Intelligent semiconductor wafer spraying control method and system

Abstract

The application provides a kind of intelligent semiconductor wafer spray control method and system, comprising: according to the material and surface roughness of the wafer, obtain the recommended cleaning time range of the wafer from the preset spray time-wafer attribute correlation database;The sub-area spray parameters obtained in the first three steps are integrated to form a wafer surface cleaning scheme, the cleaning scheme includes the nozzle layout, nozzle quantity, spray liquid composition and temperature of each sub-area, and the cleaning scheme is transmitted to the spray control unit;After cleaning, compare the wafer surface image with the preset standard image, calculate the similarity of the two, if the similarity is lower than the preset threshold, the number and cleaning parameters of the wafer are recorded in the abnormal cleaning log, and the wafer is transferred to the unqualified product buffer area, the spray control of the application improves the cleaning efficiency and quality, reduces the waste of resources, and provides an advanced cleaning solution for semiconductor manufacturing.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, in particular to the field of semiconductor wafers, and more particularly to an intelligent semiconductor wafer spraying control method and system. BACKGROUND

[0002] In the wafer cleaning process, there is a technical contradiction between real-time monitoring of spraying pressure and flow and automatic adjustment of spraying parameters according to the wafer surface contamination level. On the one hand, in order to ensure the cleaning effect, it is necessary to divide sub-regions according to the contamination distribution and pattern structure characteristics of the wafer surface, and to select the best spraying parameters for each sub-region by using clustering and decision tree algorithms. This requires the system to be able to flexibly adjust the spraying pressure, flow and angle parameters. On the other hand, in order to ensure the stability and controllability of the cleaning process, it is necessary to monitor the actual operating parameters of the spraying device in real time and compare them with the preset values. When the monitoring values deviate, the system needs to dynamically adjust the nozzle opening to ensure that the parameters meet the requirements. However, frequent adjustment of the spraying parameters may affect the continuity and uniformity of the cleaning, and even cause some areas to be insufficiently cleaned. How to ensure the cleaning effect while maintaining the relative stability of the spraying parameters has become a technical problem that needs to be solved. In addition, how to organically combine real-time monitoring data with pollution analysis results to establish a dynamically optimized cleaning parameter model to adapt to different types and levels of wafer contamination is also an important challenge faced by the system. SUMMARY

[0003] The present application provides an intelligent semiconductor wafer spraying control method, mainly comprising:

[0004] Obtain the surface contamination distribution map of the wafer to be cleaned, divide the wafer surface contamination area by using a clustering algorithm, obtain several sub-regions with different contamination levels, and for each sub-region, obtain the corresponding spraying pressure and flow parameters from a preset spraying parameter database;

[0005] Obtain the wafer surface pattern structure data, divide the wafer surface into sub-regions by using a decision tree algorithm, obtain several sub-regions with different pattern structure complexity, and for each sub-region, obtain the corresponding spraying angle parameters from the preset spraying parameter database;

[0006] According to the material and surface roughness of the wafer, obtain the recommended cleaning time range of the wafer from a preset spraying time-wafer attribute correlation database;

[0007] Integrate the sub-region spraying parameters obtained in the previous three steps to form a wafer surface cleaning scheme, the cleaning scheme including the nozzle layout, nozzle number, spraying liquid composition and temperature of each sub-region, and transmit the cleaning scheme to a spraying control unit;

[0008] The spray control unit controls the nozzles of the spray device to move on the wafer surface along a preset trajectory and time interval according to the received cleaning scheme, and monitors the pressure and flow parameters of each nozzle in real time. If the monitored values ​​deviate from the set values ​​of the cleaning scheme, the nozzle opening is dynamically adjusted;

[0009] During the wafer cleaning process, a machine vision system is used to collect images of the wafer surface in real time, process and analyze the images, and calculate the amount of residual contaminants on the wafer surface. If the residual amount is higher than a preset threshold, the wafer is re-transferred to the spray device for further cleaning.

[0010] After cleaning is completed, the wafer surface image is compared with a preset standard image and the similarity between the two is calculated. If the similarity is lower than a preset threshold, the wafer number and cleaning parameters are recorded in the abnormal cleaning log, and the wafer is transferred to the defective product buffer area.

[0011] For the wafers that have passed the cleaning, their actual cleaning parameters are uploaded to the spray parameter optimization module. The optimization module trains the historical accumulated cleaning data based on the support vector machine algorithm to obtain a dynamically updated spray parameter-cleaning effect mapping model, which is used to optimize and predict the subsequent cleaning parameters of new wafers.

[0012] The present invention provides an intelligent semiconductor wafer spray control system, which mainly includes:

[0013] Contamination analysis module, used to obtain the contamination distribution on the wafer surface and divide it into sub-areas;

[0014] Graphic structure analysis module, used to obtain the graphic structure of the wafer surface and divide it into sub-areas;

[0015] A cleaning parameter acquisition module is used to obtain a recommended cleaning time range based on wafer properties;

[0016] Cleaning plan generation module, used to integrate parameters to form a wafer surface cleaning plan;

[0017] Cleaning control module, used to control the spray device to perform cleaning and monitor and adjust in real time;

[0018] The quality detection module is used to monitor the cleaning effect in real time and perform subsequent processing.

[0019] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0020] The present invention discloses an intelligent semiconductor wafer spray control method. This method obtains contamination distribution maps and graphical structure data on the wafer surface, partitions the wafer surface using clustering and decision tree algorithms, and retrieves corresponding spray parameters from a preset database. Recommended cleaning times are determined based on the wafer material and surface roughness, and integrated into a cleaning plan. During the cleaning process, a spray control unit controls nozzle movement and adjusts parameters in real time according to the plan. A machine vision system monitors residual contaminants. After cleaning, image comparison is used to determine conformity. Cleaning data is then used for optimization model training. This method achieves intelligent and refined control of the wafer cleaning process. The adjustable spray design ensures spray uniformity and reliability, improves cleaning efficiency and quality, reduces resource waste, and provides an advanced cleaning solution for semiconductor manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of an intelligent semiconductor wafer spray control method of the present invention.

[0022] Figure 2 This is a schematic diagram of the framework structure of an intelligent semiconductor wafer spray control system of the present invention. DETAILED DESCRIPTION

[0023] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0024] like Figure 1-Figure 2 In this embodiment, an intelligent semiconductor wafer spray control method may specifically include:

[0025] Step S101, obtain the surface contamination distribution map of the wafer to be cleaned, use a clustering algorithm to divide the contaminated area of ​​the wafer surface to obtain several sub-areas with different contamination levels, and for each sub-area, obtain the corresponding spray pressure and flow parameters from a preset spray parameter database.

[0026] An image of a surface contamination distribution of a wafer to be cleaned is acquired. The image of the surface contamination distribution is preprocessed to obtain a preprocessed image of the surface contamination distribution. A K-means clustering algorithm is used to perform clustering analysis on the preprocessed image of the surface contamination distribution to obtain a plurality of sub-regions of different contamination levels. A preset spraying parameter database is acquired, wherein the spraying parameter database stores spraying pressure and flow parameters corresponding to different contamination levels. According to the contamination levels of the plurality of sub-regions, the spraying pressure and flow parameters corresponding to each sub-region are acquired from the spraying parameter database. The wafer is placed in a cleaning device, and the spraying head of the cleaning device is controlled to perform differential cleaning on each sub-region according to the acquired spraying pressure and flow parameters, to obtain a cleaned wafer. The surface of the cleaned wafer is detected to determine whether a preset cleanliness requirement is met. If the cleanliness requirement is not met, the cleaned wafer is re-placed in the cleaning device for cleaning. If the cleanliness requirement is met, the cleaned wafer is sent to a next process.

[0027] Specifically, an image of a surface contamination distribution of a wafer to be cleaned is acquired. The image can be acquired by a dark field microscope imaging technology. The working principle of the dark field microscope is to tilt the light source to the sample surface, only the scattered light can enter the objective lens, so that the contaminants appear bright in the dark background, and the small contamination particles on the wafer surface can be clearly displayed. For example, a 12-inch silicon wafer is imaged by a dark field microscope with a wavelength of 500 nm LED light source and a magnification of 500 times, a high-resolution contamination distribution image can be obtained, and the higher the brightness in the image represents the higher the concentration of contaminants. The reason for using the dark field microscope is that it has a better imaging effect on small particles, and can more accurately reflect the contamination condition of the wafer surface.

[0028] The contamination distribution image is preprocessed to remove noise and interference. Image smoothing and filtering techniques can be used to remove noise from the image. For example, using a Gaussian filter to smooth the wafer surface contamination distribution image effectively removes random noise. Morphological operations, such as erosion and dilation, can also be used to remove interference from the image, such as dust or scratches. Suppose a small area on the wafer surface is mistakenly identified as contamination, but is actually caused by light reflection. Erosion can remove this misidentified area. The purpose of image preprocessing is to improve the accuracy of subsequent cluster analysis. The K-means clustering algorithm is used to perform cluster analysis on the preprocessed contamination distribution image, dividing the contaminated area into several sub-regions with varying degrees of contamination. The K-means clustering algorithm is an unsupervised learning algorithm that divides data points into K clusters, with the center of each cluster being called the centroid. The brightness value of each pixel in the preprocessed image is used as a feature vector to divide the pixels into K clusters, each representing a sub-region with varying degrees of contamination. For example, setting the K value to 3 divides the contaminated area on the wafer surface into three sub-regions: lightly contaminated, moderately contaminated, and heavily contaminated. The K-means algorithm was chosen for its simplicity and efficiency, effectively classifying contaminated areas. Based on the contamination level of each sub-region, the corresponding spray pressure and flow rate parameters are retrieved from a pre-set spray parameter database.

[0029] A preset spray parameter database stores the optimal spray pressure and flow rates for different levels of contamination. For example, the spray pressure for lightly contaminated areas is 1 MPa and the flow rate is 1 L / min; the spray pressure for moderately contaminated areas is 2 MPa and the flow rate is 2 L / min; and the spray pressure for heavily contaminated areas is 3 MPa and the flow rate is 3 L / min. The parameters in the database are derived from extensive experimental data to ensure optimal cleaning results. The purpose of establishing this database is to achieve differentiated cleaning, improve cleaning efficiency, and minimize wafer damage. The wafer is placed in the cleaning equipment, and the spray head is controlled based on the acquired spray parameters to perform differentiated cleaning on the different contaminated sub-areas. For example, for each of the three sub-areas, the spray head will apply a pressure of 3 MPa and a flow rate of 3 L / min to the heavily contaminated area, a pressure of 2 MPa and a flow rate of 2 L / min to the moderately contaminated area, and a pressure of 1 MPa and a flow rate of 1 L / min to the lightly contaminated area, according to the preset parameters. This differentiated cleaning method can precisely clean areas with different levels of contamination, avoiding wafer damage caused by excessive cleaning and improving cleaning efficiency.

[0030] After cleaning, the wafer surface is inspected to determine whether it meets the preset cleanliness requirements. A darkfield microscope or other surface inspection equipment can be used to inspect the cleaned wafers. For example, the number and size of residual particles on the wafer surface can be measured and compared to a preset cleanliness standard. The cleanliness standard can be set as less than 10 particles per square centimeter and a maximum particle size of less than 1 micron. Inspection is performed to ensure that the cleaning effect meets the requirements and prevent unqualified wafers from being transferred to the next process. If the cleanliness requirements are not met, the wafer is returned for re-cleaning until they are met. If the inspection results indicate that the wafer surface cleanliness does not meet the requirements, for example, if residual particles still remain in heavily contaminated areas, the spray parameters are adjusted, such as increasing the spray pressure or extending the cleaning time, and then the cleaning process is repeated. This cycle continues until the wafer surface meets the preset cleanliness requirements. If the requirements are met, the cleaned wafer is sent to the next process. If the inspection results indicate that the wafer surface cleanliness meets the requirements, the cleaned wafer can be sent to the next process, such as photolithography or etching. This ensures the smooth progress of subsequent processes and ultimately produces high-quality chips.

[0031] Step S102, obtaining the graphic structure data of the wafer surface, partitioning the wafer surface using a decision tree algorithm to obtain several sub-regions with different graphic structure complexities, and obtaining corresponding spray angle parameters from the preset spray parameter database for each sub-region.

[0032] The wafer surface's graphic structure data is acquired and preprocessed to extract graphic structure features. Based on the extracted graphic structure features, a decision tree algorithm is used to partition the wafer surface into several sub-regions with varying degrees of graphic structure complexity. For each sub-region, the corresponding spray angle parameters are retrieved from a pre-set spray parameter database based on the complexity of its graphic structure. Based on the acquired spray angle parameters, the spray angle for each sub-region is determined, obtaining spray angle data for each sub-region on the wafer surface. The spray angle data for each sub-region on the wafer surface is input into the spray equipment, which controls the spray equipment to perform differentiated spray treatment on each sub-region on the wafer surface. During the spray treatment process, image data of the wafer surface is collected in real time and analyzed to determine whether the spray effect meets expectations. If the spray effect does not meet expectations, the spray angle parameters are adjusted based on the image analysis results, and the adjusted parameters are updated to the spray parameter database to optimize the subsequent wafer spray treatment process.

[0033] Specifically, wafer surface structure data can be acquired using instruments such as optical microscopes and scanning electron microscopes. For example, scanning the wafer surface with a scanning electron microscope produces high-resolution surface topography images, which contain information about various surface structures, such as lines, grooves, and holes. The acquisition of structure data is necessary because different structures have varying effects on the flow and coverage of the cleaning solution, thus affecting cleaning effectiveness. The acquired structure data often contains a significant amount of noise and redundant information, requiring preprocessing. For example, image smoothing and filtering techniques, such as Gaussian filtering or median filtering, can be used to remove noise from the image. Edge detection algorithms, such as the Canny operator or Sobel operator, can also be used to extract edge information about the structure. The purpose of preprocessing is to improve the accuracy and efficiency of subsequent feature extraction. For example, consider a long, narrow scratch on the wafer surface. This scratch is noise and needs to be removed using image processing techniques. Structure features can be extracted from the preprocessed structure data. For example, features such as structure density, complexity, and directionality can be extracted. Structure density can be defined as the proportion of structure per unit area. Complexity can be defined as the total length of the edges of a graph structure. Directionality can be defined as the dominant direction of the graph structure. These features can be used to describe the complexity of a graph structure. For example, areas with dense lines and complex directions are considered to have high complexity. The purpose of extracting these features is to provide a basis for subsequent partitioning.

[0034] A decision tree algorithm is used to partition the wafer surface. This tree-based classification algorithm can divide the wafer surface into several sub-regions of varying complexity based on graph structure characteristics. For example, the wafer surface can be divided into low-complexity, medium-complexity, and high-complexity regions based on the density and complexity of the graph structure. The decision tree algorithm was chosen because it can effectively process multi-dimensional feature data and offers good interpretability. For example, suppose one portion of the wafer surface is densely packed with circuit patterns, while another portion is relatively sparse with connecting lines. The decision tree algorithm can divide these two regions into different regions based on the density of the patterns. For each sub-region, the corresponding spray angle parameter is retrieved from a pre-set spray parameter database based on the complexity of the graph structure. The pre-set spray parameter database stores the optimal spray angle parameters for different levels of graph structure complexity. For example, the spray angle for the low-complexity region is vertically downward, the spray angle for the medium-complexity region is 45 degrees, and the spray angle for the high-complexity region is 30 degrees. Different spray angles are used to adjust the flow and coverage of the cleaning liquid across the wafer surface, thereby improving cleaning efficiency. The parameters in the database are derived from extensive experimental data to ensure optimal cleaning results. For highly complex areas, using a smaller spray angle allows the cleaning liquid to better penetrate tiny grooves and holes.

[0035] Based on the acquired spray angle parameters, the spray angle for each sub-area is determined, generating spray angle data for each sub-area on the wafer surface. For example, for the three sub-areas, spray angles corresponding to vertical downward, 45 degrees, and 30 degrees are obtained. These angle data are stored in a data structure for subsequent use by the spray equipment. The spray angle data for each sub-area on the wafer surface is input into the spray equipment, controlling the spray equipment to perform differentiated spray processing on each sub-area on the wafer surface. The spray equipment adjusts the position and angle of the spray head based on the input spray angle data to provide differentiated spraying for different sub-areas. For example, for highly complex areas, the spray head may spray at a 30-degree angle. This differentiated spraying approach enables precise cleaning of areas with varying levels of structural complexity, improving cleaning efficiency and minimizing wafer damage. During the spray process, real-time image data of the wafer surface is collected. High-speed cameras or other imaging devices can be used to capture real-time image data of the wafer surface during the spray process. This image data can be used to monitor the cleaning process and evaluate spray effectiveness. The image data is analyzed to determine whether the spray effect meets expectations. For example, the spray effect can be judged by analyzing the number and size of residual contaminants in the image. When the number of residual contaminants exceeds a preset threshold, it is considered that the spray effect has not met expectations. If the spray effect does not meet expectations, the spray angle parameters are adjusted based on the image analysis results. For example, if it is found that the cleaning effect of a highly complex area is not ideal, you can try to reduce the spray angle, such as from 30 degrees to 20 degrees, or increase the spray time. The adjusted parameters are updated to the spray parameter database to optimize the spray processing process of subsequent wafers. Updating the adjusted spray angle parameters to the database can achieve better cleaning effects on subsequent wafers with the same graphic structure. This is a data-driven optimization method that can continuously improve the cleaning process.

[0036] Step S103 : obtaining a recommended cleaning time range for the wafer from a preset spray time-wafer attribute association database according to the material and surface roughness of the wafer.

[0037] Obtain the material information and surface roughness value of the wafer and use them as query conditions. Access the preset spray time-wafer attribute association database and perform a matching query based on the material and roughness value. If a matching record is found in the association database, extract the corresponding recommended cleaning time range. Determine the target cleaning time for the wafer based on the obtained recommended cleaning time range. Pass the target cleaning time to the spray control module, and clean the wafer for the corresponding length of time. After cleaning is completed, perform a roughness test on the wafer surface to obtain the roughness value after cleaning. Compare the roughness value after cleaning with the preset threshold to determine whether the cleaning effect meets the standard, so as to determine whether secondary cleaning or adjustment of cleaning parameters is required.

[0038] Step S104, the sub-area spraying parameters obtained in the previous three steps are integrated to form a wafer surface cleaning scheme, the cleaning scheme including the nozzle layout, nozzle quantity, spraying liquid composition and temperature of each sub-area, and the cleaning scheme is transmitted to the spraying control unit.

[0039] Obtain the contamination level data and material characteristic data of the wafer surface, determine the spraying parameters of each sub-area of the wafer surface according to the contamination level data and material characteristic data, the spraying parameters including the nozzle layout, nozzle quantity, spraying liquid composition and temperature, integrate the spraying parameters of each sub-area to obtain a wafer surface cleaning scheme, use a machine learning algorithm to optimize each parameter in the cleaning scheme according to the preset historical cleaning data and current wafer surface condition data to obtain an optimized cleaning scheme, convert the optimized cleaning scheme into control instructions and transmit the control instructions to the spraying control unit, obtain the real-time spraying parameters and wafer surface condition data of each sub-area during the cleaning process, if the real-time spraying parameters and wafer surface condition data do not conform to the preset range, dynamically adjust the cleaning scheme, obtain the cleaning effect data of the wafer surface after the cleaning is completed, input the cleaning effect data into the machine learning algorithm, train the machine learning algorithm to obtain an optimized machine learning algorithm for the optimization of subsequent cleaning schemes.

[0040] Step S105, the spraying control unit controls the nozzles of the spraying device to move on the wafer surface at a preset trajectory and time interval according to the received cleaning scheme, and monitors the pressure and flow parameters of each nozzle in real time, if the monitoring values deviate from the set values of the cleaning scheme, the nozzle opening degree is dynamically adjusted.

[0041] At step S1051, the spray control unit receives the cleaning plan, obtains the preset nozzle trajectory and time interval parameters, and controls the nozzle movement across the wafer surface based on these parameters. At step S1052, during nozzle movement, the nozzle pressure and flow rate data are collected in real time and compared with the pressure and flow rate thresholds set in the cleaning plan. At step S1053, if the comparison results indicate that the real-time pressure or flow rate data deviates from the set thresholds, an abnormality is determined, triggering the nozzle opening adjustment mechanism. At step S1054, the nozzle opening adjustment value is dynamically calculated based on the direction and amount of pressure deviation, and the nozzle opening is adjusted via a control signal to restore the pressure within the set range. At step S1055, the nozzle opening adjustment value is dynamically calculated based on the direction and amount of flow rate deviation, and the nozzle opening is adjusted via a control signal to restore the flow rate within the set range. Within a cleaning cycle, steps S1052-S1055 are executed continuously until the cleaning cycle ends, completing the cleaning operation on the wafer surface. The pressure and flow rate monitoring data for a cleaning cycle are recorded and compiled to form a cleaning quality report for evaluating cleaning results and optimizing the cleaning plan.

[0042] Specifically, the spray control unit plays a vital role in the wafer surface cleaning process. It receives instructions from the cleaning plan and accurately controls the movement trajectory and time interval of the nozzle to ensure that the cleaning liquid is evenly sprayed on the wafer surface. For example, the cleaning plan stipulates that the nozzle moves in a spiral trajectory with a time interval of 5 seconds between each circle. The spray control unit will accurately control the nozzle to execute according to this plan. The setting of the nozzle trajectory and time interval parameters needs to take into account factors such as the size of the wafer, the distribution of contaminants, and the characteristics of the cleaning liquid. For example, for a wafer with a diameter of 300 mm, there are usually more contaminants in the edge area, so the nozzle can be set to move slower in the edge area and stay longer to ensure the cleaning effect of the edge area.

[0043] During nozzle movement, real-time nozzle pressure and flow data must be collected. This data reflects the operating status of the spray system and is crucial for ensuring cleaning quality. For example, pressure sensors and flow meters can be used to measure nozzle pressure and flow, respectively, and transmit the data to the spray control unit. The collected real-time data must be compared with the pressure and flow thresholds specified in the cleaning plan. The thresholds must be set based on factors such as the cleaning fluid characteristics, nozzle type, and wafer surface material. For example, for highly corrosive cleaning fluids, a lower pressure threshold may be required to prevent corrosion of the wafer surface. If the comparison results show that the real-time pressure or flow data exceeds the set threshold, it indicates an abnormality in the spray system and requires immediate adjustment. For example, if the real-time pressure exceeds the 2 MPa threshold, the nozzle opening must be reduced to lower the pressure. The nozzle opening adjustment requires dynamic calculation based on the direction and amount of pressure or flow deviation. For example, if the real-time pressure falls below the set value by 1 MPa, the nozzle opening must be increased to increase the pressure. The specific adjustment values ​​are calculated using a PID control algorithm to ensure that the pressure or flow returns quickly and stably to the set range.

[0044] During a cleaning cycle, data collection, comparison, and adjustment steps must be performed in a continuous loop until the cycle completes. The cleaning cycle setting needs to take into account factors such as the wafer's contamination level and the efficiency of the cleaning fluid. For example, for lightly contaminated wafers, the cleaning cycle can be set to 1 minute, while for heavily contaminated wafers, the cleaning cycle needs to be extended to 3 minutes or longer. After the cleaning cycle, the pressure and flow monitoring data for the cycle are recorded and compiled into a cleaning quality report. This report can be used to evaluate cleaning effectiveness and provide a reference for optimizing the cleaning solution. For example, if the cleaning quality report indicates that the cleaning pressure in a certain area is consistently below the set value, the nozzle layout can be adjusted, and the number of nozzles in that area can be increased to improve cleaning effectiveness. Through continuous monitoring, adjustment, and optimization, an intelligent, adaptive wafer surface cleaning solution can be established to improve production efficiency and product yield.

[0045] Step S106, during the wafer cleaning process, the wafer surface image is collected in real time by a machine vision system, the image is processed and analyzed, and the current residual amount of pollutants on the wafer surface is calculated. If the residual amount is higher than a preset threshold, the wafer is re-transferred to the spray device for further cleaning.

[0046] A real-time image of the wafer surface is acquired through a machine vision system and transmitted to the image processing module. In the image processing module, the acquired wafer surface image is preprocessed, including image denoising, enhancement, and correction operations, to improve image quality. Based on the preprocessed wafer surface image, an image segmentation algorithm is used to segment the image and extract the contaminant area on the wafer surface. For the segmented contaminant area, its area ratio is calculated to obtain the current contaminant residue on the wafer surface. The calculated contaminant residue is compared with the preset threshold to determine whether the wafer surface meets the cleaning requirements. If the contaminant residue is higher than the preset threshold, the conveyor device is controlled to re-convey the wafer to the spray device for re-cleaning. If the contaminant residue is lower than the preset threshold, the wafer surface is judged to be qualified for cleaning and the wafer is conveyed to the next process for subsequent processing.

[0047] Specifically, during the wafer manufacturing process, machine vision systems are used to monitor wafer surface cleanliness in real time. For example, a high-resolution camera captures images of the wafer surface, which are then transmitted to the image processing module. Within the image processing module, image denoising is first performed. This step uses software algorithms to reduce random noise in the image and improve image usability. Next, image enhancement adjusts contrast and brightness to enhance image detail for easier processing. Image correction corrects for distortion caused by camera angle or lighting variations, ensuring the resulting image accurately reflects the wafer's condition. The processed image is then processed using an image segmentation algorithm. This algorithm identifies contaminated areas within the image. This often involves complex edge detection techniques to accurately separate contaminants from the wafer itself. After contaminants are identified, the system calculates the area percentage of these areas. This is determined by analyzing the ratio of contaminated areas to the total wafer surface in the segmented image. If this percentage exceeds a factory-set threshold, the wafer surface cleanliness does not meet production standards and requires re-cleaning. At this time, the control system will instruct the conveyor to send the wafer back to the spray device for further cleaning to ensure that each wafer meets the cleaning standards required for production.

[0048] Once a wafer passes cleaning, meaning the residual contaminant level falls below a preset threshold, it is transferred to the next process. This series of operations not only ensures product quality but also improves production efficiency. Precise control of the cleaning process reduces the number of wafers requiring repeated cleaning, saving both cost and time. This automated inspection and cleaning system, powered by image processing, demonstrates the application of advanced technology in modern semiconductor manufacturing, ensuring high product quality and efficient production.

[0049] Step S107: After cleaning is completed, the wafer surface image is compared with a preset standard image to calculate the similarity between the two. If the similarity is lower than a preset threshold, the wafer number and cleaning parameters are recorded in the abnormal cleaning log, and the wafer is transferred to the unqualified product buffer area.

[0050] After wafer cleaning, a surface image of the wafer is acquired, converted to a grayscale image, and image enhancement is performed to obtain a processed surface image. A standard image matching the current wafer type is obtained from a pre-set standard image library and the standard image is subjected to the same image enhancement process. A feature extraction algorithm, such as SIFT or SURF, is used to extract key feature points from the processed surface image and the standard image. A feature point matching algorithm, such as RANSAC, is used to match feature points between the wafer surface image and the standard image. The ratio of matching point pairs to the total number of feature points is calculated to obtain a similarity value. The calculated similarity value is compared with a pre-set similarity threshold. If the similarity value falls below the threshold, the wafer is deemed to have failed cleaning quality. If the wafer cleaning quality fails, the wafer number and corresponding cleaning parameters are recorded in the abnormal cleaning log database, and the wafer is transferred to a failed product buffer for further processing. If the wafer cleaning quality passes, the wafer is transferred to the next step for further processing, and wafer processing status information is updated to facilitate real-time tracking of wafer production progress.

[0051] Specifically, after wafer cleaning is complete, its cleaning quality needs to be evaluated. This solution uses a machine vision-based image comparison method to determine whether the cleaning quality meets standards by comparing the image of the wafer under test with a standard image. First, an image of the cleaned wafer surface is acquired. For example, a high-resolution industrial camera is used to capture the wafer surface, producing a 1920x1080 pixel color image. The color image is then converted to a grayscale image for subsequent processing. This reduces computational effort, and the grayscale image better characterizes contaminants on the wafer surface. Image enhancement is performed to highlight surface features. For example, histogram equalization can be used to increase image contrast and make contaminant areas more visible. Next, a standard image matching the wafer type is selected from a standard image library. The library contains pre-stored images of ideal cleaning conditions for various wafer types. For example, for a 12-inch silicon wafer, the corresponding standard image is selected. The standard image also undergoes the same image enhancement processing as the wafer under test, ensuring consistency with the image of the wafer under test and avoiding comparison errors caused by different processing methods. A feature extraction algorithm, such as the Scale-Invariant Feature Transform (SIFT), is used to extract key feature points from the processed wafer surface image and the standard image. The SIFT algorithm extracts stable feature points from images that are unaffected by image rotation, scaling, and lighting changes. For example, 500 feature points were extracted from the wafer surface image and 600 from the standard image.

[0052] Then, a feature point matching algorithm, such as the RANSAC (Random Sample Consensus) algorithm, is used to match the feature points of the wafer surface image with the feature points of the standard image. The RANSAC algorithm can effectively eliminate mismatched points and improve matching accuracy. Assume that 400 feature points are matched. Calculate the ratio of the number of matching point pairs to the total number of feature points to obtain the similarity value. In this example, the similarity value is 400 / (500+600)≈34%. Compare the calculated similarity value with the preset similarity threshold. Assume that the preset similarity threshold is 80%. Since 34% is less than 80%, the wafer cleaning quality is judged to be unqualified. When the wafer cleaning quality is unqualified, the wafer number (for example, SN123456789) and the corresponding cleaning parameters (for example, cleaning time 5 minutes, cleaning liquid concentration 10%) are recorded in the abnormal cleaning log database for subsequent analysis of the causes of problems in the cleaning process and for improvement.

[0053] At the same time, the wafer is transferred to the defective product buffer area, awaiting further processing, such as re-cleaning or scrapping. When the wafer cleaning quality is qualified, the wafer is transferred to the next process (for example, the photolithography process) for subsequent processing. At the same time, the wafer processing status information (for example, cleaning completed, waiting for photolithography) is updated to track the wafer production progress in real time and improve production efficiency.

[0054] In step S108, for the wafers that have passed the cleaning process, their actual cleaning parameters are uploaded to the spray parameter optimization module. The optimization module trains the historically accumulated cleaning data based on the support vector machine algorithm to obtain a dynamically updated spray parameter-cleaning effect mapping model, which is used to optimize and predict the cleaning parameters of subsequent new wafers.

[0055] The actual cleaning parameters of wafers that have passed the cleaning process are obtained and uploaded to the spray parameter optimization module. The spray parameter optimization module obtains historical accumulated cleaning data and uses the support vector machine algorithm to train the cleaning data. Through the training of the support vector machine algorithm, a dynamically updated spray parameter-cleaning effect mapping model is obtained. When a new wafer needs to be cleaned, the initial cleaning parameters of the new wafer are input into the spray parameter-cleaning effect mapping model. The spray parameter-cleaning effect mapping model predicts the optimized cleaning parameters based on the input initial cleaning parameters. The optimized cleaning parameters are applied to the cleaning process of the new wafer and the new wafer is cleaned. After the cleaning is completed, the actual cleaning parameters and cleaning effect of the new wafer are obtained and added to the historical cleaning data for subsequent model updates and optimization.

[0056] The above embodiment is only one of the preferred implementation methods of the present invention and should not be used to limit the scope of protection of the present invention. Any changes or modifications that have no substantive meaning made to the main design concept and spirit of the present invention, as long as the technical problems they solve are still consistent with the present invention, should be included in the scope of protection of the present invention.

Claims

1. An intelligent semiconductor wafer spray control method, characterized in that: The method comprises: S101: Obtaining a surface contamination distribution map of a wafer to be cleaned, dividing the contaminated area on the wafer surface using a clustering algorithm to obtain a number of sub-areas with different contamination levels, and obtaining corresponding spray pressure and flow parameters for each sub-area from a preset spray parameter database; S102: Obtaining the wafer surface graphic structure data, partitioning the wafer surface using a decision tree algorithm to obtain a plurality of sub-regions with different graphic structure complexities, and obtaining corresponding spray angle parameters for each sub-region from the preset spray parameter database; S103 obtains a recommended cleaning time range for the wafer from a preset spray time-wafer attribute association database according to the material and surface roughness of the wafer; S104 integrates the sub-region spray parameters obtained in the first three steps to form a wafer surface cleaning plan, the cleaning plan including the nozzle layout, number of nozzles, spray liquid composition and temperature of each sub-region, and transmits the cleaning plan to the spray control unit; S105: The spray control unit controls the nozzles of the spray device to move on the wafer surface along a preset trajectory and time interval according to the received cleaning scheme, and monitors the pressure and flow parameters of each nozzle in real time. If the monitored values ​​deviate from the set values ​​of the cleaning scheme, the nozzle opening is dynamically adjusted; S106: During the wafer cleaning process, a machine vision system is used to collect images of the wafer surface in real time, the images are processed and analyzed, and the amount of residual contaminants on the wafer surface is calculated. If the residual amount is higher than a preset threshold, the wafer is re-transferred to the spray device for further cleaning. After the cleaning is completed in step S107, the wafer surface image is compared with a preset standard image to calculate the similarity between the two. If the similarity is lower than a preset threshold, the wafer number and cleaning parameters are recorded in the abnormal cleaning log, and the wafer is transferred to the unqualified product buffer area. S108 For the wafers that have passed the cleaning, their actual cleaning parameters are uploaded to the spray parameter optimization module. The optimization module trains the historical accumulated cleaning data based on the support vector machine algorithm to obtain a dynamically updated spray parameter-cleaning effect mapping model, which is used to optimize and predict the subsequent cleaning parameters of new wafers.

2. The method according to claim 1, characterized in that The step S101 includes: Acquire a surface contamination distribution image of the wafer to be cleaned; Preprocessing the pollution distribution image to obtain a preprocessed pollution distribution image; Performing cluster analysis on the pre-processed pollution distribution image using a K-means clustering algorithm to obtain a number of pollution sub-regions with different pollution degrees; Obtaining a preset spray parameter database, wherein the spray parameter database stores spray pressure and flow parameters corresponding to different pollution levels; According to the pollution degree of the plurality of pollution sub-areas, the spray pressure and flow rate parameters corresponding to each pollution sub-area are obtained from the spray parameter database; Placing the wafer in a cleaning device, and controlling the spray head of the cleaning device to perform differential cleaning on each contaminated sub-area based on the acquired spray pressure and flow parameters, thereby obtaining a cleaned wafer; Inspecting the cleaned wafer surface to determine whether it meets the preset cleanliness requirements; If the cleanliness requirement is not met, the cleaned wafer is placed back into the cleaning equipment for cleaning; If the cleanliness requirement is met, the cleaned wafer is sent to the next process.

3. The method according to claim 1, characterized in that The step S102 includes: Obtaining graphic structure data on the wafer surface, preprocessing the graphic structure data, and extracting graphic structure features; Based on the extracted graphic structure features, a decision tree algorithm is used to partition the wafer surface to obtain several sub-regions with different graphic structure complexities. For each sub-area, according to the complexity of its graphic structure, the corresponding spray angle parameters are obtained from the preset spray parameter database; Determine the spray angle of each sub-area based on the acquired spray angle parameters, and obtain the spray angle data of each sub-area on the wafer surface; Input the spray angle data of each sub-area on the wafer surface into the spray equipment, and control the spray equipment to perform differentiated spray treatment on each sub-area on the wafer surface; During the spray treatment process, real-time image data of the wafer surface is collected and analyzed to determine whether the spray effect meets expectations; If the spray effect does not meet expectations, the spray angle parameters are adjusted according to the image analysis results, and the adjusted parameters are updated to the spray parameter database to optimize the subsequent wafer spray processing process.

4. The method according to claim 1, wherein The step S103 includes: Obtain the wafer's material information and surface roughness value and use them as query conditions; Access the preset spray time-wafer attribute association database and perform matching queries based on material and roughness values; If a matching record is found in the associated database, the corresponding recommended cleaning time range is extracted; Determine the target cleaning time for the wafer based on the obtained recommended cleaning time range; The target cleaning time is passed to the spray control module, and the wafer is cleaned for a corresponding period of time; After cleaning is completed, the roughness of the wafer surface is tested to obtain the roughness value after cleaning; The roughness value after cleaning is compared with the preset threshold to determine whether the cleaning effect meets the standard and whether secondary cleaning or adjustment of cleaning parameters is required.

5. The method according to claim 1, wherein The step S104 includes: Acquiring contamination level data and material property data of the wafer surface, and determining spray parameters for each sub-area of ​​the wafer surface based on the contamination level data and material property data, wherein the spray parameters include nozzle layout, number of nozzles, spray liquid composition, and temperature; Integrate the spray parameters of each sub-area to obtain the wafer surface cleaning plan; Using a machine learning algorithm, based on preset historical cleaning data and current wafer surface condition data, various parameters in the cleaning solution are optimized to obtain an optimized cleaning solution; Converting the optimized cleaning scheme into a control instruction and transmitting the control instruction to a spray control unit; During the cleaning process, real-time spray parameters and wafer surface condition data of each sub-area are obtained; If the real-time spray parameters and wafer surface condition data do not conform to a preset range, dynamically adjusting the cleaning solution; After cleaning is completed, the cleaning effect data of the wafer surface is obtained, and the cleaning effect data is input into the machine learning algorithm, and the machine learning algorithm is trained to obtain an optimized machine learning algorithm for subsequent optimization of the cleaning scheme.

6. The method according to claim 1, characterized in that The S105 step includes: The spray control unit receives the cleaning plan, obtains the preset nozzle trajectory and time interval parameters, and controls the movement of the nozzle on the wafer surface according to the parameters; During the movement of the nozzle, the pressure and flow data of the nozzle are collected in real time, and the collected real-time data are compared with the pressure and flow thresholds set in the cleaning plan; If the comparison results show that the real-time pressure or flow data deviates from the set threshold, it is determined to be an abnormal situation and the nozzle opening adjustment mechanism is triggered; According to the direction and amount of pressure deviation, the nozzle opening adjustment value is dynamically calculated, and the nozzle opening is adjusted through the control signal to restore the pressure to the set range; According to the flow deviation direction and deviation amount, the nozzle opening adjustment value is dynamically calculated, and the nozzle opening is adjusted through the control signal to restore the flow rate to the set range; The pressure and flow monitoring data within a cleaning cycle are recorded and counted to form a cleaning quality report for evaluating the cleaning effect and optimizing the cleaning plan.

7. The method according to claim 1, characterized in that The step S106 includes: Acquire real-time images of the wafer surface through a machine vision system and transmit the images to an image processing module; In the image processing module, the acquired wafer surface image is preprocessed, including image denoising, enhancement and correction operations, to improve image quality; Based on the pre-processed wafer surface image, an image segmentation algorithm is used to segment the image and extract the contaminant area on the wafer surface; For the segmented contaminant areas, calculate their area ratio to obtain the residual contaminant amount on the current wafer surface; Compare the calculated contaminant residue with the preset threshold to determine whether the wafer surface meets the cleaning requirements; If the residual contaminant level is higher than a preset threshold, the conveyor device is controlled to re-transfer the wafer to the spray device for cleaning again; If the residual contaminant level is lower than the preset threshold, the wafer surface is judged to be qualified and the wafer is transferred to the next process for subsequent processing.

8. The method according to claim 1, characterized in that The step S107 includes: Obtaining a wafer surface image after wafer cleaning is completed, converting it into a grayscale image, and performing image enhancement processing to obtain a processed wafer surface image; Obtain a standard image matching the current wafer type from a preset standard image library, and perform the same image enhancement processing on the standard image; Using feature extraction algorithms, such as SIFT or SURF, key feature points are extracted from the processed wafer surface image and the standard image. Through the feature point matching algorithm, the RANSAC algorithm is used to match the feature points of the wafer surface image with the standard image, and the ratio of the number of matching point pairs to the total number of feature points is calculated to obtain the similarity value; The calculated similarity value is compared with a preset similarity threshold. If the similarity value is lower than the threshold, the wafer cleaning quality is judged to be unqualified. If the wafer cleaning quality is unqualified, the wafer number and corresponding cleaning parameters will be recorded in the abnormal cleaning log database, and the wafer will be transferred to the unqualified product buffer area; If the wafer cleaning quality is qualified, the wafer will be transferred to the next step for subsequent processing, and the wafer processing status information will be updated to track the wafer production progress in real time.

9. The method according to claim 1, characterized in that The step S108 includes: Obtain the actual cleaning parameters of the wafers that have passed the cleaning process and upload them to the spray parameter optimization module; The spray parameter optimization module obtains historical accumulated cleaning data and uses the support vector machine algorithm to train the cleaning data; Through the training of support vector machine algorithm, a dynamically updated spray parameter-cleaning effect mapping model is obtained; When a new wafer needs to be cleaned, the initial cleaning parameters of the new wafer are input into the spray parameter-cleaning effect mapping model; The spray parameter-cleaning effect mapping model predicts the optimized cleaning parameters based on the input initial cleaning parameters; Applying the optimized cleaning parameters to a cleaning process of a new wafer to clean the new wafer; After cleaning is completed, the actual cleaning parameters and cleaning results of the new wafer are obtained and added to the historical cleaning data for subsequent model updates and optimization.

10. An intelligent semiconductor wafer spray control system, characterized in that: The system is used to execute the intelligent semiconductor wafer spray control method according to any one of claims 1 to 9, and the spray control system includes: Contamination analysis module, used to obtain the contamination distribution on the wafer surface and divide it into sub-areas; Graphic structure analysis module, used to obtain the graphic structure of the wafer surface and divide it into sub-areas; A cleaning parameter acquisition module is used to obtain a recommended cleaning time range based on wafer properties; Cleaning plan generation module, used to integrate parameters to form a wafer surface cleaning plan; Cleaning control module, used to control the spray device to perform cleaning and monitor and adjust in real time; The quality detection module is used to monitor the cleaning effect in real time and perform subsequent processing.

Images