Direct Observation of Monolayer MoS2 Prepared by CVD Using In-Situ Differential Reflectance Spectroscopy

The in-situ observation is of great significance to the study of the growth mechanism and controllability of two-dimensional transition metal dichalcogenides (TMDCs). Here, the differential reflectance spectroscopy (DRS) was performed to monitor the growth of molybdenum disulfide (MoS2) on a SiO2/Si substrate prepared by chemical vapor deposition (CVD). A home-built in-situ DRS setup was applied to monitor the growth of MoS2 in-situ. The formation and evolution of monolayer MoS2 are revealed by differential reflectance (DR) spectra. The morphology, vibration mode, absorption characteristics and thickness of monolayer MoS2 have been confirmed by optical microscopy, Raman spectroscopy, ex-situ DR spectra, and atomic force microscopy (AFM) respectively. The results demonstrated that DRS was a powerful tool for in-situ observations and has great potential for growth mechanism and controllability of TMDCs prepared by CVD. To the best of the authors’ knowledge, it was the first report in which the CVD growth of two-dimensional TMDCs has been investigated in-situ by reflectance spectroscopy.


Introduction
Two-dimensional (2D) materials have attracted enormous attention in optoelectronic devices, flexible sensors, catalysis, and energy conversion due to their outstanding physical and chemical properties [1][2][3][4][5][6][7][8]. In particular, 2D transition metal dichalcogenides (TMDCs) have tunable band gaps, which make their application more likely in semiconductor optoelectronic devices than graphene [4]. The preparation of large-area and high-quality 2D thin film materials is the basis for their extensive development and application. The growth mechanism is not well understood yet, although graphene and MoS 2 have been successfully synthesized at the wafer-level size [9,10]. The efforts about in-situ observation during growth are conducted to explore the growth mechanism and control the quality of 2D materials. However, it is quite challenging to investigate the growth process in-situ because the 2D materials normally require high temperature and a wide range of temperature variations.

Preparation of MoS 2 by CVD
The atmospheric-pressure CVD was applied to prepare MoS 2 samples in a two-zone tube furnace. A silicon substrate with 300 nm thick SiO 2 was selected as the growth substrate. The substrate was ultrasonically cleaned in acetone, alcohol and DI water. Then it was dried by nitrogen gas gun. Finally, the substrate was placed face-up on a crucible boat and loaded into the temperature-zone IIof a tube furnace. 15 mg MoO 3 power (>99.5%, Sigma-Aldrich, St. Louis, MO, USA) was placed upstream on the same crucible boat with the substrate. The distance between MoO 3 and the substrate is approximately 6 cm. After weighing the sulfur powder (>99.5%, Sigma-Aldrich, St. Louis, MO, USA), a second crucible boat containing 1 g of sulfur was located in the temperature-zone Iof the furnace. During the growth process, high-purity argon gas was injected with a flow rate of 40 sccm. It aimed at maintaining an inert atmospheric-pressure environment in a tube furnace and acted as the carrier gas. The tube furnace was heated up to 750 • C during 50 min. The temperature was maintained for 15 min before cooling down to room temperature.

Ex-Situ Characterization Experiment
Optical images were obtained using an optical microscope (GFM-550, Shanghai Guangmi Instrument Co., Ltd., Shanghai, China) to observe the morphology of MoS 2 on the SiO 2 /Si substrate. Raman spectra were performed by a Raman microscope (Renishaw inVia, Gloucestershire, UK) with a laser excitation wavelength of 532 nm at the room temperature. The detection was carried out using a 2400 grooves mm −1 grating. The ex-situ DRS system was built on an optical platform. It consists of a tungsten light source, two converging mirrors and a detector. A spectrometer (QE Pro, Ocean Optics, Dunedin, FL, USA) as the detector was used to collect the reflected light. The ex-situ DR spectra were defined by: [30,31] ∆R where R 0 and R S denote the reflected light intensity of the bare substrate and regions covered with MoS 2 , respectively. The thickness and surface topography of the sample were conducted by AFM (Dimension Icon, Bruker, Santa Barbara, CA, USA) in PeakForce tapping mode. We used an AFM cantilever (ScanAsyst-Air, Bruker, Santa Barbara, CA, USA) with a resonant frequency of approximately 75 kHz and a spring constant of approximately 0.4 N/m. All ex-situ characterization experiments were carried out at room temperature.

In-Situ DRS Experiment
The in-situ experiment setup consists of CVD preparation equipment and the in-situ DRS equipment. This is a universal detector to in-situ detect the growth of 2D materials for the CVD preparation. Figure 1 shows the overall experiment setup. An Ar-gas supplier and a vacuum pump are mounted on left and right side of a two-zone tube furnace respectively, which provides the growth environment for CVD preparation of 2D materials. Optical images were obtained using an optical microscope (GFM-550, Shanghai Guangmi Instrument Co., Ltd., Shanghai, China) to observe the morphology of MoS2 on the SiO2/Si substrate. Raman spectra were performed by a Raman microscope (Renishaw inVia, Gloucestershire, UK) with a laser excitation wavelength of 532 nm at the room temperature. The detection was carried out using a 2400 grooves mm −1 grating. The ex-situ DRS system was built on an optical platform. It consists of a tungsten light source, two converging mirrors and a detector. A spectrometer (QE Pro, Ocean Optics, Dunedin, FL, USA) as the detector was used to collect the reflected light. The ex-situ DR spectra were defined by: [30,31] where R0 and RS denote the reflected light intensity of the bare substrate and regions covered with MoS2, respectively. The thickness and surface topography of the sample were conducted by AFM (Dimension Icon, Bruker, Santa Barbara, CA, USA) in PeakForce tapping mode. We used an AFM cantilever (ScanAsyst-Air, Bruker, Santa Barbara, CA, USA) with a resonant frequency of approximately 75 kHz and a spring constant of approximately 0.4 N/m. All ex-situ characterization experiments were carried out at room temperature.

In-Situ DRS Experiment
The in-situ experiment setup consists of CVD preparation equipment and the in-situ DRS equipment. This is a universal detector to in-situ detect the growth of 2D materials for the CVD preparation. Figure 1 shows the overall experiment setup. An Ar-gas supplier and a vacuum pump are mounted on left and right side of a two-zone tube furnace respectively, which provides the growth environment for CVD preparation of 2D materials. For the in-situ DRS measurements, a home-built setup is implemented above the tube furnace. A light beam is obtained by a broadband light source that is either a xenon lamp or a tungstenhalogen lamp. It enters into the furnace via a quartz window after the light beam is collimated by the lens (L1) and focused by the mirror (M1). The reflected light beam also passes through the quartz window. Following, the mirror (M2) and the lens (L2) gather the reflected light beam, which is transmitted into the fiber. Additionally, the lenses (L1 and L2) and mirrors (M1 and M2) are installed on an optical adjustable bracket to achieve their four-axial (x, y, z, θ) movement. A high-resolution spectrometer (QE Pro, Ocean Optics, Dunedin, FL, USA) connected to the fiber is acted as a spectra collector. The computer averages 1600 successive spectral data at a time to reduce the statistical noise of spectra, which spends around 120 s. We take such a set of cumulative data as one spectrum For the in-situ DRS measurements, a home-built setup is implemented above the tube furnace. A light beam is obtained by a broadband light source that is either a xenon lamp or a tungsten-halogen lamp. It enters into the furnace via a quartz window after the light beam is collimated by the lens (L1) and focused by the mirror (M1). The reflected light beam also passes through the quartz window. Following, the mirror (M2) and the lens (L2) gather the reflected light beam, which is transmitted into the fiber. Additionally, the lenses (L1 and L2) and mirrors (M1 and M2) are installed on an optical adjustable bracket to achieve their four-axial (x, y, z, θ) movement. A high-resolution spectrometer Nanomaterials 2019, 9, 1640 4 of 9 (QE Pro, Ocean Optics, Dunedin, FL, USA) connected to the fiber is acted as a spectra collector. The computer averages 1600 successive spectral data at a time to reduce the statistical noise of spectra, which spends around 120 s. We take such a set of cumulative data as one spectrum acquired during the in-situ experiment. The overall setup achieves an in-situ observation without interrupting CVD preparation, which ensures the growth and characterization of 2D materials at the same time.
High temperature and a wide range of temperature variations are needed during the CVD experiment, which lead to the thermal deformation of mechanical structures and changes in the optical coefficient of the substrate. Thus, the general DRS (Formula (1)) can not satisfy the requirements of measurement accuracy during CVD growth. It is assumed that the variations of light intensity induced by thermal deformation and changes in substrate optical coefficient are repeatable. Here, the new DR spectra suitable for variable temperature environment are calculated by the equation as follow: where R 0 (T i ) and R S (T i ) denote the reflectivity of the bare substrate without MoS 2 growth and the one of the substrate with the MoS 2 growth at the same temperature T i , respectively. Before the 2D materials growth experiment, we have heated the bare substrate in the range of the preparation temperature to obtain the reflectivity of the bare substrate with the same time interval, temperature and airflow.
The new DRS signal ∆R/R thus inhibits the measurement error caused by temperature and reveals the optical properties of the surface on a substrate.

Ex-Situ Characterization
The MoS 2 sample was prepared by atmospheric pressure CVD in a two-zone tube furnace, which was characterized ex-situ by optical microscopy, Raman spectroscopy, DRS and AFM. After cooling down to the room temperature, optical microscopy was performed to observe the distribution and morphology of MoS 2 on the SiO 2 /Si substrate. Figure  We employed Raman spectroscopy to characterize the sample structure at room temperature. Raman spectra of MoS 2 sample are plotted in Figure 2b. The excitation wavelength is 532 nm. There are two characteristic peaks of MoS 2 called E 1 2g and A 1g in the spectra. The peak E 1 2g at 382.1 cm −1 represents the in-plane reverse vibration mode of S and Mo atoms [21]. And another peak A 1g at 401.6 cm -1 is attributed to the out-of-plane vertical vibration mode of S atoms [21]. In particular, the wavenumber difference of two peaks is approximately 19.5 cm −1 , which indicates that the MoS 2 structures of films and independent triangular crystals are monolayer structures [32].
In order to study the optical absorption of MoS 2 and as a reference for in-situ spectra, ex-situ DR spectra were performed at room temperature. As shown in Figure 2c, ex-situ DR spectra were measured on an optical platform. The absorption peaks at 1.84 eV and 2.0 eV correspond to characteristic peaks A and B, respectively. They are induced by exciton transitions at the point K of the Brillouin zone. The energy interval between two exciton peaks is approximately 0.16 eV, owning to the splitting of valence bands at K point as a result of spin-orbit interaction [33].
AFM was selected to confirm the thickness and surface topography. Figure 2d presents the height profile across the edge of the MoS 2 thin film. The inset in Figure 2d shows an AFM image. The height Nanomaterials 2019, 9, 1640 5 of 9 profile was measured along the black solid line in the AFM image. It displays a high gradient of approximately 0.76 nm, which consists of the thickness of monolayer MoS 2 [34]. Therefore, the AFM result also agrees well with the DRS and Raman characterizations. We employed Raman spectroscopy to characterize the sample structure at room temperature. Raman spectra of MoS2 sample are plotted in Figure 2b. The excitation wavelength is 532 nm. There are two characteristic peaks of MoS2 called E 1 2g and A1g in the spectra. The peak E 1 2g at 382.1 cm −1 represents the in-plane reverse vibration mode of S and Mo atoms [21]. And another peak A1g at 401.6 cm -1 is attributed to the out-of-plane vertical vibration mode of S atoms [21]. In particular, the wavenumber difference of two peaks is approximately 19.5 cm −1 , which indicates that the MoS2 structures of films and independent triangular crystals are monolayer structures [32].
In order to study the optical absorption of MoS2 and as a reference for in-situ spectra, ex-situ DR spectra were performed at room temperature. As shown in Figure 2c, ex-situ DR spectra were measured on an optical platform. The absorption peaks at 1.84 eV and 2.0 eV correspond to characteristic peaks A and B, respectively. They are induced by exciton transitions at the point K of the Brillouin zone. The energy interval between two exciton peaks is approximately 0.16 eV, owning to the splitting of valence bands at K point as a result of spin-orbit interaction [33].
AFM was selected to confirm the thickness and surface topography. Figure 2d presents the height profile across the edge of the MoS2 thin film. The inset in Figure 2d shows an AFM image. The height profile was measured along the black solid line in the AFM image. It displays a high gradient of approximately 0.76 nm, which consists of the thickness of monolayer MoS2 [34]. Therefore, the AFM result also agrees well with the DRS and Raman characterizations.

In-Situ DRS
CVD growth process of MoS2 introduced in the experimental section was observed using in-situ DRS. According to the temperature in a two-zone tube furnace, the preparation process of MoS2 can

In-Situ DRS
CVD growth process of MoS 2 introduced in the experimental section was observed using in-situ DRS. According to the temperature in a two-zone tube furnace, the preparation process of MoS 2 can be separated into three sequential stages in Figure 3a. The stage I(the heating-up stage) began at room temperature and stopped at 750 • C of the temperature zone II. Almost no MoS 2 film on the substrate was observed in this stage. For the stage II(the thermostatic stage) at the temperature of 750 • C, MoS 2 thin film began to deposit on the substrate with the increased lateral size. During the stage III(the cooling-down stage), the growth of MoS 2 thin film slowed down and ended gradually. The temperature dropped to room temperature. be separated into three sequential stages in Figure 3a. The stage І (the heating-up stage) began at room temperature and stopped at 750 °C of the temperature zone ІІ. Almost no MoS2 film on the substrate was observed in this stage. For the stage ІІ (the thermostatic stage) at the temperature of 750 °C , MoS2 thin film began to deposit on the substrate with the increased lateral size. During the stage ІІІ (the cooling-down stage), the growth of MoS2 thin film slowed down and ended gradually. The temperature dropped to room temperature. The spectral evolution of reflected light from the substrate during the thermostatic stage is displayed in Figure 3b. The time interval between successive spectra is 120 s. The formation and evolution of the absorption peak at 1.83 eV are observed along the arrow direction. It can be noted that the absorption peak is in accordance with the optical characteristic peak A of monolayer MoS2. This observation implies that the reaction between molybdenum trioxide and sulfur has occurred and MoS2 has begun to grow on a SiO2/Si substrate during the thermostatic stage. In addition, the DR intensity at 1.83 eV of this stage decreases continuously. It could be inferred that DR intensity correlates with the increase of growth area and absorption for MoS2 sample under the constant temperature. Figure 3c exhibits DR spectra of MoS2 on a SiO2/Si substrate during the cooling-down stage in which the development of spectral intensity and shape is displayed. The successive spectra were recorded with the temperature interval of 40 °C . The temperature range is 730-50 °C . Each spectrum we collected is the average of cumulative spectra over a short period of time (around 120 s), which represents the average state of MoS2 during each time interval. Different from the thermostatic stage, the spectral amplitudes of the cooling-down stage rise regularly. This finding indicates that the amplitude is related to variation of temperature. Apparently, there are two negative peaks around 1.83 eV and 1.99 eV appearing during this process. The two peaks are consistent with the characteristic peaks A and B of monolayer MoS2, respectively. This observation indicates that monolayer MoS2 has grown on the SiO2/Si substrate, agreeing with the results of ex-situ characterizations. The width of the absorption peak at approximately 1.83 eV becomes narrower The spectral evolution of reflected light from the substrate during the thermostatic stage is displayed in Figure 3b. The time interval between successive spectra is 120 s. The formation and evolution of the absorption peak at 1.83 eV are observed along the arrow direction. It can be noted that the absorption peak is in accordance with the optical characteristic peak A of monolayer MoS 2 . This observation implies that the reaction between molybdenum trioxide and sulfur has occurred and MoS 2 has begun to grow on a SiO 2 /Si substrate during the thermostatic stage. In addition, the DR intensity at 1.83 eV of this stage decreases continuously. It could be inferred that DR intensity correlates with the increase of growth area and absorption for MoS 2 sample under the constant temperature. Figure 3c exhibits DR spectra of MoS 2 on a SiO 2 /Si substrate during the cooling-down stage in which the development of spectral intensity and shape is displayed. The successive spectra were recorded with the temperature interval of 40 • C. The temperature range is 730-50 • C. Each spectrum we collected is the average of cumulative spectra over a short period of time (around 120 s), which represents the average state of MoS 2 during each time interval. Different from the thermostatic stage, the spectral amplitudes of the cooling-down stage rise regularly. This finding indicates that the amplitude is related to variation of temperature. Apparently, there are two negative peaks around 1.83 eV and 1.99 eV appearing during this process. The two peaks are consistent with the characteristic peaks A and B of monolayer MoS 2 , respectively. This observation indicates that monolayer MoS 2 has grown on the SiO 2 /Si substrate, agreeing with the results of ex-situ characterizations. The width of the absorption peak at approximately 1.83 eV becomes narrower when the temperature decreases. This can be attributed to the widening of the peak induced by high temperature, which is in accordance with the previous reports [35][36][37].
In order to further detect the increment of DR spectra during the cooling-down stage, the ∆(∆R/R) spectrum is shown in Figure 3d. It can be clearly noted that two prominent absorption peaks of MoS 2 are observed by the increment of DR signals. The peak A at 1.83 eV is relatively narrow, which is generated by exciton transition of electron-hole recombination. Another peak B induced by exciton transition at the lower valence band of K point is also obvious at 1.99 eV. The appearance of two peaks is a clear growth signature for monolayer MoS 2 . It should be pointed out that the positions of absorption peaks have red shifts in this stage, which are compared with the ex-situ DR spectra at room temperature. We can attribute it to the high temperature of the substrate. This observation is consistent with the previous report that high temperature could cause the red shifts of peaks [27]. In addition, the DT increments of characteristic peaks A and B in the [27] are very weak (less than 0.1). Both the front surface and the back surface of a substrate were covered by monolayer MoS 2 . However, the DR increment of peak A and B is approximately 0.19 and 0.23 respectively. Monolayer MoS 2 was only deposited on the front surface of a substrate and didn't completely cover the front surface. The signal increases more obvious from reflectance measurements compared with that from transmittance measurements, which indicates that the reflectance measurement has a higher sensitivity for 2D materials.
Furthermore, the DR intensity at 1.83 eV and 1.99 eV are respectively plotted as a function of time in Figure 3e. This is helpful to understand the detailed evolution of the two peaks. We take the initial time of the thermostatic stage as the founding moment. As shown in Figure 3e, the values of ∆R/R drop during the growth of the thermostatic stage and then the values rise obviously in the cooling-down stage. This result is the same as the tendency shown in Figure 3b,c. It displays that the growth of MoS 2 is closely related to temperature.

Conclusions
In summary, the growth of MoS 2 prepared by CVD on a SiO 2 /Si substrate was in-situ characterized by DR spectra. An in-situ DRS setup monitoring CVD growth of 2D materials was established in order to achieve this aim. The process of the growth and cooling down of monolayer MoS 2 was investigated by the in-situ DRS. An obvious optical feature of MoS 2 in the in-situ DR spectra was observed during thermostatic growth. DR spectra of the cooling-down stage showed the evolution of the MoS 2 optical characteristics, indicating the formation of monolayer MoS 2 . Moreover, ex-situ DRS also exhibited the characteristic peaks of monolayer MoS 2 . It agrees well with other ex-situ characterization results such as optical microscopy, Raman spectroscopy and AFM, which confirmed the monolayer structure of MoS 2 . The results emphasize that DRS is effective and has a high sensitivity for in-situ observations of 2D TMDCs during the CVD growth.