Method for carrying out crystal growth by utilizing silicon carbide crystal growth excess material

Abstract

The invention discloses a method for carrying out crystal growth by utilizing silicon carbide crystal growth excess materials, and particularly relates to the technical field of silicon carbide material reutilization and crystal preparation. A reutilization unit body with crystal orientation consistency is obtained; based on the crystal orientation parameters and the thermal conductivity characteristics of the unit bodies, constructing a crystal guide model suitable for the high-temperature thermal field; arranging the plurality of unit body arrays in a multi-stage temperature zone set in a crystal growth furnace to form a quasi-oriented seed crystal array; under the conditions of temperature control and atmosphere adjustment, directional crystallization and epitaxial growth of a silicon carbide gas-phase precursor generated by sublimation on the surface of the seed crystal array are realized; the crystallization behavior of each unit body in the crystal growth process is monitored, a feedback regulation and control mechanism is constructed based on stress deviation and lattice orientation change, and the thermal field gradient and the raw material supply rate are adjusted in real time; the method realizes high-value recycling of the silicon carbide excess material, improves the crystal orientation consistency and structural integrity of crystal growth, and has the advantages of stable growth quality, high resource utilization rate and the like.

Description

Technical Field

[0001] This invention relates to the field of silicon carbide material recycling and crystal preparation technology, specifically to a method for crystal growth using silicon carbide grain growth surplus. Background Technology

[0002] Currently, silicon carbide (SiC) crystal materials are widely used in power electronics, high-frequency devices, and new energy vehicles. With the continuous growth in demand for large-size, high-quality silicon carbide single crystals, the large amount of scrap and defective crystals generated during crystal growth will lead to resource waste and increased costs if they cannot be efficiently recycled.

[0003] Existing methods for recycling silicon carbide crystals mainly include physical crushing for use as abrasives or high-temperature gasification to recover carbon. These methods are energy-intensive, have low conversion rates, and make it difficult to achieve complete regeneration of the silicon carbide crystal structure. Furthermore, silicon carbide crystal growth requires strict control over the structure of the seed crystal, thermal field distribution, and impurities. Due to crystal defects and impurity accumulation in the remaining material, it is difficult to directly use it as a seed crystal or raw material for recycling. Summary of the Invention

[0004] The purpose of this invention is to provide a method for crystal growth using silicon carbide grain growth residue, in order to overcome the shortcomings of the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for crystal growth using silicon carbide grain growth residue, comprising:

[0006] Pre-treatment of silicon carbide crystal growth waste includes removing surface oxide layer, impurity residue and structural defect areas to obtain reusable unit cells with crystal orientation consistency;

[0007] Based on the crystal orientation parameters and thermal conductivity of the reusable unit, a crystal guiding model suitable for high-temperature thermal fields is constructed.

[0008] Multiple reusable unit cells are arranged in arrays within a multi-level temperature zone set in a crystal growth furnace to form a quasi-oriented seed array.

[0009] Under controlled temperature and atmosphere conditions, the quasi-oriented seed array guides the silicon carbide vapor precursor to undergo directional crystallization, thereby achieving continuous growth of high-quality crystals.

[0010] The crystallization behavior of each reused unit cell during crystal growth is monitored, and a feedback control mechanism is constructed based on stress shift and lattice orientation changes to adjust the thermal gradient and raw material supply rate in real time.

[0011] Preferably, the pretreatment of the silicon carbide grain growth residue includes:

[0012] Selective mechanical grinding and chemical etching are combined to remove the surface oxide layer and expose the intrinsic crystal structure of the silicon carbide crystal growth residue.

[0013] It also removes metallic and non-metallic impurities, including oxygen, sodium, and iron, that remain on the crystal surface and in microcracks and affect the purity of subsequent crystallization.

[0014] Crystal orientation consistency is tested and graded to select surplus unit cells that meet the specified lattice orientation error range;

[0015] The edges of the remaining material unit are adjusted and cut to construct a reusable seed array module for crystal guidance control in the subsequent crystal growth process.

[0016] Preferably, the construction of a crystal guiding model suitable for high-temperature thermal fields includes:

[0017] Collect crystal orientation parameters, thermal conductivity, thermal expansion coefficient and crystal plane orientation information of the reuse unit cells to construct a multidimensional crystal orientation-thermal conductivity coupled database;

[0018] Based on the database, the growth trend of crystals under different thermal gradients and crystal orientation combinations is simulated and modeled using the finite element thermal field simulation method, and the thermal field stability index and the guiding gain factor are extracted.

[0019] By combining the coupling evolution relationship between the heat flux density field and the crystal orientation deviation angle, a crystal guidance propagation path function is established to optimize the arrangement and initial orientation of the seed array in the thermal field.

[0020] Preferably, the arrangement of multiple reusable unit cells within a multi-level temperature zone within the crystal growth furnace to form a quasi-oriented seed array includes:

[0021] Based on the radial and axial thermal field distribution patterns within the crystal growth furnace, the furnace cavity is divided into multiple temperature gradient zones, and a temperature control range and heat flow guidance direction are set for each zone.

[0022] Based on the crystal orientation parameters, thermal conductivity, and orientation factor of the reused unit cells, multiple unit cells are classified and graded according to the intensity of their orientation effect.

[0023] Units with high orientation strength are preferentially arranged in the critical hot region with a large temperature change rate, while units with low orientation strength are arranged in the thermal stable region to construct a multi-level linked quasi-oriented seed array.

[0024] Preferably, under controlled temperature and atmosphere conditions, the quasi-oriented seed array guides the silicon carbide vapor precursor to undergo directional crystallization, comprising:

[0025] Adjusting the temperature gradient of each temperature zone in the crystal growth furnace forms a heat flow channel that steadily decreases from the high-temperature source area to the crystal growth area, thus constructing a stable sublimation transport path;

[0026] High-purity argon or hydrogen-argon mixture is introduced into the furnace, and the gas pressure and flow rate parameters are controlled so that the sublimated silicon carbide gas phase precursor is transported to the surface of the quasi-oriented seed array along the heat flow direction.

[0027] Adjusting the local micro-atmosphere composition and pressure distribution in the crystal growth region allows the gaseous precursor to preferentially deposit on the intrinsic crystal surface exposed by the seed array, promoting directional nucleation and epitaxial expansion at the gas-solid interface.

[0028] During crystal growth, the crystal growth rate and crystal surface flatness are monitored in real time, and the local thermal gradient and airflow guidance are adjusted based on the feedback of the seed normal direction.

[0029] Preferably, the adjustment of the local thermal field gradient and airflow guidance based on the feedback of the seed crystal normal direction includes:

[0030] During crystal growth, the spatial variation and orientation offset of the normal direction of the crystal plane on the seed array surface are monitored in real time.

[0031] The monitoring data is compared with the standard model of the target crystal orientation, and the correlation factor between the crystal orientation deviation angle and thermal disturbance is calculated to form a dynamic offset spectrum for regulation.

[0032] Based on the offset spectrum, the power distribution of the heating source, the angle of the heat reflection structure, and the gas flow path are adjusted in the local area;

[0033] Based on the crystal orientation recovery rate and the crystallization interface reconstruction trend, the control strategy is updated to achieve adaptive thermal-fluid field coordinated feedback regulation during crystal growth.

[0034] Preferably, the monitoring of the crystallization behavior of each reused unit cell during crystal growth, and the construction of a feedback control mechanism based on stress shift and lattice orientation changes, includes:

[0035] The crystallization rate, crystal plane migration state, and micro-defect evolution of each reuse unit are monitored in real time to obtain its structural evolution data during the growth process;

[0036] Detect the changes in lattice stress distribution and lattice orientation, and establish a dynamic response model of thermal stress crystal orientation for each unit cell;

[0037] Based on the aforementioned thermal stress crystal orientation dynamic response model, a feedback control algorithm is constructed to quantitatively analyze the stress shift trend and lattice rotation behavior, and generate control command signals.

[0038] Based on the control signal, the local thermal field gradient distribution and the gas phase precursor supply rate are dynamically adjusted.

[0039] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0040] 1. This invention achieves directional, stable, and continuous growth of silicon carbide crystals by constructing a quasi-oriented seed array and introducing multi-level temperature zone thermal field control, precise transport of gas-phase precursors, and an in-situ crystal orientation monitoring and feedback mechanism. Compared with traditional crystal growth methods, this invention significantly improves the crystal orientation retention capability and thermal field response efficiency during crystal growth, effectively suppresses interface instability, crystal orientation drift, and structural defects, and ensures the crystallization consistency and growth uniformity of the crystal under high-temperature conditions.

[0041] 2. This invention constructs a feedback control mechanism that links multiple physical fields (thermo-mass-mechanical) by dynamically identifying the stress evolution and lattice orientation changes of each reuse unit. This mechanism can adjust the local thermal gradient and precursor supply rate in real time, achieving intelligent control of crystal growth behavior. This approach significantly reduces material costs and energy consumption while improving the efficiency of waste material reuse. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0043] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] For examples, please refer to Figure 1 As shown, the method for crystal growth using silicon carbide grain growth residue described in this embodiment includes:

[0046] Pre-treatment of silicon carbide crystal growth waste includes removing surface oxide layer, impurity residue and structural defect areas to obtain reusable unit cells with crystal orientation consistency;

[0047] Based on the crystal orientation parameters and thermal conductivity of the reusable unit, a crystal guiding model suitable for high-temperature thermal fields is constructed.

[0048] Multiple reusable unit cells are arranged in arrays within a multi-level temperature zone set in a crystal growth furnace to form a quasi-oriented seed array.

[0049] Under controlled temperature and atmosphere conditions, the quasi-oriented seed array guides the silicon carbide vapor precursor to undergo directional crystallization, thereby achieving continuous growth of high-quality crystals.

[0050] The crystallization behavior of each reused unit cell during crystal growth is monitored, and a feedback control mechanism is constructed based on stress shift and lattice orientation changes to adjust the thermal gradient and raw material supply rate in real time.

[0051] In the crystal growth method of the present invention, in order to achieve efficient reuse of silicon carbide crystal growth waste, it is necessary to perform systematic pretreatment operations to obtain reusable unit cells with requirements of crystal orientation consistency, structural integrity and high purity, so as to ensure their guiding function and crystallization quality in the subsequent crystal growth process.

[0052] First, the silicon carbide crystal growth residue undergoes preliminary physical treatment, employing precision mechanical grinding to selectively polish its surface, removing the oxide layer, rough areas on the crystal surface, and mechanically damaged layers. During the process, diamond abrasives with a particle size between 5 and 15 micrometers are used, and multi-axis synchronous grinding is performed under low-pressure conditions to prevent the formation of new microcracks or structural disturbances. The mechanical grinding preferably employs a wet polishing method, coupled with an appropriate coolant to suppress localized overheating.

[0053] Subsequently, the mechanically ground silicon carbide residue sample is immersed in an alkaline or acidic etching solution of a specific concentration, preferably a mixture of hydrofluoric acid and nitric acid, with a volume ratio controlled between 1:3 and 1:5. The chemical etching reaction is carried out at a temperature controlled between 60 and 80 degrees Celsius. This process effectively dissolves and removes the residual silicon oxide layer and further strips away the heterogeneous layer on the crystal surface, thereby exposing the intrinsic crystal structure of silicon carbide and providing a reliable data basis for subsequent crystal orientation detection.

[0054] After removing the oxide layer, a plasma-assisted cleaning process is introduced to further address the easily residual metallic impurities and impurities adsorbed within microcracks in silicon carbide crystals. A low-temperature argon or oxygen plasma source is preferably used to bombard the crystal surface under vacuum conditions for 3 to 7 minutes, with a power range of 100 to 200 watts. This method can effectively remove impurities such as metal ions, oxygen, sodium, and iron adsorbed on the crystal surface and in microcracks without damaging the crystal structure, improving the purity of the material and reducing its impact on the crystallization quality during subsequent crystal growth.

[0055] If necessary, the treated sample can be further ultrasonically cleaned with deionized water to thoroughly remove residual corrosion products and powder contaminants, ensuring that the crystal surface is uniform and clean.

[0056] To ensure that the reusable unit cells have a uniform crystal orientation and can serve as stable seed sources, the pretreated silicon carbide scrap undergoes lattice orientation detection and hierarchical management. Specific detection methods include Raman spectroscopy and X-ray diffraction (XRD) analysis of the sample orientation. The crystal orientation error is calculated by comparing the deviation of the Raman scattering peak shift with standard crystal plane data. Unit cells with crystal orientation errors controlled within ±0.5 degrees are preferred as qualified materials.

[0057] In X-ray diffraction testing, the crystal plane normal direction is determined by precisely measuring the diffraction angle change of the sample along the crystal plane (e.g., the 0001 plane). Samples meeting set standards are numbered and classified, and grouped according to their crystal orientation consistency to facilitate the subsequent construction of standardized seed array modules.

[0058] Based on the crystal orientation screening results, the remaining material unit cells undergo edge fine-tuning and cutting. Specifically, a laser micro-cutting system or high-precision wire cutting equipment is used to trim the edges of the unit cells, ensuring they have standard geometric dimensions (e.g., side length controlled between 10 and 20 mm) and precise alignment angles. This step helps to integrate multiple remaining material unit cells into an array-style arrangement module, enhancing its ability to guide the growth direction of new crystals during the crystal growth process.

[0059] During the construction of the seed array, a precision alignment platform is used to install each unit cell into a predetermined position inside the crystal growth furnace, maintaining the consistency of the crystal plane orientation in three-dimensional space. This allows them to collectively serve as the crystal growth initiation interface, improving the overall uniformity and controllability of the crystal growth process. This array structure can form a stable multi-point guiding interface in the thermal field, helping to suppress crystal growth deviations and defect generation caused by thermal disturbances.

[0060] In this invention, to effectively guide newly grown crystals in a high-temperature thermal environment using recycled silicon carbide units, and to ensure uniform and stable crystal growth along a specific crystal orientation, a crystal guidance model needs to be constructed based on the physical properties of the crystal material itself. This model not only considers the crystal's geometry and lattice orientation but also fully integrates the material's thermal conductivity, thermal expansion characteristics, and heat flow distribution behavior in a high-temperature thermal field, thereby establishing a crystal orientation guidance system suitable for complex thermal environments.

[0061] First, parameters need to be collected from the pretreated silicon carbide reuse unit. The collected data includes, but is not limited to:

[0062] Crystal orientation parameters: The principal crystal plane orientations of silicon carbide unit cells, such as (0001) and (1-100) crystal plane orientations, are obtained by X-ray diffraction testing technology, and their spatial angles are calibrated.

[0063] Thermal conductivity: The thermal conductivity of the crystal along different crystal orientations was tested using the laser scintillation method, especially the trend of its thermal conductivity change under high temperature conditions (such as 1800℃).

[0064] Coefficient of thermal expansion: The linear expansion behavior of the crystal in each principal crystal direction was tested using a high-temperature differential thermal analyzer, and its coefficient of thermal expansion as a function of temperature was recorded.

[0065] Crystal plane orientation information: Using electron backscatter diffraction technology, the crystal lattice orientation pattern is obtained and three-dimensional crystal planes are reconstructed.

[0066] All the above parameters are indexed by the crystal sample number to construct a multidimensional crystal orientation-thermal conductivity coupling database. The database adopts a multi-field structure, with fields including crystal orientation number, direction cosine, thermal conductivity matrix, thermal expansion tensor, and crystal plane normal vector, etc., to support subsequent thermal field simulation input and parameter matching. This database not only provides high-precision basic data for the model, but also serves to select the unit cell combination that best matches the target thermal field.

[0067] Based on the aforementioned coupled database, the thermal field evolution behavior during crystal growth is numerically simulated using a finite element simulation platform (such as COMSOL Multiphysics or ANSYS). The simulation model includes the following sub-modules:

[0068] Thermal conduction module: Using the thermal conductivity tensor corresponding to each crystal orientation as input variables, it simulates the dynamic distribution of heat flux density under high temperature thermal field;

[0069] Thermal stress module: Combines the coefficient of thermal expansion and temperature gradient to evaluate the distribution trend of thermal stress field within the crystal;

[0070] Interface evolution module: Set interface migration criteria at the solid-liquid interface to simulate the growth rate and crystal orientation shift trend of crystals under the action of different heat flow directions;

[0071] Crystal orientation shift simulation module: Establishes an evolution function of the angle between the crystal growth interface normal and the crystal orientation, and simulates the behavior trajectory and recovery mechanism when the crystal deviates from the ideal crystal orientation.

[0072] During the simulation, multiple thermal gradient boundary conditions and multiple crystal orientation configurations were set to evaluate the stability indices of crystal growth (such as interface perturbation amplitude and crystal orientation offset) and the guidance gain factor. The guidance gain factor is defined as the ability of the seed unit to influence the growth direction of the nascent crystal, and can be quantified by the cosine similarity between the crystal interface normal vector and the crystal orientation consistency function.

[0073] For example, if the crystal interface normal vector is N and the target crystal orientation unit vector is G, then the guidance gain factor can be expressed as:

[0074] The orientation factor is equal to the dot product of vectors N and G divided by the product of the magnitudes of the two vectors. In other words, the orientation factor is equal to the cosine of the angle between N and G.

[0075] By extracting the distribution maps of the guiding factors under different arrangement methods, the geometric configuration and initial arrangement angle of the seed array are optimized to provide data support for subsequent optimization.

[0076] To further guide the spatial arrangement and guiding strategy of seed crystals in a thermal field, this invention establishes a crystal guiding propagation path function. This function comprehensively considers:

[0077] Distribution function of heat flux density field;

[0078] Evolution function of crystal orientation deviation angle with time and position;

[0079] The guiding weight distribution between the center and edge units of the seed array;

[0080] Propagation behavior of crystal interface disturbances caused by thermal disturbances.

[0081] The propagation path function uses three-dimensional spatial coordinates as independent variables and the direction of minimizing crystal orientation deviation angle as the objective value to construct a guiding path that dynamically tracks the crystal growth front. An optimal arrangement path model can be established using variational methods or multi-objective optimization functions, resulting in a thermal field adaptive adjustment scheme.

[0082] Furthermore, by combining the symmetry and arrangement density of the seed array, a cooperative guidance coefficient parameter is introduced to describe the ability of multiple units to cooperatively influence the orientation of new crystals in a thermal field, thereby achieving:

[0083] Improved stability along the long axis of the crystal;

[0084] Morphological control of the growth interface;

[0085] The multi-point guidance enhances the adaptability to thermal field disturbances.

[0086] Finally, the initial arrangement angle and position coordinates of the seed array are deduced based on the path function, thus achieving the unity of crystal orientation consistency and thermal field coordination.

[0087] To achieve stable growth of silicon carbide crystals under complex thermal conditions, this invention proposes a method for constructing a quasi-oriented seed array based on the coupled arrangement of reusable unit cell characteristics and thermal field distribution. This method identifies multi-level temperature gradient regions within the crystal growth furnace cavity and, combined with the crystal orientation, thermal conductivity, and guiding factor of the unit cells, optimizes the spatial distribution and arrangement of multiple reusable seed cells, thereby forming a thermally guided structural array that achieves highly consistent and stable crystal growth path control.

[0088] First, the thermal field inside the crystal growth furnace is modeled and measured to clarify the temperature gradient distribution in the radial and axial directions. This thermal field analysis is usually achieved using thermocouple arrays, infrared temperature measurement systems, or heat flow simulation software (such as ANSYS Thermal or Fluent).

[0089] Under typical operating conditions of a crystal growth furnace, its axial temperature variation usually exhibits a non-linear distribution, with the central high-temperature zone decreasing gradually upwards and downwards, and a radial temperature difference also existing from the furnace core to the furnace wall. Based on these distribution characteristics, the furnace cavity is divided into several temperature gradient sections along the crystal growth direction (usually perpendicular to the axial direction), such as a high-temperature gradient region, an intermediate transition region, and a low-temperature stable region.

[0090] The following parameters are set for each segment:

[0091] Temperature control range: for example, the high temperature zone is controlled between 2200℃ and 2400℃, the intermediate zone is between 1800℃ and 2100℃, and the low temperature zone is controlled below 1600℃;

[0092] Heat flow guidance direction: By adjusting the power of the heating element and the layout of the heat screen reflector, the heat flow is controlled to be concentrated and conducted along a preset direction, forming a thermal gradient environment that is conducive to the directional growth of crystals.

[0093] After division, these hot zones are used as the basic environmental structure for subsequent seed array layout, providing the spatial basis for matching the seed with the thermal field.

[0094] After completing the thermal field division, key parameters of the seed crystal unit are obtained and organized for reuse, including crystal orientation, thermal conductivity, and coefficient of thermal expansion. In this invention, a "guiding factor" is specifically introduced as a core indicator to evaluate the seed crystal's ability to control the crystal growth direction in the thermal field.

[0095] The guidance factor is calculated as follows:

[0096] Let the normal vector of the seed crystal plane be N, and the target growth crystal direction be G. The normalized guidance factor is defined as the cosine of the angle between the two, that is:

[0097] The guidance factor = N·G / (|N|×|G|), and its value ranges from -1 to +1.

[0098] When the guidance factor approaches +1, it indicates that the seed normal is highly consistent with the target crystal orientation, and the guidance capability is strong.

[0099] Furthermore, considering the anisotropic behavior of thermal conductivity in different crystal orientations, the ratio of thermal conductivity along the principal axis (e.g., the C-axis) to that along the transverse direction is used as the thermal conductivity uniformity coefficient to help determine the crystal's ability to guide heat flow. High thermal conductivity and principal orientation uniformity contribute to stable thermal field conduction, thereby improving the guiding ability.

[0100] Based on the above parameters, a multi-factor weighting model is used to comprehensively score the unit, and the units are sorted from high to low according to the guidance factor values, and divided into:

[0101] Primary seed crystal (strongly guided type);

[0102] Secondary seed crystal (medium-guided type);

[0103] Tertiary seed crystals (auxiliary or stabilizing type).

[0104] This classification provides a basis for subsequent hot zone matching and arrangement, ensuring that each unit operates in the optimal orientation state in its suitable thermal environment.

[0105] After determining the orientation level of each unit cell, and combining the results of the furnace cavity thermal field division, a quasi-oriented seed array is constructed. The arrangement rules are as follows:

[0106] First-order seed crystals with high guidance factors are preferentially arranged in the critical hot zone with the largest temperature change rate, such as the boundary between the high-temperature zone and the intermediate zone. Due to the intense thermal gradient and significant changes in crystal growth rate in this region, stronger guidance ability is required to suppress crystal orientation deviation and interface disturbance.

[0107] Secondary seed crystals with moderate guidance factors are arranged in the middle hot zone as a transition guiding layer for crystal orientation continuation, maintaining the consistency of the growth direction.

[0108] Tertiary seed crystals with low guidance factors are arranged in the thermally stable region (e.g., the furnace bottom or outer boundary) to stabilize the edge thermal environment and reduce the disturbance of thermal stress on the growth direction of the main crystal.

[0109] During the arrangement process, the three-dimensional angles and spacing of each unit are finely adjusted by a high-precision positioning device to ensure that the normal directions of each seed crystal in the array are approximately coaxial in space, forming a "quasi-oriented seed crystal array" with an overall consistent orientation.

[0110] The role of this array structure in the thermal field is not only to physically guide the growth direction of new crystals, but also to form a "thermal flow barrier" and "thermal channel" through its bulk thermal conductivity and crystal orientation characteristics, thereby improving the directionality and uniformity of the thermal field.

[0111] In this invention, a directional vapor phase crystallization method based on a quasi-directional seed array is proposed to achieve high-quality continuous growth of silicon carbide crystals. This method controls the temperature gradient distribution, atmosphere composition, and flow field state within the crystal growth furnace, allowing the sublimated silicon carbide vapor phase precursor to be transported along a predetermined direction under controlled conditions, preferentially undergoing directional nucleation and epitaxial growth on the intrinsic crystal planes of the seed array. Simultaneously, a crystal plane normal feedback mechanism is introduced to achieve adaptive adjustment of the thermal field and airflow, thereby maintaining the consistency of the crystal growth direction and the flatness of the crystal planes, improving the integrity and uniformity of the crystal structure.

[0112] First, the temperature of the multi-level temperature zones within the crystal growth furnace is controlled to create a stable, decreasing temperature gradient from the source region (raw material sublimation region) towards the crystal growth region. The formation of this temperature gradient is fundamental to realizing the sublimation-migration-deposition process of the vapor precursor.

[0113] In practical operation, the temperature zones can be divided into: a high-temperature heat supply zone (temperature range 2200℃~2400℃), an intermediate heat conduction zone (temperature range 2000℃~2200℃), and a low-temperature deposition zone (seed array zone, temperature controlled at 1800℃~2000℃). Through multi-segment control of the heating device and precise arrangement of the heat shield reflection structure, local heat flow is concentrated towards the seed array, forming a directional and continuous sublimation transport path.

[0114] This heat flow path can be considered a heat-driven transport channel, and its stability plays a crucial role in the uniformity of crystal deposition. It is recommended that the temperature gradient be kept within a stable range of 20 to 40 degrees Celsius per centimeter to ensure that the silicon carbide precursor migrates and deposits at a moderate rate, avoiding vapor nucleation or interface overcooling.

[0115] After the thermal field is constructed, a high-purity inert gas, preferably argon (Ar) or a hydrogen-argon mixture (H2 / Ar), is introduced into the furnace cavity. Its functions include:

[0116] To maintain the chemical inertness of the reaction system and suppress side reactions;

[0117] Stabilize the gas phase flow field and promote the directional movement of the sublimated precursor;

[0118] Controlling local pressure affects the deposition efficiency and crystallization kinetics of precursors at the interface.

[0119] The gas pressure is generally controlled between 100 and 800 Torr, and the gas flow rate is maintained within the range of 2 to 5 standard liters per minute (SLM). The airflow direction and rate in the source area, sedimentation area and exhaust area are controlled separately through a multi-channel gas duct system.

[0120] Under this heat-flow synergistic control, silicon carbide sublimation products (mainly including Si, Si2C and SiC vapor clusters) can be uniformly transported to the surface of the quasi-oriented seed array along the heat flow direction, and non-oriented deposition is avoided on amorphous surfaces or furnace walls.

[0121] Once the gaseous precursor is transported to the surface of the seed array, the local atmosphere and pressure conditions will determine whether effective nucleation and epitaxial deposition occur. Therefore, this invention employs a fine-tuning approach, using gas composition adjustment and localized gas extraction techniques to create a micro-atmosphere distribution in the deposition zone.

[0122] The crystal faces exposed after pretreatment of the seed array (e.g., the 0001 face or the 11-20 face) are intrinsically low-energy faces with highly preferential deposition properties. Under the combined effect of a slightly positive pressure environment (slightly higher than atmospheric pressure) and high directional heat flow, Si and C in the gas phase components are directionally deposited on the seed crystal faces in solid form, forming initial crystal nuclei and gradually growing into complete single crystals.

[0123] To avoid deposition of non-intrinsic planes or crystal orientation shift, the proportion of hydrogen in the atmosphere should be controlled between 5% and 15% to adjust deposition kinetics and the generation of volatile byproducts, thereby creating a deposition environment conducive to maintaining crystal orientation.

[0124] To continuously ensure the stability of the crystal growth direction and the flatness of the crystal surface, this invention introduces a seed normal feedback adjustment mechanism during the crystal growth process. Specifically, this includes:

[0125] Using an in-situ laser interferometry system or high-resolution optical imaging equipment, the normal direction of the crystal planes on the surface of the seed array can be monitored in three dimensions. The changing trend and orientation shift of the normal direction at the crystal interface can be determined by information such as changes in interference fringes and the offset angle of the reflecting surface.

[0126] The collected normal vector data is compared with the standard model of the target crystal orientation to calculate the crystal orientation deviation angle and the corresponding thermal perturbation factor. This deviation angle is defined as the angle between the actual normal and the desired crystal orientation; the smaller the cosine value, the more severe the deviation. By combining data from multiple crystal types, a spatially distributed "crystal orientation shift map" can be generated.

[0127] Based on the offset map results, the output power of the heating source, the position and angle of the thermal reflector are automatically adjusted in the corresponding area, and the angle and outlet size of the gas duct are finely adjusted to achieve point-to-point adjustment of the local thermal gradient and airflow direction.

[0128] As crystal growth progresses, internal stress, interface migration, and local deposition behavior continuously change. The system dynamically updates control parameters based on the crystal orientation recovery rate and interface reconstruction trend, achieving adaptive and coordinated feedback between the thermal and fluid fields to prevent continuous crystal orientation shift, interface coarsening, or step instability.

[0129] In this invention, in order to ensure stable lattice orientation and reasonable stress distribution during crystal growth, and to avoid the impact of defect accumulation on crystal quality, a dynamic control mechanism based on real-time monitoring and feedback adjustment is designed.

[0130] First, for each reused unit cell during the crystal growth process, real-time data acquisition of crystallization behavior is performed using a non-contact in-situ monitoring system. The monitoring content mainly includes:

[0131] Crystallization rate: The velocity of the crystal front is monitored using laser reflection interferometry or area array phase detection technology, and the change in crystal growth layer thickness per unit time is calculated.

[0132] Crystal plane migration state: The morphological evolution of the crystal interface is tracked and analyzed using a high-resolution optical imaging system to identify whether there are structural instabilities such as distortion, fluctuation, and step aggregation on the crystal plane.

[0133] Microscopic defect evolution: The formation and expansion process of defect centers are detected by methods such as scattered light analysis, infrared thermography or real-time Raman displacement spectroscopy, especially the dynamic evolution trend of structural defects such as dislocations, voids and stacking faults.

[0134] By using the above methods, a continuous structural evolution data chain is established as input for the subsequent thermal-stress response model, reflecting the actual response state of the reused unit during the growth process.

[0135] Building upon structural evolution monitoring, this study further introduces methods for detecting changes in internal crystal stress and lattice orientation to achieve a deeper level of physical state perception. The following technical approach is employed:

[0136] Lattice stress detection: Using a high-temperature in-situ Raman spectroscopy analysis system, the shift of characteristic peaks is detected. Based on the shift and linewidth changes of the Raman peaks, the magnitude and distribution of internal stress are quantitatively calculated. The stress shift is related to temperature, thermal gradient, and crystal defects, and can be converted to actual units (e.g., megapascals) using the corresponding calibration curve.

[0137] Lattice orientation monitoring: In-situ X-ray diffraction (XRD) or electron backscatter diffraction (EBSD) methods are used to precisely measure the lattice orientation of micro-regions on the crystal surface and capture crystal orientation rotation, shift and orientation distortion behavior;

[0138] Dynamic response model construction: Input multiple variables such as lattice stress change rate, crystal orientation shift angle, thermal gradient, and local growth rate into a multiphysics modeling system (such as COMSOL Multiphysics) to establish a unit cell-level thermal stress crystal orientation dynamic response model. This model can predict what kind of directional shift or stress concentration will occur in the crystal under given thermal field and material supply conditions, thus providing a feedforward basis for regulation.

[0139] For example, if the crystal orientation shift angle of a certain unit cell is consistently greater than 2 degrees, and accompanied by a significant increase in the redshift of the Raman peak, the model can determine that there is thermal stress concentration in this region and predict the instability trend of the interface.

[0140] Based on the above response model, a feedback control algorithm system is constructed, which integrates the following functions:

[0141] Data fusion module: Synchronously integrates monitoring data from multiple reuse units and performs partitioning processing using spatial coordinates as an index;

[0142] Stress-crystal orientation shift trend analysis: Fitting and analyzing the stress and crystal orientation shift values ​​in the time series to determine the shift trend (such as continuous growth, oscillating change or tendency to stabilize);

[0143] Control command generation rules: Based on the predetermined offset tolerance threshold (e.g., crystal orientation offset not greater than ±1°, stress not exceeding 50MPa), once a certain unit exceeds the control range, the corresponding control command signal is automatically generated.

[0144] The instruction signal includes: the required direction of thermal field adjustment, the adjustment range (expressed as power or temperature change), the gas flow rate adjustment ratio, the percentage change in the feeding rate, etc., all of which are input into the control and execution system in a standardized format.

[0145] Upon receiving the control command signal, the control system executes a coordinated adjustment of the thermal field and raw material supply:

[0146] By changing the power output of the heating section, moving the position of the heat reflector, or adjusting the angle of the heating element, rapid fine-tuning of the temperature distribution in localized areas can be achieved. For example, if a part of the crystal deviates from the main crystal's orientation upwards, the temperature above that area can be increased to correct the heat flow direction.

[0147] The flow rate of the reactive gas can be adjusted by using a mass flow controller (MFC) to reduce the supply rate to a range more suitable for epitaxial deposition; or the temperature of the sublimation source can be adjusted to indirectly control the precursor generation rate, thereby affecting the material supply balance at the growth interface.

[0148] The control effect is continuously evaluated based on the actual feedback response results, and the parameter optimization strategy is dynamically updated to achieve adaptive adjustment of the system, thereby improving the control accuracy and crystal quality stability.

[0149] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for crystal growth using silicon carbide grain growth residue, characterized in that: include: Pre-treatment of silicon carbide crystal growth waste includes removing surface oxide layer, impurity residue and structural defect areas to obtain reusable unit cells with crystal orientation consistency; Based on the crystal orientation parameters and thermal conductivity of the reusable unit, a crystal guiding model suitable for high-temperature thermal fields is constructed. Multiple reusable unit cell arrays are arranged in a multi-level temperature zone set in the crystal growth furnace to form a quasi-oriented seed array; Under controlled temperature and atmosphere conditions, the quasi-oriented seed array guides the silicon carbide vapor precursor to undergo directional crystallization, thereby achieving continuous growth of high-quality crystals. The crystallization behavior of each reused unit cell during crystal growth is monitored, and a feedback control mechanism is constructed based on stress shift and lattice orientation changes to adjust the thermal gradient and raw material supply rate in real time.

2. The method for crystal growth using silicon carbide grain growth residue according to claim 1, characterized in that: The pretreatment of the silicon carbide grain growth residue includes: Selective mechanical grinding and chemical etching are combined to remove the surface oxide layer and expose the intrinsic crystal structure of the silicon carbide crystal growth residue. It also removes metallic and non-metallic impurities, including oxygen, sodium, and iron, that remain on the crystal surface and in microcracks and affect the purity of subsequent crystallization. Crystal orientation consistency is tested and graded to select surplus unit cells that meet the specified lattice orientation error range; The edges of the remaining material unit are adjusted and cut to construct a reusable seed array module for crystal guidance control in the subsequent crystal growth process.

3. The method for crystal growth using silicon carbide grain growth residue according to claim 2, characterized in that: The construction of a crystal guiding model suitable for high-temperature thermal fields includes: Collect crystal orientation parameters, thermal conductivity, thermal expansion coefficient and crystal plane orientation information of the reuse unit cells to construct a multidimensional crystal orientation-thermal conductivity coupled database; Based on the database, the growth trend of crystals under different thermal gradients and crystal orientation combinations is simulated and modeled using the finite element thermal field simulation method, and the thermal field stability index and the guiding gain factor are extracted. By combining the coupling evolution relationship between the heat flux density field and the crystal orientation deviation angle, a crystal guidance propagation path function is established to optimize the arrangement and initial orientation of the seed array in the thermal field.

4. The method for crystal growth using silicon carbide grain growth residue according to claim 3, characterized in that: Multiple reusable unit cells are arranged in an array within a multi-level temperature zone set within the crystal growth furnace to form a quasi-oriented seed array, including: Based on the radial and axial thermal field distribution patterns within the crystal growth furnace, the furnace cavity is divided into multiple temperature gradient zones, and a temperature control range and heat flow guidance direction are set for each zone. Based on the crystal orientation parameters, thermal conductivity, and orientation factor of the reused unit cells, multiple unit cells are classified and graded according to the intensity of their orientation effect. Units with high orientation strength are preferentially arranged in the critical hot region with a large temperature change rate, while units with low orientation strength are arranged in the thermal stable region to construct a multi-level linked quasi-oriented seed array.

5. The method for crystal growth using silicon carbide grain growth residue according to claim 4, characterized in that: The process of guiding the quasi-oriented seed array to undergo directional crystallization of the silicon carbide vapor precursor under controlled temperature and atmosphere conditions includes: Adjusting the temperature gradient of each temperature zone in the crystal growth furnace forms a heat flow channel that steadily decreases from the high-temperature source area to the crystal growth area, thus constructing a stable sublimation transport path; High-purity argon or hydrogen-argon mixture is introduced into the furnace, and the gas pressure and flow rate parameters are controlled so that the sublimated silicon carbide gas phase precursor is transported to the surface of the quasi-oriented seed array along the heat flow direction. Adjusting the local micro-atmosphere composition and pressure distribution in the crystal growth region allows the gaseous precursor to preferentially deposit on the intrinsic crystal surface exposed by the seed array, promoting directional nucleation and epitaxial expansion at the gas-solid interface. During crystal growth, the crystal growth rate and crystal surface flatness are monitored in real time, and the local thermal gradient and airflow guidance are adjusted based on the feedback of the seed normal direction.

6. The method for crystal growth using silicon carbide grain growth residue according to claim 5, characterized in that: The adjustment of the local thermal field gradient and airflow guidance based on the feedback of the seed crystal normal direction includes: During crystal growth, the spatial variation and orientation offset of the normal direction of the crystal plane on the seed array surface are monitored in real time. The monitoring data is compared with the standard model of the target crystal orientation, and the correlation factor between the crystal orientation deviation angle and thermal disturbance is calculated to form a dynamic offset spectrum for regulation. Based on the offset spectrum, the power distribution of the heating source, the angle of the heat reflection structure, and the gas flow path are adjusted in the local area; Based on the crystal orientation recovery rate and the crystallization interface reconstruction trend, the control strategy is updated to achieve adaptive thermal-fluid field coordinated feedback regulation during crystal growth.

7. The method for crystal growth using silicon carbide grain growth residue according to claim 6, characterized in that: The monitoring of the crystallization behavior of each reused unit cell during crystal growth involves constructing a feedback control mechanism based on stress shift and lattice orientation changes, including: The crystallization rate, crystal plane migration state, and micro-defect evolution of each reuse unit are monitored in real time to obtain its structural evolution data during the growth process; Detect the changes in lattice stress distribution and lattice orientation, and establish a dynamic response model of thermal stress crystal orientation for each unit cell. Based on the aforementioned thermal stress crystal orientation dynamic response model, a feedback control algorithm is constructed to quantitatively analyze the stress shift trend and lattice rotation behavior, and generate control command signals. Based on the control signal, the local thermal field gradient distribution and the gas phase precursor supply rate are dynamically adjusted.

Images