Controlled oxidation of graphite to graphene oxide with novel oxidants in a bulk scale

Graphene shows excellent properties, such as high intrinsic carrier mobility (200,000 cm²/V³ s), superior thermal conductivity, and excellent mechanical strength and elasticity (Bolotin et al. 2008; Balandin et al. 2008; Lee et al. 2008). Since graphene was isolated in 2004 by Geim and coworkers (2004) using a scotch tape method, there have been many processes developed to produce graphene. Although the mechanical exfoliation of graphite leads to production of high-quality and high-mobility graphene flakes, this method is a time consuming process limited to a small scale production. One of the methods is thermal decomposition of SiC in temperature range between 1,000 and 1,500 °C. Here, Si sublimate from the silicon carbide and leave behind a carbon-rich surface (Hass et al. 2008). This method requires controlling the number of graphene layers, repeatability of large area growths, and interface effects with the SiC substrate (Choi et al. 2010). The second method of graphene synthesis is chemical vapor deposition (CVD). This process is promising in large scale production of graphene. Many papers concerning CVD growth of graphene using different catalysts (Ni, Cu, ZnS, Fe) have been reported (An et al. 2011; Wei et al. 2009; Li et al. 2009; Reina et al. 2009; Obraztsov 2009). Graphene can be also synthesized via chemical exfoliation of graphite, where interlayer van der Waals forces are eliminated: chemical derivatization, intercalation, thermal expansion, the use of surfactants, and oxidation–reduction (Chakraborty et al. 2008; Lotya et al. 2009; Lee et al. 2008). The most common route to chemically exfoliate graphite is the use of strong oxidants to produce graphene oxide in a first stage. Graphite oxide was first prepared by B.C. Brodie, where it was treated with a mixture of potassium chlorate and nitric acid (Brodie 1859). Later, Hummers and Offeman (1958) used a mixture of sulfuric acid, sodium nitrate, and potassium permanganate to oxidize graphite. Recently, many papers reporting a modification of the Hummers method have been published. For instance, Marcano et al. (2010) used a mixture of sulfuric and orthophosphoric acids and potassium permanganate in the oxidation process, and found it improved an efficiency of the oxidation. This process is still widely investigated for two reasons: (1) it gives an opportunity to produce large scale graphene or few-layered graphene, (2) it has potential to provide the samples with controlled number of graphene layers upon the process parameters optimization.

In this study, novel oxidants of graphite, toward its chemical exfoliation, were examined: a mixture of perchloric and nitric acids and potassium chromate. Furthermore, an effect of oxidation time, temperature of oxidation and ultrasonication in exfoliation degree were investigated. The obtained materials were characterized with TEM, FT-IR, Raman spectroscopy, and XRD.

In this study, novel oxidants were used to create chemical exfoliation of graphite. In order to examine the efficiency of oxygen-containing functional groups introduction to carbon lattice, FTIR spectroscopy was used. Figure 1 presents FTIR spectra of GO1, GO2, and GO3. At each spectrum, several typical modes corresponded to oxygen-functional groups are detected. Here, stretching vibration modes of C–O bonds arise at 1,099 cm⁻¹. The peaks between 1,190 and 1,380 cm⁻¹ are related with C–OH stretching vibration (Xu et al. 2008). Deformation vibration modes of O–H groups are observed at 1,431 cm⁻¹ (Stankovich et al. 2006). The bands appeared at 1,635 cm⁻¹ correspond to adsorbed water molecules indicating hydrophilic properties of the material (Paredes et al. 2008). At approximately 1,709 cm⁻¹ C=O stretching vibration of carboxyl groups are detected. The peaks at around 3,430 cm⁻¹ arise from O–H stretching vibration (Wojtoniszak et al. 2012). FTIR spectroscopy confirms successful oxidation. Furthermore, the intensities of the modes between 1,230 and 1,740 cm⁻¹ increased in order of GO1, GO2, and GO3 which suggests an increase of the oxidation efficiency.

For further analysis XRD was used to monitor chemical exfoliation of graphite in a bulk scale. Figure 2 shows XRD patterns of graphite, GO1, GO2, and GO3. XRD pattern of graphite is dominated by single peak at 26.48° corresponding to (002) plane (Liana et al. 2010). After oxidation process, the peak started to vanish, and new peaks arose at 24.28°, 24.78°, and 24.23° in XRD patterns of GO1, GO2, and GO3, respectively. On the basis of Bragg’s law, the interlayer spacing (d ₀₀₁) of GO1, GO2, and GO3 is 3.7, 3.6, and 3.7 Å, respectively. It is enhanced compared to that of graphite (3.4 Å) (Fan et al. 2011). In GO2 and GO3 XRD patterns, more distinct decay of the peak at 26.48° is observed. It suggests an enhanced exfoliation of graphite when oxidation was performed with ultrasonication treatment and when a time and a temperature was increased, also confirming the enhanced oxidation effect.

In order to determine the number of graphene layers, the samples underwent the reduction process. For reduced graphene oxide (RGO) preparation, glucose was used as a reducing agent. The obtained materials (RGO1, RGO2, and RGO3) were analyzed with Raman spectroscopy to characterize their structure, in terms of number of graphene layers and lattice defects. Figure 3 presents Raman spectra of graphite (A), RGO1 (B), RGO2 (C), and RGO3 (D). All spectra exhibit 3 typical modes related to graphite-like structure: a G band, a D band, and a 2D band (Fig. 3, left panel). The G mode at approximately 1,580 cm⁻¹ originates from the in-plane vibration of sp² carbon atoms and is a doubly degenerate phonon mode (E _2g symmetry) at the Brillouin zone center (Li et al. 2007). The D band at roughly 1,320 cm⁻¹ is a breathing mode of A _1g symmetry involving phonons near the K zone boundary (Wilson et al. 2009). The 2D band arises from a two phonon double resonance Raman process (Ferrari and Robertson 2000). The 2D band is a key mode in identification of number of graphene layers. According to its position, shape and intensity, a thickness of graphene may be determined (Ferrari and Robertson 2000; Ni et al. 2008). Charlier et al. (2008) plotted the evolution of the 2D band as a function of layers in single-, bi-, and few-layer graphene. Briefly, a 2D band of bi- and few-layer graphene has 2 components 2D₁ and 2D₂. Increasing the number of layers leads to a significant increase of the relative intensity of the 2D₂ peak and blue shift of the 2D peak. However, a 2D peak of single-layer graphene is composed of a single 2D peak, positioned at roughly 2,600 cm⁻¹. Therefore, a Lorentzian fitting of the RGOs spectra was performed and is presented in Fig. 3 (right panel). Here, 2D band of RGO1 differs from 2D band of graphite, and has two components: 2D₁ and 2D₂ with maximal intensity at 2,624 and 2,657 cm⁻¹, respectively. It may be attributed to 5-layered graphene. The 2D₂ peak of RGO2 decreases in comparison to RGO1. Furthermore, the 2D peak of RGO2 shifts to lower frequencies. These observations suggest the enhanced exfoliation of graphene when sonication was utilized leading to creation of bilayer graphene. It is clearly seen that 2D₂ peak of RGO3 vanishes, and single 2D peak appears at 2,584 cm⁻¹ attributed to single-layer graphene (Ferrari et al. 2006). It is known that a ratio between the intensities of the D and the G bands (I _D/I _G) determines relative defect content in the carbon lattice (Jorio et al. 2011). Therefore the I _D/I _G ratios of RGOs spectra were calculated and results are presented in Fig. 4. It is shown that the I _D/I _G ratio increases in order of RGO1, RGO2, and RGO3. This is related to the formation of vacancies and defects in carbon lattice, such as five- and seven-membered carbon rings, during the reduction process (Schniepp et al. 2006). This means that introduction of more oxygen-functional groups during the oxidation results in formation of more defects and vacancies in the final product—graphene.

To confirm the morphology of RGO1, RGO2, and RGO3, atomic force microscopy was applied. Figure 5 presents topography and height profiles of the prepared nanomaterials. The sizes of graphene flakes ranged from 30 to 70 nm in the each sample. The thickness of RGO1, RGO2, and RGO3 was 2.5–2.6, 1.3, and 1.0–1.1 nm, respectively. It was already reported that the thickness of a single-layer graphene on a substrate SiO₂/Si with approximately 1 nm rms roughness, would be 0.8–1.2 nm. The enhanced thickness of monolayer graphene may be attributed to the interaction between the sample and the tip (Gupta et al. 2006; Nemes-Incze et al. 2008). In bi- or few-layered graphene, the thickness arises with additional layer and therefore, adding the expected 0.35 nm height corresponding to the native van der Waals inter-layer distance (Soldano et al. 2010). It was concluded that RGO1, RGO2, and RGO3 were composed of 5-layered, bilayer, and monolayer graphene, respectively (Li et al. 2008a, b). This observation is in full agreement with Raman response of the samples.

In summary, novel method of chemical exfoliation of graphite was presented. Here, novel oxidants were examined: a mixture of perchloric and nitric acids and potassium chromate, and an effect of oxidation time, temperature of oxidation, and ultrasonication in exfoliation degree were investigated. The presented methodology leads to creation of graphene with controlled number of layers: single-, bi-, and 5-layered graphene in a bulk scale.