Application-Oriented Growth of a Molybdenum Disulfide (MoS2) Single Layer by Means of Parametrically Optimized Chemical Vapor Deposition

In the 2D material framework, molybdenum disulfide (MoS2) was originally studied as an archetypical transition metal dichalcogenide (TMD) material. The controlled synthesis of large-area and high-crystalline MoS2 remains a challenge for distinct practical applications from electronics to electrocatalysis. Among the proposed methods, chemical vapor deposition (CVD) is a promising way for synthesizing high-quality MoS2 from isolated domains to a continuous film because of its high flexibility. Herein, we report on a systematic study of the effects of growth pressure, temperature, time, and vertical height between the molybdenum trioxide (MoO3) source and the substrate during the CVD process that influence the morphology, domain size, and uniformity of thickness with controlled parameters over a large scale. The substrate was pretreated with perylene-3,4,9,10-tetracarboxylic acid tetrapotassium salt (PTAS) seed molecule that promoted the layer growth of MoS2. Further, we characterized the as-grown MoS2 morphologies, layer quality, and physical properties by employing scanning electron microscopy (SEM), Raman spectroscopy, and photoluminescence (PL). Our experimental findings demonstrate the effectiveness and versatility of the CVD approach to synthesize MoS2 for various target applications.


Introduction
Recently, transition metal dichalcogenides (TMDs) are gaining tremendous interest for their potential use in wide range of applications [1]. Among TMDs, molybdenum disulfide (MoS 2 ) has so far received the largest consideration for its promising potential in various fields such as electronics, optoelectronics, and electrocatalysis due to its unique optical, electrical, and chemical properties [2,3]. When thinned to a single layer, a strong intra-layer covalent-bonded MoS 2 is obtained with a thickness of 0.7 nm [4], whereas in few-layer MoS 2 , the different layers are stacked together by weak van der Waals interlayer interactions [5]. As a TMD in the 2H structural phase, MoS 2 undergoes a transition from an indirect bandgap (1.3 eV) in the bulk to a direct gap in the single layer (1.8 eV) [6]. Monolayer MoS 2 has outstanding properties, such as wide direct bandgap, intense light-matter interactions, strong spin-orbit coupling, and high carrier mobility [2]. In electronics and optoelectronics, such characteristics make monolayer MoS 2 particularly attractive for building field effect transistors, photodetectors, chemical OH-terminated and, hence, highly hydrophilic. Perylene-3,4,9,10-tetracarboxylic acid tetrapotassium salt (PTAS) seeding promoter crystals (0.48 mg) were dissolved in 10 mL water. Using a micropipette, a small volume of PTAS solution (~10 µL) was spread on the clean substrate. Finally, the substrate was placed on a hot plate at 110 • C for 1 min to evaporate the water content from the solution on the surface.

Growth
Molybdenum disulfide (MoS 2 ) was grown on silicon substrates via CVD, from sulfur (S) (99.98%, Sigma-Aldrich, Darmstadt, Germany) and molybdenum trioxide (MoO 3 ) (99.97%, Sigma-Aldrich, Darmstadt, Germany) powders. The amounts of S and MoO 3 powders were 0.2 g and 0.001 g, respectively, as weighed by a microbalance. Subsequently, the measured quantities of S and MoO 3 powders were placed in the center of ceramic or quartz crucibles. Our CVD system (Planartech UK Limited, Cambridge, UK) consisted of two separate thermal zones capable of providing precise control of the temperature for each precursor. In the CVD apparatus, the ceramic crucible containing the S precursor was placed in the center of the upstream furnace, while in the downstream heating zone the Si/SiO 2 substrate was positioned face down on top of the quartz crucible containing the MoO 3 powder. The MoS 2 growth zone was located within the downstream heating zone, where the substrate was located. The distance between the two crucibles was kept fixed at 24 cm, as shown in Figure 1a. The temperature profile of the MoO 3 (black line) and S (red line) precursors followed during the growth process is shown in Figure 1b, where the different process steps as a function of time are also presented. Figure 1b also shows the Ar flux (blue line) rate during the process, varying from 30 sccm to 1000 sccm in the different steps during the CVD process. The entire growth process developed under ambient pressure conditions. Materials 2020, 13, x FOR PEER REVIEW  3 of 13 hydroxylation, making it OH-terminated and, hence, highly hydrophilic. Perylene-3,4,9,10tetracarboxylic acid tetrapotassium salt (PTAS) seeding promoter crystals (0.48 mg) were dissolved in 10 mL water. Using a micropipette, a small volume of PTAS solution (~10 µL) was spread on the clean substrate. Finally, the substrate was placed on a hot plate at 110 °C for 1 min to evaporate the water content from the solution on the surface.

Growth
Molybdenum disulfide (MoS2) was grown on silicon substrates via CVD, from sulfur (S) (99.98%, Sigma-Aldrich, Darmstadt, Germany) and molybdenum trioxide (MoO3) (99.97%, Sigma-Aldrich, Darmstadt, Germany) powders. The amounts of S and MoO3 powders were 0.2 g and 0.001 g, respectively, as weighed by a microbalance. Subsequently, the measured quantities of S and MoO3 powders were placed in the center of ceramic or quartz crucibles. Our CVD system (Planartech UK Limited, Cambridge, UK) consisted of two separate thermal zones capable of providing precise control of the temperature for each precursor. In the CVD apparatus, the ceramic crucible containing the S precursor was placed in the center of the upstream furnace, while in the downstream heating zone the Si/SiO2 substrate was positioned face down on top of the quartz crucible containing the MoO3 powder. The MoS2 growth zone was located within the downstream heating zone, where the substrate was located. The distance between the two crucibles was kept fixed at 24 cm, as shown in Figure 1a. The temperature profile of the MoO3 (black line) and S (red line) precursors followed during the growth process is shown in Figure 1b, where the different process steps as a function of time are also presented. Figure 1b also shows the Ar flux (blue line) rate during the process, varying from 30 sccm to 1000 sccm in the different steps during the CVD process. The entire growth process developed under ambient pressure conditions.

Characterization
The as-grown MoS2 samples were characterized using Raman spectroscopy, scanning electron microscopy (SEM), photoluminescence (PL), atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS). Confocal Raman spectroscopy was performed using a Renishaw In-Via spectrometer (New Mills, Kingswood, Wotton-under-Edge, UK) equipped with a solid-state laser source of excitation wavelength 514 nm/2.41 eV in backscattering configuration. Particular care was put in the laser power, reduced down to 5% of the nominal power (i.e., below 1 mW range) to avoid sample damage. The morphology of the MoS2 layers was characterized using a Zeiss-SUPRA 40 field-emission SEM device (Oberkochen, Germany) in bright field mode. Complementary PL was acquired with the same Raman equipment by changing the instrument configuration; AFM topography was performed in tapping mode in a commercial Bruker Edge system (Billerica, MA,

Characterization
The as-grown MoS 2 samples were characterized using Raman spectroscopy, scanning electron microscopy (SEM), photoluminescence (PL), atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS). Confocal Raman spectroscopy was performed using a Renishaw In-Via spectrometer (New Mills, Kingswood, Wotton-under-Edge, UK) equipped with a solid-state laser source of excitation wavelength 514 nm/2.41 eV in backscattering configuration. Particular care was put in the laser power, reduced down to 5% of the nominal power (i.e., below 1 mW range) to avoid sample damage. The morphology of the MoS 2 layers was characterized using a Zeiss-SUPRA 40 field-emission SEM device (Oberkochen, Germany) in bright field mode. Complementary PL was acquired with the same Raman equipment by changing the instrument configuration; AFM topography was performed in tapping mode in a commercial Bruker Edge system (Billerica, MA, USA) using Si tips; and XPS was performed using a monochromated Al X-ray source (1486.6 eV) and a hemispherical analyzer in a PHI ESCA 5600 instrument (Chanhassen, MN, USA).

Results and Discussion
During the CVD synthesis, the growth of TMDs is governed by several factors, such as pressure, growth temperature, growing time, and substrate position. These factors are crucial for synthesizing high-quality and large-area monolayer MoS 2 nanosheets. Hereafter, the role of each of these parameters on the lateral growth and morphology of MoS 2 nanosheets is discussed in detail. Furthermore, the best growth conditions for obtaining large-area MoS 2 nanosheets to isolated triangular domains are explained.

Pressure
Keeping the growth temperature at 750 • C, we explored the effect of MoS 2 growth at different pressures, namely 400 torr (5 × 10 4 Pa), 600 torr (8 × 10 4 Pa), and 760 torr (1 × 10 5 Pa), varying the Ar flux rate from 30 sccm to 1000 sccm in multiple steps during the CVD process. At a process pressure of 400 torr, the growth results in the formation of grains that are clearly visible as dark regions in SEM images, as shown in Figure 2a. This effect is possibly due to the concentration gradient of the precursors over the substrate surface. Zheng et al. reported that such an effect occurs due to the large evaporation of MoO 3 by lowering pressure [18]. As we see from the SEM image in Figure 2a and the Raman scattering spectra shown in Figure 2e, although the lateral coverage and the 2-layer thickness are good, the existence of randomly distributed MoS 2 domains over the surface and wide grain boundaries still prevents us from setting the conditions for the desired polycrystalline MoS 2 growth regime. Increasing the pressure up to 600 torr, the bulk growth is inhibited and the formation of mixed oxides Mo x S y O z from the MoO 3 and S precursors occurs, as revealed from the SEM analysis and Raman spectrum, as shown in Figure 2b,d. Surprisingly, 600 torr pressure remains a sharp regime for oxide deposition with bright square and wire-like structures as further confirmed from Raman characterization. At atmospheric pressure (760 torr), we observe the formation of well-defined triangular MoS 2 domains with sharp edges, as shown in Figure 2c. Under such a condition, the amount of S in the gas phase is enough to promote a uniform and S-rich atmosphere in the reactor region where the substrate is placed so as to favor the chemical reaction between Mo and S. The presence of a S-rich atmosphere is confirmed by visual inspection of the quartz tube after the completion of the growth process, where residual solid S is found on the quartz tube walls. The number of layers at 600 torr was characterized by Raman spectroscopy, where a frequency difference between the Raman modes E 1 2g (381) and A 1g (405) of 24 cm −1 allowed us to identify 4-layer thick MoS 2 along with the presence of other peaks that are assigned to MoO x phases, as shown in Figure 2d. Conversely, at 400 torr and 760 torr, the MoS 2 particles and triangular domains, when characterized by Raman spectroscopy, revealed the main modes E 1 2g (382 and 383) and A 1g (402 and 404) with a frequency difference of 20 cm −1 and 21 cm −1 (blue and orange line, respectively) corresponding to a thickness of 1 MoS 2 layer (0.7 nm) and 2 layers, as shown in Figure 2e [19]. This condition is well described by a mass transport-limited regime where the growth is controlled by the amounts of precursors that are delivered, a condition that is typical of atmospheric pressure CVD [20]. In this regime, the control of the gas flow and the mean free path of reactants become particularly relevant [21]. Indeed, we confirmed that the atmospheric pressure CVD leads to a higher nucleation rate and a larger nucleation density compared to processes at lower pressures, thereby promoting the growth of a uniform monolayer MoS 2 with a thickness of 0.7 nm at 760 torr. However, to enlarge the MoS 2 growth over the substrate area and to ultimately obtain an extended continuous monolayer, other parameters apart from pressure control should be considered, as reported in the following sections.

Temperature
Taking the ambient pressure (760 torr) as a fixed parameter, the MoS2 growth was investigated by changing the temperature from 650 °C to 800 °C within a 50 °C gradient range in step 5 of the CVD process in Figure 1b. Below 650 °C, MoS2 particles with no distinct morphology were observed, that are indicative of a poor growth environment. As the temperature was increased to 650 °C, the growth of isolated triangular domains was formed with truncated triangle-shaped crystals, as shown in Figure 3a. By raising the temperature to 700 °C, we observed that a high density of MoS2 domains increased with an exceptional coverage. Although there are many MoS2 domains with sharp edges, they appear as overlapped with adjacent bulk domains leading to a characteristic roughness regime, as shown in Figure 3b. Furthermore, when we raised the growth temperature to 750 °C, the increase in the lateral coverage and morphology with a uniform thickness of monolayer was clearly observed from Raman and SEM measurements (compared to Figure 3a,b,d). At this temperature, MoS2 domains with a large area coalesce so as to form a continuous monolayer, as shown in Figure 3c. Furthermore, increasing the temperature to 800 °C resulted in the formation of 3D islands on a continuous monolayer and small bilayer isolated triangular domains on the same MoS2 layer, as shown in Figure 3d. Similar to the parametric variation of pressure, the substrate temperature proves to have a strong impact on the thickness and morphology of grown MoS2. However, while pressure variations dictate the transition from a granular regime with a vertical island growth mode to a 2D layer regime, the variation of temperature at atmospheric pressure determines the condition of coalescence and the layer overgrowth on top of the interconnected triangular domain texture. In

Temperature
Taking the ambient pressure (760 torr) as a fixed parameter, the MoS 2 growth was investigated by changing the temperature from 650 • C to 800 • C within a 50 • C gradient range in step 5 of the CVD process in Figure 1b. Below 650 • C, MoS 2 particles with no distinct morphology were observed, that are indicative of a poor growth environment. As the temperature was increased to 650 • C, the growth of isolated triangular domains was formed with truncated triangle-shaped crystals, as shown in Figure 3a. By raising the temperature to 700 • C, we observed that a high density of MoS 2 domains increased with an exceptional coverage. Although there are many MoS 2 domains with sharp edges, they appear as overlapped with adjacent bulk domains leading to a characteristic roughness regime, as shown in Figure 3b. Furthermore, when we raised the growth temperature to 750 • C, the increase in the lateral coverage and morphology with a uniform thickness of monolayer was clearly observed from Raman and SEM measurements (compared to Figure 3a,b,d). At this temperature, MoS 2 domains with a large area coalesce so as to form a continuous monolayer, as shown in Figure 3c. Furthermore, increasing the temperature to 800 • C resulted in the formation of 3D islands on a continuous monolayer and small bilayer isolated triangular domains on the same MoS 2 layer, as shown in Figure 3d. Similar to the parametric variation of pressure, the substrate temperature proves to have a strong impact on the thickness and morphology of grown MoS 2 . However, while pressure variations dictate the transition from a granular regime with a vertical island growth mode to a 2D layer regime, the variation of temperature at atmospheric pressure determines the condition of coalescence and the layer overgrowth on top of the interconnected triangular domain texture. In particular, our study proves that MoS 2 coverage increases with the increasing growth temperature for a given value of pressure. In our case, we found that 750 • C was a favorable temperature to obtain a monolayer MoS 2 growth regime with a high degree of coverage over a mm 2 area. Further Raman measurements confirm the number of layers at different temperatures, as shown in Figure 3e. The frequency difference between the two Raman modes of A 1g and E 1 2g identifies the layer number [19]. For the MoS 2 films synthesized at 650 • C, 700 • C, 750 • C, and 800 • C, the number of MoS 2 layers was revealed by measuring the Raman spectrum at different positions that are indicated by circular spots with different colors in Figure 3. The frequency differences between the Raman modes of the E 1 2g and A 1g phonon modes were~20 cm −1 (black), 22 cm −1 (red), 24 cm −1 (blue), and 26 cm −1 (pink) that corresponds to 1, 2, 4, and 6 layers, respectively. Therefore, by considering the three black circular spots from the Raman analysis at 750 • C, a uniform monolayer MoS 2 growth is confirmed, as shown in Figure 3c. Conversely, in the other cases, there is a thickness variation of 1 to 6 layers that is confirmed from the Raman spectrum with different color spots at different positions, as shown in Figure 3e. In addition, the increase in the frequency difference between the A 1g and E 1 2g phonon modes in the Raman spectra corresponding to the increase in the number of layers is indicated in Figure 3f.
Materials 2020, 13, x FOR PEER REVIEW 6 of 13 particular, our study proves that MoS2 coverage increases with the increasing growth temperature for a given value of pressure. In our case, we found that 750 °C was a favorable temperature to obtain a monolayer MoS2 growth regime with a high degree of coverage over a mm 2 area. Further Raman measurements confirm the number of layers at different temperatures, as shown in Figure 3e. The frequency difference between the two Raman modes of A1g and E 1 2g identifies the layer number [19]. For the MoS2 films synthesized at 650 °C, 700 °C, 750 °C, and 800 °C, the number of MoS2 layers was revealed by measuring the Raman spectrum at different positions that are indicated by circular spots with different colors in Figure 3. The frequency differences between the Raman modes of the E 1 2g and A1g phonon modes were ~20 cm −1 (black), 22 cm −1 (red), 24 cm −1 (blue), and 26 cm −1 (pink) that corresponds to 1, 2, 4, and 6 layers, respectively. Therefore, by considering the three black circular spots from the Raman analysis at 750 °C, a uniform monolayer MoS2 growth is confirmed, as shown in Figure 3c. Conversely, in the other cases, there is a thickness variation of 1 to 6 layers that is confirmed from the Raman spectrum with different color spots at different positions, as shown in Figure 3e. In addition, the increase in the frequency difference between the A1g and E 1 2g phonon modes in the Raman spectra corresponding to the increase in the number of layers is indicated in Figure 3f.

Time (Thermal Budget)
We extended the study of MoS2 growth by exploring the role of the growth time at 760 torr (ambient pressure) and 750 °C. We explored the following growth times: 10 min, 20 min, 30 min, and 100 min at the given values of pressure and temperature. Referring to Figure 1b, by growth time we

Time (Thermal Budget)
We extended the study of MoS 2 growth by exploring the role of the growth time at 760 torr (ambient pressure) and 750 • C. We explored the following growth times: 10 min, 20 min, 30 min, and 100 min at the given values of pressure and temperature. Referring to Figure 1b, by growth time we mean the duration of step 5 in the CVD process. Under the given condition, the growth time identifies the process thermal budget, namely the total amount of thermal energy transferred to the substrate during the process. The experimental results are shown in Figure 4a-e, which includes SEM images and Raman spectra. From the SEM images, it is evident that MoS 2 morphology after growth varies from isolated large domains to a continuous layer growth by increasing the growth time. For instance, in the case of 10 min growth time, we strikingly found isolated monolayer domains as large as 150 µm in edge length and with sharp edges, as shown in Figure 4a. The uniform growth of the MoS 2 layer is confirmed by measuring the Raman spectrum at different positions along the triangular domain area, here indicated by black circular spots. By considering the frequency difference between the E 1 2g and A 1g phonon modes, a value of~20.4 cm −1 is found indicating the presence of a monolayer, as shown in Figure 4e (black line) [22]. A slight shift (increase of~0.4 cm −1 ) in the A 1g position in the Raman mode of single-layered MoS 2 compared to the previously mentioned value of 20 cm −1 is due to the fact that the A 1g mode is strongly sensitive to electron doping. This is because the A 1g mode has a stronger electron-phonon coupling than the E 1 2g mode [23]. Raising the growth time to 20 min, a tremendous increase in the lateral coverage with a continuous MoS 2 monolayer is visible, as shown in Figure 4b. The growth time of 20 min is favorable for obtaining a uniform growth with large coverage up to an area of a few mm 2 . The above-mentioned conditions allowed us to get the best parametric conditions for uniform thickness and large-scale coverage MoS 2 layer growth. However, if we extend the growth time to 30 min and above, additional layers start to grow from the center of the monolayer MoS 2 domain, as is clearly shown in Figure 4c. Such layer generation at the center of the monolayer is also confirmed by the Raman spectrum in Figure 4e, where the frequency difference of 21.3 cm −1 (red line, 2 layers) and 22.7 cm −1 (blue line, 3 layers) is clearly enlarged with respect to the value of the monolayer, 20.4 cm −1 (black line, 1 layer). Considering a very long growth time (100 min), the MoS 2 assumes the 3D island-like shape, typical of a bulk growth, as shown in Figure 4d. Such bulk growth is confirmed by the frequency difference of the Raman modes, which is~27.5 cm −1 (Figure 4e, dark blue line) corresponding to a MoS 2 film consisting of many layers, i.e., with bulk characteristics. Figure 4f shows a linearly increasing trend in the number of layers at an increasing value of the frequency difference between the A 1g and E 1 2g phonon modes in the Raman spectra.
Materials 2020, 13, x FOR PEER REVIEW 7 of 13 mean the duration of step 5 in the CVD process. Under the given condition, the growth time identifies the process thermal budget, namely the total amount of thermal energy transferred to the substrate during the process. The experimental results are shown in Figure 4a-e, which includes SEM images and Raman spectra. From the SEM images, it is evident that MoS2 morphology after growth varies from isolated large domains to a continuous layer growth by increasing the growth time. For instance, in the case of 10 min growth time, we strikingly found isolated monolayer domains as large as 150 µm in edge length and with sharp edges, as shown in Figure 4a. The uniform growth of the MoS2 layer is confirmed by measuring the Raman spectrum at different positions along the triangular domain area, here indicated by black circular spots. By considering the frequency difference between the E 1 2g and A1g phonon modes, a value of ~20.4 cm −1 is found indicating the presence of a monolayer, as shown in Figure 4e (black line) [22]. A slight shift (increase of ~0.4 cm −1 ) in the A1g position in the Raman mode of single-layered MoS2 compared to the previously mentioned value of 20 cm −1 is due to the fact that the A1g mode is strongly sensitive to electron doping. This is because the A1g mode has a stronger electron-phonon coupling than the E 1 2g mode [23]. Raising the growth time to 20 min, a tremendous increase in the lateral coverage with a continuous MoS2 monolayer is visible, as shown in Figure 4b. The growth time of 20 min is favorable for obtaining a uniform growth with large coverage up to an area of a few mm 2 . The above-mentioned conditions allowed us to get the best parametric conditions for uniform thickness and large-scale coverage MoS2 layer growth. However, if we extend the growth time to 30 min and above, additional layers start to grow from the center of the monolayer MoS2 domain, as is clearly shown in Figure 4c. Such layer generation at the center of the monolayer is also confirmed by the Raman spectrum in Figure 4e, where the frequency difference of 21.3 cm −1 (red line, 2 layers) and 22.7 cm −1 (blue line, 3 layers) is clearly enlarged with respect to the value of the monolayer, 20.4 cm −1 (black line, 1 layer). Considering a very long growth time (100 min), the MoS2 assumes the 3D island-like shape, typical of a bulk growth, as shown in Figure 4d. Such bulk growth is confirmed by the frequency difference of the Raman modes, which is ~27.5 cm −1 (Figure 4e, dark blue line) corresponding to a MoS2 film consisting of many layers, i.e., with bulk characteristics. Figure 4f shows a linearly increasing trend in the number of layers at an increasing value of the frequency difference between the A1g and E 1 2g phonon modes in the Raman spectra.

MoO 3 Source to Substrate Relative Distance
Having systematically studied the effects of growth pressure, temperature, and time on the characteristics of the obtained MoS 2 nanosheets, the source-to-substrate position during growth also plays an important role in achieving controllable thickness and lateral coverage.
Here, we explored the growth employed under identical growth conditions in order to compare the growth pattern, number of layers, and flake size as a function of the relative vertical distance between the position of the MoO 3 precursor and the substrate as illustrated in Figure 5a-c for the different conditions explored (set-1, -2, and -3). For this purpose, we used three quartz boats with different vertical heights, indicated by h, of 2, 5, and 10 mm. In set-1 (h = 2 mm), the deposited MoS 2 nanosheets are very limited at the sample edge close to the MoO 3 precursor, propagating up to 1 cm on the substrate surface, as shown by the red dashed region in Figure 5a. Such a MoS 2 grown region consists of very thick and almost bulk MoS 2 film, as observed by Raman spectra revealing a peak difference between the E 1 2g and A 1g peaks of 27.1 cm −1 , typical of bulk MoS 2 . Therefore, such a set condition is not effective to obtain 2D-layered growth of MoS 2 . In the case of set-2 (h = 5 mm), two types of MoS 2 growth behavior can be readily discriminated, namely a plume-shaped one corresponding to a bulk deposition (red region) and a few layers all around extending up to the substrate edges where MoS 2 reduces to a single layer, as indicated by the yellow dashed lines in Figure 5b. Raman investigation of the as-grown samples reveals a difference between the position of the E 1 2g and A 1g peaks from 27.1 cm −1 down to 23.6 cm −1 , typical of bulk to three layers of MoS 2 . In addition, in a few regions, a difference of~20 cm −1 is detected, evidence of the presence of monolayer MoS 2 . It is worth noting that the region outside the red zone and within the delimiting yellow lines is the (large) region of the substrate where MoS 2 is forming as domains of a few layers down to the monolayer. In the case of the set-3 configuration (h = 10 mm), we achieved mostly monolayer and bilayer MoS 2 growth at the edges of the substrate, as indicated by the yellow dashed line. In addition, MoS 2 domains with 3 or 4 layers are obtained at the center of the substrate at some minimal positions. In this configuration, the growth of 2D MoS 2 is homogeneous and continuous compared to the previous cases with a reduced vertical gap. Raman spectroscopy investigation reveals a peak difference between the MoS 2 main phonon modes (A 1g , E 1 2g ) in the range of 22 cm −1 -19.7 cm −1 , evidence of bilayer or monolayer MoS 2 . In this configuration, it is possible to obtain large-area crystalline and monolayer MoS 2 domains. In addition, the as-grown MoS 2 morphology and Raman analysis of the A 1g and E 1 2g peak positions versus the number of MoS 2 layers are shown in Figure 6a-d. Based on SEM and Raman data, MoS 2 growth under set-1 conditions results in a bulk-like layer (7 or 8 layers). Similarly, in set-2 and set-3, the as-grown MoS 2 films tend to multilayers (3-6 layers) and mono/bilayers (1 or 2 layers) as shown in Figure 6a-c. Figure 6d shows the peak position of the A 1g and E 1 2g peak Raman modes as a function of the number of MoS 2 layers. As the number of MoS 2 layers increases, the frequency separation between the two modes tends to increase, as is clearly shown in Figure 6d. Therefore, we conclude that a 10 mm vertical gap is the best favorable condition to obtain uniformly continuous and large-area growth of single-layer and bilayer MoS 2 [24].

CVD Growth of MoS 2 Towards Target Applications
Because of the high requirement to produce and expand MoS 2 from research to production level for its integration into various applications and to tackle this challenge, the CVD method shows a reliable way for producing large-scale, high-quality, thickness-controlled MoS 2 nanosheets according to the target application. By tuning various factors like temperature, pressure, time, and vertical distance between the MoO 3 source and the substrate, it is relatively easy to optimize the MoS 2 growth. Herein, we demonstrated the easily tunable conditions that determine the high quality, thickness control, shape, and size from isolated MoS 2 domains to continuous nanosheets suitable for integration into electronics, electrocatalysis, photonics (see the SEM and TEM images in Figures S1 and S2, Supplementary  Materials, respectively). In general, monolayer MoS 2 mainly addresses electronics and optoelectronic applications, because of its tunable bandgap with a semiconducting character or, with the limit of the monolayer, due to its direct bandgap transition, whereas 3D bulk to few-layer MoS 2 growth is an essential platform for HER in electrocatalysis and supercapacitors, where surface-to-volume exposed active sites are a key parameter [25,26]. Since MoS 2 is the most consistent potential candidate for many applications, it is essential to facilitate choosing among the growth conditions using fast and more cost-effective methods. Our characterization of MoS 2 growth provides the parametric conditions that help to obtain the MoS 2 morphologies that are specifically compliant with the two last requirements.
Two different examples in this respect are discussed below for clarity.

Electronics and Optoelectronics
There has been much interest in exploiting monolayer MoS 2 nanosheets in field-effect transistors (FETs) and other potential devices. Monolayer MoS 2 FETs offer some excellent electrical performance qualities in terms of good mobility, ON/OFF ratio, and high mechanical flexibility [1,12,27]. However, it is still difficult to obtain large-scale uniform MoS 2 monolayers in order to fabricate more than a single FET device in the same integration process. For this scope, the bottom-up synthesis of atomically thin MoS 2 using the CVD method was extensively demonstrated here. By selecting the process parameters for growth temperature at 750 • C, growth duration of 20 min, with a vertical height between the MoO 3 source and substrate of h = 10 mm, and under an ambient pressure condition, it is possible to achieve the growth of highly polycrystalline monolayer MoS 2 nanosheets on SiO 2 /Si substrate. Under such conditions, in the mass transport regime, the 3D-like growth mode of MoS 2 growth is inhibited, possibly due to the minimization of the nucleation sites on the substrate; finally, MoS 2 growth is extended and continuous, from monolayer up to the mesoscale. The vertical height is essential to allow the MoS 2 film homogenous growth over the substrate surface, thus controlling the final thickness of the MoS 2 layer. This easy method provides a high-quality MoS 2 monolayer with lateral dimensions of a few millimeters with high growth yield, thus opening the opportunity to fabricate a greater number of FETs on a large area consisting of monolayer MoS 2 .
For optoelectronic applications, photoluminescence (PL) spectroscopy investigation was done to verify the optical properties of the monolayer MoS 2 nanosheet. The PL spectrum in Figure 7a indicates a PL response peaked around the optical bandgap of 1.875 eV on monolayer MoS 2 , which confirms the direct gap transition [28]. At room temperature, the pristine monolayer MoS 2 shows a high-quality strong PL peak associated with the band-to-band optical transition at the K point. The direct optical transitions are expected to take place between the conduction band minimum and the two-valence band maxima at the K point of the Brillouin zone, which splits due to the spin-orbit coupling [29]. The obtained experimental bandgap values of strong emission for a monolayer MoS 2 associated with A exciton at 1.875 eV is in agreement with the reported value [15]. The above-mentioned PL characterization confirms that the as-grown monolayer MoS 2 nanosheets acquire an excellent optical response that is favorable in various optoelectronic applications. Furthermore, the topography monolayer MoS 2 with a height profile of 0.7 nm thickness is confirmed by AFM measurement, as presented in Figure S3, Supplementary Materials. Therefore, the above-mentioned monolayer MoS 2 nanosheets with different advantages help to push forward the applications of these TMD materials in the wide range of the optoelectronics field.

Electrocatalysis (HER)
Highly dense MoS2 domains with active metal edge sites are evidence of catalytic activity. However, considerable efforts are still needed to grow MoS2 with a metallic 1T polymorph, which displays a more amenable charge conductivity throughout the layer [10]. Conversely, the mixed phase of 2H-1T MoS2 domains was also suggested as a satisfactory candidate for HER performance compared to 1T-MoS2 and 2H-MoS2 in terms of favorable free energy ΔGH* [30]. To target HER utilizing the mixed phase of 2H-1T MoS2, we established a simple strategy to tune the 2H-1T MoS2 growth on SiO2/Si substrate. 2H-MoS2 with the presence of minor 1T observed from XPS characterization is shown in Figure S4, Supplementary Materials. The favorable CVD process parameters for MoS2 growth with a suitable morphology for electrocatalysis are a growth duration of 10 min at 700 °C, 760 torr, and a vertical height between the MoO3 source and substrate of h = 5 mm. Such conditions promoted highly polycrystalline MoS2 domains with sharp active edges. Growth temperature and time favor the formation of isolated MoS2 domains over a large area. In addition, the SEM morphological image of highly dense isolated as-grown MoS2 domains with exposed sharp edges is shown in Figure 7b. This above-mentioned MoS2 growth is much desired for accelerating the electrocatalysis process.

Conclusions
Through a systematic study of the growth pressure, temperature, time, and relative distance between MoO3 precursor and substrate, we were able to define a set of standardizing growth conditions to obtain MoS2 with a certain morphology and domain size, adapted for targeted applications. Based on the optimized experimental conditions, we successfully prepared large isolated triangular MoS2 domains and extended MoS2 growth size from several micrometers to a few millimeters. Our work confirms that the CVD approach is an easy and low-cost strategy for synthesizing MoS2 with improved crystalline quality, large area coverage, and uniform thickness down to the monolayer. This study also shows the ability to parametrically tune the growth mode from a uniform and large-scale coverage with well-defined PL signature to a laterally rough and isolated pattern of MoS2 domains in order to target optoelectronics and catalysis directions.

Electrocatalysis (HER)
Highly dense MoS 2 domains with active metal edge sites are evidence of catalytic activity. However, considerable efforts are still needed to grow MoS 2 with a metallic 1T polymorph, which displays a more amenable charge conductivity throughout the layer [10]. Conversely, the mixed phase of 2H-1T MoS 2 domains was also suggested as a satisfactory candidate for HER performance compared to 1T-MoS 2 and 2H-MoS 2 in terms of favorable free energy ∆G H* [30]. To target HER utilizing the mixed phase of 2H-1T MoS 2 , we established a simple strategy to tune the 2H-1T MoS 2 growth on SiO 2 /Si substrate. 2H-MoS 2 with the presence of minor 1T observed from XPS characterization is shown in Figure S4, Supplementary Materials. The favorable CVD process parameters for MoS 2 growth with a suitable morphology for electrocatalysis are a growth duration of 10 min at 700 • C, 760 torr, and a vertical height between the MoO 3 source and substrate of h = 5 mm. Such conditions promoted highly polycrystalline MoS 2 domains with sharp active edges. Growth temperature and time favor the formation of isolated MoS 2 domains over a large area. In addition, the SEM morphological image of highly dense isolated as-grown MoS 2 domains with exposed sharp edges is shown in Figure 7b. This above-mentioned MoS 2 growth is much desired for accelerating the electrocatalysis process.

Conclusions
Through a systematic study of the growth pressure, temperature, time, and relative distance between MoO 3 precursor and substrate, we were able to define a set of standardizing growth conditions to obtain MoS 2 with a certain morphology and domain size, adapted for targeted applications. Based on the optimized experimental conditions, we successfully prepared large isolated triangular MoS 2 domains and extended MoS 2 growth size from several micrometers to a few millimeters. Our work confirms that the CVD approach is an easy and low-cost strategy for synthesizing MoS 2 with improved crystalline quality, large area coverage, and uniform thickness down to the monolayer. This study also shows the ability to parametrically tune the growth mode from a uniform and large-scale coverage with well-defined PL signature to a laterally rough and isolated pattern of MoS 2 domains in order to target optoelectronics and catalysis directions.

Conflicts of Interest:
The authors declare no conflict of interest.