The GO solution was synthesized by the modified Hummers method, the synthesis details are described in the experimental Sect. 31. The prepared solution was cast onto a commercial DVD and allowed to dry completely. The GO-coated DVD is then inserted into the lightscribe driver. The commercial DVD burner has a 780 nm laser with a diameter of 20 µm, which scans the surface of a GO-containing DVD to convert it to rGO. By designing the path, laser scanning can reduce GO into any desire rGO patterns. This reduction process can be done in few minutes and rGO can prepare quickly in a controlled manner. Figure 1a-c shows the schematic of casting GO solution on a commercial DVD and reduction process by the lightscribe driver. The photograph of the lightscribe patterned rGO on the DVD is shown in Fig. 1d, in which the dark patterned layers refer to the rGO and the lighter parts are the GO film.
In order to increase the effective surface area of rGO layers for supercapacitor applications, the prepared films were exposed to oxygen (O2) and sulfur hexafluoride (SF6) gases at room temperature. Accordingly, various controlled processes such as plasma irradiations and lightscribe reduction were integrated to get the best results. In this regard, the following samples were prepared: GO, rGO, SF6 plasma-treated GO named as FGO, SF6 + O2 plasma-treated GO labeled as (F + O)GO, lightscribe FGO named as rFGO, lightscribe (F + O)GO called r(F + O)GO, SF6 plasma-treated rGO named as FrGO, lightscribe FrGO called rFrGO, and finally, lightscribe (F + O)rGO labeled as r(F + O)rGO.
The schematic of the r(F + O)rGO fabrication process is presented in Fig. 2. Based on this, the GO solution is first cast on a DVD and after drying, it converts into rGO by the lightscribe method (Fig. 2a). The reduced part is then exposed to SF6 and O2 containing plasma for 5 min (Fig. 2b). This process led to the formation of a three-dimensional structure of rGO on the substrate. However, functional groups originating from ion bombardments can enter into the rGO layers and degrade their reduction rate32. The sample is then re-irradiated with the laser to facilitate the elimination of the functional groups introduced during plasma bombardment (Fig. 2c). In fact, these sequential processes contribute to the formation of a three-dimensional, high conductive and large surface area rGO structure.
Figure 3 shows the SEM images of all prepared samples. The morphology of the GO layers is seen in Fig. 3a. After SF6 plasma treatment, it is observed that most of these layers are etched as shown in Fig. 3b. In the case of (F + O)GO, the addition of O2 gas leads to a decrease in the etching rate of SF6 and deeper penetration of the fluorine radicals which results in the controlled formation of porosity, as can be seen in Fig. 3c. After lightscribe reduction of GO, the layered structure of rGO sheets can be seen due to the expansion of layers as a result of the gasification of oxygen functional groups33 (Fig. 3d). A similar trend is observed for rFGO in comparison with FGO as shown in Fig. 3e. By lightscribe irradiation of (F + O)GO, it expects to be as the r(F + O)GO. According to Fig. 3f, the expanded layers originate from the reduction process which shows that the underlying sheets are less exposed to plasma and subsequently have less porosity. Comparison of FGO and FrGO samples also demonstrates that in the latter there are more opened layers with higher porosity, which highlights that the plasma treatment affects more effectively on the rGO than GO in creating porosity (Fig. 3g). The rFrGO sample also displays that the second reduction process can remove some of the heavily etched layers from the film due to further expansion and gasification of fluorine functional groups, as can be seen in Fig. 3h. Generally, the expanded rGO sheets can effectively interact with applying plasma ions compared with GO sheets. Moreover, the SEM images of Sf6 and SF6 + O2 plasma-treated samples show that the latter results in a much greater porosity and larger surface area. Therefore, according to SEM results, it can be concluded that the r(F + O)rGO sample offers a more three-dimensional structure in comparison to other introduced samples (Fig. 3i).
Figure 4a shows the TEM image of GO sheets synthesized by the modified Hummers method where are exfoliated by the sonication process. The layered structure of the GO sheets is well visible in the image. TEM image of the one FGO layer is presented in Fig. 4b. Accordingly, FGO is locally etched after SF6 plasma treatment and lost its integrated layered form. Moreover, the TEM image of the r(F + O)rGO sample exhibits its layered structure with more wrinkles due to reduction and plasma treatment processes (Fig. 4c).
Figure 5a presents the XPS result of the O2 + SF6 plasma-treated rGO sample which is named as (F + O)rGO where F and O stand for SF6 and O2 gas plasmas, respectively. Accordingly, strong C-F and C-F2 bonds appear after O2 + SF6 plasma irradiation that refer to the effective fluorination of rGO sheets34. Vacancy defects in the rGO are reasons for generating these fluorine bonds in the plasma-exposed rGO35. Although the formation of fluorine bonds improves the energy storage capacity of treated rGO layers, it greatly reduces their conductivity, which will be discussed later. Therefore, the second reduction step is necessary to improve the conductivity of (F + O)rGO sample. Figure 5b refers to the XPS spectrum of the r(F + O)rGO sample in which the contents of fluorine decrease thanks to the lightscribe process that supplies the energy needed to eliminate fluorine-containing functional groups. It is also expected that the conductivity of the r(F + O)rGO significantly increases. As can be seen in Fig. 5b, after the second reduction process, a small content of fluorine bonds remain, which change the electronic structure of rGO locally (due to the high electronegativity of the fluorine atoms) and play an effective role in increasing the specific capacity of the r(F + O)rGO36. The Raman spectra of the GO, light-scribed GO (rGO), and r(F + O)rGO are shown in Fig. 5c. As can be seen, all samples exhibit typical disorder D, graphitic G, and amorphous 2D oscillation modes at around 1351, 1585, and 2590 cm− 1, respectively37. The ID/IG ratio is calculated for all three samples, which can show a degree of defects in the graphene layers8,38. According to the results, for GO, the value of this ratio is 0.77, and increases to 0.81 for rGO, which can be due to the elimination of oxygen-containing functional groups and their vacancy as defects39. Interestingly, in the r(F + O)rGO sample, this ratio increases to 0.91 arising from the formation of more defects in layers due to the bombardment and removal of F-containing bonds after the second laser scribe reduction step.
Figure 6a displays the surface resistance measurement of all prepared samples. Based on it, the r(F + O)rGO sample has a resistance of about 82.8 Ω/sq that is the lowest surface resistance among all samples. Its comparison with rFrGO sample shows that the use of combined O2 and SF6 plasma gases results in a more efficient reduction of plasma-treated rGO. The reason behind this fact is due to the impact of oxygen ions that promote the penetration of fluorine radicals into deeper layers and provide more interaction with rGO. As a result, the second lightscribe process reduces the plasma-treated rGO layers more effectively. After SF6 bombardment of GO sheets, the resistance of the FGO sample increases, which can be due to the replacement of oxygen-related functional groups with fluorine, as well as the local etching of layers. After lightscribe irradiation, the resistance decreases in rFGO due to the elimination of surface fluorine and deeper oxygen functional groups. Interestingly, the electrical resistance of r(F + O)GO is lower than that of rFGO, indicating the effective role of oxygen ions in its improvement. Furthermore, by applying plasma to the rGO samples lower resistance can be achieved compared with GO samples because the reduction process leads to the opening of the layers and provides larger surfaces for interaction with plasma ions. The electrochemical tests of all fabricated electrodes are also presented in Fig. 6b. All samples were tested in a 5 mM K3Fe(CN)6 with 100 mM KCl supporting electrolyte at a scan rate of 50 mV/s to evaluate their electrochemical performance as a reliable and standard method. To perform these tests, all samples were cut into 1×1 cm2 pieces and used in a three-electrode standard electrochemical cell as working electrodes where platinum and Ag/AgCl were used as a counter, and reference electrodes, respectively. According to the results, The r(F + O)rGO sample has the highest electrochemical activity. By comparing the reduction peaks of the samples, it can be seen that the highest current of 748 µA is measured for r(F + O)rGO followed by 412 µA for rGO. Therefore, r(F + O)rGO electrode demonstrates better electrochemical performance compared to other rGO samples. Interestingly, in the rFrGO sample, the reduction peak occurs at 373 µA, which has a degraded performance than rGO. The oxygen-free plasma means limits the porosity to the surface of graphene where not allow the fluorine radicals to penetrate the deeper layers. Therefore, the incorporation of reduction and SF6 + O2 plasma treatments which is result in r(F + O)rGO electrode offers more porosity, higher conductivity and larger surface area that deliver superior electrochemical performance.
Figure 7a-d shows the charge and discharge curves of the r(F + O)rGO based supercapacitors that experienced different plasma exposure times from 0 to 15 min. All electrochemical tests were performed using hydrogel electrolyte composed of PVA + H2SO4. The charging-discharging currents have been normalized per effective volume of rGO in the electrodes (see supporting information). As can be seen, by increasing the plasma exposure times, charging and discharging times have increased significantly. Furthermore, the coulombic efficiency parameter which is equal to the ratio of discharge time to charging time is calculated for 0, 5, 10 and 15 min plasma-exposed samples equivalent to 72%, 89% and 94%, and 96%, respectively. Moreover, prolonged plasma exposure time leads to higher supercacitors performance where the capacitance can be accurately calculated using the slope of the discharge cycle according to follow:
Cdevice = I / (-dV/dt) (1)
The normalized capacitances of the 0, 5, 10, and 15 min plasma-exposed electrodes were calculated around 1.1, 5.2, 8.0, and 10.2 F/cm3 for the first discharge cycle, respectively (see supporting information and Figure S1 and S2). The obtained results show that incorporation of laser irradiation-15 min plasma exposing-laser irradiation sequences terminates in the highest final performance of the supercapacitors as it improves the capacitance by up to 10 times compared to the case without plasma treatment.
It is noteworthy that with increasing the plasma exposure time from 15 min onwards, the effective thickness of the rGO sample decreases, and thus the capacitance decreases significantly. The Figure S3 shows the changes in capacitance per unit volume as a function of F + O plasma bombardment time. It is observed that the maximum capacitance is obtained for the plasma exposure time of 15 min and its increase is accompanied by a significant decrease in specific capacitances.
The cyclic voltammetry test of the r(F + O)rGO supercapacitor with different plasma exposure times is presented for 50 consecutive cycles in Fig. 8a. The overlapping of the CV curves for each sample refers to their stable performance. To test the supercacitors, one terminal of the supercapacitor is connected to the counter and reference ports of a potentiostat at the same time, and the other terminal is connected to the working port of the potentiostat. The area inside the cyclic voltammeter curve shows the amount of power density stored by the supercapacitor. As can be seen, the internal area of the 15 min plasma exposed r(F + O)rGO electrode (red curve) has the highest value compared to other samples that mean higher power and high capacitance. Figure 8b shows the capacitance of the r(F + O)rGO supercapacitors with different plasma exposure times for 1000 consecutive cycles. Accordingly, the capacitance remains constant as the charge and discharge cycles increase which indicates the high-stable performance of these supercapacitors. Moreover, r(F + O)rGO electrode with 15 min plasma treatment demonstrates an average capacitance of 10 F/cm3, higher than all reported supercapacitors based on lightscribe technique. Figure 8c presents the results of EIS analysis of supercapacitors upon different plasma exposure times. With increasing plasma irradiation time, EIS curves have shifted towards smaller real and imaginary values, referring to an increase in the relative conductivity of electrodes and efficient charge exchanges. Moreover, the slope of the linear part of the curves has increased as plasma time is increased, which demonstrates a larger Warburg capacitance. The 15 min plasma-treated r(F + O)rGO electrode has a larger surface area due to its higher porosity compared with other electrodes which give a higher chance of EDL (Electrical double layer) formation in the vicinity of the electrode. From the point of view of equivalent series resistance (ESR), the EIS curves show no significant change in the resistance of electrodes with different exposure times. Since the operating voltage range of aqueous hydrogel polymer electrolyte based devices can not exceed 1 V, that make them unusable for some applications. Therefore, using this devices in a tandem structure can adapt them for higher voltage applications. Figure 8d presents the 15 min-r(F + O)rGO based two-electrode supercapacitor that is charged with a polymer gel electrolyte in a charging cycle of 0.95 V. Inset of Fig. 8d also shows the practical performance of two r(F + O)rGO micro super capacitors in series with a charging cycle of 1.9V to illuminate a light-emitting diode (LED) with an average current of 820 µA, keeping it on for 23 min.
Table 1 shows the performance of lightscribe graphene-based supercapacitors and their comparison with our introduced supercapacitors. Accordingly, the specific capacitance obtained in our work is about 10.2 F/cm3, which shows a significant improvement compared to the capacitance of other works reported in the range of 134.0 mF/cm3 to 3.0 F/cm3.