Liquid Shear Exfoliation of MoS2: Preparation, Characterization, and NO2-Sensing Properties

2D materials possess great potential to serve as gas-sensing materials due to their large, specific surface areas and strong surface activities. Among this family, transition metal chalcogenide materials exhibit different properties and are promising candidates for a wide range of applications, including sensors, photodetectors, energy conversion, and energy storage. Herein, a high-shear mixing method has been used to produce multilayered MoS2 nanosheet dispersions. MoS2 thin films were manufactured by vacuum-assisted filtration. The structural morphology of MoS2 was studied using ς-potential, UV–visible, scanning electron microscopy (SEM), atomic force microscopy (AFM), energy-dispersive X-ray spectroscopy (EDX), transmission electron microscopy (TEM), X-ray diffraction (XRD), and Raman spectroscopy (RS). The spectroscopic and microscopic analyses confirm the formation of a high-crystalline MoS2 thin film with good inter-sheet connectivity and relative thickness uniformity. The thickness of the MoS2 layer is measured to be approximately 250 nm, with a nanosheet size of 120 nm ± 40 nm and a number of layers between 6 and 9 layers. Moreover, the electrical characteristics clearly showed that the MoS2 thin film exhibits good conductivity and a linear I–V curve response, indicating good ohmic contact between the MoS2 film and the electrodes. As an example of applicability, we fabricated chemiresistive sensor devices with a MoS2 film as a sensing layer. The performance of the MoS2-chemiresistive sensor for NO2 was assessed by being exposed to different concentrations of NO2 (1 ppm to 10 ppm). This sensor shows a sensibility to low concentrations of 1 ppm, with a response time of 114 s and a recovery time of 420 s. The effect of thin-film thickness and operating temperatures on sensor response was studied. The results show that thinner film exhibits a higher response to NO2; the response decreases as the working temperature increases.


Introduction
During the previous decade, different sensing materials have been studied to develop high-performance gas sensors [1][2][3]. Among them, two-dimensional transition metal dichalcogenides (TMDs) nanostructured materials have gained attention as promising candidates for gas-sensing applications due to their unique electronic, optical, and mechanical properties [4]. In particular, MoS 2 nanomaterials have demonstrated potential applications for integrated flexible electronics [5] and gas sensing [6], with superior gas sensitivity and selectivity [7]. In the past decade, various types of MoS 2 nanostructures, including nanosheets, quantum dots, flowers, and nano-meshes, have been fabricated [8]. Among these morphologies, nanosheet structures have received particular attention due to their facile preparation method, highly exposed active sites, and relatively higher specific surface area per amount of material, which is superior for sensing applications [9].
Recent years have witnessed increasing research activity on using TMD nanosheets as an excellent sensing material for the fabrication of wearable sensing devices due to their high carrier mobility, mechanical strength, and the fact that no cracks form during the transfer from a rigid substrate to a flexible substrate [10]. Efficient exfoliation of single-or few-layered MoS 2 nanosheets can be achieved by various approaches, including lithium intercalation, sonication, mechanical exfoliation, etc. [11]. The liquid-phase shear exfoliation (LPSE) method is one of the most successful methods for the preparation of TMD nanosheets from the bulk phase of the materials in liquid dispersions, making it attractive to produce functional inks for printable electronic devices and thin films for various applications, including gas sensing. Although great progress has been made in this field, there are still many challenges regarding the availability of TMD inks and the largescale fabrication requirements of flexible/wearable devices. In particular, the availability of solution-dispersible TMD nanosheets that can be used to deposit, or print, a large-scale thin film on a variety of non-conventional substrates for the fabrication of TMD-based electronic devices [12,13]. While the sensing properties of MoS 2 nanosheets to various gases have been reported, most of these studies focused on materials mechanically exfoliated [14], CVDgrown [15], and sonication-assisted liquid-phase exfoliation methods [16]. The simplicity of this later method is attractive; however, it has many drawbacks, such as time-consuming, inferior product quality, low yields, and reproducibility. Shear exfoliation is similar to sonication exfoliation and can be regarded as one of the most industrially feasible methods for mass production of functional inks based on defect-free few-layer TMD materials [17]. Besides solution TMD material synthesis, thin-film processing methods that can be used to coat the colloidal dispersion of nanosheets onto a chosen substrate as patterns and functional thin films are highly desirable. Assembly of TMD-material films from the solution phase can be accomplished by various coating techniques, among which the vacuum-assisted filtration method is an easy and effective way to produce 2D-material thin films [18]. So far, thin-film MoS 2 fabrication has been mainly limited to the drop-casting method. Although this method is useful for coating a very thin layer, it suffers from a few drawbacks, e.g., the fact that the quality in terms of thickness is not uniform, the inhomogeneity of the films, and the coffee ring effect, which may hamper the performance in many applications. A deep understanding of the effect of material production, thinfilm processing methods, and device design on the MoS 2 sensing of NO 2 is necessary in order to develop a reliable NO 2 sensor. We reasoned that the development of solutionprocessed MoS 2 nanosheets from LPSE would allow greater control over the fabrication of large-area MoS 2 thin-film nanosheets and their integration into multifunctional devices. To evaluate the feasibility of this approach, we herein present the development of LPSE of MoS 2 and a vacuum-assisted filtration method to fabricate large-area MoS 2 thin films for gas sensing applications. Besides production and thin-film processing, a multiscale physico-chemical characterization of the produced nanosheet materials by means of various techniques, including ς-potential, UV-visible, Raman spectroscopy, atomic force (AFM), scanning electron microscopy (SEM), high-resolution transmission microscopies, and Xray diffraction, coupled with electrical measurements, was conducted to gain a deep understanding of the structure-properties relationship. Finally, the NO 2 -sensing properties of MoS 2 thin-film nanosheets have been investigated at room temperature in dry nitrogen.

Preparation of MoS 2 Dispersion
The MoS 2 water dispersion was prepared by the LPSE method, following the protocol previously reported [19,20]. The process of liquid exfoliation consists of two steps: (i) Pre-treatment and (ii) final exfoliation. In a typical preparation, 4 g of MoS 2 powder was first immersed in an 80 mL aqueous solution of sodium cholate (NaC) with a concentration of 7 g/L and sonicated for 30 min, followed by shear force mixing using a Silverson L5M-A Nanomaterials 2023, 13, 2502 3 of 17 mixer was applied at maximum speed (10,000 rpm) for 4 h. During the shear, the temperature of the solution was kept constant with an ice bath to prevent the dispersion from overheating. Afterwards, the dispersion was left overnight to settle before the supernatant was discarded. In this way, all of the aqueous-soluble contaminants that are present in MoS 2 powders are eliminated during the pre-treatment stage. Then, the sediment was re-dispersed in 80 mL of a fresh solution of NaC (concentration: 7 g/L) and exfoliated for a further 6 h at the same rate using an ice bath as in the pre-treatment stage. The top 2/3 of the supernatant was collected after settling overnight and centrifuged for two hours at 1500 rpm to remove unexfoliated MoS 2 . The supernatant after centrifugation was collected for further analysis, while the sediment was discarded.

MoS 2 Thin-Film Fabrication
To avoid aggregation, the MoS 2 solution was sonicated for 30 min to obtain a uniformly dispersed MoS 2 solution before each use. Then the MoS 2 dispersion was diluted in 100 mL of water and subjected to sonication for an additional 30 min to help MoS 2 nanosheets homogenously disperse in water. The MoS 2 films were fabricated using vacuum-assisted filtration on a nitrocellulose membrane, an MF-Millipore ® (Merck, Molsheim, France), with a pore size of 25 nm. The self-assembly films were then dried in a vacuum overnight. The self-assembly MoS 2 film on the membrane was then cut into sizes of choice and pressed against a cleaned substrate surface, which had previously been wetted with several drops of isopropanol, with the MoS 2 side in contact with the substrate. To ensure that the membrane was firmly attached to the substrate, we pressed it to remove the air at the interface. Finally, the membrane was dissolved using an acetone solution overnight, followed by a thorough cleaning with sufficient acetone and ethanol to remove the membrane's residue.

Characterization Methods
UV-visible absorption spectra were carried out on an Agilent Technologies Cary 60 UVvis spectrophotometer. Dynamic light scattering (DLS) and zeta potential measurements were performed with a Malvern Zetasizer Nano. Scanning electron microscopy (SEM, Hitachi S4800 microscope, Tokyo, Japan) was used to investigate the microstructure of MoS 2 film. The chemical state of the synthesized materials was analyzed using energy dispersion spectroscopy (EDS) (Thermo Ultra Dry-Electron Corporation, Waltham, MA, USA). The surface roughness of the MoS 2 thin film was obtained by using a Dimension Icon instrument from Bruker with the scan A system, which operated in tapping mode using a standard AFM probe featuring a pyramidal silicon tip (325 kHz, 40 N/m). X-ray diffraction measurements were performed on a Bruker D8 Discover (Germany) diffractometer using a Cu Kα source (λ1 = 1.5406 A, λ2 = 1.5444 A) and 2θ scans from 5 to 80 • . Raman spectra were recorded using a high-resolution confocal Raman microscope (Labram HR800; HORIBA Jobin Yvon, Palaiseau, France) with a microscope lens of 100 (NA = 1). Micro-Raman mapping was performed in high-resolution mode using a laser excitation of λ = 532 nm with a 20 s scan time and two accumulations per spectrum. Transmission electron microscopy (TEM) (Titan-Themis Thermo Fisher electron microscope operating at 80 and 300 kV) was performed to investigate the quality of the prepared films.

Sensor Design, Fabrication and Electrical Performances
For gas sensor tests, the interdigitated electrodes were fabricated on top of 300 nm-thick MoS 2 films transferred on glass substrates via the thermal deposition of 100 nm-thick gold through a metal shadow mask. The size of the interdigitated electrode is 11 mm × 8 mm. The interspacing between the adjacent fingers is typically 1 mm, with the width of interdigitated electrodes being 0.2 mm, as shown in Figure 1a. The fabricated MoS 2 sensors were annealed for 2 h at 100 • C in N 2 at ambient pressure to desorb attached gaseous molecules on the surface and then loaded into a sensing chamber as described in our previous work [20]. The electrical measurements were performed using a semiconductor parametric Keithley 4200-SCS (Tektronix company, Les Ulis, France) analyzer under ambient conditions. The measurements were carried out by applying a scan rate of 10 mV s −1 .
mm. The interspacing between the adjacent fingers is typically 1 mm, with interdigitated electrodes being 0.2 mm, as shown in Figure 1a. The fabrica sors were annealed for 2 h at 100 °C in N2 at ambient pressure to desorb atta molecules on the surface and then loaded into a sensing chamber as describ vious work [20]. The electrical measurements were performed using a semic ametric Keithley 4200-SCS (Tektronix company, Les Ulis, France) analyzer u conditions. The measurements were carried out by applying a scan rate of 1

Gas Sensing Measurements
The gas sensing measurement system consists of a gas generator (V-OV Westport, CT, USA), a Nextron chamber with micro-probe system analysi control system (1-97% relative humidity (RH) and 0.01 resolution), and a control system (20-200 °C). The concentration of NO2 is controlled by a gas ing permeation tube technology. An NO2 permeation rate of 6025 ng/min used. Dry N2 (99.99%) was used as the dilute gas to reach the desired conce working temperature of the fabricated sensor was controlled with a comme control system (−20-300 °C) connected to a computer. All sensing tests were room temperature with a relative humidity of 30%. Residual gas was remove the chamber with N2 flow for 30 min prior to introducing the gas. Before s urements, the sample was annealed at 100 °C in order to desorb all volatile m residues on the surface of MoS2. The response of the sensor upon exposure fined as: where Sresp is the sensitivity (response) of the gas sensors, and R and R0 are of the sensors after and before exposure to NO2 gas, respectively.

UV-Visible Spectroscopy
MoS2 exfoliation was carried out in the two-step protocol following th port [19]. The bulk MoS2 powders were first mixed together with an aqueo sodium cholate to get a comparable surface energy of 75 mJ/m 2 [21]. Followin step was to exfoliate MoS2 material through a high shear process, 6 h of mi rpm. The liquid phase exfoliation parameters were varied and optimized to production of the nanosheets. The main idea was to find out the optimal and stability of MoS2 nanosheet dispersion because this would help in sub dispersion, which is a prerequisite for ensuring good film formation. The Mo

Gas Sensing Measurements
The gas sensing measurement system consists of a gas generator (V-OVG) (Owlstone, Westport, CT, USA), a Nextron chamber with micro-probe system analysis, a humidity control system (1-97% relative humidity (RH) and 0.01 resolution), and a temperature control system (20-200 • C). The concentration of NO 2 is controlled by a gas generator using permeation tube technology. An NO 2 permeation rate of 6025 ng/min at 30 • C was used. Dry N 2 (99.99%) was used as the dilute gas to reach the desired concentration. The working temperature of the fabricated sensor was controlled with a commercial Nextron control system (−20-300 • C) connected to a computer. All sensing tests were carried out at room temperature with a relative humidity of 30%. Residual gas was removed by purging the chamber with N 2 flow for 30 min prior to introducing the gas. Before sensing measurements, the sample was annealed at 100 • C in order to desorb all volatile molecules and residues on the surface of MoS 2 . The response of the sensor upon exposure to NO 2 is defined as: where S resp is the sensitivity (response) of the gas sensors, and R and R 0 are the resistance of the sensors after and before exposure to NO 2 gas, respectively.

UV-Visible Spectroscopy
MoS 2 exfoliation was carried out in the two-step protocol following the previous report [19]. The bulk MoS 2 powders were first mixed together with an aqueous solution of sodium cholate to get a comparable surface energy of 75 mJ/m 2 [21]. Following, the second step was to exfoliate MoS 2 material through a high shear process, 6 h of mixing at 10,000 rpm. The liquid phase exfoliation parameters were varied and optimized to maximize the production of the nanosheets. The main idea was to find out the optimal concentration and stability of MoS 2 nanosheet dispersion because this would help in subsequent MoS 2 dispersion, which is a prerequisite for ensuring good film formation. The MoS 2 dispersion was characterized by UV-visible spectroscopy and the DLS technique to determine the concentration, the average layer number of MoS 2 nanosheets, the average sheet size, and the stability of the solution. Figure 2a shows a typical UV-vis absorption spectrum of the exfoliated 2H-MoS 2 nanosheet dispersion. As illustrated in this figure, the UV-visible absorption spectrum shows four characteristic bands A, B, C, and D at 670 nm, 608 nm, 450 nm, and 400 nm, respectively. The absorption peaks A and B are excitonic transitions arising from the direct band gap. The theoretical calculations have shown that the A and B excitonic peaks correspond to the energy band splitting due to spin orbitals at point K between the maximum valence band and the minimum value of the conduction band at the K point in the band diagram zone [22]. The C and D absorption bands are assumed to be related to the direct excitation transition of the M point in the Brillouin zone. Additionally, from the absorption spectrum of the MoS 2 dispersion, one can also extract useful information on the concentration, length, size, and thickness of the nanosheets. It is known that the lateral size and thickness of the nanosheets can be estimated from UV-visible absorption spectrum [17]. Such behavior is related to edge confinement effects; when the electronic band structure of the layered materials is changed, as a result of edge and confinement effects, the magnitude and position of the excitonic transitions change. So, the changes in the optical absorption spectrum reflect the changes in the size and thickness of the nanosheets. Roughly, the A and B absorption bands are related to the layer number of MoS 2 nanosheets, while the C and D absorption peaks represent the absorption of MoS 2 with smaller sizes. The mean length (L) of the MoS 2 nanosheets is dependent on the ratio of extinction at B-exciton to that of E and can be estimated by the formula [23]: where Abs B and Abs E are the intensities of the B (∼605 nm) band and E (∼345 nm), respectively. A value of the nanosheet size of 138 nm and a number of layers of six were estimated according to the Coleman method [23], which are similar to those reported in the literature [19]. The Beer-Lambert law A/l = Cα was used to calculate the concentration of the dispersion, where A/l is the absorbance per cell length of peak A and α is the absorbance coefficient. A nanosheet concentration of 0.63 g/L was determined using an absorbance coefficient of 1517 L·g −1 ·m −1 at 672 nm [24].
excitonic peaks correspond to the energy band splitting due to spin orbitals at point K between the maximum valence band and the minimum value of the conduction band at the K point in the band diagram zone [22]. The C and D absorption bands are assumed to be related to the direct excitation transition of the M point in the Brillouin zone. Additionally, from the absorption spectrum of the MoS2 dispersion, one can also extract useful information on the concentration, length, size, and thickness of the nanosheets. It is known that the lateral size and thickness of the nanosheets can be estimated from UV-visible absorption spectrum [17]. Such behavior is related to edge confinement effects; when the electronic band structure of the layered materials is changed, as a result of edge and confinement effects, the magnitude and position of the excitonic transitions change. So, the changes in the optical absorption spectrum reflect the changes in the size and thickness of the nanosheets. Roughly, the A and B absorption bands are related to the layer number of MoS2 nanosheets, while the C and D absorption peaks represent the absorption of MoS2 with smaller sizes. The mean length (L) of the MoS2 nanosheets is dependent on the ratio of extinction at B-exciton to that of E and can be estimated by the formula [23]:

− Abs Abs
where AbsB and AbsE are the intensities of the B (∼605 nm) band and E (∼345 nm), respectively. A value of the nanosheet size of 138 nm and a number of layers of six were estimated according to the Coleman method [23], which are similar to those reported in the literature [19]. The Beer-Lambert law A/l = Cα was used to calculate the concentration of the dispersion, where A/l is the absorbance per cell length of peak A and α is the absorbance coefficient. A nanosheet concentration of 0.63 g/L was determined using an absorbance coefficient of 1517 L·g −1 ·m −1 at 672 nm [24]. Sequential centrifugation was used to separate the nanosheets with varying thickness and lateral size distributions from the same stock solution. Figure 2b shows the blue shift of the A-exciton peak in the UV-visible absorption spectrum with decreasing mean nanosheet thickness. The nanosheet dispersions were obtained by sequential centrifugations with increasing speeds from 1500 to 9000 rpm. Sequential centrifugation was used to separate the nanosheets with varying thickness and lateral size distributions from the same stock solution. Figure 2b shows the blue shift of the A-exciton peak in the UV-visible absorption spectrum with decreasing mean nanosheet thickness. The nanosheet dispersions were obtained by sequential centrifugations with increasing speeds from 1500 to 9000 rpm.

Dynamic Light Scattering
DLS is a common technique to estimate the mean hydrodynamic size of MoS 2 nanosheets in solution and is an alternative method to direct TEM or AFM imaging. However, the Nanomaterials 2023, 13, 2502 6 of 17 anisotropy of nanosheet materials renders the use of light scattering more complicated due to their lateral size, thickness, and shape distributions. Coleman et al. [25] characterized MoS 2 nanosheets with the DLS technique and derived an empirical relationship between the apparent hydrodynamic radius calculated by the Stokes-Einstein relationship and the size estimated from TEM images. Figure 3 shows a typical DLS output graph; the hydrodynamic radius of the dispersed MoS 2 nanosheets ranges from 44 nm to 340 nm, with a maximum distribution at 142 nm. The mean lateral size deduced from the empirical relationship [25] is 120 nm ± 40 nm, which agrees with the average size of 138 nm estimated from UV-visible absorption measurements. Moreover, the ς-potential, which measures suspension stability, was also determined via DLS. A ς-potential of −39.1 mV was observed, indicating moderate stability of MoS 2 suspension fabricated using the LPSE method. nanosheets in solution and is an alternative method to direct TEM or AFM im ever, the anisotropy of nanosheet materials renders the use of light scatterin plicated due to their lateral size, thickness, and shape distributions. Colem characterized MoS2 nanosheets with the DLS technique and derived an empir ship between the apparent hydrodynamic radius calculated by the Stokes-E tionship and the size estimated from TEM images. Figure 3 shows a typical graph; the hydrodynamic radius of the dispersed MoS2 nanosheets ranges fr 340 nm, with a maximum distribution at 142 nm. The mean lateral size dedu empirical relationship [25] is 120 nm ± 40 nm, which agrees with the averag nm estimated from UV-visible absorption measurements. Moreover, the which measures suspension stability, was also determined via DLS. A ς-pote mV was observed, indicating moderate stability of MoS2 suspension fabricat LPSE method.

Atomic Force Microscopy
AFM was employed to study the sheet thickness and the mean sheet siz liated NO2. To do this, we drop casted the MoS2 nanosheet dispersion onto a S By analyzing the AFM height on a large number of nanosheets and plotting profile, the AFM image in Figure 4, along with the layer number distribution g a clear presence of MoS2 nanosheets with a few layers. Most of these nanoshe same apparent AFM height of 2-4 nm (Figure 4e,f), with a step height betwe nm, which agrees with the value of monolayer height of chemically exfoliate The number of monolayers per nanosheet of the randomly distributed MoS is in the range of 1-6 monolayers, and this is evident from the corresponding files and statistical analysis plots (Figure 4d-f). The lateral dimensions are in ∼50-150 nm. All these observations are in agreement with UV-visible and indicating that a high-speed rotator coupled with a higher shear time is a reli to obtain few-layered nanosheets of 2H-MoS2 without any severe damage to phology.

Atomic Force Microscopy
AFM was employed to study the sheet thickness and the mean sheet size of the exfoliated NO 2 . To do this, we drop casted the MoS 2 nanosheet dispersion onto a Si/SiO 2 wafer. By analyzing the AFM height on a large number of nanosheets and plotting their height profile, the AFM image in Figure 4, along with the layer number distribution graph, shows a clear presence of MoS 2 nanosheets with a few layers. Most of these nanosheets show the same apparent AFM height of 2-4 nm (Figure 4e,f), with a step height between 0.8 and 1 nm, which agrees with the value of monolayer height of chemically exfoliated MoS 2 [26]. The number of monolayers per nanosheet of the randomly distributed MoS 2 nanosheets is in the range of 1-6 monolayers, and this is evident from the corresponding height profiles and statistical analysis plots (Figure 4d-f). The lateral dimensions are in the range of ∼50-150 nm. All these observations are in agreement with UV-visible and DLS results, indicating that a high-speed rotator coupled with a higher shear time is a reliable method to obtain few-layered nanosheets of 2H-MoS 2 without any severe damage to its final morphology.

High-Resolution Transmission Electron Microscopy
High-resolution TEM (HRTEM) images were also recorded to estimate the late and number of nanosheet layers of the exfoliated MoS2. Figure 5 shows the HRTE ages of MoS2 nanosheets, along with their corresponding FFT (fast Fourier transfo sults. The HRTEM images provide a nanosheet size of a few hundred nanometer 150 nm). The exfoliated sheets consist of 6 to 9 layers, which means that the int distance of MoS2 is 0.6 nm (Figure 5a,b). Moreover, Figure 5c

High-Resolution Transmission Electron Microscopy
High-resolution TEM (HRTEM) images were also recorded to estimate the lateral size and number of nanosheet layers of the exfoliated MoS 2 . Figure 5 shows the HRTEM images of MoS 2 nanosheets, along with their corresponding FFT (fast Fourier transform) results. The HRTEM images provide a nanosheet size of a few hundred nanometers (100-150 nm). The exfoliated sheets consist of 6 to 9 layers, which means that the interlayer distance of MoS 2 is 0.6 nm (Figure 5a,b). Moreover, Figure 5c presents a larger MoS 2 grain composed of a small monolayer within the inset of its corresponding FFT of the selected area, illustrating the MoS 2 hexagonal symmetry in the <001> direction. Figure 5d corresponds to a magnified image of the area represented by the black rectangle in Figure 5c, indicating a perfect crystalline structure with a lattice spacing of 0.272 nm. All these demonstrate that the crystalline structure of the MoS 2 nanosheets obtained via the LPSE method remains intact after exfoliation.

Raman Spectroscopy
Raman spectroscopy provides structural and electronic information about TMD materials [27]. The Raman spectra for MoS2 obtained using 532 nm excitation are shown in Figure 6. The data were obtained using the minimum power (3 mW) in order to prevent laser heating. Two sets of peaks are visible, corresponding to E 1 2g and A1g modes: For exfoliated MoS2, E 1 2g is at 383.05 cm −1 and A1g is at 407.71 cm −1 ; for bulk MoS2, E 1 2g is at 379.02 cm −1 and A1g is at 404.2 cm −1 , indicating the in-plane and out-of-plane vibrational modes of the MoS2 layer. The peak position and the frequency difference (the peak-to-peak separation) of these two modes are sensitive to the layer thickness of MoS2. In bulk MoS2, the peak-to-peak separation of these modes is 25.2 cm −1 , while it decreases to 24.7 cm −1 in exfoliated MoS2 nanosheets, indicating the presence of four to five monolayers [27]. After exfoliation, the A1g and E 1 2g modes shifted towards the higher wavenumber.

Raman Spectroscopy
Raman spectroscopy provides structural and electronic information about TMD materials [27]. The Raman spectra for MoS 2 obtained using 532 nm excitation are shown in Figure 6. The data were obtained using the minimum power (3 mW) in order to prevent laser heating. Two sets of peaks are visible, corresponding to E 1 2g and A 1g modes: For exfoliated MoS 2 , E 1 2g is at 383.05 cm −1 and A 1g is at 407.71 cm −1 ; for bulk MoS 2 , E 1 2g is at 379.02 cm −1 and A 1g is at 404.2 cm −1 , indicating the in-plane and out-of-plane vibrational modes of the MoS 2 layer. The peak position and the frequency difference (the peak-to-peak separation) of these two modes are sensitive to the layer thickness of MoS 2 . In bulk MoS 2 , the peak-to-peak separation of these modes is 25.2 cm −1 , while it decreases to 24.7 cm −1 in exfoliated MoS 2 nanosheets, indicating the presence of four to five monolayers [27]. After exfoliation, the A 1g and E 1 2g modes shifted towards the higher wavenumber.

Raman Spectroscopy
Raman spectroscopy provides structural and electronic information about TMD materials [27]. The Raman spectra for MoS2 obtained using 532 nm excitation are shown in Figure 6. The data were obtained using the minimum power (3 mW) in order to prevent laser heating. Two sets of peaks are visible, corresponding to E 1 2g and A1g modes: For exfoliated MoS2, E 1 2g is at 383.05 cm −1 and A1g is at 407.71 cm −1 ; for bulk MoS2, E 1 2g is at 379.02 cm −1 and A1g is at 404.2 cm −1 , indicating the in-plane and out-of-plane vibrational modes of the MoS2 layer. The peak position and the frequency difference (the peak-to-peak separation) of these two modes are sensitive to the layer thickness of MoS2. In bulk MoS2, the peak-to-peak separation of these modes is 25.2 cm −1 , while it decreases to 24.7 cm −1 in exfoliated MoS2 nanosheets, indicating the presence of four to five monolayers [27]. After exfoliation, the A1g and E 1 2g modes shifted towards the higher wavenumber.

Scanning Electron Microscopy
To self-assemble the MoS 2 nanosheets on the substrate, vacuum-assisted filtration was employed to produce a compact MoS 2 film from dispersion. For morphological study and electrical characterization, the MoS 2 film was transferred onto a silicon wafer by dissolving the membrane and washing the obtained MoS 2 film with acetone. To analyze the surface morphology and film thickness of the MoS 2 nanosheet thin film, AFM and SEM characterizations were carried out. As shown in Figure 7a, all the films analyzed are tightly packed, and the nanosheets completely cover the substrate without any voids or cracks, forming a continuous porous nanosheet network. As a result, this method produces a film with good inter-sheet connectivity and relative thickness uniformity. The thickness of the MoS 2 layer is measured to be approximately 250 nm, as can be seen from the cross-sectional view of the SEM image of the MoS 2 thin film in Figure 7b

Scanning Electron Microscopy
To self-assemble the MoS2 nanosheets on the substrate, vacuum-assisted filtration was employed to produce a compact MoS2 film from dispersion. For morphological study and electrical characterization, the MoS2 film was transferred onto a silicon wafer by dissolving the membrane and washing the obtained MoS2 film with acetone. To analyze the surface morphology and film thickness of the MoS2 nanosheet thin film, AFM and SEM characterizations were carried out. As shown in Figure 7a, all the films analyzed are tightly packed, and the nanosheets completely cover the substrate without any voids or cracks, forming a continuous porous nanosheet network. As a result, this method produces a film with good inter-sheet connectivity and relative thickness uniformity. The thickness of the MoS2 layer is measured to be approximately 250 nm, as can be seen from the cross-sectional view of the SEM image of the MoS2 thin film in Figure 7b. To gain more insights into the chemical composition of the exfoliated MoS2, energy-dispersive spectrometer (EDS) analyses were carried out. The EDS measurements of MoS2 display signals of Mo and S (Figure 7c), which confirm the formation of MoS2.

Atomic Force Microscopy
Further, topography analysis of the MoS2 thin film was performed by AFM characterization, as shown in Figure 8a. The average roughness Ra of the MoS2 thin film was

Atomic Force Microscopy
Further, topography analysis of the MoS 2 thin film was performed by AFM characterization, as shown in Figure 8a. The average roughness Ra of the MoS 2 thin film was determined using NanoScope Analysis, which is 9.75 nm, revealing a low surface roughness regarding the thickness of the film, 250 nm, and the absence of any appreciable aggregation throughout the filtration process.
x FOR PEER REVIEW 10 of 18 determined using NanoScope Analysis, which is 9.75 nm, revealing a low surface roughness regarding the thickness of the film, 250 nm, and the absence of any appreciable aggregation throughout the filtration process.

X-ray Diffraction
X-ray diffraction (XRD) was carried out to investigate the phase structure of the exfoliated material. Figure 9 displays the XRD patterns of exfoliated MoS2. As can be observed in this XRD pattern, four peaks at 14.38°, 29°, 44.2°, and 61.7° of MoS2 are observed, which correspond to the (002), (004), (006), and (008) planes of 2H MoS2 (JCPDS:37-1492). The strong and narrow diffraction peak at 14.38° indicates the well-crystallization of MoS2 nanosheets. No secondary crystalline phases are observed in the XRD data, indicating that the nanosheets are not oxidized during the exfoliation process.

X-ray Diffraction
X-ray diffraction (XRD) was carried out to investigate the phase structure of the exfoliated material. Figure 9  Nanomaterials 2023, 13, x FOR PEER REVIEW determined using NanoScope Analysis, which is 9.75 nm, revealing a low surface ness regarding the thickness of the film, 250 nm, and the absence of any appreci gregation throughout the filtration process.

X-ray Diffraction
X-ray diffraction (XRD) was carried out to investigate the phase structure of foliated material. Figure 9

MoS 2 Based Sensor's Performance on NO 2 at Room Temperature
MoS 2 films were transferred onto glass substrates with interdigitated Au electrodes (shown in experimental Section 2.4, Figure 1b). Before gas sensing measurements, the electrical properties of fabricated MoS 2 thin films were characterized using the two-point I-V method under an applied bias voltage ranging from −10 V to +10 V, as shown in Figure 10. Gold electrodes were selected because it has been reported that gold forms a quasi-Ohmic contact with MoS 2 monolayers [28]. Firstly, we analyzed the effect of the thickness of the network film on its electrical properties. Two films with different thicknesses were analyzed. The I-V curves for both thinner and thicker films are shown in Figure 10. Both films exhibit good conductivity and a linear I-V curve response, indicating good ohmic contact between the MoS 2 film and the gold electrodes. This has been attributed to a reduction in the Schottky barrier at the MoS 2 -gold interface for MoS 2 multilayer nanosheets. In addition, the resistance of thin films decreases from 2.54 GΩ to 1.25 GΩ as the thickness increases, which is a characteristic of the percolation effect in nanosheet networks [29]. Moreover, thicker films have more available charge carriers, which can increase their conductivity. A value of conductivity of 4.02 × 10 −5 S·m −1 was estimated for MoS 2 nanosheet network film, a value close to that previously reported for MoS 2 printed films (∼2.5 × 10 −6 S m −1 ) [29] and solution-processed MoS 2 films (7 × 10 −7 S m −1 ) [30].
anomaterials 2023, 13, x FOR PEER REVIEW of the network film on its electrical properties. Two films with different thic analyzed. The I-V curves for both thinner and thicker films are shown in Fi films exhibit good conductivity and a linear I-V curve response, indicating contact between the MoS2 film and the gold electrodes. This has been attrib duction in the Schottky barrier at the MoS2-gold interface for MoS2 multilaye In addition, the resistance of thin films decreases from 2.54 GΩ to 1.25 GΩ as increases, which is a characteristic of the percolation effect in nanosheet n Moreover, thicker films have more available charge carriers, which can incre ductivity. A value of conductivity of 4.02 × 10 −5 S·m −1 was estimated for Mo network film, a value close to that previously reported for MoS2 printed film S m −1 ) [29] and solution-processed MoS2 films (7 × 10 −7 S m −1 ) [30]. Having characterized the electrical properties of the MoS2 nanosheet n we tested these devices for gas-sensing applications. The MoS2 chemires were probed inside the test chamber. After being heated at 100 °C for 10 m oxygen and water molecules, the sensors were exposed to a wider range of trations (1-10 ppm) for ~5 min at 30 °C. The resistance of the devices was recorded by applying a voltage of 10 V. To check the reproducibility of the a than four samples were characterized for each concentration. Different area and in the central region of the samples were analyzed and typically showed acteristics.
As shown in Figure 11a, the variation of the electrical response of MoS2 be 1.5% for 1 ppm and 9.7% for 10 ppm of NO2. Figure 11b shows the respon ery time of MoS2 thin film exposed to 1 ppm NO2. The fabricated MoS2 film tively rapid response time of 114 s to 1 ppm NO2, a value comparable to th phase-exfoliated MoS2 nanosheet-based sensor [31]. However, the recover quires 420 s. Moreover, when exposed to a high concentration of NO2, the Having characterized the electrical properties of the MoS 2 nanosheet network films, we tested these devices for gas-sensing applications. The MoS 2 chemiresistive sensors were probed inside the test chamber. After being heated at 100 • C for 10 min to remove oxygen and water molecules, the sensors were exposed to a wider range of NO 2 concentrations (1-10 ppm) for~5 min at 30 • C. The resistance of the devices was continuously recorded by applying a voltage of 10 V. To check the reproducibility of the analysis, more than four samples were characterized for each concentration. Different areas at the edges and in the central region of the samples were analyzed and typically showed similar characteristics.
As shown in Figure 11a, the variation of the electrical response of MoS 2 was found to be 1.5% for 1 ppm and 9.7% for 10 ppm of NO 2 . Figure 11b shows the response and recovery time of MoS 2 thin film exposed to 1 ppm NO 2 . The fabricated MoS 2 films show a relatively rapid response time of 114 s to 1 ppm NO 2 , a value comparable to that of a liquid-phase-exfoliated MoS 2 nanosheet-based sensor [31]. However, the recovery process requires 420 s. Moreover, when exposed to a high concentration of NO 2 , the device takes longer to recover its original resistance. As shown in Figure 11c, the recovery time from exposure to 5 ppm NO 2 was 1 h, and it took more than 2 h for MoS 2 thin film to recover from 10 ppm NO 2 . The incomplete recovery at room temperature is due to the strong binding between NO 2 and the reactive sites of MoS 2 . The long-term recovery phenomenon has been widely reported [32,33] and is known to be due to the high energy binding between the NO 2 molecules and MoS 2 surface defects. The sensor shows a linear variation of the response versus the concentration, as seen in Figure 11d, with a sensor response that varies linearly with NO 2 concentrations ranging from 1-10 ppm with a slope of −0.9357.  Table 1, the NO2 sensing performance of the shear-exfoliate films is comparable to those reported in previous works, especially in terms of res response time, and recovery time. −/>60 min As displayed in Table 1, the NO 2 sensing performance of the shear-exfoliated MoS 2 films is comparable to those reported in previous works, especially in terms of response, response time, and recovery time.  Some of the recent works on gas sensors based on nanostructured materials have shown a trend of increasing the sensor's sensitivity when the film thickness decreases, indicating that film thickness is a critical parameter for nanostructured gas sensor performance [34,41]. Table 2 shows the effect of the thickness of the film of the MoS 2 nanosheet gas sensor on sensitivity. As it is seen from Table 2, two different sensors with thicknesses of 150 and 300 nm were tested. The gas sensitivity of MoS 2 nanosheet gas sensors increases with a decrease in film thickness. A thinner film was found to exhibit better gas-sensing properties. This finding is in good agreement with previous work that has shown sensors based on rGO-MoS 2 thin films compared to the thick layer exhibit high sensitivity to NH 3 at room temperature in the ppm concentration range [34]. Such behavior is commonly observed in metal oxide semiconductor-based gas sensors [41], and it has been investigated by Windischmann and Mark, who showed that the sensitivity of gas sensors based on metal oxide semiconductors is inversely proportional to the carrier concentration [42]. In the case of nanostructured materials, the sensing performance of a thin-film sensor depends on various structural parameters such as the ratio of surface-to-volume, surface roughness, porosity, initial resistance of the film, etc. [43]. Increasing the ratio of surface-tovolume, roughness, and porosity would lead to an increase in the number of adsorption sites and, consequently, the sensing capabilities of the devices. As the thickness of the film decreases, the surface-to-volume ratio of the film increases. As a result, for a thinner sensing layer, the number of sites available for the adsorption of the analyte is expected to be higher, which results in a higher sensitivity.

The Effect of the Working Temperature on Gas Sensitivity
The effect of the working temperature on gas sensitivity was investigated by increasing the working temperature from room temperature (RT) to 100 • C ( Figure 12). The response of the MoS 2 sensor to 1 ppm of NO 2 was 2.1%, 1.7%, and 1.4% when the temperature increased from 30, 50, and 100 • C, respectively. The high response obtained at RT is attributed to possible sulfur vacancies, as reported for many works on sensors based on TMD materials. Density functional theory calculations have shown that NO 2 molecules strongly adsorb on edge sites compared to terrace sites in the basal plane of the MoS 2 . Edge sites, as well as defect sites of the MoS 2 , have a larger value of adsorption energy for NO 2 [44]. Thus, these defects are essential at low temperatures as NO 2 molecules deplete the material carrier charge through the vacancies and change the electronic conduction [44].

Gas-Sensing Mechanism
Based on previous works on the 2H-MoS2 chemiresistive sensor, the se nism can be explained by molecule adsorption followed by a charge transfe sensing layer and the adsorbed gas [41]. The change in the sensor's resista of this induced charge. The magnitude of the charge transfer is mainly rela nor/acceptor characteristics of gas molecules, such as their oxidizing or red and the intrinsic electrical properties of the sensing layer, conducting mate conducting material (n-or p-type). In ambient conditions, MoS2 is a p-type se due to the intrinsic S vacancies and oxygen incorporation originating from i air during exfoliation and film processing [45]. Pre-adsorbed oxygen molecu trons from the valence band of MoS2, thus creating a hole accumulation reg represented by Equation (2) and depicted in Figure 13. When NO2 (an elec gas) molecules are adsorbed on the 2H-MoS2 surface, they act as electron acc tion (3) and Figure 13). Therefore, they attract electrons from 2H-MoS2, lea crease in hole concentration, which is consistent with an increase in the cond a decrease in the resistance of the MoS2 as the NO2 concentration increases.

Gas-Sensing Mechanism
Based on previous works on the 2H-MoS 2 chemiresistive sensor, the sensing mechanism can be explained by molecule adsorption followed by a charge transfer between the sensing layer and the adsorbed gas [41]. The change in the sensor's resistance is a result of this induced charge. The magnitude of the charge transfer is mainly related to the donor/acceptor characteristics of gas molecules, such as their oxidizing or reducing ability, and the intrinsic electrical properties of the sensing layer, conducting material, or semiconducting material (n-or p-type). In ambient conditions, MoS 2 is a p-type semiconductor due to the intrinsic S vacancies and oxygen incorporation originating from its exposure to air during exfoliation and film processing [45]. Pre-adsorbed oxygen molecules take electrons from the valence band of MoS 2 , thus creating a hole accumulation region (HAR), as represented by Equation (2) and depicted in Figure 13. When NO 2 (an electron-acceptor gas) molecules are adsorbed on the 2H-MoS 2 surface, they act as electron acceptors (Equation (3) and Figure 13). Therefore, they attract electrons from 2H-MoS 2 , leading to an increase in hole concentration, which is consistent with an increase in the conductance, e.g., a decrease in the resistance of the MoS 2 as the NO 2 concentration increases.
In MoS 2 TMDCs, defects such as S vacancies (usually denoted as V S ) have a critical fundamental rule in the sensing mechanism toward gas molecule species contributing to carrier charge transfer. It is confirmed from previous reports that the adsorption of NO 2 on the 2H-MoS 2 surface is considered unfavorable without the existence of chalcogenide vacancies. In MoS2 TMDCs, defects such as S vacancies (usually denoted as VS) have a critical fundamental rule in the sensing mechanism toward gas molecule species contributing to carrier charge transfer. It is confirmed from previous reports that the adsorption of NO2 on the 2H-MoS2 surface is considered unfavorable without the existence of chalcogenide vacancies.

Conclusions
In summary, we have demonstrated that a high-shear mixing method produces a MoS2 nanosheet dispersion with a narrow sheet-size distribution, and a vacuum-assisted filtration method is able to produce compact film suitable for gas sensing, which is easy and cost-saving compared to conventional CVD techniques. The thin films with gold interdigitated electrodes were used to develop large-area chemiresistive gas sensors to detect NO2. The MoS2-based gas sensor shows a reversible response-recovery characteristic towards NO2 at 30 °C, but it exhibits a slow recovery due to the strong bonding between gas molecules and the MoS2 surface. Moreover, the responsivity to a low NO2 concentration of 1 ppm was demonstrated, making MoS2 a promising candidate for gas sensing applications. In terms of response optimization, we investigated thin-film thickness and working temperature. By studying the NO2 responses of 150 nm and 300 nm MoS2 thin films, it shows that thinner films process a higher response, which is due to a higher surface-to-volume ratio. By increasing the working temperature from 30 °C to 100 °C, we observed that the thin-film response decreased and recovery time increased with increasing temperature.

Conclusions
In summary, we have demonstrated that a high-shear mixing method produces a MoS 2 nanosheet dispersion with a narrow sheet-size distribution, and a vacuum-assisted filtration method is able to produce compact film suitable for gas sensing, which is easy and costsaving compared to conventional CVD techniques. The thin films with gold interdigitated electrodes were used to develop large-area chemiresistive gas sensors to detect NO 2 . The MoS 2 -based gas sensor shows a reversible response-recovery characteristic towards NO 2 at 30 • C, but it exhibits a slow recovery due to the strong bonding between gas molecules and the MoS 2 surface. Moreover, the responsivity to a low NO 2 concentration of 1 ppm was demonstrated, making MoS 2 a promising candidate for gas sensing applications. In terms of response optimization, we investigated thin-film thickness and working temperature. By studying the NO 2 responses of 150 nm and 300 nm MoS 2 thin films, it shows that thinner films process a higher response, which is due to a higher surface-to-volume ratio. By increasing the working temperature from 30 • C to 100 • C, we observed that the thin-film response decreased and recovery time increased with increasing temperature.