PH₃ gas adsorption on S and Mo vacancy MoS₂ monolayer: a first principle study

Abstract

The sensing nature and change in the electron transport behavior of S vacancy armchair MoS₂ (AmS-MoS₂), Mo vacancy armchair MoS₂ monolayer (AmMo-MoS₂), S vacancy zigzag MoS₂ (ZigS-MoS₂), and zigzag Mo vacancy MoS₂ (ZigMo-MoS₂) monolayer before and after PH₃ adsorption were theoretically investigated using Density Functional Theory (DFT) in combination with Non-Equilibrium Greens Function (NEGF) based on first principle calculations. To study the feasibility of armchair and zigzag MoS₂ device as PH₃ gas sensor, we conducted an analysis of the changes in the geometrical structures, density of states (DOS), transmission spectrum, and I–V characteristic. Our results predicted that PH₃ adsorption on all four devices is through van der Waals interactions. Among the four devices, AmMo-MoS₂ shows enhanced adsorption behavior with the adsorption energy − 1.8048 eV and charge transfer of − 0.2120e. The I–V characteristic of AmMo-MoS₂ shows a significant change in the conductivity compared with the other devices. Thus, our work concluded that AmMo-MoS₂ is considered to be a better device for PH₃ adsorption compared with the other devices.

Introduction

Phosphine (PH₃) is a highly poisonous gas that is commonly used in semiconductor industries and for fumigating grains. PH₃ gas molecules are exhausted during the production of acetylene and flame-retardant industries. The emission of PH₃ gas is very harmful to human beings, which causes cancer, headache, vomiting, fatigue, and even cause damage to the heart [1,2,3,4]. Therefore, rapid and precise sensing and monitoring of PH₃ gas molecules play an essential role in prevention.

Two-dimensional materials (2D) have a tremendous attractive interest due to their unique and extraordinary mechanical, physical, and chemical properties [5]. It has been considered as a suitable material for various potential applications. It is also considered as flexible material for next-generation optoelectronic devices, electronic devices, and gas sensors [6,7,8]. For the past few years, the researchers found that graphene has earned much more attention for its unique properties, and it has been used in many applications. However, the absence of bandgap in graphene has limited their progress [6, 9]. Transition metal dichalcogenides (TMDs) are thin semiconductors of type MX₂ [10], where M is the transition metal element from groups IV, V, or VI and X represents the chalcogen elements like S, Se, and Te [11]. 2D materials have some attractive and interesting features, including high carrier mobility, high surface to volume ratio, high chemical stability, high thermal stability, low electronic temperature noise and fast response time [12, 13], and low-cost effect [14]. Materials with these characteristics are considered to be ideal for sensing applications [11, 15]. Among TMDs, MoS₂ (molybdenum disulfide) is considered one of the most suitable materials for electronic device and sensor applications due to its excellent electrical and mechanical properties [14] and tunable bandgap when compared with graphene [12, 16]. The crystal structure of MoS₂ consists of a weakly coupled S-Mo-S sandwich layer. The structure of MoS₂ monolayer can be fabricated by the micromechanical cleavage or exfoliation method [9, 17, 18].

Shokri and Salami analyzed the sensing capabilities of MoS₂ monolayer transducer with CO, CO₂ and NO gas molecules and concluded that the NO gas molecule shows more changes in the electronic properties and charge transfer while compared with CO, CO₂ gas molecules [19]. Jasmine et al. theoretically analyzed the sensing behavior of Cl₂, PH₃, AsH₃, BBr₃, and SF₄ gas molecules on Mo/S vacancy MoS₂ monolayer and concluded that the PH₃ gas molecule shows more adsorption towards S/Mo vacancy MoS₂ monolayer [10]. Ren et al. theoretically investigated the adsorption behavior of CH₃ gas molecule on S and Mo vacancy MoS₂ monolayer, and they suggested that the different vacancies have a different effect on adsorption behavior [20]. Feng et al. concluded that the material’s conductivity is increased due to the vacancy creation [21]. Zhao et al. theoretically investigated the adsorption of various gas molecules, including CO, CO₂, NH₃, NO, NO₂, CH₄, H₂O, N₂, O₂, and SO₂ on MoS₂ monolayer using DFT. The results indicate that NO and NO₂ show better adsorption than other gas molecules [22]. Wei et al. theoretically investigated the sensing behavior of Ni-doped MoS₂ monolayer towards SO₂, H₂S, and SF₆ gas molecules. The results indicate that H₂S and SO₂ tend to adsorb on the surface of Ni-MoS₂ monolayer by chemisorption, and the adsorption energy of the H₂S and SO₂ are − 1.319 eV and − 1.382 eV, respectively [23]. Kumar et al. experimentally analyzed the recent progress and remarkable development in gas sensing field by using the 2D MoS₂. They have developed various fabrication techniques for synthesizing a wide range of different nanostructures and morphologies of the MoS₂ on rigid as well as flexible substrates. They concluded that all the exciting gas sensing results of the 2D MoS₂ could be the best candidate for developing a high-performance room temperature gas sensor [24]. Chacko et al. analyzed the experimental study of Ni and Pd functionalized MoS₂ devices towards H₂S and NO. The MoS₂-based sensors showed excellent sensing performances with high sensitivity at room temperature, which can serve as an excellent alternative to the standard semiconductor metal oxide gas sensors that require high optimal working temperatures for good response [25].

The main novelty of this work is to analyze the sensing nature, adsorption behavior, and the changes in the electron transport properties of S and Mo vacancy created MoS₂ towards PH₃ gas molecule. In recent works, researchers have been using MoS₂ for gas sensing applications. But we have analyzed the sensing nature of S and Mo vacancy created MoS₂ towards PH₃ gas molecule. Moreover, we have also constructed the armchair and zigzag device with two electrodes model and analyzed the changes in the electron transport properties using DFT combined with Non-Equilibrium Green’s Function (NEGF) for PH₃ gas adsorption.

Computational details

A modeled structure of the MoS₂ device with S and Mo vacancy is represented in Fig. 1. We have constructed the MoS₂ device with 50 Mo atoms and 100 S atoms. The device consists of three parts, i.e., the left and right electrodes and central scattering region. The size of the central region is 19.55 Å (the central region is long enough to study the adsorption behavior of the gas molecule), and the size of the electrode is 3.16 Å which is chosen in such a way to study the effect of adsorption between the MoS₂ monolayer and the PH₃ gas molecule.

Fig. 1

Optimized structure of a AmS-MoS₂ device, b AmMo-MoS₂ device, c ZigS-MoS₂ device, and d ZigMo-MoS₂ device and e PH₃ gas molecule

The electron transport properties for adsorption effects of PH₃ gas molecule on S and Mo vacancy MoS₂ monolayer are performed using DFT combined with NEGF [26]. The empirical correction DFT + D2 has been used to correct the effect of van der Waals interaction [27]. We utilized the virtual NanoLab simulation tool for constructing the MoS₂ device and the QuantumWise Atomistix Toolkit (ATK) package for performing the DFT calculations [27, 28].

The optimized geometry of Am-MoS₂ and Zig-MoS₂ with S and Mo vacancy is represented in Fig. 1. For optimization, we have used Linear Combination of Atomic Orbitals (LCAO) as basic set and Double Zeta plus polarization to solve the Kohn–Sham equations. For the geometrical optimization, we have used generalized gradient approximation (GGA) as an exchange–correlation function with Perdew–Burke–Ernzerhof (PBE) functional [19]. For energy tolerance, we have set the convergence criteria of 1.0 × 10⁻⁵ and the maximum force of 0.002 Ha/Å and 0.005 Å for the displacement of geometrical optimization. For accuracy calculation, the cut-off ratio has been set as 5.0 Å in the real space grid [27]. Zone integration is sampled with 2 × 1 × 100 grid mesh along x-, y-, and z-directions, where the electron transport is along the z-direction. For relaxation calculation, the lattice parameters were set as a = b = 12.66 Å and c = 20.00 Å [10, 27]. The atomic positions of all the geometries were fully optimized until the force on each atom becomes less than 0.05 eV/Å [10, 27, 29]. The temperature of the electron is taken as 300 K throughout the calculations [29,30,31]. The transmission spectrum was calculated using the transmission function at a particular bias. The transmission function T(E,V) is given:

$$T\left( {E,V} \right) = Tr\left[ {\Gamma_{{\text{L}}} \left( {E,V} \right)G^{{\text{R}}} \left( E \right)\Gamma_{{\text{R}}} \left( {E,V} \right)G^{{\text{A}}} \left( E \right)} \right]$$

(1)

Here, $\Gamma_{{\text{L}}}$ and $\Gamma_{{\text{R}}}$ represent the contact broadening functions of the left and right electrodes respectively. G^R and G^A represent the retarded advance Green’s function. The current I(V) can be calculated using T(E,V), and it can be written as follows:

$$I\left( V \right) = \frac{{2e^{2} }}{h}\int_{{\mu_{{\text{L}}} }}^{{\mu_{{\text{R}}} }} {\left[ {f\left( {E - \mu_{{\text{L}}} } \right) - f\left( {E - \mu_{{\text{R}}} } \right)} \right]} \,T\left( {E,V} \right)dE$$

(2)

where e, h, f, and E represented the electron charge, Plank’s constant, Fermi function, energy respectively. $\mu_{{\text{L}}}$ and $\mu_{{\text{R}}}$ are the chemical potentials of the left and right electrodes.

The adsorption energy between the PH₃ gas molecule and armchair and zigzag MoS₂ monolayer with S and Mo vacancy is defined as follows:

$$E_{{{\text{ads}}}} = E_{{{\text{PH}}_{{3}} {\text{V}}_{{{{\text{S}} \mathord{\left/ {\vphantom {{\text{S}} {{\text{MO}}}}} \right. \kern-0pt} {{\text{MO}}}}}} {\text{MoS}}_{{2}} }} - \left( {E_{{{\text{V}}_{{\text{S/MO}}} {\text{MoS}}_{{2}} }} + E_{{{\text{PH}}_{{3}} }} } \right)$$

(3)

where $E_{{{\text{PH}}_{{3}} {\text{V}}_{{{{\text{S}} \mathord{\left/ {\vphantom {{\text{S}} {{\text{MO}}}}} \right. \kern-0pt} {{\text{MO}}}}}} {\text{MoS}}_{{2}} }}$ represents the total energy of S/Mo vacancy MoS₂ device after PH₃ adsorption on. $E_{{{\text{V}}_{{\text{S/MO}}} {\text{MoS}}_{{2}} }}$ represents the total energy of S/Mo vacancy created MoS₂ device. $E_{{{\text{PH}}_{{3}} }}$ represents the total energy of the gas molecule. We have used Mulliken population analysis for charge transfer (Q) calculation. The difference between the actual valence of S/Mo atom and the valency charge obtained from Mulliken population analysis gives the charge of each and every individual atom in V_S and V_Mo MoS₂ monolayer. The net charge transfer of the system is calculated by the summation of all the differences obtained from Mulliken population analysis. From the net total, the negative sign indicated the charge transfer from V_S/V_Mo MoS₂ monolayer to PH₃ gas molecule and the positive value indicated the charge transfer from the PH₃ gas molecule to the V_S/V_Mo MoS₂ monolayer [10, 29] which are represented in Table 1. To get a deeper understanding of the electronic and sensing property, the changes caused in the charges after the adsorption of gas molecule are analyzed. In Fig. 1, the bond length between Mo atom and S atom is 2.415 Å, and between two S is 3.131 Å.

Table 1 The values of adsorption distance (d), adsorption energy (E_ads), charge transfer (Q), and changes in the bond length and bond angle

Results and discussion

To understand the adsorption effects of PH₃ gas molecule towards MoS₂ device, different adsorption configurations are calculated. The results include DOS, transmission spectrum, I-V curve, adsorption energy, and charge transfer.

Figure 2a–d shows the optimized structure of PH₃ adsorption on armchair S vacancy MoS₂ monolayer (AmS-MoS₂PH₃), PH₃ adsorption on armchair Mo vacancy MoS₂ monolayer (AmMo-MoS₂PH₃), PH₃ adsorption on zigzag S vacancy MoS₂ monolayer (ZigS-MoS₂PH₃), and PH₃ adsorption on zigzag Mo vacancy MoS₂ monolayer (ZigMo-MoS₂PH₃), respectively. In order to get a more stable configuration, the optimized PH₃ gas molecule is placed vertically above the vacancy created on the MoS₂ monolayer. From Fig. 2, we observed that there is a non-bonding interaction between the device and the gas molecule. This shows that the adsorption is through physisorption [32, 33]. After the adsorption, the structure of the PH₃ gas molecule dislocates, where the bond length P–H and bond angle HPH have been increased and decreased, and the changes are listed in Table 1. The changes in the bond length and bond angles are caused due to the van der Waals interaction [33]. From Table 1, we observed that the PH₃ adsorption on AmMo-MoS₂ and ZigMo-MoS₂ shows more change in the bond length and bond angle. For AmMo-MoS₂PH₃, the changes in the bond length are 1.4428 Å, 1.4439 Å, and 1.4433 Å for P-H₁,P-H₂, and P-H₃ (Fig. 1e), and the changes in the bond angles are 98.17°, 98.19°, and 98.32° for H₁PH₂, H₂PH₃, and H₃PH₁, respectively. Similarly, for ZigMo-MoS₂-PH₃ the changes in the bond length are 1.4503 Å, 1.5142 Å, and 1.4532 Å, and the changes in the bond angle are 93.23°, 114.30°, and 85.59°, respectively. The adsorption distance, adsorption energy, and charge transfer of the four devices are also listed in Table 1. From the Table 1, AmMo-MoS₂ and ZigMo-MoS₂ show less adsorption distance of 2.3420 Å and 2.3626 Å, more adsorption energy − 1.1048 eV and − 0.3522 eV, and more charge transfer of 0.112e and − 0.18e, respectively. From this, we observed that the PH₃ adsorption on AmMo-MoS₂ and ZigMo-MoS₂ is comparatively more.

Fig. 2

Optimized structure of a AmS-MoS₂PH₃ device, b AmMo-MoS₂PH₃ device, c ZigS-MoS₂PH₃ device, and d ZigMo-MoS₂PH₃ device

Figure 3 shows the TDOS curve of PH₃ adsorption on AmS-MoS₂, AmMo-MoS₂, ZigS-MoS₂, and ZigMo-MoS₂. The Fermi Energy (E_F) is set to 0 eV, and this represents the zero carrier density at the Fermi region. All the four systems have not produced any changes in the bandgap after the adsorption of PH₃ gas molecule; this shows that adsorption does not introduce any mid-gap states, but near the Fermi region, some states of the AmS/Mo-MoS₂ has been changed after the PH₃ adsorption. This leads to a change in the electrical conductivity. For PH₃ adsorption on ZigS and ZigMo-MoS₂ (Fig. 3c, d), no changes are observed near the Fermi region but ZigS/Mo-MoS₂ shows a slight change (− 2.0 to − 0.5 eV) after the adsorption of PH₃ gas molecule. This indicates that the adsorption of PH₃ gas molecule considerably affects the electrical conductivity of the AmS/Mo-MoS₂ when compared with ZigS/Mo-MoS₂.

Fig. 3

TDOS of a AmS-MoS₂ and b AmMo-MoS₂ and c ZigS-MoS₂ and d ZigMo-MoS₂ before and after adsorption of PH₃ gas molecule

Figure 4a–d shows the projected density of states (PDOS) of AmMo-MoS₂, AmS-MoS₂, ZigS-MoS₂, and ZigMo-MoS₂ before and after the adsorption of PH₃ gas molecule. From Fig. 4, we observed that the PH₃ gas molecule affects the p and d orbitals of all the four systems, thus causes the changes in the TDOS after the adsorption on PH₃ gas molecule. Near the Fermi region, most of the peaks of s and d orbitals have been changed after the adsorption of PH₃ gas molecule. This indicates that there is a significant charge transfer occurred between PH₃ gas molecule and MoS₂ device and no bond formation between the gas molecule and MoS₂ device. This indicates that the changes in the peaks are caused due to van der Waals’s force between the gas molecule and the system [10]. Due to these changes, the electrical conductivity of the system was changed.

Fig. 4

PDOS of a AmS-MoS₂ and b AmMo-MoS₂ and c ZigS-MoS₂ and d ZigMo-MoS₂ before and after adsorption of PH₃ gas molecule

The transmission spectrum of V_S and V_MO on the armchair and zigzag MoS₂ monolayer with and without PH₃ gas molecule are illustrated in Fig. 5a–d. From the figure, we observed that at zero bias, there is zero transmission coefficient and the width of the transmission gap are about 0.22 eV and 0.17 eV for AmS-MoS₂ and AmMo-MoS₂ and 0.82 eV and 0.81 eV for ZigS-MoS₂ and ZigMo-MoS₂, respectively, and the transmission gap acts as a barrier for electron transmission. This shows that the material has semiconducting nature [34]. For AmS/Mo-MoS₂ (Fig. 5a, b), we observed that there is a decrease in the transmission after the adsorption of PH₃ gas molecule. The peaks in the transmission spectrum indicated the conduction channels. The reduction in the transmission peaks leads to the reduction in the current. This indicates that AmS/Mo-MoS₂ is considerably affected by the PH₃ gas molecule. For ZigS/Mo-MoS₂, the changes in the transmission are comparatively lower than AmS/Mo-MoS₂.

Fig. 5

Transmission spectrum of a AmS-MoS₂ and b AmMo-MoS₂ and c ZigS-MoS₂ and d ZigMo-MoS₂ before and after adsorption of PH₃ gas molecule

To clearly observe the modification in the conductivity and to qualitatively evaluate the performance of the S and Mo vacancy armchair and zigzag MoS₂ monolayer as a PH₃ sensor, the I–V characteristics of the system were analyzed, and their respective I–V graph is shown in Fig. 6. Figure 6a shows the I–V characteristics of AmS-MoS₂ with and without PH₃ adsorption. It exhibits a non-linear behavior before and after the adsorption of PH₃ gas molecule. Figure 6a shows a linear increase in the current for the bias voltage of about 0.6 to 1.2 V; the increase in the current (I_peak) reaches a maximum value of 0.1497 mA before the adsorption of PH₃ and 0.1192 mA after the adsorption of PH₃. This shows that the I_peak decreases after the adsorption of the PH₃ gas molecule. Beyond the bias voltage of 1.2 to 1.6 V, there is a rapid decrease in the current, and the non-differential resistance (NDR) phenomena were observed [35, 36]. The decrease in the current (I_vally) reaches the minimum current value of 0.7345 mA before and 0.2450 mA after the adsorption of PH₃ gas molecule. This shows that there is a change in the conductivity of the material after the adsorption of PH₃ gas molecule. For PH₃, adsorption on AmMo-MoS₂ shows an increase in the current for the bias voltage of about 0.6 to 1.2 V before adsorption and 0.6 to 1.0 V after the adsorption. Here, the I_valley starts from 1.2 V for AmMo-MoS₂ and 1.0 V for AmMo-MoS₂PH₃; after 1.0 V, there is a reduction in current for AmMo-MoS₂ PH₃ compared with AmMo-MoS₂. The current reduction indicated the increase in the resistance of the AmS/Mo-MoS₂ PH₃ material after the adsorption of PH₃ gas molecule.

Fig. 6

I–V characteristics of a AmS-MoS₂ and b AmMo-MoS₂ and c ZigS-MoS₂ and d ZigMo-MoS₂ before and after adsorption of PH₃ gas molecule

Figure 6c, d represents the I–V characteristic of PH₃ adsorption on ZigS/Mo-MoS₂ monolayer. The magnitude of the current along the ZigS/Mo-MoS₂ is smaller than AmS/Mo-MoS₂. For PH₃ adsorption on ZigS-MoS₂, there is no significant changes in the current between before and after the adsorption of PH₃ gas molecule. This shows that PH₃ gas molecule does not causes any change in the conductivity of the material. For PH₃ adsorption on ZigMo-MoS₂, the current flow is zero until the bias voltage is 0.9 V before the absorption of PH₃ and 1.0 V after the adsorption of PH₃ gas molecule. After this, the current starts increasing dramatically. Under the bias voltage of 2.0 V, the current flow through the material is 0.017 mA before and 0.014 mA after the adsorption of PH₃ gas molecule. This shows that the adsorption of PH₃ gas molecule is slightly more in ZigMo-MoS₂ when compared with the ZigS-MoS₂. The conductance for the armchair S/Mo vacancy MoS₂ monolayer is non-linear. The changes in the conductance before and absence of PH₃ gas molecule attribute to the response of the sensor. Therefore, for estimating the sensor response, the selectivity of the sensor is calculated using the following:

$$S = \left| {G - G_{0} } \right|/G_{0}$$

(4)

where G and G₀ represent the conductance of Am/ZigS/Mo-MoS₂ after and before the gas adsorption of PH₃ gas molecule, respectively [27].

The estimated percentage of sensitivity values of the system after the adsorption of PH₃ gas molecule is listed in Table 2. Here, all the system shows different sensitivity at different bias voltage. From Table 2, we observed that the PH₃ adsorption on AmMo-MoS₂ shows more adsorption energy, charge transfer, and percentage of sensitivity of − 1.8048 eV, − 0.2120e, and 96.87% under the bias voltage of 1.8 V when compared with AmS-MoS₂, ZigS-MoS₂, and ZigMo-MoS₂ [13].

Table 2 The values of percentage of sensitivity of PH₃ gas molecule on AmS-MoS₂, AmMo-MoS₂, ZigS-MoS₂, and ZigMo-MoS₂ device under voltage from 0 to 2.0 V

Conclusion

In this study, we have investigated the sensing behavior and electron transport property of S/Mo vacancy armchair and zigzag MoS₂ monolayer using NEGF-DFT techniques. The result shows that the PH₃ gas molecule is allowed to be adsorbed in all the four systems through van der Waals interaction. The structural optimization results indicate that the changes in the bond length, bond angle, and the value of the adsorption energy and charge transfer of PH₃ adsorption on AmMo-MoS₂ are more when compared with AmS-MoS₂, ZigS-MoS₂, and ZigMo-MoS₂. PH₃ adsorption on AmMo-MoS₂ shows better adsorption. Moreover, the changes in the PDOS and transmission spectrum are comparatively more in AmS/Mo-MoS₂ which leads to more changes in the I–V curve of AmS/Mo MoS₂, and ZigMo-MoS₂ indicated that there will be a change in the conductivity of the material after the adsorption of PH₃ gas molecule. Moreover, the NDR behavior was observed in both AmMo/S-MoS₂. AmMo-MoS₂ shows more changes in the current at I_vally after the adsorption of PH₃ gas molecule. Thus, we concluded that AmMo-MoS₂ shows better adsorption towards PH₃ gas molecule when compared with the other devices.