Abstract

Two-dimensional (2D) molybdenum disulfide (MoS2) films with a tunable bandgap hold great promise for next-generation electronic and optoelectronic devices. Synthesis of large areas of high-quality MoS2 monolayers lacks experimental reproducibility. Moreover, the outcome of MoS2 growth by chemical vapor deposition is dependent on several interconnected growth parameters. In this study, we present results of MoS2 monolayer growth by strategically placing water droplets on the growth substrate and/or in the source prior to its loading in the growth chamber. The volume and distribution of water on the growth substrate and in the source had a direct impact on the morphology of the as-grown MoS2. Characterized by scanning electron microscopy (SEM), Raman microscopy, and atomic force microscopy (AFM), the number and size of MoS2 layers as well as its distribution on the growth substrate were found to have a strong dependence on the positioning of the water droplet. This study on MoS2 monolayer growth using water droplets as a promoter provides a simple and reproducible experimental technique enabling growth with high reliability.

1. Introduction

Since the celebrated discovery of graphene, intense research efforts have been directed towards developing two-dimensional (2D) materials for electronic device applications [1, 2]. Graphene of one atom thickness in the vertical direction possesses excellent properties such as high strength and elasticity as well as high electron mobility [2, 3]. Due to the absence of a bandgap in the electronic structure of graphene and challenges involved in creating a bandgap in graphene, new 2D materials were explored, such as transition metal dichalcogenides (TMDs) [4]. Field-effect transistors made with 2D TMDs are expected to exhibit high performance with reduced feature sizes. In this respect, monolayers of molybdenum disulfide (MoS2) are an attractive option for device applications since they have a direct bandgap of 1.8 eV and vertical thickness of 0.8 nm [57]. To attain this goal, it is critical to synthesize the monolayer of MoS2 with large areal coverage. It is experimentally challenging to obtain large areal coverage of MoS2 in a reproducible manner and using a cost-efficient technique [8].

Monolayers of MoS2 have been synthesized by various techniques including physical vapor deposition [9], chemical vapor deposition [10], electrochemical process [11], liquid phase exfoliation [12, 13], and hydrothermal synthesis [14]. Of the growth techniques, sulfurization of molybdenum oxides by chemical vapor deposition (CVD) or powder vapor transport (PVT) has been largely successful in growing large-area (cm scale) high-quality MoS2 thin films on selected substrates [1517]. The CVD process offers controllable large-scale synthesis of MoS2 on chosen substrates (Si, quartz, mica, sapphire, etc.), thus alleviating shortcomings of manual transfer of exfoliated MoS2 flakes using scotch tape. In recent years, research on CVD-grown MoS2 monolayer and multilayers has focused on controlling growth conditions to improve film quality and domain size by precisely pretreating substrate surface, using seeding promoters or plasma, and intentionally changing substrate orientation or position relative to mass flux of precursor [18, 19]. Despite some successes in achieving controlled growth, there are challenges in terms of reproducibility, areal coverage, film quality, and results that are suggestive of interdependent experimental parameters.

In an attempt to address these challenges, this work investigates the effect of using water vapor as a promoter in CVD-grown MoS2 layers. Using MoO3 and S as source materials, Si substrates with water droplets on their surface were used to study the impact of water on the growth of MoS2 monolayers. It has been reported that a continuous supply of water vapor into the growth region promotes the growth of high-quality MoS2 layers [20]. The technique proposed in this work explores a novel way of introducing controlled amounts of water vapor into the growth region to facilitate large-area growth of MoS2 monolayers. In this communication, we followed the PVT growth technique, which uses MoO3 powder and S powder as source materials and argon as a carrier gas. Water droplets were strategically placed on the surface of cleaned SiO2/Si substrates, prior to their loading in the furnace. Since water evaporates from the substrate surface as the temperature in the growth chamber rises, the condition most likely resembles the introduction of water vapor into the growth chamber during the CVD process. A series of well-controlled water-assisted experiments on growth showed that it was possible to control the areal coverage and numbers of layers of MoS2 films. To explain these experimental results, this work presents a possible mechanism for monolayer growth based on chemical reactions between the various constituents involved in the growth process. The significance of this work is twofold. (i) It highlights the fact that though initially there was a very limited volume of water in the growth chamber, the reversible chemical reaction between H2O and MoO3 ensures that water vapor is continuously present near the growth substrates and protects source from poisoning. (ii) Monolayer MoS2 growth is critically dependent on a well-controlled balance between the partial pressure of MoO2(OH)2, MoO3, H2O, MoO2, and S.

2. Experimental Section

2.1. Strategically Placed Water on Substrates and/or MoO3 Source

In contrast to previous works, where water was introduced as vapor during the entire growth process [20], controlled amounts of water were introduced into the system before growth. Prior to its loading in the growth tube, the MoO3 source was wet with water, volumes ranging from 0 to 200 μL. The Si substrates also had water droplets of measured volume placed on their surface prior to being placed face down on the boat containing MoO3 source.

2.2. CVD Growth of MoS2 Monolayer

Monolayers of MoS2 films were synthesized on Si/SiO2 substrates by chemical vapor deposition using the setup illustrated in the schematic of Figure 1(a). Following the standard cleaning of substrates with acetone and isopropyl alcohol, they were immersed in Piranha solution (3 : 1 mixture of H2SO4 and H2O2) for 2 hours, rinsed with deionized water, and dried with compressed air. Four of these Si substrates with 300 nm of SiO2 were placed face down next to each other above a ceramic boat containing 30 mg of molybdenum oxide (MoO3) powder (≥99.5%, Alfa Aesar). A second boat containing 500 mg of sulfur powder (≥99.5%, Alfa Aesar) was placed at the low-temperature end of the furnace. The S source was positioned at the upstream end of the furnace, and its precise location was such that it ensured that S vapor was introduced to center zone (growth region where the substrates are placed) when it has reached a temperature of 630°C. The temperature-time plot in Figure 1(b) presents the temperature-ramp rate for the Mo and S source. The critical stages of the growth process are also indicated and labeled. As seen in the plot, the temperature of the MoO3 source reaches about 630°C while the temperature of the S source reaches about 115°C. It is worth noting that the time of introducing S into the growth region is critical to the end results. If S is introduced too early, then it will sulfurize the MoO3 source and no growth will result. On the other hand, if S enters the growth region too late, then the substrates will already be covered with MoO2 film over which “fin-like” structures of MoS2 will be deposited. In our experimental setup, the optimal time for introducing sulfur was determined to be at the stage when the temperature of the MoO3 source reaches ~630°C. Growth was carried out at atmospheric pressure with 10 sccm ultra-high-purity argon flowing through the tube. As shown in the temperature plot of Figure 1(b), the samples were held at a constant temperature of 700°C for 25 minutes and then cooled down to room temperature in two stages: slow cooling at a rate of −10°C min−1 under 10 sccm of Ar followed by fast cooling, achieved by opening the furnace under 100 sccm of Ar.

2.3. Characterization of MoS2

The morphology and dimensions of the MoS2 samples were studied using a scanning electron microscope (SEM) (Hitachi-SU1510) with an accelerating voltage of 10 kV. To determine the phonon spectra of the films, the as-grown films were characterized by Raman spectroscopy using an Almega XR Raman spectrometer. The Raman scattering was collected by a 50x objective with an excitation wavelength of 532 nm. A Veeco Nanoscope III AFM was used to measure the dimension and thickness of the monolayers of MoS2 under tapping mode.

3. Results

The following discussion applies to the schematic of Figure 1(a), which shows four Si substrates placed over the MoO3 powder source. A water droplet of volume 10 μL was placed on two center substrates prior to their loading into the chamber. The water droplets were placed toward the edge of the substrate surface (Figure 1(c)) resulting in an asymmetric distribution of water vapor around the MoO3 source. After growth, the substrates were removed from the growth chamber, and their surface was examined under optical microscope and SEM. The substrates were found to be covered with structures of different morphology (Figures 1(d) and 1(e)). After repeated experiments, it was verified that an asymmetric water vapor distribution over the MoO3 source results in large areal coverage of monolayers of MoS2 islands and continuous film. Figure 1(c) shows the length scales of the substrate and the position of the water droplet. It was found that MoS2 monolayers grew in the region on the substrate that was wetted by water prior to being loaded in the furnace. Toward the edge of the water droplet, as shown in Figure 1(c), isolated MoS2 islands with edge lengths of tens of microns were observed after growth. The islands merge to monolayer film and then multilayer film toward the center of the water droplet. The volume and position of strategically placed water droplets were found to be critical to MoS2 monolayer growth. These results are indicative of the promotive role that strategically placed water droplets have on the growth kinematics of MoS2.

To elucidate the role of water (placed either on the substrate surface and/or in the source material), the growth of the MoS2 under different conditions was studied and the results are shown in Figure 2 and Table 1. It is evident that the growth pattern of MoS2 in the presence of water is different from that without water (Figure 2(a)). There is also an obvious difference in morphology and areal coverage of MoS2, resulting from a change in the volume and positioning of water on the substrates and in the MoO3 source. As seen in the optical images (Figures 2(b) and 2(c)), the morphologies of the MoS2 flakes vary from multilayer distorted triangles to monolayer star-like shapes. Such shape evolution has been observed before and is attributed to varying Mo : S atom ratio [21]. As indicated by the light blue-colored regions on the substrates (shown in Figure 2(c) and Table 1), MoS2 monolayers have different areal coverage, dependent on the volume and positioning of the water droplet. A red spot (Figure 1(c)) is used to indicate the region of the substrates with the strategically placed water droplets that was the precise location of the MoS2 monolayer growth. The morphology of MoS2 changes along the droplet surface. MoS2 monolayers are formed at the contour/interface of the water droplets, whereas multilayers were observed to grow towards the center of the droplet.

The control group of samples (Figure 2(a)) with no water on the source and on substrates showed growth of multilayers of continuous MoS2 films with centimeter scale coverage. However, the introduction of water into the growth process results in a very different growth pattern. The volume of water and its placement around MoO3 source are two key factors that dictate the resulting growth pattern. Two cases were evaluated. In the first case, the experiments involved a symmetric distribution of water on substrates (Figure 2(b)), which means that the water droplets were centrally symmetric with respect to the MoO3 source. Multilayers of MoS2 along with islands, vertical nanofins or their combinations, were found on the substrates. In the second case, (Figure 2(c)), where the droplet position on the substrate is asymmetric with respect to the MoO3 source, there is higher MoS2 monolayer coverage. The exact mechanism of the dependence of water droplet positioning on substrates and its interaction with MoO3 vapor are subjects of future study. Our current understanding of large area MoS2 monolayer coverage in experiments with an asymmetric water distribution is that the asymmetric distribution causes a concentration gradient of the gaseous species, which redistributes the MoO3(g) around the Si growth substrates and consequently lowers the partial pressure of MoS2(g).

Table 1 presents a summary of our results with entire morphology and topographical images of samples surface under different water pretreatment conditions.

As the volume of water on substrates and MoO3 increases up to 200 μL, the resulting morphology is MoO2 microplates with rhomboidal shape, a result that has been reported by other groups; its origin attributed to a deficiency of S in the growth region [21, 22]. Our results indicate that the formation of MoO2 microplates could be attributed to an increase in the partial pressure of Mo, which implies a decrease in the partial pressure of S. Strategically placed water droplets thus most likely affect the partial pressure of precursors in gaseous phase and determine the morphology of as-grown MoS2.

The as-grown MoS2 monolayer and multilayers were characterized by atomic force microscopy (AFM) and Raman spectroscopy. Figures 3(a) and 3(b) show the Raman spectrum collected from the as-grown MoS2 thin films on SiO2/Si with different layer numbers. The two characteristic Raman modes attributed to two kinds of phonon vibration modes in MoS2 crystal were analyzed. The A1g mode denotes the vibration of sulfur atoms perpendicular to the 2D atom plane and the mode indicates the vibration of Mo and sulfur atoms parallel to the 2D atom plane [23]. The frequency difference () between the and A1g modes has a distinct connection with the layer number of MoS2. Monolayer MoS2 was reported to have to 20 cm−1, and the double peak difference tends to increase as the layer number of MoS2 increases. The as-grown MoS2 film shows double peak differences  cm−1 (Figure 3(a)) and 22.7 cm−1 (Figure 3(b)), indicating that the samples are monolayers and bilayers of MoS2, respectively. The thickness of the synthesized MoS2 monolayers is shown in AFM images in Figures 3(c) and 3(d). The height profile shows the triangular island-shaped monolayers of MoS2, and from the line scanning profile (inset with topography image) gathered along the green line drawn on images, the thickness along different edges of triangles is about 0.8 nm, corresponding to the thickness of MoS2 monolayers.

From an application point of view, sulfur vacancies and local oxidation of MoS2 strongly affect the electronic properties of this material [2426]. To study this effect, electrical measurement experiments are being designed to quantify the density of sulfur vacancies and diagnose the effect of partial oxidation of as-grown MoS2 on the intrinsic electronic properties of this material.

4. Discussion

Based on the schematics and optical microscopy findings of the experiments that were described in Figure 2 and Table 1, the effect of the strategically placed water droplets was explained through a sequence of chemical reactions, before and after introducing S into the growth region. The role of water pretreatment of the growth substrates is explained based on a reversible chemical reaction. This reaction has previously been reported to explain the role of a continuous flow of water vapor during the growth process [27]. The placement of water on the substrates and on the Mo source is such that before S vaporizes and gets introduced into the growth region, a reversible reaction occurs between the MoO3 source and water to form MoO2(OH)2(g). This gaseous compound is a volatile mono-hydroxide which dissociates into MoO3, as dictated by the reaction:

The reversible nature of (1) implies that excessive water conditions result in high concentrations of Mo precursors in the gas phase, whereas the concentration of S is unaffected. When the Mo : S ratio becomes greater than 1, it results in MoS2 growth under sulfur-deficient conditions. This explains our experimental results when strategically placed water volume reaches 200 μL.

As seen in the set of equations, the five kinds of partial pressure of MoO2(OH)2 (), MoO3 (), H2O (), MoO2 (), and S () influence each other, and the most favorable growth conditions for monolayer growth of MoS2 are critically dependent on a well-controlled balance between each of them. A summary of our experimental findings presented in Table 1 highlights the most favorable conditions for monolayers of MoS2 growth.

A study of the dynamics of the various stages of MoS2 growth shows that a precipitation reaction occurs as MoS2 in gaseous phase deposits on the substrate as solid MoS2 [28]. The driving force for this rate-determining step is the difference between the partial pressure of MoS2 in a gas phase () and the equilibrium vapor pressure of MoS2 in a solid phase (). For different layer number of MoS2, the equilibrium vapor pressure of MoS2 in a solid phase is different from each other and denoted with for monolayer, bilayer, tri-layer, etc., respectively. The precipitation reaction will occur when . At higher values of , the rate of precipitation is higher, resulting in a thicker film and hence an increase in . When a thermodynamic balance is attained between and for “” number of layers, the precipitation process stops at the end of the th layer growth. Thus, a critical fine control of should enable control over the number of MoS2 layers that deposit on the substrate surface. To enable monolayer growth, values should be maintained between values for a monolayer and a bilayer. In our experiments, the varying volume and position of the water droplet on the substrate surface affect through a sequence of chemical reactions. Thus, different layer number of MoS2 was grown to maintain equivalence between and . In contrast, in the absence of water pretreatment of the growth substrates and MoO3 source, experiments showed multilayer MoS2 film growth. The most likely cause for multilayer growth is that without the protection offered by water, the gaseous MoO3 is sulfurized and the sudden increase in results in pressures exceeding . This increased pressure inhibits monolayer growth and instead promotes growth of multilayers. Any variations in experimental conditions that affect through the partial pressures of the various reacting compounds would result in unpredictable and nonreproducible MoS2 growth. As seen from the results shown in Table 1, large areal growth of MoS2 monolayers occurred for very specific volume and position of water droplet in source and/or on substrate surface. The most promising growth condition is when is balanced with the other four kinds of partial pressure, , , , and in order to maintain an appropriate for monolayers of MoS2 growth. For adapting the current experimental growth recipe to enable large areal coverage of MoS2 monolayers, i.e., wafer-scale growth of monolayers of MoS2, it is important to maintain a specific balance between and the partial pressures of the four other components (discussed above) over the whole wafer area during the growth period. This continues to be experimentally challenging, though our preliminary results indicate a step in the right direction.

It has been reported that continuously supplied water vapor or oxygen can prevent source poisoning, a process of preventing sulfurization of the MoO3 source to ensure continuous evaporation of MoO3 source [29]. In our case, strategically placed water also prevents source poisoning through a reversible equilibrium reaction with MoO3. It can be seen from the set of equations in Figure 4 that at the critical temperature when S reaches the growth region, MoO3(g) will react with S to form MoO2, thereby decreasing the amount of MoO3(g). This will force the equilibrium reaction (1) to the left, resulting in the formation of MoO3(g). The supply of MoO3(g) is thus continuously maintained.

5. Conclusion

In summary, we present our results on the investigation into the role of strategically placed water droplets on the growth kinematics of MoS2 monolayers. We demonstrate that controlled amounts of water vapor in strategic locations in the growth region affect the partial pressure of MoO3 precursor. The critical positioning of the water droplets and its volume on the growth substrates and in the MoO3 source result in different growth patterns of MoS2 films. The MoS2 morphology varies from 2D films to islands of monolayers or multilayers with varying thickness. A sequence of chemical reactions combined with the dynamics of precipitation process was discussed in detail to fully understand the promotive role of water in the overall growth process. This simplified growth technique can be used for large-area growth of MoS2 monolayers that can then be used for device development.

Data Availability

No data were used to support this study.

Conflicts of Interest

The authors declare that there is no conflict of interest regarding the publication of this paper.