Residual Oxygen Effects on the Properties of MoS₂ Thin Films Deposited at Different Temperatures by Magnetron Sputtering

Abstract

Molybdenum disulfide (MoS₂) thin films were deposited at different temperatures (150 °C, 225 °C, 300 °C, 375 °C, and 450 °C) on quartz glass substrates and silicon substrates using the RF magnetron sputtering method. The influence of deposition temperature on the structural, optical, electrical properties and deposition rate of the obtained thin films was investigated by X-ray diffraction (XRD), Energy Dispersive Spectrometer (EDS), Raman, absorption and transmission spectroscopies, a resistivity-measuring instrument with the four-probe method, and a step profiler. It was found that the MoS₂ thin films deposited at the temperatures of 150 °C, 225 °C, and 300 °C were of polycrystalline with a (101) preferred orientation. With increasing deposition temperatures from 150 °C to 300 °C, the crystallization quality of the MoS₂ thin films was improved, the Raman vibrational modes were strengthened, the deposition rate decreased, and the optical transmission and bandgap increased. When the deposition temperature increased to above 375 °C, the molecular atoms were partially combined with oxygen atoms to form MoO₃ thin film, which caused significant changes in the structural, optical, and electrical properties of the obtained thin films. Therefore, it was necessary to control the deposition temperature and reduce the contamination of oxygen atoms throughout the magnetron sputtering process.

1. Introduction

Molybdenum disulfide (MoS₂) belongs to the transition metal dichalcogenides (TMDs). Structurally, MoS₂ is a stack of planes where covalently bonded S-Mo-S atoms in layers are closely packed in a hexagonal arrangement by weak van der Waals forces. Because of its tunable energy band structure (1.8 eV for monolayer to 1.29 eV for bulk) [1] and unique photoelectric properties, MoS₂ thin film has shown potential for preparation in the field of optoelectronic devices (e.g., detectors [2] and supercapacitors [3]) in recent years. As the basis for the manufacture of molybdenum disulfide-based devices, the preparation of their high-quality films is particularly important. Among all the methods for synthesizing molybdenum disulfide film, magnetron sputtering has become the mainstream synthesis method due to its advantages of simple process flow [4], fast deposition speed [5], and low damage to the film layer [6]. While research on magnetron sputtering of large-scale prepared molybdenum disulfide films has been gradually completed in recent years [7,8], studies on the influence of process parameters on film properties are not yet complete.

Previously, the effects of sputtering power [9,10], sputtering pressure [11], and incident flux angles [12] on the synthesis of molybdenum disulfide films by magnetron sputtering have been reported, and the influence of sputtering temperature on the frictional properties of molybdenum disulfide films have also been discussed [13,14]. The effect of RF power variation on film thickness as well as resistivity was investigated by Park [9] et al. As the RF power increases, the film thickness increases from 100 nm to 240 nm, and its resistivity increases from 356 to 2006 Ω/sq. Zhong et al. [10] found that the change in RF power has a significant effect on the binding energy of Mo atoms. Gong et al. [11] suggested that the huge discrepancy in electronegativity between MoS₂ molecules promotes film defects at different deposition pressures, which in turn causes changes in the bandgap. They also pointed out that the optimal pressure for sputtering is 1.0 Pa. Li et al. [12] continued to investigate the effect of incident angular flux on molybdenum disulfide films deposited by magnetron sputtering and found that the structure of MoS₂ films changed significantly with the variation of incident angular flux. Spalvins [15] deposited MoS₂ films on aluminum and nickel substrates and investigated the different morphologies of the films at 150 °C and 320 °C, respectively. Until now, investigations into the effect of deposition temperature on molybdenum disulfide films have mainly focused on the changes of film morphology, while the structural and optoelectronic properties of the films have been less studied, as well as the deposition temperature settings not being very high (mostly below 350 °C). Since the ratio of S to Mo in MoS₂ films synthesized by magnetron sputtering is far from 2, Chen et al. [16] suggested that the increase in substrate temperature decreases the stoichiometric ratio of Mo to S. The sublimation of S at high temperatures leads to the presence of a large number of vacancy defects within the film [17]. Changes in deposition temperature affect the binding of defects to impurity elements in the films [18,19,20,21] (e.g., oxygen), and the formation of impurities in the films can change their original properties. To improve the process preparation of molybdenum disulfide to expand its application in the device field, it is essential to study the effect of high temperature on the optoelectronic and structural properties of molybdenum disulfide prepared by magnetron sputtering.

In this work, MoS₂ thin films were synthesized at different sputtering temperatures. The effects of sputtering temperature on the structure, optical and electrical properties, and deposition rate of the obtained films were investigated to provide a reference for the preparation of high-quality films with excellent properties.

2. Materials and Methods

2.1. Preparation of Molybdenum Disulfide Film

In view of the application of MoS₂-based devices, their application is more useful for MoS₂ thin films deposited on the silicon substrate. However, silicon substrates are opaque and non-insulating, thus, it would be difficult to investigate the optical and electrical properties of MoS₂ thin films in our experiments. In this paper, the quartz glass substrate was simultaneously used in our experiment to investigate the optical and electrical properties of MoS₂ thin films. Before sputtering, the quartz glass substrates and silicon substrates were ultrasonically cleaned in acetone, alcohol, and deionized water and dried in nitrogen gas. MoS₂ films were grown on the substrates by the radio frequency (RF) sputtering method. To investigate the effect of deposition temperature on the film growth, the control temperatures were 150 °C, 225 °C, 300 °C, 375 °C, and 450 °C, respectively. MoS₂ ceramic target (99.99% purity) and argon (99.9% purity) were used as the sputtering target and sputtering gas, respectively. The argon flow rate was maintained at 10 sccm by a mass flow meter. The system was pumped to a base pressure of 3 × 10⁻⁴ Pa before sputtering. During sputtering, the pressure was maintained at 1.0 Pa, the distance between substrate and target was controlled at 7.5 cm, while the RF power was 80 W. Sputtering time was 30 min. After the sputtering process, the heater and argon gas pipeline were turned off, then the sample was naturally cooled to room temperature and removed for use. The principle of sputtering deposition was shown in Figure 1.

2.2. Characterization of Properties

The properties such as crystal orientation of MoS₂ thin films were measured by X-ray diffractometer with a ray wavelength of 0.15418 nm. The Raman spectra of the samples were measured by a Raman spectrophotometer with a laser wavelength of 532 nm. The absorption and transmission spectra of the films were measured between 400 nm to 800 nm by a double-beam UV-vis spectrophotometer. The thicknesses of different samples were measured with a step meter. The resistivity of the films was measured with a four-probe resistance tester.

3. Results and Discussion

3.1. Structure of the Films

Figure 2 shows the X-ray diffraction results of the obtained thin films grown at different temperatures (150 °C, 225 °C, 300 °C, 375 °C, and 450 °C). From this figure, it can be seen that there was a notable diffraction peak that appeared at 33.5° for all the samples grown at different temperatures. According to the Standard Powder Diffraction File (JCPDS:37-1492), this peak was (101) diffraction peak for 2H-MoS₂ material, proving that the obtained MoS₂ thin film was of polycrystalline with (101) preferred orientation. It can also be seen from this figure that the samples grown at 375 °C and 450 °C had a diffraction peak of high intensity at 12.7°. This peak was a (020) diffraction peak from MoO₃ material according to the Standard Powder Diffraction File (JCPDS:05-0508). To verify our speculation, the elemental composition of molybdenum disulfide films grown on silicon substrates was analyzed using an energy dispersive spectrometer (EDS), and the results are shown in Figure 3. A clear peak of silicon elements can be observed from the figure, which was due to our choice of MoS₂ films grown on silicon substrates as the sample for testing. Based on the EDS data, the stoichiometric ratio of Mo and S in the prepared MoS₂ films was calculated to be less than 1:2, and it was found that the Mo and S stoichiometric ratio decreased gradually with the increase in temperature, which was in line with the previous conclusions of Chen et al. [16]. It was also found that when the deposition temperature was not higher than 300 °C, the percentage of O elements in the films was about 4%, which was in a relatively small variation range. When the deposition temperature was higher than 300 °C, the percentage of O elements in the films increased by more than 1%. Combined with the XRD result, we concluded that when the deposition temperature was higher than 300 °C, oxygen atoms replace S vacancies in the MoS₂ films and combined with Mo atoms to form Mo-O chemical bonds, which in turn led to the generation of MoO₃ impurities. The formation of Mo-O chemical bonds allowed more Mo atoms to be immobilized inside the lattice during deposition, while the evaporation of S is not mitigated so that the stoichiometric ratio of Mo and S decreased significantly in the samples at 375 °C versus 450 °C compared to those at temperatures below 300 °C. Comparing the results with other experiments (the XRD results of the sample by Nikpay et al. [22] and Duan et al. [23]), it was found that molybdenum oxide impurities did not appear in the films when the substrate temperature was below 300 °C, confirming the reliability of our conclusions.

The average grain size (including all peaks) of the MoS₂ could be deduced from the Scherrer equation [24]:

$$D=\frac{K\gamma }{\beta \cos \theta }$$

(1)

where D represents the average grain size; $K$ is the Scherrer constant (0.89);$\gamma$ means the wavelength of X-rays (0.15406 nm); $\beta$ denotes the full width at half maxima (FWHM) of the sample diffraction peak; θ indicates the Bragg diffraction angle. The deposition rate of films with different sputtering temperatures could be calculated from the thickness of the film and the deposition time (30 min); the thickness of the obtained samples was measured using a step profiler, and the results are given in

Table 1. Thickness of MoS₂ thin films at different sputtering temperatures.

Temperature (°C)	150	225	300	375	450
Thickness (nm)	305	233	202	172	158

. The average grain size and film deposition rate of MoS₂ films deposited at different temperatures are shown in Figure 4. There was a 5% to 8% deviation in the calculated average grain size, but this did not affect the trend of the result variation. It was shown that the deposition rate of thin films decreased with the increase in temperature. The grain size became gradually larger with increasing temperature when the deposition temperature was below 300 °C, but this trend changed when the deposition temperature was higher than 300 °C. From the results, it was found that the generation of Mo oxides during high-temperature deposition changed the trend of grain size variation with deposition temperature. Cui et al. and Zhong et al. [25,26] found that MoO₃ diffraction peaks appeared in the MoS₂ samples through the calcination method when the calcination temperature exceeded 300 °C, and the generation of MoO₃ had a great influence on the absorption spectrum of MoS₂ nanoplates. Liu et al. [27] investigated the evolutions of morphology and electronic properties of the MoS₂ layer exposed to ultraviolet ozone. It was found that inhomogeneous oxidation in MoS₂ thin films can significantly change the optoelectronic properties of MoS₂ thin films. It is believed that the oxidation of MoS₂ thin films deposited by magnetron sputtering in our experiment would also affect the other properties of the films, which is discussed in the following section.

As a popular and non-destructive technique to investigate the microstructure of thin films, Raman spectroscopy with a 532 nm laser line was used to further characterize the MoS₂ films deposited at different temperatures, and the results are shown in Figure 5. From this figure, it can be seen that there were two prominent peaks centered at around 375 cm⁻¹ and 407 cm⁻¹, which correspond to the ${\mathrm{E}}_{2\mathrm{g}}^1$ and A_1g vibrational modes of the Raman spectrum of MoS₂ films, respectively. In the ${\mathrm{E}}_{2\mathrm{g}}^1$ mode, Mo and S atoms oscillated in an anti-phase parallel to the crystal plane; therefore, ${\mathrm{E}}_{2\mathrm{g}}^1$ involved the vibration of Mo and S atoms in the basal plane. In the A_1g mode, S atoms oscillated in anti-phase perpendicular to the surface plane, and Mo atoms remained fixed as shown in Figure 5a [28,29,30]. The wavenumber difference Δk between the peaks corresponds to the film thickness of the MoS₂ film. Samples with Δk values close to 18–20 cm⁻¹ correspond to monolayer MoS₂, and the value increased with the thickness of the film [31,32]. The average value of Δk for our five experimental samples was 34.38 cm⁻¹ (150 °C: 32.6 cm⁻¹; 225 °C: 33.5 cm⁻¹; 300 °C: 34.8 cm⁻¹; 375 °C: 35.3 cm⁻¹; 450 °C: 35.7 cm⁻¹), with an error of 0.38 cm⁻¹ from the standard bulk molybdenum disulfide Δk (34 cm⁻¹), which corresponds to a thickness close to the standard of bulk MoS₂, indicating that the prepared film samples are thicker in layers. By comparing the five curves in Figure 5a, it was found that the intensities of both the ${\mathrm{E}}_{2\mathrm{g}}^1$ and A_1g Raman peaks increased with the increase in the deposition temperature. Analyzing the XRD and Raman spectra together proved that higher deposition temperature (150–300 °C) was beneficial to improve the crystallization quality of the MoS₂ films. In Figure 5b, in addition to the two peaks of ${\mathrm{E}}_{2\mathrm{g}}^1$ and A_1g, a tiny peak near 1580 cm⁻¹ was found for the samples deposited at 375 °C and 450 °C, respectively. Combining previous literature [33,34,35] and the results of XRD and EDS (Figure 2 and Figure 3), the appearance of this weak peak is considered to be related to the formation of Mo-O bonds in the films.

3.2. Optical Properties of the Films

Figure 6 shows the absorption and transmission spectra of the MoS₂ films for visible light at different deposition temperatures. From these plots, it was observed that when the deposition temperature was increased from 150 °C to 300 °C, the absorption decreased and the transmission increased. A regular trend of variation in the absorption and transmission curves was observed, which was due to the increase in temperature reducing the deposition rate of the film, thus decreasing the thickness of the film. When the deposition temperature was increased above 300 °C, the original variation pattern was broken, and the curves for the 375 °C and 450 °C samples crossed the previous curves. In order to investigate the changes in the transmittance spectrum for the samples deposited at higher temperatures (375 °C and 450 °C), in Figure 6b, an additional experiment was conducted. In this experiment, a MoS₂ thin film with a thickness of 175 nm was deposited at the temperature of 450 °C by prolonging the deposition time. The thickness of 175 nm was close to that of the MoS₂ sample (172 nm), deposited at 375 °C. The transmittance spectra for the three samples (172 nm deposited at 375 °C, 175 nm deposited at 450 °C, and 158 nm deposited at 450 °C, respectively) are shown in Figure 6c. It can be seen that the transmission for the sample deposition at 450 °C is higher than that deposited at 375 °C with almost the same thickness. Thus, it can be concluded that the changes in the transmittance spectrum for the samples deposited at higher temperatures (375 °C and 450 °C) were not only caused by the thickness of the MoS₂ thin film, but also by the further formation of Mo-O bonds. As analyzed earlier, the formation of a large number of S vacancies and the presence of MoO₃ in the film at high temperatures could dramatically change the optical properties of the film sample.

The optical band gap (E_g) of the samples can be calculated from the absorption spectra using the Tauc formula [36,37]:

$${(\alpha h\nu )}^{1/2}=A\left(h\nu -{\mathrm{E}}_{\mathrm{g}}\right)$$

(2)

$${(\alpha h\nu )}^2=A\left(h\nu -{\mathrm{E}}_{\mathrm{g}}\right)$$

(3)

where $h\nu$ represents the incident photon energy, and $\alpha$ denotes the absorption coefficient. The exponent of ($\alpha$hv) is taken as 1/2 when the calculated material is an indirect bandgap; when the calculated material is a direct bandgap, the exponent of αhν is taken as 2. To determine the type of bandgap of the MoS₂ films, two sets of curves of the samples were made separately by different calculation methods. The optical band gaps of MoS₂ films calculated by different methods are given in Figure 7a,b, respectively. Molybdenum disulfide is known to be a transition metal sulfide with a tunable bandgap (1.8 eV for the single layer with a direct bandgap; 1.29 eV for bulk with an indirect bandgap [38,39]), and the bandgap decreases with increasing thickness. The results obtained from the calculations shown in Figure 7b were not under the bandgap variation range of molybdenum disulfide, and the MoS₂ bandgap values with 305 nm thickness were close to the standard 1.8 eV for a single layer, which was unacceptable. In contrast, the results are shown in Figure 7a were perfectly in line with the variation range of the bandgap of molybdenum disulfide, and the bandgap gradually increased with decreasing thickness, so the prepared MoS₂ films had an indirect bandgap. The variation of the optical band gap of films with different deposition temperatures is shown in Figure 8. With the increasing deposition temperature from 150 °C to 300 °C, the optical band gap increased from 1.350 eV to 1.378 eV gradually; this was due to the decrease in the thickness of MoS₂ thin films [1]. When the deposition temperature was raised to 375 °C, the optical band gap increased to 1.433 eV rapidly. At the temperature of 375 °C, the MoS₂ thin film had partially changed into MoO₃ thin film; the bandgap of MoO₃ (3.1 eV) [40] was larger than that of MoS₂, thus causing the increase in the bandgap of the obtained sample.

3.3. Electrical Properties of the Films

The resistivity of the obtained thin film was measured by a resistivity-measuring instrument with the four-probe method, and the results are shown in Figure 9. It was found the resistivity increased smoothly as the deposition temperature increased from 150 °C to 300 °C. When deposition temperature was raised to 375 °C, the resistivity was increased by one order. Further increased deposition temperatures resulted in a decrease in resistivity of about 20%. Combined with the results shown in Figure 3, it can be seen that there were a large number of S vacancy defects in the film. Qi et al. [41] pointed out that oxygen impurities play a crucial role in the transition from n-type to p-type conductivity of MoS₂ films. It is well known that the pure MoS₂ film with electrons as the main carriers is an n-type semiconductor. After combining the EDS results for analysis, we believe that when the sample deposition temperature was higher than 300 °C, the MoO₃ impurities produced by O atoms intruding into the film to replace S vacancies changed the type of conductivity of MoS₂ films, which is the reason for the abrupt change in the curve across orders of magnitude at 375 °C. This evidences that the formation of MoO₃ in thin films greatly affects the electrical properties of thin films. As can be seen in Figure 9, the resistivity did not show a linear change after the abrupt change at 375 °C, but produced a decrease at 450 °C. This phenomenon is related to the rate of oxygen atom intrusion into the films at high temperatures and the number of S vacancies formed. The electrical conductivity of the films is related to the carrier concentration, electron mobility, and Hall effect of MoS₂, the specific reasons of which we will continue to analyze in future experimental tests.

4. Conclusions

MoS₂ thin films were grown successfully on quartz glass and silicon substrates by magnetron sputtering. It was found that the deposition temperature greatly influenced the structural, optical, and electrical properties of the obtained thin films. Below 300 °C, the obtained MoS₂ thin films were polycrystalline with (101) preferred orientation and showed ${\mathrm{E}}_{2\mathrm{g}}^1$ and A_1g vibrational modes in the Raman spectrum. With the deposition temperature increasing from 150 °C to 300 °C, the crystal quality improved, the optical band gap and resistivity increased, and the thickness decreased. When the deposition temperature was over 375 °C, the molecular atoms were combined with the residual oxygen atoms to form MoO₃. This oxidization process influenced the purity of the MoS₂ thin films, thus causing great changes in the structural, optical, and electrical properties of the obtained thin films.