Graphene Synthesis by Ultrasound Energy Assisted Exfoliation of Graphite in Various Solvents

Graphene has created an increasing notice thanks to its appealing properties [1]. In this study, graphene was prepared from graphite by a very simple and easy process. The one-step protocol involves conversion of graphite to graphene by sonication in different types of solvents such dimethyl sulfoxide, N,N-Dimethyl formamide, Perchloric acid. The structures and properties of the obtained graphene samples were characterized via UV–vis absorption, and Atomic Force Microscopy spectroscopic techniques. According to the UV-vis spectrums of all graphene products give peak at 265 nm wavelengths that referring sp2 C=C bonds [2], which may be caused by the ultrasonication required for proper suspension using the solution-based process. Also, as a result of AFM analyses, it can be concluded that the obtained graphene samples contain a few layers; while G-DMSO has four layers, G-NNDMF has five layers. It can be understand that DMSO shows better effect on graphite for sonication process. The preparation protocol is simple, easy, eco-friendly.


Introduction
Graphene is a versatile nanomaterial with a wide range of chemical, environmental, medical, industrial, and electronical applications by means of its remarkable thermal conductivity (above 3000 W m K −1 ), superior mechanical properties with a Young's modulus of 1 TPa, an extraordinary large specific surface area (2620 m 2 g −1 ), intrinsic strength of 130 GPa, and an extremely high electronic conductivity (room-temperature electron mobility of 2.5 × 10 5 cm 2 V −1 s −1 ) [1][2][3]. This combination of spectacular properties enables its utilization in the production of different devices, such as electronics with high-speed and radio-frequency logic, sensors, membranes, composites with high thermal and electrical conductivity, displays with superior transparency and flexibility, solar cells, coatings, ultrathin carbon films, electronic circuits, etc. [4].
Few-layered graphene has been synthesized by numerous methods including mechanical cleavage, liquid-phase exfoliation, gas-phase synthesis, Hummers' method, unrolling of multi-walled carbon nanotubes (MWCNTs), chemical vapor deposition (CVD), epitaxial growth, and electrochemical reaction [2,[5][6][7][8][9][10][11][12][13][14][15]. However, these methods have some drawbacks, such as low yield ratio, high-energy consumption, the use of expensive substrates, as well as the difficulty of obtaining a high-quality product. Scientists have studied the development of more efficient synthesis methods apart from the most popular Hummers' method employing the oxidative exfoliation of natural graphite by harsh acid treatments. The final products attained by Hummers' method have many functional groups, such as hydroxyls, carbonyls, and carboxyls, which cannot be completely removed by additional reduction and In the present study, we introduced one-pot synthesis of graphene via ultrasonication of graphite in different organic solvents without using ionic liquids or surfactants. The characterization of synthesized graphene samples was performed by XRD, AFM, UV-vis, DLS, and TEM analyses.

Materials
Graphite fine powder (extra pure) and commercial graphene (CG) were purchased from, Asbury Inc., New Jersey, and XG Sciences, Michigan, US, respectively. Dimethyl sulfoxide (DMSO) (Merck), N,N-dimethylformamide (DMF) (Merck), and perchloric acid 70−72% (PA) (Merck) were of analytical grade and used as received. CG was used as the control sample in order to compare the precision and purity of the synthesized graphene products.

Method
First, 0.3 g graphite were dispersed in a 50-mL solvent, such as DMSO, DMF, and PA. The prepared dispersions were sonicated for 3 h by using BANDELIN ® HD 2200 SONOPULS equipped with a vs. 190 T sonotrode made of titanium alloy, 200 W, 50% amplitude ( Figure 2). Afterwards, these dispersions were subjected to Elektromag, M 4812 P for 1 h at 3000 rpm in order to remove the unexfoliated graphite particles. Then, the large aggregates were settled down, the supernatant part of the samples was decanted, and then collected in separate vials.  In the present study, we introduced one-pot synthesis of graphene via ultrasonication of graphite in different organic solvents without using ionic liquids or surfactants. The characterization of synthesized graphene samples was performed by XRD, AFM, UV-vis, DLS, and TEM analyses.

Materials
Graphite fine powder (extra pure) and commercial graphene (CG) were purchased from, Asbury Inc., Asbury, NJ, USA and XG Sciences, Lansing, MI, USA, respectively. Dimethyl sulfoxide (DMSO) (Merck), N,N-dimethylformamide (DMF) (Merck), and perchloric acid 70−72% (PA) (Merck) were of analytical grade and used as received. CG was used as the control sample in order to compare the precision and purity of the synthesized graphene products.

Method
First, 0.3 g graphite were dispersed in a 50-mL solvent, such as DMSO, DMF, and PA. The prepared dispersions were sonicated for 3 h by using BANDELIN ® HD 2200 SONOPULS equipped with a vs. 190 T sonotrode made of titanium alloy, 200 W, 50% amplitude ( Figure 2). Afterwards, these dispersions were subjected to Elektromag, M 4812 P for 1 h at 3000 rpm in order to remove the unexfoliated graphite particles. Then, the large aggregates were settled down, the supernatant part of the samples was decanted, and then collected in separate vials. In the present study, we introduced one-pot synthesis of graphene via ultrasonication of graphite in different organic solvents without using ionic liquids or surfactants. The characterization of synthesized graphene samples was performed by XRD, AFM, UV-vis, DLS, and TEM analyses.

Materials
Graphite fine powder (extra pure) and commercial graphene (CG) were purchased from, Asbury Inc., New Jersey, and XG Sciences, Michigan, US, respectively. Dimethyl sulfoxide (DMSO) (Merck), N,N-dimethylformamide (DMF) (Merck), and perchloric acid 70−72% (PA) (Merck) were of analytical grade and used as received. CG was used as the control sample in order to compare the precision and purity of the synthesized graphene products.

Method
First, 0.3 g graphite were dispersed in a 50-mL solvent, such as DMSO, DMF, and PA. The prepared dispersions were sonicated for 3 h by using BANDELIN ® HD 2200 SONOPULS equipped with a vs. 190 T sonotrode made of titanium alloy, 200 W, 50% amplitude ( Figure 2). Afterwards, these dispersions were subjected to Elektromag, M 4812 P for 1 h at 3000 rpm in order to remove the unexfoliated graphite particles. Then, the large aggregates were settled down, the supernatant part of the samples was decanted, and then collected in separate vials.

Characterization
X-ray diffraction (XRD) was conducted by depositing the samples onto glass pieces (0.7 × 0.7 mm 2 ) and their XRD spectra were determined by a Rigaku D-Max 2200 Series device equipped with Cu-Kα radiation (λ = 1.54 Å) at a scanning rate of 3 • per minute. Its tube voltage and current were 40 kV and 40 mA, respectively. Samples for AFM were prepared by dropping the graphene dispersions onto glass pieces (0.7 × 0.7 mm 2 ) and measurements were carried out in contact (tapping) mode, with 10.00 µm scan size, and 20.35 Hz scan rate by using a Digital Instruments Nanoscope. Ultraviolet-visible (UV-vis) spectroscopy analyses were done by a Perkin Elmer Precisely Lambda 35 UV-vis Spectrometer in the region between 200-800 nm wavelengths. In order to elucidate the particle size distribution, polydispersity index (PDI) and zeta potential analyses were carried out by the dynamic light scattering (DLS) method through a Malvern Zetasizer Nano ZS Laser particle size distribution meter. The structural and morphological properties were determined by a Hitachi HT7800 model transmission electron microscope (TEM) operating at 120 kV.

Results and Discussion
The exfoliation mechanism of graphite by the sonochemical process using a horn-type device is given in Figure 3. This type of sonicator produces high-intensity ultrasound energy, which enables stable micromechanical exfoliation of graphite. Additionally, the horn-type sonicator provides the synthesis of graphene with minor functional groups and flat and defect-free morphologies. The stable dispersion of graphene into solvents can be provided by sonication that builds cavitation shear stress in the solvent. The sonication time and power can adjust the concentration of the dispersion.

Characterization
X-ray diffraction (XRD) was conducted by depositing the samples onto glass pieces (0.7 × 0.7 mm 2 ) and their XRD spectra were determined by a Rigaku D-Max 2200 Series device equipped with Cu-Kα radiation (λ = 1.54 Å) at a scanning rate of 3° per minute. Its tube voltage and current were 40 kV and 40 mA, respectively. Samples for AFM were prepared by dropping the graphene dispersions onto glass pieces (0.7 × 0.7 mm 2 ) and measurements were carried out in contact (tapping) mode, with 10.00 μm scan size, and 20.35 Hz scan rate by using a Digital Instruments Nanoscope. Ultravioletvisible (UV-vis) spectroscopy analyses were done by a Perkin Elmer Precisely Lambda 35 UV-vis Spectrometer in the region between 200-800 nm wavelengths. In order to elucidate the particle size distribution, polydispersity index (PDI) and zeta potential analyses were carried out by the dynamic light scattering (DLS) method through a Malvern Zetasizer Nano ZS Laser particle size distribution meter. The structural and morphological properties were determined by a Hitachi HT7800 model transmission electron microscope (TEM) operating at 120 kV.

Results and Discussion
The exfoliation mechanism of graphite by the sonochemical process using a horn-type device is given in Figure 3. This type of sonicator produces high-intensity ultrasound energy, which enables stable micromechanical exfoliation of graphite. Additionally, the horn-type sonicator provides the synthesis of graphene with minor functional groups and flat and defect-free morphologies. The stable dispersion of graphene into solvents can be provided by sonication that builds cavitation shear stress in the solvent. The sonication time and power can adjust the concentration of the dispersion. The stable homogeneous dispersion of graphite can be achieved in solvents, such as DMSO, DMF, and PA, possessing similar surface tensions close to that of graphene (~68 mJ/m 2 ) [38,39]. The surface energies of DMSO, DMF, and PA are 43.5, 37.1, and 70.0 mJ/m 2 , respectively. When the surface energies of the dispersed phase and dispersant are close to each other, the enthalpy of mixing is diminished, resulting in a stable dispersion of graphite. The crystal structure and the layer numbers of the natural graphite and the graphene products (G-DMSO, G-DMF, and G-PA) were determined by XRD analysis given in Figure 4. It is well-known that the peak at 2θ° = 26.5 indicates the graphene characteristic structure. The graphene peak has a lower intensity, which is decreased compared to graphite, showing that the graphite tends to have an amorphous crystal structure of graphene [40]. The stable homogeneous dispersion of graphite can be achieved in solvents, such as DMSO, DMF, and PA, possessing similar surface tensions close to that of graphene (~68 mJ/m 2 ) [38,39]. The surface energies of DMSO, DMF, and PA are 43.5, 37.1, and 70.0 mJ/m 2 , respectively. When the surface energies of the dispersed phase and dispersant are close to each other, the enthalpy of mixing is diminished, resulting in a stable dispersion of graphite. The crystal structure and the layer numbers of the natural graphite and the graphene products (G-DMSO, G-DMF, and G-PA) were determined by XRD analysis given in Figure 4. It is well-known that the peak at 2θ • = 26.5 indicates the graphene characteristic structure. The graphene peak has a lower intensity, which is decreased compared to graphite, showing that the graphite tends to have an amorphous crystal structure of graphene [40]. In the XRD pattern of graphite, the (002) crystal plane of graphite was evident, with a d-spacing of 3.4 Å, which is typical for graphite. After intercalation of graphite, the produced samples showed a d-spacing of 3.5, 3.3, and 3.3 Å for G-DMSO, G-DMF and G-PA, respectively. The peak (002) crystal plane of graphenes had an average value of 3.4 Å, proving the formation of graphenes with few-layered structures.
Crystals 2020, 10, x FOR PEER REVIEW 5 of 12 In the XRD pattern of graphite, the (002) crystal plane of graphite was evident, with a d-spacing of 3.4 Å, which is typical for graphite. After intercalation of graphite, the produced samples showed a d-spacing of 3.5, 3.3, and 3.3 Å for G-DMSO, G-DMF and G-PA, respectively. The peak (002) crystal plane of graphenes had an average value of 3.4 Å, proving the formation of graphenes with fewlayered structures. The thickness of graphene products was estimated by applying Scherrer's equation, which is stated as D002 = Kλ/Bcos θ. D002, where K, λ, B, and θ are the thickness of the graphene, a constant based on the crystal shape (0.89), the wavelength of the X-ray (0.15406 nm), the full width half maximum (FWHM) of the characteristic peak of graphene, and the scattering angle, respectively [41,42]. The number of layers was calculated by the following equation: N = D002/d002, where d002 is the interlayer distance [43,44]. The calculated layer numbers of G-DMSO, G-DMF, and G-PA are 9, 10, and 20, respectively. Although the layer numbers of G-DMSO and G-DMF were very close to each other, G-PA gave a higher layer number. This higher layer number value may be explained by the inverse segregation of graphene particles synthesized in PA.
The thickness and surface roughness values of G-DMSO, G-DMF, and G-PA were determined via AFM analysis shown in Figure 5. Roughness average (Ra) is defined as the arithmetic mean of the absolute values of the height alterations from the mean line along the profile. The square root of the arithmetic mean surface roughness (Rq) is also described to take into account the big peaks and valleys [45]. Additionally, the roughness mean square (RMS) is calculated by using height values of microscopic peaks and valleys. As seen from Figure 5  The thickness of graphene products was estimated by applying Scherrer's equation, which is stated as D 002 = Kλ/Bcos θ. D 002 , where K, λ, B, and θ are the thickness of the graphene, a constant based on the crystal shape (0.89), the wavelength of the X-ray (0.15406 nm), the full width half maximum (FWHM) of the characteristic peak of graphene, and the scattering angle, respectively [41,42]. The number of layers was calculated by the following equation: N = D 002 /d 002 , where d 002 is the interlayer distance [43,44]. The calculated layer numbers of G-DMSO, G-DMF, and G-PA are 9, 10, and 20, respectively. Although the layer numbers of G-DMSO and G-DMF were very close to each other, G-PA gave a higher layer number. This higher layer number value may be explained by the inverse segregation of graphene particles synthesized in PA.
The thickness and surface roughness values of G-DMSO, G-DMF, and G-PA were determined via AFM analysis shown in Figure 5. Roughness average (Ra) is defined as the arithmetic mean of the absolute values of the height alterations from the mean line along the profile. The square root of the arithmetic mean surface roughness (Rq) is also described to take into account the big peaks and valleys [45]. Additionally, the roughness mean square (RMS) is calculated by using height values of microscopic peaks and valleys. As seen from Figure 5 In order to confirm the number of graphene layers, AFM results were exploited as well. The following equation was used for the calculation of layer numbers: N = (tmeasured − 0.4)/0.335, where tmeasured is the vertical distance. The value of 0.335 nm is the thickness of the single-layer graphene. Since the mica was used as a substrate material in the AFM measurements, the height was accepted as 0.4 nm [46]. The thickness (tmeasured) values of the graphene samples were measured as 1.64, 2.15, and 7.28 nm for G-DMSO, G-DMF, and G-PA, respectively. The number of layers was calculated from the aforementioned equation as 3.69 ≅ 4, 5.22 ≅ 5, and 21.05 ≅ 21 for G-DMSO, G-DMF, and G-PA, respectively. When all the results were assessed, the roughness and the layer number values of G-DMSO and G-DMF are smaller than that of G-PA. The layer numbers of graphene products obtained from XRD agree with those estimated from the AFM. It can be inferred from these findings that while G-DMSO and G-DMF include fewer layers, G-PA has a multi-layered structure. In the case of preparation of graphene in PA, the RMS and layer number are 19.3 and 21.0 nm, respectively. For the sample of G-DMSO, the RMS and layer number are 5.7 and 4.0 nm, respectively. The multilayered specimen of G-PA has a higher RMS value, which is in accordance with the literature [47]. It is believed that the combined effect of high-power ultrasound energy and strong acidity of PA might trigger the formation of some functionalities on the graphene, leading to an increase in the surface roughness. Additionally, the reason for the higher thickness may be due to the out-of-plane rippling behavior of graphene [48].
In order to clarify the graphene structure, the UV-vis spectra of the graphene dispersions were recorded as seen in Figure 6. The synthesized graphene products were characterized by comparing the properties with that of CG. The graphene samples labeled as G-DMSO, G-DMF, and G-PA, showed a peak at the 265-nm wavelength referring to sp 2 C = C bonds [49]. In order to confirm the number of graphene layers, AFM results were exploited as well. The following equation was used for the calculation of layer numbers: N = (t measured − 0.4)/0.335, where t measured is the vertical distance. The value of 0.335 nm is the thickness of the single-layer graphene. Since the mica was used as a substrate material in the AFM measurements, the height was accepted as 0.4 nm [46]. The thickness (t measured ) values of the graphene samples were measured as 1.64, 2.15, and 7.28 nm for G-DMSO, G-DMF, and G-PA, respectively. The number of layers was calculated from the aforementioned equation as 3.69 4, 5.22 5, and 21.05 21 for G-DMSO, G-DMF, and G-PA, respectively. When all the results were assessed, the roughness and the layer number values of G-DMSO and G-DMF are smaller than that of G-PA. The layer numbers of graphene products obtained from XRD agree with those estimated from the AFM. It can be inferred from these findings that while G-DMSO and G-DMF include fewer layers, G-PA has a multi-layered structure. In the case of preparation of graphene in PA, the RMS and layer number are 19.3 and 21.0 nm, respectively. For the sample of G-DMSO, the RMS and layer number are 5.7 and 4.0 nm, respectively. The multilayered specimen of G-PA has a higher RMS value, which is in accordance with the literature [47]. It is believed that the combined effect of high-power ultrasound energy and strong acidity of PA might trigger the formation of some functionalities on the graphene, leading to an increase in the surface roughness. Additionally, the reason for the higher thickness may be due to the out-of-plane rippling behavior of graphene [48].
In order to clarify the graphene structure, the UV-vis spectra of the graphene dispersions were recorded as seen in Figure 6. The synthesized graphene products were characterized by comparing the properties with that of CG. The graphene samples labeled as G-DMSO, G-DMF, and G-PA, showed a peak at the 265-nm wavelength referring to sp 2 C = C bonds [49]. Next, the dynamic light scattering (DLS) technique, which gives the apparent size of the graphene samples in the aqueous dispersion, was applied to investigate the particle size [50]. The size distribution results are presented in Figure 7. Next, the dynamic light scattering (DLS) technique, which gives the apparent size of the graphene samples in the aqueous dispersion, was applied to investigate the particle size [50]. The size distribution results are presented in Figure 7. Table 1 gives the particle size diameters, polydispersity indexes, and zeta potentials. As seen in Table 1, the Z-average hydrodynamic radius (Rh) values of G-DMSO, G-DMF, and G-PA are 6938 ± 408, 3846 ± 18.5, and 7137 ± 2.5 nm, respectively. These results are parallel to the previous studies, which reported graphene samples with a few micrometers of Rh with a less defected structure. The graphene with these particle sizes is promising and favorable for applications in the electronics industry [28,51]. According to the measurements performed by the DLS technique, the particle domain sizes of G-DMSO and G-PA samples demonstrated bimodal distributions, with PDI values of 0.692 ± 0.308 and 0.629 ± 0.150, respectively. Nevertheless, the smallest particle size and PDI value were obtained as 3846 ± 18.5 and 0.307 ± 0.056, respectively, for the graphene sample exfoliated in DMF solvent. For G-DMSO, the major part of the material was detected at 2055 nm and the rest of the material was observed at 4103 nm. The majority of the material in the G-PA sample shows an average diameter of 1990 nm. The smaller concentration of particles presents an average size of about 4074 nm. As it is seen from Table 1, zeta potentials (mV) are measured as −8.76 ± 15.4, −11.7 ± 3.38, and −3.49 ± 4.09 for G-DMSO, G-DMF and G-PA, respectively. For all the sonication solvents, zeta potential values were negative due to the interfacial Lewis charge transfer between the graphene particles and the solvent molecules [30]. Figure 8 illustrates the TEM images of graphene products of G-DMSO, G-DMF, and G-PA.  Table 1 gives the particle size diameters, polydispersity indexes, and zeta potentials. As seen in Table 1, the Z-average hydrodynamic radius (Rh) values of G-DMSO, G-DMF, and G-PA are 6938 ± 408, 3846 ± 18.5, and 7137 ± 2.5 nm, respectively. These results are parallel to the previous studies, which reported graphene samples with a few micrometers of Rh with a less defected structure. The graphene with these particle sizes is promising and favorable for applications in the electronics industry [28,51]. According to the measurements performed by the DLS technique, the particle domain sizes of G-DMSO and G-PA samples demonstrated bimodal distributions, with PDI values of 0.692 ± 0.308 and 0.629 ± 0.150, respectively. Nevertheless, the smallest particle size and PDI value were obtained as 3846 ± 18.5 and 0.307 ± 0.056, respectively, for the graphene sample exfoliated in TEM analysis revealed that G-DMSO and G-DMF have fewer layers while G-PA has multilayers. The light-and dark-colored parts (Figure 8a,b) represent the few-layered structures at the edges of the sample and the multilayered agglomerates at the surface, respectively. Additionally, Figure 8a,b exhibit the sheet-like flakes of the graphene structure overlapping at some parts. Figure 8c presents a wrinkled and unordered structure of G-PA, which was synthesized by excessive cavitation in PA. Due to the high acidic character of PA, crumbled structure, and deficiencies, low-quality properties were observed on the surface morphology of the sample. Moreover, the dark-colored parts on the image of G-PA display the contamination arising from the residual solvent [52].
As it is seen from Table 1, zeta potentials (mV) are measured as −8.76 ± 15.4, −11.7 ± 3.38, and −3.49 ± 4.09 for G-DMSO, G-DMF and G-PA, respectively. For all the sonication solvents, zeta potential values were negative due to the interfacial Lewis charge transfer between the graphene particles and the solvent molecules [30]. Figure 8 illustrates the TEM images of graphene products of G-DMSO, G-DMF, and G-PA. TEM analysis revealed that G-DMSO and G-DMF have fewer layers while G-PA has multilayers. The light-and dark-colored parts (Figure 8a, b) represent the few-layered structures at the edges of the sample and the multilayered agglomerates at the surface, respectively. Additionally, Figure 8a and 8b exhibit the sheet-like flakes of the graphene structure overlapping at some parts. Figure 8c presents a wrinkled and unordered structure of G-PA, which was synthesized by excessive cavitation in PA. Due to the high acidic character of PA, crumbled structure, and deficiencies, low-quality properties were observed on the surface morphology of the sample. Moreover, the dark-colored parts on the image of G-PA display the contamination arising from the residual solvent [52].

Conclusions
In summary, the ultrasound assisted LPE method was used for the synthesis of graphene in the solvents of DMSO, DMF, and PA. According to the XRD results, the layer numbers of the samples labelled as G-DMSO, G-DMF, and G-PA were estimated as 9, 10, and 20, respectively. The UV-vis spectra of all the samples showed a peak at the 265-nm wavelength, indicating the sp 2 C = C bonds

Conclusions
In summary, the ultrasound assisted LPE method was used for the synthesis of graphene in the solvents of DMSO, DMF, and PA. According to the XRD results, the layer numbers of the samples labelled as G-DMSO, G-DMF, and G-PA were estimated as 9, 10, and 20, respectively. The UV-vis spectra of all the samples showed a peak at the 265-nm wavelength, indicating the sp 2 C = C bonds of graphene. Additionally, the results of AFM showed that the layer numbers were 4, 5, and 21. The zeta potentials (mV) were measured as −8.76 ± 15.4, −11.7 ± 3.38, and −3.49 ± 4.09 for G-DMSO, G-DMF, and G-PA, respectively, whereas the Z-average hydrodynamic radius (Rh) was 6930 ± 408, 3846 ± 18.5, and 7137 ± 2.5 nm for G-DMSO, G-DMF, and G-PA, respectively. XRD, AFM, and TEM revealed that G-DMSO and G-DMF contain few layers while G-PA has a multilayer structure. Finally, it can be concluded that DMSO is a promising solvent for the one-pot synthesis of few-layered graphene by the LPE method without using any surfactants or ionic liquids.