Exceptionally high thermal and electrical conductivity of three-dimensional graphene-foam-based polymer composites

Abstract

Graphene foams (GF) assembled with one- or few-layered ultrathin two-dimensional crystals have showed huge application potentials owing to their unique three-dimensional (3D) structure and superior properties. Here, we present a polymer-template-assisted assembly strategy for fabricating a novel class of 3D graphene architecture. A free-standing GF architecture has been built to act as thermal and electrical conduction paths in polymer composites. The obtained GF/polymer composites exhibit a high thermal conductivity (1.52 W mK⁻¹) and high electrical conductivity (3.8 × 10⁻² S cm⁻¹) at relatively low GF loading (5.0 wt%). The GF/polymer composites are potentially useful in advanced packaging materials of high power LED and microelectronic devices.

---

1. Introduction

Efficient thermal management has become a critical necessity for guaranteeing performance and reliability in next-generation electronic devices. Polymers are widely used in the electronics industry as the support and adhesive. However, these polymer materials trend to show very low thermal conductivity, for instance, only 0.10 W mK⁻¹ for epoxy resin, giving them poor capability in heat dissipation applications. Therefore, the development of polymer materials with high thermal conductivity is of great technological importance. A traditional method to enhance polymer thermal conductivity is to blend them with high thermal conductive fillers, including metals,^1–3 carbon and hybrid materials,^4,5 or ceramic particles.^6–13 Due to its outstanding thermal conductivity (5300 W mK⁻¹ in theory),¹⁴ mechanical strength,¹⁵ and electron mobility,^16,17 graphene has been known as an excellent nano-filler to greatly enhance the physical properties of polymers.¹⁸

Though, Yu et al. exfoliated natural graphite flakes into graphene nanoplatelets (GNPs) with the thermal-assisted exfoliation and prepared the GNPs/epoxy composites.¹⁹ These GNPs/epoxy composites were found to have thermal conductivity up to 6.44 W mK⁻¹ (25 vol% filler loading). The large amount of fillers always results in the degradation of other important properties of polymers, such as mechanical and optical properties. In brief, it still remains a great challenge to obtain the polymer materials having high thermal conductivity with a small addition of fillers.

Recently, a three-dimensional interconnected graphene foam (3DGF) structure has attracted intense interest due to its flexibility, as well as high electrical and thermal conductivity.^20–25 To date, the synthesis of 3DGF can be divided into three categories, including self-assembly approaches,^26–28 template-directed approaches^{21,22,25,29–31} and other approaches;³² all of which result in 3D graphene architectures with different microstructures and properties.

Although there have been a few work about the applications based on 3DGF/polymer composites,^33–38 we noted that a comprehensive investigation of 3DGF for thermal management or heat dissipation applications is still absent. Thus, in this study, we report a novel strategy for the fabrication of graphene foams (GF) by self-assembly of graphene sheets on a 3D polymer skeleton. Then the epoxy resin was impregnated into 3DGF to obtain the epoxy composites. Based on this method, the graphene loading can be easily controlled, and the interconnected 3D network of GF in the composites results in not only highly thermal conductivity but also enhanced electrical conductivity.

2. Experimental

2.1. Materials

Commercial polyurethane (PU) sponge used for kitchen supplies was utilized in this work. Graphene nanoplatelets (GNPs) as the filler were supplied by Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences (China). A cycloaliphatic epoxy resin (6105) and the hardener of methyl-hexahydrophthalic anhydride (MHHPA) were obtained from DOW Chemicals (USA) and Shanghai Li Yi Science & Technology Development Co. Ltd. (China), respectively. Neodymium(III) acetylacetonate trihydrate (Nd(III)acac) purchased from Aldrich Chemicals was used as latent catalyst. All other chemicals were of analytical reagent grade and used without further purification.

2.2. Preparation of GF/epoxy composites

GF/epoxy composites with different loadings (0, 1, 2, 3, and 5 wt%) were prepared by the following procedures. First of all, a pristine foam skeleton (about 2 × 2 × 2 cm³) was prepared using the commercial PU sponge, which was subsequently immersed and squeezed constantly in graphene/ethanol solution (4 mg ml⁻¹), followed by vacuum degassing for removal of the bubbles within foam structure. Secondly, the surface of PU foam skeleton was assembled with GNPs after drying process. Free-standing 3D graphene architecture with PU skeleton was pyrolyzed at 700 °C for 2 h at argon atmosphere to completely decompose the polymer. The obtained 3DGF was treated at 1400 °C in vacuum for 45 min to increase the interfacial interaction between GNPs. A quantity of Nd(III)acac was added into a cycloaliphatic epoxy resin and subsequently be stirred at 80 °C in the three-necked flask for 2 h. The obtained homogeneous solution was mixed with its curing agent (MHHPA) with 100 : 95 ratio. The mixture was mixed with the speedmixer at a speed of 3000 rmp for 5 min and then be microwaved for 1 min to decrease the viscosity and degassed in a vacuum chamber at room temperature for 30 min. Then the homogeneous mixture solution was infused into 3D graphene architecture. Finally, 3DGF immersed in epoxy solution was placed into a vacuum drier to remove the air bubbles from the solution and cured at 165 °C for 14 h. After the curing process, the samples were naturally cooled down to the room temperature and then diced with low speed diamond cutter for different characterizations. The preparation process of GF/epoxy composites is illustrated in Scheme 1.

Scheme 1 The preparation process of GF/epoxy composite.

2.3. Characterizations

The microstructures of the samples were obtained from JEOL JEM-2100 (TEM, JEOL, Japan) instrument with an acceleration voltage of 200 kV. The optical micrograph (OM) image was captured by optical microscope (OM, Leica DM2500M, Germany). The fractured surface of the composites was examined on field emission scanning electron microscopy (FE-SEM, QUANTA FEG250, USA) at an acceleration voltage of 10 kV. Samples were broken and the fractured surface was coated with a thin gold layer to avoid the accumulation of charge in SEM. Atomic force microscope (AFM) measurement was conducted on a multimode SPM from digital instruments with Nanoscope IA controller. X-ray photoelectron spectroscopy (XPS) was carried out with a Kratos AXIS Ultra DLD spectrometer, using Al Kα excitation radiation (hν: 1253.6 eV). Raman spectroscopy was recorded employing a laser wavelength of 532 nm (RENISHAW plc, Wotton-under-Edge, UK). Thermal conductivity of the composites was evaluated from thermal diffusivity (mm² s⁻¹), which was determined by LFA 447 Nanoflash apparatus (NETZSCH, Germany) at room temperature. The samples were prepared in square shape, with 10 mm in lateral size and a thickness about 1.2 mm. The IR-photos were captured by infrared camera (Fluke, Ti400, USA). The electrical conductivity of the composites was measured by four-probe conductivity meter (SB100A/2, Qianfeng, China).

3. Results and discussion

3.1. The characterizations of GNPs

SEM image for the powder of GNPs is shown in Fig. 1(a). GNPs have a flaky sheet structure in shape with width of about 10 μm or below. In Fig. 1(b), the typical OM image shows the morphology of GNPs deposited on Si substrate by dip-coating. The lateral size of the flakes is ranging from a few to approximately 10 μm, which agree with the observation in SEM. TEM was applied to determine the morphology and thickness of GNPs. As illustrated in Fig. 1(c), a large and transparent GNPs was observed on the copper grid. The GNPs sheet with a wrinkled surface is often credited as a favorable medium for strong interfacial interaction with the polar epoxy matrix. The inset in Fig. 1(c) indicated that the thickness of GNP is 3–4 nm and composed of about 10 graphene layers. Fig. 1(d) displays the AFM image of a selected GNPs flake. Fig. 1(e) shows the statistical results of thickness measurement for randomly selected 43 flakes from the ensemble, where the thickness ranges from 1 to 8 nm (average: 3.84 nm). Fig. 1(f) shows the statistical results of the lateral size for randomly selected 80 flakes from the ensemble, where the size ranges from 1 to 10 μm (average: 3.93 μm). Moreover, the XPS and Raman are also presented in Fig. S1 and S2,† both indicating the low defects of GNPs.

Fig. 1 (a) SEM, (b) OM, (c) and TEM images of GNPs. The HR-TEM image is shown in the inset of (c). (d) AFM image and estimated thickness of single GNP. Histograms of measured values for (e) thickness and (f) lateral size of GNPs.

3.2. The investigation of the microstructures

The SEM images of obtained GF and the samples integrated with epoxy are illustrated in Fig. 2. We choose PU foam, a porous structure with an interconnected 3D architecture, as a template for coating with GNPs. In Fig. S3,† we could see that the neat PU exhibited a 3D cross-linked structure and a smooth surface. After squeezing and vacuum degassing process, the PU foam was assembled with GNPs on the skeleton surface. From the low and higher resolution images in Fig. 2(a) and (b), we could see that GNPs were uniformly attached on the surface of PU skeleton with continuous macropores, whose size is around hundreds of micrometers. The morphology is similar to that of PU foams, but the skeleton of the network exhibit a crinkled and rough texture associated with the presence of flexible graphene sheets, which is due to the difference thermal expansion between the GNPs and PU skeletons during the drying process. After the removal of the PU skeleton by pyrolysis of PU under Ar, they remain as a 3D scaffold structure. Fig. 2(c) and (d) display low- and high-magnification SEM images of the obtained GF. Though a small shrinkage of the overall architecture occurs, the majority of graphene sheets is still overlapped and linked between each other, forming the wall of the foams. All the GNPs in 3DGF are in direct contact with one another without breaks. Fig. 2(e) and (f) show that the cross-sectional SEM image of GF/epoxy composite after careful polishing. The approximate hexagon surrounded by pores which is graphene skeleton reveals that GF structure was interconnected through the matrix. Besides, this figure indicates that there exists a good interface between GF and epoxy matrix, suggesting excellent resin permeability inside the GF due to the large surface area and the presence of interconnected macropores.

Fig. 2 SEM images of (a and b) GNPs coating on the surface of PU foam; (c and d) 3DGF; (e and f) GF/epoxy composite.

3.3. Thermal properties of GF/epoxy composites

In general, for obtaining the superior thermal conductivity of ceramic particles-filled composites, either the thermal conductive pathways should be maximized through high filler loading, or the interfacial contact resistance has to be reduced by the enhancement of filler–matrix affinity. We used a transient laser flash method to indirectly determine the thermal conductivity of the neat epoxy and GF/epoxy composites at room temperature. In order to evaluate the effect of GF on the thermal properties of GF/epoxy composites, thermal diffusivity was measured first and thermal conductivity was calculated, both as a function of GNPs loading content (0–5 wt%). As shown in Fig. 3(a), there is an extraordinary increase both in thermal diffusivity and thermal conductivity for the composites with the increase of GNPs content as expected. With the addition of 1 wt% GNPs, the thermal diffusivity of GF/epoxy composite increased from 0.13 (neat epoxy) to 0.36 mm² s⁻¹, about 2.8-fold enhancement. As the content of GNPs was further increased to 5 wt%, the thermal diffusivity was improved by one order of magnitude, from 0.13 mm² s⁻¹ to 1.05 mm² s⁻¹, and the thermal conductivity was improved from 0.18 W mK⁻¹ to 1.52 W mK⁻¹. At 5 wt% GNP loading, the thermal conductivity of GF/epoxy composite was improved significantly by 8.5-fold compared to that of neat epoxy. For GF/epoxy composite, the enhancement of thermal conductivity comes from highly thermally conductive GFs in the composites. As the density of the macropores and the amount of GNP layers in GF are decided by intrinsic structure of PU template and the coating process, the proportion of GF in the epoxy composite can be easily tailored. So it is possible to increase thermal conductivity of GF/epoxy composite to an outstanding level only by adding GF content. Moreover, the interfacial thermal resistance between GNPs and epoxy can be greatly decreased due to the formation of 3DGF structure in epoxy composite. Meanwhile, the thermal conductivity of GF/epoxy composites is much higher than that of graphene/polymer composites prepared by blending method with the similar filler content (≈5 wt%),³⁹ as shown in Fig. S4.† This might be due to the fact that the increase of thermal conductivity was heavily dependent on the integrity of the heat conductive path, based on the interconnected structure of 3DGF.

Fig. 3 (a) Thermal conductivity and thermal diffusivity as a function of GF content; (b) infrared images of GF/epoxy composites with 1–5 wt% GF loading.

The neat epoxy and GF/epoxy composites were vertically placed on the same heating stage, and the rise of their surface temperature as a function of heating time was captured by the thermal tracer, as shown in Fig. 3(b). Samples were mounted on a hotplate with the surface temperature set to 200 °C, and the temperature changes of the composites were then traced. Because the thermal transport proceeds through the heat flow direction of the specimen, the infrared images are closely related to through-plane thermal conductivity. After 60 s, neat epoxy does not show any significant heat transfer from the hot plate to the surface of the composites, as typical polymeric materials are generally thermal insulators. In contrast, GF/epoxy composites show remarkable color changes with the increasing GNPs content, indicating better heat dissipation ability. These results are in good agreement with the thermal conductivity values shown in Fig. 3(a). Moreover, GF/epoxy composite displays the fastest temperature rise, demonstrating the high effectiveness of heat transfer by the interconnected structure of GNPs.

3.4. Electrical conductivity of GF/epoxy composites

Fig. 4 shows the electrical conductivity of the epoxy composites as a function of the GF content. A variety of factors can affect the final conductive properties, such as structure and production method of graphene, impurities and defects, surface treatments including oxidation and organic modification, and dispersion state, etc. Epoxy composites based on GF show significant enhancements in electrical conductivity. The electrical conductivity (σ) of GF/epoxy composites with 5 wt% GFs filler (σ = 3.8 × 10⁻² S cm⁻¹) is twelve orders of magnitude higher than that of neat epoxy (σ = 1.0 × 10⁻¹⁴ S cm⁻¹). This significantly improved electrical conductivity is mainly attributed to the interconnection of GNPs in 3DGF structure within epoxy composite. Furthermore, electrical conductivity of GF/epoxy composites is significantly higher than that of GNPs/epoxy composites at same graphene loading, which was prepared by speed mixing graphene powder in epoxy matrix.³⁹ This indicates that the interconnected GF provides more efficient paths for electron transfer inside the polymer than conventional GNPs filler. It also indicates that there is no an insulator–conductor transition in the GF/epoxy composites. As the GF/epoxy composites retain 3D interconnected network of graphene, and once the network is formed, the electron can be moved freely along the graphene network inside the polymer matrix. As shown in the inset in Fig. 4, an electrical conductive model of the epoxy composites was proposed. Accordingly, the popular percolation theory does not play a role for this GF infused composite.^18,35

Fig. 4 Electrical conductivity of epoxy composites with the different GF loadings. The inset is the electrical conduction model of the epoxy composites.

In addition, the conductivity of the epoxy composites increased with the addition of GFs. This is because the conductive path and cross-sectional area of the conductive portion of composites also increased with the increase of GFs. As shown in Fig. 5, the photograph of conductive GF/epoxy composites as the conductor to drive a LED show that the electrical conductivity of epoxy composites increase with the increase of the filler content. Currently, highly electrically conductive graphene/polymer composites were mainly fabricated via the mixing methods using graphene. After a careful comparison, we find that the electrical conductivity of our samples is comparable or much higher than other graphene reinforced polymer composites, as shown in Fig. 6. The electrical conductivity of epoxy composites depends on the GF loading fraction and at 5 wt% loading is 3.8 × 10⁻² S cm⁻¹, which is 1–2 order higher than that of conventional graphene-based composites.^39–47 The high electrical conductivity of epoxy composites is attributed to the well interconnected GF frameworks.

Fig. 5 The photographs of GF/epoxy composites as the electrical conductor with various GF loadings to drive a LED: (a) neat epoxy (b) 1 wt%, (c) 2 wt%, (d) 3 wt%, and (e) 5 wt%. (f) LED driven by direct connection with an electric wire.

Fig. 6 Comparison of electrical conductivity as a function of graphene content in GF/epoxy composites with other works of graphene–polymer composites.

4. Conclusions

In summary, a facile route to prepare 3DGF by coating GNPs on PU foam and then removing PU at 700 °C has been demonstrated. The unique structure endows the high-throughput transport of phonon and electron, resulting in excellent thermal and electrical conductivities for GF/epoxy composites. The thermal conductivity of GF/epoxy composite with 5 wt% filler content was raised up to 1.52 W mK⁻¹, an 8.5-fold enhancement in comparison with that of neat epoxy. The electrical conductivity of GF/epoxy composite reached 3.8 × 10⁻² S cm⁻¹, which increase by 12 orders of magnitude than neat epoxy, from the region of the insulator to a semiconductor. This new 3DGF filler is expected to be useful as effective materials for applications in thermal management, electromagnetic interference shielding, and other electrical devices.

Acknowledgements

The authors are grateful for the financial support by the National Natural Science Foundation of China (No. 51303034, 51573201), Natural Science Foundation of Ningbo (No. Y40307DB05) and International Science and Technology Cooperation Program of Ningbo (No. 2015D10003) for financial support. We also thank the Chinese Academy of Science for Hundred Talents Program, Chinese Central Government for Thousand Young Talents Program, and The Key Technology of Nuclear Energy (CAS Interdisciplinary Innovation Team, 2014).

References

1. S. Yu, J. W. Lee, T. H. Han, C. Park, Y. Kwon, S. M. Hong and C. M. Koo, ACS Appl. Mater. Interfaces, 2013, 5, 11618–11622.
2. K. Pashayi, H. R. Fard, F. Lai, S. Iruvanti, J. Plawsky and T. Borca-Tasciuc, J. Appl. Physiol., 2012, 111, 104310.
3. T. Giang, J. Park, I. Cho, Y. Ko and J. Kim, Polym. Compos., 2013, 34, 468–476.
4. A. Yu, P. Ramesh, M. E. Itkis, E. Bekyarova and R. C. Haddon, J. Phys. Chem. C, 2007, 111, 7565–7569.
5. W. Dai, J. Yu, Y. Wang, Y. Song, F. Alam, K. Nishimura, C. Lin and N. Jiang, J. Mater. Chem. A, 2015, 3, 4884–4891.
6. G. W. Lee, M. Park, J. Kim, J. I. Lee and H. G. Yoon, Composites, Part A, 2006, 37, 727–734.
7. Z. Gao and L. Zhao, Mater. Des., 2015, 66, 176–182.
8. M. J. Assael, K. D. Antoniadis and D. Tzetzis, Compos. Sci. Technol., 2008, 68, 3178–3183.
9. S. Choi, H. Im and J. Kim, Composites, Part A, 2012, 43, 1860–1868.
10. P. W. May, R. Portman and K. N. Rosser, Diamond Relat. Mater., 2005, 14, 598–603.
11. Y. Wu and Z. Yu, Compos. Sci. Technol., 2015, 107, 61–66.
12. J. Zhao, F. Du, W. Cui, P. Zhu, X. Zhou and X. Xie, Composites, Part A, 2014, 58, 1–6.
13. T. Zhou, X. Wang, X. Liu and D. Xiong, Carbon, 2010, 48, 1171–1176.
14. A. A. Balandin, S. Ghosh, W. Z. Bao, I. Calizo, D. Teweldebrhan, F. Miao and C. N. Lau, Nano Lett., 2008, 8, 902–907.
15. C. Lee, X. D. Wei, J. W. Kysar and J. Hone, Science, 2008, 321, 385–388.
16. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, M. I. Katsnelson, I. V. Grigorieva, S. V. Dubonos and A. A. Firsov, Nature, 2005, 438, 197–200.
17. Y. B. Zhang, Y. W. Tan, H. L. Stormer and P. Kim, Nature, 2005, 438, 201–204.
18. Y. S. Jun, S. Sy, W. Ahn, H. Zarrin, L. Rasen, R. Tjandra, B. M. Amoli, B. Zhao, G. Chiu and A. Yu, Carbon, 2015, 95, 653–658.
19. A. Yu, P. Ramesh, X. Sun, E. Bekyarova, M. E. Itkis and R. C. Haddon, Adv. Mater., 2008, 20, 4740–4744.
20. X. Zhang, K. K. Yeung, Z. Gao, J. Li, H. Sun, H. Xu, K. Zhang, M. Zhang, Z. Chen, M. M. F. Yuen and S. Yang, Carbon, 2014, 66, 201–209.
21. Z. P. Chen, W. C. Ren, L. B. Gao, B. L. Liu, S. F. Pei and H. M. Cheng, Nat. Mater., 2011, 10, 424–428.
22. M. T. Pettes, H. Ji, R. S. Ruoff and L. Shi, Nano Lett., 2012, 12, 2959–2964.
23. M. Zhou, T. Lin, F. Huang, Y. Zhong, Z. Wang, Y. Tang, H. Bi, D. Wan and J. Lin, Adv. Funct. Mater., 2013, 23, 2263–2269.
24. N. Li, Z. P. Chen, W. C. Ren, F. Li and H. M. Cheng, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 17360–17365.
25. Z. Chen, C. Xu, C. Ma, W. Ren and H. M. Cheng, Adv. Mater., 2013, 25, 1296–1300.
26. Y. Tao, D. Kong, C. Zhang, W. Lv, M. Wang, B. Li, Z.-H. Huang, F. Kang and Q. H. Yang, Carbon, 2014, 69, 169–177.
27. L. Zhang and G. Shi, J. Phys. Chem. C, 2011, 115, 17206–17212.
28. M. A. Worsley, S. O. Kucheyev, H. E. Mason, M. D. Merrill, B. P. Mayer, J. Lewicki, C. A. Valdez, M. E. Suss, M. Stadermann, P. J. Pauzauskie, J. H. Satcher Jr, J. Biener and T. F. Baumann, Chem. Commun., 2012, 48, 8428–8430.
29. E. Singh, Z. Chen, F. Houshmand, W. Ren, Y. Peles, H. M. Cheng and N. Koratkar, Small, 2013, 9, 75–80.
30. X. Cao, Y. Shi, W. Shi, G. Lu, X. Huang, Q. Yan, Q. Zhang and H. Zhang, Small, 2011, 7, 3163–3168.
31. Y. C. Yong, X. C. Dong, M. B. Chan-Park, H. Song and P. Chen, ACS Nano, 2012, 6, 2394–2400.
32. F. Liu, S. Chung, G. Oh and T. S. Seo, ACS Appl. Mater. Interfaces, 2012, 4, 922–927.
33. H. S. Ahn, J.-W. Jang, M. Seol, J. M. Kim, D.-J. Yun, C. Park, H. Kim, D. H. Youn, J. Y. Kim, G. Park, S. C. Park, J. M. Kim, D. I. Yu, K. Yong, M. H. Kim and J. S. Lee, Sci. Rep., 2013, 3, 1396.
34. B. Hsia, M. S. Kim, L. E. Luna, N. R. Mair, Y. Kim, C. Carraro and R. Maboudian, ACS Appl. Mater. Interfaces, 2014, 6, 18413–18417.
35. G. Tang, Z. G. Jiang, X. Li, H. B. Zhang, A. Dasari and Z. Z. Yu, Carbon, 2014, 77, 592–599.
36. X. Wang, G. Sun and P. Chen, Frontiers in Energy Research, 2014, 2, 33.
37. Z. Y. Yang, L. J. Jin, G. Q. Lu, Q. Q. Xiao, Y. X. Zhang, L. Jing, X. X. Zhang, Y. M. Yan and K. N. Sun, Adv. Funct. Mater., 2014, 24, 3917–3925.
38. F. Yavari, Z. Chen, A. V. Thomas, W. Ren, H. M. Cheng and N. Koratkar, Sci. Rep., 2011, 1, 166.
39. Y. Wang, J. Yu, W. Dai, Y. Song, D. Wang, L. Zeng and N. Jiang, Polym. Compos., 2015, 36, 556–565.
40. J. E. An and Y. G. Jeong, Eur. Polym. J., 2013, 49, 1322–1330.
41. L. Cao, X. Liu, H. Na, Y. Wu, W. Zheng and J. Zhu, J. Mater. Chem. A, 2013, 1, 5081–5088.
42. Y. Huang, Y. Qin, Y. Zhou, H. Niu, Z. Z. Yu and J. Y. Dong, Chem. Mater., 2010, 22, 4096–4102.
43. J. R. Potts, S. Murali, Y. Zhu, X. Zhao and R. S. Ruoff, Macromolecules, 2011, 44, 6488–6495.
44. M. Sangermano, L. Calvara, E. Chiavazzo, L. Ventola, P. Asinari, V. Mittal, R. Rizzoli, L. Ortolani and V. Morandi, Prog. Org. Coat., 2015, 86, 143–146.
45. X. L. Wang, H. Bai, Z. Y. Yao, A. R. Liu and G. Q. Shi, J. Mater. Chem., 2010, 20, 9032–9036.
46. N. Yousefi, X. Lin, Q. Zheng, X. Shen, J. R. Pothnis, J. Jia, E. Zussman and J. K. Kim, Carbon, 2013, 59, 406–417.
47. M. Monti, M. Rallini, D. Puglia, L. Peponi, L. Torre and J. M. Kenny, Composites, Part A, 2013, 46, 166–172.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra27223h

‡ These authors contributed equally to this study.

---