A method for growing a silicon carbide crystal and a growing apparatus therefor

Abstract

The application provides a silicon carbide crystal growth method and a growth device thereof, and belongs to the technical field of silicon carbide single crystal growth. The growth method comprises the following steps: providing a crystal growth device, mounting a silicon carbide seed crystal on a seed crystal holder, and cyclically performing a preset number of redox pretreatments; the redox pretreatment comprises the following steps: using an oxidizing gas to perform oxidation treatment on a growth surface of the silicon carbide seed crystal to obtain an oxidation layer, and then using a reducing gas to perform reduction treatment on the oxidation layer and to desorb; after the cyclic execution, a pit is generated at a basal plane dislocation emergence of the seed crystal surface, and a pretreated seed crystal is obtained; an asymmetric structure is adopted to fill silicon carbide polycrystal materials at the bottom of a crucible in the crystal growth device; a physical vapor transport method is used for crystal growth, so that the growth rate of a central region of the crystal is greater than that of an edge region, and a finished product is obtained. The device for the above method is also provided. The problems that the existing BPD cannot be effectively removed, the existing process cost is high, and the yield is low are solved.

Description

Technical Field

[0001] This invention belongs to the field of silicon carbide single crystal growth technology, specifically relating to a method and apparatus for growing silicon carbide crystals. Background Technology

[0002] Silicon carbide, as a third-generation semiconductor material, possesses excellent physical properties such as a wide bandgap, high critical breakdown field strength, and high saturated electron drift velocity, making it promising for applications in power electronics, radio frequency devices, new energy vehicles, and smart grids. The physical vapor transport (PVT) method is currently the mainstream method for the industrial production of silicon carbide single crystals. Its basic principle is to sublimate the silicon carbide raw material under high temperature and low pressure conditions, allowing the gaseous components to gradually deposit on the seed crystal to achieve crystal growth.

[0003] However, silicon carbide crystals grown by the PVT method inevitably contain various crystal defects, among which basis plane dislocations (BPDs) are one of the main defects affecting device performance and reliability. Studies have shown that BPDs can expand into stacking faults during device operation, leading to increased forward voltage drop and reverse leakage current, and in severe cases, device failure. Moreover, BPDs are mostly concentrated in the crystal edge region. As the crystal growth thickness increases, edge BPDs extend along the basis plane towards the crystal center region, contaminating the originally high-quality central region.

[0004] Currently, the industry mainly uses two methods to address this problem: First, using larger seed crystals for crystal growth, and then removing areas with high BPD density at the wafer edge through grinding; second, enhancing the conversion of BPDs to through-edge dislocations (TEDs) on the substrate surface through etching, high-temperature annealing, interrupting growth, and increasing the epitaxial rate. However, the former leads to material waste and yield loss, and is costly; the latter is mainly applied in the epitaxial layer preparation stage, and has limited effect on suppressing BPDs during the crystal growth stage.

[0005] Currently, there is no systematic solution in the technology to effectively confine the boundary phase deviation (BPD) generated at the crystal edge to the removable region during silicon carbide crystal growth and prevent it from extending into the high-quality central region. There is an urgent need in the art for a crystal growth method that can suppress the extension of BPD towards the crystal center from the source. Summary of the Invention

[0006] The present invention aims to solve the technical problems of high cost and low yield caused by the lack of effective means to suppress BPD in the existing silicon carbide crystal growth process, which mainly relies on the passive method of larger seed crystals and edge removal.

[0007] The specific technical solution of the present invention is as follows: In a first aspect, the present invention provides a method for growing silicon carbide crystals, comprising the following steps: Provide a crystal growth apparatus and silicon carbide seed crystals, install the silicon carbide seed crystals on the seed crystal holder, and perform a preset number of oxidation-reduction pretreatment cycles; The oxidation-reduction pretreatment includes oxidizing the growth surface of the silicon carbide seed crystal with an oxidizing gas to obtain an oxide layer, and then reducing the oxide layer with a reducing gas and desorbing it. After the loop is completed, a pit is generated at the basis vector plane dislocation outcrop on the surface of the seed crystal, and a pre-treated seed crystal is obtained. The bottom of the crucible in the crystal growth apparatus is filled with silicon carbide polycrystalline material using an asymmetric structure. Crystal growth is performed using the physical vapor transport method, ensuring that the growth rate in the central region of the crystal is greater than that in the edge region, thus obtaining the finished product.

[0008] Optionally, the preset number of times is 2 to 3 times.

[0009] Optionally, in the oxidation treatment, the pressure is 100~1000Pa, the temperature is 900~1200℃, the time is 30~180min, and the flow rate of the oxidizing gas is 50~500sccm.

[0010] Optionally, in the reduction process, the pressure is 400~800Pa, the temperature is 800~1100℃, the time is 10~60min, and the flow rate of the reducing gas is 100~1000sccm.

[0011] Optionally, after performing a preset number of redox pretreatments, the process may also include: The seed crystal surface is subjected to a final reduction treatment using a reducing gas, wherein the flow rate of the reducing gas in the final reduction treatment is 120% of the flow rate of the reducing gas in the previous reduction treatment.

[0012] Optionally, the asymmetric structure is: the central region is filled with fine-grained silicon carbide polycrystalline material with a particle size of 100~300μm, and the edge region is filled with coarse-grained silicon carbide polycrystalline material with a particle size of 800~1500μm.

[0013] Optionally, the filling height of the silicon carbide polycrystalline material in the central region is 85% to 95% of the crucible height, and the filling height of the silicon carbide polycrystalline material in the edge region is 50% to 65% of the crucible height; the filling density ratio of the central region to the edge region is 1:0.4 to 0.6.

[0014] Optionally, the radius of the central region is 40% to 50% of the radius of the entire silicon carbide polycrystalline material filling area.

[0015] Optionally, after the crystal growth is completed, a gradient cooling process is performed so that the cooling rate of the crystal edge region is faster than that of the central region, and the difference in cooling rate between the edge region and the central region is 5~10℃ / min.

[0016] Optionally, a porous graphite plate is disposed between the raw material and the seed crystal, the porous graphite plate having different porosities and / or pore sizes in the radial direction. The porosity of the central region of the porous graphite plate is higher than that of the edge region, and the pore diameter of the central region is larger than that of the edge region.

[0017] Optionally, the porous graphite plate has a porosity of 0.55~0.65 and a pore size of 300~500μm in the central region and a porosity of 0.10~0.18 and a pore size of 5~20μm in the edge region.

[0018] A silicon carbide crystal growth apparatus is also provided for the above-described method, comprising: A crucible used to hold silicon carbide raw materials; A seed crystal holder is placed above the crucible to fix the pretreated seed crystal. An atmosphere supply component is used to provide oxidizing and reducing gases to achieve redox pretreatment of the growth surface of silicon carbide seed crystals. A porous graphite plate is disposed between the crucible and the pretreated seed crystal, wherein the porosity of the central region of the porous graphite plate is greater than that of the edge region. A heating device is used to heat the interior of the silicon carbide crystal growth apparatus.

[0019] Optionally, the porous graphite plate has a porosity of 0.55~0.65 and a pore size of 300~500μm in the central region and a porosity of 0.10~0.18 and a pore size of 5~20μm in the edge region.

[0020] Optionally, a cooling mechanism may also be included to ensure that the cooling rate of the crystal edge region is greater than that of the central region during the cooling phase. Compared with the prior art, the advantages of the present invention are as follows:

[0021] 1. The silicon carbide crystal growth method provided by the present invention locks the BPD within a limited range at the edge through the combined synergy of redox pretreatment and asymmetric structure loading, and then obtains a high-quality central crystal by processing and removing the edge locking ring.

[0022] Specifically, the redox pretreatment step involves a selective redox reaction at the BPD outcrop to generate pits. Due to the disordered atomic arrangement, numerous dangling bonds, and lattice distortion in the BPD dislocation core region, and the lower chemisorption activation energy of O2 molecules in the BPD core region compared to the intact lattice region, the oxidation reaction rate constant and physical adsorption coefficient are higher. Therefore, the oxidation reaction preferentially and kinetically occurs only at the BPD outcrop, leaving an atomic-level V-shaped pit at the BPD location. This step solves the technical challenge of precisely locating the BPD on the seed crystal surface, providing a physical structural basis for subsequent locking.

[0023] The asymmetric packing step employs an asymmetric structure at the bottom of the crucible, resulting in lower vapor transport resistance and increased sublimation rate in the central region, while hindered vapor transport and reduced sublimation rate in the edge region. This creates a difference in growth rate in the radial direction of the crystal, with faster growth at the center and slower growth at the edges. This rate difference causes the step flow in the central region to advance at high speed, driving the BPDs generated at the edges along the basal plane direction to the edge region. Simultaneously, the slow growth at the edges deprives the BPDs driven to the edges of the kinetics for further extension, ultimately causing them to fall into the pits left by the pretreatment and be locked in place.

[0024] In summary, this invention provides a capture trap through preprocessing and generates driving force through asymmetric loading. The two work together to lock the edge BPD in the edge region, actively controlling the spatial distribution of BPD from the source of growth. Experimental data show that after using the method of this invention, the BPD density in the central region is reduced to 0~150 / cm². 2 Edge locking ring (BPD) density ≥ 2500 / cm² 2 The density of the central BPD is reduced by an order of magnitude, and the process is compatible with existing PVT crystal growth equipment and does not require large-scale modification, showing good prospects for industrial application.

[0025] This solves the technical problems of high cost and low yield in existing silicon carbide crystal growth processes, which lack effective means to suppress BPD and mainly rely on passive methods such as larger seed crystals and edge removal.

[0026] 2. The silicon carbide crystal growth apparatus provided by the present invention, through the structural cooperation of crucible, seed crystal holder, atmosphere supply component, porous graphite plate and heating device, provides hardware guarantee for BPD space locking of the seed crystal after oxidation-reduction pretreatment in the above method. The apparatus has a simple structure and is easy to apply in industrial applications.

[0027] Specifically, the device of the present invention employs an asymmetric structure of a porous graphite plate, where the porosity of the central region of the porous graphite plate is greater than that of the edge region, resulting in a radially distributed pattern of gas phase components within the crucible that is centrally conductive and edge-blocked. Furthermore, the porous graphite plate, in conjunction with the asymmetric raw material loading structure at the bottom of the crucible, further reduces the supersaturation in the edge region, causing the growth rate in the edge region to be significantly lower than that in the center, thus effectively locking the edge BPD (Biological Phase Diode) within the edge region. Attached Figure Description

[0028] Figure 1 This is a schematic flowchart of the silicon carbide crystal growth method of the present invention.

[0029] Figure 2 This is a schematic diagram of the silicon carbide crystal growth apparatus of the present invention.

[0030] Figure 2 In the middle, 1-crucible, 2-seed crystal holder, 3-porous graphite plate, 4-heating device, 5-atmosphere supply component. Specific Implementation

[0031] In a first aspect, embodiments of the present invention provide a method for growing silicon carbide crystals; please refer to [link to relevant documentation]. Figure 1 This includes the following steps: Provide a crystal growth apparatus and a silicon carbide seed crystal. Install the silicon carbide seed crystal on the seed crystal holder inside the crystal growth apparatus and perform a preset number of oxidation-reduction pretreatment cycles. The oxidation-reduction pretreatment includes oxidizing the growth surface of the silicon carbide seed crystal with an oxidizing gas to obtain an oxide layer, and then reducing the oxide layer with a reducing gas and desorbing it. After the loop is completed, a pit is generated at the basis vector plane dislocation outcrop on the surface of the seed crystal, and a pre-treated seed crystal is obtained. The bottom of the crucible in the crystal growth apparatus is filled with silicon carbide polycrystalline material using an asymmetric structure. Crystal growth is performed using the physical vapor transport method, ensuring that the growth rate in the central region of the crystal is greater than that in the edge region, thus obtaining the finished product.

[0032] In some embodiments of the present invention, the preset number of times is 2 to 3 times.

[0033] In some embodiments of the present invention, the oxidation treatment is performed at a pressure of 100-1000 Pa, a temperature of 900-1200 °C, a time of 30-180 min, and a flow rate of oxidizing gas of 50-500 sccm.

[0034] In some embodiments of the present invention, the reduction process is carried out at a pressure of 400-800 Pa, a temperature of 800-1100 °C, a time of 10-60 min, and a flow rate of 100-1000 sccm for the reducing gas.

[0035] In some embodiments of the present invention, after performing a preset number of redox pretreatments, the method further includes: The seed crystal surface is subjected to a final reduction treatment using a reducing gas, wherein the flow rate of the reducing gas in the final reduction treatment is 120% of the flow rate of the reducing gas in the previous reduction treatment.

[0036] In some embodiments of the present invention, the asymmetric structure is: the central region is filled with fine-grained silicon carbide polycrystalline material with a particle size of 100~300μm, and the edge region is filled with coarse-grained silicon carbide polycrystalline material with a particle size of 800~1500μm.

[0037] In some embodiments of the present invention, the filling height of the silicon carbide polycrystalline material in the central region is 85% to 95% of the crucible height, and the filling height of the silicon carbide polycrystalline material in the edge region is 50% to 65% of the crucible height; the filling density ratio of the central region to the edge region is 1:0.4 to 0.6.

[0038] In some embodiments of the present invention, the radius of the central region is 40% to 50% of the radius of the entire silicon carbide polycrystalline material filling region.

[0039] In some embodiments of the present invention, after the crystal growth is completed, a gradient cooling is performed so that the cooling rate of the crystal edge region is faster than that of the central region, and the difference in cooling rate between the edge region and the central region is 5~10℃ / min.

[0040] In some embodiments of the present invention, a porous graphite plate is disposed between the raw material and the seed crystal, the porous graphite plate having different porosities and / or pore sizes in the radial direction. The porosity of the central region of the porous graphite plate is higher than that of the edge region, and the pore diameter of the central region is larger than that of the edge region.

[0041] In some embodiments of the present invention, the porosity of the central region of the porous graphite plate is 0.55~0.65 and the pore size is 300~500μm, while the porosity of the edge region is 0.10~0.18 and the pore size is 5~20μm.

[0042] In some embodiments of the present invention, the oxidizing gas includes, but is not limited to, oxygen.

[0043] In some embodiments of the present invention, the reducing gas includes, but is not limited to, hydrogen.

[0044] In some embodiments of the present invention, before the redox pretreatment, a step of depressurizing the growth apparatus is included to reduce the pressure inside the growth apparatus to a low pressure suitable for gas molecule transport, so as to ensure that the oxidizing gas subsequently introduced can reach the surface of the seed crystal.

[0045] Specifically, the argon pressure inside the growth apparatus is reduced from 800 mbar to 50 mbar, so that the pressure inside the growth apparatus is reduced to a low pressure suitable for gas molecule transport, so as to ensure that the oxidizing gas introduced later can reach the surface of the seed crystal, while avoiding the occurrence of SiC thermal corrosion.

[0046] A silicon carbide crystal growth apparatus is also provided; please refer to [link / reference]. Figure 2 The method described above includes: Crucible 1 is used to contain silicon carbide raw materials; Seed crystal holder 2 is disposed above the crucible 1 and is used to fix the pretreated seed crystal; Atmosphere supply component 5 is used to provide oxidizing and reducing gases to achieve redox pretreatment of the growth surface of silicon carbide seed crystals. A porous graphite plate 3 is disposed between the crucible 1 and the pretreated seed crystal, wherein the porosity of the central region of the porous graphite plate 3 is greater than that of the edge region. Heating device 4 is used to heat the interior of the silicon carbide crystal growth apparatus.

[0047] In some embodiments of the present invention, the porosity of the central region of the porous graphite plate is 0.55~0.65 and the pore size is 300~500μm, while the porosity of the edge region is 0.10~0.18 and the pore size is 5~20μm.

[0048] In some embodiments of the present invention, a cooling mechanism is also included, which is used to make the cooling rate of the crystal edge region greater than the cooling rate of the center region during the cooling stage.

[0049] Specifically, the cooling mechanism can be a partitioned coil structure, which can independently control the heating power of the central area and the edge area.

[0050] Specifically, the cooling mechanism can also be a phase change heat storage body located in the central area, which releases latent heat during the cooling process to slow down the cooling rate of the central area.

[0051] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto. Unless otherwise specified, the equipment, raw materials and testing methods used in the following embodiments are all conventionally selected in the art.

[0052] Example 1 This embodiment provides a method for growing silicon carbide crystals, the specific steps of which are as follows: (1) Pretreatment of seed crystal surface by oxidation-reduction cycle Provide a crystal growth apparatus to fix a 4H-SiC seed crystal (210 mm in diameter and 105 mm in radius) on a seed crystal holder inside the crystal growth apparatus (the seed crystal holder is above the crucible), and seal the growth apparatus.

[0053] The equipment was started up and the pressure was reduced from 950 mbar to 0 for leak detection.

[0054] After the leak detection was completed, the process was started, the power was increased to raise the temperature, 800 mbar of argon gas was supplied to maintain the pressure at 500 mbar, and then the pressure was reduced to about 1050℃, the argon gas was reduced to 50 mbar, and the pressure in the growth device was reduced to 4 mbar.

[0055] Oxidation stage: O2 is introduced into the crucible at a flow rate of 200 sccm, a pressure of 400 Pa, a temperature of 1050℃, and a duration of 60 min. O2 selectively oxidizes the BPD outcrop to form a SiO2 layer.

[0056] Reduction stage: Switch to H2 atmosphere, flow rate 500 sccm, pressure 600 Pa, temperature 950±20℃, duration 20 min. H2 reduces SiO2 to volatile SiO and desorbs it, leaving atomic-level V-shaped pits at the BPD outcrop.

[0057] The above oxidation-reduction process is repeated twice.

[0058] Final reduction: After the oxidation-reduction cycle is completed, H2 (flow rate of 600 sccm, which is 120% of the H2 flow rate of the previous reduction treatment) is introduced at a temperature of 950℃ and a pressure of 600 Pa to perform a final reduction treatment on the surface of the seed crystal for 15 min to ensure that the SiO2 on the surface is completely desorbed and to obtain the pretreated seed crystal.

[0059] (2) Asymmetric material loading Asymmetric loading is performed at the bottom of the graphite crucible: Central region: The radius is 49% of the total radius of the silicon carbide polycrystalline material filling area. It is filled with fine SiC polycrystalline material with a particle size of 100~300μm and the filling height is 90% of the crucible height.

[0060] Edge region: The radius is 51% of the entire silicon carbide polycrystalline material filling area, filled with coarse SiC polycrystalline material with a particle size of 800~1500μm, and the filling height is 60% of the crucible height.

[0061] The packing density ratio between the central area and the edge area is 1:0.5.

[0062] (3) Set up a porous graphite plate A porous graphite plate is placed between the raw material and the seed crystal. The porosity of the central region of the porous graphite plate is 0.60 and the pore size is 400 μm; the porosity of the edge region is 0.15 and the pore size is 10 μm; a transition zone with a radial width of 12 mm is set between the central region and the edge region, and the porosity of the transition zone drops sharply from 0.60 to 0.15 from the center to the periphery.

[0063] (4) Crystal growth The pretreated seed crystal is fixed on the top of the crucible, and the crucible is sealed. The heating system is turned on, and the temperature in the raw material zone reaches 2200~2300℃, and the temperature at the seed crystal reaches 1800~2000℃, to carry out PVT crystal growth, which takes about 100 hours.

[0064] (5) Gradient cooling After crystal growth is completed, the power of the induction coil is reduced step by step by partitioning the coil. First, the power output of the edge induction coil is reduced so that the edge area of ​​the crucible is cooled first. The cooling rate difference between the edge area and the center area is 8℃ / min until it is cooled to room temperature, and the finished crystal is obtained.

[0065] Example 2 The difference between this embodiment and Embodiment 1 is that: In step (1), the oxidation temperature is 980℃, the oxidation time is 45min, and the O2 flow rate is 150sccm; the reduction temperature is 930℃, the reduction time is 15min, and the H2 flow rate is 300sccm; the oxidation-reduction cycle is 3 times; and the final reduction H2 flow rate is 360sccm.

[0066] In step (2), the filling height of the central region is 85% of the crucible height, the filling height of the edge region is 50% of the crucible height, and the filling density ratio of the central region to the edge region is 1:0.4.

[0067] In step (5), the gradient cooling rate difference is 5℃ / min.

[0068] Example 3 The difference between this embodiment and Embodiment 1 is that: In step (1), the oxidation temperature is 1100℃, the oxidation time is 80min, and the O2 flow rate is 250sccm; the reduction temperature is 970℃, the reduction time is 25min, and the H2 flow rate is 800sccm; the final reduction H2 flow rate is 960sccm.

[0069] In step (2), the filling height of the central region is 95% of the crucible height, the filling height of the edge region is 65% of the crucible height, and the filling density ratio of the central region to the edge region is 1:0.6.

[0070] In step (5), gradient cooling is achieved through a phase change thermal storage ring: a phase change thermal storage type central insulation ring is set above the central region of the seed crystal. This insulation ring consists of a graphite sealed cavity (wall thickness 2mm) and an internally filled SiC particle-graphite matrix composite phase change thermal storage material. The ring structure has an inner diameter of 60mm, an outer diameter of 160mm, a thickness of 12mm, and a phase change temperature of 1950℃. During the cooling process, the phase change thermal storage ring releases latent heat, delaying the cooling rate of the central region by about 40%. There is no thermal storage compensation in the edge region, and the cooling rate remains unchanged. The difference in cooling rate between the edge and the center is 6℃ / min.

[0071] It should be noted that although specific examples are not given for every range of values ​​for the process parameters in this invention, the endpoints of each range have clear technical critical significance. Those skilled in the art, upon reading this specification, can determine the specific values ​​within each parameter range without any inventive effort.

[0072] Comparative Example 1 (lacking preprocessing) The difference between this comparative example and Example 1 is that the pretreatment and final reduction of the seed crystal surface oxidation-reduction cycle in step (1) are not performed. Instead, the 4H-SiC seed crystal without any surface pretreatment is directly fixed on the top of the crucible, and crystal growth is carried out according to the same asymmetric packing structure (fine center / coarse periphery), porous graphite plate, PVT growth conditions and gradient cooling as in Example 1.

[0073] Comparative Example 2 (lacking asymmetric loading) The difference between this comparative example and Example 1 is that: in step (2), the asymmetric packing structure is not used, but SiC polycrystalline material with a particle size of 100~300μm is uniformly packed at the bottom of the crucible (uniform packing, no center / peripheral partitioning), and the remaining steps (pretreatment, porous graphite plate, PVT growth conditions, gradient cooling) are the same as in Example 1.

[0074] Comparative Example 3 (lacking pretreatment + asymmetric loading, neither of which was present) The difference between this comparative example and Example 1 is that the seed crystal surface pretreatment in step (1) is not performed, and the asymmetric packing structure (uniformly packed with 100~300μm SiC polycrystalline material) is not used in step (2). Only porous graphite plates, PVT growth and gradient cooling are used. The rest is the same as in Example 1. This comparative example is equivalent to the baseline blank group.

[0075] Comparative Example 4 (Using homogeneous oxidation instead of selective oxidation) The difference between this comparative example and Example 1 is that: in step (1), pulsed O2 selective oxidation is not used, but a conventional uniform oxidation pretreatment method in the prior art is used: O2 (200 sccm) is continuously introduced at 400 Pa and 1050 °C for 60 min for oxidation treatment (without reduction-desorption cycle) to form a uniform oxide layer on the entire seed crystal surface; then the oxide layer is removed by one-time reduction with H2 (600 sccm) without cycle marking. The remaining steps are the same as in Example 1.

[0076] Comparative Example 5 (Number of preprocessing cycles exceeds the preferred range) The difference between this comparative example and Example 1 is that the number of oxidation-reduction cycles in step (1) is 5, while the rest is the same as in Example 1.

[0077] Comparative Example 6 (Prior Art Method) This comparative example follows the method of Example 1 in Chinese Invention Patent CN113089087B for crystal growth: a three-layer loading structure is adopted, wherein the upper layer has a central part of large-particle silicon carbide polycrystalline material (particle size 400~600μm) and a peripheral part of small-particle silicon carbide polycrystalline material (particle size 50~100μm), and the middle layer has a central part of large-pore porous graphite (pore size 700~900μm) and a peripheral part of small-pore porous graphite (pore size 80~150μm). No seed crystal surface pretreatment is performed, and no gradient cooling is performed.

[0078] Comparative Example 7 (lacking porous graphite plate) The difference between this comparative example and Example 1 is that a porous graphite plate is not set in step (3), while the rest is the same as Example 1.

[0079] Comparative Example 8 (lacking gradient cooling) The difference between this comparative example and Example 1 is that: after crystal growth in step (5), no gradient cooling is performed, and the crystal is directly cooled to room temperature naturally. The rest is the same as in Example 1.

[0080] Experimental Example 1: BPD Density Distribution Test The silicon carbide crystals grown in Examples 1-3 and Comparative Examples 1-8 were sliced ​​along the growth direction, and each wafer was subjected to KOH melt etching (530℃, 10 min). The BPD density was counted under an optical microscope (shell-shaped etch pits were identified as BPDs). The results are shown in Table 1.

[0081] Table 1. Radial BPD density distribution of crystals under different processing conditions (unit: cells / cm²)

[0082] Results analysis: By comparing the results of the embodiments and comparative examples, it is demonstrated that each step of the present invention is necessary and that the preferred parameters are reasonable. Embodiments 1-3, employing the complete technical solution of the present invention, achieve the best results and are superior to the comparative examples, with the BPD density in the central region all reduced to 150 / cm³. 2 Below, the edge locking ring BPD density is significantly lower than that of the comparative examples, ≥2500 / cm². Furthermore, comparative examples 1-8 fail to achieve the technical effect described in this case.

[0083] As can be seen from the data in Table 1, the steps of the present invention have a synergistic effect, and multiple stages form a complete technical solution. None of them can be omitted, and together they achieve spatial locking of BPD.

[0084] Experimental Example 2: Effects of Oxidation Temperature and Time on Oxidation Selectivity (1) With the oxidation time fixed at 60 min, the effect of different oxidation temperatures on oxidation selectivity was investigated. Pretreatment of 4H-SiC seed crystals was carried out under different oxidation temperatures (oxidation time was fixed at 60 min, and other conditions were the same as in Example 1). The SiO2 layer thickness ratio at the BPD outcrop and the non-defect region was compared, and the results are shown in the table below.

[0085] Table 2-1 Effect of oxidation temperature on oxidation selectivity (oxidation time: 60 min)

[0086] (2) With the oxidation temperature fixed at 1050℃, the effect of different oxidation times on oxidation selectivity was investigated. Pretreatment of 4H-SiC seed crystals under different oxidation time conditions (oxidation temperature fixed at 1050℃, other conditions the same as in Example 1) was carried out, and the SiO2 layer thickness ratio at the BPD outcrop and the non-defect region was compared. The results are shown in the table below.

[0087] Table 2-2 Effect of oxidation time on oxidation selectivity (oxidation temperature: 1050℃)

[0088] Results Analysis: The data above shows that 60 min is the optimal oxidation time. At this time, the SiO2 thickness at the BPD outcrop is 3.11 times that of the non-defect region, exhibiting optimal selectivity. When the oxidation time is less than 45 min, the oxide layer is too thin, and the channel diffusion effect of the BPD core is not fully activated, resulting in insufficient selectivity (thickness ratio < 2.3). When the oxidation time is greater than 90 min, the oxidation reaction enters the diffusion control region, and the difference in interface reaction between the BPD core and the non-defect region is smoothed by the diffusion process, leading to a significant decrease in selectivity (thickness ratio < 2.6). When the oxidation time reaches 240 min, selectivity is almost completely lost (thickness ratio < 2.0).

[0089] Therefore, the oxidation time of the present invention can be selected as 45-90 min, more preferably 60 min, to achieve optimal selective oxidation marking at the BPD outcrop.

[0090] Experimental Example 3: The Effect of Pretreatment Cycle Number on BPD Labeling Performance The 4H-SiC seed crystals were subjected to oxidation-reduction cycle pretreatment at different cycle numbers (0, 1, 2, 3, 4, 5 times) (other conditions were the same as in Example 1), followed by KOH etching. The pit depth (nm) at the BPD outcrop was counted under an optical microscope, and the pit morphology was recorded. The results are shown in the table below.

[0091] Table 3. Relationship between the number of preprocessing cycles and the effect of BPD labeling.

[0092] Results analysis: The data above shows that after two cycles, the pit depth is about 120nm, which is sufficient to pin dislocations; after more than three cycles, the bottom of the pit tends to be flat, and the pinning force decreases.

[0093] As the number of cycles increased from 1 to 3, the pit depth increased to 180-220 nm while maintaining a steep V-shaped edge, indicating that 2-3 cycles can achieve sufficient BPD labeling.

[0094] Two cycles are the optimal number of cycles. At this point, a steep V-shaped pit forms at the BPD outcrop, with a depth of 120-160 nm, steep sidewalls, and a sharp bottom. This morphology provides optimal physical protection against basal slippage of the BPD.

[0095] Specifically, when the number of cycles is 1, the pit depth is insufficient (50-80nm), presenting a shallow U-shape, and the pinning effect is limited.

[0096] When the number of cycles is ≥3, the pits are further deepened (180-280nm), but the shape gradually degenerates into an inverted trapezoid or a wide U-shape, the bottom widens, the sidewalls become gentler, and microcracks and secondary etch pits appear at the bottom of the pits, and the pinning effect decreases.

[0097] When the number of cycles is ≥4, excessively deep pits not only lose their pinning ability, but may also become new stress concentration sources and defect introduction points.

[0098] Therefore, the preferred number of oxidation-reduction cycle pretreatment cycles specified in this invention is 2 to 3. Two cycles are optimal; fewer than two cycles result in insufficient marking and failure to form effective BPD pinning structures; more than two cycles lead to excessively deep pits, morphological degradation, and a decrease in marking effectiveness.

[0099] Experiment Example 4: Verification of the Effects of Various Parameters of the Filling Structure This experimental example examines the effects of various parameters in the packing structure (particle size partitioning, packing height, density ratio, and radius percentage) on the difference in growth rate between the central and peripheral regions of the crystal. The radius percentage is the percentage of the radius of the central or peripheral region relative to the total radius of the silicon carbide polycrystalline material packing area.

[0100] All samples used the same pretreatment conditions, porous graphite plates, and PVT growth parameters as in Example 1, differing only in the raw material packing structure at the bottom of the crucible. The following different packing configurations were used: Configuration A (the conventional solution in the prior art): The central region is filled with large-particle silicon carbide polycrystalline material (particle size 300~600μm), and the edge region is filled with small-particle silicon carbide polycrystalline material (particle size 50~100μm). The filling height is uniform, the density ratio is 1:1, the radius of the central region is 40% of the radius of the entire silicon carbide polycrystalline material filling area, and the radius of the edge region is 60% of the radius of the entire silicon carbide polycrystalline material filling area.

[0101] Configuration B (Embodiment 1 of the present invention): Asymmetric filling: The radius of the central region is 49% of the radius of the entire silicon carbide polycrystalline material filling area; the filling material consists of fine SiC polycrystalline material with a particle size of 100~300μm, and the filling height is 90% of the crucible height.

[0102] Edge region: The radius is 51% of the entire silicon carbide polycrystalline material filling area, filled with coarse SiC polycrystalline material with a particle size of 800~1500μm, and the filling height is 60% of the crucible height.

[0103] The packing density ratio between the central area and the edge area is 1:0.5.

[0104] Configuration C (lacking particle size partitioning): Both the central and edge regions are filled with silicon carbide polycrystalline material with a particle size of 100~300μm; the other parameters are the same as those in Configuration B.

[0105] Configuration D (lacking filling height difference): The filling height of both the central and edge areas is 75% of the crucible height; all other parameters are the same as Configuration B.

[0106] Configuration E (lacking density ratio difference): density ratio 1:1 (uniform); all other parameters are the same as configuration B.

[0107] Configuration F (lacking reasonable radius proportion): The radius of the central region is 70% of the radius of the entire silicon carbide polycrystalline material filling area, and the radius of the edge region is 30% of the radius of the entire silicon carbide polycrystalline material filling area. The remaining parameters are the same as those of Configuration B.

[0108] The growth rates of the central region (r=0~80mm) and the edge region (r=95~105mm) of the crystal were measured for each configuration, and the center / edge rate ratio (growth rate of the central region / growth rate of the edge region) and the BPD density of the central region were calculated. The results are shown in Table 4.

[0109] Table 4. Effects of different packing configurations on growth rate and BPD density

[0110] Results analysis: As shown in Table 4, configuration B (the embodiment of the present invention) has the best effect, with a center / edge rate ratio significantly higher than all other configurations, and the lowest BPD density in the center region.

[0111] Configuration B (the present invention) achieves a maximum center / edge rate ratio of 2.80 through the synergistic effect of four factors: particle size partitioning (fine / coarse), filling height difference (high / low), density ratio difference (high / low), and a reasonable radius ratio (49% / 51%). This effectively drives and locks BPDs into the edge ring, reducing the BPD density in the central region to 0~150 / cm³. 2 .

[0112] The absence of any one factor will lead to a significant decrease in the rate ratio and a substantial increase in the central BPD density, proving that the synergistic effect among the process parameters in this invention is irreplaceable.

[0113] The above results demonstrate that the four packing structure parameters in this invention—particle size partitioning, packing height difference, density ratio difference, and radius ratio—work together to achieve the technical effect that the growth rate in the central region is greater than that in the edge region. The absence of any one parameter will decrease the center / edge rate ratio and weaken the BPD locking effect. This confirms that the parameters in the packing structure of this invention are not simply additive but exhibit significant synergistic effects.

[0114] Experimental Example 5: Verification of the Effects of Various Parameters on Porous Graphite Plates This experiment investigated the effects of different parameter configurations (porosity, pore size) on the growth rate of the crystal edge region in a porous graphite plate. All samples used the same pretreatment conditions, packing structure, and PVT growth parameters as in Example 1, with only the parameters of the porous graphite plate differing.

[0115] Configure the following different porous graphite plate configurations: Configuration A (control group without porous graphite plate): No porous graphite plate is set, and the rest is the same as in Example 1.

[0116] Configuration B (Complete Solution of the Invention): The porosity of the central region of the porous graphite plate is 0.60 and the pore size is 400 μm, while the porosity of the edge region is 0.15 and the pore size is 10 μm.

[0117] Configuration C (lacking porosity differences): The pore size in the central region is 400 μm, and the pore size in the edge region is 10 μm. The only difference is the pore size; the other parameters are the same as those in Configuration B.

[0118] Configuration D (lacking porosity difference): The porosity of the central region is 0.60, and the porosity of the edge region is 0.15 (porosity is the only difference). All other parameters are the same as Configuration B.

[0119] The growth rates of the crystal's central region (r=0~80mm) and edge region (r=95~105mm) were measured for each configuration, and the central / edge rate ratio was calculated.

[0120] The results are shown in Table 5.

[0121] Table 5. Influence of different parameter configurations on the edge / center rate ratio of porous graphite plates.

[0122] Results analysis: As shown in Table 5, configuration B (the complete scheme of the present invention) has the lowest edge / center rate ratio, indicating that it has the best effect on suppressing the growth rate of the edge region.

[0123] Configuration B (the complete solution of this invention) uses a dual asymmetric design of porosity difference (0.60 / 0.15) + pore size difference (400μm / 10μm) to suppress the growth rate of the edge region to 0.09mm / h, and the center / edge rate ratio reaches 3.11, which is significantly better than 1.12 without graphite plate, 1.62 without porosity difference, and 1.86 without pore size difference.

[0124] Experiments have shown that there is a significant synergistic effect between porosity difference and pore size difference: when both are acted on individually, the growth rate of the edge region is 0.16 mm / h and 0.14 mm / h, respectively; when both are acted on simultaneously, the growth rate of the edge region drops to 0.09 mm / h, indicating that the dual asymmetric design has a super-additive effect on the inhibition of edge growth.

[0125] The above results indicate that two parameters of porous graphite plates, porosity difference and pore size difference, affect the gas phase distribution of the center-conducting and edge-blocking processes, respectively. Porosity difference enables the high-pore center to conduct while the low-pore edge blocks it, while pore size difference affects the gas phase permeability, making the large-pore center unobstructed and the small-pore edge blocked.

[0126] The two work synergistically and complement each other, resulting in a growth rate in the edge region that is only 20% to 30% of that in the center, thus achieving the best gas phase barrier effect.

[0127] The above provides a detailed description of a silicon carbide crystal growth method provided by the present invention. Specific examples have been used to illustrate the principle and implementation of the present invention. The above description of the embodiments is only for the purpose of helping to understand the method and core idea of ​​the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made to the present invention without departing from the principle of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A method for growing silicon carbide crystals, characterized in that, Includes the following steps: Provide a crystal growth apparatus and silicon carbide seed crystals, install the silicon carbide seed crystals on the seed crystal holder, and perform a preset number of oxidation-reduction pretreatment cycles; The oxidation-reduction pretreatment includes oxidizing the growth surface of the silicon carbide seed crystal with an oxidizing gas to obtain an oxide layer, and then reducing the oxide layer with a reducing gas and desorbing it. After the loop is completed, a pit is generated at the basis vector plane dislocation outcrop on the surface of the seed crystal, and a pre-treated seed crystal is obtained. The bottom of the crucible in the crystal growth apparatus is filled with silicon carbide polycrystalline material using an asymmetric structure. Crystal growth is performed using the physical vapor transport method, ensuring that the growth rate in the central region of the crystal is greater than that in the edge region, thus obtaining the finished product.

2. The method according to claim 1, characterized in that, The preset number of times is 2 to 3.

3. The method according to claim 1, characterized in that, In the oxidation process, the pressure is 100~1000Pa, the temperature is 900~1200℃, the time is 30~180min, and the flow rate of the oxidizing gas is 50~500sccm.

4. The method according to claim 1, characterized in that, In the reduction process, the pressure is 400~800Pa, the temperature is 800~1100℃, the time is 10~60min, and the flow rate of the reducing gas is 100~1000sccm.

5. The method according to claim 1, characterized in that, After performing a preset number of redox pretreatments, the process also includes: The seed crystal surface is subjected to a final reduction treatment using a reducing gas, wherein the flow rate of the reducing gas in the final reduction treatment is 120% of the flow rate of the reducing gas in the previous reduction treatment.

6. The method according to claim 1, characterized in that, The asymmetric structure is as follows: the central region is filled with fine-grained silicon carbide polycrystalline material with a particle size of 100~300μm, and the edge region is filled with coarse-grained silicon carbide polycrystalline material with a particle size of 800~1500μm.

7. The method according to claim 6, characterized in that, The filling height of the silicon carbide polycrystalline material in the central region is 85% to 95% of the crucible height, and the filling height of the silicon carbide polycrystalline material in the edge region is 50% to 65% of the crucible height; the filling density ratio between the central region and the edge region is 1:0.4 to 0.

6.

8. The method according to claim 7, characterized in that, The radius of the central region is 40% to 50% of the radius of the entire silicon carbide polycrystalline material filling area.

9. The method according to claim 1, characterized in that, After the crystal growth is completed, a gradient cooling process is performed so that the cooling rate of the crystal edge region is faster than that of the central region. The difference in cooling rate between the edge region and the central region is 5~10℃ / min.

10. The method according to claim 1, characterized in that, A porous graphite plate is disposed between the raw material and the seed crystal, the porous graphite plate having different porosities and / or pore sizes in the radial direction. The porosity of the central region of the porous graphite plate is higher than that of the edge region, and the pore diameter of the central region is larger than that of the edge region.

11. The method according to claim 10, characterized in that, The porous graphite plate has a central region porosity of 0.55~0.65 and a pore size of 300~500μm, and an edge region porosity of 0.10~0.18 and a pore size of 5~20μm.

12. A silicon carbide crystal growth apparatus for use in the method according to any one of claims 1 to 11, characterized in that, include: A crucible used to hold silicon carbide raw materials; A seed crystal holder is placed above the crucible to fix the pretreated seed crystal. An atmosphere supply component is used to provide oxidizing and reducing gases to achieve redox pretreatment of the growth surface of silicon carbide seed crystals. A porous graphite plate is disposed between the crucible and the pretreated seed crystal, wherein the porosity of the central region of the porous graphite plate is greater than that of the edge region. A heating device is used to heat the interior of the silicon carbide crystal growth apparatus.

13. The apparatus according to claim 12, characterized in that, The porous graphite plate has a central region porosity of 0.55~0.65 and a pore size of 300~500μm, and an edge region porosity of 0.10~0.18 and a pore size of 5~20μm.

14. The apparatus according to claim 12, characterized in that, It also includes a cooling mechanism, which is used to ensure that the cooling rate of the crystal edge region is greater than that of the central region during the cooling stage.

Images