An edge-decorated submicron reduced graphite oxide nanoflake and its composites with carbon nanotubes for transparent conducting films

Abstract

We introduce a novel kind of submicron-sized reduced graphite oxide (μRGO) for the preparation of a transparent conducting film. The special NMP-decorated structure and the small size (∼400 nm) of the μRGO enable us to prepare large-area thin films composed of multiwalled carbon nanotubes (MWNTs) and the functionalized μRGO can fill up the pores of the spiderwork network of the MWNTs partly. This direct assembling composite film possesses a sheet resistance of 152 Ω per square at 84% transmittance.

---

Transparent conducting films (TCFs) have been one of the key components of various optoelectronic devices such as liquid crystal displays (LCD), solar cells, and organic light-emitting diodes (OLEDs).^1,2 Carbon-based TCFs, carbon nanotubes (CNTs) and graphene in particular, have received increasing attention due to their outstanding mechanical, optical, and electrical properties.^2–8 However, for pure graphene or CNTs films, excellent electrical conductivity and light transmittance are difficult to achieve at the same time.^8,9 The main reasons include the presence of voids in spiderweb-like network structure of CNTs⁹ and high intersheet contact resistance among small-size graphenes.⁸ In view of the above-mentioned facts, some fruitful progress in the research of hybrid TCFs utilizing CNTs and graphene together has been reported in recent years.^9–13 For example, Li's team⁹ used chemical vapor deposition (CVD) method providing a hybrid film based on the affinity binding between CNTs and graphene sheets, which presented a low resistance (735 Ω per square) with a high transmission of 90%. Hong's work¹⁰ reported a similarly layer-by-layer assembled film by solution-based method possessing a sheet resistance of 8 kΩ per square with a transmittance of 81%. It was suggested that the electrostatic interactions between CNTs and RGO provided an additional flexibility and relative stability. Moreover, Tung¹² prepared a RGO/CNTs film by solution-based method. This film showed a sheet resistance of 240 Ω per square at 86% transmittance and was built on weak conjugated and van der Waals interaction between CNTs and graphene.

In this paper, we provide a novel type of TCF consisting of functionalized submicron-sized reduced graphite oxide (μRGO) and multiwalled carbon nanotubes (MWNTs). This film can display excellent conductivity and transmittance because: (i) μRGO has a so small size that can fill up the pores of MWNTs partly. (ii) An edge-decorated structure of μRGO make MWNT prefer to contact with the edge of μRGO and strong hydrogen bonds form between μRGO and MWNTs. The edge-decorated structure is based on grafting of N-methyl-pyrrolidone (NMP) molecules onto the edge of μRGO by ball-milling. Therefore hydrogen bonds form between NMP molecules on the μRGO flakes and –COOH or –OH groups on the MWNTs. Owing to these two reasons, our film shows a moderately low sheet resistance of 152 Ω per square at 84% transmittance.

We introduce the submicron size and the edge-decorated structure of the μRGO first as they are important influencing factors for the hybrid film. In a typical experiment, ball milling was carried out in the presence of reduced graphite oxide (RGO) and NMP solvent. Detailed experimental conditions and optimization of manufacturing processes can be found in ESI.† We investigate the submicron size of the μRGO by transmission electron microscopy (TEM), high resolution TEM (HR-TEM) and atomic force microscope (AFM). Fig. 1A shows the stacks of dozens of micrometer sized pristine RGO flakes. The ball-milling process effectively decreases their sizes to submicron and exfoliates the heavy flakes to monolayer or few-layer (Fig. 1B–F). In a HR-TEM image of a transparent thin flake (Fig. 1D), the inset highlights the corresponding selected area electron diffraction (SAED) pattern to further identify monolayer graphene.^14–17 The spots {01̄00} and {1̄010} are more intense than spots {2̄110} and {12̄10}, which indicates the presence of graphene monolayers in μRGO dispersions. In order to visually observe the layers of these μRGO, HR-TEM images of μRGO embedded in epoxy resin slice are provided and the lattice fringe images represent the atomic carbon layers.^18–20 If we select one of the areas (Fig. 1E), we can clearly see the four-layer flake. Thus we can estimate the thickness of one layer is about 0.75 nm, which is consistent with our AFM measurements (0.78–0.86 nm, Fig. 1F) and reported results elsewhere (0.60–1.20 nm).^4,21 Furthermore, the statistical distributions of lateral sizes (Fig. 1G) and thicknesses (Fig. 1H) of μRGO flakes are collected from TEM and HT-TEM images, respectively. The results show that more than 68% of these flakes are in the lateral size range of 200–600 nm and nearly 78% in the thickness range of 1–5 layers.

Fig. 1 (A–C) TEM images of heavy stacks in RGO dispersion (A), small area as well as thin flakes in μRGO dispersion (B) and a multilayer μRGO (C). (D) HR-TEM image of a typical monolayer. Inset: SAED pattern taken from a monolayer. (E) HR-TEM image of μRGO embedded in epoxy resin slice to determine the number of layers per flake. (F) AFM image of μRGO flakes deposited on a mica substrate. Below is height profile along the lines shown above, the height difference between the green arrows is 0.786 nm and between the red arrows is 0.86 nm. (G and H) Histogram of lateral size distribution (G) and number of layers per flake (H) for μRGO dispersions.

We investigate the edge-decorated structure of the resulted μRGO by various characterizations, including X-ray photoelectron spectroscopy (XPS), thermogravimetric analysis (TGA), Raman spectroscopy, X-ray diffraction (XRD), Fourier Transform Infrared Spectroscopy (FTIR) and density functional theory (DFT) calculations.

As has been demonstrated by Jeon,^22,23 active carbon species (such as carbon dangling bonds) can be generated by cleavages of the graphitic C–C bonds in a long-time ball-milling process, thus leaving lots of defects on the edges of μRGO. Those defects can readily react with small molecular solid and gas (such as dry ice and CO₂ gas). During the past years our group has reported polystyrene chains can graft onto graphene sheets via non-covalent π–π interactions in ball-milling process.²⁴ Here, we suspect that large numbers of NMP molecules are coupled with μRGO flakes by stronger covalent (chemical) bonds.

In order to investigate the structure of NMP-decorated μRGO we take Raman spectroscopy and XRD measurements.²⁵ It is found that a new edge-decorated peak²² appears at 1130 cm⁻¹ (Fig. 2A), and the intensity ratios between I(D)/I(G) are very different, in the range of 1.6–2.6 for RGO and 2.2–3.4 for μRGO, respectively (Fig. S4 in ESI†). This large difference suggests μRGO has more defects than RGO. This can be interpreted as the fact that small flakes deliver more edges, or defects. XRD pattern of RGO (Fig. 2B) shows a weak broad peak between 20 and 28°. This broad peak means a Gaussian distribution of d-spacings.²² In contrast, μRGO shows a broader peak and blue shift. We can estimate the (002) interlayer spacing to be 3.89 Å for μRGO, and 3.65 Å for RGO. This difference reconfirms μRGO are further exfoliated to smaller and thinner ones from RGO.

Fig. 2 (A) Raman spectra of μRGO and RGO. The μRGO displays a sp³ peak at 1130 cm⁻¹. (B) XRD patterns of RGO and μRGO. The spectra of μRGO show a broader peak than that of RGO and the broader peak has a blue shift. (C) TGA data of μRGO and RGO obtained from the heating rate of 10 °C min⁻¹ in nitrogen. RGO and μRGO show 30% and 60% weight loss until 1000 °C. (D) FTIR spectroscopy of μRGO and RGO. A set of new peaks arise in the spectra of μRGO indicating that NMP molecules have been grafted on to the μRGO flakes. (E and F) XPS spectra of the C 1s peaks of RGO and μRGO. The spectrum of μRGO has a prominent band at 288.1 eV representing C=O species are introduced to μRGO flakes.

Further evidence for the origin of edge decoration comes from the XPS spectroscopic measurements. As shown in the C 1s XPS spectrum of RGO (Fig. 2E), the bands at 284.7, 286.2 and 289 eV can be assigned to C–C, C–O, and C=O species, respectively.²⁶ While in the C 1s XPS spectrum of μRGO (Fig. 2F), the band at 288.1 eV representing C=O species is more prominent, indicating oxygen-containing groups are attached to the μRGO flake. The XPS spectra also shows the C/O ratio of these two material, 5.56 for RGO and 3.75 for μRGO, respectively. Of the most importance is the increase of N content (1.75% for RGO and 10% for μRGO), implying that NMP molecules have been introduced to the μRGO flakes (Fig. S5 in ESI†). This can be further verified by our TGA data: μRGO shows a 60% weight loss up to 1000 °C, whereas RGO exhibits a only 30% loss (Fig. 2C).

Also FTIR spectroscopy (Fig. 2D) is employed to confirm NMP grafting. After microwave-reduction, most of the characteristic features of GO disappear, indicating that a large proportion of oxygen-containing groups had been removed. After ball-milling, however, a set of new peaks arose, which could be attributed to (N–CH₃) 1114 cm⁻¹, (C–N) 1291 cm⁻¹ and (C=O) 1715 cm⁻¹ stretches of grafted NMP^27–29 (the detals of FTIR can be seen in ESI†).

While the XPS and FTIR spectra have given evidence that NMP molecules have grafted on to the μRGO flakes, the type of interactions between NMP and μRGO flakes is further characterized by DFT calculations. We calculated the adsorption energy between adsorbates (NMP molecules, hydrogen free radical and NMP free radical) and μRGO edge, which have been published in the ESI† of the previous work.³⁰ The results show that the interactions between them are ascribed to the physisorption, H-bonding, and chemical bonds, as shown in the Fig. 3. There exists physisorption between NMP molecule and hydrogen-terminated carbon bond (Fig. 3A) or dangling carbon bond (Fig. 3B) because the adsorption energy is only about 0.048 or 0.126 eV, respectively. Hydrogen bonds exist between NMP molecules and oxygen-containing groups (Fig. 3C and D), as the energies at the μRGO edge are estimated to be 0.617 and 0.493 eV. The strongest force (i.e., chemical bonds) form between NMP˙ and carbon dangling bonds (Fig. 3F). The adsorption energy is up to 3.898 eV.

Fig. 3 Different kind of adsorptions of NMP molecule, H˙and NMP˙on edge of μRGO flakes. The blue lines represent carbon dangling bond.³⁰

The above results enlighten us about the experimental design to prepare large-area thin films by adding these edge-decorated μRGO to fill the pores of MWNTs. Our TCFs were produced by simple solution-casting. The μRGO/MWNTs hybrid dispersion (the weight ratio of these two components is kept at 1 : 1, see ESI† for details) was deposited onto the quartz substrates (1.5 × 1.5 cm²). The substrates were completely covered with the hybrid dispersion, allowed to spin coated at 100 rpm for 2 seconds. Then the substrates were dried at 90 °C in a convection oven for 30 min to obtain the transparent conducting film. Fig. 4A shows optical transmittance of μRGO-MWNTs films. Depending on the composite contents, the high optical transmittances at 550 nm are ranging from 63 to 84%. By plotting the sheet resistance of the hybrid films as a function of optical transmittance at 550 nm (Fig. 4B), we find that the relationship between sheet resistance and optical transmittance is in accord with equation.¹³

Fig. 4 (A) Optical transmittance of μRGO/MWNTs hybrid films. (B) Sheet resistance and transmittance measured at 550 nm for transparent conductive film. The inset is the μRGO/MWNT hybrid film on quartz.

This trend is consistent with pure CNT or graphene films.^7,8 We measure the thickness of the film with 84% transmittance by AFM (Fig. 5). The results show the average thickness of μRGO/MWNTs film is 16 nm. The TCF shows a low sheet resistance of 152 Ω per square at 84% transparency. Compared with CNTs–graphene composite films requiring complex or delicate processing,^9–13 our film is produced by a simple method and shows similar transparency and electrical conductivity.

Fig. 5 AFM images (A and B are10 μm × 10 μm, C is 1 μm × 1 μm) of μRGO/MWNTs film. The average thickness of the film is 16 nm.

Such high performance is attributed to the formation of a non-conjugated network by stitching two-dimensional μRGO with one-dimensional MWNTs. The SEM and TEM images can be used to understand generally the morphology of the films. As illustrated in Fig. 6A, the entire film exhibits a rough surface which the μRGO flakes are stitched by a randomly distributed network of MWNTs. As shown in the SEM and TEM images of local area of μRGO/MWNTs composites (Fig. 6B–D), the μRGO flakes are so small that can fill the pores of the net of MWNTs partly. This unique structure enhance optoelectrical properties of μRGO/MWNTs composites.¹³

Fig. 6 Representative SEM (A–C) and TEM (D) images of μRGO/MWNTs hybrid films. (A) The morphology of the network of μRGO/MWNTs composites. (B–D) The morphology of local area of μRGO/MWNTs composites. Note that (i) the μRGO flakes are so small that can fill the pores of the net of MWNTs; (ii) the MWNTs prefer to be located on the edge of μRGO.

Although this work also takes advantage of the synergistic effect of CNTs and RGO, e.g., CNTs are used to bridge the intersheet junctions. However, there exists delicate difference from the existed works associated with the ultralarge RGO and MWNTs hybrid films,^8,31–34 in which the use of ultralarge RGO sheets aims to reduce the number of intersheet tunneling barriers while CNTs act as conductive bridges to connect RGO sheets. Taking advantage of these two mechanisms, this kind of hybrid TCFs can achieve better electrical conductivity. In this work, our aim is to form conductive network by CNTs themselves and fill the pores between CNTs by the smaller-sized RGO in order to further enhance conductivity. The results (Fig. 5C and 6) have demonstrated that the CNTs themselves have formed a conductive network in the prepared films (the weight ratio of μRGO sheets to CNTs was kept at 1 : 1 herein). Then, RGO sheets of small size are used herein, which are prone to be efficient to fill up the pores of spider-web-like network of CNTs. In this way, the small-sized RGO sheets can promote internanotube charge transport. As the size of RGO sheets increases, the RGO sheets probably tend to swell the CNT network, thus increasing the distance between CNTs and the corresponding resistance between CNTs.¹³

In Fig. 6B–D, we also find that the MWNTs prefer to be located on the edge of μRGO instead of on its surface. This indicates that the interactions between μRGO flakes and MWNTs are not π–π conjugate actions. We speculate hydrogen bonds form between MWNTs and μRGO flakes. Because there's a certain content of oxygen-containing groups on MWNTs and NMP moleculars decorated around μRGO flakes, hydrogen bonds can form between NMP molecules and –COOH or –OH groups on MWNTs.

In order to show significant advantage of the structure of μRGO/MWNTs composites, we take a comparing experiment among μRGO/MWNTs composite, GO/MWNTs, pure μRGO and pure MWNTs films prepared by the same method. The transparency and electrical properties of above films are shown in Fig. 7. Obviously, μRGO/MWNTs films show the lowest sheet resistance at the same transmittance. The MWNTs used alone are not dispersible in water, even after sonication for 100 hours. So we can not use them to prepare transparent films (the film is very inhomogeneous). And the pure μRGO films can not reach high conductivity due to many junctions between flakes. When two components are mixed in an appropriate proportion, however, μRGO flakes will carry MWNTs into water. In such way, a well-dispersed composite can be obtained. In addition, a conductive network form between μRGO and MWNTs.

Fig. 7 Function of sheet resistance vs. transmittance changes of μRGO/MWNTs, GO/MWNTs, pure μRGO and pure MWNTs films. Among them, μRGO/MWNTs films show the lowest sheet resistance at the same transmittance.

The transmittance and sheet resistance of GO/MWNTs by the same method are also studied. The GO/MWNTs films can not reach high transmittance (>80%) and very thick GO/MWNTs film only have a average sheet resistance of 1000 Ω per square. There are two reasons for the worse electrical property: (i) when the solvent (water) evaporates at a temperature of 200 °C, we find that GO is reduced and agglomeration occur. This kind of agglomeration of flakes will lead to opacity and in homogeneity of films. But μRGO is stable under the temperature of 300 °C, no agglomeration occur. (ii) The size of GO flakes are above 2 μm (Fig. S7 in ESI†) while μRGO is about 400 nm. We hold the opinion that it is the small-area of μRGO that can fill the voids between nanotube bundles.

Here we must explain that edge-decorated structure is necessary. In other words, we explain why we use μRGO decorated with NMP to composite with MWNTs, instead of RGO with small size. We have used nano graphite to prepare nano RGO by the Hummers oxidation and hydrazine hydrate chemical reduction method, which average sizes are about 40 nm. The nano RGO flakes are severely agglomerated and can not be dispersed well in water. The SEM images (Fig. S8 in ESI†) can be used to illustrate the aggregation. For the composites, the flakes need to be isolated with each other in order to prevent van der Waals force. So in our experiment, NMP edge-decorated structure also plays a part in preventing flaks from aggregation.

Conclusions

In conclusion, we succeed in preparing a kind of NMP-decorated submicron-sized reduced graphite oxide. Based on the strong hydrogen bonds, these functionalized μRGO have filled up the pores of spiderwork network of MWNTs partly. This direct assembling composite present an excellent optoelectrical property. So the composite is used to prepare a film with a very low sheet resistance of 152 Ω per square at transmittance of 84%. We believe that this transparent conducting film is more competent for the practical application and some studies with special requirements. The proposed mechanism can shed some light and play an important role in engineering more complex nano-architectures.

Acknowledgements

Grateful acknowledgments go to Prof. Li Tan (Department of Mechanical & Materials Engineering, University of Nebraska, Lincoln, USA) for his fruitful discussion and careful review. This work was funded by Natural Science Foundation of China (51373059, 50373015) and Science Foundation of Fujian Province (2013H6014).

Notes and references

1. A. Kumar and C. Zhou, ACS Nano, 2010, 4, 11.
2. D. M. Sun, C. Liu, W. C. Ren and H. M. Cheng, Small, 2013, 9, 1188.
3. J. Zhao, S. Pei, W. Ren, L. Gao and H. Cheng, ACS Nano, 2010, 4, 5245.
4. J. Geng, L. Liu, S. B. Yang, S. C. Youn, D. W. Kim, J. S. Lee, J. K. Choi and H. T. Jung, J. Phys. Chem. C, 2010, 114, 14433.
5. C. Y. Su, A. Y. Lu, Y. Xu, F. R. Chen, A. N. Khlobystov and L. J. Li, ACS Nano, 2011, 5, 2332.
6. L. Gomez De Arco, Y. Zhang, C. W. Schlenker, K. Ryu, M. E. Thompson and C. Zhou, ACS Nano, 2010, 4, 2865.
7. B. J. Lee, E. C. Shin and G. H. Jeong, Appl. Phys. A: Mater. Sci. Process., 2012, 107, 843.
8. Q. Zheng, W. H. Ip, X. Y. Lin, N. Yousefi, K. K. Yeung, Z. G. Li and J. K. Kim, ACS Nano, 2011, 5, 6039.
9. C. Li, Z. Li, H. Zhu, K. Wang, J. Wei, X. Li, P. Sun, H. Zhang and D. Wu, J. Phys. Chem. C, 2010, 114, 14008.
10. T. K. Hong, D. W. Lee, H. J. Choi, H. S. Shin and B. S. Kim, ACS Nano, 2010, 4, 3861.
11. S. B. Bon, L. Valentini, J. M. Kenny, L. Peponi, R. Verdeio and M. A. Lopez-Manchado, Phys. Status Solidi A, 2010, 207, 2461.
12. V. C. Tung, L. M. Chen, M. J. Allen, J. K. Wassei, K. Nelson, R. B. Kaner and Y. Yang, Nano Lett., 2009, 9, 1949.
13. P. J. King, U. Khan, M. Lotya, S. De and J. N. Coleman, ACS Nano, 2010, 4, 4238.
14. J. C. Meyer, A. K. Geim, M. I. Katsnelson, K. S. Novoselov, T. J. Booth and S. Roth, Nature, 2007, 446, 60.
15. J. C. Meyer, A. K. Geim, M. I. Katsnelson, K. S. Novoselov, D. Obergfell, S. Roth, C. Girit and A. Zettl, Solid State Commun., 2007, 143, 101.
16. A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov, S. Roth and A. K. Geim, Phys. Rev. Lett., 2006, 97, 187401.
17. Y. Hernandez, V. Nicolosi, M. Lotya, F. M. Blighe, Z. Sun, S. De, I. T. McGovern, B. Holland, M. Byrne, Y. Gun'ko, J. Boland, P. Niraj, G. Duesberg, S. Krishnamurthy, R. Goodhue, J. Hutchison, V. Scardaci, A. C. Ferrari and J. N. Coleman, Nat. Nanotechnol., 2008, 3, 563.
18. J. Chen, M. Duan and G. Chen, J. Mater. Chem., 2012, 22, 19625.
19. W. Zhao, M. Fang, F. Wu, H. Wu, L. Wang and G. Chen, J. Mater. Chem., 2010, 20, 5817.
20. S. F. Lin, F. X. Liu and G. H. Chen, RSC Adv., 2014, 4, 45885.
21. J. Long, M. Fang and G. H. Chen, J. Mater. Chem., 2011, 21, 10421.
22. I. Y. Jeon, Y. R. Shin, G. J. Sohn, H. J. Choi, S. Y. Bea, J. Mahmood, S. M. Jung, J. M. Seo, M. J. Kim, D. W. Chang, L. Dai and J. B. Beak, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 5588.
23. I. Y. Jeon, H. J. Choi, S. M. Jung, J. M. Seo, M. J. Kim, L. Dai and J. B. Baek, J. Am. Chem. Soc., 2012, 135, 1386.
24. H. Wu, W. Zhao, H. Hu and G. Chen, J. Mater. Chem., 2011, 21, 8626.
25. A. C. Ferrari and D. M. Basko, Nat. Nanotechnol., 2013, 8, 235.
26. C. Mattevi, G. Eda, S. Agnoli, S. Miller, K. A. Mkhoyan, O. Celik, D. Mastrogiovanni, G. Granozzi, E. Garfunkel and M. Chhowalla, Adv. Funct. Mater., 2009, 19, 2577.
27. S. Stankovich, R. D. Piner, S. B. T. Nguyen and R. S. Ruoff, Carbon, 2006, 44, 3342.
28. J. Y. Niu, M. L. Wei, J. P. Wang and D. B. Dang, Eur. J. Inorg. Chem., 2004, 160.
29. M. Tabbal, P. Merel, S. Moisa, M. Chaker, E. Gat, A. Ricard, M. Moisan and S. Gujrathi, Surf. Coat. Technol., 1998, 98, 1092.
30. R. Su, S. F. Lin, D. Q. Chen and G. H. Chen, J. Phys. Chem. C, 2014, 118, 12520.
31. Q. Zheng, Z. Li, J. Yang and J. K. Kim, Prog. Mater. Sci., 2014, 64, 200.
32. Q. Zheng, B. Zhang, X. Lin, X. Shen, N. Yousefi, Z. D. Huang, Z. Li and J. K. Kim, J. Mater. Chem., 2012, 22, 25072.
33. C. Y. Su, Y. Xu, W. Zhang, J. Zhao, X. Tang, C. H. Tsai and L. J. Li, Chem. Mater., 2009, 21, 5674.
34. X. Zhou and Z. Liu, Chem. Commun., 2010, 46, 2611.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and supporting results. See DOI: 10.1039/c5ra02704g

---