In the preparation of two-dimensional (2D) materials, large-area fabrication has been extensively investigated to replace the con- ventional scotch tape method. Among 2D materials with excellent mechanical and electronic properties, transition metal dichalcogenides (TMDs), whose band gap is in the ultraviolet–visible–near infrared light range [1,2], have been highlighted as next-generation optoelectronic devices. In particular, as molybdenum disulfide (MoS₂) is a naturally abundant material and its physical properties have been studied for lubrication [3], MoS₂ has the potential to be readily applied to optoelectronic devices.

The parameters of the large-area fabrication of MoS₂ films have been investigated in detail to optimize chemical vapor deposition [4,5]. Furthermore, MoS₂ films synthesized by physical vapor deposition (PVD) have been reported [6–13]; however, fabrication parameters have not been explored in detail. As the fabrication parameters of PVD significantly affect the structure [14], atomic composition [6,15], residual stress [16], and optical properties [17] of synthesized thin films, it is important to determine the fabrication parameters that can provide large-area MoS₂ films with high uniformity and crystallinity. In addition, the thickness of MoS₂ films must be precisely controlled because their band gap is influenced by the number of atomic layers [18].

In this work, we demonstrated the effect of sputtering parameters on MoS₂ films synthesized by radio frequency (RF) magnetron sput- tering and electron-beam irradiation (EBI). MoS₂ films were deposited at different RF powers, working distances, and working pressures. Then, EBI treatment was applied on as-deposited MoS₂ films to induce atomic rearrangement. The EBI parameters for inducing atomic rearrangement were derived from our previous reports [19]. The structural properties of synthesized MoS₂ films were characterized through Raman spectroscopy and atomic force microscopy (AFM). Raman spectroscopy confirmed the change in crystallinity and the number of atomic layers according to sputtering parameters. After EBI treatment, the surface of MoS₂ films remained flat and smooth. Finally, transmission electron microscopy (TEM) showed six to eight atomic layers of MoS₂ arranged in a layered structure. As EBI treatment provides low-temperature and large-area fabrication and is compatible with sputtering, we believe that our results extend the overall understanding of the sputtering process and the preparation of MoS₂ films.

First, MoS₂ thin films were prepared on SiO₂ (100 nm)/Si wafers by utilizing an RF magnetron sputtering system (Infovion Inc., Seoul, Korea), as shown in Fig. 1. A MoS₂ target with a purity of 99.99 % and a diameter of 5.08 cm was used. An RF power of 5 60 W, a working distance of 11–18 cm, a working pressure of 1–10 mTorr, and a deposition time of 2 min–2 h were used as sputtering parameters. The ault values for R power, working distance, working pressure, and deposition time were 20 W, 13.5 cm, 5 mTorr, and 5 min, respectively. After the deposition of MoS₂, EBI was applied in the same chamber of the RF magnetron sputtering system. A DC power of 1 kV, an RF power of 300 W, a working pressure of 0.8 mTorr, a working distance of 7.5 cm, and a process time of 1 min were used for EBI treatment. There was no additional heating during sputtering and EBI treatment.

Figure 1. Schematic of the radio frequency (RF) magnetron sputtering system with electron-beam-irradiation (EBI) source.

Raman spectroscopy (alpha300 S, WITec GmbH, Ulm, Germany) was performed to investigate the properties of MoS₂ films synthesized under different sputtering and EBI parameters. A laser wavelength of 532 nm was used for excitation, and a grating with 1800 grooves/mm was utilized. AFM (alpha300 S, WITec GmbH, Ulm, Germany) was applied over an area of 5 × 5 μm² to characterize the surface of synthesized MoS₂ films. The accelerating voltage for scanning electron microscopy (SEM; JSM-6700F, Jeol, Akishima, Tokyo, Japan) was 10 kV. The samples for TEM (JEM-2100F, Jeol, Akishima, Tokyo, Japan) were prepared via focused ion beam milling. The accelerating voltage for TEM was 200 kV.

Raman spectroscopy is a versatile method for analyzing the crystallinity [19,20], strain [21], vacancy [22], and number of layers [13] of TMDs. These characteristics can be distinguished through various changes such as the blue shift, red shift, transition, and increase and decrease in the Raman intensity. The crystal structure of MoS₂ after EBI treatment was examined by analyzing the two typical Raman peaks of MoS₂, i.e., E¹ (in-plane) and A_1g (out-of-plane) [23]. Based on preliminary information, the main sputtering parameters were adjusted to evaluate the change in the properties of the film. We primarily used Raman spectroscopy because it is nondestructive and requires simple sample preparation and analysis.

Figure 2 shows the Raman spectra of as-deposited MoS₂ films synthesized at different RF powers and deposition times. As RF power increased, the two typical Raman peaks of MoS₂ appeared more clearly. At an RF power of 5 W, the Raman peaks were not observed even for a deposition time of 1 h. However, at an RF power of 60 W, the E¹ and A_1g peaks were strongly observed at ~380 and 405 cm^-1 [24], respectively, for a deposition time of 2 min. The number of atoms sputtered from the target increase with RF power; this increases the deposition rate. In addition, as sputtered atoms reach the substrate with higher kinetic energy, they have sufficient energy to form crystalline structures [14]. In contrast, the sputtered atoms with lower kinetic energy form amorphous structures. We considered working distance, working pressure, and RF power because these parameters are related to the kinetic energy of sputtered atoms.

Figure 2. Raman spectra of as-deposited molybdenum disulfide (MoS2) films sputtered at different RF powers and deposition times.

Figure 3 shows the Raman spectra of as-deposited and 1 kV EBI-treated MoS₂ films sputtered at working distances of 11.0, 13.5, and 18.0 cm. As shown in Fig. 1, as-deposited MoS₂ films exhibited extremely weak Raman peaks, indicating an amorphous state [20]. However, after EBI treatment, the intensity of Raman peaks increased remarkably, which indicated the atomic rearrangement induced by EBI [8]. For EBI-treated MoS₂ films, the differences between the positions of the E¹ and A_1g (Δk) peaks were 27.1, 25.9, and 25.9 cm^-1 for working distances of 11.0, 13.5, and 18.0 cm, respectively. The value of Δk for a bulk MoS film with six or more atomic layers is 24– 27 cm^-1 with Δk [7,13,20]. The intensity of the Raman peaks slightly decreased with working distance, which indicated a decrease in the amount of Mo–S bonding [12]. The possibility of scattering increases with working distance; this decreases the deposition rate.

Figure 3. Raman spectra of as-deposited and 1 kV EBI-treated MoS2 films sputtered at working distances of (a) 11.0, (b) 13.5, and (c) 18.0 cm.

Figure 4 shows the Raman spectra of as-deposited and 1 kV EBI-treated MoS₂ films sputtered at working pressures of 1, 5, and 10 mTorr. Distinguishable Raman peaks were observed for as-deposited MoS₂ films at a working pressure of 1 mTorr, whereas no Raman peaks were observed at working pressures of 5 and 10 mTorr. A lower working pressure results in a longer mean free path and less collisions and enables sputtered atoms to retain more kinetic energy when they reach the substrate. Therefore, the MoS₂ film sputtered at 1 mTorr showed distinguishable Raman peaks before EBI treatment. As the crystallinity of MoS₂ films varies with working pressure during sputtering, kinetic energy should be considered in the design of MoS₂ films for manufacturing amorphous materials. In particular, as the ultrabroadband photoresponse of amorphous MoS₂ has been reported in recent years [2], working pressure can be considered as an important factor in MoS₂ synthesis. The values of Δk for EBI-treated MoS₂ films sputtered at working pressures of 1, 5, and 10 mTorr were 28.3, 25.9, and 24.6 cm^-1 (gray solid lines in Fig. 4), respectively. Additionally, the intensity of the Raman peaks decreased as working pressure increased. As explained in Fig. 2, lower Δk and intensity of Raman peaks indicate less number of layers and amount of Mo–S bonding [12,13]. Even though a higher working pressure reduces the mean free path of sputtered atoms but increases the sputtering rate, the number of layers for the MoS₂ film sputtered at a working pressure of 10 mTorr appeared to be less than that for the MoS₂ films sputtered at working pressures of 1 and 5 mTorr.

Figure 4. Raman spectra of as-deposited and 1 kV EBI-treated MoS2 films sputtered at working pressures of 1, 5, and 10 mTorr.

The surface morphology of as-deposited and 1 kV EBI-treated MoS₂ films sputtered at different working pressures is shown in Fig. 5. AFM was carried out over an area of 5 × 5 μm²; all scale bars in Fig. 5 represent 2 μm. The average roughness (Ra) of the SiO₂/Si wafer was 0.520 nm. The surface of the as-deposited MoS₂ films was extremely flat and smooth, similar to the wafer. Ra was 0.515, 0.526, and 0.532 nm for working pressures of 1, 5, and 10 mTorr, respectively. After EBI treatment, Ra increased to 0.619, 0.646, and 0.671 nm for working pressures of 1, 5, and 10 mTorr, respectively. As the atomic rearrange- ment induced by EBI treatment changed the microstructure of MoS₂ films, the roughness of the surface increased slightly; however, it was still flat and smooth. The increase in roughness caused by EBI treat- ment is less compared to other post treatments [20,24]. The roughness of each layer is a critical factor in sputtered films [25]. The properties of sputtered films can be expected to become more ideal as layers become flatter.

Figure 5. Surface morphology of (a)–(c) as-deposited and (d)–(f) 1 kV EBI-treated MoS₂ films. Working pressures are (a), (d) 1, (b), (e) 5, and (c), (f) 10 mTorr. All scale bars represent 2 μm.

SEM and TEM were carried out to confirm the surface morphology and layered structure of synthesized MoS₂ films. Figures 6(a)–(c) show the SEM images of the surfaces of the bare SiO₂/Si wafer, as-deposited MoS₂ film, and EBI-treated MoS₂ film. As the Ra of the SiO₂/Si wafer and the MoS₂ films calculated via AFM were similar, the surface morphology observed in SEM images was uniform and smooth. The layered structure of the EBI-treated MoS₂ film was observed using cross-sectional TEM. As shown in Fig. 5(d), the MoS₂ interlayer distance was 0.64 nm. In addition, numerous changes, such as cut edges close to the interface according to the boundary and nonuniformity of the substrate, indicated the close relationship between the substrate structure and deposited layer. Six to eight atomic layers of MoS₂ films formed the layered structure; this was consistent with the thickness derived using Δk.

Figure 6. Scanning electron microscopy (SEM) images of the surface of (a) bare SiO₂/Si wafer, (b) as-deposited MoS₂ film, and (c) EBI-treated MoS₂ film. (d) Cross-sectional transmission electron microscopy (TEM) image of EBI-treated MoS₂ film. Scale bars in SEM and TEM images represent 100 and 5 nm, respectively.

We examined MoS₂ films synthesized through RF magnetron sputtering and EBI treatment. The effects of sputtering parameters, such as RF power, working distance, and working pressure, on the deposition rate and structural properties of synthesized MoS₂ films were investigated by Raman spectroscopy and AFM. The crystal structure of as-deposited MoS₂ films strongly depended on RF power, and the thickness of EBI-treated MoS₂ films was affected by working distance and pressure. The crystallinity of as-deposited MoS₂ films increased with RF power. The number of atomic layers of EBI-treated MoS₂ films decreased as working distance and pressure increased. Additionally, the atomic rearrangement induced by EBI treatment was confirmed by changes in Raman spectra. AFM and SEM revealed that the surface of EBI-treated MoS₂ films was flat and smooth after atomic rearrangement. Finally, the layered structure of EBI-treated MoS₂ films was observed using cross-sectional TEM.

This study was supported by the Basic Science Research Program of the National Research Foundation (NRF) of Korea funded by the Ministry of Science and ICT (NRF-2017R1D1A1B03032923) and the Business Support R&D Program funded by the Korea Institute of Ceramic Engineering and Technology (KCS2003-1).