The Raman band shift of suspended graphene impacted by the substrate edge and helium ion irradiation

The 20 μm × 20 μm freestanding silicon nitride (SiN_x) membrane supported on 400 μm thick silicon (100) substrate with the thickness of 20 nm was created by conventional optical photolithography followed by anisotropic wet etching of silicon in KOH solution. Then, the holey SiNx with a diameter of 1 μm wells array was fabricated through the gallium ion (Ga⁺) focused ion beam (FIB) technology with an acceleration voltage of 30 kV. The Ga⁺ ion beam current was about 6 pA and the beam does was 0.3 nc μm⁻².

Graphene sample was obtained grown by large area (14 inches) grown method through chemical vapor deposition (CVD) technology on a Cu foil. Then cut it into small pieces (0.5 cm × 0.5 cm) and transferred it to the prepared holey SiN_X substrate by the PMMA assistant wet transfer. The poly(methyl-methacrylate) PMMA layer about 200 nm was spin-coated onto the sample. The Cu foil was etched away with ammonium persulfate ((NH₄)₂S₂O₈). The sample was then transferred into fresh DI water. Then, the sample was taken out of the DI water slowly with the holey SiNx substrate and heated on a hot plate at 150 ℃ for ∼20 min to make better adhesion to the substrate. The graphene/PMMA layers were transferred to the SiNx membrane. Finally, the PMMA layer was dissolved in the acetone and then heated on the hot plate at 150 ℃ for ∼5 min to evaporate residual solvents.

Graphene nanopore fabrication by a focused helium ion beam under the pressure of 1 × 10⁻⁷ Torr and ion beam current of 0.10 pA at 10 μm aperture size based on helium ion microscopy (HIM, Zeiss Orion Nano Fab). Two low beam doses (8 × 10⁴ ions, 1 × 10⁵ ions) and two high doses (6.5 × 10⁵ ions, 7.5 × 10⁵ ions) were used to fabricate graphene nanopore.

Raman spectra of graphene were acquired under ambient conditions with a Renishaw confocal micro Raman microscope equipped with a 532 nm laser. A 100× objective lens and a diffraction grating with 1200 lines mm⁻¹ were used. The spatial resolution of the laser was ∼0.76 μm. A 1 s acquisition time and four spectrum averages were used at each point. Unless otherwise specified, the integration time was 1 s, the number of acquisitions was 1 time, and the laser power on the sample was about 2 mW for each Raman measurement. The most prominent features of the Raman spectrum of graphene are the characteristic G band at ∼1580 cm⁻¹ and the G' band at ∼2700 cm⁻¹. A third band, the D band at ∼1350 nm⁻¹, becomes Raman active in defective graphene.

The setup of the suspended graphene film sample was shown in figure 1(A). At first, 4 through holes with a diameter of 1 μm and a non-through hole were drilled on the SiNX window using a focused ion beam (FIB). Based on PMMA assistant wet transfer method, the monolayer graphene film was transferred onto the patterned SiNx substrate. The FIB and optical image of the prepared SiN_X holes as shown in figures 1(B), (C). Then, the PMMA was removed cleanly from the graphene film for the next experiment, and the continuous suspended graphene film was obtained. Figure 1(D) is the optical image of the SiN_X holes with graphene film. After the PMMA layer was removed, the graphene edges could be clearly identified on the SiNx surface [24]. The insert in figure 1(D) shows that the PMMA was removed cleanly from the graphene film. Next, the suspended graphene nanopores were prepared on the SiNx holes by optimizing the aperture and beam current of the helium ion microscope based on focused helium ion etching technology.

Zoom In Zoom Out Reset image size

To get the transferred graphene film information, Raman measurements were performed for the unirradiated graphene film. Figure 2(A) demonstrated the experimental setup for the suspended and supported film. There are two strong Raman bands for this unirradiated suspended and supported graphene film as shown in figure 2(B). The G band (∼1580 cm⁻¹) is generated from in-plane sp² C–C stretching mode. The G' band, (∼2700 cm⁻¹), which is a second-order Raman process originating from the in-plane breathing-like mode of the carbon rings. Another important band of graphene is the D band (∼1350 cm⁻¹), which is arising from the breathing modes of six-atom rings and requires a defect for its activation [20, 21]. All the Raman spectra show a sharp single Lorentzian profile for G bands, the G' to G peak intensity ratio (I_G'/I_G) over than 1, and the D bands are very low. These results indicated that the presence of the transferred graphene film is a single layer and there is no obvious defect [25]. However, the Raman spectral for suspended and supported graphene film is different. These changes are probably come from the effect of the substrate. Furthermore, the non-through hole was introduced for the further investigation of the substrate effect on suspended graphene. As shown in figure 2(B), the Raman intensity of suspended graphene film are higher than the supported film. Take the G' band, for example, the intensity of graphene on the through-hole (c) shows 3 times compared to the graphene on the non-through hole (b), and 9 times for the supported graphene (a). It also demonstrates that the defect-related D band intensity of monolayer graphene on the non-thorough hole is obviously enhanced compared to those on SiNx and thorough holes. This result may attribute to the large area of the non-through hole. The monolayer graphene film couldn't be suspended very well on such a large hole, and introduce some defects. Besides, figure S1 (available online at stacks.iop.org/NANOX/2/010001/mmedia) shows the intensity of the G band and G' band mapping of the continuous suspended graphene film. The previous studies on the stress effect of the monolayer graphene film show that the G' band intensity can increase with the tensile strain [26].

Zoom In Zoom Out Reset image size

Figures 2(C), (D) shows the Raman mapping results of the G band and the G' band Raman shifts of the continuous graphene film on the SiNx holes array. The yellow dash ranges are through holes, the green dash range is the non-through hole. The zone-center LO phonon around 1580 cm⁻¹ (G band) and the two-phonon peak around 2700 cm⁻¹ (G' band). The phonon frequencies of graphene might be influenced by mechanical strain and charge [27–29]. Raman spectra and the peak fitting of different locations are shown in figures S2(A), (B). These results show that the G band is 1591 cm⁻¹ and G' band is 2683 cm⁻¹ on the SiNx substrate. According to the previous report, the impurities in the substrate and the surrounding environment would lead to Raman band shift. Water and oxygen molecules absorbed on the graphene surface can introduce a p-type dopant, resulting in a blue-shift [29]. For the through-hole, compared with the Raman shift of graphene on a substrate, both G and G' bands of monolayer graphene on hole edge red-shift with the shift of 3 and 9 cm⁻¹, respectively. In the center of the suspended area, the G' band is shift back, and G band shows a blue-shift (3 cm⁻¹) compared with the graphene on the substrate. As for the non-through hole, it is almost the same phenomenon as the through-hole. Compared with graphene on the substrate, both G and G' bands of monolayer graphene on non-through hole edge show red-shift (5 cm⁻¹, 5 cm⁻¹). In the center of the suspended area, both G and G' bands are blue shift with 4 and 7 cm⁻¹, respectively. These results show that the shift of G' band is much larger than that of G band. The redshift of the Raman Band can be attributed to the stretching of the carbon-carbon bond [30]. It is reported that no matter n-type or p-type, the G band position of single-layer graphene shifts to a higher wave position with the weakening of the Kohn anomaly effect. The G band shifts to the low wave when the n-type is mixed [31]. Thus, these phenomena also indicated that the substrate of the substrate edge may introduce n-type mixing for suspended graphene approaching the substrate edge.

The suspended graphene nanopores were prepared on the SiNx through wells by optimizing the aperture and beam current of helium ion microscope based on focused helium ion etching technology. First, the 10 μm through holes were selected, and the helium ion beam current is about 1.5 pA at the pressure of 1 × 10⁻⁶ Torr. Then, according to our early work [12], two low beam doses ( 8 × 10⁴, 1 × 10⁵ ions) and two higher doses (6.5 × 10⁵, 7.5 × 10⁵ ions) were set to study the helium ions influences on suspended graphene film during the fabrication of graphene nanopore. The non-through hole didn't irradiate for the control experiment.

Figure 3(A) shows the Raman spectral of the helium ion treated of suspended graphene film on SiNx through wells. Raman defect-induced D and D' bands in defective monolayer graphene appear at 1350 and 1620 cm⁻¹, respectively. The D band requires a defect for its activation and originating from the breathing modes of six-atom rings and requires defect for its activation.

Zoom In Zoom Out Reset image size

The defect density of graphene can be quantified by measuring the intensity ratio (I_D/I_G) of the D and G band. Figure 3(B), shows the relationship between Raman intensity ratio of I_D/I_G , I_G'/I_G, and irradiation dose. With the increase of helium ion does, the I_G'/I_G of suspended graphene decrease gradually. However, the intensity ratio of I_D/I_G increases with the dose at first, and then down to 1.5 when the irritation dose at 7.5 × 10⁵ ions. It demonstrated that the trend of defect I_D/I_G ratio of graphene monolayer doesn't simply change with the increase of the dose of the helium ion beam. This phenomenon evidences the evolution of disordered graphene from graphite to nanocrystalline graphite [32–34]. Although the FIB imaging (figure S3) of graphene nanopore array changes obviously with the increase of the helium ion dose. Since I_D/I_G intensity ratio does not depend on the geometry of the defects, the intensity ratio between the D and D' can be used to characterize defect types [35]. I_D/I_D' ratios measured in a range of 1.9–2.5, which are smaller than those for point defects [36], and show an increase with an increase of the irritation dose (figure 4(A)). Raman mapping has been done to directly probe the change with the helium doses irritation. Figure 3(C) shows the Optical image of the Raman spectra mapping area. The yellow dash ranges are through holes, the green dash range is the non-through hole. Figures 3(D), (E) shows the Raman mapping results of the G band and G' band Raman shifts of graphene film on the SiNx holes array after the helium doses irradiation. Raman spectra and the peak fitting of different locations are shown in figures S2(C), (D).

Zoom In Zoom Out Reset image size

The G band and G' band peak position of graphene on the substrate are 1596 and 2686 cm⁻¹. After the ion irritation, the Raman showed a blue-shift with a shift of 5 and 3 cm⁻¹, which are compared to the untreated graphene film on the substrate. This phenomenon has also been confirmed by the other location graphene film (on through hole edge, non-through hole edge, suspended on the non-through hole), where are undirectly bombed by helium ions. Compared with the same location untreated graphene, the Raman shifts of all the undirectly bombed graphene are bule-shift. It has been reported that the G band peak position increases as a function of both electron and hole doping. The blue-shifted G' peak position observed in graphene is attributed to unintentional doping in graphene by the charge impurities present on the SiNx substrate [34]. As for the helium ions directly irritated graphene film, take the irritation dose at 7.5 × 10⁵ ions (i) for example. Compared with the Raman shift of graphene on this substrated, all of the D, G and G' bands peak position of the suspended graphene are red-shift with the shift of 5, 6, 10 cm⁻¹, respectively. As shown in figure 4(B), the G band always around 1590 cm⁻¹ under different helium does irrational. The G' mode shifts almost linearly with the increase of irritation dose in a range of 2673–2677 cm⁻¹ (figure 4(C)). The G frequency shift is similar to the doping effects, which is may due to the He⁺ irradiation of graphene on SiNx substrate introduce n-type doping. That is to say, with the He⁺ irradiation dose increase the hole and n-type doping are increased. Besides, the G' band shift range of the helium ions bombed suspended graphene is much smaller than the untreated suspended graphene. These results may due to the removal of water and oxygen molecules adsorbed on graphene during the fabrication of nanopore in suspended graphene.