Development of MoS 2 doping strategy for ehanced SO 2 detection at room temperature

Despite significant e ff orts to limit the use of fossil fuels, SO 2 remains a major air pollutant that adversely a ff ects human health and the environment, especially in heavily industrialized regions of developing countries. Consequently, e ff ective and sustainable methods of SO 2 monitoring remain vital issues for detector development. However, despite decades of progress, current solutions based on resistive sensors remain limited by poor sensing performance of semiconducting materials when operated at room temperature and thus su ff er from high power consumption due to heating. One solution could be to employ novel 2D semiconductor nanomaterials like MoS 2 , which have shown good room-temperature performance for gases such as NO 2 and NH 3 . However, they have also shown limited response to other gases, which on the other hand, can be improved by substitutional doping. Consequently, this work investigates, employing density functional theory, the doping of MoS 2 with Si, P, Cl, Ge, and Se to improve its SO 2 sensing capability. The results show that P doping facilitates all desired e ff ects with the electronic bandgap of MoS 2 preserved at 0.72 nm –2 doping concentration, molecule-sheet charge transfers enhanced by 300%, and moderate binding energies that enable e ff ective surface di ff usion of SO 2 at 300 K.


Introduction
Sulfur dioxide (SO 2 ) is a toxic, colorless, and irritating gas with a strong odor.It is a potent atmospheric pollutant capable of causing severe health and environmental hazards and thus is extensively monitored with ongoing efforts to reduce its emissions.However, despite the efforts, SO 2 pollution remains a persistent issue.On one hand, natural processes like volcanic activity, geothermal springs, organic matter decomposition, and anaerobic bacterial reduction remain constant sources of SO 2 [1,2].On the other, anthropogenic activities such as power generation from fossil fuels, oil refining, and heating via combustion of coal and natural gas remain the major contributors to the total emissions [3][4][5][6].Typical concentrations of SO 2 in the atmosphere range from tens to hundreds of ppt in uninhabited regions, through tens of ppb in cities, up to hundreds of ppb in industrial areas [7,8].SO 2 released into the atmosphere is solubilized in water and slowly oxidized by oxygen, ozone, or Criegee intermediates to SO 3 , which subsequently leads to the formation of sulfuric acid [9,10].Consequently, precipitations in the form of rain, snow, fog, or hail containing sulfuric acid fall to the ground, exerting a harmful effect on the environment and infrastructure [11][12][13].SO 2 absorbed through the respiratory system or skin can be a threat to human health or even life.Air contamination with this gas is one of the causes of many respiratory (laryngitis, chronic bronchitis, respiratory tract infections), and cardiovascular diseases [14,15].Exposure of a human organism to a concentration of SO 2 equal to 100 ppm is related to a direct threat to life [16][17][18].The maximum exposure can not exceed 175 ppb per 10 minutes and 2 ppm per 10-hour time period according to air quality standards [17].The harmful effect of SO 2 on human health and the environment combined with its high emissions makes its rapid detection one of the challenges of modern science.In particular, novel materials, including nanomaterials, with the potential to improve SO 2 detection have been in the area of interest of researchers in recent years.
The successful exfoliation of graphene from graphite and the subsequent discovery of its unique chemical, physical, and electronic properties have attracted much attention to 2D-layered materials [19,20].One of the particularly promising areas of their application is gas sensing due to their atomic thin-layered structure and large surface-to-volume ratio.Although graphene shows excellent gas-sensing capabilities, its application in gas sensors is limited by its zero-band gap resulting in poor switching characteristics [21].However, other 2D-layered materials, such as transition-metal-dichalcogenide (TMD) sheets, possess a sizeable band gap and thus are free of this inherent limitation of graphene [19][20][21].TMD gas-sensing platforms appear to be an attractive alternative to devices based on metal oxide semiconductors (MOS) that have been widely investigated in recent years in the context of SO 2 detection [22][23][24][25][26][27].The latter are important low-cost and highsensitivity environmental gas detectors, but are known for problems with detection at room temperature and relatively high power consumption.Simultaneously, a growing number of promising reports on effective detection at room temperature using TMD-based sensors has recently emerged in the literature [28][29][30][31].Individual TMD sheets have a three-layer structure reflected in their chemical formula, which is MY 2 .A single sheet consists of two chalcogen layers (Y) and a transitionmetal layer (M) sandwiched between them.Similarly to graphene, TMD sheets interact in bulk crystals via weak van der Waals (vdW) forces and thus can be effectively exfoliated.Atoms within individual sheets are covalently bonded and have satisfied valence.This makes TMD surfaces chemically inert under most conditions, which limits their surface interactions to physisorption.

Computational Details
The doping of the MoS 2 sheet with the selected p-block elements and SO 2 adsorption on its surface were studied on the DFT level of theory and the plane-wave/pseudopotential (PW/PP) formalism implemented in the Quantum ESPRESSO code [62][63][64].All calculations were performed applying the Rappe-Rabe-Kaxiras-Joannopoulos pseudopotentials with included scalar-relativistic and nonlinear core corrections [65].Electron exchange and correlation were addressed with the Perdew-Burke-Ernzerhof (PBE) parameterization adopted within the generalized gradient approximation (GGA) [66,67].When it comes to vdW forces, the DFT-D3 method developed by Grimmy was used [68].The set cut-off energies of the wave function and the electron density had values of 75 and 600 Ry, respectively.A Monkhorst-Pack (MP) [69] k-point grid of 10 × 10 × 1 and a Gaussian smearing of 0.001 Ry were selected for integration of a Brillouin zone surface.
The pristine and doped MoS 2 sheets were modeled using a 2D periodic slab.The test calculations were performed on 3 × 3, 4 × 4 and 5 × 5 unit cells.The doping and adsorption energies calculated employing 4 × 4 and 5 × 5 structures were almost identical.The computations carried out with the use of a 3 × 3 structure gave notably different results.Therefore, a 4 × 4 unit cell was chosen for further calculations.The adopted cell height of 25 Å combined with the truncation of the Coulomb forces in the z direction allowed to avoid interactions between neighboring units.Relaxation of atomic positions of all atoms in the system was achieved as a result of the total energy optimization with the convergence treshold on forces of < 10 −3 Ry/au.The XCrySDen software was employed to visualize modeled atomic structures [70].
The SO 2 desorption process from the doped MoS 2 surface was investigated using ab initio molecular dynamics (AIMD).The AIMD calculations were performed based on the Born-Oppenheimer theory at a temperature of 300 K with a time step of 20 au (0.9676 fs).The motion equations of the SO 2 molecule atoms were integrated by employing the Verlet approximation to project its trajectory in time after desorption.Almost all the computational parameters adopted in the other calculations performed for this work were preserved in the AIMD runs.An exception constituted the MP grid reduced to 4 × 4 × 1.

Doping of the MoS 2 sheet
The doping of the MoS 2 sheet (Figure 1a) was carried out in a twostep process.In the first step, one S atom per supercell was removed from the upper layer of the sheet (Figure 1b), while in the next one, a dopant atom (X) was adsorbed into the created S-vacancy (Figure 1c).The used doping model results in a doping concentration of ≈ 2% and can be successfully implemented in an experiment, as the studies of Komasa et al. allow us to anticipate [71,72].The authors have shown that single-atom vacancies can be formed exclusively in the S-sublattice of MoS 2 employing a transmission electron microscope (TEM).This possibility is a result of a significant difference in the threshold energies required for the displacement of the S and Mo atoms from MoS 2 .An electron beam with a minimal energy of 80 and 560 keV is able to produce vacancies in the S-and Mo-sublattices, respectively [71,72].Hence, its energy can be adjusted to be simultaneously high enough to remove S atoms and too low for the formation of the Mo-vacancies.Komasa et al. have also demonstrated the filling of the created vacancies in the S-sublattice of MoS 2 with impurity atoms introduced into a TEM chamber.
This work investigates doping of MoS 2 with the p-block elements for enhanced detection of SO 2 contrary to the majority of research focused on the d-block elements [52][53][54][55][56][57][58][59].The reason behind it is the typically lower adsorption energy of small molecules on p-doped MoS 2 compared to a d-doped sheet [45,47,73].Se, P, Cl, Ge, and Se atoms used in the current studies as doping agents were selected from the different groups and periods of the p-block.Such an approach was expected to impose differences in the binding of dopant atoms as a consequence of the dissimilarity of their electron structures and covalent radii.The parameters describing their adsorption process into the Svacancies of the MoS 2 sheet are compared in Figure 2. All dopants except Si favor positions above the S-layer, which is reflected in the positive values of ∆h X-Mo (Figure 2a) defined as follows: where h(X) and h(S) are the vertical positions of the X and Mo atoms, respectively.∆h(X-S) was shown to increase in the order of Si < P < Cl < Se < Ge.Thus, two dopants with the largest covalent radii-Se and Ge-relaxed at the highest position above the S-sublattice.The notable variation in the values of ∆h(X-Mo) suggests differences in X-Mo bonding configuration.This is further supported by the values of X-Mo binding energy, determined using the formula.
The symbols E(X-MoS 2 ), E(vac-MoS 2 ), and E(X) in equation 2 correspond to the total energies of X-doped MoS 2 , MoS 2 with the S-vacancy, and a free atom of X, respectively.The values of E b for all the dopants are negative (see Figure 2b), which means that the doping of MoS 2 is an exothermic process.More negative E b is associated with greater stability of the X-doped system and stronger bonds between the X and Mo atoms.Here, P-and Se-doping are shown to facilitate the strongest X-Mo interactions (≈-7 eV).The remaining elements are characterized by significantly weaker binding, suggesting differences in the bonding.Still, the E b values in all cases are below -4 eV, which indicates the formation of strong chemical bonds between all dopants and Mo atoms after the adsorption of the former into the S-vacancies of MoS 2 .
Given the covalent character of the X-Mo interactions, the bonding symmetry will cause orbitals of X to adopt sp λ hybridization (2 < λ ≤ 3), which will demand changes in its orbital population.This, in turn, gives rise to charge transfer between a dopant atom and an S-vacant sheet.The Löwdin population analysis was employed to determine the per orbital transfers (δQ(X)) in X according to equation: where Q(X-MoS 2 ) and Q(X) are the charges on the individual valence subshells of the X atoms adsorbed in the S-vacancies of MoS 2 , and the free X atoms, respectively.The results are presented in Figure 2c.
They show that all dopants except Si gain electronic charge as a result of doping.In all cases, the depopulation of the s-orbitals occurs, which is in line with sp-type hybridization.However, the differences between dopants arise from changes in the population of their p-subshells.Si loses charge in its p z and accumulates it in p x -and p y .It also, as the only element among the investigated doping agents, relaxes below the Ssublattice (0 > ∆h(X-S)).For these reasons, a strong in-plane character of the Si-Mo interactions is anticipated.P, Cl, and Se gain, in turn, charge in p x -and p y as well as in p z -orbitals.This is more characteristic of an sp 3 hybridization, and thus a more out-of-plane X-Mo interaction (also indicated by the positive values of ∆h(X-Mo)).Finally, Ge loses charge in p x -and p y and gains it in p z .This coincides with the largest ∆h(X-S) determined for Ge among the X-atoms.Thus, Ge favors the interactions with Mo with the strongest out-of-plane character.

Adsorption of SO 2 on pristine and doped MoS 2 sheet
A gas-sensing mechanism of 2D TMDs is based on charge transfer between analyte molecules and a sheet, which modifies its electrical resistance proportionally to the concentration of a monitored gas.Therefore, to improve the SO 2 sensing capability of MoS 2 , first of all, Xdoping should increase SO 2 -doped sheet charge transfer.Effective gas detection at room temperature can be achieved, in turn, only when the enhanced charge transfer is not accompanied by a significant decrease in the adsorption energy.Hence, charge transfer and adsorption energy are two of the most important parameters that allow for a quantitative evaluation of the X-doping effect on MoS 2 sensitivity towards SO 2 .Information on charge transfer to/from X-MoS 2 upon SO 2 adsorption can be ascertained from the total change in the electronic population of a gas molecule given as follows: Q tot (SO 2 @sub) and Q tot (SO 2 ) in equation 4 are the total pseudocharges of the adsorbed and isolated gas molecule, respectively.The adsorption energy is defined as: where E(SO 2 @sub), E(SO 2 ), and E(sub) are the total energies of the adsorbate-substrate system, the free SO 2 molecule, and the isolated substrate, i.e. the MoS 2 sheet, respectively.Small molecules typically only undergo physisorption when interacting with pristine MoS 2 , and any doping of the material should maintain this behavior in order to allow for effective room temperature detection.The dispersive nature of the vdW forces results in the lack of one well-defined adsorption site.In such a case, studies of numerous initial adsorption configurations are required.Finding the one with the lowest energy is challenging due to the existence of multiple local minima with total energies similar to the global one.Most adsorption configurations result in similar E ads and charge transfers.However, it is still important from the point of view of sensor characteristics to consider whether any configurations are leading to large E ads , or whether charge transfer depends significantly on the initial geometry of an adsorbate molecule.The former would lower the sensor recovery rate, while the latter would cause that only part of the gas molecules would give rise to the sensor response.
This study explores how doping with X affects the ability of MoS 2 to detect SO 2 .Four initial geometries were selected to probe the adsorption of SO 2 molecule near a doping site (Figure 3).The relaxation process began with the SO 2 positioned either vertically or horizontally with either its sulfur and oxygen atom above the sulfur or X atom of the pristine or doped MoS 2 , respectively.The initial vertical adsorption configurations are shown in Figure 3a (vert-S) and Figure 3b (vert-O), while the horizontal ones are presented in Figure 3c (horiz-S) and Figure 3d (horiz-O).The values of E ads and δQ tot (SO 2 ) are given in Table 1.The optimized structures with the lowest E ads values for the adsorption of SO 2 onto, both, pristine and X-doped MoS 2 are shown in Figure 4. Figure 5 provides, in turn, the maximum and average δQ tot (SO 2 ) values.It is important to note that the actual charge transfer per molecule will be lesser than the maximum value of δQ tot (SO 2 ) since SO 2 molecules will not be limited solely to one adsorption configuration.At room temperature, a variety of surface sites and adsorption geometries may be available to SO 2 .On the other hand, the actual charge transfer will also be greater than the average value of Q tot (SO 2 ) due to the predominant contribution of molecules in the lowest-energy adsorption configuration, which exhibits a stronger binding and thus is statistically longer-lasting.Although the exact charge transfer falls between the maximum and average Q tot (SO 2 ), the simultaneous analysis of these two parameters can provide reliable insight into its magnitude.The average Q tot (SO 2 ) was calculated as an arithmetic mean of charge transfers determined for all adsorption configurations considered.
The adsorption of SO 2 on pristine MoS 2 is shown to have energies ranging from -124 meV to -223 meV depending on the initial adsorption configuration of the gas molecule.Table 1 shows that the adsorption of SO 2 initially located above the sheet in the horiz-S configuration  5).However, the transfer is relatively limited (Table 1), revealing a weak response of the pristine MoS 2 toward SO 2 .The X-doping of MoS 2 impacts the adsorption process of SO 2 and modifies the sheet sensitivity towards its molecules.The strongest adsorption of the analyte is promoted by Si and Ge-doped MoS 2 .E ads of SO 2 on the Si-MoS 2 is between -2382 meV and -1381 meV, while on Ge-MoS 2 is between -883 meV and -313 meV.Thus, upon the sheet doping with Si and Ge, E ads lowers significantly with respect to its values calculated for the pristine MoS 2 (see Table 1).The primary reason for the enhanced gas adsorption on sheets doped with Si and Ge appears to be the valence configuration of these two elements.As members of the 14th group of the periodic table, Si and Ge favor tetrahedral bonding configuration via sp 3 hybridization.However, upon adsorption into S-vacancy of MoS 2 , the atoms bond with only three Mo atoms.Hence, to satisfy their valence, they require a fourth bond, which can be formed upon the adsorption of SO 2 .In both cases, the lowest-energy configuration of SO 2 is achieved for the horiz-O optimization, indicating that Si-O and Ge-O are the favorable interactions.Other configurations facilitate significantly weaker binding.Thus, at room temperature, the transition into the lowest-energy configuration should dominate over surface diffusion.Figures 4b and Figures  Despite all the similarities, the differences in the adsorption energy between Si and Ge-doped MoS 2 are quite notable (see Table 1).The effect can be attributed to differences in the Si-Mo and Ge-Mo bonding geometry.As discussed in Section 3.1, Si favors a horizontal interaction with Mo, while Ge adopts one with a greater vertical component.This results in the p z orbital electrons of Si being less involved in the Si-Mo bonding and thus more readily available in the interaction with SO 2 , compared to Ge.Nevertheless, both dopants facilitate the chemisorption of SO 2 with almost 11 (Si-MoS 2 ) and 4 (Ge-MoS 2 ) times stronger binding compared to pristine MoS 2 .Such an interaction could hinder the recovery of a sensor after evacuating the analyte, potentially impeding its response to rapid changes in concentration.Further issues arise from the affected charge transfers.Due to the stronger molecule-sheet interactions, the absolute value of maximum δQ tot (SO 2 ) increases to 0.258 e for Si-MoS 2 and 0.174 e for Ge-MoS 2 .As such, the response of the Si and Ge doped sheets to the gas evacuation would likely be obscured by the large charge transfers resulting from the molecules still bonded to the dopant unless the sensor is heated to achieve enhanced desorption.However, even if the latter is implemented, it will introduce further challenges.The extent of charge transfer between SO 2 and the sheets varies across different configurations, as indicated in Table 1.Specifically, for the Si-and Ge-doped MoS 2 , the average value of charge transfer is lower than that of the pristine MoS 2 , as demonstrated in Figure 5. Thus, regardless of the approach, the benefits of the enhanced charge transfer would be limited.
In contrast, the Se-doped MoS 2 promotes the weakest adsorption of SO 2 among investigated sheets, resulting in the values of E ads ranging between -183 meV and -98 meV.The calculated E ads values indicate that the SO 2 binding in the vicinity of the Se site is weaker than on the pristine surface of MoS 2 .Such a weak interaction is also in line with the optimized geometries because the adsorption has visually no impact on the molecule (Figure 4).Furthermore, SO 2 in the lowest-energy configuration (horiz-O) relaxes at a larger distance from the Se-MoS 2 (3.62 Å) compared to the pristine sheet (3.07 Å).This, in turn, is shown to reduce the values of both average and maximum charge transfers (see Figure 5).Consequently, Se doping of MoS 2 does not enhance the sensing properties of the sheet towards SO 2 , as the adsorption of SO 2 on Se-MoS 2 leads to weak physisorption.
On the other hand, P and Cl doping results in moderately strengthened molecule-sheet interactions.The adsorption of SO 2 on P-MoS 2 is associated with E ads values ranging from -514 meV to -136 meV, while Cl-doping results in E ads between -362 meV and -246 meV.Thus, the SO 2 binding on P-and Cl-doped sheets is stronger than that on pristine MoS 2 , but significantly weaker than that on Si-or Ge-doped MoS 2 .As a result, even in the lowest-energy adsorption configuration, the analyte molecule relaxes relatively far from the P-and Cl-doped sheets with little impact on its structure (see Figures 4c and 4d).However, the molecule-sheet interaction is still notably enhanced, which promotes the increase in maximum and average δQ tot (SO 2 ) (Figure 5).In the case of P-MoS 2 , the doping results in a 300% (230%) increase in the average (maximum) δQ tot (SO 2 ).On the other hand, Cl doping leads to an average (maximum) increase of 460% (250%).Thus, P and Cl doping of MoS 2 could improve its SO 2 detection capability.In order to better understand the underpinnings of the enhanced charge transfers, Figure 6 illustrates the changes in the total electron density of the lowest energy configurations of the SO 2 @P-MoS 2 and SO 2 @Cl-MoS 2 .The differences are given by: where n tot is the total density of pseudoelectrons.Upon inspection, it becomes evident that, despite both systems showing similar transfer values, the interactions between the gas molecules and the sheets doped with P and Cl lead to considerably different charge distributions.The main difference is that the adsorption of SO 2 on P-MoS 2 results in more pronounced accumulation and depopulation regions compared to Cl-MoS 2 .In SO 2 @P-MoS 2 , the charge accumulation between the S atom of the gas molecule and the P atom of the sheet is facilitated, but the region is separated from the analyte molecule and the P-MoS 2 by the two electron-depletion areas.An increase in the total electron density is also present in the vicinity of the O atoms of the gas molecule, as well as in the sheet near the P atom and the neighboring Mo atoms.Conversely, the depletion regions are localized in the P-MoS 2 sheet on the S atoms closest to the gas molecule and the core region of the SO 2 between its S and O atoms.The latter suggests weaker S-O bonds, which is in line with the 0.017 Å elongation.This could be attributed to the accumulation of electronic charge causing electrostatic repulsion between the S (-0.04 e) and O (-0.06 e) atoms of the gas molecule.
In contrast, the SO 2 adsorption on the Cl-MoS 2 facilitates an increase in total electron density mainly on its S and O atoms.Changes in the total electron density of the sheet are much more delocalized and thus less visible, with only regions near the Mo atoms neighboring with Cl showing some localized changes.Still, in both cases, a noticeable accumulation of charge between the S atom of the SO 2 molecule and the P or Cl atom of the sheets may, in part, contribute to enhanced interaction and charge transfer between the analyte and the P-or Cldoped MoS 2 .In both cases, changes in the total charge density are limited to the gas molecule, the atoms of the dopants, and other atoms of the sheets in the close vicinity of P or Cl, making charge-transfer enhancement a relatively local effect.Doping of MoS 2 with P and Cl can have an impact on its band structure and thus also on a sensing mechanism of a detector based on it.An interaction between SO 2 and the doped sheets resulting in charge transfer upon gas adsorption can also influence their electronic properties.For these reasons, in Figure 7, the contours of the total densities of states (DOS) of pristine (a, b) and doped with P (c, d) and Cl (e, f) MoS 2 before (left column) and after gas adsorption (right column) are compared.The grey ones represent the total DOS of the sheets, whereas the green and red ones correspond to the partial DOS of the s and p orbitals of the dopants.Figures 7b, d, and f present the DOS contours after the adsorption of SO 2 in the lowest-energy configuration.Contributions of S and O orbitals of the gas molecule are not included because of their low density in the vicinity of the Fermi level.Figure 7a shows that the pristine MoS 2 sheet is a semiconductor.Its doping with P (15 group of the periodic table) having a valence electron less than S (16 group) shifts down the Fermi level (see Figures 7a and  c).The sheet becomes a p-type semiconducting nanomaterial as a result.In contrast, MoS 2 doping with Cl (group 17) with a higher number of valence electrons compared to S moves up its Fermi level and causes a complete reduction in its band gap (see Figures 7a and e).This indicates that the Cl-doping of the sheet alters its electronic properties and makes it a metallic system.The closing of the MoS 2 band gap is a negative effect in terms of its usage as a resistive sensor because it leads to reduction of changes in its resistance upon exposure to gases, which in turn lowers its sensitivity toward them.Hence, although Cl-doping facilitates the excellent charge transfer and weak binding of the analyte, it may be less suitable for SO 2 sensing than P-doping as it would demand a lower doping concentration of the sheet to retain its electronic band gap.The SO 2 adsorption has limited impact on the position of the Fermi levels and band gaps of P-and Cl-MoS 2 similarly as in the case of the pristine MoS 2 .On the other hand, Figures 7c-f show that the gas adsorption results in the reduced contribution of the p orbitals of both dopants in the vicinity of the Fermi levels to the total DOS.This is in line with the enhanced electron transfer to the gas molecule upon adsorption (see Figure 5).The research performed in this work suggests that P-doping of MoS 2 holds promise as the most effective strategy for improving SO 2 detection.However, it is important to note that the observed effects are due to increased interaction between molecules and the sheet, which may impede gas desorption at room temperature.The decreased effectiveness of desorption would result, in turn, in a reduced sensor recovery rate.Therefore, to gain further insight into this matter, we have utilized AIMD simulations at a temperature of 300 K.The simulations have encompassed three AIMD runs.In each of the simulations, the gas molecule was initially adsorbed at the P-doping site.Figures 8a  and b show that the P-SO 2 in-plane distance increased gradually from 3 Å to 20 Å after 5.5 ps during one of the runs, while the height of the molecule fluctuated between 2 Å and 4 Å.The remaining AIMD runs have shown similar SO 2 behavior, illustrating its effective surface diffusion despite the enhanced adsorption energy.Consequently, P-doping appears to have a limited influence on the sensor recovery rate and should not impede its fast responses to rapid changes in gas concentration.Finally, Figure 8c illustrates the deviation in total energy (E − E avg ) of the SO 2 @P-MoS 2 system during one of the AIMD simulations.The calculated value of the mean total energy (E avg ) was equal to -796.159 eV/atom.The standard deviation was found, in turn, to be 4.589 × 10 −4 eV/atom.The total energy experienced a drift of 1.27 × 10 −5 eV/atom-fs.The rest of the performed simulations resulted in qualitatively identical fluctuations in the total energy.

Conclusions
In this study, the DFT computations were utilized to explore the potential of Si-, P-, Cl-, Ge-, and Se-doping for enhancing the SO 2 detection capability of MoS 2 at room temperature.To be effective as a resistive sensor, doped MoS 2 should exhibit improved charge transfer upon analyte adsorption without a significant impact on its adsorption energy.However, our findings indicate that Si-, Ge-, and Se-doping results in a decreased average charge transfer.Moreover, Si-and Gedoping facilitates strong gas binding that could impede sensor recovery.On the other hand, Cl-doping shows promise with large values of charge transfer and moderately increased molecule binding but may lead to loss of the semiconducting properties by MoS 2 and its metallic behavior at Cl doping concentrations ≥0.72 nm -2 .P, however, stands out as a suitable dopant with an almost 300% and 230% increase in average and maximum charge transfer, respectively, with the semiconductor bandgap of MoS 2 preserved at the same doping concentration as in the case of Cl.Furthermore, AIMD simulations suggest that Pdoping would not significantly impact the sensor recovery rate, making it a highly attractive candidate for enhancing MoS 2 sensitivity towards SO 2 at room temperature.

Figure 1 :Figure 2 :
Figure 1: Schematics of atomic structures of a pristine MoS 2 sheet (a), a MoS 2 sheet with S-vacancy (b), and a doped MoS 2 sheet.
4e show the favorable configurations of SO 2 @Si-MoS 2 and SO 2 @Ge-MoS 2 , respectively.The post-adsorption distance of SO 2 from the doped sheets is ≈ 1.71 Å (Si-MoS 2 ) and ≈ 1.91 Å (Ge-MoS 2 ).The values are 1.8 and 1.6 times lower compared to the pristine MoS 2 .Additionally, in both cases, the adsorption results in an elongation of the S-O bond closest to the sheet (≈ 1.6 Å), along with a decrease in the O-S-O angle of the SO 2 molecule (≈ 110°).

Figure 5 :
Figure 5: Average and maximum changes in the total charge of the SO 2 molecule upon its adsorption on pristine and X-doped MoS 2 .

Figure 6 :
Figure 6: Changes in total electron density upon SO 2 adsorption on P-MoS 2 (a, c) and Cl-MoS 2 (b, d) calculated according to the formula 6. Areas in red and blue indicate, respectively, regions of electron accumulation and depletion.The isovalue cutoff is 1 × 10 -3 electrons Bohr -3 .

Figure 7 :
Figure 7: DOS contours before (left column) and after (right column) SO 2 adsorption on pristine MoS 2 (a,b), P-MoS 2 (c,d), and Cl-MoS 2 (e,f).Grey contours correspond to the total DOS.Green and red colors represent the partial DOS of the s and p orbitals of a dopant.The Fermi energy is denoted as E F .

Figure 8 :
Figure 8: Results of AIMD simulations performed for SO 2 adsorbed at the doping site of P-MoS 2 : (a) projection of gas molecule trajectory, (b) time dependencies of P-SO 2 distance and SO 2 height above the sheet, (c) variation of total energy during simulation.

Table 1 :
Adsorption parameters of SO 2 on pristine and X-doped MoS 2 .
Init. conf.E ads [-meV] δQ tot (SO 2 ) [-e] • , respectively).The relatively weak binding and the large molecule-sheet distance without significant changes in the geometry of SO 2 indicate its physisorption on the undoped sheet, even in the case of the lowest-energy configuration.The calculated values of maximum and average δQ tot (SO 2 ) show the electron accumulation in the molecule (see Figure