVC, PMMA and PS matrixes. The aim of this study was to investigate the change of connecting type in the PPy polymerization process as influenced by MWNTs and to explore a method to reduce the amount of MWNTs in the derived nanocomposites. In addition, the effect of tunneling distance and doping effect on the carrier concentration, carrier mobility and conductivity have been evaluated.

2. Experimental section

2.1 Preparation of PPy, MPPy

The nanocomposites were synthesized by one-step in situ polymerization of pyrrole in the presence of MWNTs with ammonium peroxidisulfate (APS) as the oxidant. The pyrrole monomer (98%, Sinopharm Chemical Reagent Co., Ltd) was purified by distillation under reduced pressure. The MWNTs (95%, Chengdu Organic Chemicals Co., Ltd) were 20–30 nm in diameter and 5–15 μm in length. A solution of 100 mL of 1 mol L⁻¹ HCl containing different masses of MWNTs (0, 0.045, 0.220, 0.450, and 0.714 g) was sonicated for 30 minutes. Another 20 minutes of sonication was needed after adding the pyrrole monomer (4.19 mL) into the solution. It was then transferred to a three-necked, round-bottomed flask in an ice bath. The 30 mL HCl solution containing APS (1.53 g, keeping the molar ratio of pyrrole 1 : 1) was slowly added dropwise into the suspension with constant magnetic stirring. After an additional 4 h reaction, the mixture was filtered and the residue was washed thoroughly with deionized water and ethanol. The cleaning procedure above was repeated three times. The obtained black powders were dried at 40 °C under a vacuum for at least 24 h. The resulting composites were named PPy, MPPy₁, MPPy₅, MPPy₁₀ and MPPy₁₅ respectively.

2.2 Polymer doping

The MPPy₁₀ and MPPy₁₅ hybrids were separately blended into the PVC (Shanghai Aladdin Chemical Co.), PMMA and PS powders by grinding them together in an agate mortar. PMMA and PS powders were synthesised in our laboratory according to a previous study.^32,33 The hybrid powders were pressed into wafer-shaped samples and ground in the agate mortar again. This procedure was repeated three times. After they were fully blended, these samples were pressed into wafers with a thickness of 2 mm and a diameter of 13 mm at a constant pressure of 20 MPa, with a pressure holding time of 1 min.

2.3 Characterization

The resistance (R) of each composite was measured by using a DC Kelvin bridge and a standard four-probe technique at room temperature, and then the DC electric conductivity (σ_DC) of the composite was calculated from the relationship σ_DC = L/RS , where L and S are the thickness and sectional area of the composites, respectively. The carrier concentration and carrier mobility were measured by a Hall effect measurement instrument (HM2000) at room temperature. The chemical structure and property of the composites were analyzed by X-ray Photoelectron Spectroscopy (Perkin-Elmer PHI 5000C ESCA System) using non-mono chromatized Mg-Kα radiation for excitation. Field emission scanning electron microscopy (FESEM) imaging was performed with a JSM6700F (JEOL) field emission scanning electron microscope. Dielectric properties were measured on a broadband dielectric spectrometer (Novocontrol Concept 80, Hundsangen, Germany) at room temperature over a frequency ranging from 10⁴ to 10⁵ Hz.

3. Results and discussion

3.1 MPPy hybrid nanotubes

Fig. 1(a) shows the microstructures of pure MWNTs. The outer diameter of the MWNTs is about 25–30 nm. Fig. 1(b)–(e) show the microstructures of MPPy₁, MPPy₅, MPPy₁₀ and MPPy₁₅, respectively, at the same magnification. Fig. 1(f) is the amplification of Fig. 1(e), and shows that the tubular structure of MPPy₁₅ more defined. As shown in Fig. 1(b) and (c), MPPy₁ and MPPy₅ consist of micrometer-sized spherical particles and no tubular structures can be found. However, the bulk morphology of MPPy₁₀ changes from granular particulates to a mat of 1D nanotubes with an average diameter of 60–65 nm (see Fig. 1(d)). MPPy₁₅ exhibited a more defined tubular structure at the same magnification, as shown in Fig. 1(e). This idealized uniform tubular structure of MPPy₁₅ could be observed clearly in Fig. 1(f) with a greater magnification. These results indicate that the MWNTs acted as templates for the polymerization of Py molecules to form homogeneous layers.

Fig. 1 FESEM images of MWNTs (a), MPPy₁ (1 wt% MWNT/PPy) (b), MPPy₅ (5 wt% MWNT/PPy) (c), MPPy₁₀ (10 wt% MWNTs/PPy) (d) and MPPy₁₅ (15 wt% MWNTs/PPy) (e). (f) An image of MPPy₁₅ (15 wt% MWNTs/PPy) with a greater magnification.

The electrical conductivity of the MPPy complex starts with a flat increase and tends to rise sharply from 10 wt% to 15 wt% (see Fig. 2). The enhancement of conductivity of MPPy may come from two parts: one is mainly attributed to the incorporation of MWNTs, and the other may come from the ordered chain packing of PPy, which was confirmed as follows by XPS. Accordingly, the phase when the 1D tube structures have formed (10 wt% to 15 wt% of MWNTs) shows a more precipitous rise in conductivity compared with a relative mild increase when the MWNT content is 0 wt% to 10 wt%.

Fig. 2 Electrical conductivity of MPPy with different MWNT contents.

Previous studies indicated that there was π–π interaction between the MWNTs and PPy in the MWNT/PPy composites.^30,34 This interaction was confirmed more clearly in the 1D MPPy hybrid nanotubes by XPS results. The main peak of the C1s spectrum of pure PPy was located at 285.83 eV, but it shifted down to 285.44 eV when doped by MWNTs. This indicated that the conjugated system of the PPy layer was affected by the π-electrons of the MWNTs. This interaction may be favorable for the uniform coating of PPy on the surface of the MWNTs.

In order to explore the configuration of the PPy layer, we fitted the PPy spectrum with three chemical components (Fig. 3). The two peaks with lower electron binding energies correspond to the α and β type carbon atoms of the pyrrole monomer.^25,35 The wider peaks located at 288.3 eV probably came from the disordered carbon structures (e.g., C at terminal groups) in the pyrrole rings. The areas of α-C and β-C in Fig. 3 represent the content of carbon atoms with C–H bonds, which have not participated in the construction of polymer chains. In MPPy₁₅, the ratio of S_α/S_β was reduced by 0.3 with respect to pure PPy. The content of α-C with C–H bonds decreased, meaning that more α-C had participated the construction of polymer chains. So, the Py monomers were induced to connect in an α–α order in the presence of MWNTs. This order significantly promoted a more planar configuration of PPy, which would be in favor of electrical transport between PPy and the MWNTs, and guaranteed high electrical conductivity of 1D MPPy₁₀ and MPPy₁₅ hybrid nanotubes.

Fig. 3 C1s spectra of the samples: (a) PPy (b) MPPy₁₅. The α and β peaks represent the α type and β type of carbon atoms of the pyrrole monomer.^25,35

3.2 MPPy reinforcement of polymers

The dispersions of the MPPy nanotubes into the polymer matrixes were evaluated by FESEM (see Fig. 4(a)–(c) with 10 wt% MPPy₁₅ nanotubes in the three matrixes). We could see a similar dispersion of the MPPy₁₅ hybrid nanotubes in Fig. 4(a)–(c), so the dispersion of the MPPy₁₅ nanotubes is not the main factor to impact the distinct electrical conductivity of composites.

Fig. 4 FESEM images of (a) MPPy/PS, (b) MPPy/PMMA and (c) MPPy/PVC composites with 10 wt% of MPPy₁₅.

The electrical conductivity over the three families of composites (PS, PMMA and PVC were used as the matrixes) is plotted in Fig. 5, as a function of the MWNT content. MPPy₁₀ and MPPy₁₅ hybrid nanotubes were respectively used as fillers in Fig. 5(a) and (b). In the MPPy₁₀/polymer composites, the electrical percolation thresholds appeared at 0.3 wt%, 0.8 wt%, 1.5 wt% respectively in the PVC, PMMA, PS matrixes. The three curves in Fig. 5(b) show similar trends and percolation thresholds, but the MPPy₁₅/polymer composites process better conductivity at their corresponding percolation thresholds. This could be attributed to the better tubular structure of MPPy₁₅, as shown in Fig. 1. In contrast, the electrical percolation thresholds of the MWNT/polymer composites appeared at 5 wt%, 10 wt%, 10 wt% in the PVC, PMMA, PS matrixes, respectively (Fig. 5(c)). Apparently, the dosage of MWNTs can be greatly reduced with an order of magnitude when using MPPy hybrid nanotubes as fillers.

Fig. 5 Electrical conductivity of composites based on PS, PMMA and PVC matrixes versus the mass fraction of MWNTs, (a–c) were MPPy₁₀, MPPy₁₅ and MWNTs used as fillers, respectively.

The three curves in Fig. 6(a) and (b), corresponding to the MPPy₁₅/PS, MPPy₁₅/PMMA and MPPy₁₅/PVC composites, represent the carrier concentration and mobility of composites with different mass fractions of MPPy₁₅, respectively. All three carrier concentration curves possess two phases: first a sharp increase and then a plateau. Obviously, the three families of composites reach their inflexion with different speeds. A sharp increase of carrier concentration with 8 orders of magnitude is observed in the PVC matrix when the doping content was only 5%. In the PMMA and PS matrixes, the carrier concentration curves finish their first phase of increase later at doping contents of 10% and 20%, respectively. After the doping contents exceed their inflexions, the carrier concentration remains almost unchanged, but their carrier mobility starts to increase (see Fig. 6(b)). With the doping content increased to 30%, the carrier mobility in the PS, PMMA, PVC matrixes reached 109.6, 466.3, 810.1 cm⁻² s⁻¹ V⁻¹, respectively.

Fig. 6 Relationships between the carrier concentration and MPPy₁₅ mass fraction (a), and the carrier mobility and MPPy₁₅ mass fraction (b) of samples.

The dielectric permittivity of composites had discrepant trends with the doping content of the MPPy₁₅ hybrid (see Fig. 7). In the PVC matrix, with a small addition of MPPy₁₅ nanotubes, the dielectric permittivity of the resultant composites sharply increased, different from those in the PS and PMMA matrixes with relatively languid uptrends. Specifically, the dielectric permittivity at 10² Hz of the MPPy₁₅/PVC composite had already reached 504 with a MPPy₁₅ hybrid content of 10%, while the dielectric consistents of MPPy₁₀/PMMA and MPPy₁₀/PS were merely 23.8 and 5.06. The dielectric properties of the three families are consistent with the electrical conductivity.

Fig. 7 Relationship between the dielectric permittivities of the composites and the mass fraction of the MPPy₁₅ hybrid nanotubes.

3.3 Interaction between MPPy and the matrixes

As the above-mentioned results demonstrated, matrixes with different polarities resulted in discrepant percolation thresholds. To investigate the influence of interaction between the MPPy hybrid nanotubes and the matrixes on the electrical percolation thresholds, we compared the C1s spectra of pure matrixes with the resultant composites [w(MPPy₁₅ hybrid nanotubes) = 30%] (see Fig. 8(a)–(c)). Table 1 lists the shift of the main peak position distinctly. The main peaks of the resultant composites moved to the direction of lower binding energy relative to the pure matrixes, except for the MPPy₁₅/PS composite, which shifted to the opposite direction. The range of main peak shifts is also different: the MPPy₁₅/PVC composites display a remarkable shift of the main peak of 0.9 eV, compared with the MPPy₁₅/PMMA composites with a moderate shift of 0.4 eV and a weak shift of 0.15 eV in the MPPy₁₅/PS composites. Larger shifts may imply relatively higher degrees of doping effect. Regarding the interaction between PS and the MPPy₁₅ hybrid nanotubes, the phenyl group acted as an electron-donator to the hybrid nanotube, which is consistent with Samarajeewa’s research.³⁶ PMMA and PVC possess stronger electronegativity with the ester group and chlorine group respectively, leading to the π electron cloud shift to the matrixes, and a downshift of the binding energy in the MPPy₁₅/PMMA and MPPy₁₅/PVC composites. The XPS results indicated that the level of doping effect between the MPPy hybrid nanotubes and the matrixes were varied for matrixes with different polarities.

Fig. 8 C1s spectra of the pure matrixes and the resultant composites with 30 wt% of MPPy₁₅ hybrid nanotubes.

Table 1 The main peak positions of the samples

Matrix	BE (eV) (wt% = 0%)	BE (eV) (wt% = 30%)	ΔBE (eV)
PS	285.5	285.65	0.15
PMMA	286.2	285.80	−0.4
PVC	286.0	285.10	−0.9

---

3.4 Conductive mechanism analysis

As stated above, our work reported herein has two targets: one is to prepare highly conductive MPPy/polymer composites with reduced amounts of MWNTs, and the other is to reveal the origin of the different percolation thresholds in the three matrixes. Taking all the results into consideration, including the structure and properties of the hybrid nanotubes, doping effect and the discrepant percolation thresholds, we deduced the whole conductive mechanism in Fig. 9.

Fig. 9 The conductive mechanism for different polymer composites in the entire fabrication process.

MPPy hybrid nanotubes with a 1D conjugate tubular structure were prepared by in situ polymerization (Fig. 9(a) and (b)). This expanded chain conformation and the highly ordered PPy layer endowed the hybrid nanotubes with high electrical properties. Mechanical blending of the MPPy hybrid nanotubes and matrixes (PS, PMMA and PVC) yielded three families of composites (c–e). The content of MWNTs needed to reach percolation threshold was an order of magnitude lower compared with that of MWNT/polymer composites. In addition, the three matrixes with different polarities, obtained variant percolation thresholds.

Kyrylyuk and coworker³⁷ reported that the tunneling distance could be estimated based on the time-independent Schrodinger equation:

Where, ℏ is the reduced Planck constant, m_e is the electron mass, E_f and E_e are the energies of a conduction electron on a tube and in the matrixes, respectively. For E_f and E_e, we presume the estimated Fermi energy of an electron on a bare MPPy hybrid nanotube and the Born energy of an electron in a matrix to be the relevant energy scale.³⁷ The equation of Born energy is given in eqn (2), where l_B is the Bjerrum length, which is the inverse of the dielectric constant of the polymer matrix, and Λ is the thermal wavelength of the spatially delocalized electrons in the matrixes.³⁸

For the three matrixes PS, PMMA, and PVC, the relative dielectric permittivities are 3.3, 4.5 and 7.8, respectively. The calculated Born energies of the pure matrixes are: E_PVC = 0.288 eV, E_PMMA = 0.5 eV, E_PS = 0.68 eV. As the same MPPy nanotubes were used as fillers, E_f keeps a constant value in the three families of composites. So the tunneling distances keep the order: ξ_PVC > ξ_PMMA > ξ_PS. On the other hand, the doping effect represents the interaction between the MPPy hybrid nanotube and matrix, which did favor the connectedness of the MPPy hybrid nanotubes. The XPS results manifest that the doping effect of the MPPy/PVC families was stronger than that of the MPPy/PMMA and MPPy/PS families.

Combining the tunneling distances and doping effect, we could sum up that in the PVC matrix with a doping of 3 wt% MPPy nanotubes, the distance between the two nanotubes had reached the tunneling distance, and the strong doping effect enabled a global connection of the MPPy nanotubes (see Fig. 9(e)). So the carrier concentration was largely improved with 8 orders of magnitude and the electrical percolation threshold appeared. For the MPPy/PMMA and MPPy/PS families, with short tunneling distances and poor doping effects, the MPPy₁₅ hybrid nanotubes in the PMMA and PS matrixes were connected at a more intensive scale (see Fig. 9(d) and (c)), which resulted in a gentle increase of the carrier concentration and higher percolation thresholds. After the doping contents passed their percolation thresholds, all the MPPy nanotubes in the three matrixes were connected and the conductive networks were formed, so platforms of the carrier concentration appeared in the three families and the carrier mobility began to rise. The level of doping effect affects the increment of the carrier mobility and therefore the electrical conductivity in the second phase.

The situation changed when it came to the MWNT/polymer system, where the percolation thresholds of the three matrixes didn’t show such obvious differences, as shown in Fig. 5(c). This may attributed to the poor doping effect between the MWNTs and the three matrixes. The connectedness criterions of the MWNTs in the three matrixes were similar. So these three curves corresponding to the electrical conductivity of the three families coincided when the MWNTs were used as fillers.

4. Conclusion

A study associated with 1D MPPy hybrid nanotubes and the electrical properties of the derived nanocomposites was carried out in this work. MWNTs acted as 1D hard templates for generating an ordered PPy layer, as confirmed by SEM and XPS results. A more planar configuration of PPy, influenced by MWNTs, may be in favor of good performance of MPPy in terms of electrical conductivity. The amount of MWNTs needed to reach the threshold could be reduced by an order of magnitude but the resulting composites exhibited a similar electrical conductivity compared to using MWNTs as fillers. This finding implies that highly ordered hybrid nanotubes with a low content of MWNTs could be used as fillers to reduce the consumption of MWNTs.

The doping effects and tunneling distance were combined to clarify the diversity of the percolation threshold in the three matrixes. It may useful to develop highly conductive CNT/polymer composites by an optimized choice of matrix with a higher dielectric permittivity and a stronger doping effect with fillers.

Acknowledgements

The authors thank East China Normal University for financial support. The authors also thank Zhao-yang Zhang for technical support.

Notes and references

1. T. Sekitani, Y. Noguchi, K. Hata, T. Fukushima, T. Aida and T. Someya, Science, 2008, 321, 1468–1472.
2. L. Larrimore, S. Nad, X. Zhou, H. Abruna and P. L. McEuen, Nano Lett., 2006, 6, 1329–1333.
3. C. Yu, Y. S. Kim, D. Kim and J. C. Grunlan, Nano Lett., 2008, 8, 4428–4432.
4. M. M. Hamedi, A. Hajian, A. B. Fall, K. Hakansson, M. Salajkova, F. Lundell, L. Wagberg and L. A. Berglund, ACS Nano, 2014, 8, 2467–2476.
5. T. McNally, P. Potschke, P. Halley, M. Murphy, D. Martin, S. E. J. Bell, G. P. Brennan, D. Bein, P. Lemoine and J. P. Quinn, Polymer, 2005, 46, 8222–8232.
6. O. Breuer and U. Sundararaj, Polym. Compos., 2004, 25, 630–645.
7. J. K. W. Sandler, J. E. Kirk, I. A. Kinloch, M. S. P. Shaffer and A. H. Windle, Polymer, 2003, 44, 5893–5899.
8. J. Zhu, B. S. Shim, M. Di Prima and N. A. J. Kotov, J. Am. Chem. Soc., 2011, 133, 7450–7460.
9. S. N. Kourkouli, A. Siokou, A. A. Stefopoulos, F. Ravani, T. Plocke, M. Müller, J. Maultzsch, C. Thomsen, K. Papagelis and J. K. Kallitsis, Macromolecules, 2013, 46, 2590–2598.
10. Y. Zhao, J. Wei, R. Vajtai, P. M. Ajayan and E. V. Barrera, Sci. Rep., 2011, 10, 1038–1052.
11. Z. Qiang, G. Liang, A. Gu and L. Yuan, Composites, Part A, 2014, 64, 1–10.
12. B. Yu and X. Xu, RSC Adv., 2014, 10, 2–8.
13. T. Zhou, J.-W. Zha, Y. Hou, D. Wang, J. Zhao and Z.-M. Dang, ACS Appl. Mater. Interfaces, 2011, 3, 4557–4560.
14. M. Jouni, J. Faure-Vincent, P. Fedorko, D. Djurado, G. Boiteux and V. Massardier, Carbon, 2014, 76, 10–18.
15. M. A. Topinka, M. W. Rowell, D. Goldhaber-Gordon, M. D. McGehee, D. S. Hecht and G. Gruner, Nano Lett., 2009, 9, 1866–1871.
16. L. Hu, D. S. Hecht and G. Gruener, Chem. Rev., 2010, 110, 5790–5844.
17. P. N. Nirmalraj, P. E. Lyons, S. De, J. N. Coleman and J. J. Boland, Nano Lett., 2009, 9, 3890–3895.
18. C. M. Jiang, A. Saha, C. C. Young, D. P. Hashim, C. E. Ramirez, P. M. Ajayan, M. Pasquali and A. A. Marti, ACS Nano, 2014, 8, 9107–9112.
19. C. Jiang, A. Saha, C. Xiang, C. C. Young, J. M. Tour, M. Pasquali and A. A. Marti, ACS Nano, 2013, 7, 4503–4510.
20. A. Saha, C. Jiang and A. A. Marti, Carbon, 2014, 79, 1–18.
21. Q. Yao, L. Chen, W. Zhang, S. Liu and X. Chen, ACS Nano, 2010, 4, 2445–2451.
22. Q. Yao, L. D. Chen, X. C. Xu and C. F. Wang, Chem. Lett., 2005, 34, 522–523.
23. Y. Hiroshige, M. Ookawa and N. Toshima, Synth. Met., 2006, 156, 1341–1347.
24. Y. Zhang and H. J. Dai, Appl. Phys. Lett., 2000, 77, 3015–3017.
25. P. M. Ajayan, O. Stephan, P. Redlich and C. Colliex, Nature, 1995, 375, 564–567.
26. M. J. O’Connell, P. Boul, L. M. Ericson, C. Huffman, Y. H. Wang, E. Haroz, C. Kuper, J. Tour, K. D. Ausman and R. E. Smalley, Chem. Phys. Lett., 2001, 342, 265–271.
27. F. Balavoine, P. Schultz, C. Richard, V. Mallouh, T. W. Ebbesen and C. Mioskowski, Angew. Chem., Int. Ed., 1999, 38, 1912–1915.
28. D. Kim, Y. Kim, K. Choi, J. C. Grunlan and C. Yu, ACS Nano, 2010, 4, 513–523.
29. J. Liu and H.-Q. J. Yu, J. Electron. Mater., 2014, 43, 1181–1187.
30. Y. Yu, C. Ouyang, Y. Gao, Z. Si, W. Chen, Z. Wang and G. J. Xue, J. Polym. Sci., Part A: Polym. Chem., 2005, 43, 6105–6115.
31. C. Bin, L. Jun, L. Delzeit and P. Qibing, Proc. SPIE-Int. Soc. Opt. Eng., 2007, 6423, 1–8.
32. J.-M. Thomassin, M. Trifkovic, W. Alkarmo, C. Detrembleur, C. Jérôme and C. Macosko, Macromolecules, 2014, 47, 2149–2155.
33. R. Xu and X. Xu, RSC Adv., 2014, 4, 42226–42233.
34. N. G. Sahoo, Y. C. Jung, H. H. So and J. W. Cho, Synth. Met., 2007, 157, 374–379.
35. R. D. Sherman, L. M. Middleman and S. M. Jacobs, Polym. Eng. Sci., 1983, 23, 36–46.
36. D. R. Samarajeewa, G. R. Dieckmann, S. O. Nielsen and I. H. Musselman, Carbon, 2013, 57, 88–98.
37. A. V. Kyrylyuk and P. P. van der Schoot, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 8221–8226.
38. J. N. Israelachvili, Intermolecular and Surface Forces, Academic, London, 1985.

---