Preparation of polyvinylidene fluoride/expanded graphite composites with enhanced thermal conductivity via ball milling treatment

Abstract

In recent decades, great efforts have been devoted to prepare materials with enhanced thermal conductivity due to the growing interest in thermal conductive materials. Herein, we illustrate a facile strategy to improve the thermal conductivity of polyvinylidene fluoride/expanded graphite (PVDF/EG) composites by pre-treatment of EG via ball milling. Before incorporating EG into PVDF via conventional melt processing, EG powders were treated by shear-force-dominated ball milling. In this way, the loose and porous vermicular structure of EG could be effectively destroyed and exfoliated to graphite nanosheets (GNSs). As a result, the PVDF/GNSs composites show improved thermal conductivity owing to their larger specific surface area. With the filler content fixed at 15 wt%, the thermal conductivity of treated PVDF/GNSs composites can reach 1.29 W m⁻¹ K⁻¹, 42.5% higher than that of PVDF/EG (0.90 W m⁻¹ K⁻¹). Moreover, the electromagnetic interference (EMI) shielding property and tensile strength of PVDF/CNSs composites are also remarkably improved. Our work proves to be a simple and easily industrialized method for EG treatment which has great potential for improving the thermal conductivity of polymer composites in lighting devices and electromagnetic shielding applications.

---

1. Introduction

With the rapid development of electronics, automotive, and aerospace industries,^1,2 heat dissipative materials has aroused great attention. Polymer materials are widely used due to their low density, corrosion resistance and ease of processing. However, it is well known that the thermal conductivity of polymers is very low to meet the application demands. For most of the polymers, the thermal conductivity at room temperature is lower than 0.5 W m⁻¹ K⁻¹.³ Therefore, many efforts have been devoted to improve the thermal conductivity of polymers for the purpose of releasing the heat generated in the devices to maintain the temperature of the products at a desired level.⁴

Recent studies have revealed that the incorporation of thermally conductive fillers, such as carbon-based fillers, ceramic or metal particles,^5–7 has effectively improved the thermal conductivity of polymers. Compared with ceramic and metal particles suffering from the drawback of high density, carbon-based fillers appear to be the most promising fillers because they are lightweight and have high thermal conductivity. Graphite,^8,9 carbon fiber,¹⁰ carbon black¹¹ and carbon nanotubes¹² are commonly used carbon-based fillers. Among these fillers, two-dimensional graphite nanoplates have been regarded as the most promising conductive filler owing to its high thermal conductivity and low cost. Single graphene sheets show thermal conductivity as high as 5300 W m⁻¹ K⁻¹,^13,14 which determines the high thermal conductivity of graphite. Expanded graphite (EG), an exfoliated form of graphite with layers of 20–100 nm thickness,^15,16 is regarded as the most economical and effective candidate for conductive enhancements. Since EG is prepared from graphite intercalation compounds by rapid heating, a sudden increase in the dimension perpendicular to the carbon layers results in the loose and porous vermicular structures of initial EG. However, the “wormlike” EG is tough to be added directly to improve thermal conductivity of polymer matrices where high loading is usually needed.

Various approaches including ultrasound, solution shear, melt shear, high-speed mixing have been developed to destroy and exfoliate these layered structures. For instance, Yu and coworkers reported that EG could be exfoliated and dispersed into epoxy with high thermal conductivity up to 6.44 W m⁻¹ K⁻¹ at 25 vol% by ultrasound for 24 h.¹⁷ Coupling agents were also used to noncovalently functionalize EG to improve thermal conductivity.¹⁸ Seokwoo Jeon and coworkers sonicated EG with 1-pyrenebutyric acid in pyridine for 12 h to prepare non-covalent functionalized graphene flakes and to fabricate epoxy nanocomposites with improved thermal conductivity (∼1.53 W m⁻¹ K⁻¹).¹⁹ In most cases, EG/polymer composites with high thermal conductivity are often produced by solution mixing under the effect of ultrasound. Although ultrasound is simple, the fact that the batches of hundred or thousand milliliter and process time up to even dozens of hours has make it difficult to be utilized in practical industrial application because of high energy consumption and low efficiency. Therefore, it is significant to find a strategy to deal with EG in a large scale to meet commercial use.

Lately, ball milling, a common technique in powder production industry, is known to be a well scalable technique and a good candidate for generating shear force.²⁰ Since Knieke et al.²¹ and Zhao et al.^22,23 initiatively explored the milling method to exfoliate graphite in 2010, ball milling techniques of graphite have been developed continuously. For example, milling of graphite in the presence of NH₃BH₃,²⁴ Na₂SO₄,²⁵ melamine,²⁶ a mixture of 1-pyrene carboxylic acid and methanol²⁷ or other exfoliation media were adopted to obtain graphene. These studies mainly concentrate on fabrication of single layered-graphene involving complicated separation process after ball milling. Additionally, yield of single layered-graphene remains low regardless of great efforts devoted. Even though, exfoliation of graphite to few-layered graphite nanosheets is easily achieved by milling. Taking these into consideration, we attempt to incorporate EG milled without subsequent separation into matrices, which helps to simplify procedure and make full use of EG milled.

Herein, in order to improve thermal conductivity of polymer/EG composites, we employed ball milling method to destroy the vermicular structure of EG and obtain exfoliated graphite nanosheets (GNSs) in a large-scale without subsequent separation. Just by adjusting milling time, the size of GNSs can be easily controlled. Poly(vinylidene fluoride) (PVDF) is widely used in the field of thermal conduction for its excellent mechanical properties, chemical resistance and oxidative stability. So it's of great significance to improve thermal conductivity of PVDF in practical application. Thus we make an attempt to incorporate ball milled GNSs into PVDF though melting process, trying to improve the thermal conductivity of PVDF/EG composites. The electrical conductivity, electromagnetic interference (EMI) shielding, rheological and mechanical properties of PVDF/CNSs composites is also investigated.

2. Experimental

2.1 Raw materials

The original graphite intercalation compounds (GICs, KP32) were acid-intercalated natural crystalline graphite, purchased from Qingdao Super Graphite Co. Ltd. It was put in a muffle furnace at 900 °C for 60 s to obtain expanded graphite (EG, 80 mesh in diameter). After rapid heating, the GICs were explosively expanded for hundreds of times along the c-axis direction and emerged an enormous increase in volume due to the evaporation of the intercalators. PVDF (solef@6010) is bought from Solvay company, and available in powders with a M_w (g mol⁻¹) of 300–320. N,N-Dimethylformamide (DMF), is purchased from Tianjin Bodi Chemical Co. Ltd and used as received.

2.2 Preparation of GNSs, PVDF/EG and PVDF/GNSs

Firstly, using DMF as solvent, EG powders was ball milled inside milling container (QM-3SP2, Nanjing University Instrument Co. Ltd) at 500 rpm for 5 h, 10 h and 15 h, respectively. After that, the EG suspension was filtrated, then washed with alcohol to get rid of residual DMF, and finally put into the oven for 10 h to insure the solvent is evaporated. EG powders ball milled for 5 h, 10 h and 15 h are defined as GNSs-5, GNSs-10, and GNSs-15, respectively. To make a comparison, we also use high-speed rotating mixer (Linda Mechanical Co., Ltd., China) at 25 000 rpm to treat EG. Fixed filler content at 15%, PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10 and PVDF/GNSs-15 powders were melted in an internal mixer (Rheocord 9000, Haake Co., Ltd., Germany) operated at 60 rpm and 210 °C for 10 min. The as prepared composites were compression molded (KT-0701, Beijin Kangsente Co., Ltd., China) under 210 °C and at low pressure (2 MPa) for 120 s (2 degassing cycles), then a high pressure (10 MPa) for 180 s followed by air cooling.

2.3 Characterizations

To evaluate Brunauer–Emmett–Teller (BET) specific surface area of EG and GNSs, nitrogen physisorption measurements were conducted on Tristar-3020 (USA). The structures of EG and GNSs were observed with a field-emission scanning electron microscope (FE-SEM, Inspect F, USA) with an acceleration voltage of 20 kV. While PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10 and PVDF/GNSs-15 composites were observed with an acceleration voltage of 5 kV. The thermal conductivity of composites was measured with a thermal constant analyzer (Hot Disk TPS 2500, Sweden). The thermal constants analyzer utilizes the transient plane source (TPS) method to measure the thermal conductivity of materials. Specimens with diameter of 25 mm and thickness of 4 mm were prepared through compression molding. The average values were calculated from at least eight samples. Two points method was used to measure direct current (DC) electrical conductivity for these composites at room temperature. Silver paint was applied onto both ends of these samples to ensure good contact. As a result, contact resistance was negligible comparing with the resistance of the specimen. The electrical resistance was measured with a Keithley 6487 picoammeter at a voltage of 10 V. The EMI shielding property of composites was measured with an E5071C ENA series network analyzer (Agilent Technologies) using an industrial standard method. Samples with diameter of 12 mm and thickness of 3 mm were placed in the specimen holder, measured in the X band frequency range. Rheological measurement was performed on a dynamic rheometer (Bohlin Gemini 2000, Malvern, British) with parallel plate geometry (diameter 25 mm). The parallel-plate fixture with a diameter of 25 mm and fixed space of 1.5 mm was used at angular frequencies from 0.01 to 100 Hz. Tensile tests were carried out using an Instron universal tensile testing machine with a speed of 50 mm min⁻¹ according to GB/T1040-92 standard. The measured temperature was around room temperature. TGA was conducted on a thermogravimetric analyzer (TG 209F1 Iris, Netzsch, Germany) under dry nitrogen. These samples were heated at a rate of 10 °C min⁻¹, and the relative mass loss was recorded from 30 °C to 600 °C. The crystallization behavior of these composites was studied with a DSC-204 (Netzsch, Germany). Experiments were performed with 6–10 mg samples under dry nitrogen.

3. Results and discussion

3.1 The effect of ball milling on the structure of expanded graphite

In the milling process, there are a lot of process parameters influencing final particle size of graphite. The most straightforward processing parameter is the milling time. It has been reported that DMF, whose surface tension is close to that of graphene layers (40 mJ m⁻²), is easy to overcome van der Waals force of adjacent graphene flakes to exfoliate graphite and also protect the exfoliated sheets from agglomeration.²⁸ Choosing DMF as the dispersing medium, initial EG was ball milled for 5 h, 10 h and 15 h to obtain GNSs-5, GNSs-10 and GNSs-15, respectively. To detect the morphology change of EG, SEM was carried out. As demonstrated in Fig. 1(a), the plates of original EG are stacked together to form the loose and porous vermicular structures, which is difficult to incorporate into polymer matrices directly because of its greatly expanded volume and poor wettability to polymer matrices. After treated with high-speed rotating mixer, the vermicular structure of EG is destroyed to some extent, whose lateral size is about 219 μm. While milled with planetary ball mill, the vermicular structure of EG is destroyed completely thus the plates are isolated from each other.

Fig. 1 SEM images of (a) initial EG, (b) EG treated by high speed mixing, (c) GNSs-5 (ball milled for 5 h); (d) GNSs-10, (e) GNSs-1.

The lateral size distributions of EG and GNSs are demonstrated in Fig. 2. The lateral size of graphite decrease sharply to 49 μm after 5 hours' milling, while the sizes of GNSs-10 and GNSs-15 gradually decrease with prolonged milling time. When milling process is carried out, the balls show complex behaviour but generate two forces on layered graphite, namely, shear force and compression force²⁰ illustrated in Fig. 3. Original EG is of large size and easy to be broken by compression force. So the size of graphite decreases quickly at the very beginning. Upon milling, whether exfoliation happens is still not clear since the images of SEM do not give the information of thickness. Therefore, further characterization is necessary.

Fig. 2 Lateral size distributions of (a) EG treated by high speed mixing, (b) GNSs-5, (c) GNSs-10, (d) GNSs-15 calculated form SEM.

Fig. 3 The illustration of the exfoliation and fracture effects of ball milling. (a) Shear force-induced exfoliation; (b) compression force-induced fracture.

The N₂ adsorption isotherms of EG and GNSs are shown in Fig. 4, all the curves shows the similar trend. Because of weak interaction between N₂ and EG, the isotherm curve is flat at low P/P₀ regions. Nevertheless, once the adsorption of N₂ molecule begins, N₂–N₂ interaction tends to aggrandize the adsorption of more N₂ molecules thus adsorption isotherms undergo a sharp rise with respect to the relative pressure values. The BET specific surface area of them is calculated shown in Table 1. The surface area of EG is 23.167 m² g⁻¹. Comparing with EG, the GNSs possess much larger specific surface area. And surface area of GNSs with longer milling time is larger than that of GNSs with shorter milling time. Surface area of GNSs-15 can reach 35.493 m² g⁻¹. Larger special surface area is an indication of higher degree of graphite exfoliation in the latter. So it can be concluded that not only is the vermicular structure of EG destroyed to isolated plates, but exfoliation does happen in milling process.

Fig. 4 N₂ adsorption isotherms for EG, GNSs-5, GNSs-10, GNSs-15. The inset is N₂ adsorption isotherms at low P/P₀ regions.

Table 1 Surface area for EG, GNSs-5, GNSs-10, GNSs-15

Specimens	SBET (m^2 g^−1)
EG	23.1671
GNSs-5	24.4760
GNSs-10	30.5918
GNSs-15	35.4932

---

3.2 Thermal and electrical conductivity of PVDF/EG and PVDF/GNSs composites

EG from high-speed rotating treatment, GNS-5, GNS-10 and GNS-15 were melting mixed with PVDF to obtain PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10 and PVDF/GNSs-15 composites, respectively. Then compression molding was carried out. Compared with solution mixing, melt mixing is free of solvent, simple and easy industrialized. Fig. 5 shows thermal conductivity of PVDF/EG and PVDF/GNSs composites at filler weight content of 15 wt%. The thermal conductivity of PVDF/EG is 0.90 W m⁻¹ K⁻¹ much higher than that of pure PVDF (0.19 W m⁻¹ K⁻¹), which is ascribed to the high intrinsic thermal conductivity of EG. Nevertheless, the thermal conductivities of PVDF/GNSs composites undergo a further improvement when ball milling is adopted before blending.

Fig. 5 Thermal conductivity of PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10, PVDF/GNSs-15 composite.

As can be seen from Table 2, the thermal conductivity of PVDF/GNSs-15 is up to 1.29 W m⁻¹ K⁻¹, 42.5% higher than that of PVDF/EG and the thermal conductivity of treated PVDF/GNSs composites increases with milling time. This results is in opposite with other results reported in the literature,²⁹ saying the thermal conductivity of composites filled with EG of large sizes is significantly higher than for the smaller ones. In our study, the samples containing nanoplates of smaller size show higher thermal conductivity. The mechanism responsible for such phenomenon is that specific surface area of fillers plays a crucial role in thermal conductivity enhancement. Since surface area of GNSs increases with milling time (as described above), the probability of the formation of thermally conductive networks is increased, leading an enhancement of thermal conductivity.^30–32 Although the surface area of GNSs-15 is larger than that of GNSs-10, there is a slight enhancement of thermal conductivity when milling time is prolonged from 10 h to 15 h. The explanation for this is that thermal interface resistance contributions across the filler–polymer matrix interface should be taken into account. The lateral particle size of GNSs-15 is smaller than that of GNSs-10, causing greater thermal interface resistance. So there is a slight increase of thermal conductivity. Electrical conductivity of PVDF/EG and PVDF/GNSs were also tested (Table 2), showing the same increase trend with thermal conductivity. Incorporating 15 wt% of EG into PVDF doesn't make it conductive. While adding GNSs, PVDF/GNSs become electrical conductive, indicating that the PVDF/GNSs composites have already formed an electrically conductive network. Our approach can improve both thermal and electrical conductivity, which is simple, effective and easily industrialized.

Table 2 Thermal and electrical conductivity of PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10, PVDF/GNSs-15 composites

Specimens	Thermal conductivity (W m^−1 K^−1)	Electrical conductivity (S m^−1)
PVDF/EG	0.904 ± 0.0784	—
PVDF/GNSs-5	1.192 ± 0.0750	0.698 ± 0.0156
PVDF/GNSs-10	1.264 ± 0.0596	0.952 ± 0.0534
PVDF/GNSs-15	1.285 ± 0.0721	0.926 ± 0.0847

---

3.3 The morphology of PVDF/EG and PVDF/GNSs composites

To provide insights into the structures of composites, cryo-fractured PVDF composites are observed using SEM shown in Fig. 6. The PVDF/EG composite exhibits a fractured surface with random dispersion of EG aggregates and EG plates remain stacked. More precisely, EG aggregates are isolated from each other, failing to form conductive pathways. On the contrary, the stacked structure of EG is exfoliated to graphite nanosheets after ball milling and GNSs show good dispersion in all the three PVDF/GNSs composites forming a network throughout the polymer matrix in macro size. In the meantime, GNSs exhibit a homogeneous dispersion in micro size and an orientation of plates in the lateral direction, which further promotes interconnected conductive networks. As can been seen from Fig. 6(b₁)–(d₁), the sizes of GNSs within composites decrease and the probability of plates contact increase with longer milling time, which is a result of fragmentation and exfoliation derived from ball milling.

Fig. 6 SEM images of PVDF/EG containing 15% EG (a and a₁), PVDF/GNSs-5 containing 15% GNSs milling for 5 h (b and b₁), PVDF/GNSs-10 (c and c₁) containing 15% GNSs milling for 10 h, PVDF/GNSs-15 containing 15% GNSs milling for 15 h (d and d₁). The red circles indicate the dispersion of EG.

To further illustrate that conductive networks are formed, rheological measurement was performed. Storage modulus (G′) was measured as a function of frequency in the linear viscoelastic region, as demonstrated in Fig. 7. Results have shown the storage modulus of PVDF nanocomposites is higher than that of neat PVDF. Neat PVDF exhibits the characteristic response of a viscous polymer melt with a monotonous increase in G′ as a function of sweeping frequency, indicating a terminal flow behavior. Addition of 15 wt% of EG increased the storage modulus without changing the viscoelastic behavior, indicating that GNSs network hasn't been formed to restrain polymer chain relaxation yet. When GNSs are incorporated into PVDF, G′ starts to develop a plateau and G′ of PVDF/GNSs gradually increases upon increasing milling time at low frequency, which is indicative of a transition from liquid-like to solid-like viscoelastic behavior. This behavior at low frequency can be attributed to the formation of an interconnected GNSs network in PVDF which restrains the long-range motion of polymer chains.³³ The result of rheological measurement coincides with SEM, providing evidence that conductive networks are surely formed, so PVDF/GNSs composites show higher thermal and electrical conductivities.

Fig. 7 Storage modulus (G′) of pure PVDF, PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10, PVDF/GNSs-15 composites as a function of frequency.

3.4 EMI shielding property

EMI SE is a measure of the capacity of materials to attenuate EMI wave intensity.³⁴ Fig. 8 shows the variation of EMI SE over the frequency range of X-band for PVDF/EG and PVDF/GNSs with various milling time. It is observed that over the entire frequency range, SE of PVDF/GNSs composite is higher than that of PVDF/EG composites, which is mainly attributed to the formation of conducting networks in the insulating PVDF matrix. It has been reported that the target value of the EMI SE needed for commercial applications is around 20 dB. As presented in Fig. 8, our PVDF/GNSs composites exhibited SE > 25 dB in the X-band for 15 wt% loadings, indicating the composites can meet the commercial application demands. In conclusion, EMI shielding results, combined with the advantages of a cheap and abundant supply of graphite and the easy pretreatment of EG, indicate that polymer/GNSs composites can be used commercially as effective and lightweight shielding materials for electromagnetic radiation.

Fig. 8 EMI SE over the frequency range of X-band for PVDF/EG and PVDF/GNSs with various milling time.

3.5 Mechanical property

It is known that in order to obtain the composites with high thermal conductivity, the thermally conductive fillers with high loading are usually added into polymer matrix. Unfortunately, high loading makes the mechanical property of the thermally conductive composites worse. The stress–strain curves of PVDF nanocomposites are presented in Fig. 9. It is evident from the figure that for PVDF/GNSs at a fixed value of strain, there is an increase of tensile stress compare to PVDF/EG. This indicates an increase in the stiffness of the composites due to pretreatment of EG. With the increase of milling time, the tensile strength of PVDF/GNSs undergoes an enhancement, up to 47 MPa. Nevertheless, the tensile strength of PVDF/EG composites is only 36 MPa. This very large increase of tensile strength is attributed to the exfoliation of EG during milling process and uniform dispersion of GNSs in the composites. Our method provides an effective way to guarantee the mechanical property of composites at high loading.

Fig. 9 Mechanical property of PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10, PVDF/GNSs-15 composites.

3.6 Thermal analysis

Typical DSC crystal curves of the PVDF/EG and PVDF/GNSs composites are shown in Fig. 10. Both T_c of PVDF/EG and PVDF/GNSs composites show an obvious increase, which indicates that EG can play the role of nucleating agent for PVDF matrix. However, T_c of PVDF/GNSs composites is higher owing to their large surface area for adsorption of PVDF chain and thereby causing easier nucleation, which benefits the thermal conductivity of related composites. It has reported that in fillers/semicrystalline composites, if fillers can provide nucleation sites for polymers, the interfacial thermal resistivity can be reduced and the thermal conductivity can be improved by increasing the nucleation of crystal at the fillers/polymer interface.^35,36

Fig. 10 DSC curves of pure PVDF, PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10, PVDF/GNSs-15 composites.

TGA thermograms of the PVDF composites are also presented in Fig. 11. As shown in Fig. 11(a), all composites exhibit similar thermal degradation behavior. Fig. 11(b) shows a more remarkable improvement of PVDF/GNSs in the maximum decomposition temperature suggesting that GNSs act as an effective thermal barrier due to its better dispersion and thus hinder the degradation of PVDF.

Fig. 11 TGA curves of PVDF/EG, PVDF/GNSs-5, PVDF/GNSs-10, PVDF/GNSs-15 composites.

4. Conclusion

A facile strategy to improve thermal conductivity of PVDF/EG composites has been developed by pre-treatment of EG via ball milling. Treated by shear-force-dominated ball milling, the loose and porous vermicular structure of EG was effectively destroyed and partly exfoliated to GNSs. The thermal and electrical conductivities of PVDF/GNSs composites were improved owing to larger specific surface area of GNSs. Moreover, the electromagnetic interference (EMI) shielding was also remarkably enhanced to meet the commercial application demands. In addition, tensile strength of PVDF/CNSs composites at high loading is better than that of PVDE/EG composites. Taking both the advantages of a cheap and abundant supply of graphite and the easy pretreatment of EG into consideration, our work provides a simple and effective way for EG treatment which is easily industrialized in heat conducting field.

Acknowledgements

We would like to express our sincere thanks to the National Natural Science Foundation for financial support (Grant. no. 51421061 and 51210005).

Notes and references

1. S. Mallik, N. Ekere, C. Best and R. Bhatti, Appl. Therm. Eng., 2011, 31, 355–362.
2. D. Miracle, Compos. Sci. Technol., 2005, 65, 2526–2540.
3. Z. Han and A. Fina, Prog. Polym. Sci., 2011, 36, 914–944.
4. Y. Yoo, H. L. Lee, S. M. Ha, B. K. Jeon, J. C. Won and S. G. Lee, Polym. Int., 2014, 63, 151–157.
5. H. O. Pierson, Handbook of carbon, graphite, diamonds and fullerenes: processing, properties and applications, Noyes publications, New York, 1993.
6. G. Wypych, Handbook of fillers: physical properties of fillers and filled materials, Chem. Tech. Publishing, Toronto, 2000.
7. J. E. Fischer, Carbon Nanomater., 2006, 41–75.
8. A. Yu, P. Ramesh, X. Sun, E. Bekyarova, M. E. Itkis and R. C. Haddon, Adv. Mater., 2008, 20, 4740–4744.
9. S. Nozato, A. Nakasuga, T. Wada, H. Yoshitani and H. Ihara, RSC Adv., 2016, 6, 25776–25779.
10. R. Patton, C. Pittman, L. Wang, J. Hill and A. Day, Composites, Part A, 2002, 33, 243–251.
11. F. El-Tantawy, K. Kamada and H. Ohnabe, Mater. Lett., 2002, 56, 112–126.
12. J. N. Coleman, U. Khan, W. J. Blau and Y. K. Gun'ko, Carbon, 2006, 44, 1624–1652.
13. S. Stankovich, D. A. Dikin, G. H. 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.
14. L. M. Veca, M. J. Meziani, W. Wang, X. Wang, F. Lu, P. Zhang, Y. Lin, R. Fee, J. W. Connell and Y. P. Sun, Adv. Mater., 2009, 21, 2088–2092.
15. W. Zheng, S.-C. Wong and H.-J. Sue, Polymer, 2002, 43, 6767–6773.
16. G. Chen, C. Wu, W. Weng, D. Wu and W. Yan, Polymer, 2003, 44, 1781–1784.
17. A. Yu, P. Ramesh, M. E. Itkis, E. Bekyarova and R. C. Haddon, J. Phys. Chem. C, 2007, 111, 7565–7569.
18. W. S. Sarsam, A. Amiri, S. Kazi and A. Badarudin, Energy Convers. Manage., 2016, 116, 101–111.
19. S. H. Song, K. H. Park, B. H. Kim, Y. W. Choi, G. H. Jun, D. J. Lee, B. S. Kong, K. W. Paik and S. Jeon, Adv. Mater., 2013, 25, 732–737.
20. C. F. Burmeister and A. Kwade, Chem. Soc. Rev., 2013, 42, 7660–7667.
21. C. Knieke, A. Berger, M. Voigt, R. N. K. Taylor, J. Röhrl and W. Peukert, Carbon, 2010, 48, 3196–3204.
22. W. Zhao, M. Fang, F. Wu, H. Wu, L. Wang and G. Chen, J. Mater. Chem., 2010, 20, 5817.
23. W. Zhao, F. Wu, H. Wu and G. Chen, J. Nanomater., 2010, 2010, 1–5.
24. L. Liu, Z. Xiong, D. Hu, G. Wu and P. Chen, Chem. Commun., 2013, 49, 7890–7892.
25. Y. Lv, L. Yu, C. Jiang, S. Chen and Z. Nie, RSC Adv., 2014, 4, 13350–13354.
26. V. Leon, M. Quintana, M. A. Herrero, J. L. Fierro, A. de la Hoz, M. Prato and E. Vazquez, Chem. Commun., 2011, 47, 10936–10938.
27. R. Aparna, N. Sivakumar, A. Balakrishnan, A. Sreekumar Nair, S. V. Nair and K. R. V. Subramanian, J. Renewable Sustainable Energy, 2013, 5, 033123.
28. Y. Hernandez, V. Nicolosi, M. Lotya, F. M. Blighe, Z. Sun, S. De, I. McGovern, B. Holland, M. Byrne and Y. K. Gun'Ko, Nat. Nanotechnol., 2008, 3, 563–568.
29. J. Kratochvíla, A. Boudenne and I. Krupa, Polym. Compos., 2013, 34, 149–155.
30. X. Sun, P. Ramesh, M. E. Itkis, E. Bekyarova and R. C. Haddon, J. Phys.: Condens. Matter, 2010, 22, 334216.
31. L. E. Nielsen, Ind. Eng. Chem. Fundam., 1974, 13, 17–20.
32. D. Bigg, Polym. Compos., 1986, 7, 125–140.
33. H. Kim and C. W. Macosko, Macromolecules, 2008, 41, 3317–3327.
34. H. B. Zhang, Q. Yan, W. G. Zheng, Z. He and Z. Z. Yu, ACS Appl. Mater. Interfaces, 2011, 3, 918–924.
35. B. Shen, W. Zhai, C. Chen, D. Lu, J. Wang and W. Zheng, ACS Appl. Mater. Interfaces, 2011, 3, 3103–3109.
36. D. Cai and M. Song, Carbon, 2008, 46, 2107–2112.

---