A method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on the flowing gas method

Abstract

This invention provides a method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on a flowing gas method, relating to the field of microelectronics and device process simulation technology. The method includes the following steps: constructing a model of a silicon carbide crystal; setting a vacuum layer on the surface of the silicon carbide crystal in the model; using a reaction force field molecular dynamics method to bring the silicon carbide crystal to an initial state; heating the silicon carbide crystal to the reaction temperature; under isothermal conditions, at first preset time intervals, repeatedly emitting a group of O2 molecules at random locations at the gas rate corresponding to the reaction temperature onto the surface of the silicon carbide crystal, causing an oxidation reaction on the surface; after the oxidation reaction ends and reaches equilibrium, annealing the model and optimizing it to obtain an oxidized sample; and obtaining the structural characteristic parameters of the oxidized sample. This invention can simulate the generation process of interface defects in the oxidized sample formed by the dry oxygen thermal oxidation process, which is beneficial for obtaining control methods for the generation and reduction of interface defects.

Description

Technical Field

[0001] This invention relates to the field of microelectronics and device process simulation technology, and more specifically, to a method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on the flowing gas method. Background Technology

[0002] The rapid development of silicon-based semiconductor devices in the 20th century benefited from in-depth research on materials and interface structures. However, the inherent physical properties of silicon limit the theoretical performance of Si power devices. Silicon carbide (SiC), a next-generation wide-bandgap semiconductor, is considered an ideal candidate material for high-power devices due to its superior physicochemical properties, attracting significant attention in microelectronics, particularly in aerospace, new energy vehicles, high-speed rail, industrial motors, and smart grids. One of the main failure mechanisms of MOSFET devices is dielectric breakdown caused by oxide degradation. The strength of dielectric breakdown is highly correlated with the model structure at the interface, making the study of semiconductor interface models a widely studied scientific issue. The most crucial part of a silicon carbide MOSFET is the SiC / SiO2 interface structure. Due to the flexible Si-O bonding in silicon dioxide, it typically exists in amorphous forms of varying densities. Numerous interface states exist between the interface and the insulating layer. Within a 3-5 nm range from the interface, there exists a transition layer with varying composition and containing different structural defects. Many of these electrically active defects are the root cause of device performance degradation. During MOSFET device fabrication, the type and concentration of interface states and defects are closely related to the actual manufacturing process. Exploring the formation mechanism of defects at the SiC / SiO2 interface and clarifying the origin and reduction techniques of interface states have significant scientific and engineering practical value. The main method for obtaining oxide insulating gates (SiO2) from SiC materials is thermal oxidation. However, the chemical reactions and interface structures involved in the thermal oxidation of silicon carbide are more complex than those in Si / SiO2, and many major scientific questions remain to be solved.

[0003] Therefore, developing reliable simulation methods to link the microstructure of the MOSFET interface with the actual process and explore the relationship between interface structure and device performance can provide a strong guarantee for developing more effective semiconductor device fabrication methods. Summary of the Invention

[0004] The problem addressed by this invention is how to understand and mitigate the causes of interface defects formed by thermal oxidation.

[0005] To address the above problems, this invention provides a method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on a flowing gas method, comprising the following steps:

[0006] Step S1: Construct a silicon carbide crystal model, set a vacuum layer on the surface of the silicon carbide crystal in the model, and use the reactive force field molecular dynamics method to bring the silicon carbide crystal to its initial state;

[0007] Step S2: Heat the silicon carbide crystal to the reaction temperature. Under constant temperature, at a first preset time interval, repeatedly emit a group of O2 molecules at random positions at the gas rate corresponding to the reaction temperature to the surface of the silicon carbide crystal until an oxidation reaction occurs on the surface of the silicon carbide crystal. After the oxidation reaction ends and reaches equilibrium, anneal the model and optimize it to obtain an oxidized sample.

[0008] Step S3: Obtain the structural characteristic parameters of the oxidized sample.

[0009] Furthermore, in step S1, the shape of the model is a cuboid.

[0010] Furthermore, in step S1, the initial state of the silicon carbide crystal includes optimizing the positions of the silicon carbide atoms, relaxing them to atmospheric pressure, and making the force state of each atom zero.

[0011] Further, in step S2, heating the silicon carbide crystal to the reaction temperature includes: slowly heating from 300K to the reaction temperature and maintaining the reaction temperature until the model reaches equilibrium.

[0012] Further, in step S2, the step of emitting a group of O2 molecules at random locations at a gas rate corresponding to the reaction temperature toward the surface of the silicon carbide crystal, until an oxidation reaction occurs on the surface of the silicon carbide crystal, includes the following steps:

[0013] The process involves emitting a group of O2 molecules at random locations at a gas rate corresponding to the reaction temperature onto the surface of the silicon carbide crystal. The O2 molecules collide with atoms on the silicon carbide crystal surface, and a second preset time is waited. If the O2 molecules react with atoms on the silicon carbide crystal surface, the O2 molecules are retained; if they do not react, the unreacted O2 molecules are removed. This process is repeated until an oxidation reaction occurs on the surface of the silicon carbide crystal.

[0014] Further, in step S2, the equilibrium state includes the state after the oxidation reaction of the model has ended, the reaction temperature has been maintained at 300 ps, ​​and the model has been relaxed.

[0015] Furthermore, in step S2, the oxidized sample includes unoxidized silicon carbide crystals, a transition layer, and an oxide layer.

[0016] Furthermore, in step S3, the structural characteristic parameters of the oxidized sample include at least the density of the oxide layer, the silicon-oxygen ratio, the defect type, and the defect concentration.

[0017] Furthermore, in step S2, the reaction temperature ranges from 900 to 1100°C.

[0018] Further, in step S2, the first preset time and / or the second preset time is 0.3-1ps.

[0019] The method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on the flowing gas method described in this invention has the advantage of the prior art. By constructing a silicon carbide crystal model and employing reactive force field molecular dynamics, a group of O2 molecules is repeatedly emitted at random positions in the vacuum layer of the model at a gas rate corresponding to the reaction temperature towards the silicon carbide crystal surface. This simulates the reaction process between the flowing gas and the silicon carbide crystal surface in actual dry oxygen thermal oxidation experiments. Appropriately increasing the number of activated molecules in the O2 molecules increases the probability of effective collisions with atoms on the silicon carbide crystal surface, thereby increasing the reaction rate. Repeated addition of O2 molecules achieves rapid oxidation with high efficiency. After the oxidation reaction ends and reaches equilibrium, the model is annealed and optimized to obtain an oxidized sample. The structural characteristic parameters of the oxidized sample are then obtained, thus understanding the structural characteristic parameters of the oxidized sample formed by the dry oxygen thermal oxidation process. Through process control and analysis of the structural characteristic parameters of the oxidized sample, the causes of interface defects in the oxidized sample formed by the dry oxygen thermal oxidation process can be accurately controlled. Through repeated simulation verification, the process control method for reducing interface defects can be mastered. This invention simulates the actual dry oxygen thermal oxidation process to understand the formation process and results of the atomic structure layer of the oxidized sample, enabling in-depth analysis and control. At the same time, it avoids conducting actual dry oxygen thermal oxidation experiments, reducing costs, improving efficiency, and making it suitable for large-scale application. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on the flowing gas method in an embodiment of the present invention;

[0021] Figure 2 This is a structural diagram of the silicon carbide crystal before dry oxygen thermal oxidation in an embodiment of the present invention.

[0022] Figure 3 This is a structural configuration diagram of silicon carbide crystal during the dry oxygen thermal oxidation process in an embodiment of the present invention;

[0023] Figure 4 This is a structural diagram of the silicon carbide crystal after dry oxygen thermal oxidation in an embodiment of the present invention.

[0024] Figure 5 This is a statistical chart showing the concentration of various defects in the oxide layer at 1300K in an embodiment of the present invention;

[0025] Figure 6 This is a statistical analysis diagram of the bond angles of atoms forming bonds in the oxide layer in this embodiment of the invention;

[0026] Figure 7 This is a statistical analysis diagram of the atomic bond lengths formed in the oxide layer in an embodiment of the present invention. Detailed Implementation

[0027] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] It should be noted that, in the description of the embodiments of this application, the term "some specific embodiments" means that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same implementation or instance. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0029] like Figure 1 As shown, this embodiment of the invention provides a method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on the flowing gas method, comprising the following steps:

[0030] Step S1: Construct a model of silicon carbide crystal, set a vacuum layer on the surface of the silicon carbide crystal in the model, and use the reactive force field molecular dynamics method to bring the silicon carbide crystal to its initial state;

[0031] Step S2: Heat the silicon carbide crystal to the reaction temperature. Under constant temperature, at a first preset time interval, repeatedly emit a group of O2 molecules at random positions to the surface of the silicon carbide crystal at the gas rate corresponding to the reaction temperature, so that an oxidation reaction occurs on the surface of the silicon carbide crystal. After the oxidation reaction ends and reaches equilibrium, anneal the model and optimize it to obtain the oxidation sample.

[0032] Step S3: Obtain the structural characteristic parameters of the oxidized sample.

[0033] The method for simulating the dry oxygen thermal oxidation process of single-crystal silicon carbide based on the flowing gas method described in this invention constructs a silicon carbide crystal model and uses the reaction force field molecular dynamics method to repeatedly launch a group of O2 molecules at random positions in the vacuum layer of the model at the gas rate corresponding to the reaction temperature towards the silicon carbide crystal surface. This simulates the reaction process between the flowing gas and the silicon carbide crystal surface in the actual dry oxygen thermal oxidation process. By appropriately increasing the number of activated molecules in the O2 molecules, the probability of effective collisions with atoms on the silicon carbide crystal surface is increased, thereby improving the reaction rate. Repeated addition of O2 molecules can achieve rapid oxidation with high efficiency. After the oxidation reaction ends and reaches equilibrium, the model is annealed and optimized to obtain an oxidized sample. The structural characteristic parameters of the oxidized sample are then obtained to understand the structural characteristic parameters of the oxidized sample formed by the dry oxygen thermal oxidation process. Through process control and analysis of the structural characteristic parameters of the oxidized sample, the causes of interface defects in the oxidized sample formed by the dry oxygen thermal oxidation process can be accurately controlled. Through repeated simulation verification, the process control method for reducing interface defects can be mastered. This invention simulates the actual dry oxygen thermal oxidation process to understand the formation process and results of the atomic structure layer of the oxidized sample, achieving in-depth analysis and control. At the same time, it avoids conducting actual dry oxygen thermal oxidation experiments, reducing costs, improving efficiency, and making it suitable for large-scale application.

[0034] Specifically, in step S1 of this embodiment, a model of a silicon carbide crystal with periodic boundary conditions is constructed in the analysis kinetic simulation software. A vacuum layer is set on the surface of the silicon carbide crystal in the model for oxidation reaction. The reaction force field molecular dynamics method is used to put the silicon carbide crystal in the initial state, which helps to reduce the influence of silicon carbide atoms on the oxidation process and is more suitable for mastering the process reaction data.

[0035] Specifically, in step S2 of this embodiment, the silicon carbide crystal is slowly heated to the reaction temperature, causing the molecules or atoms in the reaction to become activated molecules. Under constant temperature conditions, at first preset time intervals, a group of O2 molecules is repeatedly emitted from random positions at a rate corresponding to the gas velocity at the reaction temperature onto the surface of the silicon carbide crystal, resulting in an oxidation reaction. This increases the effective number of collisions between O2 molecules and silicon carbide in the same amount of time, thus increasing the reaction rate. Emitting O2 molecules in groups at random positions results in a faster reaction rate and higher oxidation efficiency compared to emitting individual O2 molecules each time. After the oxidation reaction is complete, the model is held for a period of time to reach equilibrium before annealing. The annealing time and temperature are controlled, and after annealing to 0K, the positions of the atoms in each part after the reaction are optimized to finally obtain the oxidized sample.

[0036] Specifically, in step S3 of this embodiment, the structural characteristic parameters of the oxidized sample are acquired and statistically analyzed, especially the defect types and defect concentrations generated by the dry oxygen thermal oxidation process. The actual dry oxygen thermal oxidation process is optimized by using the guidance of simulation.

[0037] In some specific embodiments, in step S1, the model is rectangular in shape. Therefore, defining the boundary conditions of the silicon carbide crystal model facilitates its periodic expansion and application.

[0038] In some specific embodiments, step S1, where the silicon carbide crystal is in its initial state, involves optimizing the positions of the silicon carbide atoms and relaxing them to ambient pressure so that the force state of each atom is zero. This controls and precisely determines the atomic force state in the initial state, reducing interference from the initial state on the oxidation process, while also accurately controlling the influence of various process parameters on the final oxidized sample.

[0039] In some specific embodiments, step S2, heating the silicon carbide crystal to the reaction temperature, includes: slowly heating from 300K to the reaction temperature and maintaining the reaction temperature until the model reaches equilibrium. Therefore, starting heating from 300K reduces the ineffective time generated by heating from 0K, improving efficiency. The slow heating process helps improve the stability of the atomic state of the silicon carbide crystal and ensures uniform temperature diffusion, reducing local temperature unevenness that could affect the ordered atomic movement of the silicon carbide crystal. After reaching the reaction temperature, maintaining the atomic movement in the silicon carbide crystal model at that temperature for a period of time allows the atoms to reach a more stable and ordered equilibrium state, which helps define the state before the oxidation reaction, making the reaction process clear and reproducible.

[0040] In some specific embodiments, step S2 involves emitting a group of O2 molecules at random locations at a gas rate corresponding to the reaction temperature onto the silicon carbide crystal surface, causing an oxidation reaction to occur on the silicon carbide crystal surface. This includes the following steps:

[0041] A group of O2 molecules are emitted at random locations at a gas rate corresponding to the reaction temperature onto the surface of a silicon carbide crystal. The O2 molecules collide with the atoms on the surface of the silicon carbide crystal and wait for a second preset time. If the O2 molecules react with the atoms on the surface of the silicon carbide crystal, the O2 molecules are retained. If the O2 molecules do not react with the atoms on the surface of the silicon carbide crystal, the unreacted O2 molecules are removed. The above steps are repeated until an oxidation reaction occurs on the surface of the silicon carbide crystal.

[0042] Therefore, setting a second preset time is beneficial for observing whether O2 molecules collide with atoms on the silicon carbide crystal surface again, thereby reducing the number of times O2 molecules are repeatedly emitted and increasing the effective number of O2 molecules per group. O2 molecules that do not undergo an effective reaction within the second preset time, or other gas molecules generated by the reaction, are removed to avoid affecting the trajectory of re-emitted O2 molecules, increasing the effective collision with atoms on the silicon carbide crystal surface, and accelerating the oxidation reaction rate.

[0043] In some specific embodiments, in step S2, the equilibrium state includes the state after the oxidation reaction of the model has ended, with the reaction temperature maintained at 300-600 ps and the model relaxed. Therefore, maintaining the reaction temperature at 300 ps after the reaction ends allows the atoms in the model to relax to a more stable and ordered equilibrium state, which is beneficial for defining the state of each step, making the reaction process clearer and more repeatable.

[0044] In some specific embodiments, in step S2, the oxidized sample includes unoxidized silicon carbide crystal, a transition layer, and an oxide layer. Thus, after the oxidation reaction, atoms on the surface of the silicon carbide crystal collide with O2 molecules and are oxidized, forming an oxide layer. Due to the randomness of the reaction, a partially oxidized transition layer also appears. Below the transition layer is unoxidized silicon carbide crystal that O2 molecules have not reached. Precisely dividing the oxidized sample is beneficial for studying the structural formation of each layer and the causes of defects.

[0045] In some specific embodiments, in step S3, the structural characteristic parameters of the oxidized sample include at least the oxide layer density, silicon-to-oxygen ratio, defect type, and defect concentration. Therefore, statistical analysis of the structural characteristic parameters that significantly affect performance is beneficial for understanding the causes of interface defects formed by thermal oxidation and for summarizing methods to reduce defects.

[0046] In some specific embodiments, the reaction temperature in step S2 is in the range of 900-1100°C. Therefore, a reaction temperature within this range is beneficial for increasing the number of activated molecules in the model, increasing the number of effective collisions, reducing reaction time, and improving reaction efficiency.

[0047] In some specific embodiments, in step S2, the first preset time and / or the second preset time is 0.3-1 ps. This provides sufficient reaction time for the reaction to proceed fully, while also helping to maintain the stability of the atomic states in the model and improving the repeatability of the simulation.

[0048] Example 1

[0049] A silicon carbide crystal model was established using the molecular dynamics simulation software LAMMPS, and the actual dry oxygen thermal oxidation process was simulated. The interatomic interaction potential function was described using the ReaxFF reaction force field developed by Larsson et al. A 2.7 nm × 2.7 nm × 2.7 nm silicon carbide crystal model was constructed as the substrate, containing 1800 Si and 1800 O atoms, with a 5.5 nm thick vacuum layer added. Figure 2 As shown. A group of O2 molecules were emitted along the reverse

[100] crystal orientation, i.e., perpendicular to the silicon carbide crystal surface downwards, for a total of 2000 cycles. The first preset time was 1 fs, for a total of 600,000 steps, 600 ps, ​​and the reaction temperature was controlled at 1300 K. The O2, CO gas and CO2 gas generated during the reaction and escaping from the surface were removed and cleaned; the generated oxide layer was a SiO2 thin film. The configuration in the dry oxygen thermal oxidation process is shown in the figure. Figure 3 As shown, the final atomic configuration is as follows Figure 4 As shown, the thickness of the obtained SiO2 oxide film is 1.7 nm. The statistical results of the defect types and concentrations of the oxidized oxide layer are as follows: Figure 5 As shown in the figure, the NBOHC and POR defects account for the largest proportions in the silicon dioxide film after dry oxidation of silicon carbide, at 2.65 × 10²⁰ defects / cm³ and 7.95 × 10²⁰ defects / cm³, respectively. The statistical results of bond lengths and bond angles in the oxide layer obtained after oxidation are shown in the figure. Figure 6 and Figure 7 As shown, the bond angles of O-Si-O are mainly distributed around 95.95 degrees, while the bond angles of Si-O-Si are mainly distributed around 149.19 degrees.

[0050] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method of simulating a dry thermal oxidation process of a single crystal silicon carbide based on a flowing gas method, characterized in that, The method comprises the following steps: Step S1: a model of a silicon carbide crystal is constructed by using a molecular dynamics simulation software LAMMPS, a vacuum layer is arranged on the surface of the silicon carbide crystal in the model, and a reaction force field molecular dynamics method is used to make the silicon carbide crystal in an initial state; Step S2: the silicon carbide crystal is heated to a reaction temperature, under a constant temperature state, a group of O2 molecules are emitted to the surface of the silicon carbide crystal at a random position at a gas rate corresponding to the reaction temperature at the reaction temperature, and an oxidation reaction occurs on the surface of the silicon carbide crystal; after the oxidation reaction ends and reaches an equilibrium state, the model is annealed, and an oxidation sample is obtained after optimization; Step S3: structure characteristic parameters of the oxidation sample are obtained; the structure characteristic parameters of the oxidation sample at least include density, silicon-oxygen ratio, defect type and defect concentration of the oxidation layer; the structure characteristic parameters of the oxidation sample are used to master the cause of interface defects formed by a thermal oxidation method and summarize a method of reducing defects; In step S2, the emission of a group of O2 molecules to the surface of the silicon carbide crystal at a random position at a gas rate corresponding to the reaction temperature at the reaction temperature, and the occurrence of an oxidation reaction on the surface of the silicon carbide crystal comprise the following steps: The emission of a group of O2 molecules to the surface of the silicon carbide crystal at a random position at a gas rate corresponding to the reaction temperature at the reaction temperature makes the O2 molecules collide with atoms on the surface of the silicon carbide crystal and wait for a second preset time; if the O2 molecules react with the atoms on the surface of the silicon carbide crystal, the O2 molecules are retained; if the O2 molecules do not react with the atoms on the surface of the silicon carbide crystal, the unreacted O2 molecules are removed; and the above steps are repeated until the oxidation reaction occurs on the surface of the silicon carbide crystal; The equilibrium state includes a state after the oxidation reaction of the model ends, the reaction temperature is maintained for 300-600 ps, and relaxation is performed; and the oxidation sample includes unoxidized silicon carbide crystal, a transition layer and an oxidation layer.

2. The method of simulating a dry thermal oxidation process of single crystal silicon carbide according to claim 1, wherein, In step S1, the shape of the model is a cuboid shape.

3. The method of simulating a dry thermal oxidation process of single crystal silicon carbide according to claim 1, wherein, In step S1, the initial state of the silicon carbide crystal includes optimization of the atomic positions of the silicon carbide, and relaxation to a normal pressure state, so that the force state of each atom is zero.

4. The method of simulating a dry thermal oxidation process of single crystal silicon carbide according to claim 1, wherein, In step S2, the heating of the silicon carbide crystal to the reaction temperature comprises: slowly heating from 300K to the reaction temperature, and maintaining the reaction temperature until the model balances.

5. The method of simulating a dry thermal oxidation process of single crystal silicon carbide according to claim 1, wherein, In step S2, the reaction temperature ranges from 900-1100℃.

6. The method of simulating a dry thermal oxidation process of monocrystalline silicon carbide according to claim 1, wherein, In step S2, the first preset time and / or the second preset time is 0.3-1 ps.

Images