Graphite foam was prepared via the dehydration of sugar. Under fume hood extraction, commercial granulated sugar (50 g) was added to a 500-mL beaker, which was then placed in a metal pan filled with sand. Concentrated H₂SO₄ was then added to the beaker under the continuous stirring of a glass rod. Eq. (1) shows the reaction involved in the dehydration reaction of sugar.(1) C12H22O11+H2SO4=12C (graphite)+11 H2O+Heat 

The color of the reaction mixture changed from white to yellow and then to black; then, the mixture began to extend outside the beaker to form a so-called graphite column, as shown in Fig. 1A. The clean graphite foam containing residual H₂SO₄ and water resulting from the reaction was then placed in a ball mill for wet milling. Zirconia balls of 3.0–5.5 mm in diameter were added to the ball mill, and the graphite foam was milled in the presence of ethanol (99.7%) at a rotation speed of 3000 rpm for 15 h to ensure that the shear stress was dominant, with the weight of the milled graphite powder measured as 30 g. The small volume of ethanol added to the milling process improves the intercalation of the exfoliating material within the graphene layers, leading to a special type of exfoliation. The dark graphite slurry obtained from the ball milling process was ultrasonicated at 750 W for 30 min to break down any aggregated particles in the slurry before being filtered through pre-moistened 100-μm filtration paper in a Büchner funnel to remove the hard clusters formed in the milling procedure. The resulting reaction mixture was then subjected to centrifugation at 3000 rpm for 15 min to remove any macroscopic aggregates. The shear force of the ball milling/ultrasonication in the presence of ethanol led to the formation of a uniform, highly stable, black suspension of the graphene. After evaporating the solvent and washing the product multiple times using ethanol, a high-purity dehydrated powder was obtained. The product was structurally characterized by X-ray diffractometry (XRD) and Raman spectroscopy. Furthermore, the morphologies of the graphite powder and the prepared graphene sheets were analyzed using scanning electron microscopy (SEM). The experimental and practical steps involved in the preparation of the GFs are shown in Fig. 1.

The application of the prepared GFs to improve the flow of crude oil was tested in a circulating crude oil flow system comprising a pump, an oil pipe with an internal diameter of 4.5 cm and a length of 1.5 m, valves, pressure transducers, and a voltage regulation device. An electric field was applied to the system using an electrical capacitor, the design and specifications of which are detailed in previous work [10]. Fig. 1B shows a schematic diagram of the circulating flow system. In accordance with the central composite rotatable design (CCRD) method, 20 experiments were implemented to investigate the impact that three experimental parameters have on the process, namely, the applied voltage (150, 170, and 220 V), the stimulation time (15, 30, and 60 s), and the concentration of the GFs (100, 200, and 400 mg/L), in addition to their interaction effects. GF concentrations were moderated using an ultrasonic mixer. All of the experiments were performed at 30°C ± 2°C. The differential pressure inside the pipe and crude oil viscosity were investigated using a suitable measuring device to probe the drag reduction and flow improvement efficiency. Crude oil samples were obtained from the Al-Doora refinery in Baghdad, Iraq, the physical properties and specifications of which are listed in Tables 1 and 2. In addition, the specifications of the circulating flow system equipment and capacitor used in the experiments are listed in Table 3. The optimization of the process was achieved using a systematic method in the experimental design employing the CCRD technique. The optimization method used has been previously described in detail in the literature [7,12,33].

The experimentally measured pressure drop and viscosity values were used to calculate the Reynolds number (the ratio of inertial to viscous forces in a fluid), the shear stress, friction factor, and the corresponding pumping power consumption as follows [7]:(2) Re. No.=u×d×ρμ  where u is the fluid velocity, which remains constant at 1.253 cm/s; d represents the internal diameter of the pipe, which is 4.3 cm; and μ is the dynamic viscosity (c.p).(3) Shear rate (γ)=8×ud 

(4) Shear stress (τ)=γ×μ

(5) Jf=8Re. No.

where J_f is the friction factor.(6) ε=22.12×Δpξ

where ε represents the pumping power consumption in W; Δp represents the drop in pressure in Pa; and ξ is the pump efficiency, which has a value of 0.85.

Structural analysis of the graphite powder (starting material) and produced graphene sheets was achieved by XRD and Raman spectroscopy analysis. Figs. 2A and 2B show the XRD patterns of the bulk graphite and graphene, from which it can be seen that bulk graphite exhibits an obvious peak at 2θ = 26.5°. This peak, shown in Fig. 2A, is extremely sharp and intense, thus indicating that the graphite powder is highly crystalline. In contrast, from Fig. 2B, it can be seen that the XRD pattern of the graphene features peaks that are very wide with low intensity, thus revealing it to be polycrystalline, which can be attributed to the decrease in its crystal size due to it being milled at high speed (300 rpm) for a long time period (15 h) in the presence of a solvent. Raman spectroscopy can be used to determine valuable structural information on graphene, in terms of its defects, the stacking of its layers, and crystallite size. Carbon materials usually exhibit fingerprints in Raman spectra in the form of D-, G-, and 2D-band peaks at around 1350, 1580, and 2700 cm⁻¹, respectively, because of the variation in their electron bands. Figs. 2C and 2D show a comparison of the Raman spectra of graphite and graphene. The two most intense peaks in the spectra are the G-band peak, which results from the doubly degenerate zone center E_2g mode of the material and is located at around 1580 cm⁻¹, and the 2D-band peak that represents second-order zone-boundary phonons and is located at around 2700 cm⁻¹. The latter peak can be deconvoluted into four bands in the case of few-layered graphene.

Furthermore, the G-band intensity of graphene is lower than that of graphite as the intensity of this band increases linearly with an increase in the number of layers in the measured material. The D-band peak at around 1350 cm⁻¹ can be attributed to the breathing modes of sp² rings in the graphene and is only present if the material being measured features defects. The intensity of this peak is related not only related to the amount of disorder in the graphene but also to the edges of sub-micrometer flakes. Fig. 3 shows atomic force microscopy (AFM) data that indicate the surface morphology of the three (3D) and two-dimensional (2D) particles of graphene as distribution plots. The AFM data shows that the measured sample features irregularly shaped graphene sheets of non-uniform thickness that have lateral dimensions with an average grain size of 90.21 nm. From the SEM images shown in Figs. 4 and 5, it can be observed that the AFM images indicate the non-uniform distribution of the particles in terms of their size and shape, meaning that the prepared graphene might be characterized by a group of high-density 3D clusters with small agglomerations on their surface. Various shapes, heights, and sizes of particles can be observed, indicating the crumpled and wavy structure of the graphene due to its deformity during the exfoliation process. The exfoliation of graphite layers occurs as there is a strong interaction between the basic planes of the graphite and the solvent, where the solvent minimizes the interfacial tension between the liquid and the GFs. Based on the results, the product synthesized in this study was proven to be graphene. The synthesis process was enhanced using an organic solvent as a milling medium to promote non-destructive exfoliation, which was confirmed in the characterization of the prepared samples. These results are in accordance with those of Al-Sherbini et al. [34].

Table 4 shows the experimental pressure drop values and crude oil viscosity measurements in addition to the results of the calculations of fluid flow characteristics processed using the CCRD technique. The Reynolds number, shear stress, friction factor, and power consumption results show the obvious impact that the GFs have on the drop in pressure in the crude oil system. Moreover, the electric field intensity had a significant impact on reducing the viscosity of the crude oil. The best flow characteristics are observed under optimal operating conditions. The Reynolds number was enhanced when the viscosity of crude oil was reduced at a constant pumping flow rate. The reduction in viscosity is a result of the application of an electric field, which leads to the GFs and asphaltene and paraffin molecules being arranged in short chains that slide over each other to minimize flow resistance and friction. Shear stress dramatically decreases with a reduction in viscosity. The effects of the operating conditions on the drop in pressure and power consumption are discussed in the following sections.

Table 4 shows the optimal conditions calculated using the WinQSB version 1.0 software. The objective function as a drop in crude oil pressure was devised using a polynomial formulation, represented by the following:(7) Y=a0+a1X1+a2X2+a3X3+a4X1X2+a5X1X3+a6X2X3+a7X12+a8X22+a9X32

The STATISTICA version 8.0 software was used to obtain the polynomial coefficients. The values of these coefficients and analysis of variance (ANOVA) results are listed in Table 5, which are in good agreement with the experimental data. The results indicate that higher R² values were obtained using the model than the adjusted R² values for all of the investigated conditions. Thus, the statistical acceptance of the model was proven.

Fig. 6A shows the impact that the addition of the GFs has on the drop in crude oil pressure under the optimal conditions according to the stimulation time and applied voltage. In the absence of the GFs, the pressure drop was 0.985 Pa. When specific amounts of the GFs were added to the system, there was a sharp increase in the pressure drop, with the lowest pressure drop of 0.683 Pa observed for 120 mg/L of GFs. The concentrations of GFs were tested, with the pressure drop observed to increase gradually at the concentrations of (0, 100, 200, and 400 mg/L). The exegesis of the positive impact of the GFs can be attributed to two factors. First, the increased dielectric constant of the GFs of around 5200 at 30 kHz had a condensing impact on the electrical field. Second, the nanoscale-sized GFs act as a flow improver, with their shape and morphology providing layers for fluid sliding and lubricant properties in addition to their dispersion as suspended particles during fluid flow. The physical basis of the enhancement in the flow rate of crude oil can be explained by the theory of liquid suspensions. The crude oil flows in straight lines according to the effect and orientation of the electric field. The adverse effects observed on the crude oil flow for GF concentrations above 120 mg/L is a result of greater doses that lump the flow rate and increase the viscosity of the oil. In addition, nanoparticles are present in the liquid that form clusters and hinder the flow of the oil. It can be concluded that the pressure drop decreases by 30.66% under optimal conditions. Fig. 6B shows the effect that the flow improvement under the influence of the GFs has on the pumping power consumption. It has been previously mentioned that the power consumption is based on the drop in pressure inside the pipe. Thus, the lowest power consumption of 17.77 W for the pumping of heavy crude oil was achieved using an optimal amount of the GFs of 120 mg/L, compared with 25.5 W in the absence of the GFs. The power consumption increased in line with an increase in the concentration of the GFs above 120 mg/L due to an increase in the value of the pressure drop.

Figs. 7A and 7B show that the voltage applied has a great impact on increasing the pressure drop and meeting the required power consumption. Moreover, a high voltage was observed to have an adverse effect on the values of the pressure drop and power consumed increased gradually beyond 150 V. The reason for this trend is that when crude oil is not subjected to an electric field, the nanoparticles in the oil are randomly distributed, whereas when the crude oil is exposed to an electric field, the particles suspended in the oil are polarized (polarization effects) and driven in a certain direction by the effect of the electric field [35]. The particles suspended in solution become less randomly arranged and rearrange into short chains according to the electric field lines. Strong electric fields also interact with the fluid medium in which the particles are suspended, generating forces that act on the fluid and its flow, in addition to the high dielectric constant of the suspended GF leading to a concentration of the electric field. Also, the electric field presents a great advantage since it generates an effective repulsive force between the nanoparticles, somewhat preventing their agglomeration. These results are significant as not only is the flow rate of the crude oil improved, but also turbulence is inhibited in the pipeline. These results are in accordance with those of Tao et al. [1], and Alwasiti et al. [7].

The stimulation time has a great effect on the impact that the electric field on the pressure drop and power consumption, where the optimal value was determined as 55 s, with the best pressure drop and lowest pumping power consumption values being 0.683 Pa and 17.77 W, respectively. When the stimulation time was far from optimal, the pressure drop and pumping power consumption values increased gradually until values of 0.8 Pa and 20.8 W were reached after 65 s, respectively. This is because over a short time period, there is not enough time for the arrangement of suspended asphalt and paraffin particles in addition to GFs, while the optimal time of 55 s represents the most effective time for the straight chains to configurate. These short chains of nanoparticles act as a lubricant to promote fluid flow. During longer test periods, the chains of short particles chains thicken and increase the viscosity of the system, leading to the agglomeration of particles and a build-up of clusters, thus resulting in increased shear stress and drag. Figs. 8A and 8B show the relationship between the stimulation time, the pressure drop, and the pumping power consumption, where the results are in accordance with those of Tao et al. [1].