Photodetection Properties of MoS2, WS2 and MoxW1-xS2 Heterostructure: A Comparative Study

Layered transition metals dichalcogenides such as MoS2 and WS2 have shown a tunable bandgap, making them highly desirable for optoelectronic applications. Here, we report on one-step chemical vapor deposited MoS2, WS2 and MoxW1-xS2 heterostructures incorporated into photoconductive devices to be examined and compared in view of their use as potential photodetectors. Vertically aligned MoS2 nanosheets and horizontally stacked WS2 layers, and their heterostructure form MoxW1-xS2, exhibit direct and indirect bandgap, respectively. To analyze these structures, various characterization methods were used to elucidate their properties including Raman spectroscopy, X-ray diffraction, X-ray photoelectron spectrometry and high-resolution transmission electron microscopy. While all the investigated samples show a photoresponse in a broad wavelength range between 400 nm and 700 nm, the vertical MoS2 nanosheets sample exhibits the highest performances at a low bias voltage of 5 V. Our findings demonstrate a responsivity and a specific detectivity of 47.4 mA W−1 and 1.4 × 1011 Jones, respectively, achieved by MoxW1-xS2. This study offers insights into the use of a facile elaboration technique for tuning the performance of MoxW1-xS2 heterostructure-based photodetectors.


Introduction
Semiconductor photodetectors, namely photodiodes, are the most common types of detectors used in optical communication systems owing to their compact size, fast detection speed and high detection efficiency. Practical photodiodes can have a variety of device structures, beyond the basic PN junction construction, to improve their efficiency [1,2]. High-performance photodetectors have been used in a wide range of applications, including electro-optical displays [3], imaging [4], environmental monitoring [5], optical communication [6], military applications and security checks [7]. In these domains, two-dimensional (2D) materials, especially transition-metal dichalcogenides (TMDs), are becoming more attractive for designing photodetectors [8][9][10] due to their unique properties such as their ability to operate in the full range of visible light while having high photodetection polarization sensitivity, a fast photoresponse and high spatially resolved imaging [9]. This class of materials exhibits a layer-dependent electronic band structure in terms of unique physical characteristics and detection mechanisms [11,12]. Photodetectors based on 2D-TMDs materials are more sensitive throughout a wide range of the electromagnetic spectrum compared to photodetectors based on conventional bulk semiconductors [8]. However, an enhanced absorption coefficient and a longer lifespan of photoexcited carriers are preferred for optimal photocurrent generation and photodetector operation. For instance, MoS 2 thin films have a high light absorption coefficient of 10 7 m −1 , a high absorption of 95% of total light [13] and a direct bandgap of 1.8 eV. Moreover, photodetectors based on 2D-TMDs possess a good current on/off ratio, efficiency,

Materials and Methods
Among several techniques used to fabricate 2D materials, CVD is the commonly employed technique to control defects, crystallinity, and morphology of this class of materials. In particular, CVD sulfurization process is a facile one-step processing route that allows the fabrication of several sulfur based 2D-materials. The fabrication control is often ensured by monitoring several processing parameters such as gas flow, temperature, heating rate, the distance between precursors, and the position and height of the collecting substrate [29]. The synthesis of all samples was obtained in a one-step CVD process using a single-heating zone furnace at atmospheric pressure, as shown in Figure 1. The CVD system mainly consists of a quartz tube connected to high-purity (99.999%) argon cylinder streaming at flow rates of 70 sccm and 50 sccm for MoS 2 fabrication and for WS 2 and MoS 2 /WS 2 respectively. The SiO 2 /Si (1 × 1 cm 2 ) substrates were rinsed successively in deionized water, acetone, and ethanol in an ultrasonic bath for 10 min each.
A powder consisting of WO 3 (≥99.8%), MoO 3 (≥99.8%), or mixed WO 3 /MoO 3 with ratio 1:1 was mixed with Sulphur ≥99.98% using a ball-milling machine. All chemicals were purchased from Sigma Aldrich (Saint Louis, MO, USA). A diluted suspension solution with a concentration of 100 mg/mL was prepared using either the WO 3 /S, the MoO 3 /S, or the MoO 3 -WO 3 /S with ethanol and subsequently sonicated to enhance the homogeneity of the solution. Prior to the CVD process, a drop of 10 µL of the suspension solution was directly dropped onto the cleaned substrate using a pipette, as shown in Figure 1. Subsequently a 250 mg of sulfur powder was introduced at the edge of the sealed end of the quartz tube. Initially, the furnace was heated from room temperature up to 400 • C at a 20 • C/min heating rate, then to 850 • C at 5 • C/min for MoS 2 fabrication and 950 • C at 5 • C/min for WS 2  A powder consisting of WO3 (≥99.8%), MoO3 (≥99.8%), or mixed WO3/MoO3 with ratio 1:1 was mixed with Sulphur ≥99.98% using a ball-milling machine. All chemicals were purchased from Sigma Aldrich (Saint Louis, MO, USA). A diluted suspension solution with a concentration of 100 mg/mL was prepared using either the WO3/S, the MoO3/S, or the MoO3-WO3/S with ethanol and subsequently sonicated to enhance the homogeneity of the solution. Prior to the CVD process, a drop of 10 μL of the suspension solution was directly dropped onto the cleaned substrate using a pipette, as shown in Figure 1. Subsequently a 250 mg of sulfur powder was introduced at the edge of the sealed end of the quartz tube. Initially, the furnace was heated from room temperature up to 400 °C at a 20 °C/min heating rate, then to 850 °C at 5 °C/min for MoS2 fabrication and 950 °C at 5 °C/min for WS2 and MoS2/WS2. During the growth of either the MoS2, the WS2, or the MoxW1-xS2, the temperature was maintained at 850 °C or 950 °C for 30 min. Then, the furnace was allowed to cool down naturally to room temperature.
The morphology of the fabricated specimens was analyzed using a dual beam focused-ion beam and scanning electron microscope (FIB-SEM) Scios 2 ThermoFisher Scientific (Waltham, MA, USA) microscope. The same tool was also used for the preparation of thin lamella for the transmission electron microscopy (TEM) study. TEM study was carried out using Tecnai and Titan systems from ThermoFisher Scientific (Waltham, MA, USA). TEM samples were prepared on a thin carbon coated Cu mesh grid by transferring the grown TMD samples through a gentle physical exfoliation. The vibrational modes of the processed samples were examined with a micro-Raman spectrometer Renishaw (Wotton-under-Edge, UK), using a laser excitation of 532 nm. The crystalline structure was investigated by X-ray diffraction (XRD) using a D8 Discover diffractometer Bruker (Billerica, MA, USA); KaCu = 1.54 Å. The optical properties were investigated using a UV-Visnear IR spectrometer JASCO V-670. An X-ray photoelectron spectroscopy (XPS) study was carried out using a PHI VersaProbe III scanning XPS microprobe Physical Electronics (Chanhassen, MN, USA), equipped with a monochromatic and microfocused Al K-Alpha The morphology of the fabricated specimens was analyzed using a dual beam focusedion beam and scanning electron microscope (FIB-SEM) Scios 2 ThermoFisher Scientific (Waltham, MA, USA) microscope. The same tool was also used for the preparation of thin lamella for the transmission electron microscopy (TEM) study. TEM study was carried out using Tecnai and Titan systems from ThermoFisher Scientific (Waltham, MA, USA). TEM samples were prepared on a thin carbon coated Cu mesh grid by transferring the grown TMD samples through a gentle physical exfoliation. The vibrational modes of the processed samples were examined with a micro-Raman spectrometer Renishaw (Wotton-under-Edge, UK), using a laser excitation of 532 nm. The crystalline structure was investigated by X-ray diffraction (XRD) using a D8 Discover diffractometer Bruker (Billerica, MA, USA); K aCu = 1.54 Å. The optical properties were investigated using a UV-Vis-near IR spectrometer JASCO V-670. An X-ray photoelectron spectroscopy (XPS) study was carried out using a PHI VersaProbe III scanning XPS microprobe Physical Electronics (Chanhassen, MN, USA), equipped with a monochromatic and microfocused Al K-Alpha X-ray source (1486.6 eV). During the experiment, an E-neutralizer (1 V), was implemented. CasaXPS processing software 2.3. was used for the calibration and the curve fitting. Finally, electrical measurements were performed using Palmsens-4 electrochemical workstation under ambient conditions. Figure 2a presents the Raman spectra of the MoS 2 sample. The observed main Raman vibrational modes indicate the presence of hexagonal 2H-MoS 2 such as E 1 2g (382 cm −1 ) and A 1g (409.8 cm −1 ), which correspond to the in-plane and out-of-plane atomic vibrations, respectively [30].

Material Characterization
The main Raman vibration modes recorded for WS 2 correspond to a 2H-WS 2 structure such as 2LA(M), E 1 2g and A 1g as shown in Figure 2b. The strongest peak at 350 cm −1 may be fitted with two sub-peaks with maximum frequencies of 323.6 cm −1 and 351.3 cm −1 leading to 2LA(M) and E 1 2g , respectively. The first-order vibrational mode E 1 2g represents the in-plane vibration between sulfur and tungsten atoms while the A 1g vibrational mode at 420 cm −1 corresponds to the out-of-plane vibration of sulfur atoms. It is worth noting that the A 1g is sensitive to the number of WS 2 layers [31]. The main Raman vibration modes recorded for WS2 correspond to a 2H-WS2 s ture such as 2LA(M), E 1 2g and A1g as shown in Figure 2b. The strongest peak at 350 may be fitted with two sub-peaks with maximum frequencies of 323.6 cm −1 and 351.3 leading to 2LA(M) and E 1 2g, respectively. The first-order vibrational mode E 1 2g repre the in-plane vibration between sulfur and tungsten atoms while the A1g vibrational m at 420 cm −1 corresponds to the out-of-plane vibration of sulfur atoms. It is worth n that the A1g is sensitive to the number of WS2 layers [31].
Regarding the MoxW1-xS2 heterostructure, the Raman peaks corresponding to MoS2 and 2H-WS2 are present, as shown in Figure 2c. The positions of the E 1 2g and vibrational modes do not seem to shift compared to the observed peaks in individual ples. This indicates that WS2 and MoS2, obtained through our preparation route, hav effect on each other's long-range Coulomb interactions between the effective charg previously reported [24]. The x-ray diffraction (XRD) diagram illustrated in Figure 3a shows clear diffra peaks at 14.25°, 25.81°, 32.15°, 44.13° and 60.21° corresponding to 2H-MoS2. They ar tributed, respectively, to the (002), (004), (103), (006) and (008) planes of the hexagona MoS2. For WS2, the XRD diagram shows several significant diffraction peaks at 14.3°, 2 43.9°, 59.8°, and 77.13° as can be seen in Figure 3b. They are attributed to 2H-WS2 p (002), (004), (006), (008) and (00,10).  Regarding the Mo x W 1-x S 2 heterostructure, the Raman peaks corresponding to 2H-MoS 2 and 2H-WS 2 are present, as shown in Figure 2c. The positions of the E 1 2g and A 1g vibrational modes do not seem to shift compared to the observed peaks in individual samples. This indicates that WS 2 and MoS 2 , obtained through our preparation route, have no effect on each other's long-range Coulomb interactions between the effective charges as previously reported [24].
The x-ray diffraction (XRD) diagram illustrated in Figure   The main Raman vibration modes recorded for WS2 correspond to a 2H-WS2 st ture such as 2LA(M), E 1 2g and A1g as shown in Figure 2b. The strongest peak at 350 may be fitted with two sub-peaks with maximum frequencies of 323.6 cm −1 and 351.3 leading to 2LA(M) and E 1 2g, respectively. The first-order vibrational mode E 1 2g repres the in-plane vibration between sulfur and tungsten atoms while the A1g vibrational m at 420 cm −1 corresponds to the out-of-plane vibration of sulfur atoms. It is worth no that the A1g is sensitive to the number of WS2 layers [31].
Regarding the MoxW1-xS2 heterostructure, the Raman peaks corresponding to MoS2 and 2H-WS2 are present, as shown in Figure 2c. The positions of the E 1 2g and vibrational modes do not seem to shift compared to the observed peaks in individual s ples. This indicates that WS2 and MoS2, obtained through our preparation route, hav effect on each other's long-range Coulomb interactions between the effective charge previously reported [24].
The x-ray diffraction (XRD) diagram illustrated in Figure   For the MoxW1-xS2 heterostructure, the MoS2 and WS2 peaks are present in the co sponding XRD diagram given in Figure 3c. This confirms the successful fabrication o heterostructure. The sharp diffraction peaks observed on the spectra are a clear indica of the high crystallinity of the fabricated nanosheets.
The SEM images show different morphologies for the MoS2, WS2 and the heterost ture samples. The MoS2 flakes ( Figure 4a) are observed to grow vertically. This is h lighted at a higher magnification in Figure 4b. Similar results were reported previo [32]. On the other hand, WS2 shows accumulated crystals stacked on top of the subst For the Mo x W 1-x S 2 heterostructure, the MoS 2 and WS 2 peaks are present in the corresponding XRD diagram given in Figure 3c. This confirms the successful fabrication of the heterostructure. The sharp diffraction peaks observed on the spectra are a clear indication of the high crystallinity of the fabricated nanosheets.
The SEM images show different morphologies for the MoS 2 , WS 2 and the heterostructure samples. The MoS 2 flakes (Figure 4a) are observed to grow vertically. This is highlighted at a higher magnification in Figure 4b. Similar results were reported previously [32].  As can be seen in Figure 4e,f, the MoxW1-xS2 heterostructure exhibits a mixed morphology. It consists of both vertically aligned MoS2 nanosheets and stacked layers of WS2. A coherence between the two phases is observed with no visible segregation between the two compounds.
The change in the MoS2 and WS2 morphology could be attributed to the following hypotheses: (1) The high CVD reaction temperature used to process the WS2 could enhance the nucleation kinetics of the first WS2 seeds allowing the coalescence process to occur horizontally. In contrast, for MoS2 the reaction temperature is lower leading to dispersed seeds in the surface of the substrate favoring the coalescence on the top of the first layers; (2) The WS2 weight may impede the vertical shape of WS2; (3) The flow rate used for the processing of WS2 (50 sscm) is lower compared to the one for MoS2 fabrication (70 sccm), which may not allow the evacuation of sulfur excess.
In order to comprehend the mixing mechanism of MoS2 and WS2 structures, we have conducted further microstructure analysis using HRTEM for the three samples as shown in Figure 5. As can be seen in Figure 4e,f, the Mo x W 1-x S 2 heterostructure exhibits a mixed morphology. It consists of both vertically aligned MoS 2 nanosheets and stacked layers of WS 2 . A coherence between the two phases is observed with no visible segregation between the two compounds.
The change in the MoS 2 and WS 2 morphology could be attributed to the following hypotheses: (1) The high CVD reaction temperature used to process the WS 2 could enhance the nucleation kinetics of the first WS 2 seeds allowing the coalescence process to occur horizontally. In contrast, for MoS 2 the reaction temperature is lower leading to dispersed seeds in the surface of the substrate favoring the coalescence on the top of the first layers; (2) The WS 2 weight may impede the vertical shape of WS 2 ; (3) The flow rate used for the processing of WS 2 (50 sscm) is lower compared to the one for MoS 2 fabrication (70 sccm), which may not allow the evacuation of sulfur excess.
In order to comprehend the mixing mechanism of MoS 2 and WS 2 structures, we have conducted further microstructure analysis using HRTEM for the three samples as shown in Figure 5.  Figure 5c indicates the nature of the flakes' growth, where certain layers tend to be continuous and few sheets get terminated due to the absence of growth space, which could be at the origin of the vertically aligned 2H-MoS 2 . Figure 5e-g shows the cross-sectional views of the WS 2 sample. The smooth planar growth of the WS 2 is clearly visible compared to MoS 2 , in agreement with the observation on the SEM images. An interplanar distance of 2H-WS 2 is determined at~0.65 nm. Figure 5h shows typical TEM bright field images obtained from the Mo x W 1-x S 2 heterostructure showing an overlapping region (red and purple boxes). Zoomed images of both boxes indicate the presence of 2H-MoS 2 (red box, Figure 5i) and 2H-WS 2 (purple box, Figure 5j), which is a signature of the MoS 2 -WS 2 heterostructure form with a crystallographic relationship (200) MoS2 //(101) WS2 . Moreover, it is worth noting that 2D materials in their single-phase form are often considered to be thermally stable. Nevertheless, in their heterostructure form, they may suffer from thermal defects and induced stresses that could affect their electronic and optical properties. In the current work, these inherent stresses have not been evaluated as well as their impact on the heterostructure physical properties. It is believed that the effects of these stresses on the heterostructure structural and electronic properties cannot be neglected.   To precisely investigate the chemical composition of the fabricated samples, we have conducted XPS analyses, as shown in Figure 6.
suffer from thermal defects and induced stresses that could affect their electronic and optical properties. In the current work, these inherent stresses have not been evaluated as well as their impact on the heterostructure physical properties. It is believed that the effects of these stresses on the heterostructure structural and electronic properties cannot be neglected.
To precisely investigate the chemical composition of the fabricated samples, we have conducted XPS analyses, as shown in Figure 6.  [33]. For Sulfur, S 2p peaks are recorded and shown in Figure 6c. The two strong peaks at 162.54 eV and 163.73 eV are attributed to S 2− 2p3/2 and p1/2 states. Regarding the WS2, the XPS survey scan (Figure 6d) indicates the presence of W and S in the material, translated by the strong peaks for W 4f7/2 and W 4f5/2 appearing at 33.28 eV and 35.43 eV, respectively (Figure 6e). An additional peak appears at 38.72 eV that could be attributed to W p3/2. The S peaks appearing at 162.88 eV and 164.07 eV are due to the S 2− 2p3/2 and p1/2 states, respectively (Figure 6f). Moreover, the XPS survey scan obtained from the MoxW1-xS2 heterostructure sample is provided in Figure 6g, indicating the presence of W, Mo, and S in the heterostructure. The Mo 3d peaks are found at 229.79 eV and 232.95 eV (Figure 6h), and the S 2s peak is found at 227.2 eV Additionally, strong peaks of W appear at 33.3 eV and 35.42 eV, corresponding to W 4f7/ and W 4f5/2. The peak at 35.91 eV could be attributed to W-O bond, while the 38.80 eV peak is attributed to W 5p3/2 (Figure 6i). The S peaks appearing at 162.84 eV and 164.0 eV correspond to the S 2− 2p3/2 and p1/2 states, respectively (Figure 6j).

Optical Properties
Density functional theory (DFT) calculations, using generalized gradient approxima-  (Figure 6b). A small peak appearing at 236.2 eV indicates a minor oxidation of the Mo material [33]. For Sulfur, S 2p peaks are recorded and shown in Figure 6c. The two strong peaks at 162.54 eV and 163.73 eV are attributed to S 2− 2p 3/2 and p 1/2 states. Regarding the WS 2 , the XPS survey scan (Figure 6d) indicates the presence of W and S in the material, translated by the strong peaks for W 4f 7/2 and W 4f 5/2 appearing at 33.28 eV and 35.43 eV, respectively (Figure 6e). An additional peak appears at 38.72 eV that could be attributed to W p 3/2 . The S peaks appearing at 162.88 eV and 164.07 eV are due to the S 2− 2p 3/2 and p 1/2 states, respectively (Figure 6f). Moreover, the XPS survey scan obtained from the Mo x W 1-x S 2 heterostructure sample is provided in Figure 6g, indicating the presence of W, Mo, and S in the heterostructure. The Mo 3d peaks are found at 229.79 eV and 232.95 eV (Figure 6h), and the S 2s peak is found at 227.2 eV. Additionally, strong peaks of W appear at 33.3 eV and 35.42 eV, corresponding to W 4f 7/2 and W 4f 5/2 . The peak at 35.91 eV could be attributed to W-O bond, while the 38.80 eV peak is attributed to W 5p 3/2 (Figure 6i). The S peaks appearing at 162.84 eV and 164.0 eV correspond to the S 2− 2p 3/2 and p 1/2 states, respectively (Figure 6j).

Optical Properties
Density functional theory (DFT) calculations, using generalized gradient approximation (GGA) and Perdew Burke Ernzerhof (PBE) methods (implemented in Quantum Espresso) were first used to estimate the structural optimizations and electronic attributes (see Supplementary Materials). We have summarized in Table 1 the crystal structures of MoS 2 and WS 2 , as well as the cell parameters, the cutoff wave function and charge densities for both materials implemented in DFT calculations. For all configurations (monolayer and bilayer), the cell parameters and atomic locations were fully relaxed using the Broyden-Fletcher-Goldfarb-Shanno (BFGS) approach until the remaining force on each atom was less than 10 −3 Ryd/Bohr (see Supplementary Materials). To simulate monolayers in our calculations, a vacuum space of 15 was created along both sides of the z-axis to isolate the crystal and prevent interactions between the adjacent layers. For the sampling of the Brillouin zone, a Monkhorst-Pack technique is used, with k-point meshes of 9 × 9 × 2 for the bulk and 9 × 9 × 1 for the monolayer and bilayer structures. The generated optimal structure was utilized to compute the band structures for various setups. The results revealed a transition from an indirect bandgap to a direct bandgap as the structure changed from bulk to monolayer for both WS 2 and MoS 2 . Similar results have been reported [29,[34][35][36]. The band structure of the MoS 2 and WS 2 layers were well conserved, as seen in Figure 7.  To simulate monolayers in our calculations, a vacuum space of 15 Å was creat along both sides of the z-axis to isolate the crystal and prevent interactions between t adjacent layers. For the sampling of the Brillouin zone, a Monkhorst-Pack technique used, with k-point meshes of 9 × 9 × 2 for the bulk and 9 × 9 × 1 for the monolayer a bilayer structures. The generated optimal structure was utilized to compute the ba structures for various setups. The results revealed a transition from an indirect bandg to a direct bandgap as the structure changed from bulk to monolayer for both WS2 a MoS2. Similar results have been reported [29,[34][35][36]. The band structure of the MoS2 a WS2 layers were well conserved, as seen in Figure 7. The conduction band minimum (CBmin) and the valence band maximum (VBmin) the MoS2 and WS2 monolayers are positioned at the K point, respectively. MoxW1-xS2 is indirect semiconductor with a 1.45 eV indirect bandgap. Unlike their homogeneous layers' counterparts, the CBmin of MoxW1-xS2 heterostructures is positioned at the K poi whilst the VBmin is located at the point . Through van der Waals interactions, the Mo and WS2 monolayers produce an atomically sharp type-II heterointerface, which may favorable for electron-hole pair separation. Free electrons and holes will spontaneou separate in a type II heterostructure, which is useful for optoelectronics and solar ener conversion applications [37][38][39].
Furthermore, the optical reflectance of all samples was measured at room tempe ture in the wavelength range 400-800 nm as shown in Figure 8a. The conduction band minimum (CB min ) and the valence band maximum (VB min ) of the MoS 2 and WS 2 monolayers are positioned at the K point, respectively. Mo x W 1-x S 2 is an indirect semiconductor with a 1.45 eV indirect bandgap. Unlike their homogeneous bilayers' counterparts, the CB min of Mo x W 1-x S 2 heterostructures is positioned at the K point, whilst the VB min is located at the point Γ. Through van der Waals interactions, the MoS 2 and WS 2 monolayers produce an atomically sharp type-II heterointerface, which may be favorable for electron-hole pair separation. Free electrons and holes will spontaneously separate in a type II heterostructure, which is useful for optoelectronics and solar energy conversion applications [37][38][39].
Furthermore, the optical reflectance of all samples was measured at room temperature in the wavelength range 400-800 nm as shown in Figure 8a.
MoS 2 shows the lowest reflectance compared to the other samples, with the presence of both excitons, A and B, clearly visible at 636 nm and 688 nm positions, respectively [40,41]. This is due to the high optical absorption of the vertical morphology of the MoS 2 nanosheets with high specific area and light trapping via the multiple scattering effects [42]. On the other hand, WS 2 exciton appears clearly at 620 nm, showing the highest reflectance caused by its planar morphology. Finally, the reflectance of the Mo x W 1-x S 2 sample shows a mixed behavior between MoS 2 and WS 2 with an enhancement of the MoS 2 excitons. This validates the successful fabrication of the heterostructure as confirmed by the Raman spectroscopy and HRTEM analyses discussed earlier. MoS2 shows the lowest reflectance compared to the other samples, w of both excitons, A and B, clearly visible at 636 nm and 688 nm positio [40,41]. This is due to the high optical absorption of the vertical morphol nanosheets with high specific area and light trapping via the multiple s [42]. On the other hand, WS2 exciton appears clearly at 620 nm, showin flectance caused by its planar morphology. Finally, the reflectance of the ple shows a mixed behavior between MoS2 and WS2 with an enhancem excitons. This validates the successful fabrication of the heterostructure the Raman spectroscopy and HRTEM analyses discussed earlier.

PEER REVIEW
To obtain the optical bandgap, the reflectance measurements record To obtain the optical bandgap, the reflectance measurements recorded for all investigated samples are implemented in Kubelka-Munk model as per the following equation [43]: where K represents the molar absorption coefficient, S is the scattering factor, and R is the reflectance. Our results show that both MoS 2 and WS 2 exhibit a bandgap of 1.77 eV and 1.85 eV, respectively, approximately equivalent to bandgap obtained using DFT calculations. In contrast, the Mo x W 1-x S 2 sample shows a low bandgap of 1.63 eV compared to DFT calculations, which is probably due to the heterostructure construction and implementation in DFT that does not seem to reproduce the effective form of the heterostructure.

Photoresponse Measurements
To evaluate the optoelectronic properties of the MoS 2 , WS 2 , and Mo x W 1-x S 2 nanocomposite films, we deposited a pair of Au electrodes onto the device surface, as illustrated in Figure 9. The electrical measurements were conducted at room temperature under dark and illumination conditions using a halogen lamp (70 mW/cm 2 ), and under different excitation wavelengths ranging from 400 nm to 700 nm. The effective detection area of the samples was 0.075 cm 2 . The responsivity (R ) and the relative detectivity (D*) [ obtained using the following equations: where q is the absolute value of an electron charge (1.6 × 10 − sivity given in units of mA W −1 , and D* is the relative detec In Figure 10, compared to the other samples, MoS2 ex 2 Figure 9. Schematic of the photoresponse measurements set up used for all samples.
The J-V curves were collected using a voltage sweep program −/+ 5 V at 0.1 V step. The J Ph at 5 V bias was subsequently computed using the following formula: where I light and I dark represent the current obtained in the light and the dark conditions, respectively. A is the active detection area. The photoresponse (P) of our samples was computed using the following equation: The responsivity (R λ ) and the relative detectivity (D*) [17] of the photodetectors were obtained using the following equations: where q is the absolute value of an electron charge (1.6 × 10 −19 Coulombs), R λ is the responsivity given in units of mA W −1 , and D* is the relative detectivity given in units of Jones.
In Figure 10, compared to the other samples, MoS 2 exhibits the highest J Ph achieving 4.8 mA/cm 2 , compared to the other samples, while WS 2 and the heterostructure Mo x W 1-x S 2 have shown lower values of 0.8 mA/cm 2 and 3.7 mA/cm 2 , respectively. Moreover, the highest photoresponse is also achieved by MoS 2~6 .8 × 10 4 %, while WS 2 exhibits the lowest one of 1.5 × 10 3 %. The photoresponse of Mo x W 1-x S 2 heterostructure of~5.8 × 10 3 % is similar to previously reported values [44]. This strong photoresponse is due to the high optical absorbance of the vertical MoS 2 nanosheets, known for possessing a high ability to capture light and a quick charge transfer [45] (e.g., Figure 10a For MoS2, we obtain maximum values of R and D*, respectively at 68 m 6.3 × 10 3 Jones, with a 5 V bias voltage. It is worth noting that higher R values reported [18,19,46,47], however in those works the considered active area was smaller (~10 −7 cm 2 ) and a high applied bias was considerably higher (~50 V) co our present study. Instead, our findings concur that WS2 and MoxW1-xS2 heter exhibit R and D* of 8.9 mA W −1 , 2.1 × 10 10 Jones and 47.4 mA W −1 , 1.4 × 10 11 Jon tively compared to MoS2.
For further examinations of the photoresponse of our samples, we conduc of photoresponse measurements under monochromatic light excitations in the tween 400 nm and 700 nm wavelengths. Figure 11d shows the relative detectivi puted at various wavelengths using the above-mentioned formula (Equations ( which is in agreement with the measured values in Figure 10d. From the latt notice that R and D* are decreasing from 77.2 to 10.9 mA W −1 and from 7.2 × 1 10 10 Jones, respectively, with an increasing excitation wavelength from 400 nm For MoS 2 , we obtain maximum values of R λ and D*, respectively at 68 mA W −1 and 6.3 × 10 3 Jones, with a 5 V bias voltage. It is worth noting that higher R λ values were also reported [18,19,46,47], however in those works the considered active area was extremely smaller (~10 −7 cm 2 ) and a high applied bias was considerably higher (~50 V) compared to our present study. Instead, our findings concur that WS 2 and Mo x W 1-x S 2 heterostructure exhibit R λ and D* of 8.9 mA W −1 , 2.1 × 10 10 Jones and 47.4 mA W −1 , 1.4 × 10 11 Jones, respectively compared to MoS 2 .
For further examinations of the photoresponse of our samples, we conducted a series of photoresponse measurements under monochromatic light excitations in the range between 400 nm and 700 nm wavelengths. Figure 11d shows the relative detectivity D* computed at various wavelengths using the above-mentioned formula (Equations (4) and (5)), which is in agreement with the measured values in Figure 10d. From the latter, one can notice that R λ and D* are decreasing from 77.2 to 10.9 mA W −1 and from 7.2 × 10 11 to 1.8 × 10 10 Jones, respectively, with an increasing excitation wavelength from 400 nm to 700 nm. The maximum responsivity R of 77.2 mA W −1 and the relative detectivity D* of 10 11 Jones, were obtained at 400 nm excitation for MoS2 sample as reported elsew [18,19,46,47].

PEER REVIEW 12
To further correlate our investigation with existing works, we conducted a surv available data for sole MoS2, WS2 and the heterostructure made out of these compou The survey is summarized in Table 2.  The maximum responsivity R λ of 77.2 mA W −1 and the relative detectivity D* of 7.2 × 10 11 Jones, were obtained at 400 nm excitation for MoS 2 sample as reported elsewhere [18,19,46,47].
To further correlate our investigation with existing works, we conducted a survey of available data for sole MoS 2 , WS 2 and the heterostructure made out of these compounds. The survey is summarized in Table 2.
The photodetection obtained for MoS 2 shows similar performances in terms of detectivity with slightly better performance of our CVD-fabricated MoS 2 compared to the sample processed by PLD. However, there is a difference in responsivity, mainly attributed to applied high bias voltages (~50 V) and very low active area used 10 −7 cm 2 , compared to our active area of 10 −2 cm 2 . This shows that our one-step fabricated MoS 2 has exhibited high photodetection performances despite the larger effective area used. To the best of our knowledge, no previous results were reported on WS 2 better than our results on CVDgrown samples achieving a responsivity of 11.8 mA W −1 and a detectivity of 10 10 Jones, obtained at a very low bias voltage of 5 V. The control of CVD parameters to grow a high quality Mo x W 1-x S 2 monolayer was already reported [28]. Similarly, we have fabricated the Mo x W 1-x S 2 using a one-step CVD fabrication method achieving a responsivity and a detectivity of 47 mA W −1 and 10 11 Jones, respectively. Similar results were reported before using a mixture of MoS 2 /WS 2 monolayers and graphene or using MoS 2 core shell containing WS 2 . We believe that our results indicate that our processing route consisting of a one-step and low-cost CVD fabrication technique of MoS 2 , WS 2 and Mo x W 1-x S 2 hold a strong promising potential for the future development of scalable photodetector devices.

Conclusions
A Mo x W 1-x S 2 heterostructure was successfully synthesized by a one-step CVD route. The high purity and high quality of the samples have been confirmed by multiple characterization techniques. The incorporation of the MoS 2 , WS 2 and the Mo x W 1-x S 2 heterostructure into photoconductive devices demonstrated a promising potential of these compounds to be used as broadband photodetectors. In particular, the Mo x W 1-x S 2 heterostructure has achieved a responsivity of 47.4 mA W −1 and a relative detectivity of 1.4 × 10 11 Jones under visible light excitation ranging from 400 to 700 nm.  Data Availability Statement: Data would be made available upon request to the corresponding author.