Fabricating conductive poly(ethylene terephthalate) nonwoven fabrics using an aqueous dispersion of reduced graphene oxide as a sheet dyestuff

Abstract

A simple one-step dyeing-like approach was presented for fabricating conductive poly(ethylene terephthalate) (PET) nonwoven fabrics using reduced graphene oxide (rGO) as a sheet dyestuff. Using polyurethane (PU) as a middle adhesive layer, the rGO can be adsorbed and immobilized on the surface of the PET fabrics due to the strong attraction between the two adjacent components. As a result, conductive fabrics with high structural stability were successfully prepared in an aqueous dispersion of rGO. The as-prepared composite fabrics were characterized using scanning electron microscopy, Fourier transform infrared spectroscopy, Raman spectroscopy, and thermogravimetric analysis. The effects of the weight fraction of the rGO on the morphology, structure and electrical properties of the composite fabrics were discussed. These composite fabrics exhibited a rather low electrical percolation threshold due to the homogeneous coverage of graphene sheets on the surface of the PET nonwoven fabrics, and the highest conductivity was about 2.0 × 10⁻⁵ S sq⁻¹, ensuring feasibility to manufacture efficient heating elements. Importantly, the direct use of an aqueous dispersion of rGO in dyeing opens up new possibilities for the mass production of graphene-based innovative fabrics.

---

1. Introduction

The importance of conductive textiles made by integrating electronic functionality into textile structures has been increasing because of their various applications in many fields, such as the areas of smart and e-textiles, chemical sensing, wearable electronics and energy conversion and storage.^1–4 Their potential application as biomedical devices is also expected, especially as heating elements in medical fields such as electrotherapy treatment, and as medical blankets for maintaining the body temperature of patients.⁵ Conductive textiles are textile materials combined with electrical materials through different methods such as polymerization, metal wire insertion and coating techniques.^1,6,7 Coating techniques have become attractive in the field of conductive textiles, because textile materials with a large surface area possess several advantages over conventional materials, such as high flexibility and mechanical properties, making them good light-weight substrates to deposit some functional materials for different applications. Moreover, the process of coating is relatively simple, cheap and materials obtained maintain their mechanical properties, along with electrical properties.^6–9

Graphene, a promising versatile nanofiller for polymers, is defined as a single layer of sp²-bonded carbon atoms arranged in a honeycomb lattice and can be produced in large quantity at low cost by chemical conversion from graphite.^1,10–12 In particular, because it is highly conductive, has a large surface area, and is bendable, graphene has been considered as an excellent candidate for flexible conductive platforms.^13,14 As an intermediate in the manufacture of graphene, graphene oxide (GO) has structural features similar to common textile dyes, such as the presence of an abundant number of hydroxyl and carboxyl groups.¹⁰ Thus their homogeneous colloidal suspensions can be easily obtained without the need for either polymeric or surfactant stabilizers, which are widely used for carbon blacks or carbon nanotubes.^1,5,15,16 An approach similar to the common dyeing and finishing method has been performed for fabricating conductive fabrics by directly placing fabrics such as polyarylate,¹⁷ cotton^6,7,18 and polyester^8,9 in a solution containing the GO as a dyestuff. It should be pointed out that this work was performed in both aqueous and polar organic solvents without the need for foreign stabilizers or chemical modification.^12,19 However, GO is an electrically insulating material due to its disrupted sp² bonding networks.¹² In order to obtain conductive fabrics, further reduction of GO to reduced graphene oxide (rGO) must be carried out at a relatively high temperature.^10,20 It is noted that rGO consists of electronically conductive graphitic sheets which are a few layers thick, making it suitable for the synthesis of conducting nanocomposites.^10,12 Unlike perfect graphene sheets which tend to aggregate with each other and are hard to disperse in aqueous solution, rGO with a considerable number of sheet defects exhibit relative stability in an aqueous dispersion, and have the versatility of being amenable to a range of solution processing techniques.¹⁹

Poly(ethylene terephthalate) (PET) is one of the most widely used, low-cost polymers due to its excellent thermal and chemical resistance and mechanical performance, and PET nonwoven fabrics are one of the most important fabrics. The above mentioned properties of PET film are encouraging for its application as a structural material. Although local electrostatic interactions between graphene and PET can occur,²¹ these interactions cannot withstand subsequent washing due to poor fixing of the graphene sheets on the textile surface. Improved bonding of rGO with textile surfaces not only increases durability, but also opens up new possibilities for the mass production of graphene-based innovative fabrics. In this work, a low-cost approach for dyeing PET nonwoven fabrics by anchoring rGO on the surface of fabrics in an aqueous system is presented.

2. Materials and methods

2.1 Preparation of an aqueous dispersion of rGO

rGO was obtained by chemical reduction of readily available exfoliated GO with N₂H₄·H₂O, which was synthesized by a modified Hummers method.^1,6 In a typical procedure, 10.0 g of graphite and 7.5 g of NaNO₃ powder were added to 230 ml of cooled (0 °C) H₂SO₄ (98%). 30.0 g of KMnO₄ was added gradually with stirring at 35 °C for 45 min. 460 ml of distilled water was then slowly added to the mixture. After 15 min, the reaction was terminated by adding 1400 ml of distilled water followed by 100 ml of 30% H₂O₂ solution. Therefore, GO was obtained by washing the suspension with deionized water until the pH reached 7. Afterwards, 1.0 g GO was dispersed in 1000 ml deionized water, and subsequently 1 ml hydrazine hydrate was added. The reaction was kept at 98 °C with mechanical stirring for 1 h, and the black product was filtered with a Buchner funnel and washed with deionized water until the pH reached 7. Only a small fraction of the as-prepared rGO was dried in the oven at 60 °C for 24 h for comparison, and most of it was dispersed again in deionized water under ultrasonication to prepare an aqueous suspension containing weight fractions of the rGO ranging from 0.001 to 0.150 wt%.

2.2 Preparation of conductive PET fabrics

PET nonwoven fabrics were firstly immersed in 0.1 wt% PU in DMF at room temperature for 60 s. After drying, the fabrics were washed with deionized water to remove residual DMF, and then dried under ambient conditions overnight. As shown in Fig. 1, the as-prepared composite fabrics were immersed in the above-mentioned aqueous suspension with a liquor ratio of 100 : 1. The fabric dyeing was carried out in a bath of ice and water under ultrasonication for 10 min. After drying at 80 °C in an oven overnight, dark-brown PET composite fabrics were obtained.

Fig. 1 Schematic for the fabrication of rGO-coated PET nonwoven fabrics.

2.3 Apparatus

The morphologies of the PET nonwoven fabrics and as-prepared composite fabrics were observed by field emission scanning electron microscopy (FE-SEM) (S-3000N, Hitachi, Japan). The morphologies of the rGO samples were observed by Atomic Force Microscopy (AFM) (Veeco, Santa Barbara, CA), using a NanoScope IV Multimode microscope equipped with an E scanner and NSC11/AIBS cantilevers (MikroMasch, Tallinn, Estonia). The samples were prepared by casting a heavily diluted aqueous rGO suspension on the surfaces of silicon chips. The images were obtained using the tapping mode at a scanning rate of 1 Hz, and a cantilever with a spring constant of 40 N m⁻¹ was used.

The chemical structures were characterized by Fourier transform infrared (FT-IR) spectroscopy on an attenuated total reflectance accessory of the spectrometer (Nicolet 8700, American) and a micro-Raman spectrometer (inVia+Reflex, Renishaw, England). FT-IR spectra were obtained with a resolution of 4 cm⁻¹, a scan number of 200 and a scan area from 400 to 4000 cm⁻¹. Raman spectra were recorded using the 532 nm line of an argon ion laser with 50 mW output power.

Thermal stability was tested using a thermo-gravimetric analyzer (TG209F1, Netzsch, Germany) with a heating rate of 20 °C min⁻¹ under a nitrogen atmosphere. The structural stability of the composite fabrics was evaluated by treating the fabrics in water by ultrasonication at different powers, and stirring at different rotation rates.

Electrical measurements were carried out by a two-electrode method, applying a voltage of 100 V using an Insulation Resistance Meter (BH828, China). The dimensions of the applied fabrics were 3 × 3 cm². The surface conductivity of the composite fabric was calculated using the equation, σ = D/(R_sL), where D is the distance between the two electrodes, R_s is the resistance of the sample, and L is the width of the electrode. Heat generation characteristics were determined using a DC power supply (WYJ, 3 A, 30 V). A thermometer was inserted into six fabric samples (5 × 5 cm²) and positioned half way, and then the edges of the fabrics were regulated by two cramps and connected to the terminals of the power supply.

3. Results and discussion

3.1 Morphology

Using the dyeing-like process, the surfaces of the PET nonwoven fabrics were covered successively with rGO at different weight fractions in an aqueous dispersion. As shown in Fig. 1, the as-prepared black composite fabrics maintained the excellent mechanical properties of the PET nonwoven fabrics. Furthermore, AFM was used to evaluate the morphologies of the resultant rGO. The apparent height for the observed rGO with different geometric shapes was found to be around 3 nm as shown in Fig. 1, indicating that the rGO consisted of three-layer graphene sheets, and the width of most of the rGO was more than 10 micrometers. Due to partial removal of the hydroxyl and carboxyl groups, rGO is easily exfoliated in water forming stable colloidal dispersions. In order to achieve uniform coverage of the rGO on the surface of the fabrics, the preparation process was performed under ultrasonication. Herein, PET nonwoven fabrics were treated with PU solution in DMF before dyeing, ensuring that the rGO could be adsorbed and immobilized on the surface of the PET fabrics due to a strong attraction between the two adjacent components.^4,16,22–24 Without this, it is very easy to remove the rGO from the surface of the fabrics through a simple washing procedure. Through a very thin PU adhesion layer, rGO as a sheet-shaped dyestuff is firmly attached on the surfaces of multifilament layers through the traditional dyeing approach.

SEM was used to observe the morphologies of the PET nonwoven fabrics and the resulting composite fabrics prepared in aqueous dispersions with different rGO content. There was almost no difference in the morphology and diameter of the PET nonwoven fabrics before and after PU treatment. Furthermore, a similar morphology was found with a low weight fraction of rGO as shown in Fig. 2, and it was difficult to identify whether the very thin sheets of two-dimensional rGO provide efficient coverage of the surfaces of the fabrics. With the increase in rGO fraction, the colour of the fabrics turned from dark-brown to black, and some aggregates of the folds developed during the deposition process could be observed on the surface of the fabrics. In particular, a larger number of rGO aggregates were commonly observed at the intersection of the fibers for larger rGO fractions.

Fig. 2 SEM images of the PET nonwoven fabric and the resulting composite fabrics prepared in aqueous dispersions with different rGO content. (a) PET nonwoven fabric; (b) 0.001 wt%; (c) 0.010 wt%; (d) 0.030 wt%; (e) 0.080 wt% and (f) 0.150 wt%.

3.2 Electrical percolation transition

Fig. 3 displays the change in surface conductivity of the composite fabrics with the increase in the weight fraction of rGO in the aqueous dispersion. Generally, when conductive nanofillers form a conductive network in an insulating matrix, the direct current electrical conductivity of the composites increases rapidly, and subsequently the electrical percolation threshold of the composites can be measured. As shown in Fig. 3, the conductivity of the composite fabrics increases sharply once the rGO fraction exceeded a threshold of about 0.0012 wt%. At a nanofiller content of about 0.010 wt%, the conductivity of the composite fabrics had already satisfied the antistatic criterion (10⁻⁶ S sq⁻¹), which is about 8 orders of magnitude higher than that of PET fabrics (2.2 × 10⁻¹⁴ S sq⁻¹). The conductivity of the composite fabrics then continued to increase gradually with the increase in nanofiller content, and reached a saturation value of 2.0 × 10⁻⁵ S sq⁻¹ when the rGO fraction reached 0.080 wt%, which is comparable with the value for the conductive polyester fabric (9.1 × 10⁻⁵ S sq⁻¹).¹⁷ It is worth noting that in contrast with the one-step fabrication performed in this work, a further reduction was involved in the preparation of conductive polyester fabrics, using GO as a dyestuff, in which the GO-dyed fabrics were again immersed in an aqueous solution containing 0.5 wt% sodium hydrosulfite at about 363 K for 30 min.¹⁷

Fig. 3 Change in the surface conductivity of the composite fabrics with the increase in the weight fraction of rGO in the initial aqueous dispersion. The insets display SEM images of the composite fabrics prepared at rGO fractions of 0.001 (I), 0.030 (II) and 0.080 wt% (III).

Furthermore, to better understand the formation of conducting pathways over the fabric, the insets in Fig. 3 give three different surface morphologies for the composite fabrics prepared with various rGO fractions in the initial aqueous dispersion. At a low rGO fraction, fewer folds appeared on the relatively smooth fabric surface as shown in Fig. 3(I), implying that most of the deposited rGO was present in a two-dimensional form. When the rGO fraction exceeded 0.001 wt%, an abrupt increase in the conductivity of the composite fabrics occurred. With more rGO deposited onto the surface of the fabrics, more folds could be found, as shown in Fig. 3(II). It was found that irreversible agglomeration took place for all rGO samples during the reduction process due to their hydrophobicity.²⁵ These wrinkled and overlapped graphene sheets linked the individual graphene sheets effectively and carried a high current density, resulting in high electrical conductivity. This folded rGO became increasingly interconnected and finally formed a continuous conducting network as shown in Fig. 3(III), and did not show an increase in conductivity with a further increase in the density of the conductive network.

The conductivity of the composite fabrics can be further rationalized in terms of a modified classical percolation theory, σ = σ₀(φ − φ_c)^t, where σ represents the conductivity of the composite, σ₀ represents the conductivity of the rGO, φ is the rGO fraction, φ_c the percolation fraction, and t is the critical exponent.^12,23–26 For a single percolation system, the critical exponent depends only on the dimensionality of the composites, and follows a power-law dependence of approximately 2 in a three-dimensional system and 1–1.3 in a two-dimensional system.^12,23,24 Using the data in Fig. 4, t was estimated to be 1.35, indicating the presence of a two-dimensional conductive network on the surface of the composite fabrics. The electrical percolation threshold was about 0.0012 wt%. Such a low electrical percolation threshold can be explained by the extraordinarily large specific surface area of the graphene sheets and the homogeneous coverage of the graphene sheets on the surface of the PET nonwoven fabrics. However, as only partial reduction of the GO occurred under the mild conditions, some oxygen-containing functional groups remained on the surface of rGO, which resulted in a low electrical conductivity.

Fig. 4 Linear fitting for the electrical percolation transition occurred for the composite fabrics prepared in aqueous dispersions with increasing rGO content.

3.3 Chemical structure

FT-IR and Raman spectra for rGO, PU, PET, PU-treated PET and the composite fabrics prepared in aqueous dispersions with different rGO content were recorded and are shown in Fig. 5(a) and (b), respectively. For rGO, all IR bands arising from oxidized groups were substantially reduced although they did not completely disappear as shown in Fig. 5(a). For rGO, besides a band around 1533 cm⁻¹, assigned to the skeletal vibration of the graphene sheets,⁸ a band around 1723 cm⁻¹ and a weak band around 1025 cm⁻¹, mainly resulting from oxygen-containing groups at the edges of the rGO,^8,10 were attributed to stretching vibrations of C=O, and stretching vibrations of C–O, respectively. When rGO was deposited on PU-treated PET fabrics, no substantial variation was observed compared with the infrared spectra of the pure PET fabric, and only a very slight decrease occurred for the different bands of PET. In addition, the baseline shift caused by rGO became more obvious with the increase in rGO content.

Fig. 5 FT-IR (a) and Raman (b) spectra for rGO, PU, PET nonwoven fabric, PU-treated fabric and the resulting composite fabrics prepared in aqueous dispersions with different rGO content.

Similarly, it was difficult to identify the PU characteristic bands in the composite fabrics. As mentioned above, it was essential to treat the PET nonwoven fabrics in the PU solution in DMF before dyeing. It was found that the treatment time had only a slight influence on the conductivity of the composite fabrics when the time exceeded 1 min. To avoid damage to the PET fabrics in DMF, treatment should be completed within 1 min to prepare the PU-treated PET fabrics. This implies that a very thin PU layer would be formed on the surface of the PET fabrics, which has almost no influence on the diameter of the fabrics observed by SEM, as mentioned previously. It has been demonstrated that strong interface adhesion can be achieved through the interaction between the carbonyl group of PET and the –NH group of PU,²⁷ and by dipolar interactions or hydrogen bonding between oxygen-containing functional groups in rGO and PU.^16,22 However, the changes in the infrared spectra of the composite fabrics induced by these interactions cannot be properly discerned, because very thin layers of both PU and rGO covered the PET fabric surface.

Raman scattering is highly sensitive to electronic structure, and has proven to be an essential tool for characterizing graphene-based materials.^11,20 As shown in Fig. 5(b), the rGO exhibits strong D and G bands, and a weak and broad 2D band. The G band corresponds to an E_2g mode of graphite which is related to the vibration of sp²-bonded carbon atoms in a two-dimensional hexagonal lattice (at 1575 cm⁻¹), the D band (at 1343 cm⁻¹) is caused by defects and disorder in the hexagonal graphitic layers, while the 2D band is attributed to a double-resonance process resulting in the production of two phonons with opposite momentum.^6,11,28 The D band, which is defect-related, appears as a strong band for a thick graphite sample with a large number of layers. The ratio (I_G/I_D) of the intensities of the G and D bands is related to the sheet thickness of few-layer graphene samples,¹¹ and a relatively low intensity ratio between the G peak and D peak reveals that most of the oxygen-containing functional groups are removed and a more efficient reduction of the GO.^20,25 For the rGO used in this work, the ratio was about 0.84. According to the equation, L_a = 4.4(I_G/I_D),¹¹ the rGO sheet thickness (L_a) was found to be 3.70 nm, which is consistent with that determined by AFM. Moreover, similar to the infrared spectral observations, all of the characteristic bands for PET appeared in the spectra of the composite fabrics, but no obvious PU bands could be identified. In addition, with the increase in rGO fraction in the dispersion, an increase in the intensity of the rGO bands was noticed, indicating more rGO deposition on the surface of the PET fabrics.

3.4 Structural stability

Fig. 6(a) shows TGA curves of rGO, PU, PET, PU-treated PET, and the composite fabrics prepared in aqueous dispersions with rGO content of 0.010, 0.030 and 0.080 wt%. The TGA curve of rGO shows that there is slight weight loss below 100 °C, indicating the removal of oxygen-containing functional groups as well as water.²⁵ Furthermore, weight loss is observed in the TGA curve from 150 to 300 °C, possibly due to the loss of CO and CO₂ from the decomposition of oxygen-containing functional groups.²⁸ The initial decomposition temperature (T₀) and maximum decomposition temperature (T_max) for PET nonwoven fabric were about 380 and 450 °C, respectively. Compared with neat PET, the composite fabrics exhibited a slight decrease in thermal stability, but were clearly superior to the PU-treated PET fabric. Furthermore, the composite fabric with an rGO content of 0.030 wt% showed the best thermal stability, and the highest T_max appeared for the composite fabric with a 0.080 wt% rGO content. Usually, the weight fraction for the deposited rGO can be estimated from TGA curves by analyzing the residual weight for different rGO fractions at high temperature. Unfortunately, the actual weight fraction of the rGO could not be determined in a small error range due to the extremely low rGO content on the surface of the fabrics.

Fig. 6 Evaluation of the thermal stability (a) for the composite fabrics prepared in aqueous dispersions with rGO content of 0.010, 0.030 and 0.080 wt%, and structural stability (b and c) for an rGO content of 0.080 wt%. (b) Ultrasonication, (c) stirring (R/R₀: ratio of the resistance for the treated/untreated composite fabrics).

It is important for the composite fabrics to ensure long-term structural and electrical stability in practical applications. To evaluate the structural stability, the composite fabrics prepared with an rGO content of 0.080 wt% were treated with deionized water under ultrasonication at different powers, and washed with water at 40 °C through mechanical stirring at different speeds, respectively. Fig. 6(b) shows only a slight change in the conductivity of the composite fabrics under ultrasonic irradiation with a power of 50 W. When the power was changed to 275 W, the change in the relative resistance increased roughly 5 fold, and then remained stable after 5 cycles. Fig. 6(c) displays the effect of the stirring speed on the change in the relative resistance. It was found that a large change in the relative resistance occurred at a high stirring speed before washing 5 times with water, and then remained almost unchanged. These tests demonstrated that such thin rGO-skin functioning as electrically conductive layers had excellent adhesion, flexibility, and durability.

3.5 Heat generation

By controlling the weight fraction of rGO in aqueous dispersion, a wide range of electrical conductivities for the composite fabrics could be obtained. It has been reported that fabrics with a conductivity of 10⁻⁹ S sq⁻¹ can be expected to be used as brushes in photocopying machines, and with 10⁻⁵ S sq⁻¹ for anti-static clothing, and as personal heating garments for generating sufficient heat to warm up the body under a relatively low applied voltage.¹⁷ Fig. 7 gives the electrothermal performances for the composite fabrics with a valid heating area of 5 × 5 cm² under ambient conditions. It is found that a corresponding exponential rise in temperature occurs over time until a steady-state temperature is reached, which can increase when a larger voltage is applied. For the composite fabrics prepared in an aqueous dispersion containing 0.080 wt% rGO, a steady-state temperature of 59 °C can be reached in less than 30 min at 30 V. These results demonstrate the feasibility of efficiently fabricating heating elements using these composite fabrics for maintaining patient body temperature. What is more, the direct use of rGO as a sheet dyestuff in an aqueous dispersion opens up new applications of nanotechnologies in fiber, fabric or textile production. For example, the interaction between graphene and metal oxide nanocrystals or conducting polymers provides hybrids with additional properties, offering rich opportunities for the composite fabrics to tune the materials’ structure and properties.^11,13,14,29

Fig. 7 Time dependence of temperature for the composite fabrics prepared in an aqueous dispersion with an rGO content of 0.080 wt%.

4. Conclusion

Conductive PET nonwoven fabrics were successively prepared through a conventional dyeing approach by utilizing rGO in an aqueous dispersion as a sheet dyestuff. The use of a thin PU adhesion layer enabled large, uniform coverage of rGO and structural stability of the composite fabrics. By controlling the weight fraction of rGO in the initial aqueous dispersion, a wide range of electrical conductivities, from 10⁻¹³ to 10⁻⁵ S sq⁻¹, for the composite fabrics could be obtained. With the increase in rGO fraction, a continuous interconnected network was formed over the surface of the PET nonwoven fabrics with a low percolation fraction for the rGO. Moreover, the feasibility of using the composite fabrics as efficient elements for thermal generation was demonstrated. Our findings make it possible to process graphene materials in a high volume using low-cost solution processing techniques for the preparation of flexible conductive textiles. Possible practical applications using interconnected conductive networks as the functional elements for heat generation have been evaluated. What is more, flexible conductive platforms which can provide a support for metal oxides and conductive polymers can be manufactured at very low cost and integrated into smart textiles.

Acknowledgements

This work has been financially supported by the National Natural Science Foundation of China (21274019), Fundamental Research Funds for the Central Universities (2232013A3-02), and the Program for Changjiang Scholars and Innovative Research Team in the University (IRT1221).

References

1. L. Hu and Y. Cui, Energy Environ. Sci., 2012, 5, 6423–6435.
2. C. Xiang, W. Lu, Y. Zhu, Z. Sun, Z. Yan, C. Hwang and J. M. Tour, ACS Appl. Mater. Interfaces, 2012, 4, 131–136.
3. G. Yu, L. Hu, M. Vosgueritchian, H. Wang, X. Xie, J. R. McDonough, X. Cui, Y. Cui and Z. Bao, Nano Lett., 2011, 11, 2905–2911.
4. Z. Tai, X. Yan and Q. Xue, J. Power Sources, 2012, 213, 350–357.
5. B. Fugetsu, E. Akiba, M. Hachiya and M. Endo, Carbon, 2009, 47, 527–544.
6. W. Liu, X. Yan, J. Lang, C. Peng and Q. Xue, J. Mater. Chem., 2012, 22, 17245–17253.
7. L. Hu, M. Pasta, F. La Mantia, L. Cui, S. Jeong, H. D. Deshazer, J. K. Choi, S. M. Han and Y. Cui, Nano Lett., 2010, 10, 708–714.
8. J. Molina, J. Fernández, J. C. Inés, A. I. del Ríom, J. Bonastre and F. Cases, Electrochim. Acta, 2013, 93, 44–52.
9. J. Molina, J. Fernández, A. I. del Río, J. Bonastre and F. Cases, Appl. Surf. Sci., 2013, 279, 46–54.
10. S. Park and R. S. Ruoff, Nat. Nanotechnol., 2009, 4, 217–224.
11. C. N. R. Rao, K. Biswas, K. S. Subrahmanyam and A. Govindaraj, J. Mater. Chem., 2009, 19, 2457–2469.
12. S. Stankovich, D. A. Dikin, G. H. B. Dommett, K. M. Kohlhaas, E. J. Zimney, E. A. Stach, R. D. Piner, S. T. Nguyen and R. S. Ruoff, Nature, 2006, 442, 282–286.
13. S. Cui, S. Mao, G. Lu and J. Chen, J. Phys. Chem. Lett., 2013, 4, 2441–2454.
14. I. V. Lightcap and P. V. Kamat, Acc. Chem. Res., 2013, 46, 2235–2243.
15. A. G. Goncalves, B. Jarrais, C. Pereira, J. Morgado, C. Freire and M. F. R. Pereira, J. Mater. Sci., 2012, 47, 5263–5275.
16. J. Gao, M. Hu, Y. Dong and R. K. Y. Li, ACS Appl. Mater. Interfaces, 2013, 5, 7758–7764.
17. B. Fugetsu, E. Sano, H. Yu, K. Mori and T. Tanaka, Carbon, 2010, 48, 3340–3345.
18. M. Shateri-Khalilabad and M. E. Yazdanshenas, Carbohydr. Polym., 2013, 96, 190–195.
19. D. Li, M. B. Müller, S. Gilje, R. B. Kaner and G. G. Wallace, Nat. Nanotechnol., 2008, 3, 101–105.
20. Y. Shen, S. Yang, P. Zhou, Q. Sun, P. Wang, L. Wan, J. Li, L. Chen, X. Wang, S. Ding and D. W. Zhang, Carbon, 2013, 62, 157–164.
21. W. Park, S. Noh, M. Joo, T. Kim, W. Park, M. Jung, J. Moon and K. Park, Appl. Phys. Lett., 2013, 103, 033107.
22. N. Yousefi, M. M. Gudarzi, Q. Zheng, X. Lin, X. Shen, J. Jia, F. Sharif and J. K. Kim, Composites, Part A, 2013, 49, 42–50.
23. N. Yousefi, M. M. Gudarzi, Q. Zheng, S. H. Aboutalebi, F. Sharif and J. K. Kim, J. Mater. Chem., 2012, 22, 12709–12717.
24. K. H. Liao, Y. Qian and C. W. Macosko, Polymer, 2012, 53, 3756–3761.
25. S. Bai, X. Shen, G. Zhu, A. Yuan, J. Zhang, Z. Ji and D. Qiu, Carbon, 2013, 60, 157–168.
26. H. Zhang, W. Zheng, Q. Yan, Y. Yang, J. Wang, Z. Lu, G. Ji and Z. Yu, Polymer, 2010, 51, 1191–1196.
27. C. K. Samios, K. G. Gravalos and N. K. Kalfoglou, Eur. Polym. J., 2000, 36, 937–947.
28. J. Shen, T. Li, Y. Long, M. Shi, N. Li and M. Ye, Carbon, 2012, 50, 2134–2140.
29. J. Zhu, M. Chen, Q. He, L. Shao, S. Wei and Z. Guo, RSC Adv., 2013, 3, 22790–22824.

---