Ultrahigh sensitivity with excellent recovery time for NH3 and NO2 in pristine and defect mediated Janus WSSe monolayers

Rajneesh Chaurasiya and Ambesh Dixit *
Department of Physics and Center for Solar Energy, Indian Institute of Technology, Jodhpur, 342037, India. E-mail: ambesh@iitj.ac.in

Received 16th April 2020 , Accepted 26th May 2020

First published on 26th May 2020


We demonstrated ultrahigh sensitivity with excellent recovery time for H2S, NH3, NO2, and NO molecules on the sulfur and selenium surfaces of Janus WSSe monolayers using density functional theory. The selenium surface of the WSSe monolayer showed strong adsorption in comparison to the sulfur surface. The respective adsorption energies for H2S, NH3, NO2 and NO molecules are −0.193 eV, −0.220 eV, −0.276 eV, and −0.189 eV. These values are higher than the experimentally reported values for ultrahigh sensitivity gas sensors based on MoS2, MoSe2, WS2, and WSe2 monolayers. The computed adsorption energy and recovery time suggest that the desorption of gas molecules can be achieved easily in the WSSe monolayer. Further, the probable vacancy defects SV, SeV, and (S/Se)V and antisite defects SSe, and SeS are considered to understand their impact on the adsorption properties with respect to the pristine WSSe monolayer. We observed that the defect-including WSSe monolayers showed enhanced adsorption energy with fast recovery, which makes the Janus WSSe monolayer an excellent material for nanoscale gas sensors with ultrahigh sensitivity and excellent recovery time.


1. Introduction

The detection and mitigation of toxic gases are vital because of their adverse impact on the environment as well as human health, causing medical issues at mass scales leading to public health and safety concerns. NH3 and NO2 gases are contributing significantly to the pollution of the environment. The sources of NH3 pollutants are emission sources from the chemical industry, automobile exhaust, and decomposition of fertilizers used in agriculture.1 NO2 sources are fossil fuel consumption, automobile exhaust, smoke, and kerosene heaters. These are contaminating not only the environment but also affect the health of living beings. The detection of these polluting gases has become essential for monitoring the environment so that crucial measures can be considered in advance or at least on time. A variety of structures like zero, one, and two dimensional, and even bulk materials, have been investigated intensively for gas sensing.2–4

Metal oxide semiconductors like ZnO and SnO2 materials are the most widely used gas sensing materials and have also been commercialized for various gas sensor applications.5 These metal oxide semiconductor based gas sensors exhibit multiple advantages such as high sensitivity and relatively low fabrication cost, and resist environmental degradation. However, they also suffer from several drawbacks such as the requirement for higher operating temperatures, large power consumption, and relatively poor selectivity.6 Such problems are mitigated to some extent using carbon nanostructures, e.g., nanotubes, yet, the enormous recovery time and complex handling of such materials has impacted their use at large scales.7–9 In the family of carbon, the one-atom thin monolayer with high surface to volume ratio, i.e., graphene, showed relatively high sensitivity, selectivity, and low power consumption.10,11 Here, the low electrical noise provides a massive signal to noise ratio even for a tiny fraction of gases, showing very high sensitivity.11

Other two dimensional (2D) materials such as MXenes, phosphorene, and boron nitride are also attracting attention for their potential in various nanodevices.12 2D material based gas sensors work on the basis of charge transfer between the adsorbate and adsorbent, causing significant changes in electrical response.13 Moreover, the semiconducting behavior of these monolayers with respect to their counterpart graphene leads to a relatively large variation in transport properties, thus improving the sensing performance.12,14 TMD based monolayers are the most studied systems in the family of 2D materials after graphene, for gas sensing application.15 Further, 2D TMD based gas sensors are very sensitive to NO, NO2 and NH3 gases and functionalized monolayers have shown improved sensitivity for other gases as well.16–27

Recently, a new type of TMD, the Janus MoSSe monolayer, has been experimentally synthesized using chemical vapor deposition based on the bottom-up approach.28,29 It is a derivative of MoS2 and MoSe2 monolayers, where one side of chalcogen atoms is entirely replaced by different chalcogen atoms. The confirmation of the Janus MoSSe monolayer is investigated using Raman spectroscopy and scanning tunneling electron microscopy measurements.29 It showed exceptional physicochemical properties with respect to the parent TMD monolayers, and thus attracted wider attention.30–41 Further, Janus monolayers are explored for numerous other applications such as hydrogen storage,42 water splitting,35 solar cells,34,43 Li-ion batteries,39 gas sensors,32 thermoelectrics,36 piezoelectrics,44 field-effect transistors,45 the hydrogen evolution reaction,31 photocatalysis,46etc. The experimental realization of defect-free Janus monolayers may not be possible and will rely on the growth process used for synthesizing such layers. The synthesis of TMD based Janus monolayers is carried out using chemical vapor deposition, which is prone to various defects.47,48 These defects showed an improvement in gas sensing properties over the conventional TMDs.49 Firstly, Jin et al.32 explored the gas sensing application of the Janus MoSSe monolayer. Moreover, the external strain improved the internal electric field, which enhanced the sensitivity and selectivity. Our group also investigated the gas sensing properties of pristine and defect-including Janus MoSSe monolayers.50 The chalcogen defects also resulted in improved gas sensing properties. Further, the intrinsic electric field because of crystallographic asymmetry in Janus MoSSe monolayers improves the gas sensing properties with respect to the parent monolayers. However, the gas sensing properties of the Janus WSSe monolayer are still unexplored.

In this article, we considered pristine and defect-including WSSe monolayers as hosts for gas sensing applications. The defects may provide active sites to adsorb the toxic gas molecules. Janus WSSe monolayers have sulfur and selenium ending surfaces, where the adsorption of gas molecules may take place. The different orientation of toxic gas molecules is also investigated on these surfaces to understand the probable preferential adsorption sites for different gas molecules. The adsorption on the surface of pristine and defect-including monolayers is analyzed by computing the adsorption energy, charge density difference (CDD), adsorption length, recovery time, and Bader charge analysis.

2. Computational details

The adsorption calculations for toxic gases are performed using density functional theory as implemented in Quantum ESPRESSO (QE).51,52 A 4 × 4 supercell along the X and Y directions with a 15 Å vacuum along the Z direction is used for minimizing the interlayer interactions. Monolayer and gas molecules are optimized without any constraints considering total energy and force minimization criteria.53 The spin-polarized calculations are carried out using ultrasoft pseudopotentials under the generalized gradient approximation together with the Perdew–Burke–Ernzerhof (GGA-PBE) approximation.54 The energy cutoff values are 50 Ry and 400 Ry for the wave functions and charge density, respectively. We used 6 × 6 × 1 K-points for the sampling of the Brillouin zone (BZ) as per the Monkhorst Pack scheme for the optimization of these structures.55 Further, the total energy and force are constrained at 1 × 10−4 Ry and 1 × 10−3 a.u., respectively, between two consecutive self-consistent field (SCF) cycles for optimization. We also carried out Bader charge analysis to investigate the charge transfer between the monolayer and molecules.56 Further, 6 × 6 × 1 K-points with tetrahedral occupation are used for computing the density of states.57 The DFT-D2 approximation, as implemented in QE, is used to compute accurate energy values, where vdW interactions are included by adding the semi-empirical damped dispersion term into the Kohn–Sham energy.58 The accuracy of the DFT-D method depends on the damped dispersion term. The total energy of the unit cell after DFT-D2 correction can be expressed as59

E DFT-D2 = EKS-DFT + E(2)disp, where E(2)disp can be written as

image file: d0cp02063j-t1.tif
Here Nat is the total number of atoms, Cij6 is the dispersion coefficient, Rij is the bond distance between the atom pair i and j, and S6 is a scaling factor, dependent on the XC-functional. A damping factor fdmp is used to prevent singularities at small distances. Zhao et al.60 predicted the gas sensing properties of CO, CO2, NH3, NO, NO2, CH4, H2O, N2, O2 and SO2 molecules on the MoS2 monolayer using the DFT-D2 and DFT-D3 methods and noticed a similar trend with a small change (<0.05 eV) in the adsorption energy. Zhang et al.61 investigated the catalytic properties of gold nanoclusters on the MoS2 monolayer by considering DFT-D3 and DFT-D2 and also reported a similar trend with a small change in the adsorption energy. The adsorption energy computed using the DFT-D3 method is a bit larger than the DFT-D2 method. However, most of the reports on TMD monolayers are based on the DFT-D2 method for the adsorption of gas molecules. That's why, for relative comparison, we used the DFT-D2 method to involve the vdW interactions.

After optimization of the structural parameters, the stability of all the considered points defects is evaluated using the formation energy as62

Formation energy of a vacancy defect (Evacf) = EdeftotEpristot + EXremoved

Formation energy of an antisite defect (Eantif) = EdeftotEpristot + EXremovedEXadded
where Edeftot is the total energy of the defective monolayer, Epristot is the total energy of the pristine monolayer, EXremoved is the total energy of the removed atom and EYadded is the total energy of the added atom. Further, the magnetic behavior of the defect-including monolayer can be predicted using the polarization energy calculated using the expression
ΔE = Enon-spinEspin
where Enon-spin and Espin are the total energy of the defect-including monolayer without and with spin polarization, respectively. ΔE = 0, confirming non-magnetic behavior.

The adsorption behavior of different gas molecules is investigated using the adsorption energy, which defines the stability of gas molecules on the host surface. The adsorption energy is expressed as63

Eads = Emonolayer+gas − (Emonolayer + Egas)
where Emonolayer+gas and Emonolayer are the total energy of the monolayer with adsorbed gas molecules and the monolayer without adsorbed gas molecules, respectively. Egas is the total energy of the gas molecule. Adsorption of gas molecules on 2D materials usually relies on the charge transfer mechanism, which can be analyzed in terms of the CDD and Bader charge analysis. The CDD is computed as64 Δρ = ρmonolayer+gas − (ρmonolayer + ρgas), where ρmonolayer+gas, ρmonolayer, and ρgas are the charge density of the gas molecule-including monolayer, monolayer and gas molecule, respectively.

Additionally, the gas sensing can be quantified by noticing the change in electrical conductivity of the monolayer after the adsorption of gas molecules. Here, the change in electrical conductivity is attributed to the charge transfer between the adsorbate and host materials. The change in the electrical conductivity can be estimated as65image file: d0cp02063j-t2.tif, where Eg is the band gap of the host material. kB and T are the Boltzmann constant and temperature, respectively. Further, the recovery time is also an important parameter for gas sensing application, and is defined as66image file: d0cp02063j-t3.tif. Here, ν, Ead, kB and T are the attempt frequency for bond breaking, the adsorption energy, Boltzmann's constant and temperature, respectively. Peng et al.67 predicted a 1 THz attempt frequency at room temperature for NO2 gas molecules adsorbed on a carbon nanotube.

3. Results and discussion

The optimization of the structural parameters and electronic properties of the pristine Janus WSSe monolayer are reported in our previous work.68 In brief, the optimized lattice constants are a = b = 3.25 Å, together with bond lengths W–S (2.43 Å) and W–Se (2.55 Å).68 The Janus WSSe monolayer consists of tungsten, which is covalently bonded with the sulfur and selenium atoms. Here, tungsten exhibits six-fold coordination, whereas chalcogen elements S and Se are in three-fold coordination.32 This WSSe monolayer breaks the out of plane symmetry and possesses C3v point group symmetry. The pristine WSSe monolayer exhibits non-magnetic semiconducting behavior with a 1.71 eV band gap.68

TMD based monolayers are defect prone; defects arise during synthesis because of their low formation energy. However, the growth of the WSSe monolayer may also create antisite substitution, causing antisite defects. For example, the MoSSe monolayer is synthesized by the sulfurization/selenization of MoSe2/MoS2 monolayers grown using the chemical vapor deposition method.28,29 Thus, there is a high probability of switching atomic sites because of the large thermal energy available during the process. Here, these possible point defects are created by removing an atom or replacing one atom with another atom in the 4 × 4 supercell of the WSSe monolayer. The considered vacancy defects are the sulfur vacancy (SV), selenium vacancy (SeV), sulfur/selenium vacancy [(S/Se)V], and tungsten vacancy (WV). Moreover, the antisite defects are a selenium atom replaced by a sulfur atom (SeS), a sulfur atom replaced by a selenium atom (SSe), a selenium atom replaced by a tungsten atom (SeW), a sulfur atom replaced by a tungsten atom (SW) and sulfur/selenium atoms replaced by tungsten atoms ((S/Se)W). The optimized structures of the defect-including monolayers together with the respective CDD are shown in Fig. S1 (ESI). The optimized structural, electronic, and magnetic parameters of the defect-including WSSe monolayers are listed in Table 1.

Table 1 Bond length, formation energy, electronic behavior, band gap, magnetic behavior, magnetic moment, and polarization energy of the vacancy and antisite defect induced WSSe monolayers. (SC – semiconductor, HM – half metal, NM – non-magnetic, FM – ferromagnetic, Eg – band gap, and ΔE – polarization energy)
Defects Bond length near defect (Å) E F (eV) Electronic behavior E g (eV) Magnetic behavior Magnetic moment (μB) ΔE (meV)
W–S W–Se
Vacancy defect WSSe_SV 2.45 2.57 −2.72 SC 1.63 NM 0.00 0.00
WSSe_SeV 2.43 2.59 −2.85 SC 1.67 NM 0.00 0.00
WSSe_(S/Se)V 2.41 2.59 −4.22 SC 1.67 NM 0.00 0.00
WSSe_WV 2.38 2.60 −5.31 SC 1.62 NM 0.00 0.00
Antisite defect WSSe_SeS 2.45 2.59 −0.76 SC 1.61 NM 0.00 0.00
WSSe_SSe 2.45 2.59 0.03 SC 1.57 NM 0.00 0.00
WSSe_SeW 2.46 2.62 −5.17 SC 1.35 FM 2.00 297
WSSe_SW 2.45 2.64 −10.1 HM FM 2.04 112
WSSe_(S/Se)W 2.36 2.51 −2.50 M FM 0.7 20


3.1 Electronic and magnetic properties of point defect-including WSSe monolayers

3.1.1 Vacancy defects. The optimized structure of the SV-including WSSe monolayer is shown in Fig. S1(a) (ESI). The defect percentage of SV in the WSSe monolayer is considered as 2.08%. The atomic position near to the defect is displaced slightly, and the bond length observed is 2.45 Å and 2.57 Å for W–S and W–Se, respectively. The formation energy for the SV-including WSSe monolayer was found to be −2.72 eV, as reported in Table 1. The band structure, along with the total density of states (TDOS) of the SV-including WSSe monolayer, is shown in Fig. 1(a), showing that the valence band maximum (VBM) and conduction band minimum (CBM) are both located at the K point of the BZ. The presence of defect states 0.3 eV below the CBM is substantiating the n-type semiconductor behavior. The defect states are mainly originating because of unsaturated tungsten and chalcogen bonds near the defect points. Similar results are observed for SV-including WS2 and MoS2 monolayers.69 Rafik et al.70 reported experimental evidence substantiating the n-type semiconductor behavior in MoS2, and the source of the n-type behavior is attributed to the sulfur vacancy in the MoS2 monolayer. The CDD of the SV-including WSSe monolayer is shown in Fig. S1(a) (ESI), showing a uniformly positive charge distribution near to the defect points. We have also plotted the spin-polarized partial density of states (PDOS) for the SV-including WSSe monolayer, Fig. 1(e), to gain more insight about the electronic properties, showing the equal distribution of spin up and spin down states confirming the non-magnetic semiconducting behavior of the monolayer. The states in the VBM and CBM are mainly originating from W-d orbitals, where a small contribution from S/Se-p orbitals is also noticed. W-d orbitals and S/Se-p orbitals are hybridized in the valence band and conduction band because of the covalent bonding between the W–S and W–Se atoms. The defect states below the CBM are mainly originating due to the presence of unsaturated atoms of W-d and S/Se-p orbitals. The optimized structure for the selenium vacancy (SeV)-including WSSe monolayer, together with the CDD, is shown in Fig. S1(b) (ESI). The defect percentage of SeV is equal to the SV-including WSSe monolayer. The optimized structure shows displacement in the position of atoms near to the defect site. The computed bond length is 2.43 Å and 2.59 Å for W–S and W–Se, respectively. The formation energy for the SeV-including monolayer is −2.85 eV higher than the SV defect because of its high atomic weight. The band structure along with the TDOS for the SeV-including WSSe monolayer is shown in Fig. 1(b), certifying similar behavior to that of the SV-including WSSe monolayer. The optimized structure of the (S/Se)V-including WSSe monolayer along with the CDD is shown in Fig. S1(c) (ESI). Two S and Se atoms are removed to obtain (S/Se)V in the WSSe monolayer and hence the defect percentage of the (S/Se)V defect is 4.16%. In this case, tungsten atoms near to the defects are trying to come close to each other, causing a lowering of the W–W bond length. The formation energy of the (S/Se)V-including WSSe monolayer is found to be −4.22 eV higher than the SV and SeV defects. Moreover, the formation energy of the (S/Se)V defect is −2.11 eV per atom, close to the SV and SeV defects. The band structure together with the TDOS is shown in Fig. 1(c) for the (S/Se)V-including WSSe monolayer. Here we noticed that the defect states are shifted 0.42 eV below the CBM more towards the Fermi energy as compared to the SV and SeV-including WSSe monolayers. Its electronic behavior is similar to the S2V-including WS2 and MoS2 monolayers, as reported by Zhao et al.71 and Zheng et al.,72 respectively. The CDD plot for the (S/Se)V-including WSSe monolayer is shown in Fig. S1(c) (ESI), showing the accumulation of positive charge near the (S/Se)V defect site, and this is attributed to the presence of unsaturated bonds of tungsten atoms. Additionally, the defect states are originating because of the unsaturated tungsten and chalcogen bonds, showing the contribution of W-d and S/Se-p orbitals to the defect states, Fig. 1(g). Similar orbital contributions are observed in the valence and conduction bands of the SV and SeV-including WSSe monolayers. The onset of the defect levels below the conduction band in the (S/Se)V-including WSSe monolayer makes it an n-type semiconductor. The optimized structure of the WV-including WSSe monolayer is shown in Fig. S1(d) (ESI). The considered defect percentage of WV is 2.08%, which is equal to SV and SeV defects. The formation energy of WV is −5.31 eV, which is the highest among the considered vacancy defects and thus the probability of the WV defect is relatively lower as compared to the other vacancy defects. Thus, tungsten vacancy defects may not be easily created and probably may be absent during experimental realization. The obtained result is validated with experimental results, which show that the Mo vacancy is rarely observed.73 The change in the bond length near to the defect site is reported in Table 1. The respective band structure is shown in Fig. 1(d), suggesting that the defect states are lying deep or close to the Fermi level in the valence and conduction bands. The TDOS confirms that equal amounts of energy states are occupied by spin up and spin down; as a result, the WV-including WSSe monolayer showed non-magnetic semiconductor behavior. The CDD of the WV-including WSSe monolayer is shown in Fig. S1(d) (ESI). This is also substantiated by the spin-polarized PDOS spectra of defect states, Fig. 1(h). The onset of these defect states close to the Fermi level in the valence and conduction bands is attributed to W-d and Se-p orbitals and their respective hybridization. Moreover, W-d orbitals are mainly contributing to the VBM and CBM.
image file: d0cp02063j-f1.tif
Fig. 1 (a–d) Spin-polarized band structure along with the TDOS and (e–h) spin-polarized DOS of vacancy defect-including WSSe monolayers.
3.1.2 Antisite defects. The computed formation energy of the considered antisite defects is summarized in Table 1. The defect percentage for the considered antisite defects is 2.08% in the supercell. The CDD plot is shown in Fig. S1(e) (ESI) for the SeS-including WSSe monolayer, showing little charge accumulation and depletion because of the same number of valence electrons in sulfur and selenium atoms. The electronic configurations for sulfur ([Ne]3s23p4) and selenium ([Ar]3d104s24p4) atoms are identical and that's why the SeS antisite defect is not affecting the electronic properties significantly. Fig. 2(a) shows that the band gap of the SeS-including monolayer is about 1.61 eV, which is slightly less than the pristine MoSSe monolayer. The symmetric spin up and spin down contributions, as evident from the TDOS, for the SeS-including WSSe monolayer confirm the non-magnetic behavior. The respective PDOS contribution is shown in Fig. 2(f) and it can be noticed that W-d orbitals are mainly contributing to the VBM and CBM, as some of the occupied tungsten orbitals lie in the valence band, and unoccupied orbitals lie in the conduction band. SSe is another antisite defect created by replacing a sulfur atom with selenium, as shown in Fig. S1(f) (ESI), together with the respective CDD plot. The band structure for this monolayer is shown in Fig. 2(b). Here, both the VBM and CBM are located at the K point of the BZ and are separated by 1.57 eV, confirming the direct band gap semiconducting behavior. The band structure is not showing any significant change because of identical valence electrons for S and Se atoms. The contribution of the PDOS is similar to that of SeS in the WSSe monolayer, Fig. 2(g). The formation energies for the SeS and SSe-including WSSe monolayers are relatively smaller, Table 1. Thus, these antisite defects are more stable among the other considered vacancies and antisite defects in WSSe monolayers. Further, SeW and SW antisite defects are created by replacing the respective chalcogen atom with a tungsten atom. The computed formation energies for these defects are relatively higher than the other defects, suggesting that the formation of these defects is less probable. The higher formation energies of these defects make them less stable with respect to the SSe and SeS defects in the WSSe monolayer. Due to substitution of W atoms, large distortion is observed and the bond lengths are 2.45 Å and 2.62 Å for W–S and W–Se, respectively. The CDD plots for the SeW and SW-including WSSe monolayers are shown in Fig. S1(g) and (h) (ESI), respectively. This explains the charge accumulation at the defect site for both defects. The polarization energy for the SeW and SW-including WSSe monolayers is 297 meV and 112 meV, respectively, which confirms the magnetic behavior. Moreover, the other considered defects have zero polarization energy, confirming the non-magnetic behavior. The band structure (Fig. 2(c)) for the SeW-including WSSe layer shows that the defect states lie near to the Fermi energy because of the spin-up states’ contribution, consistent with the respective CDD plot. Further, the PDOS suggests that defect states are originating because of the presence of W-d and Se-p orbitals, Fig. 2(h). Here, interestingly, spin splitting in the up and down spins is observed, causing the onset of magnetic behavior. The spin splitting is also evident in the respective CDD plot, supporting the magnetic semiconducting behavior for such defects included in the WSSe monolayer. The magnetic moment of the SeW-including WSSe monolayer is ∼2.0 μB. The band structure along with the TDOS of the SW-including WSSe monolayer is shown in Fig. 2(d). The spin-polarized band structure suggests that defect states of up spins are crossing the Fermi level while defect states of down spins are near to the Fermi energy in the conduction band, confirming the half-metallic property. We also plotted the PDOS, confirming that the defect states in the up and down spins are mainly due to the presence of tungsten and chalcogen atoms, Fig. 2(i). The PDOS spectra exhibit the spin splitting, causing the onset of magnetic behavior with a total magnetic moment of around ∼2.04 μB. The SW-including WSSe monolayer has half-metallic ferromagnetic behavior, suitable for spintronic application. The last defect, (S/Se)W, is created by replacing the sulfur/selenium atoms with tungsten atoms. Fig. S1(i) (ESI) shows the optimized structure with the CDD plot. The optimized structure shows distortion near to the defect site and the bond lengths W–S and W–Se were found to be 2.36 Å and 2.51 Å, respectively. The band structure with the TDOS of (S/Se)W shows that the spin up and spin down states cross the Fermi level, confirming the metallic behavior. Fig. 2(j) shows the PDOS, confirming the asymmetry in the spin up and spin down states. The polarization energy and asymmetry in the PDOS suggest the onset of a magnetic moment in the (S/Se)W-including WSSe monolayer. The formation energy and other electronic and magnetic properties are reported in Table 1.
image file: d0cp02063j-f2.tif
Fig. 2 (a–d) Spin-polarized band structure along with the TDOS and (e–j) spin-polarized DOS of antisite defect-including WSSe monolayers.

The formation energy of SV vacancy point defects is relatively lower, whereas the SSe and SeS antisite defect formation energies are the lowest among all the considered intrinsic point defects. The studies suggest that antisite SSe and SeS point defects may be the most favorable antisite defects in the WSSe monolayer. We extended our studies on the most stable vacancy defects such as SV, SeV, and (S/Se)V and antisite defects SeS and SSe to investigate the sensing behavior of H2S, NH3, NO2, and NO gases. We considered these favourable defects further in the WSSe monolayer together with the stable orientation of gas molecules for analyzing their impact on the gas sensing characteristics.

3.2 Adsorption behavior

The optimized structural parameters of the gas molecules are in agreement with a previous report,50 as shown in Fig. 3. Jin et al.32 examined the adsorption of gas molecules (CO, CO2, NH3, NO, and NO2) on sulfur and selenium surfaces of MoSSe monolayers at different sites such as the center of the hexagon, top of the chalcogen element, and top site of the transition metal and chalcogen bond (Mo–S/Se) and observed that the center of the hexagon is the most stable configuration for the adsorption of gas molecules. We also observed similar results for the adsorption of gas molecules (H2S, NH3, NO2 and NO) on MoSSe monolayers.50 The electronic configuration and crystal symmetry of the WSSe monolayer are similar to those of the MoSSe monolayer. This is the reason for considering the most stable site (hexagon) in the WSSe monolayer for adsorption of the gas molecules with two probable exposures (i) sulfur and (ii) selenium surfaces as the top side. The gas molecules i.e. H2S, NH3, NO2, and NO are adsorbed at the most stable site i.e. hexagon sites of the WSSe monolayer. The different adsorption sites for gas molecules are presented in Fig. 4 with respective adsorption heights from the adsorbent surface.
image file: d0cp02063j-f3.tif
Fig. 3 Optimized molecules along with the structural parameters of (a) H2S, (b) NH3, (c) NO2, and (d) NO.

image file: d0cp02063j-f4.tif
Fig. 4 Optimized structure of H2S (a–d), NH3 (e–h), NO2 (i–l) and NO (m–p) molecules adsorbed on the selenium and sulfur surfaces of the WSSe monolayer. (The WSSe_Se (C1) configuration shows the gas molecule H in H2S, H in NH3, O in NO2 and O in NO pointing towards the top of the selenium surface. The WSSe_Se (C2) configuration shows the gas molecule S in H2S, N in NH3, N in NO2 and N in NO pointing towards the top of the selenium surface. The WSSe_S (C1) configuration shows the gas molecule H in H2S, H in NH3, O in NO2, and O in NO pointing towards the top of the sulfur surface. The WSSe_S (C2) configuration shows the gas molecule S in H2S, N in NH3, N in NO2, and N in NO pointing towards the top of the sulfur surface.)
3.2.1 Adsorption on the WSSe pristine monolayer. The H2S molecule is allowed to adsorb from the hydrogen site (hydrogen atom pointing toward the monolayer) and sulfur site (sulfur atom pointing toward the monolayer) at the top of the hexagon of the sulfur and selenium surface of the WSSe monolayer. The optimized geometries of the H2S molecule adsorbed on WSSe monolayers together with the adsorption distances (heights) and the respective charge transfer are shown in Fig. 4(a–d). The adsorption distances vary from 2.49 Å to 3.37 Å, suggesting the impact of weak covalent and vdW interactions, referring to physiosorption between the gas molecule and WSSe monolayer. The adsorption distance between H2S and the WSSe monolayer is found to be a minimum when the hydrogen atom of H2S is allowed to adsorb on the top of the selenium surface. Moreover, the maximum adsorption distance is observed when H2S is adsorbed from the sulfur site on the top of the sulfur surface of the WSSe monolayer. The adsorption height reduces for H2S adsorption from the hydrogen site and this is attributed to the different nature of charges on H and S/Se atoms, showing enhanced Coulomb attraction. Further, we also observed a higher charge sharing of about 0.011e and 0.008e for sulfur as compared to H in H2S on selenium and sulfur surfaces, respectively. The optimized configurations show that the hydrogen in the H2S molecule is located at the top of the tungsten atom, whereas S in H2S is at the middle of the hexagon of the WSSe monolayer. In all these configurations, the charge is transferred from the monolayer to the gas molecule, suggesting acceptor behavior of the H2S molecule. We also plotted the CDD to substantiate the Bader charge analysis and here also the maximum charge accumulation is observed near the H2S molecule, Fig. 5(a–d). The spin-polarized DOS is shown in Fig. 6(a and b) for the pristine monolayer together with the adsorbed H2S molecule in different configurations. We do not observe the onset of any additional states in the forbidden region (i.e. the band gap) of the WSSe monolayer and thus H2S may not exhibit any significant change in electrical conductivity during adsorption. Further, the noticed equal spin up and spin down contribution in the DOS suggests the absence of any magnetic state after the adsorption of the H2S molecule.
image file: d0cp02063j-f5.tif
Fig. 5 Charge density difference of H2S (a–d), NH3 (e–h), NO2 (i–l), and NO (m–p) molecules adsorbed on the selenium and sulfur surfaces of the WSSe monolayer. (Isosurface value 10−4 e Å−3, the yellow color indicates electron accumulation and the blue color indicates electron depletion.) (The WSSe_Se (C1) configuration shows the gas molecule H in H2S, H in NH3, O in NO2, and O in NO pointing towards the top of the selenium surface. The WSSe_Se (C2) configuration shows the gas molecule S in H2S, N in NH3, N in NO2, and N in NO pointing towards the top of the selenium surface. The WSSe_S (C1) configuration shows the gas molecule H in H2S, H in NH3, O in NO2, and O in NO pointing towards the top of the sulfur surface. The WSSe_S (C2) configuration shows the gas molecule S in H2S, N in NH3, N in NO2, and N in NO pointing towards the top of the sulfur surface.)

image file: d0cp02063j-f6.tif
Fig. 6 Comparison of the pristine monolayer and adsorbed gas monolayer density of states: (a and b) WSSe (H2S), (c and d) WSSe (NH3), (e and f) WSSe (NO2), and (g and h) WSSe (NO).

We also considered the adsorption of the NH3 molecule with different orientations such as nitrogen and hydrogen sites on the sulfur and selenium surfaces of the WSSe monolayer. The top and side view of the optimized geometry of the NH3 molecule adsorbed on the WSSe monolayer along with the adsorption distance and Bader charges are shown in Fig. 4(e–h). The adsorption distance varies between 2.58 Å and 2.82 Å for different configurations. The minimum adsorption distance is observed for the adsorption of the NH3 molecule from the hydrogen site at the top of the selenium surface of the WSSe monolayer. However, the maximum adsorption distance is observed when the NH3 molecule is adsorbed from the nitrogen site at the top of the selenium surface of the WSSe monolayer, confirming the physiosorption between them. We also observed relatively more charge transfer from the NH3 molecule of about 0.023e and 0.029e towards the WSSe monolayer on the selenium and sulfur surfaces of the WSSe monolayer, as shown in Fig. 4(f) and (h), respectively. The charge transfer from the NH3 molecule to the WSSe monolayer confirms the donor behavior of the gas molecule. The Fermi level of the Janus WSSe monolayer may lie between the highest occupied molecular orbital (HOMO) and lowest occupied molecular orbital (LUMO) of NH3. This suitable position of energy levels is the main source of the noticed charge transfer from the NH3 molecule to the WSSe monolayer. When the NH3 molecule is allowed to adsorb from the hydrogen site on the selenium or sulfur surface of the WSSe monolayer then the hydrogen atoms of the optimized NH3 molecule are pointing towards the Se/S atoms of the WSSe monolayer. The adsorption distances are relatively lower, and the charge sharing is more for the NH3 molecule as compared to that of the H2S molecule. This is attributed to the superposition of the intrinsic dipoles from NH3 and the WSSe monolayer, leading to an enhanced electric field.32 This enhanced electric field will exhibit improved gas sensing characteristics. This is similar to the observations in MoS2 by Yue et al.74 Further, Bader charge analysis also supports the charge accumulation, as illustrated in Fig. 4(e–h), thus validating the CDD observations, Fig. 5(e–h). The spin-polarized DOS for the pristine monolayer is compared with that of NH3 adsorbed on the WSSe monolayer, Fig. 6(c and d). The DOS is not showing any change in the bandgap due to NH3 adsorption. The absence of energy explains that there may not be any significant change in electrical conductivity during NH3 adsorption together with non-magnetic semiconducting behavior.

The top and side views of the optimized structure of NO2 adsorbed on the WSSe monolayer are shown in Fig. 4(i–l) for both oxygen and nitrogen sites at the top of the selenium and sulfur surfaces of a WSSe monolayer. The top view of the optimized geometry shows that the oxygen atoms in NO2 are shifted towards the top of the transition metal. The adsorption distance between NO2 and the WSSe monolayer is in the range of 2.76 Å to 2.81 Å. These configurations are confirming the vdW interactions together with the minimum adsorption distance 2.76 Å for oxygen atoms facing towards S atoms in the WSSe monolayer, suggesting physiosorption between the adsorbate and host. The CDD plots are shown in Fig. 5(i–l), illustrating charge accumulation near to NO2, confirming the acceptor nature. Here, the LUMO of NO2 lies below the Fermi level of the WSSe monolayer and this is the main source of the observation of charge accumulation near to NO2. The same result is also substantiated using Bader charge analysis. That's why the NO2 molecule shares more charge of about 0.122e for oxygen atoms of molecules towards the selenium surface, and the lowest charge transfer of about 0.057e is observed for the nitrogen atom facing towards the sulfur surface of the WSSe monolayer. The spin-polarized DOS is shown in Fig. 6(e and f) for the NO2 molecule adsorbed on the WSSe monolayer. Flat bands are observed for both the valence and conduction bands due to NO2 molecular orbitals. Spin splitting between spin up and spin down is observed for the NO2 molecule, showing a ∼1 μB magnetic moment for such adsorption.

Adsorption of the NO molecule is also evaluated on the WSSe monolayer, and tilted orientations of NO molecules with respect to the monolayer are noticed, Fig. 4(m–p). The optimized geometries for NO adsorbed on the WSSe monolayer under different configurations suggest that the NO molecule is stable at the center of the hexagon on the WSSe monolayer. The adsorption distances vary from 2.86 Å to 3.22 Å, confirming the vdW interactions with physiosorption under different NO configurations on the WSSe monolayer. The adsorption of the NO molecule from the nitrogen site on the top of the selenium surface of the WSSe monolayer is the most stable among the other configurations, as the lowest adsorption height is noticed for this configuration. Bader charge analysis also supports the acceptor behavior for the NO molecule because of the relative LUMO position of the NO molecule, lying below the Fermi level of the WSSe monolayer. The CDD plots, Fig. 5(m–p), also support the noticed charge accumulation at the NO molecule from Bader charge analysis. Here also, like NO2, flat bands are observed in the forbidden region of the spin-polarized DOS between the valence band and conduction band and this is attributed to asymmetry of the spin up and spin down states, which is also inducing about a 1 μB magnetic moment for the adsorbed NO molecule on the WSSe monolayer.

The adsorption energy and recovery time are important parameters to understand the gas sensing characteristics of any material. We calculated these quantities using the approach described in the computational section, and the results are summarized in Fig. 7(a–d). The adsorption energy (recovery time) of the H2S molecule lies in the range of −0.13 eV (1.5 × 10−4 μs) to −0.19 eV (1.5 × 10−3 μs) for the sulfur and selenium surfaces in the WSSe monolayer with different H2S configurations, respectively. The maximum adsorption energy on the selenium surface is observed when the H2S molecule is adsorbed from the hydrogen site. Moreover, the minimum adsorption energy is noticed when the H2S molecule is adsorbed from the sulfur site at the selenium surface of the WSSe monolayer. The adsorption energy of the H2S molecule on the WSSe monolayer is higher than the MoSSe monolayer.50 The adsorption energies (recovery time) for the NH3 molecule are −0.16 eV (4.5 × 10−4 μs) for H at the selenium surface, −0.22 eV (4.9 × 10−3 μs) for N at the selenium surface, −0.19 eV (1.5 × 10−3 μs) for H at the sulfur surface, and −0.17 eV (7.1 × 10−4 μs) for N at the sulfur surface in the WSSe monolayer. We noticed that the N at the selenium surface configuration for the adsorption is the most stable among the other NH3 configurations. The NH3 adsorption energy for the WSSe monolayer is higher than the WS2, WSe2, MoS2, MoSe2, and MoSSe monolayers, as can be inferred from the values summarized in Table 2. The recovery time is much smaller with respect to other monolayer systems, suggesting that the Janus WSSe monolayer may be very suitable for NH3 gas sensing application. Further, the adsorption energy and recovery time for the NO2 molecule for various considered configurations are shown in Fig. 7(a) and (b–d), respectively, with a maximum adsorption (recovery time) of −0.28 eV (4.3 × 10−2 μs) for the oxygen at the sulfur surface configuration. The WSSe monolayer has more NO2 adsorption capability along with low recovery time with respect to other studied TMD monolayers, as illustrated in Table 2. The NO molecule has an adsorption energy (recovery time) of −0.10 eV (4.8 × 10−5 μs) for O on the selenium surface, −0.19 eV (1.5 × 10−3 μs) for N on the selenium surface, −0.14 eV (2.2 × 10−4 μs) for O on the sulfur surface, and −0.17 eV (7.1 × 10−4 μs) for N on the sulfur surface of the WSSe monolayer. The selenium surface showed a high sensitivity over the sulfur surface of the WSSe monolayer because of the large electrostatic potential on the selenium surface with respect to the sulfur surface of the WSSe monolayer. Here, the difference in the electrostatic potential induces an electric field along the out of plane direction. The gas molecules share more charge on the selenium surface due to its large electrostatic potential with respect to the sulfur surface. This large charge on Se surface induces relatively larger dispersive force.32 That's why the adsorption energy for gas molecules on the selenium surface is larger than that on the sulfur surface. The gas sensing characteristics for the WSSe monolayer are compared with other 2D materials, summarized in Table 2. The noticed WSSe monolayer adsorption energies are higher than other 2D monolayer and heterobilayer systems. Further, the recovery time is on the order of a microsecond, suggesting ultrahigh reversibility for WSSe monolayer based sensors. The higher adsorption energy for the NO2 molecule with respect to the other considered molecules, together with fast recovery, suggests the suitability of the WSSe monolayer for NO2 gas sensing applications. High temperature is used to improve the recovery time of gases. The effect of temperature is considered in the 300 K to 400 K range, and the respective recovery times are computed. The temperature provides additional thermal energy to break the bonds between the adsorbate and host, resulting in fast recovery of gases. The temperature dependent recovery is plotted in Fig. 7(b–d). The recovery time is decreasing with increasing temperature.


image file: d0cp02063j-f7.tif
Fig. 7 (a) Adsorption energy and (b–d) temperature-dependent recovery time of gas molecules adsorbed on the selenium and sulfur surfaces of the Janus WSSe monolayer.
Table 2 Comparison of the sensing characteristics i.e. adsorption energy Eads and charge Q of the Janus WSSe monolayer with reported highly sensitive gas sensing materials
Adsorbent Properties H2S NH3 NO2 NO Theory Ref.
WSSe monolayer E ads (eV) −0.193 −0.220 −0.276 −0.189 GGA & DFT-D2 This work
Q (e) 0.008 −0.023 0.122 0.020
MoSSe monolayer E ads (eV) −0.156 −0.203 −0.252 −0.117 GGA & DFT-D2 50
Q (e) 0.011 −0.028 0.137 0.006
MoSSe monolayer E ads (eV) NA −0.200 −0.242 −0.150 GGA & DFT-D3 32
Q (e) NA −0.031 0.107 0.038
WS2 monolayer E ads (eV) NA −0.180 −0.206 −0.132 GGA & DFT-D2 16
Q (e) NA NA NA NA
WSe2 monolayer E ads (eV) NA −0.042 −0.067 −0.025 GGA 75
Q (e) NA −0.017 0.116 0.034
MoS2 monolayer E ads (eV) NA −0.160 −0.169 −0.114 GGA & DFT-D2 16
Q (e) NA NA NA NA
MoS2 monolayer E ads (eV) NA −0.195 −0.208 −0.138 GGA & DFT-D3 32
Q (e) NA −0.036 0.055 −0.005
MoSe2 monolayer E ads (eV) NA −0.195 −0.252 −0.140 GGA & DFT-D3 32
Q (e) NA −0.025 0.117 0.017
MoS2/WS2 heterobilayer, MoS2 side E ads (eV) NA −0.166 −0.116 −0.177 GGA & DFT-D2 16
Q (e) NA NA NA NA
MoS2/WS2 heterobilayer, WS2 side E ads (eV) NA −0.186 −0.213 −0.135 GGA & DFT-D2 16
Q (e) NA NA NA NA
HfSe2 monolayer E ads (eV) −0.865 −0.149 −0.171 −0.193 GGA & DFT-D2 76
Q (e) −0.005 0.002 0.053 −0.054
Graphene E ads (eV) NA −0.031 −0.067 −0.029 GGA 77
Q (e) NA 0.027 −0.102 0.017
h-BN monolayer E ads (eV) NA −0.10 −0.125 −0.117 LDA 78
Q (e) NA 0.017 0.039 0.005
P2C2 E ads (eV) −0.449 −0.483 −1.318 NA GGA & DFT-D2 79
Q (e) 0.002 −0.032 0.117 NA
SnSe E ads (eV) NA −0.263 −0.770 NA GGA & DFT-D2 80
Q (e) NA −0.099 0.307 NA


3.2.2 Sensing characteristics of defect-including WSSe monolayers.
Selenium replaced by sulfur. The H2S molecule is placed on the top of the SeS-including WSSe monolayer, and the structure is relaxed. The optimized geometry is shown in Fig. 8(a), illustrating a 3.35 Å adsorption distance between the adsorbate and host, supporting the physiosorption between them. The optimized geometry shows that the sulfur atom of the H2S molecule is pointed towards the host and the bond length of the adsorbate is elongated by 0.002 Å. The calculated adsorption energy (recovery time) is −0.176 eV (8.9 × 10−4 μs), which is higher than that for the H at the selenium and sulfur atomic plane configurations in the WSSe monolayer. The CDD plot, Fig. 8(e), shows the sharing of charge between the adsorbate and the host material. The magnitude of the charge sharing is calculated using Bader charge analysis and we noticed a negligible 0.0001e charge transfer from the adsorbed molecule to the host, confirming the donor behavior. The spin-polarized DOS is plotted in Fig. 8(i), for H2S adsorbed on the SeS-including host. We noticed strong hybridization of the orbital contribution of H2S with the SeS-including WSSe monolayer. However, the small charge transfer will not lead to any significant change in electronic conductivity after the adsorption of the H2S molecule. Adsorption of the NH3 molecule is also examined and an optimized structure is shown in Fig. 8(b), explaining the adsorption distance of about 2.94 Å. It's confirming the vdW interactions along with physiosorption between them. The optimized configuration shows that the NH3 molecule is adsorbed from the nitrogen site towards the SeS-including WSSe monolayer. The CDD plot, Fig. 8(f), shows the strong charge overlap between the adsorbate and the host. Further, Bader charge analysis illustrates about 0.024e charge transfer from NH3 to the SeS-including WSSe monolayer, suggesting the donor nature of the NH3 molecule. The adsorption energy (recovery time) is −0.20 eV (2.8 × 10−3 μs), which is less than N at the selenium and sulfur surfaces of the WSSe monolayer. The spin-polarized DOS is calculated to see the change in the electrical properties, but no additional states are noticed after the adsorption of the NH3 molecule, Fig. 8(j). The optimized geometry of the NO2 molecule adsorbed on the top of the SeS-including WSSe monolayer is shown in Fig. 8(c). The optimized configuration shows an adsorption distance of 2.83 Å, whereas Bader charge analysis confirms about 0.12e charge transfer from the host to the NO2 molecule, suggesting acceptor behavior of the NO2 molecule. The computed adsorption energy (recovery time) is −0.28 eV (4.7 × 10−2 μs), which is higher than that of the pristine WSSe monolayer. The spin-polarized DOS for NO2 adsorbed on the host is compared with the SeS-including WSSe monolayer and the onset of additional states in the valence and conduction bands is noticed for the adsorbed NO2 molecule. The optimized structure of NO adsorbed on the SeS-including host is shown in Fig. 8(d), and about a 3.08 Å adsorption distance is noticed, confirming the vdW interactions with physiosorption between them. The CDD plot, Fig. 8(h), shows the charge accumulation near to the NO molecule, validating the Bader charge results of a recorded charge transfer of about 0.01e from the SeS-including WSSe monolayer to the NO molecule, showing acceptor character of the NO molecule. The adsorption energy (recovery time) is −0.15 eV (3.1 × 10−4 μs) and the spin-polarized DOS suggests that the electronic states are not changing significantly; however, the Fermi level is shifted due to adsorption of the NO molecule.
image file: d0cp02063j-f8.tif
Fig. 8 (a–d) Adsorption distance and charge transfer between the gas molecules and the SeS-including WSSe monolayer. (e–h) Charge density difference, and (i–l) total density of states of the SeS-including WSSe monolayer compared with gas molecules adsorbed on the SeS-including WSSe monolayer. (CDD plot: isosurface value 10−4 e Å−3, the yellow color indicates electron accumulation and the blue color indicates electron depletion.)

Sulfur replaced by selenium. The optimized geometry of the H2S molecule adsorbed on the SSe-including WSSe monolayer shows an adsorption distance of 3.08 Å, implying the presence of physisorption between the host and adsorbate, Fig. 9(a). The adsorption energy (recovery time) is −0.13 eV (1.7 × 10−4 μs) and 0.0006e charge transfer is noticed from the H2S molecule to the SSe-including WSSe monolayer. Similar behavior is observed in the pristine and SeS-including WSSe monolayers. The CDD plot, Fig. 9(e), shows charge sharing between the H2S molecule and host, confirming the donor behavior of the molecule, whereas the spin-polarized DOS does not reflect any significant change for H2S adsorbed on the SSe-including WSSe monolayer and this is similar to that of the SeS-including WSSe monolayer. Fig. 9(b) shows the optimized geometry of NH3 adsorbed on the top of the SSe-including WSSe monolayer and also shows an adsorption distance 3.02 Å, suggesting vdW interactions and physisorption between them. The adsorption energy (recovery time) is around −0.20 eV (2.1 × 10−3 μs) and a Bader charge of about 0.028e is transferred from the NH3 molecule to the host, inferring donor characteristics of the NH3 molecule.
image file: d0cp02063j-f9.tif
Fig. 9 (a–d) Adsorption distance and charge transfer between the gas molecules and the SSe-including WSSe monolayer. (e–h) Charge density difference, and (i–l) total density of states of the SSe-including WSSe monolayer compared with gas molecules adsorbed on the SeS-including WSSe monolayer. (CDD plot: isosurface value 10−4 e Å−3, the yellow color indicates electron accumulation and the blue color indicates electron depletion.)

The spin-polarized DOS is shown in Fig. 9(j), signifying the strong overlap between the NH3 molecule and the host. The NH3 molecule adsorbed on the top of the SSe-including WSSe monolayer is found to be more sensitive as compared to the H2S molecule. The adsorption of the NO2 molecule on top of the SSe-including WSSe monolayer is shown in Fig. 9(c) together with a 2.58 Å adsorption distance, which is less than the SeS-including WSSe monolayer. The small adsorption distance confirms the weak covalent and vdW interactions. The computed NO2 adsorption energy (recovery time) is −0.23 eV (8.7 × 10−3 μs), which is higher than H2S and NH3 molecules adsorbed on the SSe-including WSSe monolayer. Bader charge analysis depicts 0.10e charge transfer from the monolayer to the NO2 molecule, confirming the acceptor behavior of the NO2 molecule. The CDD plot, Fig. 9(g), validates the Bader charge results. The optimized geometry for the NO molecule adsorbed on the top of the SSe-including WSSe monolayer is shown in Fig. 9(d), where the molecule is inclined by an angle together with a 3.26 Å adsorption distance, confirming vdW interactions between them. The computed adsorption energy (recovery time) is −0.09 eV (3.2 × 10−5 μs), which is smaller than the other considered molecules i.e. H2S, NH3, and NO2 adsorbed on the SSe-including WSSe monolayer. We noticed charge accumulation near to the NO molecule, as can be noticed in the CDD plot, Fig. 9(h), together with charge transfer of about 0.008e from the host to the NO molecule, suggesting acceptor characteristics of the NO molecule.


Sulfur vacancy. The optimized geometry of H2S adsorbed on the SV-including WSSe monolayer is shown in Fig. 10(a), which shows a 1.57 Å adsorption distance, confirming the weak covalent interaction between the adsorbate and host material. The relatively large adsorption energy (recovery time) of −0.30 eV (1.04 × 10−1 μs) also confirms the strong sensitivity and good recovery time. Large charge sharing is noticed in the CDD plots for the H2S molecule, where the Bader charge results also support the charge transfer of about 0.021e from the host to the adsorbate, showing acceptor characteristics of the H2S molecule. The bond length for the H2S gas molecule is elongated slightly by 0.001 Å. The H2S electronic states are strongly overlapping with the host material, as can be inferred from the spin-polarized DOS, Fig. 10(i). Further, the optimized geometry of the adsorbed NH3 molecule on top of the SV-including WSSe monolayer is shown in Fig. 10(b). The adsorption distance is found to be around 0.95 Å, and this led to covalent interactions. The adsorption energy (recovery time) is about −0.31 eV (1.4 × 10−1 μs), higher than that of the H2S molecule. The Bader charge analysis shows acceptor behavior of the NH3 molecule where a charge transfer of about 0.027e is noticed from the host to the molecule. Charge sharing is also evident from the CDD plot, as shown in Fig. 10(f). The spin-polarized DOS for NH3 adsorbed on the SV-including WSSe monolayer also shows similar behavior to H2S adsorbed on the SeV-including WSSe monolayer. The DOSs of the spin up and spin down states are symmetric, inferring non-magnetic semiconducting behavior. The adsorption height, adsorption energy, and charge sharing values confirm that the SV-including WSSe monolayer is more sensitive to the NH3 molecule with respect to the H2S molecule. The NO2 molecule is adsorbed from an oxygen side on the surface of the SV-including WSSe monolayer, as shown in Fig. 10(c).
image file: d0cp02063j-f10.tif
Fig. 10 (a–d) Adsorption distance and charge transfer between the gas molecules and the SV-including WSSe monolayer. (e–h) Charge density difference, and (i–l) total density of states of SV compared with gas molecules adsorbed on the SV-including WSSe monolayer. (CDD plot: isosurface value 10−4 e Å−3, the yellow color indicates electron accumulation and the blue color indicates electron depletion.)

The optimized monolayer shows that the NO2 molecule is dissociating and forming the oxygen doped NO adsorption configuration, confirming chemisorption. The adsorption energy (recovery time) is −0.26 eV (2.80 × 10−2 μs) for the NO2 molecule. The adsorption distance between the adsorbate and the host material is 2.56 Å. The Bader charge analysis shows that the adsorbed molecule is accepting 1.09e charge from the host, thus confirming acceptor behavior. More interestingly, the observed splitting of the spin-polarized DOS and the onset of additional states are attributed to the adsorption of NO2 on the host. The noticed spin splitting induces an about 1 μB magnetic moment for the adsorbate system. The adsorption of the NO molecule from the oxygen site at the top of the SV-including WSSe monolayer is investigated and the optimized geometry is shown in Fig. 10(d). The optimized geometry of the NO molecule is tilted by 45° from the host material and they are separated by 1.57 Å. The adsorption distance confirms the covalent interaction. The adsorption energy is about −0.27 eV and about 0.05e charge transfer from Bader charge analysis is noticed from the host to the adsorbed NO molecule, confirming the acceptor behavior, and the results are consistent with the CDD plots. The onset of flat states is noticed from the spin-polarized DOS, Fig. 10(l), after adsorption of the NO molecule, and thus the system will exhibit significant changes in electrical conductivity.


Selenium vacancy. Fig. 11(a–d) show the optimized structures for the considered gas molecules on the SeV-including WSSe monolayer. We noticed an absorption height of about 1.62 Å and 0.022e charge transfer from the host to the molecule, confirming the acceptor behavior for the H2S molecule. The calculated adsorption height confirms the covalent interaction between the molecule and the host. In the SeV-including WSSe monolayer, more charge transfer is noticed to the H2S molecule as compared to the SV-including WSSe monolayer. The adsorption energy (recovery time) is −0.305 eV (1.3 × 10−1 μs), which is also higher than that of the H2S molecule adsorbed on the SV-including WSSe monolayer. The absence of any additional density of states due to the adsorption of H2S suggests that the absorption behavior of the H2S molecule on the SeV-including WSSe monolayer is similar to that on the SV-including WSSe monolayer. However, the adsorption properties show that the SeV-including WSSe monolayer is more sensitive to the H2S molecule with respect to the SV-including WSSe monolayer. The optimized geometry of the NH3 molecule adsorbed on the SeV-including WSSe monolayer is shown in Fig. 11(b) together with a 0.85 Å adsorption height, which is smaller than that of the SV-including WSSe monolayer. The calculated adsorption energy and Bader charge are −0.36 eV and 0.030e, respectively, which are higher than those of the SV-including monolayer. The spin-polarized DOS is also not changing significantly after NH3 adsorption. The recovery time has increased to 1.2 μs for the NH3 molecule and also the bond length of the adsorbed NH3 molecule has elongated by 0.04 Å. Thus, the energetics suggests that the SeV-including WSSe monolayer is more sensitive to the NH3 molecule with respect to the SV-including WSSe monolayer. When the NO2 molecule adsorbs on the SeV-including WSSe monolayer then the NO2 dissociated and formed oxygen doped NO adsorption. The adsorption distance confirms the chemisorption between them. We observed similar behavior of NO molecules to that on the SV-including monolayer. The onset of the new density of states in the forbidden region for both these molecules suggests that significant changes in electronic conductivity may be observed.
image file: d0cp02063j-f11.tif
Fig. 11 (a–d) Adsorption distance and charge transfer between the gas molecules and the SeV-including WSSe monolayer. (e–h) Charge density difference, and (i–l) total density of states of the SeV-including WSSe monolayer compared with gas molecules adsorbed on the SeV-including WSSe monolayer. (CDD plot: isosurface value 10−4 e Å−3, the yellow color indicates electron accumulation and the blue color indicates electron depletion.)

Sulfur/selenium vacancy. The optimized geometry for the adsorbed H2S molecule on the top of the (S/Se)V-including WSSe monolayer is shown in Fig. 12(a), where a hydrogen atom is pointing towards the site. The adsorption distance between the adsorbate and host is about 1.61 Å together with a −0.31 eV (1.6 × 10−1 μs) adsorption energy (recovery time) for the H2S molecule. The noticed adsorption energy in this configuration is more than the SV and SeV-including WSSe monolayer configurations. Bader charge analysis depicts about 0.023e charge transfer from the host material to the adsorbed molecule. It's slightly higher than that observed in other vacancy-including WSSe monolayer configurations. The CDD plot, Fig. 12(e), validates the Bader charge analysis results. Here, no change in the DOS is observed from the computed spin-polarized DOS, as shown in Fig. 12(i). The optimized geometry of the NH3 molecule adsorbed on the (S/Se)V-including WSSe monolayer, Fig. 12(b), shows that the hydrogen atom is pointing towards the defect site of the host material and the adsorption distance is about 0.94 Å. The calculated adsorption energy of −0.36 eV is equal to that of the SeV-including WSSe monolayer. The CDD plot shows a strong interaction between the adsorbate and the host, similar to the SV and SeV-including WSSe monolayers, together with acceptor behavior for NH3 where 0.029e charge transfer is noticed from the host to the NH3 molecule. The optimized structure of the NO2 molecule adsorbed on the surface of the (S/Se)V-including WSSe monolayer shows that the NO2 molecule dissociates into NO and O and possesses chemisorption. Due to the presence of strong overlapping of NO and O orbitals with the monolayer, the WSSe monolayer shares a huge 2.10e charge with the NO2 molecule. The CDD plot validates the results predicted by the Bader charge analysis, confirming the covalent interaction. The adsorption energy is observed to be around −2.58 eV, which is very high. The recovery time is very long so the desorption of the molecule will be very tough. The spin-polarized DOS of the (S/Se)V-including WSSe monolayer with NO2 adsorption shows asymmetry in the spin up and spin down states, which leads to a change in the electrical conductivity and the onset of a magnetic moment of around 1 μB. The optimized structure shows that the NO molecule breaks into N and O atoms, forming bonding at the bottom and top layer, respectively, and forming covalent bonding with chemisorption. The CDD illustrates the strong orbital overlap, confirming the huge charge sharing and covalent bonding. The spin-polarized DOS of NO absorbed on the WSSe monolayer shows symmetry between the spin up and spin down states, suggesting non-magnetic semiconductor behavior.
image file: d0cp02063j-f12.tif
Fig. 12 (a–d) Adsorption distance and charge transfer between the gas molecules and the (S/Se)V-including WSSe monolayer. (e–h) Charge density difference, and (i–l) total density of states of the (S/Se)V-including WSSe compared with gas molecules adsorbed on the (S/Se)V-including WSSe monolayer. (CDD plot: isosurface value 10−4 e Å−3, the yellow color indicates electron accumulation and the blue color indicates electron depletion).

Further, we summarized the adsorption energy for all considered defect-including WSSe monolayer configurations, as shown in Fig. 13a. The NO2 molecule adsorbed on the SeV and (S/Se)V-including WSSe monolayers has high adsorption energy. Moreover, the NO molecule also has high adsorption energy when it adsorbs on the (S/Se)V-including WSSe monolayer. That's why the recovery time for these configurations is very high. The recovery time for all molecules adsorbed on defective monolayers is plotted in Fig. 13(b–d) for different temperatures. The recovery time is decreasing with increasing temperature because of additional thermal energy, assisting in extracting the molecule away from the host materials’ surface. The adsorption energy and Bader charge for gas molecules adsorbed on defect-including monolayers are listed in Table 3.


image file: d0cp02063j-f13.tif
Fig. 13 (a) Adsorption energy and (b–d) temperature dependent recovery time of gas molecules adsorbed on defect-including monolayers.
Table 3 Comparison of the gas sensing performance of defective WSSe monolayers with defective MoSSe monolayers
Adsorption configuration Properties H2S NH3 NO2 NO Ref.
WSSe_SeS E ads (eV) −0.176 −0.205 −0.279 −0.148 This work
Q (e) −0.0001 −0.024 0.12 0.01
WSSe_SSe E ads (eV) −0.133 −0.198 −0.235 −0.090 This work
Q (e) −0.0006 −0.028 0.10 0.008
WSSe_SV E ads (eV) −0.299 −0.307 0.265 −0.268 This work
Q (e) 0.021 0.027 1.10 0.05
WSSe_SeV E ads (eV) −0.305 −0.363 −2.075 −0.276 This work
Q (e) 0.022 0.030 1.11 0.06
MoSSe_SeV E ads (eV) −0.232 −0.281 −3.36 −2.788 50
Q (e) 0.017 |0.019 1.027 0.915
WSSe_(S/Se)V E ads (eV) −0.310 −0.363 −2.581 −3.510 This work
Q (e) 0.023 0.029 2.10 2.24
MoSSe_(S/Se)V E ads (eV) −0.238 −0.274 −3.404 −2.894 50
Q (e) 0.018 0.022 1.077 1.058


4. Conclusion

We systematically investigated the gas sensing properties of H2S, NH3, NO2, and NO molecules on the sulfur and selenium surfaces of the Janus WSSe monolayer in terms of adsorption height, charge transfer, adsorption energy, recovery time and electrical conductivity. The selenium surface showed greater adsorption characteristics as compared to the sulfur surface. These observations suggest that the WSSe monolayer is highly sensitive to H2S, NH3, NO2, and NO together with very fast recovery times as compared to the most commonly reported MoS2, MoSe2, WS2 WSe2, graphene and h-BN monolayers for gas sensing application. The SeS and SSe-including WSSe monolayers showed a 70–80% improvement in the adsorption energies of H2S and NH3 molecules as compared to the pristine monolayer. NO2 and NO molecules are ultrasensitive to vacancy defects and their adsorption energy increases significantly. These studies on gas sensing behavior will provide a strong foundation for the development of Janus WSSe monolayer based nanoscale highly sensitive gas sensors.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Author Ambesh Dixit acknowledges Department of Science and Technology, Government of India, through project DST/INT/Mexico/P-02/2016 and IIT Jodhpur for providing computational resources to carry out this work.

References

  1. X. Li, X. Li, Z. Li, J. Wang and J. Zhang, WS2 nanoflakes based selective ammonia sensors at room temperature, Sens. Actuators, B, 2017, 240, 273–277 CrossRef CAS .
  2. K. Patel, B. Roondhe, S. D. Dabhi and P. K. Jha, A new flatland buddy as toxic gas scavenger: A first principles study, J. Hazard. Mater., 2018, 351, 337–345 CrossRef CAS PubMed .
  3. S. H. Mir, S. Chakraborty, J. Wärnå, S. Narayan, P. C. Jha, P. K. Jha and R. Ahuja, A comparative study of hydrogen evolution reaction on pseudo-monolayer WS2 and PtS2: Insights based on the density functional theory, Catal. Sci. Technol., 2017, 7, 687–692 RSC .
  4. N. Sakhuja, R. K. Jha, R. Chaurasiya, A. Dixit and N. Bhat, 1T-Phase Titanium Disulfide Nanosheets for Sensing H2S and O2, ACS Appl. Nano Mater., 2020, 3(4), 3382–3394,  DOI:10.1021/acsanm.0c00127 .
  5. J. Zhang, Z. Qin, D. Zeng and C. Xie, Metal-oxide-semiconductor based gas sensors: Screening, preparation, and integration, Phys. Chem. Chem. Phys., 2017, 19, 6313–6329 RSC .
  6. C. Wang, L. Yin, L. Zhang, D. Xiang and R. Gao, Metal oxide gas sensors: Sensitivity and influencing factors, Sensors, 2010, 10, 2088–2106 CrossRef CAS PubMed .
  7. L. Camilli and M. Passacantando, Advances on sensors based on carbon nanotubes, Chemosensors, 2018, 6, 1–17 CrossRef .
  8. Y. Wang and J. T. W. Yeow, A review of carbon nanotubes-based gas sensors, J. Sens., 2009, 493904,  DOI:10.1155/2009/493904 .
  9. I. V. Zaporotskova, N. P. Boroznina, Y. N. Parkhomenko and L. V. Kozhitov, Carbon nanotubes: Sensor properties. A review, Mod. Electron. Mater., 2016, 2, 95–105 CrossRef .
  10. T. Wang, D. Huang, Z. Yang, S. Xu and G. He, A Review on Graphene-Based Gas/Vapor Sensors with Unique Properties and Potential Applications, Nano-Micro Lett., 2016, 8, 95–119 CrossRef PubMed .
  11. A. Nag, A. Mitra and S. C. Mukhopadhyay, Graphene and its sensor-based applications: A review, Sens. Actuators, A, 2018, 270, 177–194 CrossRef CAS .
  12. S. Yang, C. Jiang and S. H. Wei, Gas sensing in 2D materials, Appl. Phys. Rev., 2017, 4, 021304,  DOI:10.1063/1.4983310 .
  13. B. Cho, M. G. Hahm, M. Choi, J. Yoon, A. R. Kim, Y. Jeong, K. Nam, S. Lee, T. J. Yoo and C. G. Kang, Charge-transfer-based Gas Sensing Using Atomic-layer, Sci. Rep., 2015, 1–20 Search PubMed .
  14. G. Yang, Y. Gu, P. Yan, J. Wang, J. Xue, X. Zhang, N. Lu and G. Chen, Chemical Vapor Deposition Growth of Vertical MoS 2 Nanosheets on p-GaN Nanorods for Photodetector Application, ACS Appl. Mater. Interfaces, 2019, 11, 8453–8460 CrossRef CAS PubMed .
  15. B. L. Li, J. Wang, H. L. Zou, S. Garaj, C. T. Lim, J. Xie, N. B. Li and D. T. Leong, Low-Dimensional Transition Metal Dichalcogenide Nanostructures Based Sensors, Adv. Funct. Mater., 2016, 26, 7034–7056 CrossRef CAS .
  16. V. Babar, H. Vovusha and U. Schwingenschlögl, Density Functional Theory Analysis of Gas Adsorption on Monolayer and Few Layer Transition Metal Dichalcogenides: Implications for Sensing, ACS Appl. Nano Mater., 2019, 2, 6076–6080 CrossRef CAS .
  17. A. Abbasi and J. J. Sardroodi, Adsorption of O 3, SO 2 and SO 3 gas molecules on MoS 2 monolayers: A computational investigation, Appl. Surf. Sci., 2019, 469, 781–791 CrossRef CAS .
  18. X. Deng, X. Liang, S. P. Ng and C. M. L. Wu, Adsorption of formaldehyde on transition metal doped monolayer MoS2: A DFT study, Appl. Surf. Sci., 2019, 484, 1244–1252 CrossRef CAS .
  19. M. Nayeri, M. Moradinasab and M. Fathipour, The transport and optical sensing properties of MoS2, MoSe2, WS2 and WSe2 semiconducting transition metal dichalcogenides, Semicond. Sci. Technol., 2018, 33(2), 025002,  DOI:10.1088/1361-6641/aaa168 .
  20. H. G. Abbas, T. T. Debela, S. Hussain and I. Hussain, Inorganic molecule (O2, NO) adsorption on nitrogen- and phosphorus-doped MoS2 monolayer using first principle calculations, RSC Adv., 2018, 8, 38656–38666 RSC .
  21. L. Cabral, F. P. Sabino, M. P. Lima, G. E. Marques, V. Lopez-Richard and J. L. F. Da Silva, Azobenzene Adsorption on the MoS 2 (0001) Surface: A Density Functional Investigation within van der Waals Corrections, J. Phys. Chem. C, 2018, 122, 18895–18901 CrossRef CAS .
  22. D. Ma, B. Ma, Z. Lu, C. He, Y. Tang, Z. Lu and Z. Yang, Interaction between H2O, N2, CO, NO, NO2 and N2O molecules and a defective WSe2 monolayer, Phys. Chem. Chem. Phys., 2017, 19, 26022–26033 RSC .
  23. D. Ma, B. Ma, Z. Lu, C. He, Y. Tang, Z. Lu and Z. Yang, Interaction between H2O, N2, CO, NO, NO2 and N2O molecules and a defective WSe2 monolayer, Phys. Chem. Chem. Phys., 2017, 19, 26022–26033 RSC .
  24. J. Sun, N. Lin, H. Ren, C. Tang, L. Yang and X. Zhao, Gas adsorption on MoS 2/WS 2 in-plane heterojunctions and the I-V response: A first principles study, RSC Adv., 2016, 6, 17494–17503 RSC .
  25. H. Li, M. Huang and G. Cao, Markedly different adsorption behaviors of gas molecules on defective monolayer MoS2: A first-principles study, Phys. Chem. Chem. Phys., 2016, 18, 15110–15117 RSC .
  26. H. Luo, Y. Cao, J. Zhou, J. Feng, J. Cao and H. Guo, Adsorption of NO2, NH3 on monolayer MoS2 doped with Al, Si, and P: A first-principles study, Chem. Phys. Lett., 2016, 643, 27–33 CrossRef CAS .
  27. R. Cao, B. Zhou, C. Jia, X. Zhang and Z. Jiang, Theoretical study of the NO, NO2, CO, SO2, and NH3adsorptions on multi-diameter single-wall MoS2nanotube, J. Phys. D: Appl. Phys., 2016, 49, 045106,  DOI:10.1088/0022-3727/49/4/045106 .
  28. J. Zhang, S. Jia, I. Kholmanov, L. Dong, D. Er, W. Chen, H. Guo, Z. Jin, V. B. Shenoy, L. Shi and J. Lou, Janus Monolayer Transition-Metal Dichalcogenides, ACS Nano, 2017, 11, 8192–8198 CrossRef CAS PubMed .
  29. A. Y. Lu, H. Zhu, J. Xiao, C. P. Chuu, Y. Han, M. H. Chiu, C. C. Cheng, C. W. Yang, K. H. Wei, Y. Yang, Y. Wang, D. Sokaras, D. Nordlund, P. Yang, D. A. Muller, M. Y. Chou, X. Zhang and L. J. Li, Janus monolayers of transition metal dichalcogenides, Nat. Nanotechnol., 2017, 12, 744–749 CrossRef CAS PubMed .
  30. R. Chaurasiya, G. K. Gupta and A. Dixit, Ultrathin Janus WSSe buffer layer for W(S/Se)2 absorber based solar cells: A hybrid, DFT and macroscopic, simulation studies, Sol. Energy Mater. Sol. Cells, 2019, 201, 110076 CrossRef CAS .
  31. D. Er, H. Ye, N. C. Frey, H. Kumar, J. Lou and V. B. Shenoy, Prediction of Enhanced Catalytic Activity for Hydrogen Evolution Reaction in Janus Transition Metal Dichalcogenides, Nano Lett., 2018, 18(6), 3943–3949,  DOI:10.1021/acs.nanolett.8b01335 .
  32. C. Jin, X. Tang, X. Tan, S. C. Smith, Y. Dai and L. Kou, A Janus MoSSe monolayer: a superior and strain-sensitive gas sensing material, J. Mater. Chem. A, 2019, 7, 1099–1106 RSC .
  33. R. Peng, Y. Ma, B. Huang and Y. Dai, Two-dimensional Janus PtSSe for photocatalytic water splitting under the visible or infrared light, J. Mater. Chem. A, 2019, 7, 603–610 RSC .
  34. M. M. Idrees, H. U. Din, R. Ali, G. Rehman, T. Hussian, C. Nguyen, I. Ahmad and B. Amin, Optoelectronic and solar cell applications of janus monolayers and their van der Waals heterostructure, Phys. Chem. Chem. Phys., 2019, 21, 18612–18621 RSC .
  35. X. Ma, X. Wu, H. Wang and Y. Wang, A Janus MoSSe monolayer: A potential wide solar-spectrum water-splitting photocatalyst with a low carrier recombination rate, J. Mater. Chem. A, 2018, 6, 2295–2301 RSC .
  36. S.-D. Guo and J. Dong, Biaxial strain tuned electronic structures and power factor in Janus Transition Metal Dichalchogenide monolayers, Semicond. Sci. Technol., 2018, 419, 545–553 Search PubMed .
  37. Z. Guan, S. Ni and S. Hu, Tunable Electronic and Optical Properties of Monolayer and Multilayer Janus MoSSe as a Photocatalyst for Solar Water Splitting: A First-Principles Study, J. Phys. Chem. C, 2018, 122, 6209–6216 CrossRef CAS .
  38. R. Peng, Y. Ma, S. Zhang, B. Huang and Y. Dai, Valley Polarization in Janus Single-Layer MoSSe via Magnetic Doping, J. Phys. Chem. Lett., 2018, 9, 3612–3617 CrossRef CAS PubMed .
  39. C. Shang, X. Lei, B. Hou, M. Wu, B. Xu, G. Liu and C. Ouyang, Theoretical Prediction of Janus MoSSe as a Potential Anode Material for Lithium-Ion Batteries, J. Phys. Chem. C, 2018, 122(42), 23899–23909,  DOI:10.1021/acs.jpcc.8b07478 .
  40. H. Cai, Y. Guo, H. Gao and W. Guo, Tribo-piezoelectricity in Janus transition metal dichalcogenide bilayers: a first-principles study, Nano Energy, 2018, 56, 33–39 CrossRef .
  41. R. Li, Y. Cheng and W. Huang, Recent Progress of Janus 2D Transition Metal Chalcogenides: From Theory to Experiments, Small, 2018, 14, 1802091 CrossRef PubMed .
  42. N. N. Som and P. K. Jha, Hydrogen evolution reaction of metal di-chalcogenides: ZrS2, ZrSe2 and Janus ZrSSe, Int. J. Hydrogen Energy, 2019 DOI:10.1016/j.ijhydene.2019.09.033 .
  43. J. Wang, H. Guo, J. Xue, D. Chen, G. Yang, B. Liu, H. Lu, R. Zhang and Y. Zheng, Janus Ga2SeTe: A Promising Candidate for Highly Efficient Solar Cells, Sol. RRL, 2019, 3, 1–5 Search PubMed .
  44. L. Dong, J. Lou and V. B. Shenoy, Large In-Plane and Vertical Piezoelectricity in Janus Transition Metal Dichalchogenides, ACS Nano, 2017, 11, 8242–8248 CrossRef CAS PubMed .
  45. Y. Li, J. Wang, B. Zhou, F. Wang, Y. Miao, J. Wei, B. Zhang and K. Zhang, Tunable interlayer coupling and Schottky barrier in graphene and Janus MoSSe heterostructures by applying an external field, Phys. Chem. Chem. Phys., 2018, 20, 24109–24116 RSC .
  46. H. Yang, P. Zhao, Y. Ma, X. Lv, B. Huang and Y. Dai, Janus single-layer group-III monochalcogenides: A promising visible- light photocatalyst, J. Phys. D: Appl. Phys., 2019, 52, 22408–22418 Search PubMed .
  47. S. Wang, A. Robertson and J. H. Warner, Atomic structure of defects and dopants in 2D layered transition metal dichalcogenides, Chem. Soc. Rev., 2018, 47, 6764–6794 RSC .
  48. H. Taghinejad, D. A. Rehn, C. Muccianti, A. A. Eftekhar, M. Tian, T. Fan, X. Zhang, Y. Meng, Y. Chen, T. V. Nguyen, S. F. Shi, P. M. Ajayan, J. Schaibley, E. J. Reed and A. Adibi, Defect-Mediated Alloying of Monolayer Transition-Metal Dichalcogenides, ACS Nano, 2018, 12, 12795–12804 CrossRef CAS PubMed .
  49. R. Kumar, N. Goel, A. V. Agrawal, R. Raliya, S. Rajamani, G. Gupta, P. Biswas, M. Kumar and M. Kumar, Boosting Sensing Performance of Vacancy-Containing Vertically Aligned MoS2 Using rGO Particles, IEEE Sens. J., 2019, 19, 10214–10220 CAS .
  50. R. Chaurasiya and A. Dixit, Defect engineered MoSSe Janus monolayer as a promising two dimensional material for NO2 and NO gas sensing, Appl. Surf. Sci., 2019, 490, 204–219 CrossRef CAS .
  51. S. Equations and C. Effects, Self-Consistent Equations Including Exchange and Correlation Effects, Phys. Rev., 1965, 140, 1133–1138 CrossRef .
  52. P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni, I. Dabo, A. Dal Corso, S. De Gironcoli, S. Fabris, G. Fratesi, R. Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari, F. Mauri, R. Mazzarello, S. Paolini, A. Pasquarello, L. Paulatto, C. Sbraccia, S. Scandolo, G. Sclauzero, A. P. Seitsonen, A. Smogunov, P. Umari and R. M. Wentzcovitch, QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials, J. Phys.: Condens. Matter, 2017, 29, 465901,  DOI:10.1088/0953-8984/21/39/395502 .
  53. Y. H. Dai, Convergence properties of the BFGS algoritm, SIAM J. Optim., 2003, 13, 693–701 CrossRef .
  54. M. Perdew, J. P. Burke and K. Ernzerhof, Generalized Gradient Approximation Made Simple, Phys. Rev. Lett., 1996, 77, 3865–3868 CrossRef PubMed .
  55. J. D. Pack and H. J. Monkhorst, ‘Special points for Brillouin-zone integrations’-a reply, Phys. Rev. B: Solid State, 1977, 16, 1748–1749 CrossRef .
  56. G. Henkelman, A. Arnaldsson and H. Jónsson, A fast and robust algorithm for Bader decomposition of charge density, Comput. Mater. Sci., 2006, 36, 354–360 CrossRef .
  57. G. Lehmann and M. Taut, On the Numerical Calculation of the Density of States and Related Properties, Phys. Status Solidi, 1972, 54, 469–477 CrossRef CAS .
  58. S. Grimme, Semiempirical GGA-Type Density Functional Constructed with a Long-Range Dispersion Correction, J. Comput. Chem., 2006, 27, 1787–1799 CrossRef CAS PubMed .
  59. J. B. A. Davis, F. Baletto and R. L. Johnston, The Effect of Dispersion Correction on the Adsorption of CO on Metallic Nanoparticles, J. Phys. Chem. A, 2015, 119, 9703–9709 CrossRef CAS PubMed .
  60. S. Zhao, J. Xue and W. Kang, Gas adsorption on MoS2 monolayer from first-principles calculations, Chem. Phys. Lett., 2014, 595–596, 35–42 CrossRef CAS .
  61. H. Yang, Y. Liu, C. Gao, L. Meng, Y. Liu, X. Tang and H. Ye, Adsorption Behavior of Nucleobases on Doped MoS2 Monolayer: A DFT Study, J. Phys. Chem. C, 2019, 123, 30949–30957 CrossRef CAS .
  62. R. Chaurasiya and A. Dixit, Point defects induced magnetism in CdO monolayer: A theoretical study, J. Magn. Magn. Mater., 2018, 469, 279–288 CrossRef .
  63. M. S. Khan, A. Srivastava, R. Chaurasiya, M. S. Khan and P. Dua, NH3 and PH3 adsorption through single walled ZnS nanotube: First principle insight, Chem. Phys. Lett., 2015, 636, 103–109 CrossRef CAS .
  64. P. Ou, P. Song, X. Liu and J. Song, Superior Sensing Properties of Black Phosphorus as Gas Sensors: A Case Study on the Volatile Organic Compounds, Adv. Theory Simul., 2019, 2, 1800103 CrossRef .
  65. M. Samadizadeh, A. A. Peyghan and S. F. Rastegar, Sensing behavior of BN nanosheet toward nitrous oxide: A DFT study, Chin. Chem. Lett., 2015, 26, 1042–1045 CrossRef CAS .
  66. B. Roondhe, K. Patel and P. K. Jha, Two-dimensional metal carbide comrade for tracing CO and CO2, Appl. Surf. Sci., 2019, 496, 143685 CrossRef CAS .
  67. S. Peng, K. Cho, P. Qi and H. Dai, Ab initio study of CNT NO2 gas sensor, Chem. Phys. Lett., 2004, 387, 271–276 CrossRef CAS .
  68. R. Chaurasiya, A. Dixit and R. Pandey, Strain-mediated stability and electronic properties of WS 2, Janus WSSe and WSe 2 monolayers, Superlattices Microstruct., 2018, 122, 268–279 CrossRef CAS .
  69. S. Salehi and A. Saffarzadeh, Atomic defect states in monolayers of MoS2and WS2, Surf. Sci., 2016, 651, 215–221 CrossRef CAS .
  70. R. Addou, L. Colombo and R. M. Wallace, Surface Defects on Natural MoS2, ACS Appl. Mater. Interfaces, 2015, 7, 11921–11929 CrossRef CAS PubMed .
  71. X. Zhao, X. Dai, C. Xia and T. Wang, Superlattices and Microstructures Structural defects in pristine and Mn-doped monolayer WS 2: A first-principles study, Superlattices Microstruct., 2015, 85, 339–347 CrossRef CAS .
  72. H. Zheng, B. Yang, D. Wang, R. Han, X. Du and Y. Yan, Tuning magnetism of monolayer MoS2 by doping vacancy and applying strain, Appl. Phys. Lett., 2014, 104, 1–6 Search PubMed .
  73. J. Hong, Z. Hu, M. Probert, K. Li, D. Lv, X. Yang, L. Gu, N. Mao, Q. Feng, L. Xie, J. Zhang, D. Wu, Z. Zhang, C. Jin, W. Ji, X. Zhang, J. Yuan and Z. Zhang, Exploring atomic defects in molybdenum disulphide monolayers, Nat. Commun., 2015, 6, 1–8 Search PubMed .
  74. Q. Yue, Z. Shao, S. Chang and J. Li, Adsorption of gas molecules on monolayer MoS2 and effect of applied electric field, Nanoscale Res. Lett., 2013, 8, 425 CrossRef PubMed .
  75. T. Wang, R. Zhao, X. Zhao, Y. An, X. Dai and C. Xia, Tunable donor and acceptor impurity states in a WSe2 monolayer by adsorption of common gas molecules, RSC Adv., 2016, 6, 82793–82800 RSC .
  76. G. Yang, P. Yan, C. Zhu, Y. Gu, N. Lu, J. Xue, X. Zhang, R. Sun and X. Fang, Selenium Vacancy–Enhanced Gas Adsorption of Monolayer Hafnium Diselenide (HfSe 2) from a Theoretical Perspective, Adv. Theory Simul., 2019, 2, 1900052 CrossRef .
  77. O. Leenaerts, B. Partoens and F. M. Peeters, Adsorption of H2 O, N H3, CO, N O2, and NO on graphene: A first-principles study, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 77, 1–6 Search PubMed .
  78. M. Pashangpour, Electronic Structure Calculation of Adsorbate Gas Molecules on a BN Nanolayer, a First Principle Study, Int. Proc. Chem. Biol. Environ. Eng., 2011, 2, 113–116 Search PubMed .
  79. J. Wang, J. M. Lei, G. F. Yang, J. J. Xue, Q. Cai, D. J. Chen, H. Lu, R. Zhang and Y. D. Zheng, An ultra-sensitive and selective nitrogen dioxide sensor based on a novel P2C2 monolayer from a theoretical perspective, Nanoscale, 2018, 10, 21936–21943 RSC .
  80. J. Wang, G. F. Yang, J. J. Xue, J. M. Lei, D. J. Chen, H. Lu, R. Zhang and Y. D. Zheng, A Reusable and High Sensitivity Nitrogen Dioxide Sensor Based on Monolayer SnSe, IEEE Electron Device Lett., 2018, 39, 599–602 CAS .

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/d0cp02063j

This journal is © the Owner Societies 2020