Method for simulating thermal characteristics of semiconductor device under action of multiple physical fields

Abstract

The invention discloses a method for simulating thermal characteristics of a semiconductor device under the action of multiple physical fields, which relates to the field of semiconductor simulation, and comprises the following steps of: obtaining process structure parameters and material parameters of the semiconductor device, and constructing a two-bit cellular structure model in simulation software for analogue simulation to obtain a heat source; then, a three-dimensional multi-gate finger structure model of the semiconductor device is constructed according to the structural characteristics of the semiconductor device, a simulation object is expanded to a more complex three-dimensional multi-gate finger structure from a two-dimensional cellular structure, and a heat source obtained through simulation of the two-dimensional cellular structure is guided into a three-dimensional model for multi-physical field simulation; and carrying out thermodynamic analysis on a multi-physics field simulation result. The device can simulate the interaction of a plurality of physical fields such as a thermal field and a stress field, is used for multi-field coupling analysis in a complex structure and environment, and can more comprehensively analyze the thermal characteristics of a semiconductor device under the action of the plurality of physical fields. And a more accurate simulation basis is provided for optimizing the device design and improving the performance and reliability of the device in high-pressure, high-temperature and radiation environments.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor devices, and in particular to a method for simulating thermal characteristics of semiconductor devices under the action of multiple physical fields. Background Art

[0002] In modern power electronic equipment, power semiconductor devices play a vital role and are widely used in aerospace, communications, radar and other fields. The application environment of semiconductors is usually very complex. In actual work, devices are subject to the combined effects of multiple physical fields, including electric fields, temperature fields, radiation fields, etc. The coupling of these physical fields often leads to complex thermal distribution inside the device, which has become an important factor affecting its reliability and long-term stability. At present, the research on the thermodynamic core performance of semiconductor devices mainly relies on two means: experiments and simulations. The experimental method can provide an intuitive understanding of its thermodynamic performance, but the test cycle is long, and it is difficult to understand the changes in the thermal field of the device at the microscopic level. Therefore, simulation of semiconductor devices has become the main direction of research.

[0003] At present, many studies are mainly focused on the single physical field simulation of semiconductor devices, such as only considering the electric field or thermal field. However, the application environment of semiconductors is usually very complex. The simulation of a single physical field cannot fully reflect the complex thermal coupling phenomenon inside the device, and cannot fully analyze the thermal distribution of semiconductor devices under multi-physical field coupling conditions, especially in high pressure, high temperature and radiation environments. The complex thermal coupling phenomenon inside the device has not been accurately simulated.

[0004] Traditional simulation methods, such as using semiconductor simulation software TCAD to simulate and analyze semiconductor devices, can simulate the electrical characteristics of the device in detail, such as electric field distribution, carrier transport and electrothermal effect, etc. Researchers can accurately analyze the electrical behavior of semiconductor devices under different working conditions, especially the performance optimization under high voltage and radiation conditions. However, a significant limitation of TCAD is that it can only simulate the cellular structure of the device and cannot obtain the overall thermal distribution at the device level. In addition, TCAD has limited simulation capabilities in mechanics. In multi-field coupling simulation, it is impossible to simulate the distribution of stress and strain. For example, under the combined action of multiple conditions such as high temperature, high pressure or radiation, complex stress distribution may be generated inside the semiconductor. It is impossible to directly simulate these mechanical characteristics, and it is difficult to reflect the thermal coupling phenomenon of the device under the action of multiple physical fields. There is a lack of simulation of the overall force field distribution of the complex gate finger structure. These problems will lead to deviations in the evaluation of the long-term stability and reliability of the device. Summary of the invention

[0005] In order to solve the limitations of cellular model simulation under the action of multiple physical fields, the present invention provides a method for simulating the thermal characteristics of semiconductor devices under the action of multiple physical fields, which can more comprehensively analyze the thermal characteristics of semiconductor devices under the action of multiple physical fields, and provide a more accurate simulation basis for optimizing device design and improving its performance and reliability in high pressure, high temperature and radiation environments.

[0006] To achieve the above object, the technical solution proposed by the present invention is: a method for simulating thermal characteristics of semiconductor devices under the action of multiple physical fields, comprising the following steps: S1 builds a gridded two-dimensional structural model in semiconductor simulation software TCAD; S2 adds an electrical simulation model, builds an electrical model of the device, and applies a bias voltage to the model to obtain an electrical characteristic curve; The electrical characteristic curve of the S3 simulation model is compared with the electrical characteristic curve in the actual device product manual to optimize the model; S4 sets the collision ionization model, sets the incident particle information parameters to construct a single particle burnout effect simulation model, and performs simulation; S5 adds a lattice heating model and inputs physical parameters for thermal effect analysis; S6 constructs a three-dimensional multi-gate finger structure model of the semiconductor device in the finite element simulation software COMSOL according to the structural characteristics of the actual device, sets the corresponding material properties, and meshes the three-dimensional structure model; S7 sets boundary conditions in the heat conduction module and structural mechanics module of the finite element simulation software COMSOL, adds a thermal expansion model in the multi-physics field, and simulates the thermal distribution of the device; S8 performs thermodynamic analysis based on multiphysics simulation results.

[0007] Furthermore, the step S1 is specifically as follows: S11 obtains process structure parameters of actual semiconductor devices; S12 constructs a cellular structure model in TCAD based on the parameters of the actual semiconductor device; S13 divides each area into grids and constructs a gridded two-dimensional simulation structure model; S14 defines the process structure and material parameters of each area and matches them with the structure of the device.

[0008] Specifically, the electrical simulation model in step S2 includes a mobility model, a carrier statistics model and a composite model, and the electrical characteristic curve includes a transfer characteristic, an output characteristic and a breakdown characteristic.

[0009] Furthermore, in step S3, the electrical characteristic curve of the simulation model is compared with the electrical characteristic curve in the actual device product manual, and the comparison result is obtained according to the following steps: If the comparison result error is within 20%, the next simulation is performed; If the comparison result error exceeds 20%, the parameters of the two-dimensional simulation structure model in step S1 are optimized to make the electrical characteristic curve of the two-dimensional structure model consistent with the electrical characteristic curve of the actual device, and then proceed to the next step.

[0010] Specifically, the optimization includes calibrating key process parameters of the device substrate, epitaxial layer size, and doping concentration.

[0011] Specifically, in step S4, the incident particle information includes the incident position, the incident angle, the track width, the linear charge deposition (LCD), and the peak time and width of the charge generation pulse.

[0012] Specifically, in step S5, the physical parameters include lattice thermal conductivity and specific heat capacity; the thermal effect analysis is to generate data on the variation of thermal power density with time and position, and fit the data to a Gaussian distribution function.

[0013] Specifically, the Gaussian distribution function formula is: Where: Q(t) is the function of the thermal power density changing with time, x is the horizontal coordinate of the heat source position, x(t) is the function of the horizontal coordinate of the heat source position changing with time, and R is the radius of the heat source.

[0014] Specifically, in step S6, the three-dimensional model includes a gate, a drain, a source, a channel layer and a substrate; the material properties include density, thermal conductivity, specific heat capacity, Young's modulus and thermal expansion coefficient of GaN, AlGaN and Si materials.

[0015] Specifically, the specific contents of step S7 include: in the structural mechanics module, a fixed constraint is imposed on the bottom of the model to limit the displacement of the bottom of the device; in the heat conduction module, an initial temperature is set, and a Gaussian distribution function is input to simulate the thermal distribution of the device during a single particle burning process; in the thermal expansion multi-physics field, the stress distribution change caused by a single particle impact is simulated; a plurality of probes are set to observe the temperature and stress change process of each layer of material; in the transient simulation process, the time step is in the picosecond level, and the dynamic evolution process of temperature and stress during the single particle burning process is recorded by the transient solver.

[0016] Specifically, in step S8, the thermal effect analysis outputs the temperature field distribution diagram, stress field distribution diagram, local overheating areas and stress concentration areas in the device through multi-physical field simulation, and analyzes the thermal field distribution at the moment of burning, the stress distribution during the temperature rise process, and the temperature and stress changes of different material layers.

[0017] Beneficial effects of the present invention: 1. The scope of thermal simulation has been expanded. By combining the semiconductor simulation software TCAD and the finite element simulation software COMSOL, it can accurately simulate the thermal distribution of semiconductor devices under the joint action of multiple physical fields (electric field, thermal field, stress field, etc.), and expand the simulation scope from two-dimensional cellular structure to more complex three-dimensional multi-gate finger structure, thereby improving the comprehensiveness of the simulation.

[0018] 2. Improved stress field analysis. By introducing the finite element simulation software COMSOL, detailed analysis of the stress distribution of semiconductor devices under high temperature, high pressure and radiation conditions can be performed, solving the problem that traditional semiconductor simulation software TCAD tools cannot handle the stress field distribution inside the device. This improvement provides technical support for avoiding potential failure problems caused by thermal-mechanical coupling effects.

[0019] 3. It can accurately simulate transient thermal changes. In the process of transient thermal changes caused by single particle effects, it can accurately simulate the thermal distribution and stress changes of the device, and better show the transient overheating and stress concentration phenomena in the process of single particle burning. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. For those skilled in the art, other drawings can be obtained based on the drawings without paying creative work.

[0021] Figure 1 It is a flow chart of the method for simulating the thermodynamic characteristics of semiconductor devices under the action of multiple physical fields of the present invention; Figure 2 A schematic diagram of a GaN HEMT device unit structure according to an embodiment of the present invention; Figure 3 This is a simulation cell structure diagram of a GaN HEMT device according to an embodiment of the present invention; Figure 4 A comparison diagram of transfer characteristic curves of GaN HEMT devices according to an embodiment of the present invention; Figure 5 The electric field distribution of the TCAD electrothermal distribution two-dimensional cloud map according to the embodiment of the present invention; Figure 6 It is a two-dimensional cloud diagram of TCAD electric heat distribution - temperature field distribution in an embodiment of the present invention; Figure 7 The Gaussian heat source of the embodiment of the present invention; Figure 8 A gate finger structure simulation model according to an embodiment of the present invention; Fig. 9 The thermal field distribution and temperature field distribution at the burning moment of the embodiment of the present invention; Fig.10 The stress field distribution at the time of burning of the embodiment of the present invention; Fig.11 The temperature changes of different material layers in the embodiment of the present invention; Fig.12 FIG. 1 is a diagram showing stress changes in different material layers according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] like Figure 1 As shown, a specific embodiment of the method for simulating thermodynamic characteristics of semiconductor devices under the action of multiple physical fields of the present invention comprises the following specific steps: Step S1, constructing a gridded two-dimensional structure model in semiconductor simulation software TCAD: First, the process structure parameters of the actual semiconductor device are obtained; Then, a cellular structure model is constructed in TCAD according to the parameters of the actual semiconductor device; Then, each area is divided into grids to construct a gridded two-dimensional simulation structure model; Finally, the process structure and material parameters of each area are defined and matched with the structure of the device.

[0024] Step S2, adding an electrical simulation model to the constructed two-dimensional simulation mechanism model, constructing an electrical model of the device, applying a bias voltage to the model to obtain an electrical characteristic curve, wherein the electrical simulation model includes a mobility model, a carrier statistical model and a recombination model, and the electrical characteristic curve includes transfer characteristics, output characteristics and breakdown characteristics.

[0025] Step S3, compare the electrical characteristic curve of the simulation model with the electrical characteristic curve in the actual device product manual to optimize the model, and proceed according to the following steps: compare the electrical characteristic curve of the simulation model with the electrical characteristic curve in the actual device product manual. If the comparison result error is within 20%, proceed to the next simulation; if the comparison result error exceeds 20%, optimize the parameters of the two-dimensional simulation structure model in step S1, including optimizing and calibrating the substrate, epitaxial layer size, doping concentration key process parameters of the device, so that the electrical characteristic curve of the two-dimensional structure model is consistent with the electrical characteristic curve of the actual device, and then proceed to the next step.

[0026] Step S4, setting the collision ionization model and setting the incident particle information parameters to construct a single particle burnout effect simulation model for simulation; wherein the incident particle information includes the incident position, incident angle, track width, linear charge deposition (LCD), peak time and width of the charge generation pulse.

[0027] Step S5, adding a lattice heating model to the two-dimensional structure model, inputting physical parameters for thermal effect analysis, the physical parameters including lattice thermal conductivity and specific heat capacity; the thermal effect analysis is to generate data on the variation of thermal power density with time and position, and fitting the data to a Gaussian distribution function to represent the Gaussian heat source distribution, the formula of which is: Where: Q(t) is the function of the thermal power density changing with time, x is the horizontal coordinate of the heat source position, x(t) is the function of the horizontal coordinate of the heat source position changing with time, and R is the radius of the heat source.

[0028] Step S6, constructing a three-dimensional model of a semiconductor device with a multi-gate finger structure in finite element simulation software COMSOL according to the structural characteristics of the actual device, setting corresponding material properties, and meshing the three-dimensional model.

[0029] Among them, the three-dimensional model includes the gate, drain, source, channel layer and substrate; the material properties include the density, thermal conductivity, specific heat capacity, Young's modulus and thermal expansion coefficient of GaN, AlGaN and Si materials.

[0030] Step S7, setting boundary conditions in the heat conduction module and structural mechanics module of the finite element simulation software COMSOL, adding a thermal expansion model in the multi-physics field, and simulating the thermal distribution of the device, specifically: in the structural mechanics module, a fixed constraint is imposed on the bottom of the model to limit the displacement of the bottom of the device; in the heat conduction module, the initial temperature is set, and a Gaussian distribution function is input to simulate the thermal distribution of the device during the single-particle burning process; in the thermal expansion multi-physics field, the stress distribution change caused by the single-particle impact is simulated; by setting multiple probes, the temperature and stress change process of each layer of material is observed; in the transient simulation process, the time step is in the picosecond level, and the transient solver is used to record the dynamic evolution of temperature and stress during the single-particle burning process.

[0031] Step S8, performing thermodynamic analysis based on the multi-physics field simulation results, specifically: thermal effect analysis through the temperature field distribution diagram, stress field distribution diagram, local overheating area and stress concentration area in the device output by the multi-physics field simulation, analyzing the thermal field distribution at the moment of burning, the stress distribution during the temperature rise process, and the temperature and stress changes of different material layers.

[0032] The present invention provides a specific simulation embodiment to introduce the simulation process in detail.

[0033] The semiconductor device of this embodiment is a p-GaN gate structure GaN-based HEMT device, such as Figure 2 The p-GaN gate structure GaN-based HEMT device unit structure shown in the figure has a GaN HEMT device process structure of a lateral planar structure, in which the gate, source, and drain are located on the same plane.

[0034] like Figure 3 The cell structure used for GaN HEMT device simulation shown has a lateral width of 11 μm, a thickness of p-GaN of 0.07 μm, a thickness of the AlGaN layer of 15 nm, a thickness of the channel layer of 0.2 μm, and a thickness of the buffer layer of 2 μm.

[0035] According to the design and process parameters of the GaN HEMT device mentioned above, the semiconductor device two-dimensional modeling tool (semiconductor simulation software TCAD) is used to perform two-dimensional modeling of the GaN HEMT device to be simulated, and a two-dimensional model of the device to be simulated is obtained. Then, the established two-dimensional model is meshed to generate a meshed device structure, define the process structure and material parameters of each area in the two-dimensional model, and match the mesh structure with the device structure to effectively ensure the simulation accuracy and calculation efficiency.

[0036] An electrical simulation model is added to the constructed two-dimensional structure model of the GaN HEMT device, including a mobility model, a carrier statistics model, a recombination model, etc., and a suitable bias voltage is applied to the device structure model for TCAD device electrical simulation. Appropriate voltages are applied to the source and drain of the device model respectively, so as to obtain the typical electrical parameters or curves of the device, such as the transfer characteristic curve and the output characteristic curve. The device electrical simulation results are then compared with the electrical parameters or curves measured in the experiment. If the deviation is within 20%, it is considered to be consistent, and the next simulation can be carried out directly; if the deviation exceeds 20%, it is considered to be inconsistent, and the various parameters defined when constructing the two-dimensional structure model of the device are optimized, and the key process parameters such as the substrate, epitaxial layer size, and concentration of the device are optimized and calibrated. Finally, the electrical simulation results are consistent with those measured in the experiment before the next simulation is carried out.

[0037] like Figure 4 The comparison between the simulated value and the actual measured value of the transfer characteristic curve of the GaN HEMT device shown in the figure shows that the electrical simulation curve is basically consistent with the measured curve, with a deviation of less than 20%, which ensures the accuracy of the model and the reliability of subsequent simulation results.

[0038] The optimized electrical simulation model of GaN HEMT device is used to set a collision ionization model. The information parameters of the incident particles are defined by assembly language, including the incident position, angle, track width, linear charge deposition (LCD), peak time and width of the charge generation pulse, and a single event burnout (SEB) model is established for the GaN HEMT device. The bias voltage V GS Set to 0V, drain-source voltage V DS The voltage is set to 220V, the heavy ion LCD value is 0.8pC / μm, the operating temperature is set to 300K, and the single-particle burnout simulation is performed by vertically incident through the entire device from the most sensitive part of the single-particle burnout effect (gate-drain spacing gate 0.5μm). like Figure 5 and Figure 6 As shown, the distribution of electric field and temperature field at different times is obtained through simulation, showing the electric and thermal distribution cloud map when single particle burning occurs.

[0039] The lattice heating model is introduced in the single particle burnout effect simulation. This model is used to simulate the thermal effect on the lattice of the device material when the particle is incident. By inputting appropriate physical parameters, including lattice thermal conductivity, specific heat capacity, etc., the heat conduction process inside the device when the single particle burnout effect occurs can be more realistically described; like Figure 7 As shown in the figure, after the simulation of the single particle burnout effect, the data of thermal power density changing with time and position are output from TCAD, and the data are fitted into a Gaussian distribution function to represent the Gaussian heat source distribution. The formula is: Where: Q(t) is the function of the thermal power density changing with time, x is the horizontal coordinate of the heat source position, x(t) is the function of the horizontal coordinate of the heat source position changing with time, and R is the radius of the heat source.

[0040] The Gaussian distribution function acts as a heat source load in transient simulations to more accurately reproduce the heat diffusion behavior after particle impact.

[0041] like Figure 8 As shown, a three-dimensional model of a GaN HEMT device with a multi-gate finger structure is constructed in the finite element simulation software COMSOL. The model is composed of five cells crossed by an interpolation structure. The size of the entire model is 53μm×260μm×3.415μm. The multi-gate structure can better reflect the heat conduction path of the device when it is locally impacted by particles, which is helpful to analyze the local burning characteristics and stress changes of the GaN HEMT device when the single particle burning effect occurs.

[0042] In this embodiment, the channel layer and the buffer layer are combined in the modeling process. Both layers are made of GaN material and have similar physical properties. Material parameters are set for each component of the GaN HEMT device, including density, specific heat capacity, thermal conductivity, thermal expansion coefficient, elastic modulus (Young's modulus), etc. The established three-dimensional model is meshed, and the AlGaN barrier layer, GaN layer, and SiC substrate layer of the model are divided into hexahedral meshes by sweeping, with a maximum unit size of 0.15 μm. The remaining parts use free tetrahedral meshes, and the average unit mass of the mesh is 0.8138, which achieves a high quality and meets the simulation accuracy requirements.

[0043] Boundary conditions are added to the heat conduction module and the structural mechanics module to perform single-particle burnout thermal simulation. Specifically, in the heat conduction module, the initial temperature is set to 300K, corresponding to the starting temperature in TCAD, and the Gaussian distribution function obtained by analysis is applied to the middle cell as a heat source load between the gate and the drain to simulate the heat distribution of the device during the single-particle burnout effect. In the structural mechanics module, a fixed constraint is applied to the bottom of the model to limit the displacement of the bottom of the device to ensure that there is no free expansion or structural deformation of the bottom during the heat conduction process, so as to accurately simulate the thermal stress distribution under the single-particle burnout effect. At the same time, a thermal expansion multi-physics field is added to the three-dimensional model to simulate the stress distribution change caused by a single particle impact to ensure accurate coupling between the multi-physics fields. Multiple probes are added to observe the temperature and stress change process of each layer of material. In the transient simulation process, a picosecond time step is used to record the dynamic evolution of temperature and stress during the single-particle burnout process through the transient solver, which can capture the heat diffusion path and the formation of stress concentration areas to ensure that the simulation data fully reflects the transient response characteristics of the device.

[0044] After the multi-physics field simulation is completed, the temperature field and stress field distribution diagrams of the GaN HEMT device are output according to the multi-physics field simulation results, especially the local overheating areas and stress concentration areas in the device, and the thermal field distribution at the moment of burning is analyzed, including the temperature field distribution and the stress field distribution; and the temperature and stress changes of different material layers are analyzed, including temperature changes and stress changes.

[0045] like Fig. 9 The temperature field distribution at the moment of burning is shown. A transient high-temperature area is generated at the single particle hitting point. Initially, the temperature is concentrated near the gate, and then the high-temperature area moves toward the drain, and the temperature continues to rise.

[0046] like Fig.10 The stress field distribution at the burning moment is shown in the figure. During the gradual increase in temperature, due to the different thermal expansion coefficients of various materials, stress first appears near the channel, and then the stress is concentrated in the high temperature area until it burns out.

[0047] like Fig.11 The temperature variation curves of different material layers are shown. Fig.12 As shown in the stress variation curves of different material layers, when single particle burnout occurs, GaN and AlGaN have the highest temperatures, followed by the drain metal electrode, and the Si substrate has the lowest temperature; the maximum stress occurs in AlGaN, followed by the metal electrode, and then GaN.

[0048] This embodiment combines the semiconductor simulation software TCAD and the finite element simulation software COMSOL to construct a GaN HEMT device simulation model with a multi-gate finger structure, and uses the finite element simulation software COMSOL to simulate the interaction of multiple physical fields such as thermal fields and stress fields, which is used for multi-field coupling analysis under complex structures and environments. By importing the results of electrothermal generation (Joule heat) obtained by the semiconductor simulation software TCAD simulation into the finite element simulation software COMSOL three-dimensional model, the simulation object is expanded from a two-dimensional cell structure to a more complex three-dimensional multi-gate finger structure, and the thermal characteristics of the GaN HEMT device under the action of multiple physical fields can be more comprehensively analyzed, and the local hot spots and stress concentration areas inside the GaN HEMT device can be better captured, providing a more accurate simulation basis for optimizing the device design and improving its performance and reliability under high pressure, high temperature and radiation environments.

[0049] Although the preferred implementation cases of the present invention have been described above in conjunction with the accompanying drawings, the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms of specific changes without departing from the scope of protection of the invention spirit and claims, which all fall within the scope of protection of the present invention.

Claims

1. A method for simulating thermal characteristics of semiconductor devices under the action of multiple physical fields, characterized in that: The following steps are involved: S1 builds a gridded two-dimensional structural model in semiconductor simulation software TCAD; S2 adds an electrical simulation model, builds an electrical model of the device, and applies a bias voltage to the model to obtain an electrical characteristic curve; The electrical characteristic curve of the S3 simulation model is compared with the electrical characteristic curve in the actual device product manual to optimize the model; S4 sets the collision ionization model, sets the incident particle information parameters to construct a single particle burnout effect simulation model, and performs simulation; S5 adds a lattice heating model and inputs physical parameters for thermal effect analysis; S6 constructs a three-dimensional multi-gate finger structure model of the semiconductor device in the finite element simulation software COMSOL according to the structural characteristics of the actual device, sets the corresponding material properties, and meshes the three-dimensional structure model; S7 sets boundary conditions in the heat conduction module and structural mechanics module of the finite element simulation software COMSOL, adds a thermal expansion model in the multi-physics field, and simulates the thermal distribution of the device; S8 performs thermodynamic analysis based on multiphysics simulation results.

2. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: The step S1 is specifically as follows: S11 obtains process structure parameters of actual semiconductor devices; S12 constructs a cellular structure model in TCAD based on the parameters of the actual semiconductor device; S13 divides each area into grids and constructs a gridded two-dimensional simulation structure model; S14 defines the process structure and material parameters of each area and matches them with the structure of the device.

3. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: The electrical simulation model in step S2 includes a mobility model, a carrier statistics model and a composite model, and the electrical characteristic curve includes a transfer characteristic, an output characteristic and a breakdown characteristic.

4. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: In step S3, the electrical characteristic curve of the simulation model is compared with the electrical characteristic curve in the actual device product manual, and the comparison result is obtained according to the following steps: If the comparison result error is within 20%, the next simulation is performed; If the comparison result error exceeds 20%, the parameters of the two-dimensional simulation structure model in step S1 are optimized to make the electrical characteristic curve of the two-dimensional structure model consistent with the electrical characteristic curve of the actual device, and then proceed to the next step.

5. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 4, characterized in that: The optimization includes calibrating the key process parameters of the device substrate, epitaxial layer size, and doping concentration.

6. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: In step S4, the incident particle information includes the incident position, the incident angle, the track width, the linear charge deposition (LCD), the peak time and the width of the charge generation pulse.

7. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: In the step S5, the physical parameters include lattice thermal conductivity and specific heat capacity; the thermal effect analysis is to generate data on the variation of thermal power density with time and position, and fit the data to a Gaussian distribution function.

8. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 7, characterized in that: The Gaussian distribution function formula is: Where: Q(t) is the function of the thermal power density changing with time, x is the horizontal coordinate of the heat source position, x(t) is the function of the horizontal coordinate of the heat source position changing with time, and R is the radius of the heat source.

9. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: In step S6, the three-dimensional model includes a gate, a drain, a source, a channel layer and a substrate; the material properties include density, thermal conductivity, specific heat capacity, Young's modulus and thermal expansion coefficient of GaN, AlGaN and Si materials.

10. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: The specific contents of step S7 include: in the structural mechanics module, a fixed constraint is imposed on the bottom of the model to limit the displacement of the bottom of the device; in the heat conduction module, an initial temperature is set, and a Gaussian distribution function is input to simulate the thermal distribution of the device during a single particle burnout process; in the thermal expansion multi-physics field, the stress distribution change caused by a single particle impact is simulated; a plurality of probes are set to observe the temperature and stress change process of each layer of material; in the transient simulation process, the time step is in the picosecond level, and the dynamic evolution process of temperature and stress during the single particle burnout process is recorded by a transient solver.

11. The method for simulating thermal characteristics of semiconductor devices under multi-physical fields according to claim 1, characterized in that: In step S8, the thermal effect analysis is performed by outputting the temperature field distribution diagram, stress field distribution diagram, local overheating areas and stress concentration areas in the device through multi-physical field simulation, analyzing the thermal field distribution at the moment of burning, the stress distribution during the temperature rise process, and the temperature and stress changes of different material layers.

Images