Since the last century, a large amount of graphite (GR) and carbon products are now synthetic and these products have been improved continuously . Synthetic graphitic products are manufactured by a compaction process mixing carbon and organic binders followed by a graphitizing treatment, leading to commercial products such as the electrodes for secondary steel-making industry. Since the discovery and characterization of graphene (G) in 2004  , research has been directed to find new manufacturing techniques    to promote industrial applications of graphene together with graphene oxide (GO). Graphene films grown by chemical vapor deposition (CVD) exhibit high quality but are not suitable for large-scale production  . Chemical exfoliation methods based on the Hummers’ method promoting graphite oxidation to obtain graphene oxide (GO) followed by chemical or thermal reduction may have the advantages of low-cost fabrication process  . Electrochemical exfoliation of high-quality graphite in aqueous solutions has led to graphene and/or graphene oxide sheets depending on the experimental parameters   . Besides, at present industrial research is looking for large-scale polymer/graphene-product nanocomposites which promise to attain materials with better electronic, thermal and/or mechanical properties  . In the present preliminary work, graphene oxide/multilayer-graphene (GO-MG) flakes were obtained by electrochemical exfoliation of industrial steel-making graphite electrodes following the procedures reported in the literature   with some slight modifications. The exfoliated GO-MG flakes were processed to finally be incorporated in a domestic isocyanate/polyol commercial system required to achieve the corresponding elastomeric polyurethane (PU) matrix. The mechanical properties of the PU/GO-MG composites were evaluated and compared with equivalent composites made of PU/GR powder to exhibit the enhanced chemical bonding between GO-MF flakes and polyurethane chains.
2. Experimental Procedure
Transmission electron microscopy (TEM) and X-ray diffraction (XRD) measurements were performed on powder samples obtained from the electrochemical exfoliation of graphite. The TEM measurements were carried out using an electron microscope JEOL (Jem-1400) in bright field mode at 100 kV. The powder XRD data were collected with BRUKER (D8 Advance) diffractometer in a Bragg-Brentano configuration using Cu-sealed tube (CuKa = 0.15418 nm) operating at 40 kV and 30 mA. Measurements were performed in the scattering 2θ range of 5˚ to 60˚ with a step size of 0.01˚ and counting rate of 2 s/step. UV-Vis absorption spectra were recorded on a PERKIN-ELMER UV/Vis (Lambda-25) spectrometer to gain further information into the quality of electrochemically prepared GO-MG flakes. Raman spectroscopy (RS) was carried out using RENISHAW system with an excitation laser source of 532 nm and power below 0.1 mW on the sample to avoid induced laser heating. To evaluate the mechanical properties, tension test was performed using a PHYSICAL TEST SOLUTIONS (MMD-30k) universal material testing machine using a cross-head speed of 1 mm/min at room temperature.
2.2. Materials Manufacture
Exfoliated graphene oxide/multilayer-graphene (GO-MF) flakes were obtained by electrochemical exfoliation of industrial graphite performed in a two-electrode system using tungsten (99.5% W) as the counter electrode and graphite as the working electrode. The separation between the cathode and the anode was maintained at 3 cm. Usually there is the presence of impurities on industrial graphite-like Fe, Co, Cu and Ni  , consequently in aqueous acid conditions the cationic species should remain dissolved, so the chosen electrolyte solution was prepared from concentrated sulfuric acid (H2SO4) and deionized water (pH adjusted to 1.50). A direct DC voltage bias of +2.5 V was first applied to the graphite for 1 hour to cause soft intercalation of the SO42− ions in the graphite layers   , and at this conditions, the graphite electrode remained solid. Then the voltage bias was increased to +6 V and the graphite electrode started to dissociate in small flakes distributing all over the solution. The experimental setup for electrochemical exfoliation of graphite electrode is shown in Figure 1(a) and the final electrochemical exfoliated product after 4 hours is also shown in Figure 1(b). Test specimens for evaluation of the tensile properties of elastomeric PU/GO-MG composites were obtained washing and drying the GO-MG flakes, then dispersing the product in 1,4-butanediol, drying once again to mix it with the polyol compound, then the slurry was added to the isocyanate compound and finally cast to the molds. Figure 2 illustrates the size and shape of the samples used to determine the mechanical behavior of the elastomeric PU composites filled with different concentrations of GR and GO-MG.
3. Results and Discussion
Figure 3 illustrates the morphology of the GO-MG flakes which were observed by transmission electron microscopy (TEM). The TEM images revealed two different type of morphologies, one disordered with scrolls and wrinkles (Figure 3(a)) and the other with a transparent and sheet-like structure (Figure 3(b)).
Figure 1. Photographs of the electrochemical exfoliation of graphite. (a) System at a voltage bias of +2.5 V; (b) System after 4 hours at a voltage bias of +6 V.
Figure 2. Photographe showing the shape and size of the tensile elastomeric PU composite samples fabricated without (white) and with 10 wt% GO-MG (black).
Figure 3. TEM bright field images obtained from the electrochemical exfoliation of graphite. (a) Graphene oxide (GO); (b) Multilayer graphene (MG).
Both type of flakes remained stable under the electron beam. In general, graphene oxide (GO) has been assigned to the morphology having scrolls and wrinkles while the sheet-like morphology has usually been assigned to multilayer graphene structure (MG)   .
To corroborate the presence of both GO and MG compounds, XRD were performed on exfoliated flakes. Figure 4 clearly shows the difference between the XRD patterns of the graphite powder obtained from mechanical grinding the GR electrode and the flakes obtained from electrochemical exfoliation of the GR electrode. Figure 4 indicates that graphite electrode is highly crystalline and the most intense peak is the one associated to the (002) crystallographic planes of graphite (JCPDS 41-1487). On the other hand, the XRD pattern of the exfoliated graphite in Figure 4 shows two broad peaks at 2θ = 11.28˚ and 2θ = 26.19˚. The first peak at 11.28˚ has frequently been associated to the presence of graphene oxide      as a consequence of the intercalation of different functional groups. The second broad peak should correspond to the (002) graphite planes, but the usually interlayer distance of 0.337 nm found in Figure 4 for graphite powder, in the case of GO-MG flakes seems to be slightly larger than that of the d-spacing of bulk graphite.
Figure 4. XRD patterns obtained from graphite electrode powder and GO-MG exfoliated flakes.
UV-Vis spectra of aqueous electrochemical exfoliated graphite dispersions were used to determined the presence of C = C and C - O bonding. The spectrum obtained by UV-Vis spectroscopy of the electrochemically exfoliated graphite is in agreement with the previously graphene oxide reported results     , where an absorption peak appears at about 230 nm and a broad peak at approximately 290 - 310 nm as shown in Figure 5. The absorption peak at 230 nm has been assigned to the π ® π* transition of the C = C bonds by other authors  . Then, the present assumption of the absorption peak observed at 226 nm to the same π ® π* transition seems quite reasonable. On the other hand as in previous works    , it is possible to assign the broad peak observed at 305 nm to the n ® π* transitions due to the presence of C - O - C or C - OOH bonding, usually reported for graphene oxide. On the other hand, there is experimental evidence suggesting that π ® π* transition shifts to higher wavelength (red-shift), displaying the peak at about 250 nm when graphene is present  . Because the peak observed at 226 nm in Figure 5 is quite wide, there was no possibility to discern by this technique the presence of graphene in the GO-MG flakes.
Also, Raman spectroscopy (RS) was used to know about structural changes occurred during the electrochemical exfoliation of graphite, mainly because RS is a sensitive technique to characterize disorder in sp2 carbon materials. The G-band observed in Figure 6 (at 1585 cm−1) arises from the stretching of the C - C bond in graphitic materials and is common to all sp2 carbon systems.
The peak at 1355 cm−1 know as the D-band represents the resonance of aromatic rings arising due to defects in the structure and its presence suggest the existence of GO . On the other hand in Figure 6, the 2D peak at around 2700 cm−1 is active in the absence of any defects and the peak shifts with the number of graphene layers, becoming similar to that of graphite if there is more than 10 layers of graphene, in which case the peak appears at 2730 cm−1  . Consequently, from reported experimental data  the number of layers in the graphene flakes should be approximately between 2 and 5 in our samples. Then, all experimental data indicated the presence of both, graphene oxide and multilayer-graphene compounds in the electrochemically exfoliated graphite flakes. From XRD peak analysis using the Rietveld refinement method and considering from Figure 4 that only was possible to assign the presence of two different structures, it was concluded that GO-MG flakes were approximately composed of 67 wt% GO and 33 wt% MG. If there are other compounds in the GO-MG flakes, they should be below the detection limit of the XRD technique.
Figure 7 shows a microphotograph of the employed graphite particles together with the tensile properties of the PU/GR composites with different graphite powder content. In this case, the maximum attained strength was of 750 kPa and corresponds to 20 wt% of graphite (PU/20GR). On the other hand, Figure 8(a) shows the tensile properties of the PU/GO-MG composites with different filler content. From stress-strain curves, it can be seen a remarkable difference in the mechanical behavior between GR and GO-MG composites. In both cases, it can be seen that although the elongation at break decreases with increasing filler content, the overall elongation of the composite materials still is significant. Besides, Figure 7 strongly suggests that the reinforcement contribution of graphite is not as effective as the one observed in Figure 8(a) for graphene oxide/multilayer-graphene (GO-MG). The relationship between the breaking stress of composites and type of filler content is shown in Figure 8(b). The pure PU showed a breaking stress of 570 kPa as compared to PU/10GO-MF composite exhibiting a breaking stress of 1060 kPa.
Figure 5. UV-Vis absorption spectrum of the electrochemically exfoliated graphite solution in a quartz cuvette.
Figure 6. Raman spectrum of the GO-MG flakes showing the G (1355 cm−1), D (1585 cm−1) and 2D (2700 cm−1) bands.
Figure 7. (a) Photograph of the graphite powder obtained by milling of the electrode; (b) Stress-strain curves of the elastomeric PU/GR composites with different filler content.
Figure 8. (a) The stress-strain curves of the elastomeric PU/GO-MG composites with different filler content; (b) The relationship between the breaking stress and the filler content; (c) Magnification of the stress-strain curve of the sample with 10 wt% GO-MG showing the presence of strain hardening.
The distinctive stress-strain curves obtained from the PU/GO-MG composites is presented for just one composition in Figure 8(c). The results indicated that GO-MG fillers are capable of improving the breaking stress of elastomeric PU with the presence of strain hardening behavior . The phenomenon of strain hardening in polymers is a consequence of orientation of molecular chains in the stretch direction. The large difference in mechanical properties between GR and GO-MG could be attributed to efficient load transfer between the graphene oxide sheets and the PU matrix, resulting from the covalent bonding and the cross-linking between the isocyanate chains and the functional groups attached to GO   , leading to the observed strain hardening behavior.
Exfoliated graphene oxide/multilayer-graphene (GO-MF) flakes were obtained by electrochemical exfoliation of industrial graphite performed in a two-electrode system using tungsten (99.5% W) as the counter electrode and graphite as the working electrode. From different analytical techniques and XRD results, it was concluded the presence of only these two structures: graphene oxide (GO) and multilayer-graphene (MG). It was calculated from XRD Rietveld refinement that flakes were approximately composed of 67 wt% GO and 33 wt% MG. Comparing Raman results with those reported previously, the number of layers in the graphene should be approximately between 2 and 5 sheets in our samples.
Besides, the relationship between the breaking stress of composites and type of filler content strongly suggests that the reinforcement contribution of graphite (GR) is not as effective as the one observed for graphene oxide/multilayer-graphene (GO-MG). The pure elastomeric PU showed a breaking stress of 570 kPa, while the PU/20GR attained a maximum stress of 750 kPa as compared to PU/10GO-MF composite exhibiting a breaking stress of 1060 kPa. Finally, results indicated that GO-MG fillers are capable of improving the breaking stress of elastomeric PU by means of strain hardening behavior, probably improving orientation of molecular chains as a consequence of covalent bonding leading to cross-linking between the polymer and the functional groups attached to GO.