Method for predicting electric transport property of MXene-MSi2N4 heterojunction based on theoretical calculation

Abstract

The invention discloses a method for predicting the electrical transport property of an MXene-MSi2N4 heterojunction based on theoretical calculation. Belongs to the technical field of electronic devices and comprises the following operation steps that a material database is screened, MXene and MSi2N4 models are built respectively, VASP calculated based on the density functional theory is adopted to calculate work functions of MXene and MSi2N4 of different functionalizations, and preliminary screening of heterojunction materials is carried out; an MXene-MSi2N4 heterojunction structure is constructed, and the property of a contact interface is explored based on VASP software; screening heterojunctions with stable thermodynamics, longitudinal ohmic contact and good tunneling performance for next calculation; a two-dimensional double-gate field effect transistor device model is built based on the screened MXene-MSi2N4 heterojunction, the electric transport properties including a zero-bias transmission spectrum, projection local state density and contact resistance are calculated by using Quantum ATK, and the optimal device model is screened. According to the invention, the electric contact material with excellent performance is effectively predicted through a theoretical calculation method, and the two-dimensional field effect transistor component with low contact resistance and full ohmic contact is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of electronic devices and relates to a theoretical method for predicting the electrical transport properties of heterojunction electrical contact materials; specifically, it relates to a method for predicting the electrical transport properties of MXene-MSi2N4 heterojunctions based on theoretical calculations. Background Art

[0002] With the advent of the post-Moore era, traditional short-channel devices using silicon-based materials as channel materials are facing problems such as the short-channel effect. Two-dimensional semiconductors have attracted much attention due to their inherent ultra-thin layered structure, ultra-flat, clean surface without dangling bonds, and excellent electrical properties. They are expected to replace traditional silicon-based materials, further reduce device feature sizes, and effectively suppress the short-channel effect. They are highly promising channel materials for short-channel components. However, two-dimensional semiconductor devices face several technical barriers. Two-dimensional semiconductor materials are prone to high contact resistance when in contact with metal electrodes. This is mainly due to the mismatch between the work functions of the semiconductor and the electrode, and the presence of Fermi-level pinning caused by metal-induced interstitial states at the interface. The Schottky contact caused by these problems increases the contact resistance in the component, reducing the performance indicators of the component. Therefore, how to achieve low-resistance ohmic contact between two-dimensional semiconductor materials and metals is a key technical challenge in improving the performance of two-dimensional semiconductor components.

[0003] To address this technical challenge, researchers have explored a variety of strategies. Some studies have attempted to enhance the interaction between metals and 2D semiconductors to increase orbital overlap and achieve strong interface hybridization, thereby reducing the Schottky barrier. Alternatively, a thin dielectric layer has been introduced to weaken the interface coupling between the metal and the 2D semiconductor, thereby suppressing Fermi-level pinning. However, large van der Waals gaps or thin dielectric layers create large interfacial tunneling barriers, severely hindering carrier injection from the metal electrode into the 2D semiconductor. Furthermore, attempts have been made to heavily dope the 2D semiconductor through surface charge transfer doping to achieve lower contact resistance, but this technology is currently difficult to implement in terms of process.

[0004] In summary, the contact between two-dimensional semiconductors and metal electrodes currently studied still has problems such as large tunneling barriers and the need for additional doping. Summary of the Invention

[0005] In response to the above problems, the purpose of the present invention is to propose a method for predicting the electrical transport properties of MXene-MSi2N4 heterojunctions with low-resistance ohmic contacts based on theoretical calculations.

[0006] The technical solution of the present invention is: a method for predicting the electrical transport properties of MXene-MSi2N4 heterojunction based on theoretical calculations, the operating steps of which are as follows:

[0007] Step (1): Based on the Materials Project database and VESTA software, a two-dimensional (2D) single-layer MXene and MSi2N4 structure model was preliminarily constructed, and the structure file was converted into a VASP readable file format;

[0008] Step (2): Use the density functional theory (DFT)-based software VASP to perform structural relaxation on the preliminarily constructed single-layer two-dimensional MXene and MSi2N4 materials. After self-consistent convergence, perform static self-consistent calculations and non-static self-consistent calculations. Analyze the results to extract the work functions of MXene and MSi2N4, and select MXene and MSi2N4 materials with matching work functions for the next step of calculation;

[0009] Step (3): constructing a two-dimensional MXene / MSi-2N4 van der Waals heterojunction model based on the MSi2N4 and MXene selected in step (2), optimizing the interlayer spacing and configuration, and screening a thermodynamically stable contact system;

[0010] Step (4): Use VASP to calculate the electronic properties of the MXene-MSi2N4 two-dimensional van der Waals heterojunction and confirm the size of the contact interface barrier (Schottky barrier and tunneling barrier); combine the metal and semiconductor interface theory to determine the type of MXene-MSi2N4 contact interface, and select a heterojunction material with longitudinal ohmic contact (no Schottky barrier) and excellent tunneling performance for the next calculation;

[0011] Step (5): constructing a two-dimensional dual-gate field effect transistor (FET) device model based on the heterojunction material selected in step (4);

[0012] Step (6): Combine density functional theory with non-equilibrium Green's function (DFT+NEGF) method and use QuantumATK software to calculate the electrical transport properties of the FET model, including zero-bias transmission spectrum, device contact resistance, projected localized density of states (PLDOS), etc.

[0013] Step (7): Based on the calculation results of step (6), the electronic transport properties of FET are studied and the FET device model with the optimal electronic transport properties is selected.

[0014] Furthermore, in step (1), the VESTA software imports a structure file as a .cif structure file and exports a structure file as a .vasp structure file;

[0015] The MSi2N4 structure file is from a published literature, and the space group number is 187;

[0016] The MXene is a single-layer surface functionalized MXene with a space group number of 164, which was screened from the C2DB database;

[0017] For two-dimensional materials, the above models are constructed with consideration of the length greater than or equal to Vacuum layer;

[0018] The specific operation process is as follows:

[0019] Step (1.1): Obtain the lattice parameters, atomic coordinate positions, bond lengths and angles of the initial structure of single-layer 2D MSi2N4 and 2DMXene based on literature and the Materials Projects database;

[0020] Step (1.2): Use VESTA software to preliminarily construct a single layer of 2D MSi2N4 and 2D MXene, obtain and convert its structure file into .vasp file format.

[0021] Furthermore, the specific operation process of step (2) is as follows:

[0022] Step (2.1): Use the DFT-based VASP software to optimize the structure of single-layer 2D MSi2N4 and 2D MXene. The specific settings of its parameters include: the energy criterion of the plane wave is 550eV (the wave function cutoff energy is set to 550eV), the convergence standard of the force is (Convergence criteria of the force of structural optimization ), the energy convergence criterion is 10-6eV for structural optimization;

[0023] Step (2.2): Based on the above calculations, a stable monolayer 2D MSi2N4 and 2D MXene structure is obtained; then a static self-consistent calculation is performed, and the energy convergence standard is increased to 10-7eV to ensure the calculation accuracy. The planar electrostatic potential of the two along the z axis is obtained, and the calculated work function is collected. The work function of MXene is recorded as the metal work function. The work function of MSi2N4 is recorded as the semiconductor work function

[0024] The specific settings of its parameters include: the convergence standard of the electronic self-consistent cycle is 1×10 -7 eV, the sampling method centered on Γ was used to sample the K point with a sampling density of 21×21×1. The high symmetry point path in the K space of the energy band calculation was Γ-MK-Γ. The plane electrostatic potential of MXene and MSi2N4 along the z axis was extracted after the self-consistent calculation converged, and the work function value was extracted;

[0025] Step (2.3): The prerequisite for n-type ohmic contact is: Screening MXenes that meet the prerequisites for the next step of building a van der Waals heterojunction model;

[0026] The specific steps are: align the vacuum energy levels of all structural models, record the energy difference from the vacuum energy level to the Fermi level of each model as the work function, and record the work function of MXene as the metal work function The work function of MSi2N4 is recorded as the semiconductor work function Filter to meet the conditions The structure of the single-layer MXene used in the subsequent steps is shown in Figure 5.

[0027] Furthermore, in step (3), the MXene-MSi2N4 van der Waals heterojunction model is constructed; the specific operation process is as follows:

[0028] Step (3.1): Based on the results of step (2), VESTA was used to build a MXene / MSi2N4 heterojunction model and determine that the lattice mismatch between the two is <5%.

[0029] Step (3.2): Use VESTA software to convert the structure file format into .vasp format file for subsequent calculations;

[0030] Step (3.3): Use VESTA to construct structural models with different interlayer spacings, optimize the interlayer spacing, screen the structure with the lowest interlayer binding energy, and determine the optimal interlayer spacing;

[0031] Step (3.4): Determine 6 structural configurations through translation and rotation operations, perform configuration optimization, and find the stable configuration with the lowest binding energy for subsequent contact property calculations;

[0032] Specifically, in step (3.1), the heterojunction model is formed by stacking the MXene layer and the MSi2N4 layer along the c-axis, and controlling the interlayer spacing between Between them, the interlayer spacing is guaranteed to be within the van der Waals contact range, and the length along the c-axis is greater than or equal to Vacuum layer;

[0033] In step (3.3), the specific settings of the calculation parameters include: the wave function cutoff energy is set to 550 eV, the electronic self-consistent cycle convergence standard is set to 1×10 -6 eV, the conjugate gradient algorithm is used for ion relaxation, and the maximum atomic force is Dispersion correction was performed using the DFT-D3 method to better describe the van der Waals force. The K point was sampled using a Γ-centered sampling method with a sampling density of 15×15×1. After the structure optimization reached the set accuracy, the total energy of the system at each interlayer spacing was calculated, and the interlayer spacing model with the lowest total energy was selected for subsequent calculations.

[0034] Furthermore, the specific operation process of step (4) is as follows:

[0035] Step (4.1): Perform self-consistent calculation, state density and energy band calculation on the MXene / MSi2N4 van der Waals heterojunction after completing the structural optimization of step (3), where the wave function cutoff energy is set to 550 eV and the electronic self-consistent energy convergence standard is 1×10 -7 eV, DFT-D3 (IVDW = 11) was used to perform dispersion correction for the van der Waals heterojunction (LDIPOLE = .TURE.), and the self-consistent calculation used a Γ-centered sampling method to sample the K points with a sampling density of 15 × 15 × 1, and the parameter LVHAR = .TRUE. The high-symmetry point path in K space for the band structure calculation was Γ-MK-Γ.

[0036] Step (4.2): Extract the plane electrostatic potential (plane electrostatic potential along the z-axis), Fermi level (Fermi level value), binding energy value, and electronic structure of the heterojunction, including the projected density of states and projected band diagram, obtained in the above calculations;

[0037] Thermodynamic stability is measured by the interlayer binding energy of the heterojunction, the interface contact properties are determined by the longitudinal interface Schottky barrier, and the tunneling properties at the interface are determined by the interface tunneling barrier (the tunneling barrier is determined by the plane electrostatic potential along the z-axis and the Fermi level);

[0038] A heterojunction model with negative interlayer binding energy, a longitudinal interface Schottky barrier less than 0, and a small interlayer tunneling barrier is selected.

[0039] Furthermore, the specific operation process of step (5) is as follows:

[0040] Step (5.1): Based on the MXene / MSi2N4 heterojunction selected above, a dual-gate FET device was constructed using Quantum ATK. It is a dual-probe device consisting of left and right electrode regions, a central scattering region, and left and right electrode extension regions sandwiched between the two. The MXene / MSi2N4 heterojunction is the minimum repeating unit of the left and right electrode regions (the left and right electrode regions and the left and right electrode extension regions are all formed by simple repetition of the selected heterojunction model along the transmission direction c-axis), and the central scattering region is repeated by intrinsic 2D MSi2N4 (the central region is an intrinsic single-layer MSi2N4); the a-axis is the device width direction, the b-axis is the thickness direction, and the longitudinal van der Waals heterojunction contact interface perpendicular to the electrode part and the vacuum layer thickness is greater than The c-axis is the device transmission direction;

[0041] Step (5.2): Set the device parameters as follows: the effective oxide layer thickness of the dielectric layer is 0.41nm, the dielectric constant is selected as 3.9, the gate thickness is 0.2nm, and the channel length is 5.1nm; the left and right electrode regions constitute the source and drain ends of the FET device, and the central scattering region is the channel region of the FET device.

[0042] Furthermore, in step (6), the electron transport properties of each two-dimensional FET model are calculated, including the zero-bias transmission spectrum, the total device resistance, and the lateral interface contact properties; the specific operation process is as follows:

[0043] Step (6.1): Combine density functional theory calculations with non-equilibrium Green's functions to perform electrical transport model calculations on the non-equilibrium device. Import the device model into the Quantum ATK workflow and set the device's DeviceLCAO calculation parameters: boundary conditions, pseudopotential basis set selection, electron temperature, K-space mesh density, convergence criteria, dispersion correction parameters, etc.

[0044] Step (6.2): Based on the above settings, the transmission probability spectrum T(E, V) of the device is obtained by solving the non-equilibrium Green's function in the central region;

[0045] Step (6.3): Based on the calculated transmission probability spectrum, extract the zero-bias transmission spectrum, differential conductance, and projected local density of states (PLDOS);

[0046] Specifically, in step (6.1), density functional theory calculations and non-equilibrium Green's function are combined, the device model built in step 5 is used, the Quantum ATK workflow is introduced, and the DeviceLCAOCalculator calculation module is used to set the calculation parameters;

[0047] The specific parameter settings are as follows: the pseudopotential basis set is PseudoDojo, the electron temperature is set to 300K, the K-space grid density is set to 17*1*300, and the convergence criterion is 10 -5 eV, the dispersion correction parameter is set to Grimme DFT-D3;

[0048] In step (6.2), the Green's function of the central region is solved based on the above calculation parameters, and the quantum transmission coefficient is calculated based on the Green's function; the calculation parameter settings of the transmission spectrum include an energy range of -3.0 eV to 1.0 eV and a point setting of 201;

[0049] In step (6.3), the zero-bias transmission spectrum, differential conductance, and lateral Schottky barrier are extracted from the quantum transmission coefficient.

[0050] Furthermore, in step (7), the optimal two-dimensional ohmic contact FET model is selected; the specific operation process is as follows:

[0051] Step (7.1): Extract the lateral Schottky barrier based on the projected state density of the device model calculated in step (6), select the model in which the lateral interface is an ohmic period, and combine the model selected in step (4) to obtain a device model in which both the lateral and vertical interfaces are ohmic contacts;

[0052] Step (7.2): Calculate the zero-bias differential conductance of the selected device model, using the formula R c =G / 2w to get the intrinsic contact resistance of the device. In this formula, R c represents the intrinsic contact resistance, G is the differential conductance, extracted from the zero-bias transmission spectrum, and w is the device width;

[0053] Step (7.3): Use Quantum ATK to calculate the current-voltage transfer characteristic curve and extract the device indicators of various transport performances, namely the on-state current (I ON ), subthreshold swing (SS), delay time (τ), and power delay product (PDP) to fully evaluate the application potential of MXene / MSi-2N4 contacts in short-channel FET devices;

[0054] Step (7.4): Select the FET device model with the best electrical transport properties based on the above calculations;

[0055] Specifically, step (7.1) is: plotting the calculated PLDOS of the device model, extracting the difference between the bottom energy level of the conduction band in the channel region and the Fermi level, which is the lateral Schottky barrier; based on the lateral Schottky barrier of each device model, selecting a device model with a negative Schottky barrier at the lateral interface, i.e., the lateral contact is also an ohmic contact. The selected model is the device model with ohmic contacts at both the longitudinal and lateral interfaces;

[0056] Step (7.2) is: Based on the above screening system, calculate the zero-bias transmission spectrum of the device model. The Transmission Analyzer analysis module of QuantumATK can obtain the device conductance and normalize the total device resistance by the device width, that is, the formula R device =W / G, the source / drain contact resistance can be approximated to half of the total device resistance. The device model with the minimum total device resistance or contact resistance is the optimal two-dimensional ohmic contact FET model.

[0057] The beneficial effects of the present invention are as follows: 1. The present invention uses density functional theory to calculate the work function values of both the monolayer functionalized MXene and the two-dimensional semiconductor MSi2N4. The moderate interlayer van der Waals interaction can obtain a smaller tunneling barrier and a higher carrier injection efficiency while ensuring ohmic contact, thereby realizing a two-dimensional FET device with lower contact resistance, which has positive significance for the further development and application of two-dimensional semiconductor FETs; 2. The present invention uses the new two-dimensional semiconductor MSi2N4 as the channel material of the field effect transistor device. The special structure and clean surface without dangling bonds of this material make it show important potential in short channel devices; 3. The present invention adopts M Xene is used as an electrode material. Since MXene has rich functional groups that can fully tune the work function and has good conductivity, it can form ohmic contact, and the presence of hydroxyl groups on its surface can ensure a high probability of carrier tunneling; 4. The interlayer charge from MXene to MSi2N4 is spontaneously transferred, resulting in effective doping on the two-dimensional semiconductor side, without the need for additional doping through ion implantation, etc., which simplifies the process flow; 5. The NEGF+DFT method is combined to realize the electrical transport simulation of the device model based on MXene-MSi2N4 under zero bias, and the lateral Schottky barrier and contact resistance of the system are calculated, which effectively evaluates the transport performance of the device model. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is an operational flow chart of the present invention;

[0059] Figure 2 3D images of the MXene and MSi2N4 models selected in the embodiment of the present invention; (a) is the MSi2N4 model, and (b) is the MXene model;

[0060] Figure 3 This is the energy band alignment diagram of MXene and MSi2N4 in an embodiment of the present invention;

[0061] Figure 4 Schematic diagram of a heterojunction model constructed by a single layer of MXene and a single layer of MSi2N4 in an embodiment of the present invention;

[0062] Figure 5 This is a schematic diagram of a 5nm two-dimensional FET device model based on a MXene-MSi2N4 heterostructure in an embodiment of the present invention;

[0063] Figure 62 are electron transport characteristic diagrams of the device model in the embodiment of the present invention; wherein, (a) is the zero-bias transmission spectrum diagram ((a) left) and the projected local density of states diagram ((PLDOS diagram) (a) right) of the device model based on the Ti2N(OH)2-MoSi2N4 heterojunction, and (b) is the zero-bias transmission spectrum diagram ((b) left) and the PLDOS diagram ((b) right) of the device model based on the Ti2N(OH)2-WSi2N4 heterojunction. DETAILED DESCRIPTION

[0064] The specific technical solutions of the present invention are further described in detail below with reference to specific examples.

[0065] As shown in the figure, the method of predicting the electrical transport properties of MXene-MSi2N4 heterojunction based on theoretical calculations according to the present invention includes the following steps:

[0066] Step (1): Construct a single layer of MXene and MSi2N4 and build the structure. The specific process includes:

[0067] Step (1.1): Construct a single-layer MSi2N4 structure model with the chemical formula MSi2N4 and space group number 187, where M represents a transition metal element. The specific structure model is as follows: Figure 2 As shown in (a), its geometric structure is composed of seven atomic layers stacked together. The outer layer is specifically composed of two layers of Si-N layers composed of silicon (Si) and nitrogen (N), with a layer of MN2 sandwiched in the middle. This structure is similar to inserting an MN2 monolayer into a Si2N2 monolayer to form a "sandwich" structure.

[0068] Single-layer MSi2N4 belongs to the hexagonal crystal system and has a spatial symmetry group of P63 / mmc. Each M atom is usually surrounded by six N atoms, forming a planar hexagonal structure, while Si atoms are located in the upper and lower layers, forming stable chemical bonds with N atoms.

[0069] The structural model is built by VESTA and converted into .vasp structure file type; the unit cell has a length of The vacuum layer; the selected MSi2N4 is MoSi2N4 and WSi2N4;

[0070] Step (1.2): Construct a single-layer MXene structure model with a chemical formula of M2AT2 and a space group number of 164, where M represents a transition metal element, A represents an element C or N, and T represents a surface functional group. The unit cell has a length of vacuum layer; combined Figure 2 In (b), the MXene model consists of two layers of surface functional groups -OH on the outermost sides, two layers of transition metal atoms on the outermost sides, and a central layer of C atoms along the z-axis. The unit cell is as follows: Figure 2 As shown in the black diamond box in the top view (b), the MXenes selected contain 1 C atom, 2 transition metal atoms, 2 O atoms, and 2 H atoms, including V2C(OH)2, Ti2C(OH)2, and Ti2N(OH)2;

[0071] Step (2): Use the density functional theory (DFT)-based software VASP to perform structural relaxation on the preliminarily constructed single-layer two-dimensional MXene and MSi2N4 materials, perform static self-consistent calculations and non-static self-consistent calculations, analyze the results to extract the work functions of MXene and MSi2N4, and screen MXene and MSi2N4 materials with matching work functions for the next step of calculation. The specific selection process includes:

[0072] Step (2.1): Structural optimization parameter setting: wave function cutoff energy is set to 550eV, and the convergence standard of the structural optimization force is The energy convergence criterion is 1×10 -6 eV, the K point was sampled using the Γ-centered sampling method with a sampling density of 15 × 15 × 1;

[0073] Step (2.2): The specific settings of the self-consistent calculation and electronic structure calculation parameters include: the electronic self-consistent cycle convergence standard is 1×10 -7 eV, the Γ-centered sampling method is used to sample the K point with a sampling density of 21×21×1. The high symmetry point path in the K space of the energy band calculation is Γ-MK-Γ. The plane electrostatic potential of MXene and MSi2N4 along the z axis is extracted after the self-consistent calculation converges, and the work function value is extracted;

[0074] Step (2.3): Align the vacuum energy levels of all structural models, record the energy difference from the vacuum energy level to the Fermi level of each model as the work function, and record the work function of MXene as the metal work function The work function of MSi2N4 is recorded as the semiconductor work function Filter to meet the conditions The structure of the single-layer MXene is used as a model for the subsequent steps. Figure 3 , the work functions of all MXenes are smaller than that of MSi-2N4, and the selected MXenes can be used to construct the van der Waals heterojunction model.

[0075] Step (3): construct a two-dimensional MXene / MSi2N4 van der Waals heterojunction model based on the selected MSi2N4 and MXene, perform structural optimization of the interlayer spacing and configuration, and screen a thermodynamically stable contact system;

[0076] Step (3.1): Calculate the lattice mismatch of the van der Waals heterojunction constructed based on MXene and MSi2N4, and select the structure with lattice mismatch ε < 5% for the next modeling;

[0077] Step (3.2): Use VESTA software to construct the initial MXene-MSi2N4 van der Waals heterojunction model; convert the structure file format to .vasp format file for subsequent calculations. The heterojunction model is to build the MXene layer and the MSi2N4 layer along the c-axis, where the interlayer spacing range is set to Between, ensure that the interlayer interaction is combined by van der Waals force, and along the c-axis has a length greater than The vacuum layer; heterojunction model such as Figure 4 As shown, the black rectangular box represents the unit cell boundary;

[0078] Step (3.3): Use VASP software to optimize the interlayer spacing, lattice constant, and atomic position of the MXene layer and the MSi2N4 layer in the heterojunction model; the wave function cutoff energy is set to 600 eV, and the electronic self-consistent cycle convergence standard is 1×10 -6 eV, the conjugate gradient algorithm is used for ion relaxation, and the maximum atomic force is Dispersion correction was performed using the DFT-D3 method to better describe the van der Waals force. K points were sampled using a Γ-centered sampling method with a sampling density of 15×15×1. The structural configuration with the lowest interlayer binding energy was selected for subsequent calculations.

[0079] Step (3.4): Determine 6 structural configurations through translation and rotation operations, perform configuration optimization, and find the stable configuration with the lowest binding energy for subsequent contact property calculations.

[0080] Step (4): Use VASP to calculate the longitudinal contact interface properties of the MXene-MSi2N4 two-dimensional van der Waals heterojunction; the specific steps include:

[0081] Step (4.1): Perform self-consistent calculation and band structure calculation on the optimized MXene-MSi2N4 van der Waals heterojunction, where the wave function cutoff energy is set to 550 eV and the electronic self-consistent cycle convergence standard is 1×10 -7 eV, dispersion correction was performed using the DFT-D3 method (IVDW = 11), and dipole correction was enabled (LDIPOLE = .TURE.). Self-consistent calculations used a Γ-centered sampling method to sample K points with a sampling density of 15 × 15 × 1, and the parameter LVHAR = .TRUE. The high-symmetry point path in K space for band structure calculations was Γ-MK-Γ.

[0082] Step (4.2): Extract the plane electrostatic potential along the z-axis, the Fermi level value, the binding energy value, and the electronic structure of the heterojunction, including the projected energy band and the projected state density map, obtained from the above calculations;

[0083] The thermodynamic stability is measured by the interlayer binding energy of the heterojunction, the interface contact properties are determined by the longitudinal interface Schottky barrier, and the tunneling barrier is determined by the plane electrostatic potential along the z-axis and the Fermi level. A heterojunction model with negative interlayer binding energy, a longitudinal interface Schottky barrier less than 0, and a small interlayer tunneling barrier is selected.

[0084] The interlayer binding energy, longitudinal interface Schottky barrier, and tunneling barrier of the constructed heterojunction are shown in Table 1 below:

[0085] Table 1 Interlayer binding energy, vertical interface Schottky barrier, and tunneling barrier of heterojunction

[0086]

[0087] It can be seen from Table 1 above that the constructed heterojunctions all have negative interlayer binding energy and longitudinal interface Schottky barriers, but the tunneling barrier of the heterojunction formed by Ti2N(OH)2 and MSi2N4 is significantly smaller than that formed by the other two MXenes. Therefore, Ti2N(OH)2-MoSi2N4 and Ti2N(OH)2-MoSi2N4 are selected to construct the two-dimensional FET device model;

[0088] Step (5): Based on the selected van der Waals heterojunction model, a two-dimensional FET device model that meets the ITRS2013 standard is constructed. For example, the gate length of the device model constructed here is 5.1nm; wherein, the constructed two-dimensional FET device model has the following characteristics: the model is constructed by Quantum ATK software, and is a dual-probe device, consisting of left and right electrode regions, a central scattering region, and left and right electrode extension regions sandwiched between the two, wherein the left and right electrode regions and the left and right electrode extension regions are all formed by simple repetition of the screened heterojunction model along the transmission direction c-axis, and the central region is an intrinsic single-layer MSi2N4; the a-axis is the device width direction, the b-axis is the thickness direction, and the van der Waals contact interface perpendicular to the electrode portion heterojunction has a length greater than The vacuum layer is located in the c-axis, and the device transmission direction is the c-axis. The device parameters are set as follows: the effective oxide layer thickness of the dielectric layer is 0.41nm, the dielectric constant is selected as 3.9, the gate thickness is 0.2nm, and the gate length is 5.1nm. The left and right electrodes constitute the source and drain terminals of the FET device, and the single-layer intrinsic MSi2N4 in the central area constitutes the conductive channel.

[0089] Step (6): Calculate the electron transport properties of each two-dimensional FET model, including zero-bias transmission spectrum, total device resistance, and lateral interface contact properties; the specific process includes:

[0090] Step (6.1): Combine density functional theory calculations with non-equilibrium Green's function, use the device model built in step (5), introduce the Quantum ATK workflow, and use the DeviceLCAOCalculator calculation module to set the calculation parameters;

[0091] The specific parameters are set as follows: the pseudopotential basis set is PseudoDojo, the electron temperature is set to 300K, the K-space grid density is set to 17×1×300, and the convergence criterion is 10 -5 eV, the dispersion correction parameter is set to Grimme DFT-D3;

[0092] Step (6.2): Solve the Green's function of the central region based on the above calculation parameters, and calculate the quantum transmission coefficient based on the Green's function; the calculation parameters of the transmission spectrum are set as follows: the energy range is -3.0 eV to 1.0 eV, and the sampling point is set to 201. The boundary condition is set to use Dirichlet boundary conditions along the transmission direction;

[0093] Step (6.3): solving the Green's function of the central region based on the above calculation parameters, and calculating the quantum transmission coefficient based on the Green's function;

[0094] Step (7): Selecting the optimal two-dimensional ohmic contact FET model based on the electron transport properties of each two-dimensional FET model; the specific process includes:

[0095] Step (7.1): Plot the calculated PLDOS of the device model and extract the difference between the bottom energy level of the conduction band in the channel region and the Fermi level, which is the lateral Schottky barrier.

[0096] According to the size of the lateral Schottky barrier of each device model, the device model with a negative lateral interface Schottky barrier is selected, that is, the lateral contact is also an ohmic contact. The selected model is the device model with ohmic contacts on both the longitudinal and lateral interfaces. The zero bias transmission spectrum is as follows: Figure 6 (a) Figure 6 As shown in the left figure of (b), PLDOS Figure 6 (a) Figure 6 As shown in the right figure of (b);

[0097] Step (7.2): Based on the above screening system, calculate the zero-bias transmission spectrum of the device model. The Transmission Analyzer analysis module of QuantumATK can obtain the device conductance and normalize the total device resistance by the device width, that is, the formula R device=W / G, the source / drain contact resistance can be approximated to half of the total device resistance. The device model with the minimum total device resistance or contact resistance is the optimal two-dimensional ohmic contact FET model. The total resistance and source / drain contact resistance of the device model are calculated as shown in Table 2 below:

[0098] Table 2 Total device resistance and source / drain contact resistance of the device model

[0099]

[0100] It can be seen from Table 2 above that the device model based on the Ti2N(OH)2-MoSi2N4 heterojunction has a lateral interface ohmic contact and the smallest contact resistance. Therefore, the two-dimensional FET device model based on the Ti2N(OH)2-MoSi2N4 heterojunction is the optimal device model.

Claims

1. A method for predicting the electrical transport properties of MXene-MSi2N4 heterojunctions based on theoretical calculations, characterized in that: The steps are as follows: Step (1): Based on the Materials Project database and VESTA software, a two-dimensional single-layer MXene and MSi2N4 structure model was preliminarily constructed, and the structure file was converted into a VASP readable file format; Step (2): Use the density functional theory-based software VASP to perform structural relaxation on the preliminarily constructed single-layer two-dimensional MXene and MSi2N4 materials. After self-consistent convergence, perform static self-consistent calculations and non-static self-consistent calculations. Analyze the results to extract the work functions of MXene and MSi2N4, and screen MXene and MSi2N4 materials with matching work functions for the next step of calculation. Step (3): constructing a two-dimensional MXene / MSi-2N4 van der Waals heterojunction model based on the MSi2N4 and MXene selected in step (2), optimizing the interlayer spacing and configuration, and screening a thermodynamically stable contact system; Step (4): Use VASP to calculate the electronic properties of the MXene-MSi2N4 two-dimensional van der Waals heterojunction and confirm the size of the contact interface barrier; combine the metal and semiconductor interface theory to determine the type of MXene-MSi2N4 contact interface, and select heterojunction materials with longitudinal ohmic contact and excellent tunneling performance for the next step of calculation; Step (5): constructing a two-dimensional dual-gate field-effect transistor device model based on the heterojunction material selected in step (4); Step (6): Combine density functional theory with the nonequilibrium Green's function method to calculate the electrical transport properties of the FET model using Quantum ATK software; Step (7): Based on the calculation results of step (6), the electronic transport properties of FET are studied and the FET device model with the optimal electronic transport properties is selected.

2. The method for predicting the electrical transport properties of a MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: The specific operation process of step (1) is as follows: Step (1.1): Obtain the lattice parameters, atomic coordinate positions, and bond lengths and angles of the initial structures of single-layer 2D MSi2N4 and 2D MXene based on literature and the Materials Projects database; Step (1.2): Use VESTA software to preliminarily construct a single layer of 2D MSi2N4 and 2D MXene, obtain and convert its structure file into .vasp file format.

3. The method for predicting the electrical transport properties of MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: The specific operation process of step (2) is as follows: Step (2.1): Use the DFT-based VASP software to optimize the structure of single-layer 2D MSi2N4 and 2D MXene. The energy criterion of the plane wave is 550eV and the convergence standard of the force is The energy convergence criterion is 10-6eV for structural optimization; Step (2.2): Based on the above calculations, a stable monolayer 2D MSi2N4 and 2D MXene structure is obtained; then a static self-consistent calculation is performed, and the energy convergence standard is increased to 10-7eV to ensure the calculation accuracy. The planar electrostatic potential of the two along the z axis is obtained, and the calculated work function is collected. The work function of MXene is recorded as the metal work function. The work function of MSi2N4 is recorded as the semiconductor work function Step (2.3): The prerequisite for n-type ohmic contact is: Screen MXenes that meet the prerequisites for the next step of building a van der Waals heterojunction model.

4. The method for predicting the electrical transport properties of a MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: The specific operation process of step (3) is as follows: Step (3.1): Based on the results of step (2), VESTA was used to build a MXene / MSi2N4 heterojunction model and determine that the lattice mismatch between the two is <5%. Step (3.2): Use VESTA software to convert the structure file format into .vasp format file for subsequent calculations; Step (3.3): Use VESTA to construct structural models with different interlayer spacings, optimize the interlayer spacing, screen the structure with the lowest interlayer binding energy, and determine the optimal interlayer spacing; Step (3.4): Determine 6 structural configurations through translation and rotation operations, perform configuration optimization, and find the stable configuration with the lowest binding energy for subsequent contact property calculations.

5. The method for predicting the electrical transport properties of MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: The specific operation process of step (4) is as follows: Step (4.1): Perform self-consistent calculation, state density and energy band calculation on the MXene / MSi2N4 van der Waals heterojunction after completing the structural optimization of step (3), where the wave function cutoff energy is set to 550 eV and the electronic self-consistent energy convergence standard is 1×10 -7 eV, DFT-D3 was used to perform dispersion correction on the van der Waals heterojunction, and the self-consistent calculation adopted a Γ-centered sampling method to sample the K points with a sampling density of 15×15×1; the high symmetry point path in the K space in the band structure calculation was Γ-MK-Γ; Step (4.2): Extract the plane electrostatic potential, Fermi level, projected state density and projected energy band diagram obtained in the above calculations; the interface contact properties are determined by the longitudinal interface Schottky barrier, and the tunneling properties at the interface are determined by the interface tunneling barrier; select a heterojunction model with negative interlayer binding energy, a longitudinal interface Schottky barrier less than 0, and a small interlayer tunneling barrier.

6. The method for predicting the electrical transport properties of a MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: The specific operation process of step (5) is as follows: Step (5.1): Based on the MXene / MSi2N4 heterojunction selected above, a dual-gate FET device was constructed using Quantum ATK, where the MXene / MSi2N4 heterojunction is the minimum repeating unit of the left and right electrode regions, and the central scattering region is composed of the repeating intrinsic 2D MSi2N4. The a-axis is the device width direction, the b-axis is the thickness direction, and the longitudinal van der Waals heterojunction contact interface perpendicular to the electrode part and the vacuum layer thickness is greater than The c-axis is the device transmission direction; Step (5.2): The specific parameters are that the effective oxide layer thickness of the dielectric layer is 0.41nm, the dielectric constant is selected as 3.9, the gate thickness is 0.2nm, and the channel length is 5.1nm; the left and right electrode regions constitute the source and drain ends of the FET device, and the central scattering region is the channel region of the FET device.

7. The method for predicting the electrical transport properties of a MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: In step (6), the electrical transport properties include zero-bias transmission spectrum, device contact resistance and projected localized state density.

8. The method for predicting the electrical transport properties of a MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: The specific operation process of step (6) is as follows: Step (6.1): Combine density functional theory calculations with non-equilibrium Green's functions to perform electrical transport model calculations on the non-equilibrium device. Import the device model into the Quantum ATK workflow and set the device's DeviceLCAO calculation parameters. Step (6.2): Based on the above settings, the transmission probability spectrum T(E, V) of the device is obtained by solving the non-equilibrium Green's function in the central region; Step (6.3): Based on the calculated transmission probability spectrum, extract the zero-bias transmission spectrum, differential conductance, and project the local density of states.

9. The method for predicting the electrical transport properties of a MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 8, characterized in that: In step (6.1), the device LCAO calculation parameters include boundary conditions, pseudopotential basis set selection, electron temperature, K-space grid density, convergence criteria and dispersion correction parameters.

10. The method for predicting the electrical transport properties of a MXene-MSi2N4 heterojunction based on theoretical calculation according to claim 1, characterized in that: The specific operation process of step (7) is as follows: Step (7.1): Extract the lateral Schottky barrier based on the projected state density of the device model calculated in step (6), select the model in which the lateral interface is an ohmic period, and combine the model selected in step (4) to obtain a device model in which both the lateral and vertical interfaces are ohmic contacts; Step (7.2): Calculate the zero-bias differential conductance of the selected device model, using the formula R c =G / 2w to get the intrinsic contact resistance of the device. In this formula, R c represents the intrinsic contact resistance, G is the differential conductance, extracted from the zero-bias transmission spectrum, and w is the device width; Step (7.3): Use Quantum ATK to calculate the current-voltage transfer characteristic curve and extract the device indicators of various transport performances, namely on-state current, subthreshold swing, delay time, and power-delay product, to evaluate the application potential of MXene / MSi2N4 contacts in short-channel FET devices; Step (7.4): Based on the above calculations, select the FET device model with the optimal electrical transport properties.

Images