RF Sputtered Nb-Doped MoS2 Thin Film for Effective Detection of NO2 Gas Molecules: Theoretical and Experimental Studies.

Doping plays a significant role in affecting the physical and chemical properties of two-dimensional (2D) dichalcogenide materials. Controllable doping is one of the major factors in the modification of the electronic and mechanical properties of 2D materials. MoS2 2D materials have gained significant attention in gas sensing owing to their high surface-to-volume ratio. However, low response and recovery time hinder their application in practical gas sensors. Herein, we report the enhanced gas response and recovery of Nb-doped MoS2 gas sensor synthesized through physical vapor deposition (PVD) toward NO2 at different temperatures. The electronic states of MoS2 and Nb-doped MOS2 monolayers grown by PVD were analyzed based on their work functions. Doping with Nb increases the work function of MoS2 and its electronic properties. The Nb-doped MoS2 showed an ultrafast response and recovery time of t rec = 30/85 s toward 5 ppm of NO2 at their optimal operating temperature (100 °C). The experimental results complement the electron difference density functional theory calculation, showing both physisorption and chemisorption of NO2 gas molecules on niobium substitution doping in MoS2.

Introduction

Industrial technologies over the past several decades have significantly increased the amount of toxic gases in the atmosphere, which has dramatically affected the natural environment and human health.[1−3] Toxic gases in the atmosphere such as NO2, SO2, H2S, CO, H2, NH3, and CH4, can seriously affect human health and lead to global warming.[4−7] NO2 has attracted significant attention because it can affect human health even at low ppb (parts per billion) levels. Moreover, inhalation of NO2 leads to asthma, bronchitis, pulmonary edema, and respiratory problems.[8−11] Hence, real-time environmental monitoring of NO2 gas sensors has become important in day-to-day life. Semiconductor gas detectors, electrochemical devices, mass sensors, and piezoelectric devices are normally utilized to monitor air quality. However, complex fabrication, low sensitivity, slow response, high power consumption, poor stability, and high device cost hinder these types of sensors in a wide range of applications.[5,7,12] This has stimulated researchers to focus on gas sensors with high response and selectivity. Recently, two-dimensional (2D) materials have gained considerable attention in gas sensing because of their physical and chemical properties.[13,14] Transition-metal dichalcogenide (TMD)-based materials are widely used in transistors, gas sensors, wearable devices, energy storage, and catalysis because of their unique thickness-dependent bandgap and excellent thermal properties.[13−17] TMDs particularly have large surface-area-to-bulk ratio thus enhance the adsorption of gases which can significantly modify their properties, making them promising materials for gas detection.[18,19] In particular, monolayer MoS2 has attracted significant interest because of its direct bandgap of ≅1.83 eV and has proved to be versatile material for a wide range of applications in sensors.[18,19] Monolayer MoS2 consists of three atomic layers, covalently bonded Mo and six S atoms, forming a sandwiched structure. However, the absence of a dangling bond in defect-free monolayer MoS2 leads to chemically inert, terminal by S atoms.[20] Various methods, especially the introduction of defects and metal dopants have been explored, that improve the chemical activity and sensitivity of the basal plane of the MoS2 monolayer.[20−22] Jia et al. investigated the theoretical structural stability and gas adsorption of monolayer MoS2 doped with V, Nb, and Ta. They showed that this doping can significantly improve the adsorption of gas molecules.[20] Suh et al. synthesized Nb-doped MoS2 by substitution cation doping, which is highly stable for the adsorption of volatile species.[23] Doping with V and Cr in monolayer MoS2 has been studied by Alex et al., showed that doping affected the electronic and mechanical properties of monolayer MoS2.[24] This means that doping with metals improves the physical and chemical properties of monolayer MoS2.[25] Nb is a transition metal with good solubility and one less d-orbital electron occupancy than Mo. Doping MoS2 with Nb results in the injection of a high concentration of hole carriers. In addition, MoS2 and NbS2 have similar lattice parameters which do not cause a significant change in the lattice structure even when a NbS2 covalent bond is formed.[26−29] The incorporation of metals such as Nb or V should enhance the strong interaction between the metal and gas molecules (NO2). Thus, such dopant materials promise to enhance the selectivity and gas response of monolayer MoS2 gas sensors.[20] However, Nb-doped MoS2 monolayer NO2 gas sensors have not yet been reported. Herein, we present theoretical and experimental investigations on the adsorption properties of monolayer MoS2 and Nb-doped MoS2 for gas molecules (acetone, NH3, toluene, NO2, and CO). Controlled doping of the Nb-doped MoS2 monolayer was performed using a physical vapor deposition (PVD) method. Furthermore, we demonstrated the behavior of doped and undoped MoS2 gas sensors toward different concentrations of NO2 at 100 °C. Finally, we report on density functional theory (DFT) simulations to understand the adsorption energetics and changes in the electronic structure of MoS2 and Nb-doped MoS2 after interaction with NO2 gas molecules. DFT simulation complements the experimental results and also helps to understand the gas-sensing mechanism of MoS2 and Nb-doped MoS2.

Results and Discussion

A MoS2 and Nb-doped MoS2 monolayer was fabricated using radio frequency (RF)-sputtering and was transferred onto the bare polyethylene terephthalate (PET) substrate, as shown in the schematic representation in Figure and verified by atomic force microscope (AFM). The thickness of MoS2 and Nb-doped MoS2 was approximately 0.75 and 0.77 nm, as shown in Figure b–e. Figure S1a,b shows the SEM top view and energy dispersive X-ray analysis (EDX) of MoS2 and Nb-doped MoS2 fabricated device. The result shows that the perfect nanochannel is made up from the monolayer MoS2 and Nb-doped MoS2 thin film. EDX analysis reveals the presence of Nb, Mo, and S elements in the RF sputtered monolayer thin film. We have analyzed the electrical transport of MoS2 and Nb-doped MoS2 fabricated devices, as shown in Figure . The results indicate that the current of Nb-doped MoS2 has sharply decreased compared with pure MoS2, which indicates that substitutional Nb doping significantly increased hole injection and also suppressed the heavily n-type MoS2 characteristics (Figure a,b). Moreover, the slight movement in the threshold voltage to positive voltage indicated the doping of Nb in the MoS2 monolayer. Figure c,d shows I–V curves of MoS2 and Nb-doped MoS2 by applying different gate biases (0 to 50 V), the results indicate that restraint of electrical transport conductivity with Nb doping.[25] The electronic states of the MoS2 and Nb-doped MoS2 thin films were investigated using Kelvin probe force microscopy (KPFM). The work function of MoS2 increased from 4.583 ± 0.005 to 4.98 ± 0.005 eV, and the new states were formed near the valence band, maximum owing to substitution doping of Nb in the MoS2 films (Figure a,b). Raman spectroscopy analysis was performed to determine the formation of monolayer MoS2 and Nb-doped MoS2. As shown in Figure c,d, two major peaks at 384 and 404 cm–1 were observed in the samples, corresponding to in-plane E2g1 and out-of-plane Ag1 vibrations of Nb-doped MoS2. The distance between the vibrational peaks was 20 cm–1, which indicates the formation of monolayer MoS2 that agrees with the AFM image.[30,31]

Figure 1

(a) Schematic representation of fabrication and transfer process of monolayer MoS2 and Nb-doped MoS2 using the RF magnetron sputtering technique. (b–e) AFM image and line profile analysis of MoS2 and Nb doped MoS2.

Figure 2

(a, b) Id vs Vg characteristics of monolayer MoS2 and Nb doped MoS2 (linear scale). (c, d) Id vs Vd characteristics of monolayer MoS2 and Nb-doped MoS2.

Figure 3

(a, b) KPFM image and band diagrams of pristine MoS2 and Nb-doped MoS2. (c, d) Raman spectrum of MoS2 and Nb-doped MoS2.

X-ray photoelectron spectroscopy (XPS) analysis was performed to determine the oxidation state of the elements via binding energy and surface element composition. Binding energies have distinct values for each element, which are used to identify the individual elements in materials. Figure shows the high-resolution spectra of Mo 3d, Nb 3d, and S 2s core levels. Figure a illustrates the wide spectra of Nb-doped MoS2 which clearly indicates the presence of Nb, Mo, S, and C. Three major peaks are featured at 228.1 eV (Mo 3d5/2), 231.2 eV (Mo 3d3/2), and 225.5 eV (S 2s) as shown in Figure b. The intense peak at 228.1 eV is attributed to Mo4+ (i-Mo4+) and the charge state of molybdenum in Nb-doped MoS2. The binding energy at 234.5 eV is due to Mo6+ of the unreacted precursor of MoO3, which may be a contaminant in the Nb-doped MoS2.[32−35] The peak at 225.5 eV is due to an overlapping S 2s peak, corresponding to sulfur close to a defect, which agrees well with the previous reports.[26] The sulfur environment can be more clearly studied by employing the S 2p core level spectrum, as shown in Figure c, and the convoluted XPS shows two doublet peaks at 160.8 and 161.9 eV, respectively. The higher binding energy (160.8 eV) is attributed to sulfur vacancies (d-s) and the lower binding energy to intrinsic S.[35−38] The convoluted XPS spectrum of Nb 3d is composed of doublets attributed to the spin–orbital splitting of 3d5/2 (205.7 eV) and 3d3/2 (208.7 eV), as depicted in Figure d. The results indicate that the as-grown Nb-doped MoS2 has a uniform high-quality intrinsic structure on the substrate.[23]

Figure 4

(a) Wide spectrum of Nb-doped MoS2. (b, c) XPS spectra of Mo 3d and S 2p of Nb-doped MoS2. (d) XPS spectrum of Nb 3d.

We fabricated a flexible monolayer gas sensor on PET substrate, as shown in the schematic representation in Figure a,b, and investigated the gas sensing behavior of MoS2 and Nb-doped MoS2 toward NO2 gas in air atmosphere at different temperatures. First, the relative responses of MoS2 and Nb-doped MoS2 gas sensors to different gas at various temperatures were examined. From Figure c,d, it was observed that the MoS2 and Nb-doped MoS2 showed higher response toward NO2 than other gases. In addition, the selectivity of both sensing materials were evaluated to determine the efficiency of the sensor. The selectivity of the as-synthesized sensing materials was examined at different operating temperatures (50–200 °C) for the most commonly interfering gases such as toluene, CO, acetone, and NH3 in an air atmosphere. The resistance change was assessed at 5 ppm concentration for each target gas, and the results indicated the highly selective nature of monolayer MoS2 and Nb-doped MoS2 toward NO2 gas compared with other gases. It was observed that monolayer MoS2 showed a slight response to the interfering gases, and the observed response values were 2.2 (150 °C) and 2.35 (50 °C) for CO and acetone, respectively. However, Nb-doped MoS2 showed high selectivity to NO2 and no response (S = 1) to the interfering gases, making the Nb-doped MoS2 sensor selectively operable during low-temperature operation for NO2 detection even when it coexists with other reducing gases in the air. Figure a compares the typical dynamic gas response–recovery characteristics of monolayer MoS2 and Nb-doped MoS2 gas sensors with different concentrations of NO2 gas exposure at 100 °C. The gas response is represented in by normalized change in resistance of the sensor. For monolayer MoS2, the corresponding gas response values to 5, 7, 10, 12, and 16 ppm of NO2 gas were noted to be 5.2, 8.1, 10.5, 12.35, and 16.23, respectively. Notably, the gas response of the Nb-doped MoS2 monolayer was enhanced compared to that of the undoped monolayer MoS2 over the full gas concentration range. The corresponding gas response values for the same sequence of NO2 gas concentrations were 40.22, 50.12, 61, 70.9, and 80.5, respectively. Figure b shows the estimated sensitivity (S) of Nb-doped MoS2 (S = 10) to be higher than that of undoped MoS2 monolayer (S = 2), and a linear trend is observed for the undoped MoS2 sensor as compared to Nb-doped MoS2, where a slightly curved fitting is observed. In addition, undoped MoS2 showed a good response to 3 ppm of NO2 in air (Figure S2). Figure c shows the dynamic gas-sensing response and recovery curves of monolayer MoS2 upon exposure to 5 ppm of NO2 at operating temperatures of 50, 100, 150, and 200 °C. Monolayer MoS2 showed an increased sensing response at 100 °C compared to other temperatures, along with rapid recovery upon the exclusion of NO2. However, when the temperature rose beyond 150 °C, there was a gradual decrease in the gas response due to the significant desorption rate compared to the adsorption rate (representing a typical volcano curve). In addition, the low desorption rate of the molecules at a lower temperature of MoS2 may be due to the high binding energy of NO2 or strong bonding between NO2 and MoS2. At lower temperatures, the chemical activity of NO2 gas molecules on the MoS2 surface is relatively low, which leads to a low response. At the optimum temperature (100 °C), the desorption rate of the NO2 gas molecules enhances the recovery from the Nb-MoS2 by removing accumulated moisture before reacting with targeted gas and initiating the surface catalyst at elevated temperatures. However, a further increase in temperature leads to a quick reaction and penetration of NO2 gas molecules in the Nb-MoS2-sensing film, which results in a decrease in the gas response.[39,40]

Figure 5

(a) Schematic representation of Nb-doped MoS2 device under NO2 gas sensor. (b) Crystal structure in which Nb is doped in the substitutional sites of Mo. (c, d) Selectivity of MoS2 and Nb-doped MoS2.

Figure 6

Measurement of sensing performance: (a, b) Dynamic response and recovery curves of MoS2 and Nb-doped MoS2 to various concentrations of NO2. (c, d) Gas response curve of MoS2 and Nb-doped MoS2 at different temperature. (e, f) Response and recovery times toward 5 ppm at different operating temperatures for both samples. (g, h) Several cycles to determine the repeatability in detection of NO2 for MoS2 and Nb-doped MoS2.

Further, Figure d shows the dynamic gas sensing response and recovery curves of Nb-doped MoS2 upon exposure to 5 ppm of NO2 at different operating temperatures of 50, 100, 150, and 200 °C. Compared to monolayer MoS2, Nb-doped MoS2 shows a significant increase in sensing response to 5 ppm of NO2 with an enhanced stable behavior at 100 °C compared to that at other temperatures. A trend similar to that observed in monolayer MoS2 with a decrease in gas response beyond 100 °C was seen in the sensing performance of Nb-doped MoS2. However, Nb-doped MoS2 displayed a higher and stable response with almost full recovery when compared to monolayer MoS2, suggesting that doping with Nb plays a significant role in gas sensing enhancement. We further investigated the time required for both samples to acquire 90% response and recovery on gas exposure and removal, respectively, at different operating temperatures. The transient dynamic response of the monolayer MoS2 and Nb-doped MoS2 sensors in Figure e,f shows that the corresponding response–recovery times are (tres = 110/80 s) and (trec = 30/85 s) toward 5 ppm of NO2 at their optimal operating temperature (100 °C). The Nb-doped MoS2 sensor displays a faster response and more stable recovery than monolayer MoS2. To further verify the stability and reproducibility of the sensors, both sensors were kept in ambient conditions and inspected after one month. Both monolayer MoS2 and Nb-doped MoS2 sensors were subjected to 4 and 10 reversible cycles (each cycle comprised of one response and recovery process) under exposure to 5 ppm of NO2 at 100 °C, as displayed in Figure g,h. The response was similar for each cycle. Furthermore, both sensors displayed excellent stability and reversibility in each cycle. In Figure g, the sensor response of monolayer MoS2 toward 5 ppm of NO2 at 100 °C for each cycle was 5.2, 4.96, 5.42, and 4.98, which can be considered a very stable response. Figure h shows that Nb-doped MoS2 exhibited a highly stable sensing behavior toward 5 ppm of NO2 at 100 °C with cycle responses of 40.22, 40.48, 40.52, and 39.95, which is <2% variability. These results reveal that the sensor has good stability toward NO2 gas. The results of previously reported MoS2 gas sensors are compared with those of our gas sensor are shown in Table .[41−47]

Table 1

Comparison between the Literature-Based MoS2 Gas Sensors and Present Work

sensing materials	S (%)	NO2 conc. ppm	T (°C)	tres (s)	trec (s)	ref
MoS2 aerogel	12	0.5	200	33	107	(41)
MoS2/Pt	18	5	–	1600	1600	(42)
atomic layered MoS2	25	50	100	–	–	(43)
MoS2 monolayer	12	5	100	71	310	(1)
MoS2 hallow sphere	40.3	100	150	79	225	(44)
graphene/MoS2	8	5	150	–	–	(45)
sputtered MoS2	32	100	150	56	80	(46)
MoS2 nanowire	2	1	120	16	172	(47)
Nb-MoS2	44	5	100	30	85	this work

To date, the exact gas-sensing mechanism is unknown, but in this work, its elucidation is based on the degree of charge perturbation on the surface between NO2 and monolayer MoS2 interaction. The change in the resistance of MoS2 directly corresponds to the concentration and the amount of adsorption of NO2 gas molecules on the MoS2 surface. The gas-sensing properties are strongly affected by the surface stoichiometry and the intrinsic properties of MoS2. The mechanism of gas sensing is briefly explained below with the help of DFT.

DFT Study of Adsorption of NO2 Molecules

Initially, the atomic structure of Nb atoms placed in the sulfur vacancy position of MoS2 (MoS2-S-Vac-Nb) and substitutional Nb atoms in the Mo atom location of MoS2 (MoS2-Mo-Subs-Nb) were optimized by fully relaxing the atomic structure until the remaining residual force is smaller than 0.05 eV/Å. In the resultant atomic structure of Nb-doped MoS2, a NO2 molecule was adsorbed and then optimized as shown in Figure c,e. Similarly, the MoS2-NO2 structure was geometrically optimized to have a residual force smaller than 0.05 eV/Å. Figure a shows the geometrically optimized MoS2-NO2 atomic structure. The binding energy of the NO2 molecule on the MoS2 or Nb-doped MoS2 supercell is calculated as Ebind = E(MoS – (EMoS+ ENO), where E(MoS is the total energy of the NO2 molecule adsorbed on MoS2 or MoS2Nb-doped supercells, EMoS is the total energy of the MoS2 or MoS2Nb-doped supercells, and ENO is the total energy of the NO2 molecule. The calculated binding energies for MoS2-NO2, MoS2-S-Vac-Nb-NO2, and MoS2-Mo-Subs-Nb-NO2 atomic structures are −0.681, −3.9, and −0.658 eV, respectively. These binding energies indicate that (i) the NO2 molecule is physisorbed onto MoS2 and MoS2-Mo-Subs-Nb atomic structures and (ii) NO2 molecules are chemically bonded onto the Nb-doped sulfur vacancy of MoS2 as the binding energy is higher than 1.0 eV. In the experimental work, the sensors were recovered after NO2 sensing measurements, which indicates that NO2 molecules were physisorbed onto Nb-doped MoS2. This analysis rules out the possibility of using Nb-doped sulfur vacancy MoS2 crystals in sensing measurements. To assess the amount of charge and transfer between NO2 molecules and MoS2 or Nb-doped MoS2 channels, Mulliken population analysis was performed. It was found that the NO2 molecules accept 0.066, 0.198, and 0.009 electrons for MoS2-NO2, MoS2-S-Vac-Nb-NO2, and MoS2-Mo-Subs-Nb-NO2 atomic structures, respectively. The electron difference density plots are shown in Figure b,d,f, depicting this charge transfer process. In the case of MoS2-NO2 and MoS2-Mo-Subs-Nb-NO2 structures, charge transfer from the MoS2 layer to NO2 molecules occurs via the van der Waals bonding. For the MoS2-S-Vac-Nb-NO2 system, chemical bonding-induced charge transfer happens.

Figure 7

Geometrically optimized atomic configuration of (a) MoS2-NO2, (c) Nb-doped in sulfur vacancy MoS2-NO2, and (e) substitutional Nb-doped in Mo atom position of MoS2-NO2. These electron difference density isosurface plots are shown in (b), (d), and (f), respectively. Isovalue: 0.11 e/Å3.

To understand the atomistic origin of the high sensing response of Nb-doped MoS2 channel, the total density of states (TDOS) and projected density of states (PDOS) analyses were carried out. Figures a,b show that NO2 adsorption leads to p-type doping in MoS2 and substitutional Nb in the Mo position of MoS2 (MoS2-Mo-Subs-Nb) channels. Furthermore, Nb doping in the sulfur vacancy position of MoS2 (MoS2-S-Vac-Nb) does not show p-doping characteristics, which is attributed to the formation of stable chemical bonds. In the case of the MoS2-NO2 structure, the NO2 molecule orbitals hybridize with the d-orbitals of MoS2 close to the Fermi energy (Figure c). In the MoS2-Mo-Subs-Nb-NO2 system, Nb d-orbitals hybridize well with the p- and d-orbitals of MoS2. Additionally, these d-orbitals strongly hybridize with NO2 molecule orbitals around the Fermi energy, as shown in Figure d. Owing to this well-localized hybridization, more scattering occurs in the experimental results.

Figure 8

(a) TDOS plot of MoS2-NO2, Nb-doped in sulfur vacancy MoS2-NO2, and substitutional Nb-doped in Mo atom position of MoS2-NO2 and their PDOS in (b), (c), and (d), respectively.

In addition, oxygen atoms of the NO2 molecule move closer to the sulfur atoms of MoS2 (shown in Figure S3). The charge transfer of 0.009 electrons transfer indicates that scattering is dominant rather than doping. These characteristics lead to high response to NO2 sensing in the Nb-doped MoS2 channel.[48−54]

From the above experimental and theoretical studies, we propose the gas sensing mechanism as follows. The adsorption of NO2 gas molecules on Nb-doped MoS2 prefers to be physisorbed on the Nb-MoS2surface via the two oxygen atoms. These oxygen atoms bound to the metal sites by forming an M–O–N–O four numbered ring structure. The DFT analysis showed higher adsorption energy for Nb-doped MoS2, which suggests that the doping Nb center can drastically enhance the adsorption of gas molecules due to the catalytic activity. In addition to this, the Nb metal center has strong activation energy to attract the NO2 gas molecules. This leads to the enhanced gas response, reproducibility, and stability of the Nb-doped gas sensor. However, in the case of an undoped MoS2 monolayer, the interaction between the NO2 gas molecules and MoS2 is weak with low corresponding adsorption energy, which leads to poor gas response and stability. In addition, in an oxygen environment, a p-type doping effect is observed owing to the extraction of electrons from MoS2. Hence, the absorbed oxygen extracts electrons from the MoS2 and formed O2– ions, as shown in the following equations:[1,55,56]

The electron carrier density of MoS2 decreased due to the depletion of its electrons by atmospheric oxygen. Moreover, the several reactive sites in MoS2 are occupied by oxygen molecules. Upon exposure to NO2 at 100 °C, fewer gas molecules are absorbed on the surface, and there is less electron transfer due to the limited number of active sites. In addition, the interaction between oxygen molecules and the MoS2 layer will be weak due to lower adsorption energy. This lower adsorption energy leads to less adsorption of oxygen molecules.

Conclusion

We showed the effect of the flexible Nb-doped gas-sensing properties of monolayer MoS2 using the PVD method. The NO2 gas sensing response was enhanced dramatically, by five times compared to an undoped MoS2 monolayer, by employing the substitutional doping of Nb. In addition, Nb-doped MoS2 showed excellent stability and reproducibility compared to undoped MoS2. These effects on gas response and stability can be attributed to substitutional cation doping of Nb in the host MoS2, which enhances the number of grains and surface-to-volume-ratio. In addition, the substitutional cation showed physisorption of NO2, which induces faster response and recovery compared to the undoped monolayer MoS2. We hope these results will be useful in understanding the role of Nb dopants (metals) in gas-sensing applications.

Experimental Section

Synthesis of Monolayer MoS2 and Nb-Doped MoS2

Nb-doped MoS2 and MoS2 were grown controllably on 1 cm × 1 cm sapphire (Al2O3) substrates, using a 4-in. pure MoS2 target (MoS2, 99.9%) and Nb-doped MoS2 targets (5%: 95%) by PVD at 700 °C for 30 s. Prior to the growth of monolayer MoS2, the sapphire substrate was cleaned by using a standard chemical process. The substrate was placed precisely in a face-down position in the chamber. Approximately 30 W of RF power was utilized to sputter monolayer MoS2 under an Ar (13 sccm) flow atmosphere, and the working pressure was approximately 1 Pa during the sputtering process. The samples were postannealed at 1000 °C in an H2S (5 sccm)/Ar (80 sccm) atmosphere for 30 min.

Transfer of MoS2 and Nb-Doped MoS2 Thin Film onto a PET Substrate

The MoS2 thin film formed on the sapphire substrate was spin-coated with a poly(methyl methacrylate) (PMMA) solution. The substrate was then dried at 120 °C on the hot plate for 2 min. A few hours later, the substrate was immersed in a buffered oxide etchant (NH4F:HF, 6:1) solution at 80 °C for 1 h. After etching the substrate, MoS2/PMMA was transferred to deionized water several times to remove the buffer solution. Finally, the delaminated MoS2/PMMA was transferred onto a PET substrate. PMMA was removed by immersing the PET substrate in acetone for 5 min.

Characterization and Sensor Fabrication

The thicknesses of the MoS2 and Nb-doped MoS2 monolayer were measured using an AFM system (XE100, Park Systems) in contact mode under a contact force of 30 nN and at a scan rate of 0.5 Hz. Raman analysis was performed using a JASCO NR1800 Raman spectrometer equipped with a Nd:YAG laser. The deposited Au/Ti/Cr electrodes with a length and thickness of 50 μm were deposited by using photolithography and electron beam evaporation. The thicknesses of the coated Au (40 nm), Ti (8 nm), and Cr (2 nm).

Gas Sensing Measurement

Gas sensing tests were conducted using our in-house gas sensing system. Gas sensing characteristics were investigated using nitrogen oxide (NO2), toluene (C6H5CH3), carbon monoxide (CO), and acetone (CH3COCH3), diluted with synthetic air using mass flow controllers. Joule heating was used to control the operating temperature of the gas sensors by using a ceramic heater connected to a power supply and varying the temperature from 50 to 200 °C. The responses of each test were defined as Rg/Ra or Ra/Rg when they were reacted with oxidizing or reducing gases, respectively, where Ra and Rg represent resistance in ambient and analyte gas, respectively. In addition, the response and recovery times for each test were determined by calculating the time for both the ambient and analyte gas atmospheres. The schematic representation of the gas sensing setup is shown in the Figure S4

DFT Simulation

DFT calculations were performed using the Quantum ATK DFT package,[53,54] which is based on a linear combination of numerical atomic orbitals. FHI pseudopotentials with double-ζ double polarized basis sets are employed. To account for the long-range van der Waals interactions more accurately, Grimme DFT-D3 van der Waals corrections were utilized.[54] The revised Perdew–Burke–Ernzerhof exchange–correlation functional was used to obtain a more accurate molecule to MoS2 layer distance and binding energy. A 20 Å vacuum distance was used above and below the MoS2 layer to overcome any spurious interactions with the adjacent supercell. MoS2 supercell dimensions of 31.604 Å × 27.3698 Å × 30 Å were employed in these simulations. A density mesh cutoff of 75 Ha was used.