Dft study on gas adsorption of dichalcogenide monolayers towards gas sensing applications

Summary of the adsorption energy, adsorption distance and response length for toxic gases adsorption on monolayer MoTe2 .... The favorable configurations of CO2-MoTe2 top view a and sid

`

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

NGUYEN HOANG HUNG

DFT STUDY ON GAS ADSORPTION OF DICHALCOGENIDE MONOLAYERS: TOWARDS GAS SENSING APPLICATION

MASTER'S THESIS

`

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

NGUYEN HOANG HUNG

DFT STUDY ON GAS ADSORPTION OF DICHALCOGENIDE MONOLAYERS: TOWARDS GAS SENSING APPLICATION

MAJOR: NANOTECHNOLOGY CODE: 8440140.11QTD

RESEARCH SUPERVISORS:

Specially Appointed Professor, Dr DINH VAN AN

Prof Dr YOJI SHIBUTANI

Hanoi, 2021

`

ACKNOWLEDGMENT

First of all, I would like to express my special thanks to my supervisors

Specially Appointed Professor, Dr Dinh Van An and Prof Dr Yoji Shibutabi

Thank you for all your thorough and supportive instructions, your courtesy, and your encouragement This thesis absolutely could not be conducted well without your dedicated concerns

Second of all, I would like to show my gratitude to the MNT program and VNU Vietnam Japan University (VJU) professors, lecturers and staff for all of your

support

Third of all, I would like to express my warm thanks to my seniors, Pham Trong Lam, Ta Thi Luong, Pham Ba Lich for significant advices from my first step

in VJU to the last days

Last but not least, I would like to send thanks from the bottom of my heart to

Japan International Cooperation Agency (JICA) and VJU for providing me with

financial support during my study in VJU

1.2 Two-dimensional (2D) material-based toxic gas sensors 4 1.3 2D Transition Metal Dichalcogenides (TMDCs) and monolayer MoTe2 7

`

i

LIST OF TABLES

Table 1.1 Adsorption characteristics of toxic gases on some typical 2D materials 5

Table 1.2 Adsorption characteristics of common gases on MoTe2 monolayer 11

Table 3.1 The structure parameters of monolayer MoTe2 21

Table 3.2 The band gap and spin orbit splitting values comparting by non-empirical vdW functionals 22

Table 3.3 The values of bond-length (l), the distance between the lowest atom of molecule and the nearest atom of absorbent (da), the angle between molecule and the absorbent surface () and the bond angle () of favorable configurations 28

Table 3.4 Summary of the adsorption energy, adsorption distance and response length for toxic gases adsorption on monolayer MoTe2 34

Table 3.5 Bader charge analysis of the CO – MoTe2 system 35

Table 3.6 Bader charge analysis of the CO2 – MoTe2 system 37

Table 3.7 Bader charge analysis of the NO – MoTe2 system 39

Table 3.8 Bader charge analysis of the NO2 – MoTe2 system 41

Table 3.9 Bader charge analysis of the NH3 – MoTe2 system 43

`

ii

LIST OF FIGURES

Figure 1.1 The molecular structure of CO2 2

Figure 1.2 The molecular structure of CO 2

Figure 1.3 The molecular structure of NO2 3

Figure 1.4 The molecular structure of NO 3

Figure 1.5 The molecular structure of NH3 4

Figure 1.6 Illustration of the charge transfer mechanism of MoS2 to gas molecules 5

Figure 1.7 Illustration some synthesis methods for few-layer TMDCs (a) MBE (b) CVD (c) Metal-organic CVD 9

Figure 1.8 The illustration of MoTe2 FET sensor (a) Optical microscope image (b) AFM topography image d) High resolution TEM image (e) SAED pattern taken were reported by Feng’s group……….9

Figure 2.1 DFT computational scheme 16

Figure 3.1 (a) side view (b) top view of the hexagonal monolayer MoTe2 and (c) the considered adsorption sites on monolayer MoTe2 20

Figure 3.2 The Band structure and DOS of the monolayer MoTe2 with (a) and without (b) SOC taken into account 23

Figure 3.3 The C≡O bond-length (l), the distance between C atom and the nearest atom of absorbent (da), the angle between molecules and the absorbent surfaces () of CO-MoTe2 favorable configurations 24

Figure 3.4 The favorable configurations of CO2-MoTe2 top view (a) and side view (b) with the adsorption distance (d), the C=O bond length (l), the distance between C atom and the nearest Te atom of absorbent (da), the bond angle () 25

Figure 3.5 Top view (a) and side view of the favorable configurations of NO-MoTe2 (b) with the adsorption distance (d), the N O bond length (l), the distance between N atom and the nearest atom of absorbent (da), the angle between molecules and the absorbent surfaces () 25

Figure 3.6 Top view (a) and side view of the favorable configurations of NO2-MoTe2 (b) with the adsorption distance (d), the N O bond length (l), the distance between O atom and the nearest Te atom of absorbent (da), the bond angle O N O () 26

Figure 3.7 The favorable configurations of NH3-MoTe2 top view (a) and side view (b) with the adsorption distance (d), the N-H bond length (l), the distance between C atom and the nearest Te atom of absorbent (da), the bond angle N-H-N () 27

Figure 3.8 The adsorption energy Ead as a function of distance dz between (a) CO (b) CO2 (c) NO (d) NO2 (e) NH3 and monolayer MoTe2 achieved by differerent vdW-DFs 33

Figure 3.9 The comparison of the magnitude of adsorption energies and the adsorption distance of toxic gases adsorption on monolayer MoTe2 achieved by optB88 vdW functional 33

`

iii

Figure 3.10 Comparison of the charge transfer of the toxic gases adsorbed on

monolayer MoTe2 achieved by optB88 vdW functional 46

Figure 3.11 Charge density difference for CO– MoTe2 system achieved by optB88 vdW functional Top view (left) and side view (right) (isosurface level = 0.0002) 47

Figure 3.12 Charge density difference for CO2 – MoTe2 system achieved by optB88 vdW functional Top view (left) and side view (right) (isosurface level = 0.0002) 47

Figure 3.13 Charge density difference for NO– MoTe2 system achieved by optB88 vdW functional Top view (left) and side view (right) (isosurface level =0.0022) 48

Figure 3.14 Charge density difference for NO2 – MoTe2 system achieved by optB88 vdW functional Top view (left) and side view (right) (isosurface level: 0.0006) 48

Figure 3.15 Charge density difference for NH3 – MoTe2 system achieved by optB88 Top view (left) and side view (right) (isosurface level = 0.0004) 48

Figure 3.16 BAND-DOS of CO– MoTe2 calculated by optB88 50

Figure 3.17 BAND-DOS of CO2 – MoTe2 calculated by optB88 51

Figure 3.18 BAND-DOS of NH3 – MoTe2 calculated by optB88 51

Figure 3.19 BAND-DOS of NO– MoTe2 calculated by optB88 52

Figure 3.20 BAND-DOS of NO2 – MoTe2 calculated by optB88 52

Figure 3.21 The variation in work function () for different gas adsorption systems.53

`

iv

LIST OF ABBREVIATIONS

DFT Density Functional Theory

GGA Generalized gradient approximation

LDA Local density approximation

VASP Vienna Ab initio Software Package

PAW Projector Augmented Wave

vdW van der Waals

vdW-DFs van der Waals density functionals

TMDCs Transition metal dichalcogenides

BAND Energy band structure

DOS Density of state

HF Hartree-Fock

VBM Valence Band Maximum

CBM Conduction Band Minimum

FET Field Effect Transistor

in today's era These gas pollutants may cause climate change, ozone pollution, harmful direct and indirect human health and even threaten food security Therefore, the detection of toxic gases has become extremely essential for environmental controlling, human safety and manufacturing monitoring products

In general, a proficiency of the gas sensor is decided by several factors, that include, (1) selectivity, (2) sensitivity (3) faster response and recovery time, (4) reversible i.e., return to its original state after gas is removed (5) low cost Recently, it can be seen that the gas sensor is generally based on the semiconducting metal oxide which is high sensitivity and low cost However, there are some drawbacks such as the high operating temperature requirement, large power consumption and low selectivity Hence, the exploration for a novel material to produce a proficiency gas sensor that can operate at room temperature is broadened

`

2

CHAPTER 1 LITERATURE REVIEW

1.1 Overview toxic gases

The toxic gases have a direct and indirect effect on human health as well as the environment In our work, we investigate the five most effective candidate gases: CO,

CO2, NO, NO2, NH3

a) Carbon dioxide (CO 2 ): CO2 is a “greenhouse gas” described as the worst climate pollutant It was proved that high concentration of carbon dioxide is responsible for a variety of health effects when inhaled which include headaches, dizziness, increased heart rate, etc Carbon dioxide mainly produced from burning fossil fuels (coal, natural gas, and oil), land-use changes and industrial processes In structure, CO2 is a flat molecule with sp hybridization, carbon linked to two oxides by

a double bond with the bond length equal to 1.163 Å. [1]

Figure 1.1 The molecular structure of CO2

b) Carbon monoxide (CO):

Carbon monoxide is a colorless, odorless, tasteless, flammable, high toxicity gas that has a main source from burning carbon-based fuel of vehicles, regime, etc High concentration of carbon monoxide entering your bloodstream immediately combined with hemoglobin produces carboxyhenmologin, which prevents carrying oxygen in blood, leading to serious nervous system damage, or even death For chemical properties, CO is linked by the triple bond including one  bond and 2  bond with the bond length around 1.128 Å [2]

Figure 1.2 The molecular structure of CO

`

3

d) Nitrogen dioxide (NO 2 ): NO2 is a reddish-brown gas at high temperature which use

in industrial to produce of nitric acid Every year, millions of tons of NO2 were emitted

to environment by combustion fuels The high toxicity of NO2 lead to a serious harm

to health such as heart failure and even to death by directly absorbed through human lungs and its inhalation Basically, NO2 is a paramagnetic molecule with a single unpaired electron NO2 has bond length of 1.197 Å, the bond angle of 134.3 and sp2hybridization [3]

Figure 1.3 The molecular structure of NO2

c) Nitrogen monoxide (NO): NO gas is a colorless gas that is a major part of acid rain

In general, nitrogen oxides are mainly produced from natural phenomena - lightning in thunderstorms, for instance However, in the atmosphere, NO transforms naturally into

NO2 during the oxidation process Besides, the burning of fuels contributes a quite significant NO to the air Breathing low levels of NO leads to some symptom like cough, tiredness, and nausea Nonetheless, NO can seriously damage our lungs over the next one to two days after breathing Fundamentally, NO is a paramagnetic molecule with bond length of 1.15 Å and has sp2 hybridization with one unpaired electron [4]

Figure 1.4 The molecular structure of NO

e) Ammonia (NH 3 ): Ammonia is a pungent smell, colorless alkaline gas which is

known as the abundant nitrogen-based substances in the air Over breathing the threshold of NH3 concentration expose you to sinusitis, upper airway irritation, and eye irritation Some diseases of the lower 11 airways and interstitial lung are found as

Figure 1.5 The molecular structure of NH3

1.2 Two-dimensional (2D) material-based toxic gas sensors

Gas sensors are devices which are used to identify or quantify the properties of detected toxic gases and organic vapors These devices are essential in several fields, for example environmental security, human safety, industry emission control, medical diagnostic and also military, etc Nowadays, various types of materials are used for gas sensing applications such as carbon nanotubes, conducting polymers, semiconducting metal oxides [6] Among these materials, metal oxides are mainly used due to their high sensitivity and economic efficiency Nevertheless, there are some remaining disadvantages, for instance, the high operating temperature (>200°C), large power consumption and low selectivity [6] Therefore, it is vital to discover novel gas sensing materials with the proficiency: sensitivity, selectivity, low operating temperature, fast response and recovery time, reversibility and low cost

2D materials possess a large surface to volume rate, unique electronic properties and performance at ambient temperature The most common fundamental mechanism

of 2D material sensing is the charge transfer mechanism The charge transfer mechanism based on the charge transfers between sensing material and molecules during the molecule adsorption The difference of direction and quantities of charges when exposing different gases leads to different variations of resistance For example, the charge transfer process between n-type MoS2 and toxic gases combined with the charge difference was shown in Figure 1 [7] When the O2, H2O, NO, NO2, and CO adsorbed on MoS2 the charges transfer from MoS2 to gases which lead to decrease of

Figure 1.6 Illustration of the charge transfer mechanism of MoS2 to gas molecules [7]

Table 1.1 Adsorption characteristics of toxic gases on some typical 2D materials

Materials Gas

molecules

Adsorption Sites

E a (meV) Q (e) d (Å) Method Ref

`

8

and fullerenes, in the 1990s, Reshef Tenne and others led vigorous efforts in the field

of inorganic fullerenes and nanotubes, beginning with the discovery of WS2 nanotubes and nested particles, followed by synthesis of MoS2 Nanotubes and nanoparticles The rapid growth of graphene-related research starting in 2004 has stimulated the development of well-supported techniques for working with layered materials, opening avenues for further research in TMDCs and especially their ultrathin films [17]

2D TMDCs have several favorable properties which have been interested in the sensor application Firstly, 2D TMDCs have a large surface to volume ratio similar to the 2D materials such as graphene The large surface to volume ratio led to abundant effective sites on the surface, as well as more unsaturated surface sites Therefore, TMDCs are completely sensitive to the surrounding environment and could be used in adsorption and gas sensing Secondly, in contrast with graphene which has zero bandgap, 2D TMDCs have a direct tunable bandgap, which could be tuned by thickness, defects, dopants, and mechanical deformation (by applying strain), giving rise to the electronic properties Thirdly, the defect engineering can be easy to implement in 2D materials, which have been confirmed to be an efficient method for intensifying the catalytic activities in TMDCs Last but not least, TMDCs group has a variety of members which have abundant amount in nature and 2D TMDCs could be simply synthesized [18] The main metals W and Mo are both abundant, cheap, and widely used in industry So far, MoS2, WS2, MoSe2, and ReS2 have been naturally found Specifically, MoS2 exists as molybdenite in nature and is the main source of molybdenum with a large amount Two-dimensional TMDCs were originally fabricated by the top-down exfoliation from bulk crystals such as mechanical exfoliation and liquid phase exfoliation While exfoliation can be applied simply in a laboratory, the preparation of 2D TMDCs materials is now becoming possible through Chemical Vapor Deposition (CVD), Physical Vapor Deposition (PVD), Molecular beam epitaxy (MBE) and Metal–organic CVD techniques, allowing large area fabrication of any two-dimensional material Using these techniques, it is even possible to grow 2D materials directly into heterostructures With several favorable properties and large-scale processing, TMDCs hold great promise in terms of the development of gas sensors One evidence is that several gas sensors based on MoS2,

1.3.2 Monolayer MoTe 2 based sensor

Monolayer MoTe2 is a new addition to TMDCs group which inherited outstanding properties from TMDCs In detail, 2H semiconducting phase MoTe2 has a direct near- infrared band gap around 1.1 eV Besides, it has high electron mobility up

to 137 cm2 V-1 at 77K, nature strong spin orbit coupling, spin valley coupling, ambipolar behavior making it to be a promising candidate for applications in nanoelectronics, near-infrared optoelectronic, valleytronic, spintronics [22]

`

10

To our best knowledge, the research for MoTe2 application in gas sensors is still the drawback In 2018, the few-layer MoTe2 based NH3 (NO2) FET sensor (Figure 3) with high sensitivity of detection limit to 1 ppm (100 ppb) were reported by Feng’s group; Nevertheless, the insight mechanism is not understood at all [23]

Figure 1.8 The illustration of MoTe2 FET sensor(a) Optical microscope image (b) AFM topography image d) High resolution TEM image (e) SAED pattern taken were

reported by Feng’s group [23]

On the other hand, there was limited reliable theoretical investigation which was shown in Table 1.2 As can be seen, the potential of MoTe2 for detecting SO2from SF6 decomposition and Methane are evaluated It was proved from previous studies that the vdW interaction is vital in weak bonded systems such as physisorption [24-26] In recent, the computational research of NO, NO2 and NH3 adsorption on monolayer MoTe2 took the vdW interaction into account but they used the semi-empirical vdW correction It can be seen from previous studies that the non-empirical vdW functionals give a higher accuracy than the vdW correction for vdW description [25, 26] Furthermore, the comparative calculation of vdW methods for adsorption of

`

11

CO2 on metal organic framework shows that the non-empirical vdW functional group reproduces nearest the values of adsorption enthalpy [27] In conclusion, the systematic comparative investigation of the adsorption of the most influenced toxic gases on the monolayer MoTe2 using non-empirical vdW functionals is essential

Table 1.2 Adsorption characteristics of common gases on MoTe2 monolayer

Gas

molecules

Adsorption sites E a (meV) Q (e) d (Å) Method Ref

`

12

CHAPTER 2 METHODOLOGY

2.1 Density Functional Theory

Density Functional theory (DFT) is a quantum computational modeling method applied in physics, chemistry and materials science to investigate the electronic structure of many-body systems In 1964, Pierre Hohenberg and Walter Kohn gave the first Hohenberg-Kohn definition proving that for molecules with a fundamental non-degenerate state, the ground energy state, wave function and other electron properties could be determined by the electron density function 𝜌0(𝑥, 𝑦, 𝑧) , a function with only three variables It was followed by the second Hohenberg-Kohn theorem defined the ground-state energy at the by a functional 𝐸0 = 𝐸[𝜌0] [31, 32] However, Hohenberg-Kohn held only the non-degenerate ground-state for their theorem Consequently, in

1967, John P.Perdew and Mel Levy proved the theorem for degenerate ground-state [31, 32]

According to Hohenberg-Kohn theorem, in principle, we could find out all ground-state properties of molecules by ground-state electron density 𝜌0(𝑟) without wavefunction However, Hohenber and Kohn did not point out the way to calculate 𝐸0from 𝜌0 as well as find 𝜌0 without wave function Following this incompliance, in

1965, Kohn and Sham developed the approximate method to treat this problem [31, 32]

2.1.1 Kohn-Sham method

Kohn and Sham considered an abstract system, which is described by the s

index and usually called the non-interaction system, of n non interacting electrons

electron probability density function determines the external potential Therefore,

Despite of s reference system include non-interacting element, the

Kohn and Sham rewrite the Hohenberg-Kohn equation as follows Firstly,

molecular system and a reference system of electrons which do not interact with others

in the same electron density

energy functional is defined as𝐸𝑥𝑐[𝜌] ≔ Δ𝑇̅[𝜌] + Δ𝑉̅𝑒𝑒[𝜌], we achieve

𝐸0 = 𝐸𝜈[𝜌] = ∫ 𝜌(𝒓)𝜈(𝒓)𝑑𝒓 + 𝑇̅𝑠[𝜌] +1

2∬𝜌(𝒓1)𝜌(𝒓2)

𝑟 12 𝑑𝒓1𝑑𝒓2+ Exc[𝜌] (5)

`

14

electronic energy including the nuclear repulsion is evaluated by adding the

Hohenberg-Kohn theorem show that the value of the ground-state energy could

where the exchange-correlation potential 𝜈𝑥𝑐 is found as the functional derivative

of noninteracting electrons, so, strictly speaking, these orbitals have no physical significance other than in allowing the exact molecular ground-state 𝜌 to be calculated The density-functional molecular wave function is not a Slater

determinant of spin-orbitals In fact, there is no density-functional molecular wave function However, in practice, one finds that the occupied Kohn-Sham orbitals

resemble molecular orbitals calculated by the Hartree-Fock method, and the KohnSham orbitals can be used (just as Hartree-Fock MOs are used) in qualitative MO

`

15

discussions of molecular properties and reactivities Note that, strictly speaking, Hartree-Fock orbitals also have no physical reality, since they refer to a fictitious model system in which each electron experiences some sort of average field of the other electrons [31, 32]

For a closed-shell molecule, each Hartree-Fock occupied-orbital energy is a good approximation to the negative of the energy needed to remove an electron from that orbital (Koopmans’ theorem) However, this is not true for Kohn-Sham orbital energies The one exception is 𝜀𝑖𝐾𝑆 for the highest-occupied KS orbital, which can be proved to be equal to minus the molecular ionization energy With the currently used approximations to 𝐸𝑥𝑐ionization energies calculated from KS highest-occupied-orbital energies agree 4poorly with experiment [23, 24]

Various approximate functionals 𝐸𝑥𝑐[𝜌] are used in molecular DFT calculations To study the accuracy of an approximate 𝐸𝑥𝑐[𝜌], one uses it in DFT calculations and compares calculated molecular properties with experimental ones The lack of a systematic procedure for improving 𝐸𝑥𝑐[𝜌] and hence improving calculated molecular properties is the main drawback of the DFT method [31, 32]

In a true‖ density-functional theory, one would deal with only the electron density (a function of three variables) and not with orbitals, and would search directly for the density that minimizes 𝐸𝜈[𝜌] Because the functional 𝐸𝜈 is unknown, one instead uses the Kohn-Sham method, which calculates an orbital for each electron Thus, the KS method represents something of a compromise with the original goals of DFT [31, 32]

2.1.2 DFT computational scheme

The summary of DFT calculation steps as follows Figure 2.1

𝐸𝑥𝑐 = 𝐸𝑥𝐺𝐺𝐴+ 𝐸𝑐0+ 𝐸𝑐𝑛𝑙 (8) Here, 𝐸𝑥𝐺𝐺𝐴 is the part of exchange energy treat by GGA and 𝐸𝑐0 is the LDA correlation energy If the first two term of the equation (8) plus the 𝐸𝑐𝐺𝐺𝐴 the GGA correlation energy, it becomes the standard GGA exchange-correlation energy The nonlocal component, 𝐸𝑐𝑛𝑙 , accounts for long-ranged electron correlation effects responsible for

in ref 33-36 In this study we consider the following vdW-DFs: revPBE-vdW, which is the “original” vdW-DF introduced in ref [33], the three variants introduced by Klimes and co-workers called “opt” functional group: optB86b [34], optB88, optPBE [35], where the exchange functionals were optimized for the correlation parts and one variant introduced by Lee’s group, vdW-DF2 [36] The four variants functionals differ from the revPBE-vdW on their choice in the exchange energy, 𝐸𝑥𝐺𝐺𝐴, corresponding to the first term on the right-hand side of equation

2.2 VASP

The Vienna Ab initio Simulation Package (VASP) is a package for

performing quantum mechanical modelling from first principle using either the projected augmented wave (PAW) pseudopotentials, or Vanderbilt pseudopotentials and a plane wave basic set [30] The fundamental methodology is density functional theory (DFT), but the code also allows use of post-DFT corrections such as hybrid functionals mixing DFT and Hartree-Fock exchange such as HSE [38], PBE0 [39], B3LYP [40], many-body perturbation theory (the GW method) and dynamical electronic correlations within the random phase approximation

VASP was originally based on code written by Mike Payne (then at MIT), which was also the basis of CASTEP It was later conducted to the University of Vienna, Austria, in July 1989 by Jürgen Hafner The main program was written

by Jürgen Furthmüller and Georg Kresse VASP is currently being developed

by Georg Kresse; recent additions include the extension of methods frequently used in molecular quantum chemistry (such as MP2) to periodic systems VASP is currently

`

18

used by more than 1400 research groups in academia and industry worldwide on the basis of software license agreements with the University of Vienna

2.3 Bader Charge Analysis

Atomic charge in solid and molecules is an unobservable quantity which could not be defined by quantum theory The result of quantum calculation is continuous electronic charge density and there is not a specific scheme to partition electrons from molecules into atoms Several methods have been published based on electronic orbitals (Muliken population analysis, Density matrix based normal population analysis…) and just the charge density (Bader analysis, Hirshfeld analysis…) [41]

Bader charge analysis was founded by Richard Barder from McMaster University He published an effective approach to divide molecules into atoms based

on zero flux surface Zero flux surface is 2D surface on which the charge density is a minimum perpendicular to the surface In the molecular system, the minimum of charge density is reached between atoms, which is the natural separated place of atoms Bader’s tool is often used for charge analysis [41]

The Bader charge analysis scheme

• The electron density, ρ(x, y, z), of materials are analyzed

• Critical points of ρ(x, y, z) are determined and classified

• The 3D space is divided into subsystems, each usually containing one nucleus (but sometimes none)

• Zero-flux surfaces separate the subsystems: ∇ρ (rs) • n(rs) = 0 for every point rs on the surface S(rs) where n(rs) is the unit vector normal to the surface at rs

• The electron density can either be from experimental data (e.g., X-ray crystallography) or theoretical data (e.g., ab initio calculations)

To model the adsorption of toxic gases on monolayer MoTe2, the 4x4x1 supercell of MoTe2 containing 48 atoms was chosen The vacuum region of 25 Å was added to avoid the interaction between the periodic layers The plane-wave basis set was implemented with the cutoff energy of 500 eV and the Brioullin zone is sampled with Monkhorst–Pack grids of 2 x 2 x 1 [39] The convergence condition reaches if the atomic force acting on each atom and the energy threshold are less than 0.025 eV/A and 10-5 eV, respectively The Computational DFT-based Nanoscope [46] was used to find out the most favorable configurations and related parameters The charge transfer was analyzed by the Bader charge anlysis [41]

`

20

CHAPTER 3 RESULTS AND DISCUSSION

3.1 Monolayer MoTe 2 properties

3.1.2 Geometrical properties

(a) (b) (c)

Figure 3.1 (a) side view (b) top view of the hexagonal monolayer MoTe2 and (c) the

considered adsorption sites on monolayer MoTe2

The 2H phase structure of MoTe2 is the most stable at room temperature, hence,

we choose this phase for our investigation The 2H MoTe2 crystals crystallize in a hexagonal structure with the space group P63/mmc corresponding to group number They are formed from X-M-X monolayers by van der Waals (vdW) interaction Each MoTe2 monolayer has a honeycomb structure with a point symmetry group of D3h

consisting of two hexagonal planes of the chalcogen atom (Te) and an intermediate hexagonal plane of the transition metal atom (Mo) linked bound through ionic covalent interactions, with arranged in a hexagonal (armchair like structure) as shown in Figures 3.1 (a) and Figure 3.1 (b)

Figure 4.c shows 6 different adsorption sites which we have used to consider the most favorable adsorption configuration (i) H site (on the top of a hexagonal center) (ii) TM site (on the top of Mo atom) (iii) TTe site (on the top of Te atom)(iv) BM site(bridge between Mo atom) (v) BTe (bridge between Te atom) The directions of molecules (point toward or away the surface) are also considered

The comparison of optimized structure parameters achieved by vdW-DFs functionals is shown in the Table 3.1 It can be seen that opt functional group is good in

`

21

reproducing the structure because of the perfect matching with the experimental values

In contrast, the geometrical values were obtained by revPBE and vdW-DF2 functional are quite overestimated

Table 3.1 The structure parameters of monolayer MoTe2

`

22

could expect that the opt functionals are better for performing electronic properties than the others The values of VBM splitting and CBM splitting were also evaluated Notably, DOS of pristine MoTe2 is mainly contributed by Mo d orbitals, Te d orbitals

as well as hybridization However, the major contribution around VBM and CBM comes from d orbitals of Mo

Table 3.2 The band gap and spin orbit splitting values comparing by non-empirical

`

23

(b)

Figure 3.2 The Band structure and DOS of the monolayer MoTe2 with (a) and

without (b) SOC taken into account

3.2 Adsorption mechanism of toxic gases on monolayer MoTe 2

3.2.1 Adsorption configuration

CO adsorption on monolayer MoTe 2 (CO-MoTe 2 )

After optimization of the CO-MoTe2 system, the favorable configuration is achieved at B site with C toward the adsorption surface (Figure 3.3) The C≡O bond length is not change significantly just around 1% compare with 1.128 Å of the free-standing state (see Table 3.3) The smallest distance between lowest atom and nearest atom of adsorbent is around 3.6 Å which is longer than the sum of the C and Te covalent radii about 2.14 Å (Table 5) [49] As a result, it was shown in Figure 3.3 that there are no notable deformation of adsorption surfaces and no chemical bonding between a carbon atom and telluride atoms These evidences implies that the adsorption mechanism of this system is physisorption

`

24

Figure 3.3 The C≡O bond-length (l), the distance between C atom and the nearest

atom of absorbent (da), the angle between molecules and the absorbent surfaces () of

CO-MoTe2 favorable configurations

CO2 adsorption on monolayer MoTe2 (CO2-MoTe2)

Figure 8 shows the stable configuration of CO2 adsorption on monolayer MoTe2 systems As can be seen from Figure 3.4, Carbon atom prefers to locate at the B site when one Oxide atom lies on the top of the Mo atom and other direct to hollow site making the molecule nearly lie parallel to the adsorbent surface The C=O bond-length

is witnessed by a tiny stretching approximate 1.8 % from 1.163 Å in the isolated system Beside that the C=O=C bond angle bends a little bit from the flat structure depending on the functionals Similar to the CO adsorption case, there is no notable surface deformation and chemical bonding between carbon and telluride because of the large distance from nearest Te and C around 3.93 Å [49] Those could conclude that the adsorption mechanism is physical adsorption

`

25

Figure 3.4 The favorable configurations of CO2-MoTe2 top view (a) and side view (b) with the adsorption distance (d), the C=O bond length (l), the distance between C atom

and the nearest Te atom of absorbent (da), the bond angle ()

NO adsorption on MoTe 2 monolayer (NO-MoTe 2 )

After optimization of the adsorption structure for different sites and orientations, the favorable configuration is achieved at B site with Nitrogen pointing at the surface

of adsorbent (Figure 3.5) The N O bond-length shrinks from 1.152 Å at the standing state to 1.141-1.146 Å depending on the functionals (Table 3.3) On the other hand, the smallest distance from the N atom to the nearest Te atom is approximately 2.64 Å larger than the sum of covalent radius N and Te (around 2.09 Å) [49] Combining with no deformation of the surface of adsorbent, we could evaluate that the

free-NO is physiosorbed on monolayer MoTe2

Figure 3.5 Top view (a) and side view of the favorable configurations of NO-MoTe2 (b) with the adsorption distance (d), the N O bond length (l), the distance between

N atom and the nearest atom of absorbent (da), the angle between molecules and the

absorbent surfaces ()

d

