Polishing solution automatic proportioning and supplying system applied to silicon wafer and control method

Abstract

The invention discloses a polishing solution automatic proportioning and supplying system applied to a silicon wafer and a control method, and relates to the technical field of semiconductor manufacturing, and the control method comprises the following steps: collecting sound wave and chemical data in real time, firstly analyzing sound wave amplitude and calculating an abnormal coefficient; if the sound wave is abnormal, immediately tightening the judgment threshold value of the chemical parameter of the mixed liquid so as to improve the monitoring sensitivity, and then performing chemical state verification; and if the verification is abnormal, giving out early warning. And if the sound wave is normal, periodically adjusting the acquisition frequency to optimize resources. According to the process, a self-adaptive cooperation mechanism of sound wave front early warning and chemical secondary verification is constructed, and early capture and accurate judgment of pollution pulses are achieved. According to the invention, abnormity can be sensed in advance before pollutant diffusion, and the chemical monitoring sensitivity can be dynamically adjusted, so that instant pollution pulses can be effectively captured, false alarms and missing alarms can be obviously reduced, batch silicon wafer defects caused by out-of-control proportioning can be prevented from the source, and the process consistency and the product yield can be guaranteed.

Description

Technical Field

[0001] This invention relates to the field of semiconductor manufacturing technology, specifically to an automatic proportioning and supply system and control method for polishing slurry applied to silicon wafers. Background Technology

[0002] In nanoscale manufacturing, even the slightest fluctuations in concentration or pH can directly lead to defects in silicon wafer polishing. Therefore, it is necessary to monitor the state of the mixture in real time using high-precision sensors. The controller compares the mixture with the preset formula and calculates the deviation, then instructs the precision actuator to dynamically adjust the flow rate of each component. This enables continuous fine-tuning and constant output of the polishing solution, ultimately ensuring high yield and process consistency in chip production.

[0003] When the valve is closed, its mechanical structure forms a sealed cavity. This cavity is a stagnant area that water cannot flush out. Water stagnating in this area for a long time is prone to bacterial growth. When the valve is reopened or the pipeline pressure fluctuates, this wastewater will mix into the main pure water system. Due to the sampling delay of the online monitoring sensor, and the difficulty in ensuring that the contaminated stagnant water is captured or that only a very small amount of contaminated water is collected during the sampling process, it will create a false impression that the monitoring parameters are compliant. This misjudgment will cause the polishing fluid to be continuously delivered in the wrong proportion, ultimately leading to batch product defects that are difficult to trace. Summary of the Invention

[0004] Technical problems to be solved To address the shortcomings of existing technologies, this invention provides an automatic proportioning and supply system and control method for polishing slurry applied to silicon wafers. This solves the problem that the dead water zone formed by the mechanical structure of the valve releases high concentrations of pollutants in the form of pulses when it operates. This instantaneous pollution cannot be accurately captured due to the sampling delay and spatial limitations of online monitoring, causing the control system to maintain an incorrect proportion based on distorted signals, ultimately resulting in batch-specific and difficult-to-trace product defects.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an automatic proportioning control method for polishing slurry applied to silicon wafers, comprising the following specific steps: Step 1: Real-time acquisition of acoustic wave amplitude data and mixed slurry state data, and preprocessing thereof; Step 2: Calculation based on acoustic wave amplitude data to obtain an acoustic wave anomaly coefficient, and then analysis of whether the acoustic wave is abnormal. If the acoustic wave is abnormal, the mixed slurry state threshold is adjusted, and Step 3 is executed; if the acoustic wave is normal, the acoustic wave amplitude data acquisition frequency is periodically adjusted; Step 3: Analysis based on the mixed slurry state data value and the mixed slurry state threshold. If the analysis indicates that the mixed slurry state is normal, the process returns to Step 1 or ends directly; if the analysis indicates that the mixed slurry state is abnormal, an early warning is issued.

[0006] Furthermore, the specific method for obtaining the acoustic anomaly coefficient is as follows: each value of the acoustic amplitude data is calculated with a preset acoustic amplitude threshold to obtain an amplitude difference. The positive or negative value of the amplitude difference is analyzed, and the values ​​of the acoustic amplitude data corresponding to negative or zero amplitude differences are statistically analyzed to obtain the value to be detected. The value to be detected is summed to obtain the sum to be detected. The amplitude difference and the sum to be detected are combined to obtain the acoustic anomaly coefficient.

[0007] Furthermore, the specific method for obtaining the acoustic anomaly coefficient is as follows: ;in, Represents the acoustic anomaly coefficient. Indicates the amplitude difference. This indicates the value to be detected.

[0008] Furthermore, the specific method for obtaining the amplitude difference is as follows: calculate the difference between each value of the sound wave amplitude data and the sound wave amplitude threshold to obtain the amplitude difference.

[0009] Furthermore, the specific method for analyzing the positive or negative value of the amplitude difference is as follows: compare the amplitude difference with zero. If the amplitude difference is greater than zero, it indicates a positive amplitude difference; if the amplitude difference is less than or equal to zero, it indicates a negative amplitude difference or an amplitude difference of zero.

[0010] Furthermore, the specific method for analyzing whether the sound wave is abnormal is as follows: the sound wave abnormality coefficient is compared with a preset sound wave abnormality threshold. If the sound wave abnormality coefficient is greater than the sound wave abnormality threshold, it indicates that the sound wave is abnormal. The difference between the sound wave abnormality coefficient and the sound wave abnormality threshold is calculated to obtain the sound wave abnormality difference. If the sound wave abnormality coefficient is less than or equal to the sound wave abnormality threshold, it indicates that the sound wave is normal.

[0011] Furthermore, the specific method for adjusting the state threshold of the mixture is as follows: normalize the acoustic anomaly difference and the state threshold of the mixture, set a reduction value, assign the acoustic anomaly difference to the reduction value, calculate the difference between the state threshold of the mixture and the reduction value, and obtain a new state threshold of the mixture.

[0012] Furthermore, the specific method for periodically adjusting the sound wave amplitude data acquisition frequency is as follows: when the sound wave anomaly coefficient is less than or equal to the sound wave anomaly threshold, the sound wave amplitude data acquisition frequency is increased, and after a period of time, the sound wave amplitude data acquisition frequency is restored, and this process is repeated.

[0013] Furthermore, the specific method for analyzing the mixture state data based on the mixture state threshold is as follows: compare the mixture state data value with the mixture state threshold. If the mixture state data value is greater than the mixture state threshold, it indicates that the mixture state is abnormal; if the mixture state data value is less than or equal to the mixture state threshold, it indicates that the mixture state is normal.

[0014] An automatic proportioning and supply system for polishing slurry applied to silicon wafers includes the following specific modules: a sensor preprocessing module, an acoustic diagnostic adjustment module, and a chemical verification and early warning module. The sensor preprocessing module collects and preprocesses acoustic wave amplitude data and mixed slurry state data in real time. The acoustic diagnostic adjustment module calculates an acoustic wave anomaly coefficient based on the acoustic wave amplitude data, then analyzes whether the acoustic wave is abnormal. If the acoustic wave is abnormal, it adjusts the mixed slurry state threshold and executes the chemical verification and early warning module; if the acoustic wave is normal, it periodically adjusts the acoustic wave amplitude data acquisition frequency. The chemical verification and early warning module analyzes the mixed slurry state data and the mixed slurry state threshold. If the analysis indicates that the mixed slurry state is normal, it returns to the sensor preprocessing module or terminates directly; if the analysis indicates that the mixed slurry state is abnormal, it issues an early warning.

[0015] Beneficial effects Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: 1. By introducing acoustic monitoring as a sentinel, and by analyzing anomalies in acoustic amplitude data, fluid disturbances and contaminant release events in the pipeline can be detected in advance before contaminants fully diffuse and significantly change the chemical parameters of the mixture. This physical signal anomaly provides a valuable early warning window for the system, realizing an improvement from post-detection to pre-warning.

[0016] 2. When the acoustic monitoring indicates an abnormality, the system does not immediately trigger an alarm. Instead, it dynamically tightens the threshold of downstream chemical monitoring to perform secondary verification with higher sensitivity. Conversely, when the acoustic monitoring is normal, the system periodically adjusts the acquisition frequency to optimize resources. This not only captures instantaneous risks but also effectively filters out occasional noise interference from a single sensor, greatly reducing false alarms and missed alarms and improving the accuracy and reliability of the entire monitoring system.

[0017] 3. Through early warning and accurate judgment, the system can intervene in time before the impact of pollution expands, such as triggering stricter formula verification, starting the rinsing procedure, or suspending the supply, thereby preventing the continuous delivery of polishing slurry with abnormal ratios and particles to the polishing machine. This directly avoids the random and batch-related defects in silicon wafers caused by this, ensuring the extreme consistency of the chemical mechanical polishing process from the source, which is of decisive significance for maintaining a high yield in chip production.

[0018] 4. When process anomalies occur, this time-series data can provide engineers with clear clues for fault analysis, helping to locate the root cause of the problem, such as a specific valve or pipeline section. This not only facilitates rapid troubleshooting but also provides a valuable data foundation for predictive maintenance, system design optimization, and further expansion of the process window.

[0019] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0020] Figure 1 This invention relates to a flowchart of an automatic proportioning control method for polishing slurry applied to silicon wafers. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0023] Example 1: like Figure 1 As shown, this embodiment of the invention provides an automatic proportioning control method for polishing slurry applied to silicon wafers, including the following specific steps: Step 1: Real-time acquisition of acoustic amplitude data and mixture state data. Acoustic amplitude data is acquired by high-frequency acoustic emission sensors or ultrasonic transducers, which are usually placed directly on the valve body, pipe wall or key locations where dead water is suspected to exist, to capture characteristic acoustic signals generated by fluid disturbance or particle release. Data on the state of the mixture is collected by online process analysis sensors, including concentration meters, pH / ORP meters, etc., which are directly installed on the main process pipeline downstream of the mixing unit to achieve real-time continuous monitoring of the chemical properties of the polishing fluid. Data cleaning is performed on the acoustic amplitude data and the mixture state data to remove redundant values ​​and improve the quality of the acoustic amplitude data and the mixture state data. Step Two: Calculate the acoustic wave anomaly coefficient based on the acoustic wave amplitude data, and then analyze whether the acoustic wave is abnormal. If the acoustic wave is abnormal, adjust the state threshold of the mixture. First, set a static threshold based on the core parameter range of the formula determined in the initial process development stage. Then, continuously collect historical data under normal operating conditions during mass production. Determine the natural fluctuation range and control limits of each parameter through statistical process control analysis, which serves as the basis for dynamic adjustment. When the acoustic diagnostic module detects a physical anomaly risk, the system will temporarily tighten the chemical parameter threshold according to the preset mapping rules. This composite threshold system will periodically perform feedback calibration based on actual yield data and fault tracing results to achieve a balance between process robustness and monitoring sensitivity, and then execute Step Three. If the acoustic wave is normal, periodically adjust the acoustic wave amplitude data acquisition frequency. Step 3: Analyze the mixed liquid status data and the mixed liquid status threshold. If the analysis shows that the mixed liquid status is normal, return to Step 1 or end directly. If the analysis shows that the mixed liquid status is abnormal, issue an early warning. This will be done by visually alerting the human-machine interface in the central control room with a flashing red pop-up, an alarm list, and highlighting abnormal points on the process flow diagram. At the same time, the audible and visual alarms on the control console and the field area will be triggered to attract attention. Meanwhile, the early warning information will be automatically pushed to the mobile terminals or workstations of relevant engineers.

[0024] Example 2 differs from Example 1 in that: The specific method for obtaining the acoustic anomaly coefficient is as follows: Each value of the acoustic wave amplitude data is calculated with a preset acoustic wave amplitude threshold to obtain the amplitude difference, which reflects the impact of contaminated water on the flowing pure water. That is, the ultrasonic waves experience severe energy attenuation after returning over a large volume of sewage. The positive and negative values ​​of the amplitude difference are analyzed, and the values ​​of the acoustic wave amplitude data corresponding to negative or zero amplitude differences are statistically analyzed to obtain the value to be detected. The value to be detected is calculated by summing the values ​​to be detected, which indicates that the contaminated water continues to accumulate even when the acoustic wave is not abnormal. That is, although the contaminated water is initially small and has a small impact when the acoustic wave anomaly coefficient is less than or equal to the acoustic wave anomaly threshold, the contaminated water accumulates slowly in the later stages, and the volume of contaminated water increases, which will have a greater impact. The acoustic wave anomaly coefficient is obtained by comprehensively calculating the amplitude difference and the value to be detected.

[0025] The specific method for obtaining the acoustic anomaly coefficient is as follows: ; in, This represents the sound wave anomaly coefficient, reflecting whether the sound wave is abnormal. Indicates the amplitude difference. This indicates the value to be detected.

[0026] The specific method for obtaining the amplitude difference is as follows: The amplitude difference is obtained by calculating the difference between each value of the acoustic wave amplitude data and the acoustic wave amplitude threshold.

[0027] The specific method for analyzing the sign of amplitude difference is as follows: The amplitude difference is compared with zero. If the amplitude difference is greater than zero, it indicates a positive amplitude difference; if the amplitude difference is less than or equal to zero, it indicates a negative amplitude difference or a zero amplitude difference.

[0028] The specific method for analyzing whether sound waves are abnormal is as follows: The sound wave anomaly coefficient is compared with the preset sound wave anomaly threshold, where the sound wave anomaly threshold is the average value of historical normal sound wave amplitude data. If the sound wave anomaly coefficient is greater than the sound wave anomaly threshold, it indicates that the sound wave is abnormal. The difference between the sound wave anomaly coefficient and the sound wave anomaly threshold is calculated to obtain the sound wave anomaly difference value. If the sound wave anomaly coefficient is less than or equal to the sound wave anomaly threshold, it indicates that the sound wave is normal.

[0029] The specific method for adjusting the state threshold of the mixture is as follows: The acoustic anomaly difference and the mixed liquid state threshold are normalized to eliminate dimensional differences and transform the values ​​of the acoustic anomaly difference and the mixed liquid state threshold, which are of different orders of magnitude, into a unified numerical range. A reduction value is set, and the acoustic anomaly difference is assigned to the reduction value. The difference between the mixed liquid state threshold and the reduction value is calculated to obtain a new mixed liquid state threshold. By reducing the mixed liquid state threshold, the sensitivity of detecting the mixed liquid state is increased, avoiding the false impression of compliance of monitoring parameters caused by a very small amount of contaminated water. This prevents the polishing fluid from being continuously delivered in the wrong proportion, which could ultimately lead to batch product defects that are difficult to trace.

[0030] The specific method for periodically adjusting the frequency of sound wave amplitude data acquisition is as follows: When the acoustic anomaly coefficient is less than or equal to the acoustic anomaly threshold, the acoustic amplitude data acquisition frequency is increased. After a period of time, the acoustic amplitude data acquisition frequency is restored, and this process is repeated. In cases where polluted water evaporates, even if no sewage is detected (i.e., the acoustic anomaly coefficient is less than or equal to the acoustic anomaly threshold), it does not mean that there was no polluted water before or after. In other words, the sewage has already evaporated. Therefore, it is necessary to increase the frequency of acquiring acoustic amplitude data to increase the probability of detecting sewage. In addition, continuously increasing the frequency of acquiring sewage will increase power consumption, so periodic frequency increases are necessary.

[0031] The specific method for analyzing the mixture state data based on the mixture state threshold is as follows: The value of the mixture status data is compared with the mixture status threshold. If the value of the mixture status data is greater than the mixture status threshold, it indicates that the mixture status is abnormal; if the value of the mixture status data is less than or equal to the mixture status threshold, it indicates that the mixture status is normal.

[0032] Example 3: An automated slurry mixing and supply system for silicon wafers includes the following specific modules: Sensor preprocessing module: Real-time acquisition of acoustic wave amplitude data and mixed liquid state data, and preprocessing thereof; Acoustic diagnostic adjustment module: Calculates the acoustic wave abnormality coefficient based on the acoustic wave amplitude data, and then analyzes whether the acoustic wave is abnormal. If the acoustic wave is abnormal, it adjusts the mixed liquid state threshold and executes the chemical verification early warning module; if the acoustic wave is normal, it periodically adjusts the acoustic wave amplitude data acquisition frequency. Chemical verification early warning module: Analyzes the mixed liquid state data and the mixed liquid state threshold. If the analysis shows that the mixed liquid state is normal, it returns to the sensor preprocessing module or ends directly; if the analysis shows that the mixed liquid state is abnormal, it issues an early warning.

[0033] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred 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 content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for automatically controlling the proportioning of a polishing solution applied to a silicon wafer, characterized by the steps of: The specific steps include the following: ​ Step 1: Real-time acquisition of acoustic wave amplitude data and mixture state data, followed by preprocessing; Step 2: Calculate the acoustic wave anomaly coefficient based on the acoustic wave amplitude data, and then analyze whether the acoustic wave is abnormal. If the acoustic wave is abnormal, adjust the mixed liquid state threshold and proceed to Step 3; if the acoustic wave is normal, periodically adjust the acoustic wave amplitude data acquisition frequency. Step 3: Analyze the values ​​of the mixed liquid state data and the mixed liquid state threshold. If the analysis shows that the mixed liquid state is normal, return to Step 1 or end directly; if the analysis shows that the mixed liquid state is abnormal, issue an early warning.

2. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 1, characterized in that: The specific method for obtaining the acoustic anomaly coefficient is as follows: Each value of the acoustic wave amplitude data is calculated with a preset acoustic wave amplitude threshold to obtain the amplitude difference. The positive or negative value of the amplitude difference is analyzed, and the acoustic wave amplitude data corresponding to negative or zero amplitude difference values ​​are statistically analyzed to obtain the value to be detected. The value to be detected is summed based on the value to be detected to obtain the sum to be detected. The amplitude difference and the sum to be detected are combined to obtain the acoustic wave anomaly coefficient.

3. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 2, characterized in that: The specific method for obtaining the acoustic anomaly coefficient is as follows: ； in, Represents the acoustic anomaly coefficient. Indicates the amplitude difference. This indicates the value to be detected.

4. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 3, characterized in that: The specific method for obtaining the amplitude difference is as follows: The amplitude difference is obtained by calculating the difference between each value of the acoustic wave amplitude data and the acoustic wave amplitude threshold.

5. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 2, characterized in that: The specific method for analyzing the sign of the amplitude difference is as follows: The amplitude difference is compared with zero. If the amplitude difference is greater than zero, it indicates a positive amplitude difference. If the amplitude difference is less than or equal to zero, it indicates a negative amplitude difference or a zero amplitude difference.

6. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 5, characterized in that: The specific method for analyzing whether sound waves are abnormal is as follows: The sound wave anomaly coefficient is compared with the preset sound wave anomaly threshold. If the sound wave anomaly coefficient is greater than the sound wave anomaly threshold, it indicates that the sound wave is abnormal. The difference between the sound wave anomaly coefficient and the sound wave anomaly threshold is calculated to obtain the sound wave anomaly difference value. If the sound wave anomaly coefficient is less than or equal to the sound wave anomaly threshold, it indicates that the sound wave is normal.

7. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 6, characterized in that: The specific method for adjusting the state threshold of the mixture is as follows: The acoustic anomaly difference and the mixed liquid state threshold are normalized, a reduction value is set, the acoustic anomaly difference is assigned to the reduction value, and the difference between the mixed liquid state threshold and the reduction value is calculated to obtain a new mixed liquid state threshold.

8. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 7, characterized in that: The specific method for periodically adjusting the frequency of sound wave amplitude data acquisition is as follows: When the acoustic anomaly coefficient is less than or equal to the acoustic anomaly threshold, the acoustic amplitude data acquisition frequency is increased, and after a period of time, the acoustic amplitude data acquisition frequency is restored, and this process is repeated.

9. The automatic proportioning control method for polishing slurry applied to silicon wafers according to claim 8, characterized in that: The specific method for analyzing the mixture state data based on the mixture state threshold is as follows: The value of the mixture state data is compared with the mixture state threshold. If the value of the mixture state data is greater than the mixture state threshold, it indicates that the mixture state is abnormal. If the value of the mixture status data is less than or equal to the mixture status threshold, it indicates that the mixture status is normal.

10. An automatic proportioning and supply system for polishing slurry applied to silicon wafers, used to implement the automatic proportioning and control method for polishing slurry applied to silicon wafers as described in any one of claims 1-9, characterized in that, The automatic proportioning and supply system for polishing slurry applied to silicon wafers includes: a sensor preprocessing module, an acoustic diagnostic adjustment module, and a chemical verification and early warning module; The sensor preprocessing module: collects acoustic wave amplitude data and mixed liquid state data in real time, and performs preprocessing; The acoustic diagnostic adjustment module calculates the acoustic anomaly coefficient based on the acoustic amplitude data, and then analyzes whether the acoustic wave is abnormal. If the acoustic wave is abnormal, it adjusts the state threshold of the mixture and executes the chemical verification early warning module. If the acoustic wave is normal, it periodically adjusts the acoustic amplitude data acquisition frequency. The chemical verification early warning module analyzes the values ​​of the mixed liquid state data and the mixed liquid state threshold. If the analysis shows that the mixed liquid state is normal, it returns to the sensing preprocessing module or ends directly; if the analysis shows that the mixed liquid state is abnormal, it issues an early warning.

Images