MOCVD (Metal Organic Chemical Vapor Deposition) wavelength uniformity optimization treatment system based on carrier gas regulation and control

Abstract

The embodiment of the invention provides an MOCVD (Metal Organic Chemical Vapor Deposition) wavelength uniformity optimization processing system based on carrier gas quantity regulation and control, and the system comprises a data collection module which is used for setting different carrier gas quantities in MOCVD equipment, carrying out an experiment, and carrying out the spectrum test of a grown epitaxial layer, so as to obtain the wavelength data of each position in a wafer; the model building module is used for calculating a wavelength variation coefficient based on the wavelength data and building a relation model between the gas carrying capacity and the wavelength variation coefficient; the optimization control module is used for determining an optimal gas carrying capacity range enabling the wavelength variation coefficient to be smaller than a preset target value according to the relation model; and the feedback execution module is used for dynamically adjusting the gas carrying capacity in the MOCVD production process according to the optimal gas carrying capacity range, and data acquisition, relation modeling, optimization control and feedback execution are organically combined by constructing an integrated processing system, so that accurate regulation and control of the gas carrying capacity and stable optimization of wavelength uniformity are realized.

Description

Technical Field

[0001] The embodiments in this specification relate to the field of semiconductor manufacturing technology, and in particular to an MOCVD wavelength uniformity optimization system based on carrier gas volume control. Background Technology

[0002] In the field of metal-organic chemical vapor deposition (MOCVD), controlling the wavelength uniformity of epitaxial layers has always been a challenge. Traditional methods often rely on fixed process parameters or manual adjustments based on experience, lacking systematic modeling and dynamic optimization mechanisms. This results in significant fluctuations in wavelength uniformity under different material systems and process conditions, affecting production efficiency and product consistency.

[0003] Therefore, a better solution is urgently needed. Summary of the Invention

[0004] In view of this, the embodiments of this specification provide an MOCVD wavelength uniformity optimization processing system based on carrier gas volume control to solve the technical defects existing in the prior art.

[0005] According to a first aspect of the embodiments of this specification, an MOCVD wavelength uniformity optimization system based on carrier gas volume control is provided, comprising: The data acquisition module is used to set different carrier gas volumes and conduct experiments in the MOCVD equipment, as well as to perform spectral testing on the grown epitaxial layer to obtain wavelength data at various locations within the wafer. The model building module is used to calculate the wavelength variation coefficient based on wavelength data and to establish a model of the relationship between carrier gas volume and wavelength variation coefficient. The optimization control module is used to determine the optimal carrier gas volume range that makes the wavelength variation coefficient less than the preset target value based on the relationship model. The feedback execution module is used to dynamically adjust the carrier gas volume according to the optimal carrier gas volume range during the MOCVD production process.

[0006] In one possible implementation, the carrier gas volume includes nitrogen flow rate and hydrogen flow rate.

[0007] In one possible implementation, the wavelength variation coefficient is obtained by dividing the standard deviation by the average wavelength.

[0008] In one possible implementation, the relationship model is a curve showing the relationship between carrier gas volume and wavelength variation coefficient.

[0009] In one possible implementation, the optimal carrier gas volume range determined by the optimization control module makes the wavelength variation coefficient less than 0.15 percent.

[0010] In one possible implementation, the feedback execution module, in conjunction with the feedback control system, monitors wavelength uniformity in real time and automatically adjusts the carrier gas volume.

[0011] In one possible implementation, the optimal carrier gas volume range is adjusted according to different material systems and process conditions.

[0012] In one possible implementation, the model building module calculates the wavelength variation coefficient using the following formula: First, calculate the average wavelength. ; in It is the first The wavelength values ​​at each measurement point are acquired by the data acquisition module, where N is the total number of measurement points, determined by the data acquisition module. It is the average wavelength; Then calculate the standard deviation. ; in It is the standard deviation of the wavelength data; Finally, the wavelength variation coefficient was calculated. ; Where CV is the wavelength variation coefficient.

[0013] In one possible implementation, when the model building module builds the relational model, it performs linear fitting using the following formula: The relationship between the wavelength variation coefficient CV and the carrier gas quantity Q is as follows: ; The parameters a and b are obtained using the least squares method, and the calculation formula is: , ; in , Where M is the number of experiments, determined by the data acquisition module. It is the first The carrier gas volume for this experiment was obtained by the data acquisition module. It is the first The wavelength variation coefficient for this experiment was calculated by the model building module. It is the average carrier gas volume. is the average wavelength variation coefficient, and a and b are fitting parameters.

[0014] In one possible implementation, the feedback execution module dynamically adjusts the carrier gas volume based on real-time monitored wavelength uniformity data, so that the wavelength variation coefficient is maintained within the preset target value range.

[0015] This specification provides an MOCVD wavelength uniformity optimization system based on carrier gas volume control, comprising: a data acquisition module for setting different carrier gas volumes in the MOCVD equipment and conducting experiments, and performing spectral testing on the grown epitaxial layer to obtain wavelength data at various locations within the wafer; a model building module for calculating the wavelength variation coefficient based on the wavelength data and establishing a relationship model between the carrier gas volume and the wavelength variation coefficient; an optimization control module for determining the optimal carrier gas volume range that makes the wavelength variation coefficient less than a preset target value according to the relationship model; and a feedback execution module for dynamically adjusting the carrier gas volume according to the optimal carrier gas volume range during the MOCVD production process. By constructing an integrated processing system, data acquisition, relationship modeling, optimization control, and feedback execution are organically combined, achieving precise control of the carrier gas volume and stable optimization of wavelength uniformity. Attached Figure Description Figure 1 This is a schematic diagram of an MOCVD wavelength uniformity optimization system based on carrier gas volume control, provided in one embodiment of this specification. Detailed Implementation

[0016] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0017] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0018] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0019] This specification provides a MOCVD wavelength uniformity optimization system based on carrier gas volume control, which will be described in detail in the following embodiments.

[0020] See Figure 1 , Figure 1 This diagram illustrates a system schematic of an MOCVD wavelength uniformity optimization system based on carrier gas volume control, according to an embodiment of this specification. Specifically, it includes a data acquisition module for setting different carrier gas volumes in the MOCVD equipment and conducting experiments, as well as performing spectral testing on the grown epitaxial layer to obtain wavelength data at various locations within the wafer; a model building module for calculating the wavelength variation coefficient based on the wavelength data and establishing a relationship model between the carrier gas volume and the wavelength variation coefficient; an optimization control module for determining the optimal carrier gas volume range that makes the wavelength variation coefficient less than a preset target value based on the relationship model; and a feedback execution module for dynamically adjusting the carrier gas volume according to the optimal carrier gas volume range during the MOCVD production process.

[0021] In practical applications, the data acquisition module refers to the component in an MOCVD system responsible for performing experiments and collecting wavelength data. It's used to set different carrier gas volumes and perform spectral measurements to obtain wavelength data at various locations within the wafer. The MOCVD equipment refers to the process apparatus used for metal-organic chemical vapor deposition, capable of growing epitaxial layers and providing carrier gas volume control. Carrier gas volume refers to the gas flow rate parameter used in the MOCVD process, which can include the flow rate of carriers such as nitrogen or hydrogen. The experiment refers to the epitaxial growth process in the MOCVD equipment by changing the carrier gas volume to verify the influence of different parameters on wavelength uniformity. The epitaxial layer refers to the thin film structure grown on the substrate, used to obtain its optical properties through spectral measurements. Spectroscopic measurements refer to the method of performing non-destructive optical measurements on the epitaxial layer to obtain wavelength data, enabling the analysis of wavelength distribution at various locations within the wafer. Wavelength data refers to the set of wavelength values ​​from multiple measurement points obtained through spectral measurements, used for subsequent calculations and analysis of wavelength uniformity. The model building module refers to the component that calculates statistical indicators based on wavelength data and constructs mathematical relationships to calculate the wavelength variation coefficient and establish a model relating the carrier gas volume to the wavelength variation coefficient. The wavelength variation coefficient (WVC) is a statistical parameter that measures wavelength uniformity, obtained by the ratio of the standard deviation to the average wavelength. A relational model refers to a mathematical expression or curve describing the relationship between carrier gas volume and the WVC, which can be used to predict and optimize carrier gas volume settings. An optimization control module determines the optimal carrier gas volume range based on the relational model, analyzing and ensuring the WVC is below a preset target value. The preset target value is a user-defined threshold for the WVC to define an acceptable level of wavelength uniformity. The optimal carrier gas volume range refers to the range of carrier gas volumes that keeps the WVC below the preset target value, guiding parameter adjustments during production. A feedback execution module controls the carrier gas volume in real-time during MOCVD production, dynamically adjusting the volume based on the optimal range to maintain wavelength uniformity. The MOCVD production process refers to a continuous epitaxial growth operation in a real industrial environment, using feedback control to achieve stable process output.

[0022] As a specific example: In the AlGaAs material system, an MOCVD device was used to conduct experiments with carrier gas volumes ranging from 100 sccm to 500 sccm in 50 sccm increments. After growing the epitaxial layer, wavelength data at nine points within the wafer were obtained through spectral testing, for example, the wavelength at the center point was 650 nm and the wavelength at the edge point was 648 nm. The model building module calculated the average wavelength to be 649 nm and the standard deviation to be 0.8 nm, obtaining a wavelength variation coefficient of 0.123%. A linear relationship model was established, showing that the wavelength variation coefficient was minimized when the carrier gas volume was 300 sccm. The optimization control module determined the optimal carrier gas volume range to be 280 sccm to 320 sccm based on the relationship model, with a preset target value of 0.15%. The feedback execution module monitored real-time wavelength data during actual production. When the wavelength variation coefficient approached 0.16%, the carrier gas volume was automatically adjusted from 310 sccm to 295 sccm, restoring the wavelength uniformity to the target range.

[0023] This invention achieves precise control of carrier gas volume and stable optimization of wavelength uniformity through an integrated processing system, thereby improving the level of production automation and product consistency, adapting to different material systems and process conditions, reducing the negative impact of process fluctuations, and enhancing the reliability and efficiency of large-scale production.

[0024] In one possible implementation, the carrier gas volume includes nitrogen flow rate and hydrogen flow rate.

[0025] In practical applications, nitrogen flow rate refers to the volumetric flow rate of nitrogen used as a carrier gas. It provides a stable transport environment and influences the uniformity of reactant distribution on the substrate surface. Hydrogen flow rate refers to the volumetric flow rate of hydrogen used as a carrier gas. Its unique thermal conductivity and surface reaction characteristics can be used to regulate the decomposition rate and migration ability of precursors.

[0026] As a specific example: When growing AlGaAs-based epitaxial structures, hydrogen is used as the main carrier gas, with a hydrogen flow rate of 8000 standard milliliters per minute and a nitrogen flow rate of 2000 standard milliliters per minute. Epitaxial growth is carried out by maintaining this mixed carrier gas ratio. When switching to the InGaP material system, the hydrogen flow rate is adjusted to 6000 standard milliliters per minute and the nitrogen flow rate is increased to 4000 standard milliliters per minute. This ratio change effectively optimizes the mass transfer process of different metal-organic sources at the growth interface. In actual production, by monitoring wavelength uniformity data in real time, when the wavelength value at the edge of the wafer deviates from the center point by more than a set threshold, the hydrogen flow rate is automatically increased by standard milliliters per minute from the current value, while the nitrogen flow rate is correspondingly decreased by standard milliliters per minute. Through this dynamic balance adjustment, the wavelength uniformity within the wafer is always maintained within the range required by the process specifications.

[0027] This invention clarifies the specific gas composition and flow ratio of the carrier gas, making the carrier gas volume control more clearly defined in terms of operational dimensions. This enhances the interpretability of the relationship between process parameters and epitaxial quality, provides a clear adjustment path for process optimization under different material systems, effectively improves the wavelength uniformity control accuracy and process repeatability, and reduces performance fluctuations between different batches of products.

[0028] In one possible implementation, the wavelength variation coefficient is obtained by dividing the standard deviation by the average wavelength.

[0029] In practical applications, the wavelength variation coefficient can be a statistical parameter characterizing the uniformity of wavelength distribution on an epitaxial wafer, used to quantify the degree of wavelength fluctuation on the wafer surface. The standard deviation is a mathematical measure of the dispersion of wavelength values ​​at each measurement point relative to its mean, reflecting the absolute range of fluctuation in the wavelength dataset. The average wavelength is the arithmetic mean of wavelength values ​​at all measurement points on the wafer surface, used to provide a reference for the central tendency of the wavelength distribution.

[0030] As a specific example: During the preparation of GaN-based epitaxial wafers, wavelength data from 49 test points on the wafer surface were collected using a fully automated spectral scanning system. The measured wavelength values ​​were distributed between 455.2 nm and 455.8 nm. The calculated average wavelength was 455.5 nm, with a standard deviation of 0.15 nm. Dividing the standard deviation by the average wavelength yielded a wavelength variation coefficient of 0.033%, which was lower than the 0.05% required by the process specifications. After replacing the reaction chamber components, the wavelength variation coefficient of the new batch of products increased to 0.06%. By adjusting the carrier gas ratio, the value was brought back under control within the acceptable range.

[0031] This invention establishes a quantifiable uniformity evaluation standard by clarifying the specific mathematical composition of the wavelength variation coefficient, making the process optimization effect measurable, improving the accuracy of quality control and the efficiency of product consistency judgment, providing a reliable basis for rapid decision-making on the production line, and enhancing the performance comparability between different production batches.

[0032] In one possible implementation, the relationship model is a curve showing the relationship between carrier gas volume and wavelength variation coefficient.

[0033] In practical applications, a relationship model refers to the mathematically established correspondence between carrier gas volume and wavelength variation coefficient, which can intuitively reflect the inherent relationship between process parameters and uniformity indicators in the form of a relationship curve. A relationship curve can be a continuous trajectory depicting the correspondence between carrier gas volume and wavelength variation coefficient in a two-dimensional coordinate system, used to reveal the optimal range and trend of parameter adjustment.

[0034] As a specific example: When optimizing GaAs-based epitaxial processes, five sets of wavelength data under different carrier gas flow rates were obtained through systematic experiments, corresponding to hydrogen flow rates of 3000 standard milliliters per minute, 4000 standard milliliters per minute, 5000 standard milliliters per minute, 6000 standard milliliters per minute, and 7000 standard milliliters per minute, respectively. The corresponding wavelength variation coefficients were measured to be 0.18%, 0.12%, 0.08%, 0.11%, and 0.16%, respectively. Connecting these data points forms a relationship curve with a clear minimum point, showing that the optimal process window is located near 5000 standard milliliters per minute. In actual production control, when a deviation in wavelength uniformity is detected, the operator adjusts the carrier gas flow rate to the optimal range corresponding to the minimum point of the curve according to the relationship curve, so that the process can be quickly restored to the optimal state.

[0035] This invention transforms complex process parameter optimization into intuitive graphical guidance by establishing a visualized relationship curve model, significantly improving the efficiency and accuracy of process debugging. It enables operators to quickly identify the optimal process window and provides a reliable reference for process transfer to different material systems, effectively shortening the new process development cycle and improving production stability.

[0036] In one possible implementation, the optimal carrier gas volume range determined by the optimization control module makes the wavelength variation coefficient less than 0.15 percent.

[0037] In practical applications, the optimization control module can refer to a functional unit that makes process parameter optimization decisions based on a relational model, used to determine the process window that meets specific uniformity requirements. The optimal carrier gas volume range can refer to the range of carrier gas volumes that can stably maintain the wavelength variation coefficient below a predetermined threshold, providing a reliable operating parameter space for the production process. 0.15 percent can refer to a specific threshold limit for the wavelength variation coefficient, serving as a quality standard for judging whether the wavelength uniformity of the epitaxial wafer meets the requirements.

[0038] As a concrete example: During the mass production of InP-based epitaxial wafers, the optimization control module analyzes the established relationship curve between carrier gas volume and wavelength variation coefficient. It determines that when the hydrogen flow rate is maintained within the range of 5500 to 6500 standard milliliters per minute, the corresponding wavelength variation coefficient can be stably kept below 0.15%. In the actual control strategy, the baseline value of the carrier gas volume is set to 6000 standard milliliters per minute, and adjustment is allowed within a range of ±500 standard milliliters per minute. When the online monitoring system detects that the wavelength variation coefficient of a certain batch of products is close to 0.14% and has an upward trend, the control module immediately adjusts the carrier gas volume from the current value to 6200 standard milliliters per minute, so that the wavelength uniformity returns to the optimized state of 0.12%.

[0039] This invention provides a clear and reliable operating guide for the production process by precisely defining the correspondence between the optimal carrier gas volume range and specific uniformity indicators. It effectively avoids the blind setting of parameters, significantly improves the product qualification rate and quality stability, reduces batch-to-batch quality fluctuations caused by parameter drift, and enhances the adaptability and overall efficiency of the production system.

[0040] In one possible implementation, the feedback execution module, in conjunction with the feedback control system, monitors wavelength uniformity in real time and automatically adjusts the carrier gas volume.

[0041] In practical applications, a feedback execution module refers to a functional unit that implements closed-loop control of process parameters, converting monitoring data into control commands to drive the actuators. A feedback control system refers to an automatic adjustment system composed of real-time data acquisition, processing, and execution, capable of maintaining the dynamic stability of process parameters within a target range. Real-time monitoring refers to the continuous and uninterrupted acquisition and analysis of data from the production process to capture instantaneous changes in the process state. Automatic adjustment refers to an operating mode where the system autonomously corrects parameters according to a preset algorithm, replacing manual intervention to achieve precise control.

[0042] As a specific example: During the continuous production of GaN-based LED epitaxial wafers, the feedback control system collects wavelength data from five key points on the wafer every second through in-situ monitoring equipment installed in the reaction chamber. When the system detects that the wavelength variation coefficient has increased from 0.10% to 0.13% for three consecutive sampling cycles, the feedback execution module immediately starts the adjustment program. According to the preset control algorithm, the carrier gas flow rate is increased from the baseline value of 5800 standard milliliters per minute to 6000 standard milliliters per minute. Subsequently, monitoring shows that the wavelength variation coefficient has fallen back to 0.09% and remained stable in the next two sampling cycles. The entire adjustment process is completed within fifteen seconds without interrupting the normal production process.

[0043] This invention significantly improves the adaptive capability and response speed of the process system by realizing closed-loop linkage between monitoring and control, effectively suppresses parameter drift in the production process, ensures the continuous stability of product quality, and greatly reduces the cost of manual monitoring and the incidence of production anomalies, providing a reliable technical guarantee for continuous large-scale production.

[0044] In one possible implementation, the optimal carrier gas volume range is adjusted according to different material systems and process conditions.

[0045] In practical applications, a material system can refer to a category of epitaxial materials composed of specific components and structures, used to determine the required optoelectronic properties and corresponding growth process requirements. Process conditions can refer to a set of adjustable parameters that affect the epitaxial quality during MOCVD growth, which can be comprehensively adjusted to achieve the ideal material growth state.

[0046] As a specific example: when the production system switches from AlGaInP red light material to GaN-based blue light material, the optimization control module adjusts the optimal carrier gas flow rate range from 6200-6800 standard milliliters per minute (SPM) dominated by hydrogen flow rate to 4500-5000 SPM dominated by nitrogen flow rate; within the same material system, when the reaction chamber pressure increases from 100 mbar to 150 mbar, the system automatically shifts the original carrier gas flow rate range by approximately 8% towards the higher flow rate; while maintaining a growth temperature of 1250 degrees Celsius, the system compensates for changes in raw material characteristics by fine-tuning the center value of the carrier gas flow rate range to ensure that wavelength uniformity always meets the requirements, in response to fluctuations in the purity of different batches of source materials.

[0047] This invention significantly enhances the flexibility and adaptability of the production system to handle the production of multiple varieties and specifications by establishing an adaptive matching mechanism between material properties, process parameters and the optimal carrier gas volume range. It effectively solves the problem of process transfer between different material systems, while improving equipment utilization and product yield, and provides strong technical support for a diversified product strategy.

[0048] In one possible implementation, the model building module calculates the wavelength variation coefficient using the following formula: First, calculate the average wavelength. ; in It is the first The wavelength values ​​at each measurement point are acquired by the data acquisition module, where N is the total number of measurement points, determined by the data acquisition module. It is the average wavelength; Then calculate the standard deviation. ; in It is the standard deviation of the wavelength data; Finally, the wavelength variation coefficient was calculated. ; Where CV is the wavelength variation coefficient.

[0049] In practical applications, the model building module refers to the functional unit that implements data processing and model construction, used to transform raw test data into a quantitative model that can guide production. The average wavelength refers to the arithmetic mean of the wavelength values ​​at all valid measurement points on the wafer surface, reflecting the concentration trend of the epitaxial layer's emission wavelength. Measurement points refer to test locations selected on the wafer surface according to a preset distribution, used to obtain spatially representative wavelength sample data. Wavelength value refers to a physical quantity characterizing the luminescence properties of the epitaxial material obtained through spectral testing, directly reflecting the composition and thickness uniformity of the epitaxial layer. The total number of measurement points refers to the number of spatial locations where data is collected in a single test, used to ensure the representativeness and reliability of the statistical results. The standard deviation refers to the dispersion of the wavelength values ​​at each measurement point relative to the average wavelength, quantifying the fluctuation range of the wavelength distribution. The wavelength variation coefficient refers to the dimensionless parameter of the ratio of the standard deviation to the average wavelength, used to eliminate dimensional influences and achieve uniformity comparison between different batches of products.

[0050] As a specific example: When monitoring an 8-inch GaAs epitaxial wafer, the data acquisition module obtains the wafer surface wavelength values ​​using a nine-point test method, recording them as 849.6 nm, 850.2 nm, 849.8 nm, 850.5 nm, 849.9 nm, 850.1 nm, 849.7 nm, 850.3 nm, and 849.5 nm respectively; the model building module first calculates the average wavelength as 850.0 nm, then calculates the standard deviation as 0.32 nm, and finally obtains the wavelength variation coefficient as 0.038%; the system stores the calculation results together with historical data in the database for subsequent updates to the model relating carrier gas volume and wavelength uniformity.

[0051] This invention establishes a standardized uniformity evaluation system by clearly defining the calculation process of the wavelength variation coefficient and the data sources of each parameter, ensuring the repeatability and comparability of the measurement results, providing a reliable data foundation for process optimization, and improving the automation and accuracy of quality monitoring, effectively supporting the quality control needs of large-scale production.

[0052] In one possible implementation, when the model building module builds the relational model, it performs linear fitting using the following formula: The relationship between the wavelength variation coefficient CV and the carrier gas quantity Q is as follows: ; The parameters a and b are obtained using the least squares method, and the calculation formula is: , ; in , Where M is the number of experiments, determined by the data acquisition module. It is the first The carrier gas volume for this experiment was obtained by the data acquisition module. It is the first The wavelength variation coefficient for this experiment was calculated by the model building module. It is the average carrier gas volume. is the average wavelength variation coefficient, and a and b are fitting parameters.

[0053] In practical applications, parameter 'a' can refer to the intercept term in the linear relationship model, reflecting the fundamental influence of factors other than carrier gas volume on wavelength uniformity. Parameter 'b' can refer to the slope coefficient in the linear relationship model, used to characterize the degree of influence of carrier gas volume variation on wavelength uniformity. Least squares method refers to a mathematical optimization method that determines model parameters by minimizing the sum of squared errors, ensuring the best fit between the fitted curve and experimental data. The number of experiments, M, refers to the total number of systematic experiments conducted to establish the model, used to provide sufficient data samples to ensure model reliability. The carrier gas volume in the j-th experiment... This can refer to the specific value of the carrier gas volume set under a particular experimental number, which can be used as input variable data for modeling. The wavelength variation coefficient of the j-th experiment. This can refer to the uniformity results measured for a specific carrier gas volume setting in an experiment, and can be used as output response data for modeling. Average carrier gas volume. This can refer to the arithmetic mean of all experimental carrier gas volume settings, used for centering when calculating fitting parameters. Average wavelength variation coefficient. It can refer to the arithmetic mean of all experimental uniformity measurements, used to characterize the overall process level benchmark. The fitting parameters a and b can refer to the final set of linear model coefficients, which can fully describe the quantitative relationship between carrier gas volume and wavelength uniformity.

[0054] As a specific example: When establishing the GaN epitaxial process model, the data acquisition module conducted five experiments with carrier gas volumes set to 3800, 4000, 4200, 4400, and 4600 standard milliliters per minute, respectively. The corresponding measured wavelength variation coefficients were 0.12%, 0.09%, 0.07%, 0.10%, and 0.13%. The model building module first calculated the average carrier gas volume as 4200 standard milliliters per minute and the average wavelength variation coefficient as 0.102%. Then, it used the least squares method to calculate the slope parameter b as 0.00002 standard milliliters per minute and the intercept parameter a as 0.014%. Finally, the complete relational model CV = 0.014% + 0.00002Q was obtained. This model showed that the optimal uniformity was achieved with a carrier gas volume of 4250 standard milliliters per minute. In actual production, the carrier gas volume was controlled within the range of 4150-4350 standard milliliters per minute.

[0055] This invention establishes a precise linear relationship model, which enables the quantitative correlation between process parameters and quality indicators. This significantly improves the scientific nature and predictability of process optimization, provides reliable theoretical guidance for the production process, enhances the system's adaptability to different process conditions, and effectively improves product quality stability and process controllability.

[0056] In one possible implementation, the feedback execution module dynamically adjusts the carrier gas volume based on real-time monitored wavelength uniformity data, so that the wavelength variation coefficient is maintained within the preset target value range.

[0057] In practical applications, the feedback execution module refers to the functional unit that implements closed-loop control of process parameters, converting real-time monitoring data into control commands and driving the actuators. Real-time monitored wavelength uniformity data refers to the data stream continuously acquired by an online detection system, reflecting the wavelength distribution characteristics of the epitaxial wafer, providing instantaneous feedback information on the process status. Dynamic adjustment refers to an operating mode that continuously optimizes process parameters based on real-time data changes to maintain a stable production process. The preset target value range refers to the allowable fluctuation range of the wavelength variation coefficient pre-set according to product quality requirements, serving as a benchmark for process control.

[0058] As a specific example: During the continuous production of InGaAlP epitaxial wafers, the feedback execution module acquires wavelength distribution data of 49 points across the entire wafer every 30 seconds through an integrated spectral monitoring system. When the wavelength variation coefficient is detected to gradually increase from 0.08% to 0.12% and approach the preset target upper limit of 0.15%, the control algorithm immediately initiates a correction program. Based on the established model of the relationship between carrier gas volume and uniformity, the hydrogen flow rate is gradually increased from the initial 5500 standard milliliters per minute to 5800 standard milliliters per minute. At the same time, the monitoring system continuously provides feedback on the uniformity changes after the adjustment until the wavelength variation coefficient falls back to the ideal range of 0.09% and remains stable. The entire adjustment process is completed within five minutes without interrupting the normal production process.

[0059] This invention significantly improves the production system's ability to suppress process fluctuations by realizing closed-loop control with real-time monitoring and dynamic adjustment, ensuring the continuous stability of product quality. At the same time, it greatly reduces the intensity of manual intervention and the incidence of product anomalies, providing a reliable technical guarantee for high-efficiency and high-consistency large-scale production.

[0060] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments in this specification are not limited to the described order of actions, because according to the embodiments in this specification, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments in this specification.

[0061] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0062] The preferred embodiments disclosed above are merely illustrative of this specification. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments described herein. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the embodiments, thereby enabling those skilled in the art to better understand and utilize this specification. This specification is limited only by the claims and their full scope and equivalents.

Claims

1. A MOCVD wavelength uniformity optimization system based on carrier gas volume control, characterized in that, The method comprises the following steps: A data acquisition module is used to set different carrier gas amounts in a MOCVD device and perform experiments, and to perform spectral testing on the grown epitaxial layer to obtain wavelength data of each position in the wafer; A model establishment module is used to calculate a wavelength variation coefficient based on the wavelength data and to establish a relationship model between the carrier gas amount and the wavelength variation coefficient; An optimization control module is used to determine an optimal carrier gas amount range that makes the wavelength variation coefficient less than a preset target value according to the relationship model; A feedback execution module is used to dynamically adjust the carrier gas amount according to the optimal carrier gas amount range in the MOCVD production process. 2.The carrier gas amount-based MOCVD wavelength uniformity optimization processing system according to claim 1, wherein, The carrier gas amount includes nitrogen flow and hydrogen flow. 3.The carrier gas amount-based MOCVD wavelength uniformity optimization processing system according to claim 1, wherein, The wavelength variation coefficient is obtained by dividing the standard deviation by the average wavelength.

4. The MOCVD wavelength uniformity optimization processing system based on carrier gas flow regulation of claim 1, wherein, The relationship model is a relationship curve between the carrier gas amount and the wavelength variation coefficient.

5. The carrier gas amount-based control MOCVD wavelength uniformity optimization processing system according to claim 1, wherein, The optimal carrier gas amount range determined by the optimization control module makes the wavelength variation coefficient less than 0.15%.

6. The carrier gas amount-based control MOCVD wavelength uniformity optimization processing system according to claim 1, wherein The feedback execution module is combined with a feedback control system to monitor the wavelength uniformity in real time and automatically adjust the carrier gas amount.

7. The carrier gas amount-based control MOCVD wavelength uniformity optimization processing system according to claim 1, wherein The optimal carrier gas amount range is adjusted according to different material systems and process conditions.

8. The carrier gas amount-based control MOCVD wavelength uniformity optimization processing system according to claim 1, wherein, The model establishing module calculates the wavelength variation coefficient by the following formula: first, calculate the average wavelength ; wherein is the wavelength value of the nth measurement point, obtained by the data acquisition module, and N is the total number of measurement points, determined by the data acquisition module, is the average wavelength; Then calculate the standard deviation ; wherein is the standard deviation of the wavelength data; Finally the coefficient of variation of the wavelength is calculated ; Wherein CV is the wavelength variation coefficient.

9. The carrier gas amount-based control MOCVD wavelength uniformity optimization processing system according to claim 1, wherein, When the model establishment module establishes the relationship model, linear fitting is performed by the following calculation formula: The wavelength variation coefficient CV is related to the carrier gas amount Q by ; wherein the parameters a and b are obtained by least square method, and the calculation formula is , ; wherein , where M is the number of experiments, determined by the data acquisition module, is the amount of carrier gas for the experiment, obtained by the data acquisition module, is the wavelength variation coefficient for the experiment, calculated by the model establishment module, is the average amount of carrier gas, is the average wavelength variation coefficient, and a and b are fitting parameters.

10. The carrier gas amount-based control MOCVD wavelength uniformity optimization processing system according to claim 1, wherein, The feedback execution module dynamically adjusts the carrier gas amount according to the real-time monitored wavelength uniformity data, so that the wavelength variation coefficient is maintained within the range of the preset target value.

Images