Simulation modeling method and device for nanoscale device

Abstract

The invention discloses a simulation modeling method and device for a nanoscale device, relates to the technical field of semiconductor modeling and simulation, and aims to solve the problem that efficiency and precision cannot be considered in nanoscale device modeling and simulation in the prior art. The method comprises the steps that key material parameters are extracted from electronic structure basic data of a target nanoscale device, a default material parameter library in TCAD software is corrected based on the key material parameters, and a corrected material parameter library is obtained; electronic structure basic data is obtained based on a density functional theory; based on the corrected material parameter library, optimizing adjustable parameters of a ballistic transport model preset in TCAD software in combination with a transport IV curve to obtain a ballistic transport model after parameter optimization; the transportation IV curve is generated by adopting a quantum transportation method based on an unbalanced Green function in combination with key material parameters; and integrating the key material parameters with the ballistic transport model after parameter optimization, and constructing a simulation parameter library of the target nanoscale device.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor modeling simulation, in particular to a simulation modeling method and device for nanoscale devices. BACKGROUND

[0002] With the semiconductor technology advancing towards sub-10 nanometer nodes, nanoscale device modeling simulation is facing systematic challenges.

[0003] For the modeling simulation of nanoscale devices, there are two mainstream modeling simulation systems in the prior art: the first modeling simulation system is an atomic scale simulation method combining density functional theory and non-equilibrium Green's function, which can accurately capture quantum behavior, but the calculation cost is too high to support large-scale design iteration in enterprises; the second modeling simulation system is to use technology computer aided design (TCAD) software for simulation, which can realize rapid iteration, but the default model of the technology computer aided design software generally lacks key parameters such as effective mass, dielectric constant, and interface barrier required by nanoscale devices, and its quantum correction model relies on empirical parameter calibration, lacks physical correlation with material thickness and energy band structure, and has insufficient universality across materials and sizes, which cannot accurately describe the quantum behavior of nanoscale devices. SUMMARY

[0004] The purpose of the present application is to provide a simulation modeling method and device for nanoscale devices to solve the problem that the existing technology cannot balance efficiency and accuracy in nanoscale device modeling simulation.

[0005] In order to achieve the above purpose, the present application provides the following technical solutions: In a first aspect, the present application provides a simulation modeling method for nanoscale devices, comprising: extracting key material parameters from electronic structure basic data of a target nanoscale device, and modifying the default material parameter library in the TCAD software based on the key material parameters to obtain a modified material parameter library; the electronic structure basic data is obtained based on the density functional theory; based on the modified material parameter library, combining the transport IV curve to optimize the adjustable parameters of the pre-set ballistic transport model in the TCAD software to obtain a parameter-optimized ballistic transport model; the transport IV curve is generated by combining the key material parameters with the quantum transport method based on the non-equilibrium Green's function; based on the key material parameters and the parameter-optimized ballistic transport model, constructing a simulation parameter library for the target nanoscale device.

[0006] Compared with existing technologies, the simulation modeling method for nanoscale devices provided by this invention extracts physical correlation parameters based on density functional theory to correct the TCAD parameter library, and combines the quantum transport method of non-equilibrium Green's function for high-precision IV curve calibration. This avoids the high computational cost of atomic-scale simulation methods that combine pure density functional theory and non-equilibrium Green's function, and also makes up for the shortcomings of traditional TCAD models, such as missing key parameters and reliance on empirical calibration. By extracting reusable key material parameters in one go, there is no need to repeat the entire atomic-level calculation process. At the same time, by leveraging the rapid iteration capability of TCAD software and the globally optimized ballistic transport model, the simulation efficiency and adaptability to large-scale design iterations are greatly improved while ensuring the accuracy of quantum behavior simulation of nanoscale devices, achieving a dual optimization of accuracy and efficiency.

[0007] Optionally, before optimizing the adjustable parameters of the ballistic transport model pre-set in the TCAD software based on the modified material parameter library and the transport IV curve, the construction method also includes: A source-channel-drain structure model corresponding to the target nanoscale device is constructed; the material property parameters of the source-channel-drain structure model are assigned values ​​using key material parameters. A quantum transport method based on the non-equilibrium Green's function was used to simulate the source-channel-drain structure model under different simulation combinations, and the current values ​​corresponding to each voltage were obtained under the same channel length; the simulation combination refers to the combination of different channel lengths and different voltages. By correlating the voltage and the corresponding current value under the same channel length, an IV curve corresponding to the same channel length is generated. The IV curves for different channel lengths are integrated into a set to obtain the transport IV curves.

[0008] Optionally, based on the modified material parameter library, and combined with the adjustable parameters of the pre-set ballistic transport model in the transport IV curve optimization TCAD software, a parameter-optimized ballistic transport model is obtained, including: Substitute the parameters from the modified material parameter library that are compatible with the drift-diffusion model into the drift-diffusion model to generate drift-diffusion IV curves for all simulation combinations. For each simulation combination, calculate the single combination error between the corresponding transport IV curve and the drift-diffusion IV curve, summarize the single combination errors of all simulation combinations, and construct the global error. With the goal of minimizing global error, the adjustable parameters of the ballistic transport model are iteratively adjusted until the single-combination error of all simulation combinations is less than a preset threshold, thus obtaining the ballistic transport model with optimized parameters. The adjustable parameters include at least one of ballistic velocity, carrier relaxation time, and quantum confinement factor.

[0009] Optionally, before extracting key material parameters from the fundamental electronic structure data of the target nanoscale device, the method further includes: determining the fundamental electronic structure data based on density functional theory; The fundamental data of electronic structure were determined based on density functional theory, including: Obtain the basic parameters of the target nanoscale device; the basic parameters include the chemical composition, crystal structure, heterojunction interface structure to be simulated, and applied external field conditions of the target nanoscale device; The basic parameters are input into first-principles software based on density functional theory to generate band structure data, charge density distribution data, potential distribution data of materials under a uniform electric field, and projected band data of metal-semiconductor heterojunctions for the target nanoscale device. The band structure data, charge density distribution data, electric potential distribution data, and projected band structure data are identified as the fundamental data for electronic structure.

[0010] Optionally, the key material parameters include at least the effective mass of charge carriers, the non-parabolic coefficient, the band width, and the energy difference between energy valleys; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The band structure data is fitted to obtain the band structure curve; the band structure curve covers the band distribution information of the entire Brillouin zone; A second-order polynomial is fitted to the band structure curve near the extreme point. The second derivative is calculated from the fitting result, and the effective mass of the charge carrier is calculated based on the second derivative. For the target region in the band structure curve that meets the preset conditions, a high-order fitting method is used to fit the band structure curve data of the target region, or the band structure curve data of the target region is fitted based on the Kane model. The non-parabolic coefficients are obtained through fitting optimization. The preset conditions include that the electron energy in the band structure curve is outside the band extremum region and the parabolic band model fails. Based on the band distribution information of the entire Brillouin zone covered by the band structure curve, the maximum and minimum energy values ​​of the target band are extracted respectively, and the difference between the maximum and minimum energy values ​​is determined as the band width; the target band includes the valence band and the conduction band, with the valence band centered on the band where the highest occupied band is located, and the conduction band centered on the band where the lowest unoccupied band is located. In the band structure curve, identify the band extrema at each high symmetry point, determine the energy value corresponding to the global extremum from the band extrema at each high symmetry point, and calculate the difference between the extremum energy of each high symmetry point and the energy value corresponding to the global extremum point to form a set of energy differences between different energy valleys of the target nanoscale device; among the band extrema at each high symmetry point, the conduction band corresponds to the band minimum, the valence band corresponds to the band maximum, and the extremum energy of each high symmetry point is the energy value corresponding to the band extrema at each high symmetry point.

[0011] Optionally, key material parameters also include the relative permittivity; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The potential distribution data is fitted to obtain the potential distribution curve; Based on the potential distribution curve, the electric field strength within the material is calculated. Combined with the applied vacuum electric field strength, the formula is used: ; The relative permittivity of the target nanoscale device was calculated; where, It is the relative permittivity; The electric field strength in vacuum; The electric field strength within the material.

[0012] Optionally, key material parameters also include the physical thickness of the device; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The charge density distribution data is fitted to obtain the charge density distribution curve; the charge density distribution curve is a charge density profile curve along the material thickness direction; The spatial range in the charge density distribution curve where the charge density value is greater than a preset charge density threshold is determined, and the size of the spatial range is used as the physical thickness of the target nanoscale device.

[0013] Optionally, key material parameters also include the Schottky barrier height; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The projected band structure of the metal-semiconductor heterojunction is obtained by fitting the projected band structure data. Based on the projected band structure of the metal-semiconductor heterojunction, the Fermi level of the metal electrode and the conduction band bottom energy or valence band top energy of the semiconductor channel material are determined, and the energy difference between the Fermi level of the metal electrode and the conduction band bottom energy or valence band top energy is calculated. The energy difference is defined as the height of the Schottky barrier.

[0014] Optionally, key material parameters also include contact resistance; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: A quantum transport method based on non-equilibrium Green's function is used to process the basic electronic structure data and device contact structure parameters to obtain carrier transmittance data. Among them, the device contact structure parameters include the contact type between the source / drain and the metal, the atomic-level structure information of the contact interface, and the applied bias voltage. Substituting the carrier transmittance data into the quantum transport current formula based on the non-equilibrium Green's function theory: ; By performing integration, the current is obtained; where, For energy; Carrier transmittance; For current; For electron charge; It is Planck's constant; The Fermi distribution function of the left electrode; The Fermi distribution function of the right electrode; By substituting the applied bias voltage and current into Ohm's law formula, the contact resistance between the source / drain of the target nanoscale device and the metal is calculated.

[0015] Secondly, the present invention also provides a simulation modeling apparatus for nanoscale devices, comprising: The extraction module is used to extract key material parameters from the basic electronic structure data of the target nanoscale device, and to correct the default material parameter library in the TCAD software based on the key material parameters to obtain the corrected material parameter library; the basic electronic structure data is obtained based on density functional theory. The optimization module is used to optimize the adjustable parameters of the ballistic transport model pre-set in the TCAD software based on the modified material parameter library and the transport IV curve, so as to obtain the parameter-optimized ballistic transport model; the transport IV curve is generated by the quantum transport method based on the non-equilibrium Green's function combined with key material parameters. The module is used to build a simulation parameter library for target nanoscale devices based on key material parameters and optimized ballistic transport models. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1 A flowchart illustrating a simulation modeling method for nanoscale devices provided in an embodiment of the present invention; Figure 2 A schematic diagram of the band structure curve provided for one embodiment of the present invention; Figure 3 A potential distribution diagram provided for one embodiment of the present invention; Figure 4 A charge density distribution diagram provided for one embodiment of the present invention; Figure 5 A schematic diagram of device contact types provided for an embodiment of the present invention. Figure 6 This is a schematic diagram of the band structure of a metal-semiconductor heterostructure. Figure 7 A comparison of IV curves for two simulation methods under different simulation combinations, provided as an embodiment of the present invention; Figure 8 A schematic diagram of the simulation process for a nanoscale device provided in one embodiment of the present invention; Figure 9 A schematic diagram of a simulation modeling device for nanoscale devices provided in an embodiment of the present invention. Detailed Implementation

[0017] To facilitate a clear description of the technical solutions in the embodiments of the present invention, the terms "first" and "second" are used to distinguish identical or similar items with essentially the same function and effect. For example, the first threshold and the second threshold are merely used to distinguish different thresholds and do not limit their order. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" are not necessarily different.

[0018] It should be noted that in this invention, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0019] In this invention, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between the associated objects, indicating that three relationships can exist.

[0020] Before introducing the embodiments of the present invention, some terms will be explained.

[0021] DFT stands for Density Functional Theory.

[0022] NEGF, Non-Equilibrium Green's Function.

[0023] EM-NEGF, Effective mass Non-Equilibrium Green's Function.

[0024] DD stands for Drift-Diffusion.

[0025] TB-NEGF, Tight-Binding Non-Equilibrium Green's Function.

[0026] KP-NEGF, k·p Non-Equilibrium Green's Function.

[0027] like Figure 1 As shown, embodiments of the present invention provide a simulation modeling method for nanoscale devices, including: Step 100: Extract key material parameters from the basic electronic structure data of the target nanoscale device, and modify the default material parameter library in the TCAD software based on the key material parameters to obtain the modified material parameter library; the basic electronic structure data is obtained based on density functional theory. Specifically, correcting the default material parameter library in TCAD software based on key material parameters means: for parameters missing in the default material parameter library, the corresponding items from the extracted key material parameters are added to the corresponding positions; for parameters that already exist in the default material parameter library but need to be optimized or updated, the corresponding items from the extracted key material parameters are used to replace the original parameters.

[0028] Understandably, before extracting key material parameters from the fundamental electronic structure data of the target nanoscale device, the method also includes: determining the fundamental electronic structure data based on density functional theory; The fundamental data of electronic structure were determined based on density functional theory, including: (1) Obtain the basic parameters of the target nanoscale device; the basic parameters include the chemical composition, crystal structure, and heterojunction interface structure of the target nanoscale device to be simulated; Specifically, crystal structure includes lattice constant and space group. Lattice constant is the spatial periodicity of the atomic arrangement, while space group is the symmetrical distribution pattern of atoms.

[0029] Heterojunction interface structure specifically refers to the interface bonding mode between metal electrode material and semiconductor channel material. The interface bonding mode can include atomic stacking order and bonding type.

[0030] The applied external field conditions mainly refer to the uniform electric field strength.

[0031] (2) Input the basic parameters into the first-principles software based on density functional theory to generate the band structure data, charge density distribution data, potential distribution data of the material under a uniform electric field and the projected band data of the metal-semiconductor heterojunction of the target nanoscale device. Specifically, first-principles software based on density functional theory is a class of software that calculates the microscopic electronic structure of materials using fundamental principles of quantum mechanics without relying on empirical parameters. Examples of first-principles software include VASP, Quantum ESPRESSO, and CASTEP.

[0032] Band structure data describes the relationship between the allowed energy states of electrons in a material and the wave vector, and is the core basis for extracting key parameters such as band gap and effective mass.

[0033] Charge density distribution data is the probability data of electron distribution in the material space.

[0034] The potential distribution data of a material under a uniform electric field is the data on how the internal potential changes with spatial position after a uniform electric field is applied to both ends of the material.

[0035] The projected band structure data of a metal-semiconductor heterojunction is the data that projects the band structure at the metal-semiconductor interface onto a specific direction.

[0036] Carrier transmittance data is the probability data that carriers such as electrons and holes can pass through the contact interface of a device, such as the source-drain interface.

[0037] The principle behind generating various electronic structure data by inputting basic information in this step lies in the core concept of density functional theory: "Electron density determines all properties of a material." After inputting the basic parameters, the first-principles software will perform the following specific logical calculations: ① Input the chemical composition and crystal structure. The first-principles software first determines the spatial arrangement of atoms and the configuration of extranuclear electrons, and then solves the energy state of electrons through the Schrödinger equation to generate band structure data.

[0038] ② Input the chemical composition and crystal structure, and the first-principles software will simultaneously solve for the spatial distribution of electrons and generate charge density distribution data.

[0039] ③ Input the uniform electric field strength. After the first-principles software applies the uniform electric field to both ends of the material, it calculates the change of internal electric potential with spatial position and generates the potential distribution data of the material under the uniform electric field.

[0040] ④ Input the heterojunction interface structure. The first-principles software generates the projected band structure of the metal-semiconductor heterojunction through the electronic interactions of the interface atoms. ⑤ Based on the generated band structure data and projected band data of the metal-semiconductor heterojunction, combined with the energy state of electrons and interface characteristics, the carrier transmittance data is further derived; that is, chemical composition, crystal structure and heterojunction interface structure jointly affect the magnitude of carrier transmittance.

[0041] (3) The band structure data, charge density distribution data, potential distribution data of materials under uniform electric field and projected band data of metal-semiconductor heterojunction are determined as the basic data of electronic structure.

[0042] Step 200: Based on the modified material parameter library, the adjustable parameters of the ballistic transport model pre-set in the TCAD software are optimized by combining the transport IV curve to obtain the parameter-optimized ballistic transport model; the transport IV curve is generated by combining the quantum transport method based on the non-equilibrium Green's function with key material parameters; Step 300: Based on the key material parameters and the optimized ballistic transport model, construct a simulation parameter library for the target nanoscale device.

[0043] The beneficial effects of this embodiment: This invention addresses the core shortcomings of two mainstream systems in existing technologies by extracting key parameters using DFT and correcting them with the TCAD parameter library, combined with a collaborative scheme of generating IV curves using NEGF to optimize the ballistic transport model. 1) To address the issues of high computational cost and inability to support large-scale iterations in the first system: key material parameters are extracted only once through DFT, and the results can be reused without repeating atomic-level full-process calculations, which greatly reduces the time consumption. At the same time, the rapid iteration capability of TCAD software is leveraged to meet the needs of large-scale design.

[0044] It needs to be explained in detail that the first modeling and simulation system in the existing technology adopts an atomic-scale simulation method that combines density functional theory with non-equilibrium Green's functions. The core reason why this method is time-consuming due to its repeated atomic-level full-process calculations is that it needs to start from the fundamental physical laws at the atomic scale and perform high-precision simulations of the device from start to finish. This involves the complex interactions of a massive number of microscopic particles and large-scale numerical calculations, which can be divided into three points: ① The microscopic scale and number of particles of the objects being calculated are enormous.

[0045] Atomic-level full-process calculations need to cover all atoms in the device (including atoms in the source, drain, channel, and interface regions). Although nanoscale devices are tiny, the number of atoms still reaches thousands to tens of thousands.

[0046] The problem requires solving the equations of motion of the extranuclear electrons of each atom (such as the Schrödinger equation), and there are Coulomb interactions between electrons. It is a many-body problem that cannot be solved directly and needs to be approximated by methods such as density functional theory (DFT). However, it still requires processing massive amounts of data such as wave functions and charge densities of electrons.

[0047] For heterojunction devices, such as metal-semiconductor interfaces, it is also necessary to consider the interfacial coupling and chemical bonding between atoms of different materials, which further increases the computational complexity.

[0048] ② The high-dimensionality of the computation process and the need for iterative solutions.

[0049] The core of atomic-level simulation is to obtain the electronic structure of materials, a process that requires multi-dimensional numerical iterations: It is necessary to traverse all the wave vector points in the Brillouin zone. Each wave vector point corresponds to an energy state of the electron. To ensure accuracy, the number of K points usually needs to reach hundreds or even thousands. Solving for electronic structures requires self-consistent iterative convergence. Each iteration requires updating the charge density and potential energy surface, and then resolving the electronic wave function until the convergence accuracy is met. This typically requires tens to hundreds of iterations. If device transport characteristics are involved, the Green's function and carrier transmittance must be solved in addition to the electronic structure, and integration must be performed across different energy ranges, further extending the computation time.

[0050] ③ There is a lack of reusable intermediate results, and each simulation must start from scratch.

[0051] Traditional atomic-level full-process calculations, such as DFT-NEGF, lack a fixed parameter library storage mechanism. Each simulation, for different channel lengths, voltages, or materials, requires repeating the following full process: The entire process lacks any reusable intermediate parameters. Even if only a single variable such as voltage or channel length is adjusted, all steps must be re-executed, resulting in linear or even exponential growth in computational costs. This makes it impossible to support large-scale design iterations, such as the rapid selection of hundreds of device structures in an enterprise.

[0052] The present invention extracts key material parameters in one go through DFT and can be reused for a long time. Subsequently, only the TCAD model needs to be modified based on the parameters, without repeating atomic-level calculations, which avoids the above-mentioned time-consuming pain points.

[0053] 2) Addressing the issues of missing key parameters in the second system, reliance on empirical calibration for quantum correction models, and insufficient universality: Physical correlation parameters extracted by DFT are used to supplement / replace the default parameter library of TCAD, filling the parameter gaps for nanoscale devices, especially new material devices. Then, based on the high-precision transport IV curves generated by NEGF, the ballistic transport model is optimized to establish a strong correlation between parameters and material physical properties, improving universality across materials and sizes, and accurately describing quantum behavior. Ultimately, a win-win situation of accuracy and efficiency is achieved: it retains the atomic-level simulation accuracy of DFT-NEGF while leveraging the rapid iteration advantage of TCAD, solving the core pain point that existing technologies cannot simultaneously address both.

[0054] In one exemplary embodiment, the key material parameters include at least the effective carrier mass, non-parabolic coefficient, band width (BW), and energy difference between valleys; Step 100: Extract key material parameters from the fundamental electronic structure data of the target nanoscale device, including: S110: The band structure data is fitted to obtain the band structure curve; the band structure curve covers the band distribution information of the entire Brillouin zone; Specifically, band structure curves can intuitively show the correspondence between electron energy and wave vector, for example... Figure 2 The illustrated band structure curve allows for the simultaneous acquisition of the band gap between the conduction and valence bands, providing a foundation for analyzing the core characteristics of the device. It should be noted that... Figure 2 The four related sub-graphs shown are essentially multi-dimensional focused presentations of the same band structure curve, not four independent curves. The separation is to clearly demonstrate the physical features directly related to the extraction of key material parameters, specifically highlighting the region near the band extrema, the non-parabolic region outside the extrema, the band span across the entire Brillouin zone, and the band extrema at high symmetry points. These correspond to the extraction logic supporting the effective carrier mass, non-parabolic coefficient, band width, and energy difference between valleys. This regional decomposition presentation method ensures the integrity of the band structure information while avoiding the interpretation confusion caused by the overlap of multiple types of information in a single graph, making the complex band structure and parameter extraction logic intuitively correspond.

[0055] S111: Perform a second-order polynomial fitting on the band structure curve near the extreme point, calculate the second derivative based on the fitting result, and calculate the effective mass of the charge carrier based on the second derivative; where the extreme point is the global minimum point of the conduction band, i.e., the bottom of the conduction band, or the extreme point is the global maximum point of the valence band, i.e., the top of the valence band.

[0056] From a physical perspective, the core significance of the effective mass of charge carriers lies in describing the inertia exhibited by electrons or holes in a crystal in response to applied electric and magnetic fields. Its magnitude directly reflects the band curvature; the greater the curvature, the smaller the effective mass, and the higher the carrier mobility. The theoretical basis for this principle is the parabolic dispersion relation near the band extrema, as follows: Near extreme points such as the conduction band trough and valence band trough, the dispersion relation between electron energy and wave vector satisfies a parabolic approximation: (1); Based on this dispersion relation, the fundamental formula for the effective mass of charge carriers can be approximated in one dimension as follows: (2); in, For wave vectors; For electrons in wave vector Energy at the location; For effective quality; Energy at the extreme point; To reduce Planck's constant; It is the second derivative of energy with respect to the wave vector (corresponding to the band curvature).

[0057] In practical application, it is necessary to first locate the extreme points. Extreme points are commonly found in... By fitting the band structure curve with a second-order polynomial near the extreme points such as the K-point and the K-point, the second derivative can be directly derived from the coefficients of the quadratic term obtained by fitting. Then, by substituting it into formula (2), the effective mass of the charge carrier can be calculated.

[0058] It should be added that if the material of the target nanoscale device is anisotropic, it is necessary to fit along different high symmetry directions to obtain the longitudinal effective mass and the transverse effective mass, so as to accurately reflect the carrier transport characteristics of the material in different directions.

[0059] S112: For the target region in the band structure curve that meets the preset conditions, a high-order fitting method is used to fit the band structure curve data of the target region, or the band structure curve data of the target region is fitted based on the Kane model, and the non-parabolic coefficients are obtained through fitting optimization; the preset conditions include that the electron energy in the band structure curve is outside the band extreme region and the parabolic band model fails. The non-parabolic coefficient is introduced to correct the band structure when the electron energy is far from the band extremum. This coefficient is commonly found in narrow bandgap semiconductor devices.

[0060] The Kane model can be written as: (3); in, The mass of free electrons; The coefficients are non-parabolic. Higher-order fitting methods can be written as: (4); In practice, the extraction process requires fitting within a certain energy range (e.g., 0-0.5 eV) above the conduction band bottom. curve; Next, we use the Kane model or Taylor expansion to the fourth-order terms: (5); in, ; These are the coefficients of the fourth-order terms.

[0061] Finally, the non-parabolic coefficients were determined by least squares fitting. .

[0062] S113: Based on the band distribution information of the entire Brillouin zone covered by the band structure curve, the maximum and minimum energy values ​​of the target band are extracted respectively, and the difference between the maximum and minimum energy values ​​is determined as the band width; the target band includes the valence band and the conduction band, with the valence band centered on the band where the highest occupied band is located, and the conduction band centered on the band where the lowest unoccupied band is located. Specifically, band width is the energy broadening of a given energy band (usually the valence band or conduction band) across the entire Brillouin zone, i.e., the difference between its maximum and minimum values, expressed by the formula: (6); in, For the bandgap; The maximum energy value of the target band; The minimum energy value of the target band.

[0063] S114: Identify the band extrema at each high symmetry point in the band structure curve, determine the energy value corresponding to the global extremum from the band extrema at each high symmetry point, and calculate the difference between the extremum energy of each high symmetry point and the energy value corresponding to the global extremum point one by one to form a set of energy differences between different energy valleys of the target nanoscale device; among the band extrema at each high symmetry point, the conduction band corresponds to the band minimum, the valence band corresponds to the band maximum, and the extremum energy of each high symmetry point is the energy value corresponding to the band extrema of each high symmetry point.

[0064] Specifically, the energy difference between energy valleys is the energy difference between different energy valleys (e.g., in the conduction band). The energy difference between (X, L valleys).

[0065] In practical implementation, identifying the band extrema at each high symmetry point means identifying the band minimum (conduction band) or maximum (valence band) at each high symmetry point, and then applying the formula: (7); Calculate the energy difference between each high-symmetry point and the global extremum point; where... The corresponding energy difference between energy valleys; For the first Energy values ​​at highly symmetric points; This represents the energy value corresponding to the global extreme point.

[0066] The beneficial effects of this embodiment are as follows: 1) For the carrier transport characteristics dominated by quantum effects in nanoscale devices, by accurately extracting key parameters such as non-parabolic coefficients and energy differences between energy valleys, the limitations of traditional TCAD models that rely solely on effective mass approximation are overcome, making it more consistent with the actual transport scenario of carriers far from the band extremum in short-channel devices; 2) A direction-separated fitting strategy is adopted for anisotropic materials, and the Kane model is introduced for narrow bandgap semiconductors to correct the extracted parameters, so that the extracted parameters can accurately reflect the intrinsic physical properties of the materials, providing physically accurate basic data for subsequent TCAD model correction and avoiding the accumulation of errors caused by empirical parameters; 3) The extraction method that combines full Brillouin zone scanning with high symmetry point identification ensures the integrity of parameters such as band width and energy differences between energy valleys, providing support for accurate simulation of key performance characteristics such as device mobility and hot electron effects.

[0067] In one exemplary embodiment, the key material parameter also includes the relative permittivity; Step 100: Extract key material parameters from the fundamental electronic structure data of the target nanoscale device, including: S120: Fit the potential distribution data to obtain the potential distribution curve; where the potential distribution curve is, for example... Figure 3 As shown, the horizontal axis represents distance, and the vertical axis represents electric potential; S121: Based on the potential distribution curve, calculate the electric field intensity within the material, and combine it with the applied vacuum electric field intensity according to the formula: (8); The relative permittivity of the target nanoscale device was calculated; where, It is the relative permittivity; The electric field strength in vacuum; The electric field strength within the material.

[0068] Understandably, before performing S120, based on the principle of macroscopic parallel-plate capacitors, a uniform electric field is first applied to both ends of the material, and then the potential distribution data inside the material is calculated using first-principles software. The uniform electric field is consistent with the external field conditions applied in determining the fundamental electronic structure data based on density functional theory.

[0069] The beneficial effects of this embodiment are: 1) It breaks through the limitations of traditional dielectric constant measurement relying on experimental or empirical estimation. By combining DFT simulation with theoretical derivation of electric field-potential distribution, it achieves accurate extraction of the relative dielectric constant of materials at the nanoscale, adapting to the microscale characteristics of sub-10 nanometer devices; 2) The extraction process is directly related to the electronic structure of the material, avoiding the size effect error of macroscopic measurement methods at the nanoscale, so that the dielectric constant parameter can accurately reflect the polarization characteristics of the material in the actual working environment of the device; 3) It provides core parameter support for the capacitance characteristic simulation in the TCAD model, solving the technical gap that the existing DFT method cannot obtain the capacitance characteristics of new material devices.

[0070] In one exemplary embodiment, key material parameters also include the physical thickness of the device; Step 100: Extract key material parameters from the fundamental electronic structure data of the target nanoscale device, including: S130: Fit the charge density distribution data to obtain a charge density distribution curve; the charge density distribution curve is a charge density profile curve along the material thickness direction; wherein, the charge density distribution curve is, for example... Figure 4 As shown, the horizontal axis represents distance, and the vertical axis represents charge density; S131: Determine the spatial range in the charge density distribution curve where the charge density value is greater than a preset charge density threshold, and use the size of this spatial range as the physical thickness of the target nanoscale device; the preset charge density threshold is set based on the intrinsic charge density characteristics of the material. The preset charge density threshold is not empty.

[0071] The beneficial effects of this embodiment are: 1) The thickness is defined based on the non-zero spatial range of charge density distribution, which overcomes the defect that traditional geometric thickness measurement cannot reflect the actual extension range of the electronic structure of the material, and is more in line with the characteristics of the close relationship between the physical thickness of the material and electron transport in nanoscale devices; 2) The extracted results are directly related to the electronic distribution characteristics of the material, so that the thickness parameter in the TCAD model is no longer a simple geometric dimension, but a functional level thickness that is strongly related to physical effects such as carrier transport and capacitance characteristics, thereby improving the physical realism of the simulation.

[0072] In one exemplary embodiment, key material parameters also include contact resistance; Step 100: Extract key material parameters from the fundamental electronic structure data of the target nanoscale device, including: S150: The quantum transport method based on the non-equilibrium Green's function is used to process the basic electronic structure data and device contact structure parameters to obtain carrier transmittance data; among which, the device contact structure parameters include the contact type between the source / drain and the metal, the atomic-level structure information of the contact interface, and the applied bias voltage; Substituting the carrier transmittance data into the quantum transport current formula based on the non-equilibrium Green's function theory: (9); By performing integration, the current is obtained; where, For energy; Let E be the carrier transmittance, which is related to the energy E. The transmission probability is different at different energies. For current; For electron charge; It is Planck's constant; Let Fermi distribution function be the function of the left electrode, which describes the occupancy probability of an electron with energy E in the left electrode. Let be the Fermi distribution function of the right electrode, which describes the occupancy probability of an electron with energy E in the right electrode. The integral variable in formula (9) is... The integral range covers the possible energy range of charge carriers.

[0073] Specifically, see Figure 5 The contact types in the device contact structure parameters specifically include top contacts, edge contacts, and semi-enclosed contacts. The calculation of carrier transmittance data requires combining the Landau formula and using the NEGF method to solve for the carrier transport probability at the source / drain and metal interface.

[0074] S151: Substitute the applied bias voltage and current into Ohm's law formula to calculate the contact resistance between the source / drain of the target nanoscale device and the metal.

[0075] The beneficial effects of this embodiment are: 1) It integrates basic electronic structure data with device contact structure parameters, and realizes physical-level calculation of contact resistance through the NEGF method, avoiding the low efficiency and high error of traditional manual extraction or empirical formula estimation; 2) It covers extraction scenarios of various contact types, adapts to the structural design requirements of different nanodevices, and enables contact resistance parameters to accurately reflect the impact of interface atomic-level characteristics on device performance; 3) The extraction process can be automatically completed through programming language scripts, solving the problem that existing manual extraction methods cannot meet the needs of rapid evaluation of large-scale device designs, and improving the efficiency and consistency of parameter extraction.

[0076] In one exemplary embodiment, step 100: the key material parameters further include the Schottky barrier height; step 100: extracting key material parameters from the basic electronic structure data of the target nanoscale device, including: S140: Fit the projected band data of the metal-semiconductor heterojunction to obtain the projected band curve of the metal-semiconductor heterojunction; Understandably, before S140, a heterojunction structure of metal and semiconductor channel materials needs to be built, and then the projected bandgap data is calculated using first-principles software.

[0077] For example, the projected band structure of a metal-semiconductor heterojunction can be referenced. Figure 6 .

[0078] S141: Based on the projected band structure of the metal-semiconductor heterojunction, determine the Fermi level of the metal electrode and the conduction band bottom energy or valence band top energy of the semiconductor channel material, and calculate the energy difference between the Fermi level of the metal electrode and the conduction band bottom energy or valence band top energy. S142: Determine the energy difference as the height of the Schottky barrier.

[0079] The beneficial effects of this embodiment are as follows: 1) Based on the direct derivation of the heterojunction projected band structure, the Schottky barrier height is extracted at the atomic level with high precision, avoiding the problems of missing interface barrier parameters or reliance on empirical calibration in traditional TCAD models; 2) The calculation logic of the energy difference between the metal Fermi level and the semiconductor conduction band bottom / valence band top is clarified, so that the extracted barrier height can accurately reflect the energy level matching relationship of the metal-semiconductor interface and adapt to the decisive influence of Schottky barrier fluctuations on performance in short-channel devices; 3) Core parameters are provided for the accurate simulation of the source-drain contact characteristics of the device, solving the problem of prediction deviation of conduction current and subthreshold slope caused by inaccurate interface barrier simulation in the prior art.

[0080] In one exemplary embodiment, before optimizing the adjustable parameters of a pre-set ballistic transport model in TCAD software based on the modified material parameter library and incorporating transport IV curves, the construction method further includes: (1) Construct the source-channel-drain structure model corresponding to the target nanoscale device; the material property parameters of the source-channel-drain structure model are assigned by the key material parameters; (2) The quantum transport method based on the non-equilibrium Green's function is used to simulate the source-channel-drain structure model under different simulation combinations, and the current values ​​corresponding to each voltage are obtained under the same channel length; the simulation combination refers to the combination of different channel lengths and different voltages. Specifically, quantum transport methods based on non-equilibrium Green's functions include, but are not limited to, DFT-NEGF, TB-NEGF, KP-NEGF, and EM-NEGF methods.

[0081] (3) Associate the voltage and the corresponding current value under the same channel length to generate the IV curve corresponding to the same channel length; (4) Integrate the IV curves under different channel lengths into a set to obtain the transport IV curve.

[0082] In summary, simulations of the source-channel-drain structure model yielded IV curves for different channel lengths and voltages, as referenced. Figure 7 The reason for needing multiple IV curves is as follows: (1) Covering all operating scenarios of devices: The performance of nanoscale devices such as carrier transport and quantum tunneling varies with channel length and voltage. The IV curve of a single channel and a single voltage can only reflect the characteristics of a specific scenario and cannot support model optimization across scenarios. Multiple curves can cover short / medium / long channels (5nm, 10nm, 15nm) and low / high voltages (0.05V, 0.7V), ensuring that the optimized model has high accuracy in the entire operating range.

[0083] (2) Capturing quantum and size effects: In sub-10 nanometer devices, effects such as quantum confinement and short-channel tunneling are enhanced as the channel length shortens, and voltage changes affect the energy distribution and transmission probability of charge carriers. Multiple curves can comprehensively capture the variation patterns of these effects and avoid the model being only suitable for a single scenario and deviating in other scenarios.

[0084] (3) Improve model universality: The goal is to build a reusable simulation parameter library that can be adapted to device designs of different sizes and operating conditions. Multi-curve optimization can enable the adjustable parameters of the ballistic transport model to have the ability to adapt across sizes and voltages, meeting the needs of large-scale device design iteration.

[0085] The goal of parameter optimization is to find a unified set of adjustable parameters so that the optimized ballistic transport model minimizes the error with the corresponding NEGF transport IV curve in all scenarios (different channel lengths, different voltages), rather than adjusting the parameters individually for each curve.

[0086] Figure 7 This is a comparison chart of IV curves for nanoscale devices under different channel lengths and voltages. Key information is as follows: On the coordinate axis, the horizontal axis represents the gate voltage in volts (V), ranging from -0.2V to 0.8V, fully covering the typical operating range of the device, including the subthreshold region and conduction region; the vertical axis represents the current in amperes (A), using a logarithmic scale design, ranging from... arrive A can clearly show the current change of nanoscale devices from the off state to the on state, and is especially intuitive for displaying the subthreshold characteristics of low current.

[0087] Figure 7 The curves are grouped around channel length, simulation method, and voltage as core dimensions, including three channel lengths (5nm, 10nm, 15nm), two simulation methods (NEGF quantum transport method, DD drift-diffusion method), and two voltages (0.05V, 0.7V). The specific combinations of curves are labeled as follows: at a channel length of 15nm, the NEGF method curve at 0.7V is represented by a red square. The DD method curve is represented by a solid red line. The NEGF method curve at 0.05V is represented by a blue triangle. The DD method curve is represented by the solid blue line. At a channel length of 10 nm, the NEGF method curve at a voltage of 0.7 V is a yellow diamond shape. The DD method curve is represented by a solid yellow line. The NEGF method curve for a voltage of 0.05V is represented by a green triangle. The DD method curve is represented by a solid green line. At a channel length of 5nm, the NEGF method curve at 0.7V is a purple pentagram (). The curve for the DD method is a solid purple line. The NEGF method curve at 0.05V is a gray circle ( The DD method curve is a gray solid line ( ).

[0088] In terms of core data characteristics, under the same channel length and voltage conditions, the discrete labeled curve of the NEGF method deviates from the solid line curve of the DD method. This deviation is more pronounced in the subthreshold region at low gate voltages, intuitively reflecting the accuracy difference between the traditional drift-diffusion model and the quantum transport model. Under the same simulation method and voltage, the shorter the channel length (e.g., 5nm), the higher the on-state current value, consistent with the inherent characteristics of short-channel devices. Under the same channel length and simulation method, the higher the voltage (0.7V), the higher the on-state current value (0.05V), clearly demonstrating the regulatory effect of voltage on the device's conduction capability. These curves are a set of transport IV curves generated by "constructing a source-channel-drain model combined with NEGF quantum transport simulation," providing a high-precision benchmark for subsequent optimization of the TCAD model.

[0089] Therefore, based on the revised material parameter library, and combined with the adjustable parameters of the pre-set ballistic transport model in the TCAD software using the transport IV curve optimization, a parameter-optimized ballistic transport model is obtained, including: S210: Substitute the parameters from the modified material parameter library that are compatible with the drift-diffusion model into the drift-diffusion model to generate drift-diffusion IV curves for all simulation combinations; where the drift-diffusion model is a model pre-set in TCAD software; and the parameter models compatible with the drift-diffusion model include at least: mobility model, effective mass model and ballistic transport model, etc.

[0090] S220: For each simulation combination, calculate the single-combination error between the corresponding transport IV curve and the drift-diffusion IV curve, summarize the single-combination errors of all simulation combinations, and construct the global error; where the error is either absolute error or relative error; Understandably, each simulation combination refers to each simulation combination with "the same channel length and the same voltage".

[0091] S230: With the goal of minimizing the global error, the adjustable parameters of the ballistic transport model are iteratively adjusted until the error of each individual simulation combination is less than a preset threshold, thus obtaining the ballistic transport model with optimized parameters. The adjustable parameters include at least one of ballistic velocity, carrier relaxation time, and quantum confinement factor. The optimized parameters can be adapted to all simulation combinations simultaneously.

[0092] The beneficial effects of this embodiment are as follows: 1) The use of multiple NEGF quantum transport methods to generate transport IV curves provides a high-precision benchmark for ballistic transport model optimization, solving the problem of traditional TCAD models lacking quantum-level calibration benchmarks; 2) Based on error comparison and iterative optimization with the same channel length and voltage, the adjustable parameters of the ballistic transport model are strongly correlated with the key material parameters, avoiding blind adjustment of empirical parameters and improving the physical rationality of the model; 3) The simulation conditions covering different channel lengths and voltages are covered, enabling the optimized ballistic transport model to adapt to the simulation requirements of cross-size devices, making up for the lack of cross-size universality of traditional quantum correction models, and providing reliable support for batch device simulation.

[0093] In one specific implementation, see Figure 8 The following is an exemplary application description of the entire process of nanoscale device simulation modeling in this invention: Process 1: Material parameter extraction.

[0094] Based on density functional theory (DFT), the core parameters of nanomaterials are extracted in two categories: Material contact characteristics: Schottky barrier and contact resistance are calculated to describe the electrical characteristics of the source-drain-metal interface; Material physical properties: Extract band gap state density, dielectric constant, and material thickness to describe the basic physical properties of the material itself.

[0095] These parameters are the basic inputs for subsequent TCAD modeling.

[0096] Process 2: TCAD-level device modeling.

[0097] Import the above material parameters into TCAD software and complete two core steps: Fitting the TCAD model: Based on the material parameters, the default TCAD model is modified to construct a simulation model that adapts to the target nanodevice; IV curve acquisition using the NEGF method: The IV curves of the device under different conditions are obtained through simulation using the NEGF quantum transport method, ensuring the accuracy of the model.

[0098] Process 3: Circuit-level application.

[0099] Import the device model exported from TCAD into the SPICE tool to complete the circuit-level application: Compact Model Acquisition: Convert TCAD device models into compact models usable for circuit design; Unit circuit evaluation: The performance of unit circuits is simulated and evaluated based on compact models, achieving full chain coverage from "materials-devices" to "circuits".

[0100] The core value of this process is to ensure the physical accuracy of material parameters through density functional theory, balance simulation accuracy and efficiency by combining TCAD with NEGF, and finally realize the circuit-level application of the device through SPICE, forming a closed loop of "microscopic parameters → device model → circuit evaluation".

[0101] In summary, the embodiments of this invention calculate the band structure of the target nanomaterial using first-principles software, accurately extracting the band gap, effective mass, non-parabolic coefficient, and energy difference between valleys, providing physical-level fundamental parameters for device simulation; based on the charge density distribution calculated by DFT, the physical thickness of the nanomaterial is determined, ensuring a strong correlation between the parameters and the material's microstructure; a uniform electric field is applied to both ends of the material to simulate the internal potential distribution, and the relative permittivity at the nanoscale is extracted by combining the ratio of the vacuum permittivity to the electric field strength within the material; the carrier transmittance at the source / drain and metal contact interface is calculated using the DFT-NEGF method, and the contact resistance and contact barrier are obtained by combining Ohm's law, thus perfecting the interface characteristic parameters.

[0102] Based on the extracted material parameters, the default material model of TCAD is corrected. At least one quantum transport method among DFT-NEGF, TB-NEGF, KP-NEGF, and EM-NEGF is used to calculate the IV current curves under different channel lengths and voltages. Combined with the IV data of the traditional drift-diffusion model, the adjustable parameters of the TCAD ballistic transport model are iteratively fitted until the IV curve error of the two models is minimized. Finally, a simulation parameter library with both accuracy and universality is constructed.

[0103] This invention addresses the industry pain point of being unable to balance accuracy and efficiency in the modeling and simulation of nanoscale devices, especially those made of new materials. On the one hand, it compensates for the lack of key material parameters in the default model of traditional TCAD software, thus solving the problem of inaccurate simulation. On the other hand, it avoids the limitation of high accuracy but low efficiency in pure DFT-NEGF atomic-scale simulation, achieving a dual balance between accuracy and efficiency, and can support large-scale device design iteration.

[0104] The apparatus provided by the present invention will be described below. The simulation modeling apparatus for nanoscale devices described below and the simulation modeling method for nanoscale devices described above can be referred to in correspondence.

[0105] like Figure 9 As shown, embodiments of the present invention also provide a simulation modeling apparatus for nanoscale devices, used to implement the simulation modeling method for nanoscale devices in any of the above embodiments. This simulation modeling apparatus for nanoscale devices may include: Extraction module 910 is used to extract key material parameters from the basic electronic structure data of the target nanoscale device, and to correct the default material parameter library in TCAD software based on the key material parameters to obtain the corrected material parameter library; the basic electronic structure data is obtained based on density functional theory. The optimization module 920 is used to optimize the adjustable parameters of the ballistic transport model pre-set in the TCAD software based on the modified material parameter library and the transport IV curve, so as to obtain the parameter-optimized ballistic transport model; the transport IV curve is generated by the quantum transport method based on the non-equilibrium Green's function combined with key material parameters. Module 930 is used to build a simulation parameter library for target nanoscale devices based on key material parameters and the optimized ballistic transport model.

[0106] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings, the disclosure, and the appended claims in carrying out the claimed invention. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. Although the invention has been described in conjunction with specific features and embodiments, it is apparent that various modifications and combinations can be made thereto without departing from the spirit and scope of the invention. Accordingly, this specification and the accompanying drawings are merely exemplary descriptions of the invention as defined by the appended claims and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Obviously, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if such modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include such modifications and modifications.

Claims

1. A simulation modeling method for nanoscale devices, characterized in that, include: Key material parameters are extracted from the basic electronic structure data of the target nanoscale device, and the default material parameter library in the TCAD software is corrected based on the key material parameters to obtain the corrected material parameter library; the basic electronic structure data is obtained based on density functional theory. Based on the modified material parameter library, the adjustable parameters of the ballistic transport model pre-set in the TCAD software are optimized by combining the transport IV curve to obtain the parameter-optimized ballistic transport model; the transport IV curve is generated by combining the key material parameters with the quantum transport method based on the non-equilibrium Green's function. Based on the key material parameters and the ballistic transport model optimized by the parameters, a simulation parameter library for the target nanoscale device is constructed.

2. The simulation modeling method for nanoscale devices according to claim 1, characterized in that, Before optimizing the adjustable parameters of the ballistic transport model pre-set in the TCAD software based on the modified material parameter library and the transport IV curve, the construction method further includes: Construct a source-channel-drain structure model corresponding to the target nanoscale device; the material property parameters of the source-channel-drain structure model are assigned values ​​using the key material parameters; The quantum transport method based on the non-equilibrium Green's function was used to simulate the source-channel-drain structure model under different simulation combinations, and the current values ​​corresponding to each voltage were obtained under the same channel length; the simulation combination refers to the combination of different channel lengths and different voltages. By correlating the voltage and the corresponding current value under the same channel length, an IV curve corresponding to the same channel length is generated. The IV curves for different channel lengths are integrated into a set to obtain the transport IV curve.

3. The simulation modeling method for nanoscale devices according to claim 2, characterized in that, Based on the revised material parameter library, and combined with the transport IV curve optimization of the adjustable parameters of the pre-set ballistic transport model in the TCAD software, a parameter-optimized ballistic transport model is obtained, including: Substitute the parameters from the modified material parameter library that are compatible with the drift-diffusion model into the drift-diffusion model to generate drift-diffusion IV curves for all simulation combinations. For each simulation combination, calculate the single combination error between the corresponding transport IV curve and the drift-diffusion IV curve, summarize the single combination errors of all simulation combinations, and construct the global error. With the goal of minimizing the global error, the adjustable parameters of the ballistic transport model are iteratively adjusted until the single-combination error of all simulation combinations is less than a preset threshold, thus obtaining the ballistic transport model with optimized parameters; the adjustable parameters include at least one of ballistic velocity, carrier relaxation time, and quantum confinement factor.

4. The simulation modeling method for nanoscale devices according to claim 1, characterized in that, Before extracting key material parameters from the electronic structure fundamental data of the target nanoscale device, the method further includes: determining the electronic structure fundamental data based on the density functional theory; The fundamental data of the electronic structure are determined based on the density functional theory, including: Obtain the basic parameters of the target nanoscale device; the basic parameters include the chemical composition, crystal structure, heterojunction interface structure to be simulated, and applied external field conditions of the target nanoscale device; The basic parameters are input into first-principles software based on density functional theory to generate the band structure data, charge density distribution data, potential distribution data of the material under a uniform electric field, and projected band data of the metal-semiconductor heterojunction of the target nanoscale device. The band structure data, the charge density distribution data, the potential distribution data, and the projected band data are determined as the basic data of the electronic structure.

5. The simulation modeling method for nanoscale devices according to claim 4, characterized in that, The key material parameters include at least the effective mass of charge carriers, the non-parabolic coefficient, the band width, and the energy difference between energy valleys; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The band structure data is fitted to obtain a band structure curve; the band structure curve covers the band distribution information of the entire Brillouin zone. The band structure curve is fitted with a second-order polynomial near the extreme point. The second derivative is calculated based on the fitting result, and the effective mass of the charge carrier is calculated based on the second derivative. For the target region in the band structure curve that meets the preset conditions, a high-order fitting method is used to fit the band structure curve data of the target region, or the band structure curve data of the target region is fitted based on the Kane model, and the non-parabolic coefficients are obtained through fitting optimization; the preset conditions include that the electron energy in the band structure curve is outside the band extreme region and the parabolic band model fails. Based on the band structure curve covering the entire Brillouin zone band distribution information, the maximum and minimum energy values ​​of the target band are extracted respectively, and the difference between the maximum and minimum energy values ​​is determined as the band width; the target band includes a valence band and a conduction band, the valence band is centered on the band where the highest occupied band is located, and the conduction band is centered on the band where the lowest unoccupied band is located. In the band structure curve, the band extrema at each high symmetry point are identified. From the band extrema at each high symmetry point, the energy value corresponding to the global extremum point is determined. The difference between the extremum energy at each high symmetry point and the energy value corresponding to the global extremum point is calculated one by one to form the energy difference set between different energy valleys of the target nanoscale device. Among the band extrema at each high symmetry point, the conduction band corresponds to the band minimum, the valence band corresponds to the band maximum, and the extremum energy at each high symmetry point is the energy value corresponding to the band extrema at each high symmetry point.

6. The simulation modeling method for nanoscale devices according to claim 4, characterized in that, The key material parameters also include the relative permittivity; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The potential distribution data is fitted to obtain the potential distribution curve; Based on the potential distribution curve, the electric field strength within the material is calculated. Combined with the applied vacuum electric field strength, the formula is used: ； The relative permittivity of the target nanoscale device was calculated; wherein, The relative permittivity is denoted as . The vacuum electric field strength is mentioned. The electric field strength within the material is denoted as .

7. The simulation modeling method for nanoscale devices according to claim 4, characterized in that, The key material parameters also include the physical thickness of the device; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The charge density distribution data is fitted to obtain a charge density distribution curve; the charge density distribution curve is a charge density profile curve along the material thickness direction; The spatial range in the charge density distribution curve where the charge density value is greater than a preset charge density threshold is determined, and the size of the spatial range is used as the physical thickness of the target nanoscale device.

8. The simulation modeling method for nanoscale devices according to claim 4, characterized in that, The key material parameters also include the Schottky barrier height; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The projected band data of the metal-semiconductor heterojunction is fitted to obtain the projected band curve of the metal-semiconductor heterojunction. Based on the projected band structure of the metal-semiconductor heterojunction, the Fermi level of the metal electrode and the conduction band bottom energy or valence band top energy of the semiconductor channel material are determined, and the energy difference between the Fermi level of the metal electrode and the conduction band bottom energy or the valence band top energy is calculated. The energy difference is determined as the Schottky barrier height.

9. The simulation modeling method for nanoscale devices according to claim 8, characterized in that, The key material parameters also include contact resistance; Key material parameters are extracted from the fundamental electronic structure data of the target nanoscale device, including: The electronic structure data and device contact structure parameters are processed using the quantum transport method based on the non-equilibrium Green's function to obtain carrier transmittance data; wherein, the device contact structure parameters include the contact type between the source / drain and the metal, the atomic-level structure information of the contact interface, and the applied bias voltage; Substituting the carrier transmittance data into the quantum transport current formula based on the non-equilibrium Green's function theory: ； By performing integration, the current is obtained; where, For energy; Carrier transmittance; The current; For electron charge; It is Planck's constant; The Fermi distribution function of the left electrode; The Fermi distribution function of the right electrode; By substituting the applied bias voltage and the current into Ohm's law formula, the contact resistance between the source / drain of the target nanoscale device and the metal is calculated.

10. A simulation modeling device for nanoscale devices, characterized in that, include: An extraction module is used to extract key material parameters from the basic electronic structure data of the target nanoscale device, and to correct the default material parameter library in the TCAD software based on the key material parameters to obtain a corrected material parameter library; the basic electronic structure data is obtained based on density functional theory. The optimization module is used to optimize the adjustable parameters of the ballistic transport model pre-set in the TCAD software based on the modified material parameter library and combined with the transport IV curve, so as to obtain the parameter-optimized ballistic transport model; the transport IV curve is generated by combining the key material parameters with the quantum transport method based on the non-equilibrium Green's function. A construction module is used to build a simulation parameter library for the target nanoscale device based on the key material parameters and the ballistic transport model optimized by the parameters.

Images