Gallium nitride aluminum avalanche detector preparation optimization method based on defect identification
Through defect identification and optimization technology, the problems of insufficient defect identification and low production efficiency in the preparation of gallium nitride avalanche detectors are solved, and a high-quality and efficient preparation process is achieved.
Patent Information
- Application Number
- CN202411274743.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-12
- Publication Date
- 2024-12-13
AI Technical Summary
The prior art lacks effective defect identification methods in the preparation process of gallium nitride avalanche detectors, resulting in unstable performance and low reliability, and traditional methods rely on experience to lead to low production efficiency and inconsistency in product.
Defect recognition technology is used to detect materials, obtain defect types and distributions, calculate defect density, establish a mapping relationship between growth conditions and defect types, adjust growth conditions through virtual simulation and iterative optimization, monitor and dynamic adjustment in real time to improve preparation quality and efficiency.
Through precise defect identification and optimization methods, the preparation quality and production efficiency of gallium nitride aluminum avalanche detector are significantly improved, ensuring stable and consistent product performance.
Smart Images
Figure CN119132994A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of detector manufacturing, and in particular to a method for optimizing the preparation of a gallium aluminum nitride avalanche detector based on defect identification. Background Art
[0002] In today's electronic technology field, gallium aluminum nitride materials have broad application prospects in the preparation of avalanche detectors due to their unique properties. However, existing technologies face many challenges in the preparation of gallium aluminum nitride avalanche detectors. On the one hand, due to the lack of effective defect identification methods, it is difficult to accurately grasp the defect types and distribution in the material, which may cause the prepared detectors to have problems such as unstable performance and low reliability, seriously affecting the preparation quality. On the other hand, traditional preparation methods often rely on experience to set growth conditions and lack scientific optimization methods, resulting in low production efficiency and difficulty in ensuring product consistency.
[0003] The prior art has technical problems of low preparation quality and production efficiency of gallium aluminum nitride avalanche detectors. Summary of the invention
[0004] The present application provides a method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification, which is used to solve the technical problems of low preparation quality and production efficiency of gallium aluminum nitride avalanche detectors in the prior art.
[0005] In view of the above problems, the present application provides a method for optimizing the preparation of gallium nitride aluminum avalanche detectors based on defect identification, the method comprising: using defect identification to detect materials in the preparation process of gallium nitride aluminum avalanche detectors, and obtaining defect identification results, wherein the defect identification results include multiple defect types and defect distributions; performing defect density calculation on the defect distribution to determine the defect density of the defect distribution; obtaining a mapping relationship between growth conditions and the defect types, and obtaining multiple defect growth condition categories corresponding to the multiple defect types based on the mapping relationship; based on the defect density, adjusting and optimizing multiple existing growth condition parameters of the multiple defect growth condition categories to obtain a final growth optimization condition set, wherein in the optimization process, the adjusted growth optimization condition set is evaluated according to the defect density; based on the final growth optimization condition set, real-time monitoring of the defect types and distribution during the material growth process is performed, and dynamic feedback and adjustment are performed.
[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0007] Defect identification is used to detect materials in the process of preparing gallium nitride aluminum avalanche detectors to obtain defect identification results; defect density is calculated for the defect distribution to determine the defect density of the defect distribution; a mapping relationship between growth conditions and defect types is obtained, and multiple defect growth condition categories corresponding to the multiple defect types are obtained based on the mapping relationship; based on the defect density, multiple existing growth condition parameters of the multiple defect growth condition categories are adjusted and optimized to obtain a final growth optimization condition set; based on the final growth optimization condition set, the defect type and distribution during the material growth process are monitored in real time, and dynamic feedback and adjustment are performed. The technical effect of significantly improving the preparation quality and production efficiency of gallium nitride aluminum avalanche detectors through precise defect identification, virtual simulation and iterative optimization is achieved. BRIEF DESCRIPTION OF THE DRAWINGS
[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0009] Figure 1 A schematic diagram of a process for optimizing the preparation of a gallium nitride aluminum avalanche detector based on defect identification provided in an embodiment of the present application;
[0010] Figure 2 A schematic diagram of a process for obtaining defect identification results in a method for optimizing the preparation of a gallium aluminum nitride avalanche detector based on defect identification provided in an embodiment of the present application. DETAILED DESCRIPTION
[0011] The present application provides a gallium aluminum nitride avalanche detector preparation optimization method based on defect identification, which is used to solve the technical problems of low preparation quality and production efficiency of gallium aluminum nitride avalanche detectors in the prior art.
[0012] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0013] Examples, such as Figure 1 As shown, the present application provides a method for optimizing the preparation of gallium nitride aluminum avalanche detectors based on defect identification, the method comprising:
[0014] Step S100: using defect identification to detect materials in the process of preparing the AlGaN avalanche detector to obtain defect identification results, wherein the defect identification results include multiple defect types and defect distributions.
[0015] Specifically, first of all, defect identification will use a variety of detection methods, using high-resolution microscopes and electron beam detection technology to observe and analyze the materials. In this process, various types of defects are identified, including crystal structure defects, such as dislocations and stacking faults; impurity defects, such as different types of impurity particles mixed in the material; and surface defects, such as scratches and pits. Each type of defect has different degrees of impact on the performance of the detector, so it is crucial to accurately identify them. At the same time, the distribution of defects in the material will be determined, including the distribution of defects in different regions and different levels of the material. For example, some defects are concentrated in a specific area of the material, which suggests that there are problems with the preparation process in this area, and understanding the distribution of defects helps to deeply analyze the causes and propagation mechanisms of defects, and provide an important basis for subsequent optimization measures. Through this step, detailed defect identification results can be obtained, including multiple defect types and defect distributions, to provide accurate data support for subsequent steps, so as to optimize the preparation process of gallium nitride aluminum avalanche detectors and improve the quality and performance of detectors.
[0016] Step S200: performing defect density calculation on the defect distribution to determine the defect density of the defect distribution.
[0017] Specifically, the defect density of the defect distribution obtained in the previous step is calculated to determine the defect density of the distribution. First, in order to calculate the defect density, it is necessary to count the number of all detected defects, accurately identify and count various defects in the material, whether they are crystal structure defects, impurity defects or surface defects, and include them in the statistical range, and record the total number of various defects. Next, measure the surface area of the gallium aluminum nitride material. This can be achieved through precise measuring instruments and methods to ensure that the surface area measurement results are accurate and reliable. Then, the defect density of the defect distribution is calculated based on the ratio of the total number of various defects obtained by statistics to the measured surface area of the material. Defect density is an important indicator that can intuitively reflect the density of defects in the material. A higher defect density means that there are major problems with the quality of the material and more active measures need to be taken to improve it; while a lower defect density indicates that the quality of the material is relatively good, but further optimization may still be required to improve the performance of the detector. By calculating the defect density of the defect distribution, it can provide a specific quantitative basis for the subsequent adjustment and optimization of the growth condition parameters, help determine the best set of growth optimization conditions, and thus improve the preparation quality of the gallium aluminum nitride avalanche detector.
[0018] Step S300: obtaining a mapping relationship between growth conditions and the defect types, and obtaining a plurality of defect growth condition categories corresponding to the plurality of defect types based on the mapping relationship.
[0019] Specifically, first, a mapping relationship between growth conditions and defect types is established by querying a historical database. A large amount of data from the past preparation of gallium nitride aluminum avalanche detectors is stored in the historical database, including various defect types that appear under different growth conditions. By analyzing and sorting these data, the potential correlation between growth conditions and defect types can be found. For example, certain specific growth temperature, pressure, gas flow rate and other conditions may have a high correlation with specific defect types. Then, based on the established mapping relationship between growth conditions and defect types, a defect growth condition mapping model is constructed. This model can more systematically represent the correspondence between growth conditions and defect types, and can quickly determine possible growth condition factors based on the input defect type. Then, using the constructed defect growth condition mapping model, the identified multiple defect types are mapped. Through this mapping, the multiple defect growth condition categories corresponding to each defect type can be determined. For example, for a specific crystal structure defect, the corresponding growth condition categories are determined to include excessive growth temperature, unstable growth pressure, etc. These defect growth condition categories provide a clear direction for the subsequent optimization of the growth condition parameters, which helps to improve the growth conditions in a targeted manner, reduce the generation of defects, and improve the preparation quality of GaN aluminum avalanche detectors.
[0020] Step S400: Based on the defect density, multiple existing growth condition parameters of the multiple defect growth condition categories are adjusted and optimized to obtain a final growth optimization condition set, wherein during the optimization process, the adjusted growth optimization condition set is evaluated according to the defect density.
[0021] Specifically, first, multiple growth condition parameter spaces of multiple defect growth condition categories are obtained. This parameter space contains various factors that affect material growth and defect generation, such as temperature, pressure, gas flow, growth time, etc. Within this space, the existing growth condition parameters are randomly adjusted to generate potential growth condition combinations. These potential growth condition combinations represent different growth condition settings, which will have different effects on defect generation and material quality.
[0022] Next, based on these potential growth condition combinations, virtual simulated growth is performed, a growth model of aluminum gallium nitride material is constructed in a virtual environment, growth parameters are set according to the potential growth condition combinations, and then numerical simulation software is used to simulate virtual simulated growth under these conditions, generate simulated defect identification results, and calculate simulated defect density. Through virtual simulated growth, the impact of different growth condition combinations on defects can be quickly evaluated without actually growing materials.
[0023] Potential growth condition combinations are evaluated based on the simulated defect density and the actual defect density. If the simulated defect density is less than the actual defect density, it means that the potential growth condition combination may help reduce the generation of defects, so this combination is retained and the next round of optimization is continued. If the simulated defect density is greater than the actual defect density, it means that this potential growth condition combination may cause more defects, so it is not retained.
[0024] During the whole optimization process, the above steps are repeated continuously to perform global iterative optimization until convergence. In each iteration, the adjusted growth optimization condition set is evaluated according to the defect density, and the change trend of the simulated defect density is monitored in real time. By continuously adjusting the growth condition parameters, the optimal growth condition combination is gradually approached to reduce the generation of defects and improve the quality of the material.
[0025] Finally, when the convergence condition is reached, the potential growth condition combination that is finally retained is determined as the final growth optimization condition set, which can be applied in the actual preparation process of gallium aluminum nitride avalanche detectors to improve the quality and performance of the product.
[0026] Step S500: Based on the final growth optimization condition set, the defect type and distribution during the material growth process are monitored in real time, and dynamic feedback and adjustment are performed.
[0027] Specifically, based on the final set of growth optimization conditions, the material growth in the preparation process of the gallium nitride aluminum avalanche detector is monitored and adjusted dynamically in real time. Once the final set of growth optimization conditions is obtained, advanced monitoring technology is continuously used to monitor the defect type and distribution in real time during the material growth process. This involves the use of high-resolution imaging equipment, sensors, etc., which can capture any possible defect information in a timely manner at various stages of material growth. Through real-time monitoring, it is possible to know at any time whether new defect types appear in the material and whether the distribution of the original defects has changed. If a change in the defect type or an abnormality in the defect distribution is found during the monitoring process, dynamic feedback will be immediately given. This feedback can be sending an alarm message to remind you to pay attention to problems in the production process. At the same time, the growth process is adjusted according to the actual situation. For example, if the frequency of a certain defect type is found to increase, it is necessary to further analyze which growth condition has deviated, and to fine-tune the corresponding growth condition parameters in a timely manner. Through real-time monitoring, dynamic feedback and adjustment, it can be ensured that the material growth is always carried out under the optimal conditions. Even if unforeseen situations occur during the production process, it can respond quickly to minimize the generation of defects and ensure the quality and performance stability of the gallium nitride aluminum avalanche detector. Such a dynamic adjustment mechanism makes the production process more flexible and controllable, helping to improve production efficiency and product quality.
[0028] In one possible implementation, Figure 2 As shown, step S100 also includes:
[0029] Step S110: performing surface imaging on the aluminum gallium nitride material, and identifying the multiple defect types by comparing with a standard material map.
[0030] Step S120: imaging multiple regions of the aluminum gallium nitride material, counting defects in each region, and generating the defect distribution.
[0031] Specifically, the surface imaging of aluminum gallium nitride materials uses high-resolution imaging technology, electron microscopy or optical microscopy to obtain detailed image information of the material surface. Through these imaging methods, the microstructure and characteristics of the material surface can be clearly observed. Next, the acquired material surface image is compared with the material standard atlas, which is usually determined after a lot of research and verification, and contains reference images of various material characteristics under normal conditions and possible defect types. Through careful comparison, multiple defect types on the material surface can be identified. For example, crystal structure defects such as dislocations and stacking faults can be identified; surface impurity defects such as different types of particles attached to the surface can also be found.
[0032] Imaging multiple areas of the gallium aluminum nitride material can be done by adjusting the field of view of the imaging device, scanning and imaging different parts of the material one by one, and performing detailed defect statistics for each imaged area, accurately identifying and counting various defects in each area, and integrating the defect statistics of each area to generate the defect distribution of the material. This defect distribution can include information such as the frequency and concentration of defects in different locations of the material. For example, it will be found that some areas have more defects, while others have relatively fewer, or certain types of defects are more likely to occur in specific areas.
[0033] In a possible implementation, step S200 further includes:
[0034] Step S210: Count all detected defects and record the total number of each type of defects.
[0035] Step S220: measuring the surface area of the aluminum gallium nitride material.
[0036] Step S230: Obtain the defect density of the defect distribution according to the ratio of the total number of each type of defects to the surface area of the material.
[0037] Specifically, all detected defects are counted. This process requires the use of advanced detection technologies and tools to ensure that each defect can be accurately identified. Whether it is a crystal structure defect, an impurity defect or a surface defect, each defect must be counted one by one, and the total number of each type of defect must be recorded. For example, different types of defects such as dislocation defects and particle impurity defects are counted separately to fully understand the overall situation of defects in the material.
[0038] The surface area of aluminum gallium nitride materials is measured using precise measuring instruments and methods to ensure that the surface area measurement results are accurate and reliable. Since defect density is an indicator related to the surface area of the material, accurate surface area measurement is crucial for subsequent calculations.
[0039] The defect density of the defect distribution is obtained based on the ratio of the total number of defects of each type to the surface area of the material. The total number of defects of each type recorded previously is divided by the measured surface area of the material. The ratio is the defect density. The defect density reflects the number of defects per unit area and is an important indicator for measuring material quality. A higher defect density means that the defects in the material are denser, which will have a greater impact on the performance of the detector; while a lower defect density indicates that the material quality is relatively good, but further optimization may still be required to improve the performance of the detector. Through this series of steps, the defect density of aluminum gallium nitride materials can be accurately calculated, providing a specific quantitative basis for the subsequent optimization of growth conditions and improvement of the preparation process.
[0040] In a possible implementation, step S300 further includes:
[0041] Step S310: Establish a mapping relationship between growth conditions and defect types by querying the historical database.
[0042] Step S320: constructing a defect growth condition mapping model based on the mapping relationship between the growth conditions and the defect types.
[0043] Step S330: obtaining a plurality of defect growth condition categories corresponding to the plurality of defect types through a defect growth condition mapping model.
[0044] Specifically, the mapping relationship between growth conditions and defect types is established through historical database queries. The historical database stores a large amount of relevant data on the preparation of aluminum gallium nitride avalanche detectors in the past, including various defect types that appear under different growth conditions. During the query process, the relationship between different growth parameter combinations, such as temperature, pressure, gas flow, etc. and the corresponding defect types that appear is carefully analyzed. After combing and summarizing a large amount of data, it can be found that certain specific growth conditions often have a high correlation with specific defect types. For example, a higher growth temperature may be associated with a certain crystal structure defect, while a specific pressure range may easily lead to the appearance of impurity defects.
[0045] Collect a large amount of historical data, which includes different combinations of growth conditions and corresponding defect types. Organize and preprocess these data to ensure the accuracy and completeness of the data. The decision tree algorithm is a classification algorithm based on a tree structure. It builds a decision tree model by gradually dividing the data. When building a defect growth condition mapping model, the growth condition can be used as the input feature and the defect type as the output category. When starting to build a decision tree, first select the best feature as the root node. The selection of this feature is usually based on indicators such as information gain and Gini coefficient. Information gain measures the degree to which a feature reduces the uncertainty of classification, and the Gini coefficient reflects the impurity of the sample set. Selecting the feature with the highest information gain or the lowest Gini coefficient as the root node can enable the decision tree to reduce uncertainty as quickly as possible during the classification process. For example, if the growth conditions include features such as temperature, pressure, and gas flow, by calculating the information gain, it can be found that the temperature feature has the highest information gain when distinguishing different defect types. Then, select temperature as the root node and divide the data set into different subsets according to different temperature values. For each subset, continue to repeat the above process and select the next best feature for division until the stopping condition is met. The stopping condition can be that the purity of the data set reaches a certain level, the depth of the tree reaches a preset value, or the number of samples in the node is less than a certain threshold. After the decision tree model is built, you can input the new growth conditions, start from the root node, and follow the branch path of the decision tree to gradually determine the corresponding defect type. An effective defect growth condition mapping model is constructed based on the mapping relationship between growth conditions and defect types, providing strong support for subsequent defect analysis and growth condition optimization.
[0046] The defect growth condition mapping model that has been constructed is trained based on a large amount of historical data and the mapping relationship between growth conditions and defect types. This model can predict the possible defect types based on the input growth condition parameters, and can also infer the growth condition categories that may cause the defect based on the known defect types. When it is necessary to determine the defect growth condition categories corresponding to multiple defect types, the identified multiple defect types are input into the model one by one, and the model will analyze and calculate according to its internal logic and algorithm. Assume that three defect types, namely crystal structure defects, impurity defects and surface defects, are identified. After the crystal structure defects are input into the model, the model will infer that the growth condition category corresponding to the defect type is too high growth temperature, unstable growth pressure, etc. For impurity defects, the model may determine that the corresponding growth condition category is insufficient raw material purity, improper gas flow control, etc. Surface defects may be related to growth condition factors such as mechanical stress and environmental humidity during the growth process. In this way, the corresponding defect growth condition category can be found for each defect type, providing specific directions and goals for the subsequent adjustment and optimization of growth condition parameters. This allows targeted improvements to be made to these specific types of growth conditions, thereby reducing the generation of defects and improving the preparation quality of gallium aluminum nitride avalanche detectors.
[0047] In a possible implementation, step S400 further includes:
[0048] Step S410: Acquire multiple growth condition parameter spaces of the multiple defective growth condition categories, and randomly adjust the multiple existing growth condition parameters in the multiple growth condition parameter spaces to obtain potential growth condition combinations.
[0049] Step S420: Based on the potential growth condition combination, perform virtual simulated growth to obtain simulated defect density.
[0050] Step S430: Evaluate the potential growth condition combination according to the simulated defect density and the defect density, and continue to perform global iterative optimization until convergence to obtain a final growth optimization condition set.
[0051] Specifically, multiple growth condition parameter spaces of multiple defect growth condition categories are obtained. These parameter spaces cover various factors that affect material growth and defect generation, such as temperature, pressure, gas flow, etc. Within these parameter spaces, the existing growth condition parameters are randomly adjusted. This random adjustment aims to explore different parameter combinations to find potential growth condition combinations that may reduce defects. Through random adjustment, a large number of potential growth condition combinations can be generated, providing rich options for subsequent evaluation and optimization.
[0052] A growth model of aluminum gallium nitride material is constructed in a virtual environment. This model is based on the understanding of the actual growth process and a large amount of experimental data. It can simulate the growth of materials under different growth conditions as accurately as possible. The model contains various physical and chemical process parameters, such as temperature, pressure, gas flow rate and other growth condition parameters, as well as the crystal structure, chemical composition and other characteristics of the material. Then, the potential growth condition combination is input into this growth model. Growth parameters are set according to the potential growth condition combination. These parameters will determine the various conditions in the virtual simulation growth process. For example, if the potential growth condition combination includes specific temperature, pressure and gas flow rate values, then these parameters will be set accordingly in the model to simulate the growth process of the material under these conditions. Then, virtual simulation growth is performed using numerical simulation software. The numerical simulation software simulates the growth dynamic process of the material under the set growth conditions by solving complex mathematical equations and physical models. During the simulation process, the software will track the microstructural changes, chemical composition evolution and the formation process of various defects of the material. After a period of simulated growth, the simulated defect recognition results are generated. By analyzing the material after simulated growth, various types of defects such as crystal structure defects, impurity defects, surface defects, etc. are identified. At the same time, the number and distribution of these defects are analyzed. Finally, the simulated defect density is calculated based on the statistical results of the simulated defects. The simulated defect density is determined by the ratio of the total number of defects detected to the simulated material surface area. This simulated defect density can be used as an important indicator to evaluate the advantages and disadvantages of potential growth condition combinations. If the simulated defect density is low, it means that the potential growth condition combination may help reduce the occurrence of defects in actual production; conversely, if the simulated defect density is high, it is necessary to further adjust the growth condition parameters to find a better potential growth condition combination.
[0053] According to the simulated defect density and the actual defect density, the potential growth condition combination is evaluated. If the simulated defect density is less than the actual defect density, it means that the potential growth condition combination may help reduce the generation of defects, so this combination is retained and the global iterative optimization is continued. During the iterative optimization process, steps S410 and S420 are repeated continuously, the growth condition parameters are continuously adjusted, a new potential growth condition combination is generated, and virtual simulated growth and evaluation are performed. If the simulated defect density is greater than the actual defect density, the potential growth condition combination is not retained. This process is continued until convergence. The condition for convergence can be that the amplitude of the change in the simulated defect density in multiple consecutive iterations is lower than a preset threshold. When converged, the potential growth condition combination that is finally retained is determined as the final growth optimization condition set, and this optimization condition set is applied in the actual preparation process to improve the quality and performance of the aluminum gallium nitride avalanche detector.
[0054] In a possible implementation, step S420 further includes:
[0055] Step S421: constructing a growth model of aluminum gallium nitride material in a virtual environment, and setting growth parameters based on the potential growth condition combination.
[0056] Step S422: using numerical simulation software to simulate virtual growth under the potential growth condition combination, generate simulated defect identification results, and calculate simulated defect density.
[0057] Specifically, a growth model of GaN aluminum material is constructed in a virtual environment. This model is based on a deep understanding of the growth process of GaN aluminum material and a large amount of experimental data. It takes into account various factors that affect material growth, including temperature, pressure, gas flow rate, growth time and other parameters in the potential growth condition combination. By accurately defining the physical and chemical processes in the model, the growth of materials under different growth conditions can be simulated as realistically as possible. Once the growth model is constructed, the growth parameters are set according to the potential growth condition combination. These parameters will directly determine the growth process of the material in the virtual environment. The growth condition combination specifies a specific temperature range, pressure value and gas flow rate, and then these parameters are adjusted accordingly in the growth model to simulate the growth dynamics of the material under these specific conditions.
[0058] The virtual simulated growth under the potential growth condition combination is simulated by numerical simulation software. The numerical simulation software can accurately simulate the growth process of the material in the virtual environment by solving complex mathematical equations and physical models. During the simulation, the software will track the microstructure changes, chemical composition evolution and various defect formation processes of the material. As the simulation proceeds, the software will generate simulated defect identification results. By conducting a detailed analysis of the material after simulated growth, various types of defects such as crystal structure defects, impurity defects, surface defects, etc. can be identified. At the same time, these defects are quantitatively counted and distributed to understand the distribution of defects in the material. Finally, the simulated defect density is calculated based on the statistical results of the simulated defects. The calculation method of the simulated defect density is similar to the actual defect density, that is, the ratio of the total number of defects detected to the simulated material surface area is used as the simulated defect density. This simulated defect density can be used as an important indicator to evaluate the advantages and disadvantages of the potential growth condition combination. If the simulated defect density is low, it means that the potential growth condition combination may help reduce the defects in actual production; conversely, if the simulated defect density is high, it is necessary to further adjust the growth condition parameters to find a better potential growth condition combination. The impact of different potential growth condition combinations on the defects in the growth process of gallium aluminum nitride materials can be evaluated in a virtual environment.
[0059] In a possible implementation, step S430 further includes:
[0060] Step S431: comparing the simulated defect density with the defect density.
[0061] Step S432: If the simulated defect density is less than the defect density, retain the potential growth condition combination and proceed to the next round of optimization.
[0062] Step S433: If the simulated defect density is greater than the defect density, the potential growth condition combination is not retained.
[0063] Step S434: repeat the iterative process until the convergence condition is reached, and determine the final retained potential growth condition combination as the final growth optimization condition set.
[0064] Specifically, the simulated defect density obtained through virtual simulated growth is compared with the actual detected defect density. This comparison is a key step in evaluating the effectiveness of a potential combination of growth conditions. By comparing these two density values, one can intuitively understand the effect of this potential combination of growth conditions on reducing defects.
[0065] If it is found that the simulated defect density is less than the defect density, it means that the potential growth condition combination has shown a better effect in reducing defects in the virtual simulation. At this time, this potential growth condition combination will be retained and continue to the next round of optimization. In the subsequent optimization process, the parameters in this combination can be further adjusted, or combined with other potential growth condition combinations for optimization in the hope of obtaining better results.
[0066] However, if it is found that the simulated defect density is greater than the defect density, it means that this potential growth condition combination does not play a role in reducing defects in the virtual simulation, and even causes more defects. In this case, this potential growth condition combination is not retained and other possible combinations are explored.
[0067] Repeat the above iterative process, continuously generate new potential growth condition combinations, perform virtual simulated growth, calculate simulated defect density, and compare with actual defect density. In each iteration, retain the potential growth condition combinations with better performance and eliminate the combinations with poor performance. This process continues until the convergence condition is reached. The convergence condition can be that the change amplitude of the simulated defect density in multiple consecutive iterations is lower than the preset threshold. When the convergence condition is reached, the potential growth condition combination finally retained is determined as the final growth optimization condition set. This optimization condition set is obtained after multiple iterations and screening, and is expected to minimize the generation of defects in the actual preparation process of gallium nitride aluminum avalanche detectors and improve the quality and performance of products.
[0068] In a possible implementation, step S434 further includes:
[0069] During each iteration, the changing trend of the simulated defect density is monitored in real time, and a dynamic convergence judgment standard is set. When the change amplitude of the simulated defect density in three consecutive iterations is lower than the preset change threshold, it is judged that the convergence condition has been reached, and the final growth optimization condition set is obtained.
[0070] Specifically, in the iterative process of optimizing the preparation of gallium nitride aluminum avalanche detectors, it is crucial to monitor the changing trend of the simulated defect density in real time. In each iteration, the dynamics of the simulated defect density are closely followed, which means that the simulated defect density obtained after the virtual simulated growth of the newly generated potential growth condition combination is continuously tracked and analyzed. Through real-time monitoring, it is possible to clearly understand how the simulated defect density changes as the iteration proceeds. If the simulated defect density continues to decrease, it means that the current optimization direction is correct and is moving towards the goal of reducing defects. In order to accurately judge whether the iterative process has reached the convergence condition, a dynamic convergence judgment standard is set. When the change amplitude of the simulated defect density in three consecutive iterations is lower than the preset change threshold, it is judged that the convergence condition has been reached. The preset change threshold here is a value determined based on actual conditions and experience, which represents the acceptable range of small changes in the simulated defect density. If the change in the simulated defect density is very small in three consecutive iterations, it means that the current potential growth condition combination is very close to the optimal solution, and continuing the iteration will not bring significant improvement. At this time, the potential growth condition combination finally retained can be determined as the final growth optimization condition set. This final set of growth optimization conditions was obtained after multiple iterations and fine-tuning. It can minimize the generation of defects in the actual preparation process and improve the quality and performance of GaN AlAAVD. In this way, the efficiency and stability of the preparation process can be ensured, providing a strong guarantee for the production of high-quality detector products.
[0071] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0072] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
[0073] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.
Claims
1. A method for optimizing the preparation of gallium nitride aluminum avalanche detectors based on defect identification, characterized in that: The method comprises: Using defect identification to detect materials in the process of preparing the gallium aluminum nitride avalanche detector, and obtaining defect identification results, wherein the defect identification results include multiple defect types and defect distributions; Performing defect density calculation on the defect distribution to determine the defect density of the defect distribution; Acquire a mapping relationship between growth conditions and the defect types, and acquire multiple defect growth condition categories corresponding to the multiple defect types based on the mapping relationship; Based on the defect density, multiple existing growth condition parameters of the multiple defect growth condition categories are adjusted and optimized to obtain a final growth optimization condition set, wherein during the optimization process, the adjusted growth optimization condition set is evaluated according to the defect density; Based on the final growth optimization condition set, the defect type and distribution during the material growth process are monitored in real time, and dynamic feedback and adjustment are performed.
2. The method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification according to claim 1, characterized in that: Defect recognition is used to detect materials in the process of preparing the gallium aluminum nitride avalanche detector to obtain defect recognition results. The method also includes: Performing surface imaging on the aluminum gallium nitride material, and identifying the multiple defect types by comparing with a standard material map; Imaging is performed on multiple regions of the aluminum gallium nitride material, and defects in each region are counted to generate the defect distribution.
3. The method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification according to claim 2, characterized in that: Performing density calculation on the defect distribution to determine the defect density of the defect distribution, the method comprising: Count all detected defects and record the total number of each type of defects; Measuring the surface area of aluminum gallium nitride materials; The defect density of the defect distribution is obtained according to the ratio of the total number of each type of defects to the surface area of the material.
4. The method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification according to claim 1, characterized in that: Acquire a mapping relationship between growth conditions and defect types, and acquire multiple defect growth condition categories corresponding to the multiple defect types based on the mapping relationship. The method includes: By querying the historical database, a mapping relationship between growth conditions and defect types is established; Based on the mapping relationship between the growth conditions and the defect types, constructing a defect growth condition mapping model; A plurality of defect growth condition categories corresponding to the plurality of defect types are obtained through a defect growth condition mapping model.
5. The method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification according to claim 4, characterized in that: Based on the defect density, multiple existing growth condition parameters of the multiple defect growth condition categories are adjusted and optimized, and the method includes: Acquire multiple growth condition parameter spaces of the multiple defective growth condition categories, and randomly adjust the multiple existing growth condition parameters in the multiple growth condition parameter spaces to obtain potential growth condition combinations; Based on the potential growth condition combination, performing virtual simulated growth to obtain simulated defect density; According to the simulated defect density and the defect density, the potential growth condition combination is evaluated, and global iterative optimization is continued until convergence to obtain a final growth optimization condition set.
6. The method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification according to claim 5, characterized in that: Based on the potential growth condition combination, virtual simulated growth is performed to obtain simulated defect density, and the method includes: Constructing a growth model of aluminum gallium nitride material in a virtual environment, and setting growth parameters based on the potential growth condition combination; Numerical simulation software is used to simulate virtual growth under the potential growth condition combination, generate simulated defect identification results, and calculate simulated defect density.
7. The method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification according to claim 5, characterized in that: Based on the simulated defect density and the defect density, evaluating the potential growth condition combination, the method comprising: comparing the simulated defect density with the defect density; If the simulated defect density is less than the defect density, retaining the potential growth condition combination and continuing to the next round of optimization; If the simulated defect density is greater than the defect density, the potential growth condition combination is not retained; The iterative process is repeated until a convergence condition is reached, and the final retained potential growth condition combination is determined as the final growth optimization condition set.
8. The method for optimizing the preparation of gallium aluminum nitride avalanche detectors based on defect identification according to claim 7, characterized in that: The iterative process is repeated until a convergence condition is reached, and the final retained potential growth condition combination is determined as the final growth optimization condition set, including: During each iteration, the changing trend of the simulated defect density is monitored in real time, and a dynamic convergence judgment standard is set. When the change amplitude of the simulated defect density in three consecutive iterations is lower than the preset change threshold, it is judged that the convergence condition has been reached, and the final growth optimization condition set is obtained.