GaN HEMT single event effect transient simulation method and system based on physical model

Abstract

The invention discloses a GaN HEMT single event effect transient simulation method and system based on a physical model, and the method comprises the steps: building a GaN HEMT single event effect physical model, carrying out the hexahedron subdivision of the whole physical model, and obtaining model structure information including the unit number and node physical coordinates of a hexahedron; based on the model structure information, single event effect model equation sets including a drift-diffusion equation set, a Poisson equation and an irradiation carrier generation rate are normalized, and a backward Euler method is adopted to carry out time discretization on the single event effect model equation sets; deriving a discrete single event effect model equation set through a time domain spectral element method, adding a boundary condition of hot electron emission in a current continuity equation, and introducing a carrier generation rate item; deriving a single event effect model equation set through a Galerkin test method, carrying out Newton iteration, and coupling the equation set into a large matrix form; and solving and calculating the large matrix. According to the invention, the anti-single event effect research efficiency of the GaN HEMT device can be improved, and a large amount of research cost can be saved.

Description

Technical Field

[0001] The present invention relates to a physical model study of third-generation semiconductor GaN, and in particular to a single-particle effect transient simulation method and system for GaN HEMT based on the physical model. Background Art

[0002] With the continuous development of aerospace technologies such as space satellites and deep space exploration, the demand for power systems that are resistant to high temperatures, high power, miniaturized, and adaptable to extreme radiation environments is becoming increasingly apparent. Gallium nitride (GaN)-based power devices, as a core representative of wide-bandgap semiconductor technology, offer advantages over traditional silicon-based devices, such as high breakdown voltage, low on-resistance, and high-temperature resistance, meeting the application requirements of next-generation spacecraft power systems. Due to their use of heterojunction layers instead of gate oxides, wide-bandgap devices like GaN inherently offer excellent immunity to total ionizing dose (TID) effects. However, they are also highly sensitive to single-event effects (SEEs) caused by high-energy particles in space. Currently, SEEs have become a research hotspot in the field of space device radiation. SEEs are radiation effects in which a single ionizing particle in the space environment passes through a sensitive region of a semiconductor device, causing abnormal changes in the device. These SEEs can potentially cause the entire device system to fail. As device design continues to move towards higher integration density and smaller device sizes, electronic devices operating in space radiation environments are more susceptible to SEEs.

[0003] The development of radiation-hardened devices involves a continuous "design-manufacture-test" process, which is extremely costly, especially in terms of time. Experimentation is far more expensive than computer-assisted calculations. Therefore, rapid numerical evaluation of GaN HEMT device SEE resistance is necessary. To provide a theoretical basis for SEE design of GaN HEMT devices, currently used equivalent circuit models require actual measurement data and are unable to describe the changes in physical quantities and their distribution within semiconductor devices. Therefore, research on SEE resistance of GaN HEMT devices based on SEE physical models is of great significance. Summary of the Invention

[0004] The present invention aims to provide a method and system for simulating the single event effect transient of GaN HEMT based on a physical model, which can improve the research efficiency of GaN HEMT devices against single event effects and save a lot of research costs.

[0005] The above-mentioned object of the present invention is achieved through the following technical solutions:

[0006] A single event effect transient simulation method for GaN HEMT based on a physical model, comprising:

[0007] Step 1: Establish a single-event effect physical model of the GaN HEMT and perform hexahedral meshing on the entire physical model to obtain model structure information including hexahedral unit numbers and node physical coordinates;

[0008] Step 2: Based on the model structure information, the SEP model equations, including the drift-diffusion equations, Poisson's equation, and irradiation carrier generation rate, are normalized and time-discretized using the backward Euler method.

[0009] Step 3: Derive the discretized single-event effect model equations using the time-domain spectral element method, add the boundary conditions for thermal electron emission to the current continuity equation, and introduce the carrier generation rate term;

[0010] Step 4: Derive the single-particle effect model equations using the Galerkin test method and perform Newton iteration to couple the equations into a large matrix.

[0011] Step 5: Solve and calculate the large matrix to obtain the electron quasi-Fermi potential, hole quasi-Fermi potential and electric potential that meet the convergence and stability conditions, and obtain the electron concentration, hole concentration, electron current, hole current and displacement current of each node. Finally, denormalize and add the electron current, hole current and displacement current to obtain the transient drain current under single-particle radiation.

[0012] Furthermore, the single event effect model equations after time discretization in step 2 are:

[0013]

[0014] Where n and p represent the electron concentration and hole concentration respectively, n m-1 and p m-1 represent the electron concentration and hole concentration at the previous moment, respectively. and are the electron current density and hole current density, τ n and τ p denote the electron and hole lifetimes, α n and α p are the ionization coefficients of electrons and holes, G n and G p represent the radiation electron generation rate and hole generation rate respectively, represents the electric potential, Γ represents the net doping concentration, is the density operator, and q represents the unit charge.

[0015] Further, the electron concentration and hole concentration are:

[0016]

[0017] Where, φ n and φ p represent the quasi-Fermi potential of electrons and holes respectively, and M represents the ratio of the intrinsic carrier concentration of different materials to that of AlGaN.

[0018] Furthermore, the carrier generation rate term in step 3 is:

[0019] G n (G p )=G LET (l)R(w,l)T(t)

[0020]

[0021] In the formula, T(t) represents the time function term, R(w,l) represents the spatial function distribution, G LET (l) represents the linear energy generation density, t0 represents the single particle impact time, s hi represents the eigenvalue of the Gaussian function, erf represents the Gaussian error function, w represents the vertical distance from the internal point of the single particle incident area to the center of the incident trajectory, and w t (l) represents the radius of the single particle incident area, LET_f(l) represents the energy deposited per unit path length, and l represents the heavy ion incident depth.

[0022] Furthermore, the step three also includes: replacing the two-dimensional electron gas of the HEMT device with the surface charge density in the Poisson equation and imposing a potential continuity condition.

[0023] Furthermore, in step 4, the single-particle effect model equations are derived by the Galerkin test method to obtain the equation:

[0024]

[0025] Where:

[0026]

[0027]

[0028] Where N i Represents the coefficient of the shape function, V n and V p represents the surface recombination velocity of electrons and holes, V n1 、V p1 and V n2 、V p2They represent the surface recombination velocities of electrons and holes in the materials on both sides of the heterojunction, respectively; X1 and χ1 represent the electron affinities of the materials on both sides of the heterojunction; n0 and p0 represent the electron concentration and hole concentration in the quasi-equilibrium state; μ n and μ p represents the electron and hole mobility, G 雪崩 represents the electron generation term caused by avalanche current, E g1 and E g2 represents the band gap of the materials on both sides of the heterojunction, ρ s represents the surface charge density, ε2 represents the dielectric constant of AlGaN, C represents the ratio of the dielectric constants of GaN to AlGaN, WXF represents the weight factor, and represents the potential at GaN and AlGaN respectively, S represents the contact surface area at the heterojunction, V represents the volume of the HEMT device structure layer, A takes the value of 1 on the gate Schottky contact surface and takes the value of 0 on the ohmic contact surface, B takes the value of 1 at the continuous quasi-Fermi level at the AlGaN / GaN heterojunction surface and takes the value of 0 at other heterojunctions.

[0029] Furthermore, in step five, PARDISO is used to solve the large matrix. The matrix is first sorted by METIS and then stored in CSR format for solution.

[0030] Furthermore, the electron concentration, hole concentration, electron current, hole current and displacement current of each node in step 5 are:

[0031]

[0032]

[0033] Where k is the Boltzmann constant, T is the ambient temperature, and N i and N j Represent the shape functions of the expanded basis function and the test basis function, I ni , I pi and represent electron current, hole current and displacement current respectively.

[0034] A single event effect transient simulation system for implementing the method, comprising:

[0035] The single-event effect physical model decomposition unit is used to establish the single-event effect physical model of GaNHEMT, perform hexahedral decomposition on the entire physical model, and obtain the model structure information including the hexahedral unit number and node physical coordinates;

[0036] The SEP model equations discretization unit normalizes the SEP model equations including the drift-diffusion equations, Poisson's equation, and irradiation carrier generation rate based on the model structure information, and uses the backward Euler method to time discretize the SEP model equations.

[0037] The SEP model equation optimization unit uses the time-domain spectral element method to derive the discretized SEP model equations, adds thermal electron emission boundary conditions to the current continuity equation, and introduces the carrier generation rate term;

[0038] The single-particle effect model equation derivation unit derives the single-particle effect model equations through the Galerkin test method and performs Newton iteration to couple the equations into a large matrix form;

[0039] The solving unit solves and calculates the large matrix to obtain the electron quasi-Fermi potential, hole quasi-Fermi potential and electric potential that meet the convergence and stability conditions, and obtains the electron concentration, hole concentration, electron current, hole current and displacement current of each node. Finally, the electron current, hole current and displacement current are denormalized and added to obtain the transient drain current under single particle radiation.

[0040] Compared with the prior art, the present invention has the following significant advantages:

[0041] (1) The method of the present invention can simulate the transient process of single-particle effects in the third-generation semiconductor GaN-based physical model;

[0042] (2) The method of the present invention adopts the SETD method and the global conformal hexahedron decomposition, which can reduce the corresponding unknown quantities while ensuring the block diagonal characteristics of the mass matrix, speed up the calculation, reduce the calculation memory, and save the calculation cost.

[0043] (3) The method of the present invention can simulate a variety of charge generation and distribution models and adapt them according to different radiation scenarios, making the simulation results more accurate and closer to the actual results.

[0044] (4) The method of the present invention can accelerate the solution speed by continuously updating the solver and calculate the single particle effect physical model that is difficult to converge due to the coupling of multiple physical fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a schematic diagram of the physical model of GaN HEMT. Figure 1 (a) is a structural diagram. Figure 1 (b) is the doping distribution diagram.

[0046] Figure 2COMSOL and SETD simulation diagrams for electron injection, hole injection, and electron-hole pair injection. Figure 2 (a) is the change of drain current over time. Figure 2 (b) is a graph showing the change of relative error over time.

[0047] Figure 3 is the potential distribution diagram when carriers are injected, Figure 3 (a) is the COMSOL potential distribution diagram when carriers are injected. Figure 3 (b) is the SETD potential distribution diagram during carrier injection.

[0048] Figure 4 is the electron concentration distribution diagram during carrier injection, Figure 4 (a) is the COMSOL electron concentration distribution diagram, Figure 4 (b) is the SETD electron concentration distribution diagram.

[0049] Figure 5 is the hole concentration distribution diagram, Figure 5 (a) is the COMSOL hole concentration distribution diagram, Figure 5 Middle (b) is the SETD hole concentration distribution diagram.

[0050] Figure 6 This is a schematic diagram of the relative errors of the potential, hole concentration, and electron concentration at X = 0.55 um during carrier injection. DETAILED DESCRIPTION

[0051] The present invention proposes a method for simulating the transient state of single-particle effects of GaN HEMT based on the time-domain spectral element method. Figure 1 Taking the GaN HEMT shown in FIG. 1 as an example, the specific steps of the present invention are further described in detail.

[0052] See also Figure 1 The following is a schematic diagram of the GaN HEMT physical model structure. The model's geometric dimensions are as follows: the SiC substrate in the GaN HEMT is 0.8μm long, 1μm wide, and 0.2μm high; the GaN channel layer is 0.8μm long, 1μm wide, and 0.2μm high; and the GaN barrier layer is 0.8μm long, 1μm wide, and 15nm high. The specific steps are as follows:

[0053] Step 1: Establish an accurate single-event effect physical model of the GaN HEMT. Use Ansys software to perform hexahedral meshing on the entire model to obtain model structure information including hexahedral unit numbers and node physical coordinates.

[0054] Step 2: Normalize the single-event effect model equations, including the drift-diffusion equations, Poisson's equation, and irradiation carrier generation rate, and use the backward Euler method for time discretization;

[0055] Step 3: Derive the discretized single-event effect model equations using SETD. In the current continuity equation, it is necessary to add boundary conditions for thermal electron emission to ensure the continuity of the current density and introduce a carrier generation rate term. In the Poisson equation, the surface charge density is used to replace the two-dimensional electron gas of the HEMT device and a potential continuity condition is imposed.

[0056] Step 4: Derive the single particle effect model equations through the Galerkin test method, and perform Newton iteration to couple the equations into a large matrix form.

[0057] Step 5: Calculate the large matrix through the PARDISO solver to obtain the electron quasi-Fermi potential, hole quasi-Fermi potential and electric potential that meet the convergence and stability conditions, obtain the electric potential, electron concentration and hole concentration of each node, and finally denormalize to obtain the transient drain current under single-particle radiation.

[0058] In the step 1, the physical model is modeled and segmented using Ansys software. Due to the multi-layer structure of the HEMT device, the model needs to be cut into three layers: the barrier layer, the channel layer, and the substrate, and the material numbers 2, 1, and 3 are attached respectively.

[0059] In the second step, the normalized physical model equations are time discretized using the backward Euler method. The discretized equations are as follows:

[0060]

[0061] Where n and p represent the electron concentration and hole concentration respectively, n m-1 and p m-1 represent the electron concentration and hole concentration at the previous moment, respectively. and are the electron current density and hole current density, τ n and τ p denote the electron and hole lifetimes, α n and α p are the ionization coefficients of electrons and holes, G n and G p denote the radiation electron generation rate and hole generation rate, respectively, n and φ p denote the electron and hole quasi-Fermi potentials, respectively. represents the electric potential, Γ represents the net doping concentration, and q represents the unit charge.

[0062] The relationship between the electron and hole concentrations and the electron and hole quasi-Fermi potentials is shown below:

[0063]

[0064] In step 3, the derivation of the carrier generation term is as follows:

[0065] G n (G p )=G LET (l)R(w,l)T(t)(5)

[0066]

[0067] Where t0 represents the time of single particle impact, s hi represents the eigenvalue of the Gaussian function, erf represents the Gaussian error function, w represents the vertical distance from the internal point of the single particle incident area to the center of the incident trajectory, and w t (l) represents the radius of the single particle incident area, and LET_f(l) represents the linear energy transfer density.

[0068] By changing the values of the above variables, any heavy ion incident situation in space can be completed, greatly reducing the cost of single-particle effect research.

[0069] In step 4, the SETD method is used as a solution platform to perform Galerkin tests on the drift-diffusion equation and Poisson equation of the single-particle effect model equations. First, the continuity equation of the electron current at the GaN material is derived. The results are as follows:

[0070]

[0071] Using vector identities and Transforming formula (8), we get:

[0072]

[0073] The contacts at the electrodes of HEMT devices are divided into ohmic contacts and Schottky contacts. A takes the value of 1 on the gate Schottky contact surface and 0 on the ohmic contact surface. The boundary conditions of the heterojunction contact surface are divided into thermal electron emission boundary conditions and continuous quasi-Fermi level boundary conditions. B is a continuous quasi-Fermi level at the AlGaN / GaN heterojunction surface and takes the value of 1, and takes the value of 0 at other heterojunctions.

[0074] The boundary conditions for the gate Schottky contact are:

[0075]

[0076] The boundary conditions for thermal electron emission are:

[0077]

[0078] Where, E c1 and E c2 Represent the conduction bands of the materials on both sides, E v1 and E v2 Represent the valence bands of the materials on both sides respectively.

[0079] Substituting equations (10) and (11) into equation (9), we can obtain:

[0080]

[0081] Similarly, the hole continuity equation can be obtained as follows:

[0082]

[0083] After normalizing the Poisson equation and performing basis function testing, we can get

[0084]

[0085] Using vector identities and Transforming formula (14) yields:

[0086]

[0087] Since there are multiple materials in a HEMT device, it is necessary to impose a potential continuity condition at the material interface:

[0088]

[0089] Use surface charge equivalent to replace the two-dimensional electron gas generated by the polarization effect at the heterojunction:

[0090]

[0091] Because of the vector transformation relationship and Derivation (18) yields:

[0092]

[0093] And because there is Derivation (19) yields:

[0094]

[0095] Substituting equations (17) and (19) into equation (16), we can obtain:

[0096]

[0097] Since the single-particle effect model equations are highly nonlinear, the Newton iteration method is used to process the equations. After Taylor expansion, the higher-order terms are ignored. The final Newton iteration format of the single-particle effect model equations is as follows:

[0098]

[0099]

[0100] in:

[0101]

[0102]

[0103] In the fourth step, the drift-diffusion equations and the Poisson equation coupled model equations of the single particle effect model equations are solved:

[0104]

[0105] The final large matrix is in the form of:

[0106]

[0107] in:

[0108] [SNN]=[SFN]-[SN]+[RN]+[RecomN]+[InterF_N1]+[InterF_N2]

[0109] [SFRN]=[RF]-[SFN]-[InterF_NF1]-[InterF_NF2]

[0110] [SNP]=[SFP]+[SP]+[RP]+[RecomP]+[InterF_P1]+[InterF_P2]

[0111] [SFRP]=[RF]-[SFP]+[InetrF_PF1]+[InetrF_PF2]

[0112] [STT]=-[S]-[T]-[InterF_B1]+[InterF_B2]

[0113] {SRXN}={SNX}-{RX}+{RecomN_RH}+{InterFN_RH}

[0114] {SRXP}=-{SPX}-{RX}+{RecomP_RH}+{InterF_PRH}

[0115] {SF}={F}+{InterF_BRH}

[0116] In step 5, due to the transient changes of single-particle effects, the unknown variables in the equations change suddenly, so the convergence difficulty is relatively high, requiring a finer grid and a smaller step. To improve the solution efficiency, the present invention uses the PARDISO solver to solve the single-particle effect model equations. The specific solution steps are as follows:

[0117] Phase 1: Fill-reduction analysis and symbolic factorization;

[0118] Phase 2: numerical decomposition;

[0119] Phase 3: forward and backward solving, including iterative optimization;

[0120] Termination and memory release phase (PHASE≤0).

[0121] After the convergence and stability conditions are met during the iteration process, the electron and hole quasi-Fermi potentials and electric potentials at the current moment are obtained. Based on this, the electron concentration, hole concentration, electron current, hole current, and displacement current of each node in the GaN HEMT physical model can be calculated:

[0122]

[0123] Finally, the electron current, hole current and displacement current are added together to obtain the transient drain current under single particle radiation.

[0124] This embodiment provides a single event effect transient simulation system for implementing the method, including:

[0125] The single-event effect physical model decomposition unit is used to establish the single-event effect physical model of GaN HEMT, perform hexahedral decomposition on the entire physical model, and obtain model structure information including the hexahedral unit number and node physical coordinates;

[0126] The SEP model equations discretization unit normalizes the SEP model equations including the drift-diffusion equations, Poisson's equation, and irradiation carrier generation rate based on the model structure information, and uses the backward Euler method to time discretize the SEP model equations.

[0127] The SEP model equation optimization unit uses the time-domain spectral element method to derive the discretized SEP model equations, adds thermal electron emission boundary conditions to the current continuity equation, and introduces the carrier generation rate term;

[0128] The single-particle effect model equation derivation unit derives the single-particle effect model equations through the Galerkin test method and performs Newton iteration to couple the equations into a large matrix form;

[0129] The solving unit solves and calculates the large matrix to obtain the electron quasi-Fermi potential, hole quasi-Fermi potential and electric potential that meet the convergence and stability conditions, and obtains the electron concentration, hole concentration, electron current, hole current and displacement current of each node. Finally, the electron current, hole current and displacement current are denormalized and added to obtain the transient drain current under single particle radiation.

[0130] This embodiment further provides a computer storage medium, wherein the computer storage medium stores an executable program, and the executable program is executed by a processor to implement the steps of the single event effect transient simulation method.

[0131] This paper utilizes the spectral element time-domain (SETD) method to simulate and analyze the generation process and underlying physical mechanisms of single-particle transient currents (SPTs) caused by heavy ion impacts in GaN HEMT devices. This method efficiently studies SPTs in GaN HEMTs through precise numerical calculations, providing a new platform for research on SPTs. This method can simulate and evaluate the transient effects of any high-energy charged particle (such as protons, heavy ions, or neutrons in cosmic rays) impacting a GaN HEMT.

[0132] Example

[0133] According to the method of the present invention, Figure 1 The GaN HEMT physical model shown in the figure is simulated. A step voltage of 0-0.3V is applied to the drain. After the drain bias voltage stabilizes at 0.3V, a single event effect is added at 40ps. The single event effect region is {0.55um≤X≤0.65um,0.1um≤Y≤0.415um}. The entire device is uniformly doped with N-type, and the donor doping concentration is 10 15 cm -3 The AlGaN layer is Gaussian doped near the source and drain, with a doping concentration of 5Δ×10 18 cm -3 . Figure 2 This is the transient simulation result of the single event effect. From the current change curve, we can see that the drain current will quickly increase to a certain value after the single event effect is added. Then, due to the recombination of carriers, the drain current will slowly return to the value before the single event effect. Figure 3 、 Figure 4 and Figure 5 It can be seen that when the single-particle effect is added, the carrier concentration at the incident trajectory increases significantly. Figure 6The relative error of the SETD method results at X = 0.55 μm compared with COMSOL is shown. Figure 6 The results demonstrate the accuracy of single-particle effect transient simulation of HEMT devices using the SETD method.

[0134] Obviously, those skilled in the art may make various changes and modifications to the embodiments of the present invention without departing from the spirit and scope of the embodiments of the present invention. Thus, if such changes and modifications of the embodiments of the present invention fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A method for simulating single event effects transient of GaN HEMT based on a physical model, characterized in that: include: Step 1: Establish a single-event effect physical model of the GaN HEMT and perform hexahedral meshing on the entire physical model to obtain model structure information including hexahedral unit numbers and node physical coordinates; Step 2: Based on the model structure information, the SEP model equations, including the drift-diffusion equations, Poisson's equation, and irradiation carrier generation rate, are normalized and time-discretized using the backward Euler method. Step 3: Derive the discretized single-event effect model equations using the time-domain spectral element method, add the boundary conditions for thermal electron emission to the current continuity equation, and introduce the carrier generation rate term; Step 4: Derive the single-particle effect model equations using the Galerkin test method and perform Newton iteration to couple the equations into a large matrix. Step 5: Solve and calculate the large matrix to obtain the electron quasi-Fermi potential, hole quasi-Fermi potential and electric potential that meet the convergence and stability conditions, and obtain the electron concentration, hole concentration, electron current, hole current and displacement current of each node. Finally, denormalize and add the electron current, hole current and displacement current to obtain the transient drain current under single-particle radiation.

2. The method for simulating single event effects transient of GaN HEMT based on a physical model according to claim 1, characterized in that: The single event effect model equations after time discretization in step 2 are: Where n and p represent the electron concentration and hole concentration respectively, n m-1 and p m-1 represent the electron concentration and hole concentration at the previous moment, respectively. and are the electron current density and hole current density, τ n and τ p denote the electron and hole lifetimes, α n and α p are the ionization coefficients of electrons and holes, G n and G p represent the radiation electron generation rate and hole generation rate respectively, represents the electric potential, Γ represents the net doping concentration, is the density operator, and q represents the unit charge.

3. The method for simulating single event effects transient of GaN HEMT based on a physical model according to claim 2, characterized in that: The electron concentration and hole concentration are: Where, φ n and φ p represent the quasi-Fermi potential of electrons and holes respectively, and M represents the ratio of the intrinsic carrier concentration of different materials to that of AlGaN.

4. The method for simulating single event effects transient of GaN HEMT based on a physical model according to claim 1, characterized in that: The carrier generation rate term in step 3 is: G n (G p )=G LET (l)R(w,l)T(t) In the formula, T(t) represents the time function term, R(w,l) represents the spatial function distribution, G LET (l) represents the linear energy generation density, t0 represents the single particle impact time, s hi represents the eigenvalue of the Gaussian function, erf represents the Gaussian error function, w represents the vertical distance from the internal point of the single particle incident area to the center of the incident trajectory, and w t (l) represents the radius of the single particle incident area, LET_f(l) represents the energy deposited per unit path length, and l represents the heavy ion incident depth.

5. The method for simulating single event effects transient of GaN HEMT based on a physical model according to claim 4, characterized in that: The step three further includes: replacing the two-dimensional electron gas of the HEMT device with the surface charge density in the Poisson equation and imposing a potential continuity condition.

6. The method for simulating single event effects transient of GaN HEMT based on a physical model according to claim 1, characterized in that: In step 4, the single-particle effect model equations are derived using the Galerkin test method to obtain the equation: Where: Where N i Represents the coefficient of the shape function, V n and V p represents the surface recombination velocity of electrons and holes, V n1 、V p1 and V n2 、V p2 They represent the surface recombination velocities of electrons and holes in the materials on both sides of the heterojunction, respectively; X1 and χ1 represent the electron affinities of the materials on both sides of the heterojunction; n0 and p0 represent the electron concentration and hole concentration in the quasi-equilibrium state; μ n and μ p represents the electron and hole mobility, G 雪崩 represents the electron generation term caused by avalanche current, E g1 and E g2 represents the band gap of the materials on both sides of the heterojunction, ρ s represents the surface charge density, ε2 represents the dielectric constant of AlGaN, C represents the ratio of the dielectric constants of GaN to AlGaN, WXF represents the weight factor, and represents the potential at GaN and AlGaN respectively, S represents the contact surface area at the heterojunction, V represents the volume of the HEMT device structure layer, A takes the value of 1 on the gate Schottky contact surface and takes the value of 0 on the ohmic contact surface, B takes the value of 1 at the continuous quasi-Fermi level at the AlGaN / GaN heterojunction surface and takes the value of 0 at other heterojunctions.

7. The method for simulating single event effects transient of GaN HEMT based on a physical model according to claim 1, characterized in that: The step five uses PARDISO to solve the large matrix. First, the matrix is sorted by METIS, and then stored in CSR format for solution.

8. The method for simulating single event effects transient of GaN HEMT based on a physical model according to claim 1, characterized in that: The electron concentration, hole concentration, electron current, hole current and displacement current of each node in step 5 are: Where k is the Boltzmann constant, T is the ambient temperature, and N i and N j Represent the shape functions of the expanded basis function and the test basis function, I ni , I pi and represent electron current, hole current and displacement current respectively.

9. A single event effect transient simulation system for implementing the method according to any one of claims 1 to 8, characterized in that: include: The single-event effect physical model decomposition unit is used to establish the single-event effect physical model of GaN HEMT, perform hexahedral decomposition on the entire physical model, and obtain model structure information including the hexahedral unit number and node physical coordinates; The SEP model equations discretization unit normalizes the SEP model equations including the drift-diffusion equations, Poisson's equation, and irradiation carrier generation rate based on the model structure information, and uses the backward Euler method to time discretize the SEP model equations. The SEP model equation optimization unit uses the time-domain spectral element method to derive the discretized SEP model equations, adds thermal electron emission boundary conditions to the current continuity equation, and introduces the carrier generation rate term; The single-particle effect model equation derivation unit derives the single-particle effect model equations through the Galerkin test method and performs Newton iteration to couple the equations into a large matrix form; The solving unit solves and calculates the large matrix to obtain the electron quasi-Fermi potential, hole quasi-Fermi potential and electric potential that meet the convergence and stability conditions, and obtains the electron concentration, hole concentration, electron current, hole current and displacement current of each node. Finally, the electron current, hole current and displacement current are denormalized and added to obtain the transient drain current under single particle radiation.

10. A computer storage medium, characterized in that The computer storage medium stores an executable program, and the executable program is executed by a processor to implement the steps of the single event effect transient simulation method according to any one of claims 1 to 8.

Images