Electron-Beam-Induced Fluorination Cycle for Long-Term Preservation of Graphene under Ambient Conditions

Abstract

The aging in air inevitably results in the accumulation of airborne hydrocarbon contaminations on a graphene surface, which causes considerable difficulties in the subsequent application of graphene. Herein, we report an electron-beam-activated fluorination/defluorination cycle for achieving a long-term preservation of CVD graphene. After experiencing such cycle, the accumulation of airborne hydrocarbon on the graphene surfaces is strongly reduced, and the initial chemical status of graphene can be restored, which is confirmed by employing atomic force microscopy and X-ray photoelectron microscopy. Our reported approach provides an efficient method for the cleaning and long-term preservation of graphene, and it is particularly useful for graphene microscopy characterizations.

1. Introduction

The emergence of graphene, an atomic thick carbon layer with honeycomb lattices, has prompted many interesting observations and promising applications [1,2]. To achieve a controlled performance of graphene in various fields, such as surface analysis, functionalization and nanoelectronic devices, a graphene sample with superior quality is highly desirable [3,4,5]. For analyzing and processing graphene on a nanoscale, the long-term endurance of graphene to the environment is of the highest importance, especially for electron beam microscopy analysis, which often requires an ultra-clean surface of graphene to achieve atomic resolution imaging [6,7,8]. Although various studies have been carried out to minimize the contamination from graphene, the surfaces cannot be totally protected from unintentional contaminations, e.g., airborne hydrocarbon [9,10]. Airborne hydrocarbon, which mainly consists of volatile organic compounds (alkanes, alkenes, aromatics, alcohols), exists almost everywhere, even in the high-end cleanroom for nanofabrication. Moreover, several studies have shown that hydrocarbon can change the wettability of a graphene surface, which affects its performance in the synthesis process. Albeit O₂ plasma treatment is efficient to prevent the contamination of the sample by hydrocarbon, such plasma approach is highly chemically reactive and thus can damage graphene, even with a low power input [11].The typical method to eliminate airborne hydrocarbon is through vacuum annealing, but the resulting surface is not sufficiently clean for many surface analysis purposes [12]. The annealing method is significantly limited by the insertion of defects under a relatively high temperature [13]. Moreover, the cleanness of the graphene surface cannot be preserved during prolonged aging. An efficient strategy is thus needed for the long-term preservation of graphene samples. Recently, Li [14] has proposed a reversible fluorination technique induced by an electron beam in vacuum. The study has shown that the initial graphene status can be restored after a fluorination/defluorination cycle. Herein, we report a technique to eliminate the airborne hydrocarbon on graphene during aging by using the electron-beam-activated fluorination/defluorination cycle. X-ray photoelectron spectroscopy (XPS) is carried out to analyze the accumulation of hydrocarbon both on pristine graphene and fluorinated/defluorinated graphene. The surface status of graphene is further investigated by atomic force microscopy. Moreover, we have studied the structural evolution of graphene after the fluorination/defluorination cycle by Raman spectroscopy and gated measurements.

2. Materials and Methods

2.1. Materials and Device Fabrication

The chemical vapor deposition of graphene on synthesized Cu foil is transferred onto a SiO₂/Si substrate with a SiO₂ thickness of 300 nm using a polymer-assisted wet chemistry method. The as-transferred graphene is then cleaned by applying a solvent (isopropanol/H₂O for 2 min at room temperature) mixture treatment to effectively remove the polymer residues on the surface. The back-gated graphene devices are fabricated through standard electron beam lithography (EBL) in a Nanobeam nB5 EBL system, oxygen plasma etching in an Advanced Vacuum Vision 320 Reactive Ion Etching (RIE) system, and metal deposition in a Lesker PVD 75 system (Kurt J. Lesker Ltd., Jefferson Hills, PA, USA).

2.2. Graphene Aging and Fluorination/Defluorination

Graphene samples are kept under ambient condition with a temperature of 20 ± 1 °C and a relative humidity of 45 ± 3% for aging. The electron-beam-activated fluorination of graphene is conducted in a FEI Strata DB235 FIB/SEM (Thermofisher, Hillsboro, OR, USA) at a vacuum condition of 10⁻⁶ mbar under an electron acceleration voltage of 5 kV. The fluorination is carried out by the irradiation of graphene with an electron beam for 300 s with a size of 100 µm × 100 µm, while simultaneously injecting the XeF₂ via GIS system in FIB (Thermofisher, Hillsboro, OR, USA). For the defluorination process, electron beam irradiation is carried out without gas injection for 100 min.

2.3. Characterizations

Raman and X-ray photoelectron spectra are acquired in Renishaw inVia Raman spectrometer (Renishaw, Kingswood, UK), with a 532 nm excitation laser, and Physical Electronics Quantera II (ULVAC-PHI, Chanhassen, MN, USA) Scanning XPS Microprobe monochrome Al Kα radiation (1486.7 eV), with a 45° angle of electron emission, respectively. The electrical measurements of the graphene back-gated devices are carried out using an Agilent B1500 semiconductor parameter analyzer (Keysight, Santa Rosa, CA, USA) inside a Faraday cage with metal probes.

3. Results

The graphene samples are synthesized by chemical vapor deposition (CVD) at a large scale and subsequently transferred onto a SiO₂/Si substrate following a solvent mixture cleaning (see the details in Materials and Methods) to minimize the contamination from the polymer residues during transferring. Figure 1 shows the X-ray photoelectron spectroscopy (XPS) results of the graphene samples C1s peak after aging for 15 days at air under a cleanroom environment. The C1s peak undoubtedly consists of 2 main components: the sp³ C peak located at 284.8 eV and the sp² C peak located at 284.0 eV. The initial larger intensity of the sp² peak can be mainly attributed to the intrinsic sp² hybridization of graphene, whereas the sp³ bonds are contributed by the defects, polymer residues, and hydrocarbon. As the aging process is underway, the concentration of sp² drops from 69.3% to 28.7%, while the sp³ bonds continuously increase, and this is mainly attributed to the accumulation of airborne contaminations on the graphene surface. Moreover, we have observed a considerable C–O peak located at 286.1 eV after 10 days of aging which can also be ascribed to the oxygen-containing contaminations coming from the air. Figure 1b,c shows the atomic force microscopy images of the graphene aged after 15 days and the initial pristine graphene, respectively. It can been seen that there is barely no topographical change on the graphene surface after aging, indicating that the evolution of the C1s signal can be explained as the accumulation of airborne contaminations, which is difficult to characterize with a normal height image.

Figure 2a illustrates the electron-beam-induced fluorination/defluorination cycle for eliminating the aging effect of the graphene surface. CVD graphene is fluorinated inside a SEM chamber by using electron-beam-activated fluorination. In this method, graphene is illuminated under the electron beam while simultaneously injecting XeF₂ gas. After the aging experiment of the as-fluorinated sample in the air, the fluorinated graphene is again exposed to the electron beam without precursor gas to perform the defluorination process (more details can be found in Materials and Methods). The C1s XPS spectra of graphene at different steps of the fluorination/aging/defluorination cycle are shown in Figure 2b (before fluorination), Figure 2c (after fluorination), and Figure 2d (after defluorination). It is clear that before fluorination the C1s peak of initial graphene is mainly composed of an sp² signal located at 284.4 eV, and after fluorination there are 4 main components in the spectrum: the sp³ peak (284.8 eV), C–F peak (287.4 eV and 289.2 eV), C–F₂ peak (290.0 eV) and C–F₃ peak (292.0 eV). The absence of sp² suggests the saturated fluorination of graphene. Figure 2e shows the XPS spectrum of graphene after the defluorination process. Interestingly, we have observed that the C1s peak is composed of the sp² signal, and the sp³ intensity is relatively low, similarly to that of the initial graphene sample. In order to confirm the reproducibility of this result, more C1s XPS spectra with corresponding AFM images of defluorinated graphene under the same process have been shown in Figure S1. The XPS results clearly illustrate the effect of the fluorination/defluorination cycle on eliminating the airborne contamination on the graphene surface resulting from the aging process. As several studies have already shown that electron beam irradiation can effectively improve the removal of polymer residues by decomposition [15,16], the mechanism of the airborne contaminants removal during electron beam irradiation is rarely reported. In the work, during the defluorination process, the C–F bonds with an ionization cross section of up to 80 eV can be dissociated with both backscattered and secondary electrons [17,18] to generate fluorine radicals. As shown in Figure S2, the F1s peak has disappeared after experiencing the fluorination/defluorination cycle, indicating that the fluorine radicals have evaporated in the high vacuum of the SEM chamber. We believe the cleaning effect can be attributed to the interactions between fluorine radicals and airborne contaminants on the surface during the electron beam irradiation.

To further investigate the evolution of the surface status of graphene during the fluorination/defluorination cycle, we have carried out an AFM analysis on the graphene surface at different processing steps, shown by Figure 2e (before fluorination), Figure 2f (after fluorination), and Figure 2g (after defluorination). It can be observed that the surface roughness of graphene has increased and the height of the impurities has decreased after fluorination. There is a large number nano-sized fluorinated clusters, which have been reported before [19,20]. Most of these clusters disappear after the defluorination process, as observed in Figure 2g, indicating the restoring of the initial status of graphene. Moreover, the number of polymer residues with a size of 30–100 nm has decreased after fluorination/defluorination, which is in agreement with the previously reported decomposition effect, and may result in a more contamination-free status of graphene.

As the accumulation of hydrocarbon will inevitably happen on the fluorinated graphene, the concentrations of sp³ bonds C_sp3 and C–F bonds C_F with different aging time are shown in Figure 3. We observe that C_sp3 slightly and linearly increases from 60.2 ± 5.1% to 72.6 ± 3.5%, whereas C_F drops accordingly with time increases. This result indicates the stability of fluorographene during aging, and it provides further proof of the hydrocarbon accumulation even on the fluorinated surface during aging.

In order to investigate the effect of the fluorination/defluorination cycle on graphene properties, the Raman spectroscopy and gated measurements have been carried out. Figure 4a shows the Raman spectra of the initial graphene and fluorination/defluorination treated graphene. It can be seen that after defluorination, there is an emerging D peak (located at 1350 cm⁻¹) besides the intrinsic G peak (located at 1580 cm⁻¹) and 2D peak (located at 2690 cm⁻¹), indicating the introduction of disordering after defluorination. Moreover, the broadening of the 2D peak also suggests the structural changes in graphene. This disordering can be mainly attributed to the damage from electron beam irradiation [21,22]. Figure 4b shows the gated measurements of pristine graphene and defluorinated graphene with the source-drain voltage of 50 mV. We have observed a significant decrease of conductivity after defluorination, and the charge carrier mobility drops from 5811 cm² V⁻¹ S⁻¹ to 233 cm² V⁻¹ S⁻¹. The substantial drop in conductivity and mobility has further provided the evidence of the disordering after electron beam irradiation during defluorination. As the transport behavior of graphene can be mainly affected by the grain boundaries [23], the decrease of mobility in this case can be attributed to the grain boundaries’ degradation after defluorination.

In conclusion, we have reported a novel technique for achieving a long-term preservation of supported graphene by using an electron-beam-activated fluorination/defluorination cycle. In this research, the accumulation of hydrocarbon on CVD graphene due to the aging in the air has been observed and characterized by an XPS analysis, with the concentration of the sp² signal dropping from 66.1% to 26.2%. We have demonstrated that by implementing the electron-beam-activated fluorination/defluorination cycle, the hydrocarbon accumulation can be eliminated, which restores the initial chemical status of graphene before fluorination. Moreover, we show that after the defluorination cycle the disordering is introduced due to the electron beam irradiation with an emerging D peak in Raman spectra and the degradation of mobility in gated measurements.