Characterization of carbon structures produced by graphene self-assembly
aDepartment of Basic and Applied Sciences for Engineering, Sapienza University of Rome, Via A. Scarpa 16, Rome, RM 00161, Italy, bDepartment of Chemical Science and Technology – MINASlab, University of Rome `Tor Vergata', Via della Ricerca Scientifica, Rome, RM 00133, Italy, and cCenter for Nanotechnology Applied to Engineering of Sapienza (CNIS), Sapienza University of Rome, Piazzale Aldo Moro 5, Rome, RM 00185, Italy
*Correspondence e-mail: email@example.com
Low-dimensional carbon-based materials, in particular two-dimensional graphenic carbon structures, have been produced from single-walled carbon nanotube disruption using high-shear mixing and/or treatments in sulfonitric acid mixtures at both room and high temperature. Among other two-dimensional graphenic carbon structures, colloidal dispersions of graphenic nanoflakes have been obtained. Different structural arrangements, resulting from the reorganization of carbon because of the disruption procedures applied, were observed through selected area electron diffraction (SAED) and through reflection high-energy electron diffraction (RHEED) analyses coupled to transmission and scanning electron microscopy observations. Such combined investigations in the real and reciprocal space provided structural information at the nanoscale on the clustering of graphene layers in nanoplatelets or/and on their assembly into highly ordered (single-crystal) nanosheets. Furthermore, a different carbon phase exhibiting an orthorhombic cell with Cmma symmetry has been detected by SAED and RHEED analyses. In addition, a variety of self-assemblies of hexagonal basal planes have been observed to occur as the result of their different rotational and/or translational stacking faults. Overall, the reported results contribute to define the conditions for a controlled self-assembly of graphene-based structures with tailored dimensions, which is an important technological challenge, as their structure at the nanoscale dramatically affects their electrical properties.
Nowadays, the shaping of two-dimensional materials into predefined structures is a subject of intense investigation aimed at the control of the effects produced by confinement of electrons. The sizes of such structures decrease down to values comparable to the wavelength of an electron at the Fermi level. Moreover, significant enhancements of surface-related effects can be obtained through an increase of the number of surface atoms able to interact with different chemical and structural entities.
In this context, graphitic carbon materials confined into a two-dimensional nanometric scale represent an extremely important class of technological materials. The interesting chemical and physical properties of low-dimensional graphites (Liao et al., 2011), as well as the capability of modulating such properties by tailoring at the nanoscale the shape and size of the carbon structures (Hernandez et al., 2008), have led to the exploration of a number of potential applications of such light, stiff and flexible materials. Among the applications, two-dimensional C atoms have been proposed as building block components for nanoelectromechanical systems and as conducting channels in complementary metal-oxide-semiconductor nano-electronic devices (Rümmeli et al., 2011).
However, the control of the routes for the fabrication of specific nanostructures with different dimensionalities is still a challenging and complex task. Various chemical approaches and self-assembly processes may produce graphitic nanostructures with shape and dimension designed for specific applications. One of the most feasible and widely used methods to produce graphitic layered materials is the unzipping of carbon naontubes (Kosynkin et al., 2009; Huang et al., 2013; Wei et al., 2013), a process which can produce a large variety of two-dimensional nanostructures, such as nanostrips (Wohlthat et al., 2010), nanoplatelets (Bin & Wei-Hong, 2011) and graphene nanoribbons (Terrones, 2009). Such nanocarbon forms are of great scientific interest, in particular for the nature of edge dislocations and the appearance of defective dangling bonds in carbon networks. As theoretically reported for graphene (Neto et al., 2009), the conductivity decreases exponentially moving from the edge into the bulk material. Moreover, the charge density distribution localized on the edge sites at the Fermi level is strongly dependent on the type of the terminations, zigzag or armchair edge sites both being possible (Ihn et al., 2010; Enoki, 2012).
The real technological applications of such nanostructures strongly depend on the evolution of the electronic structure, which is sensitive to the crystallographic orientation of the edges (Rossi et al., 2011).
Here, we focus our investigation on the structural characterization of self-assembled graphenic layers produced by various unzipping procedures. The study of low-dimensional and complex C aggregations is a long-standing scientific problem, in which interest is still growing because of the wide range of potential applications of such materials. After the opening of rolled nanostructures, such as nanotubes, their self-assembling re-aggregation produces graphitic flakes. These do not possess a long-range translational or orientational order, but rather have some degree of short- and medium-range order acting as seeding sites for self-assembly. The resulting structures are substantially different from those of the `ideal' crystalline systems, where long-range ordering is established by periodic stacking of fundamental building blocks, known as unit cells.
In order to study the local nanostructure of self-assembled graphene-like layers produced from single-walled carbon nanotubes (SWCNTs) using various unzipping procedures, transmission electron microscopy (TEM), scanning electron microscopy (SEM) and reflection high-energy electron diffraction (RHEED) (Rossi et al., 2011) have been used to characterize the morphology and to identify crystalline features. Detailed internal structural investigation has been achieved by selected area electron diffraction (SAED) combined with pattern simulation.
The starting material for the preparation of the two-dimensional graphitic structures was SWCNT powder (purchased from Carbolex), which was submitted to different physico-chemical treatments. A first batch of purified nanotubes was dispersed in dimethylformamide (DMF, 1 mg ml−1) by high-shear mixing at room temperature for 1 h (sample NP1). A second batch was treated with a solution of 60% HNO3:95.0–98.0% H2SO4 in a 1:3 volume ratio (1 mg ml−1) at room temperature for 2 h. The powder was then washed till neutrality, filtered and dried in an oven. After that, the dried powder was dispersed in DMF (1 mg ml−1) by high-shear mixing at room temperature for 1 h (sample NP2). Finally, a third batch was left to react in a solution of 60% HNO3:95.0–98.0% H2SO4 in a 1:3 volume ratio (1 mg ml−1) at 363 K for 2 h. Then, the reaction solution was cooled down to room temperature. The powder was washed till neutrality, filtered and dried in an oven. The final powder was dispersed in DMF (1 mg ml−1) by high-shear mixing at room temperature for 1 h (sample NP3).
2.2. Electron microscopy characterizations
The samples were deposited by drop casting on Si substrates for SEM and RHEED analyses and by dipping, for a very short time, holey carbon grids into the sulfonitric mixtures for TEM and SAED analyses. Specific ultrasound-based methodologies have been set up in order to prepare the samples for the different transmission and diffraction electron microscopy analyses.
The SEM investigations were performed using a field emission scanning electron microscope (Hitachi model S-4000). TEM and SAED analysis were carried out in a Hitachi model H-7100 transmission electron microscope, setting the electron beam at 125 keV. RHEED observations were performed at 60 keV on an electron optics column (AEI EM6G), equipped with a high-resolution diffraction stage.
Fig. 1 shows a typical electron diffraction pattern (EDP) obtained through the RHEED technique used to probe the NP1 sample. The EDP exhibits a number of detected rings much lower than would be expected for graphite crystallites. No diffraction rings were observed from the planes (hkl) with l ≠ 0, with the exception of the (002) planes, where the (00l) planes intercept perpendicularly the crystallographic c axis (or stacking axis of graphene nanosheets). Such planes were associated with a slightly broadened ring.
The broadening of the (002) diffraction ring, with the concomitant absence of its higher-order diffraction (004, 006 etc.) rings, indicates a loss of periodicity along the c-axis crystallographic direction of the hexagonal lattice structure (Fig. 1a). The deviations from a perfect single crystal are indicated by
(i) an interlayer distance that is not perfectly constant between the hexagonal basal planes;
(ii) rotational and/or translational stacking faults of the hexagonal basal planes;
These self-assembled graphitic platelets can be considered as an ordered stacking of n-layer graphene. The deviation associated with the bending effect decreases as the number of graphene layers increases. From the experimental RHEED patterns (one of which is reported in Fig. 1a), the measured interplanar distance of the (002) basal planes has a value of 3.43 (3) Å. This value is larger than the interplanar distance along the c axis of the typical graphite structure, but it is similar to that of the disordered form, schematically drawn in Fig. 1(b), well known in the past as `turbostratic graphite' (Li et al., 2007; Rossi et al., 2005; Terranova et al., 1994) and currently called multi-layered graphene.
For sample NP2, TEM bright-field image observation shows self-aggregation of graphite nanoribbons with layers anisotropically bent at the edges, where the degree of bending depends on the extension area of the graphitized flakes (Fig. 2a).
The inset of Fig. 2(a) shows evident boundaries between assembled nanoribbons of graphite, forming nano-mosaic arrangements of two-dimensional single pieces that differ in shape because of the variable etching directions during the unzipping process. Furthermore, the formation of high-contrast platelets can be noticed; these are probably due to the anisotropic bending of the graphene layers, which can be rationalized in terms of a minimization of the pristine surface energy.
The mesoscale EDP of these self-assembled platelets, obtained from RHEED analysis, shows the presence of polycrystalline material with a P63/mmc hexagonal symmetry (Fig. 2b). The fine structure is characterized by a perfect superposition of diffraction spots coming from random orientation of different crystallites, as shown in the inset of Fig. 2(a). Since the RHEED resolution in the reciprocal space is better than the resolution of the SAED method in transmission conditions (Rossi et al., 2011), accurate analysis of the RHEED pattern enables one to detect a further internal weak ring close to the undiffracted electron beam.
From such a diffraction ring an interplanar spacing of about 5.65 Å was measured, indicating the existence of an anomalous carbon phase. It is worth noticing that this ring detected in the RHEED pattern cannot be assigned to any reflection of the Si substrate supporting the sample, because, on the basis of the standard values in the PDF card (PDF No. 00-027-1402; JCPDS No. 27-1401 1976; International Centre for Diffraction Data, http://www.icdd.com/ ), the measured value of the diffraction ring does not belong to any Si crystalline phase. A TEM image from a wrinkled and folded area is displayed in Fig. 2(c). A part of this area is shown at higher magnification in Fig. 2(d). The corresponding SAED pattern in transmission configuration, shown in Fig. 2(e), reveals the presence of highly ordered single-crystal flakes with a relatively large surface [some thousands of nm2 from Fig. 2(d)].
The regular diffraction spots, ascribable to the presence of a single crystal, have been reproduced by EDP simulation. We started considering that these samples could be characterized by a hexagonal superstructure and we simulated the experimental EDP by considering the typical P63/mmc hexagonal symmetry.
The simulated pattern for the typical hexagonal graphite crystal structure was found to fit only roughly the experimental SAED pattern obtained from the NP2 sample. Therefore, we considered for the simulation another phase of graphite, having an orthorhombic superstructure (American Mineralogist Crystal Structure Database, amcsd 0013979, Fayos, 1999; Downs & Hall-Wallace, 2003) (Fig. 2f). A crystal model was built up considering the space group Cmma using the theoretical unit-cell parameters calculated by Fayos and Guinea (Fayos, 1999; Guinea, 2010) processed through the electron diffraction simulation software JEMS (Stadelmann, 2004). Figs. 2(f) and 2(g) display the electron diffraction simulation of graphitic layers oriented along the  zone axis parallel to the electron beam and the corresponding atomic structure model projected along the same zone axis, respectively. It is worth pointing out that the simulated EDP of an orthorhombic superstructure agrees with the extra experimental ring detected by the RHEED technique displayed in Fig. 2(b). The measured inner diffraction ring has an interplanar spacing of d001 = 5.65 Å, corresponding to the higher-intensity diffraction peak of the simulated orthorhombic phase (Fayos, 1999).
Moreover, the two experimental strong and broadened rings shown in Fig. 2(e), overlapping the single-crystal diffraction signal, indicate the presence of small graphite crystals without any preferential orientation and with larger interplanar distances among the graphene layers. Since the precision of the d-spacing determinations decreases proportionally to the crystal dimension and to the broadening effect of the ring, these experimental diffraction rings cannot be attributed to any crystalline phases. However, the RHEED analysis of Fig. 2(b) helps us to establish the existence of P63/mmc hexagonal symmetry characterizing the NP2 sample.
Overall, the structural investigation of NP2 samples using two different electron diffraction techniques has shown the presence of crystal packing constituted by graphite bilayers stacked along the c axis of an orthorhombic symmetry. According to theoretical studies (Fayos, 1999), the stacked bilayers themselves interact only through weak van der Waals interactions, determining a slight shifting regarding the ab crystallographic plane with an inter-biplanar distance of about 3.20 Å (Fig. 2h). Within each single bilayer, one nonplanar layer constituted of hexagonal benzene rings deformed to a chair conformation is bound to a similarly deformed layer in a parallel stacking configuration by strong covalent interactions (Khaliullin et al., 2011).
Finally, the NP3 sample, prepared through the same procedure as NP2 but treated at a higher temperature, has been preliminary observed through the SEM technique, showing clustering of graphite particles (Fig. 3a). To better investigate the morphological aggregation of the reassembled unzipped SWCNTs, TEM analysis has been performed on a large area, showing a number of multilayer nanoribbons and graphite platelets turbostratically stacked with several edges scrolled up or slightly folded (Fig. 3b). The morphological analysis, combining the complementary information given by the TEM and SEM techniques, indicates that the unzipping process performed at high temperature gives rise to graphitic micro-clusters with relatively smooth surfaces.
The unzipping through high-T treatment leads to uniform graphitic layers, reducing the formation of wrinkles around the step edges with a remarkable increase in the surface area and thickness of the graphite islands.
The structural differences due to the increase of temperature can be rationalized on the basis of the induced annealing processes and of a solid-state epitaxial regrowth starting from localized seeds with pristine nano and sub-nano sizes.
The structural analysis of the platelets characterizing the NP3 samples has been performed by the RHEED technique (Fig. 3c). The measurements of the d spacings corresponding to the concentric rings confirm the presence of polycrystalline material with a P63/mmc hexagonal symmetry. Also in this case, the (004) second-order diffraction ring of the (00l) planes is still absent, but those corresponding to the first-order (002) and third-order (006) rings are instead present. Following the same considerations as before, these observations undoubtedly confirm the presence of polycrystalline turbostratic graphite, also revealed in the NP1 sample, with lattice parameters practically coincident with those of the `reference' graphite reported in Fig. 1(a). A model of a disordered self-assembling of graphitic nanoribbons is sketched in Fig. 3(d).
The structural results achieved through the electronic diffraction analyses of the three different samples (NP1–NP3) can be summarized as follows.
For NP1 samples only the first-order diffraction of the hexagonal phase was detected, whereas for the NP2 samples prepared following a different procedure (but at the same room temperature), both first- and second-order diffraction has been detected, revealing an increased stacking order along the c axis. This could be explained by the formation of crystalline bilayers of the orthorhombic phase within the hexagonal phase of the NP2 graphitic platelets.
The NP3 sample, prepared with the same procedure as NP2 but at higher temperature, is characterized by a decreased order along the stacking c axis, as revealed by the missing second-order diffraction signal. In addition to the thermal-stress-induced process, the unstable phase of graphitic bilayers has not been detected. This is in agreement with the presence of a structural phase transition of orthorhombic graphite into the classical hexagonal graphite phase during annealing and recrystallization treatments.
Overall, the analysis of the EDPs of the NP1 and NP3 samples indicates that the platelets are `expanded' graphite, with a larger interplanar distance between the stacking graphene layers and slightly reduced interplanar distances of the transverse planes. It is worth noticing that no equatorial oscillations or lines, no short diffraction lines located in the vertical or horizontal direction, and no diffraction lines with arc shapes, all signals due to unprocessed SWCNTs, have been detected in any electron diffraction patterns.
The presented results indicate that the assembly processes leading to the formation of two-dimensional graphitic platelets can be tailored by modulating the methodology of unzipping SWCNTs in order to selectively produce different forms of graphene-like layers.
In all the samples produced by the SWCNT unzipping, we were able to detect different structural arrangements of conventional graphite layers. In the case of the NP2 sample, an orthorhombic graphite bilayer structure was experimentally revealed for the first time, and this can be considered a further interesting outcome of the present research.
The reported experimental data coherently lead us to think that temperature plays an important role in defining the final structure, inducing a decrease of the bending effect of the aggregating nanoribbons in a nanoplatelet's shape, with consequent formation of turbostratic arrangements.
As demonstrated by this case study, in order to solve the fine structure of low-dimensional carbon entities, it is essential to couple TEM analysis of the flake edges with electron diffraction investigations. Effort has been directed toward the use of the electron diffraction tool, because it is very sensitive to any small variations in the periodic structures or lattice imperfections, allowing one to investigate in detail phase transition phenomena such as those occurring in graphene-based systems. Layers with different morphological/structural features are found to self-assemble following different architectures, from polycrystalline aggregates to highly self-oriented mosaic-like structures that can further evolve towards single-crystal platelets.
The variety of these two-dimensional nanostructures, which are exciting objects from both scientific and technological points of view, stresses the importance of exploring the fundamental mechanisms for controlling the transitions from isolated layers to nano- and meso-crystalline organizations.
The unique properties of two-dimensional C atoms make them attractive for many potential applications (Wu et al., 2010), in many technology fields, from aerospace to medicine. Most of the applications are still at the exploratory stage, but for some of them the time to market is not likely to be long. As an example, improvements of the electrochemical performance of lithium ion batteries have been recently achieved using graphenic ribbons, exhibiting a high reversible capacity and excellent cycle stability (Bhardwaj et al., 2010).
It is expected that the identification of cooperative mechanisms acting for the self-assembly of two-dimensional carbon systems could help in opening innovative crystallization pathways for materials synthesis and device fabrication.
Bhardwaj, T., Antic, A., Pavan, B., Barone, V. & Fahlman, B. D. (2010). J. Am. Chem. Soc. 132, 12556–12558. Web of Science CrossRef CAS PubMed
Bin, L. & Wei-Hong, Z. (2011). J. Mater. Sci. 46, 5595–5614.
Downs, R. T. & Hall-Wallace, M. (2003). Am. Mineral. 88, 247–250. CAS
Enoki, T. (2012). Phys. Scr. T146, 014008. Web of Science CrossRef
Fayos, J. (1999). J. Solid State Chem. 148, 278–285. Web of Science CrossRef CAS
Guinea, F. (2010). New J. Phys. 12, 083063. Web of Science CrossRef
Hernandez, Y. et al. (2008). Nat. Nanotechnol. 3, 563–568. Web of Science CrossRef PubMed CAS
Huang, C., Grobert, N., Watt, A. A., Johnston, C., Crossley, A., Young, N. P. & Grant, P. S. (2013). Carbon, 61, 525–536. Web of Science CrossRef CAS
Ihn T., Güttinger J., Molitor F., Schnez S., Schurtenberger E., Jacobsen A., Hellmüller S., Frey T., Dröscher, S., Stampfer, C. & Ensslin, K. (2010). Mater. Today, 13, 44–50.
Khaliullin, R. Z., Eshet, H., Kühne, T. D., Behler, J. & Parrinello, M. (2011). Nat. Mater. 10, 693–697. Web of Science CrossRef CAS PubMed
Kosynkin, D. V., Higginbotham, A. L., Sinitskii, A., Lomeda, J. R., Dimiev, A., Price, B. K. & Tour, J. M. (2009). Nature, 458, 872–876. Web of Science CrossRef PubMed CAS
Li, Z., Lu, C., Xia, Z., Zhou, Y. & Luo, Z. (2007). Carbon, 45, 1686–1695. Web of Science CrossRef CAS
Liao, K. et al. (2011). Carbon, 49, 2862–2868. Web of Science CrossRef CAS
Neto, A. H., Guinea, F., Peres, N. M. R., Novoselov, K. S. & Geim, A. K. (2009). Rev. Mod. Phys. 81, 109–162.
Rossi, M., Manno, D., Orlanducci, S., Serra, A. & Terranova, M. L. (2011). Encyclopedia of Nanoscience and Nanotechnology, edited by H. S. Nalwa, Vol. 13, pp. 375–425. Valencia: American Scientific Publishers.
Rossi, M., Terranova, M., Piccirillo, S., Sessa, V. & Manno, D. (2005). Chem. Phys. Lett. 402, 340–345. Web of Science CrossRef CAS
Rümmeli et al. (2011). Adv. Mater. 23, 4471–4490.
Stadelmann, P. A. (2004). Java Electron Microscopy Software (JEMS), http://cimewww.epfl.ch/people/stadelmann/jemsWebSite/jems.html .
Terranova, M., Rossi, M., Sessa, V. & Vitali, G. (1994). Solid State Commun. 91, 55–58. CrossRef CAS Web of Science
Terrones, M. (2009). Nature, 458, 845–846. Web of Science CrossRef PubMed CAS
Wei, D., Xie, L., Lee, K. K., Hu, Z., Tan, S., Chen, W., Sow, C. H., Chen, K., Liu, Y. & Wee, A. T. (2013). Nat. Commun. 4, 1374. Web of Science CrossRef PubMed
Wohlthat, S., Reimers, J. R. & Hush, N. S. (2010). Phys. Rev. B, 81, 195125. Web of Science CrossRef
Wu, Y. H., Yu, T. & Shen, Z. X. (2010). J. Appl. Phys. 108, 071301. Web of Science CrossRef
© International Union of Crystallography. Prior permission is not required to reproduce short quotations, tables and figures from this article, provided the original authors and source are cited. For more information, click here.