Experimental comparison of the effects of the two designed probes for exfoliation of graphene by sonication

Abstract

In the present research, comparative simulation and experimental investigation were carried out for two different probes to improve the efficiency and quality of few-layer graphene utilizing liquid-phase exfoliation. Two differently designed shaped probes are fabricated for this study, i.e., a simple probe and a stepped probe. The acoustic pressure distribution in the vessel is simulated for both probes at different output powers. In the experiment, both probes exfoliate graphite powder in a mixture of deionized water and ethanol. Different sonication times and output power were studied for different pulse modes. The graphene layers were characterized using Ultraviolet–visible spectroscopy (UV‒Vis), Scanning Electron Microscopy (SEM), Transmission Electron Microscopy (TEM), and Raman Microscope. The simulation shows that the total displacement on the tip of the stepped probe is 5.7% greater than that of the simple probe. Also, the pressure difference produced by the stepped probe is 9.15 × 10⁶ pa compared to the pressure difference of the simple probe which is 8.23 × 10⁶ pa. Consequently, the stepped probe was more effective in exfoliating graphene. The experimental results show that the absorbance peak in the stepped probe is approximately 32% greater than the absorbance peak in the simple probe with the same output power. Also, the graphene with better quality is produced with the stepped probe compared to the simple probe, which verifies the simulation findings. Additionally, the optimum output power, pulse duration, and sonication time to produce graphene with less time and energy consumption are obtained. The best graphene flakes were obtained via sonication with a pulse duration of 0.7 s, an output power of 282 W, and a sonication time of 45 min using a stepped probe. UV–Vis., SEM and TEM analysis show increases in quantity and quality of the exfoliated flakes. These results are approved by the peak ratio of I_2D/I_G around 1 in Raman spectra revealed the production of the few-layer graphene by the simple probe and the bilayer graphene by the stepped probe. So, this method is suitable for the production of graphene flakes from graphite in suitable quantity and quality.

Introduction

Graphene is a single layer of graphite with atoms arranged in a hexagonal shape that has recently been extensively investigated due to its exciting properties, such as high thermal and electrical conductivity, optical transparency, high flexibility, good mechanical strength, and large surface area^1–5. It is the thinnest and lightest material and is used in a variety of applications, such as micro—and optoelectronics, energy conversion, photonics, solar cells, and transistors^6–9. Recent applications of graphene include fire detection¹⁰, improving rechargeable battery performance¹¹, and Sound detection¹².

The sp² carbon atoms are hybridized in graphene, forming 3 σ-bonds with three neighboring carbon atoms. Additionally, graphene has one π-bond that is perpendicular to the surface and causes electrical and thermal properties in graphene¹³. Several techniques have been employed to synthesize graphene sheets^14–16. Among these methods, liquid-phase exfoliation (LPE) of graphite by ultrasonic waves is one of the low-cost and easiest methods; by optimizing conditions, significant improvements in production efficiency have been reported¹⁷. Using a probe-type sonicator allows control of properties such as size, thickness, and number of layers and improvement in product yield compared to ultrasonic bath sonication¹⁸. A probe-type ultrasonic processor contains a power supply, a transducer that converts the electric signal to mechanical vibration, a booster that amplifies the amplitude of the vibration, and a probe that transmits the wave through the media¹⁹. When passing through a solution, ultrasonic waves produce periodic high- and low-pressure reigns in the solution. These pressure difference in the LPE method is the main parameter to exfoliate graphene. The method involves two steps for producing graphene. In the first step, ultrasonic waves overcome the Van Der Waals interaction between graphene layers of graphite in an appropriate solution. In the second step, large graphite particles are removed²⁰. Many techniques, such as varying the output power of the probe, solution concentration, sonication time, and controlling the temperature of the solution, can improve the efficiency of the products by using an ultrasonic processor which have been partly investigated to some extent^2,21,22.

Most of these studies have produced graphene using available normal probes. However, the effect of the probe's shape as an important parameter has not received enough attention. For the simple probe type, the cavitation region in the liquid medium is around the tip of the probe, and other areas of the liquid are less affected. Therefore, the reaction in the liquid near the tip is more intense than in the surrounding areas. Designing probes with different configurations can change the cavitation zone and cause the energy propagate in larger volumes of the liquid²³. We hypothesize that the shape of the probe has an important effect on graphene production and this is one of the important aims investigated in this research. therefore, the impact of pulse duration, power irradiation, and time duration on the quality and quantity of graphene layer is investigated.

Consequently, computational simulations were employed to design two probes with simple and stepped shapes. Additionally, the pressure distribution in the media (water–ethanol) was simulated for both probes under different output powers and the maximum pressure difference was obtained which we expected cause more exfoliation. The simulation results were verified by the experimental results for both of the fabricated probes at different output power. To determine the optimum exfoliation condition, the experiments carried out in different radiation modes, and sonication times in this research. These parameters are necessary for producing the maximum amount of graphene with minimal defects.

Materials and methods

Graphite powder with a size less than 50 μm, and 99.8% purity obtained from Sigma-Aldrich is used in this study. The LPE medium is composed of deionized water and ethanol with 99.9% purity. Sonication of the graphene is performed with two titanium probes (Sonotrode, Horn) made from an industrial titanium alloy grade 5 (90% Ti, 6% Al, 4% V). These probes were designed utilizing simulated using the finite element method (FEM), and then manufactured. The ultrasonic processor (Fapan Co. Ltd- Iran, Model 1200UPS) was used for the experimentation. The first probe has a simple shape with a tip 40 mm in diameter, and the second probe has a stepped shape with a similar tip diameter. The operation frequency of both probes is approximately 20 kHz, and the maximum nominal power is 1200 W (RMS power of 300 W). After the graphite powder was sonicated in a solution of water–ethanol, the mixture was centrifuged at 3500 rpm with a centrifuge device (Behdad Co. Ltd.–Iran, 16 branches Universal Centrifuge Model). After centrifugation, the top of the solution, which is expected to contain few layers of graphene, is separated. To analyze the solution which we expected to contain graphene, the absorbance of the gathered suspension was measured using a UV‒Vis spectrometer (Perkin-Elmer Lambda 25). A scanning electron microscope (SEM) (TESCAN, VEGA 3 operated at 150 kV) and a transmission electron microscopy (Zeiss EM10C –Germany operated at 100 kV) are used to identify the morphology of the layers. The quality of the graphene layers is subsequently characterized using Raman spectroscopy using Takram N1-541 spectrometer operated at 532 nm.

Simulation

Both probes are simulated by FEM method, optimized and then fabricated. The simulation is carried out for the probe's operation in a water–ethanol solution in the simulated vessel.

Simulation of the probes

An optimal probe must have maximum displacement at the tip and good impedance matching with the solution to achieve the best wave transfer efficiency. For operation of the probe at the resonant frequency, all the components must be multiplied by λ/2 in length²⁴. A two–dimensional axisymmetric finite element simulation model of the probes is established. The diameter of the titanium alloy rod used was 45 mm for both probes, with a tip diameter of 40 mm and the specifications including a speed of sound of 5090 m/s, a density of 4400 kg/m³ and a Young’s modulus of 114 × 10⁹ Pa.

A linear elastic material module is used for the probe simulation. Newton's second law governs the equation²⁵ used:

$$-\rho {\omega }^{2}u=\mathrm{\nabla }.s$$ 1

and

$$-i\omega =\lambda$$ 2

where u is the particle displacement, ρ is the density of the probe material, ω is the angular frequency and S is the stress. The novel stepped probe has a more irradiating surface compared to the simple probe. Additionally, the stepped probe generates higher cavitation and more mechanical waves in the solution. Figure 1 shows both of the designed probes operating at a frequency of near to 20 kHz. The total displacement in the simple probe is approximately 1.22 × 10^–4 mm at the tip of the probe, which is obtained from the line graph of the total displacement. With the stepped probe, the displacement on the tip is 1.29 × 10^–4 mm, which is 5.7% greater than that of the simple probe. As the ratio of the diameter of the upper part of the probe to the lower part of the probe be more, amplitude gain will be more. Therefore, amplitude gain of the stepped probe is more than the simple probe. For the simple probe, only lower surface of the tip of the probe moves up and down, while for the stepped probe in addition to the lower surface of the tip, an edge above the tip is also moving up and down, therefore more graphene flakes are expected.

Fig. 1.

Simulated displacement for both probes with 40 mm diameters: (a) simple probe and (b) novel stepped probe.

Simulation of the acoustic pressure distribution

To simulate pressure distribution at different ultrasonic output powers, the wave equation is solved by the finite element method. We assumed linear wave propagation in homogenous liquid media with neglected shear stress²⁶. Therefore, the wave equation gets the form:

$$\frac{1}{\rho }{\mathrm{\nabla }}^{2}p-\frac{1}{\rho c^2}\frac{{\partial }^{2}p}{\partial t^2}=0$$ 3

In this equation, P is the acoustic pressure, C is the speed of sound and ρ is the density of the media. The frequency is constant (~ 20 kHz), so the pressure is considered to be temporally harmonic, so:

$$p\left(r,,,t\right)=p\left(r\right)e^{i\omega t}$$ 4

where ω = 2πf. The space dependence of the wave equation is:

$$\frac{1}{\rho }{\mathrm{\nabla }}^{2}p+\frac{{\omega }^2}{\rho c^2}p=0$$ 5

The pressure is calculated from

$$\text{p}=\sqrt{2I\rho c}$$ 6

where (I) is the intensity of radiation transferred by the transducer to the solution in the vessel. The intensity of the active surface of the probe (A = πr²) is obtained by

$$I=\frac{p_u}{A}\left(\text{W}/{\text{m}}^2\right)$$ 7

where P_u is the output power. The simulation is carried out on a 2D axisymmetric model. The boundary conditions are assumed to be as follows: the upper edge of the vessel is a liquid medium which is in contact with the atmosphere; thus, p = 0. For another boundary where the liquid is in contacts with the walls of the container, $\frac{\partial p}{\partial n}$ = 0; then, distribution of the acoustic pressure zone is varied by changing the effective parameters in the process²⁷.

Sonication for graphene exfoliation

In this research graphene is exfoliated from the sonication of graphite using both a simple probe and a stepped probe. Graphite powder (0.07 gr) was added to a 97 mL volume of the solution in the vessel. The solution consisted of mixture of 35% ethanol and 65% deionized water. We used a water‒ethanol mixture as a green solution. The ratio of ethanol to deionized water was calculated using the Connors–Wright equation and the surface tension of the solution is 46 mJ/m². The obtained surface tension is in the range of 40–50 mJ/m² that is the optimal surface tension for exfoliation of graphene²⁸. To prevent increasing the temperature of the solution during the process, the vessel was immersed in a water–ice bath. The designed and fabricated probes and the setup of the experiment are shown in Fig. 2. The effects of the different ultrasonic output powers in pulse mode and sonication time duration given in Table 1 are the parameters that were investigated. The experiments were carried out for an output power ranging from 200 to 300 W and sonication times ranging from 25 to 65 min. The effect of the pulse duration on the optimal output power and irradiation time was studied from 0.5 to 0.9 s per second. After the sonication process, the thick flakes of graphite were removed via centrifugation. Centrifugation was performed for 30 min at 3500 rpm. The top transparent section of the supernatant was collected for further analysis. Our expectation is that more graphene will be produced for the probe, which will result in a greater pressure difference in the solution during irradiation. These pressure difference overcome the Van Der Waals force between graphite sheets to exfoliate graphene²⁹. The results for the experimentally exfoliated graphene are compared with the simulated pressure difference for both probes which is interesting.

Fig. 2.

The experimental devices used for exfoliation: (a) the setup of the experimental process, (b) the titanium stepped probe, and (c) the titanium simple probe.

Table 1.

The important ultrasonic varied parameters for ultrasonic treatment for exfoliation of graphene.

Sonication time (min)	25	35	45	55	65
Output power (W)	200	245	268	282	300
Pulse duration (% sec.)	0.50	0.60	0.70	0.80	0.90

Results and discussion

The goals of this research are to enhance the yield of graphene layers generated from graphite powder. Then, a comparison between the simple-probe and the stepped-probe simulation and experimental results was performed to determine the optimal operating parameters for producing few-layer graphene. Initially, the acoustic pressure distributions in the vessel with different output powers are simulated for both the probes. The effects of the output power, pulse duration, and sonication time duration were subsequently measured experimentally. The best conditions according to the simulations and experimental results were selected for further research.

The output power of the transducer is one of the important factors that can influence the pressure distribution in the vessel. Variations in output power causes the probe's tip pressure in the liquid medium to vary and lead to change in the pressure distribution zone²⁹. Due to the constant volume of the solution, the power density was selected to be similar for both probes for each output power. The pressure distribution in the reactor vessel was simulated for the simple probe and the stepped probe at various output powers. The simulated pressure distributions in the reactor vessel for the simple probe and the stepped probe at 282 W of output power are shown in Fig. 3. The higher and lower pressure zones are almost the same for both probes. The diameter of both probes on the tip was 40 mm, so the intensity of irradiation (W/m²) was the same for both probes. The maximum and minimum pressures and pressure differences for 282 W of output power for the simple probe and the stepped probe are presented in Table 2. The maximum and minimum pressures increase with increasing output power from 200 to 300 W for both probes. Simultaneously, cavitation and ultrasonic waves increase with increasing pressure difference, so a large amount of energy is produced in the liquid. This energy overcomes the Van Der Waals force between the graphene layers and increases the exfoliation efficiency in the reactor vessel²⁹. Figure 4 shows the simulated pressure difference for the simple probe and the stepped probe as a function of the output power. The trend of the pressure difference for both probes at different supplied powers is almost the same, although the pressure difference for the stepped probe is much greater than the pressure difference for the simple probe. The highest-pressure difference (Δp) for 282 W of output power is 9.15 × 10⁶ Pa in the medium for the stepped probe compared to that for the simple probe, which is a significant increase.

Fig. 3.

Distribution of the acoustic pressure in the reactor vessel at 282 W of output power: (a) simple probe, (b) stepped probe.

Table 2.

Simulated pressure and pressure difference (Δp) at 282 W output power for the simple probe and the stepped probe.

Output power (282 W)	Simple-probe	Stepped-probe
Maximum pressure (Pa)	3.72 × 106	4.15 × 106
Minimum pressure (Pa)	− 4.51 × 106	− 5.00 × 106
Pressure difference (Pa)	8.23 × 106	9.15 × 106

Fig. 4.

Comparative Pressure difference for different output powers for the simple probe and the stepped probe.

For characterizing the exfoliated graphene, UV‒Vis spectra of the suspensions were taken at wavelengths ranging from 200 to 400 nm. A peak centered at approximately 264 nm was attributed to π-π^* bonds for all the samples³⁰. The position of the π–π^* bond absorbance changes as a function of thickness. The intensity and position of this peak vary in different environments, particularly for thick layers where more graphene mass is in contact with the media³¹. The absorption peak in the simple probe is at 262 nm for all the output powers. This means the flakes with specific number of layers having absorbance at 262 nm are maximum number in the solution. In the UV‒Vis spectra for the stepped probe, the samples with 200 W and 245 W exhibit a peak at 262 nm. The position of this peak shifts to 266 nm for output power of 268 W, 282 W, and 300 W. For both probes, the intensity of the peak increases with increasing output power, indicating that the concentration of the graphene sheets increases remarkably with increasing output power. An increase in output power leads to more exfoliation, which confirms the simulation prediction in Fig. 4, which shows an increase in the pressure difference (ΔΡ). In Fig. 5, the UV‒Vis shows an increase in the production of graphene in the solution provided by sonication energy to overcome Van Der Waals forces to separate the graphene layers. The absorption peaks for the stepped probe are more intense than those for the simple probe at all output powers. Additionally, the position of the peak is shifts to a longer wavelength of 266 nm for the stepped probe, which means that the graphene layers produced by this probe contain thinner layers compared to those produced by the simple probe³². Therefore, the stepped probe exfoliates graphene more efficiently under the same conditions compared to the simple probe.

Fig. 5.

UV‒Vis spectra for the stepped probe and simple probe at 200 W, 245 W and 282 W output powers.

The absorption peaks for the different output powers are shown in Fig. 6. This figure shows that more graphene is exfoliated for the stepped probe than for the simple probe at all powers. Additionally, there is a greater increase in the slope between 268 and 282 W of output power. The absorbance peak in the stepped probe for 282 W is approximately 32% greater than the absorbance peak in the simple probe at the same output power. At 282 W, the maximum exfoliation is observed, and for 300 W output power, the graphene production is reduced.

Fig. 6.

Comparative absorbance of UV‒Vis spectra of suspended graphene flakes produced at different output powers for the simple probe and the stepped probe.

Figure 7 shows scanning electron microscopy (SEM) images of the graphene flakes produced by the stepped probe and simple probe at 282 W. The SEM images in Fig. 7 clearly show the exfoliated graphene and the sheet-like morphologies. The size of the flakes increases with increasing output power from 200 to 282 W (not shown). Increasing the output power to 300 W provides graphene flakes with a decreased size. The observed decrease in flake size at 300 W output power may be due to extra radiation intensity that crashes graphene sheets because of more shear stresses acting upon it or may cause agglomeration of the exfoliated layers. Furthermore, the flakes produced at 282 W output power by the stepped probe exhibit better quality compared to the other samples. Also, from the image it's clear that the layers that provided by the stepped probe are more transparent indicating that they are thinner. A transmission electron microscopy (TEM) images of the graphene flakes produced by the simple probe and the stepped probe at 282 W are shown in Fig. 8. It is obvious from the TEM micrographs that the flakes are ultrathin and transparent but the flakes produced by the stepped probe showed the good quality and appear as few-layer graphene (FLG). The TEM images confirm the SEM images in Fig. 7. No defects are visible in the layers produced by sonication, and the size of the flakes is sufficiently large, which confirms the good exfoliation conditions for the production of few-layer graphene.

Fig. 7.

SEM images of graphene flakes after 45 min. Sonication by (a) the simple probe and (b) the stepped probe at 282 W (scale bar 2 µm).

Fig. 8.

TEM micrograph of the graphene flakes exfoliated at 282 W by the (a) simple probe, (b) stepped probe for 45 min.

Another investigated parameter in this research is variation in the pulse duration with respect to the output power. The pulse duration output power is expected to differ from that of continuous ultrasonic irradiation; therefore, in this research, the effects of the pulse duration output power were tested and presented. The pulse duration was defined as the percentage of time in one second that the probe irradiated. For instance, when pulse duration of 70% is chosen, the probe irradiates for 0.7 s and does not irradiate for 0.3 s. Five different pulse durations, corresponding to the best power for exfoliation in this case being 282 W for the stepped probe, are defined in Table 1, where the other parameters are held constant. The UV‒Vis spectra in Fig. 9 demonstrate that for an output power of 282 W, the high absorbance peak at 266 nm increases with rising pulse duration from 0.50 to 0.70 s. This shows the presence of more graphene flakes.

Fig. 9.

UV‒Vis spectra of graphene layers exfoliated for different pulse durations.

Increasing the pulse duration to 0.80 s causes a change in the peak position to 264 nm, and another peak is observed at 226 nm, which corresponds to the absorption peak of graphene oxide. At higher powers, the amount of graphene oxide increases with increasing energy of collapse that provides sufficient energy for the bonding of dissolved oxygen in the solution with graphene. Also this energy, produces higher amount of radicals, which enhances the oxidation of graphite^17,33. The amount of graphene oxide in 0.90 s pulses was greater than that in case of 0.80 s pulses, and even the peak corresponding to graphene oxide dominated the peak corresponding to graphene. Though the vessel is in an ice bath, momentary and locally increase in temperature cause oxidation of the exfoliated graphene. SEM images of the samples treated with pulses of 0.60- and 0.90-s pulse width taken at the same magnification are shown in Fig. 10. The samples prepared with 0.50- and 0.70-s pulse widths had nearly the same distributions of size and quality. Increasing the pulse width to 0.90 s causes flake damage due to the additional duration of pulsed radiation. Therefore, this is confirmed by the SEM results shown in Fig. 10. In 0.60 s, pulse duration irradiation some layers are not exfoliated and in 0.9 s pulse duration the small layers are shaped because of long time emission.

Fig. 10.

SEM images of graphene layers produced with different pulse durations of (a) 0.60 s, (b) 0.90 s W (scale bar 2 µm).

According to the previous work by this group, the best exfoliation time for a simple probe in a similar setup is 55 min²⁹. To compare the results with those of the stepped probe, we varied the exfoliation time from 25 to 65 min. The UV‒Vis spectra in Fig. 11 and the SEM images in Fig. 12 were chosen to investigate the effect of exfoliation time. The quality of production decrease with increasing sonication time and also the layers adhere to each other and simultaneous agglomeration and crushing are observed in 65 min of sonication time. Increasing the sonication time leads to an increase in the amount of graphene produced. The maximum amount of graphene was obtained after 45 min of sonication. The position of the π–π^* peak shifted to 266 nm from 264 nm wavelength, and the intensity of the peak at 266 nm increased. After 55 and 65 min of sonication, the intensity of the peak decreased, and a weak peak at 228 nm was observed for the samples sonicated for 65 min, which indicating the presence of graphene oxide and graphene³³. The decrease in graphene amount with increasing sonication for more than 45 min may be attributed to the production of excess of graphene in the solution which increases tendency of flakes in the solution to adhere to each other due to longer exposure to solvent. Additionally, extra ultrasonic treatment can cause flakes to crash³⁴.

Fig. 11.

UV‒Vis spectra of graphene layers obtained by increasing the exfoliation time from 25 to 65 min for a 0.7 s pulse duration in 282W sonication power.

Fig. 12.

SEM images of the exfoliated graphene after sonication times of (a) 45 min and (b) 65 min. (Scale bar 2 µm).

Figure 13 displays the Raman spectra of the graphene layer produced by the simple probe and the stepped probe in 282W output power for 45 min sonication time. The measurements were taken within the range of 1000–3000 cm⁻¹. In the Raman spectrum of graphene obtained by both of the probe G-band peak is founded at ~ 1584 cm⁻¹, 2D band at ~ 2730 cm⁻¹ and D band at ~ 1350 cm⁻¹. The intensity of D peak is attributed to the amount of disorder graphene layer. According the previous work in mono-layer graphene the I_2D/I_G ratio being ~ 2–3, in bilayer graphene the ratio being 2 > I_2D/I_G > 1, and for multilayer graphene the ratio being I_2D/I_G < 1³⁵.

Fig. 13.

Raman spectra of graphene flakes produced by the simple probe and the stepped probe.

The I_2D/I_G ratio for the graphene produced by the simple probe and the stepped probe is calculated to be approximately 0.9, and 1.02 respectively. This suggests that the simple probe produces few-layer graphene, while the stepped probe may produce a mixture of few-layer and bilayer graphene. The Raman measurements are consistent with the TEM images, confirming the formation of few-layer graphene in this paper.

Conclusion

In the present research, two different probes i.e. the stepped probe and the simple probe are simulated and fabricated and then compared for preparing graphene. We also have investigated and discussed the effects of varying the output power, pulse duration of the probe irradiation and sonication time on the efficiency and quality of the exfoliated graphene layers. The stepped probe increased the graphene yield by approximately 32% under the same conditions compared to that of the simple probe. Additionally, based on the simulation and experimental results for both probes, the optimum conditions were identified when the maximum difference in the acoustic pressure in larger volume of the solution was achieved. The efficiency of the graphene produced was optimized by controlling the irradiation conditions, i.e., the sonication power, the pulse duration and the sonication time. The best flakes with good quality were obtained in a 0.70 s pulse duration with 282 W of output power by the stepped probe after 45 min of sonication. The result was validated using UV‒Vis spectroscopy. Based on the SEM and TEM images, the graphene layers are expected to be few-layer structures of good quality. The Raman measurement indicated the formation of few- layer graphene by the simple probe and bilayer graphene by the stepped probe. This method shows good improvement in graphene production by LPE method utilizing sonication with new designed probe suitable for higher quality and quantity graphene.

Author contributions

The practical parts and the simulations are carried on by E.Kh.D. under R.A. guidance. The article English text and discussion on scientific results are written by E.Kh.D. and English and scientific correction is done by R.A.

Data availability

Correspondence and requests for materials should be addressed to R.A.

Competing interests

The authors declare no competing interests.