Remote plasma source equipment rubber ring local cooling method and system

Abstract

The invention belongs to the technical field of plasma sources, and discloses a remote plasma source equipment rubber ring local cooling method and system, based on cooling devices and temperature sensors which are uniformly arranged on a remote plasma source equipment rubber ring in the circumferential direction, an energy loss model is determined by extracting a feature data set from a historical database, and a local cooling result is obtained. The optimal characteristic data set is determined, and cooling control is performed based on the local temperature measured in real time, so that accurate local temperature control is realized, the cooling efficiency is improved, the energy waste is reduced, and the cooling effect is optimized.

Description

Technical Field

[0001] The present application relates to the field of plasma source technology, and in particular to a method and system for local cooling of a rubber ring of a remote plasma source device. Background Art

[0002] Remote Plasma Source (RPS) is an advanced plasma generating device that generates plasma at a location far away from the processing surface through a specific method (such as radio frequency or microwave electromagnetic waves) and transmits the plasma to the surface area that needs to be processed. The main feature of this technology is its "remote" characteristic, that is, there is a certain distance between the plasma source and the processing surface, thus avoiding the direct contact and thermal damage problems that may exist in traditional plasma processing.

[0003] In remote plasma source equipment, most of the working gas is introduced into a sealed hollow cylindrical cavity. Most of these hollow cylindrical cavities are made of quartz material, and both ends are covered with rubber rings to ensure the sealing of the cavity. However, in the process of generating plasma, high energy (high pressure, high electric field or high magnetic field, etc.) is often required to ionize gas molecules. The temperature of the plasma will be very high, so the temperature of the hollow cylindrical cavity carrying the plasma will also be very high, and the temperature of the rubber ring connected to the cavity will also be very high. On the other hand, due to the remote plasma source equipment, or various external factors such as the input power adjustment is not optimal, the plasma will be unevenly distributed in the cavity, so the temperature distribution of the cavity itself will also be uneven, and the corresponding rubber ring temperature distribution will also be uneven, which will cause the rubber ring to melt locally, thereby affecting the sealing or contaminating the port of the cavity, thereby reducing the service life of the cavity.

[0004] In order to solve the practical problem of excessively high local temperature of such rubber rings, the existing solution is to set a water cooling device on the periphery of the cavity. Although this can effectively reduce the temperature of the cavity, it is a full-coverage cooling of the entire cavity. Therefore, although the local temperature can be reduced, the temperature of other areas (areas that do not need to be cooled) will also be reduced, which will increase the power consumption of the remote plasma source equipment and reduce the utilization rate of the input power. This results in the existing method not only having poor effect, low efficiency, time-consuming and labor-intensive, but also low effective utilization rate of the input power. Therefore, how to locally cool the rubber ring in the remote plasma source equipment is of great practical significance. Summary of the invention

[0005] The purpose of the present application is to provide a method and system for local cooling of the rubber ring of a remote plasma source device, which can effectively locally cool the rubber ring in the remote plasma source device, reduce energy waste and optimize the cooling effect.

[0006] In a first aspect, the present application provides a method for local cooling of a rubber ring of a remote plasma source device, based on a local cooling system, wherein the local cooling system comprises a plurality of cooling devices uniformly arranged circumferentially on the rubber ring of the remote plasma source device, and a temperature sensor is arranged between any two adjacent cooling devices; The method for local cooling of the rubber ring of the remote plasma source equipment comprises the following steps: A1. extracting a plurality of historical characteristic data sets from a historical database, wherein the historical characteristic data sets include historical input power of a remote plasma source device, historical temperature threshold deviation, historical cooling time, and historical cooling speed of a cooling device; A2. Establishing an energy loss model using the extracted historical feature data set; A3. According to the determined characteristic data, using the energy loss model, determine the optimal characteristic data group; the characteristic data group includes the input power of the remote plasma source device, the temperature threshold deviation, the cooling time and the cooling speed of the cooling device, and the determined characteristic data is at least one characteristic data in the characteristic data group; A4. Based on the local temperature measured in real time by the temperature sensor, the local temperature reduction control is performed on the rubber ring of the remote plasma source device according to the optimal characteristic data set, and the characteristic data set is collected during the control process and added to the historical database.

[0007] By extracting characteristic data groups from the historical database, establishing an energy loss model, determining the optimal characteristic data group, and performing cooling control based on the local temperature measured in real time, accurate local temperature control is achieved, cooling efficiency is improved, energy waste is reduced, and the cooling effect is optimized.

[0008] Preferably, the historical temperature threshold deviation is the deviation between the historical highest temperature threshold and the historical lowest temperature threshold; the historical cooling time is the time taken for the local temperature of the rubber ring of the remote plasma source device to drop from the historical highest temperature threshold to the historical lowest temperature threshold under the conditions of the historical input power and the historical cooling speed; Correspondingly, the temperature threshold deviation is the deviation between the highest temperature threshold and the lowest temperature threshold; the cooling time is the time it takes for the local temperature of the rubber ring of the remote plasma source device to drop from the highest temperature threshold to the lowest temperature threshold under the conditions of the input power and the cooling speed; The local temperature of the rubber ring of the remote plasma source equipment is measured by a temperature sensor.

[0009] Preferably, step A2 comprises: A201. Establish an initial energy loss model; A202. Use the extracted historical feature data set to train the initial energy loss model to obtain a confirmed energy loss model.

[0010] This method combines theoretical models with actual data, and can more accurately describe and predict the energy loss of the rubber ring of remote plasma source equipment, providing a reliable theoretical basis for local cooling control.

[0011] Preferably, in step A201, an initial energy loss model is established according to the following formula: ; ; in, is the cooling time, is the temperature threshold deviation, is the input power of the remote plasma source device, is the heat energy conversion coefficient, is the energy loss factor before cooling, is the energy loss factor of the cooling process, is the cooling speed of the cooling device, is the energy loss factor function, is the error factor.

[0012] Preferably, step A202 includes: The extracted historical characteristic data group is used as sample data, and a reference value of the energy loss factor of the corresponding cooling process is calculated according to each sample data; Input each of the sample data into the initial energy loss model in turn, and use the estimated value of the energy loss factor of the cooling process output by the initial energy loss model and the corresponding reference value to calculate the loss function, so as to iteratively optimize the model parameters of the initial energy loss model to obtain the established energy loss model.

[0013] Preferably, step A3 comprises: A301. Filtering a historical feature data group having the same and / or similar historical feature data as the determined feature data from the historical database as an alternative feature data group; A302. Calculate the energy loss quantification characterization value corresponding to each of the candidate feature data sets according to the established energy loss model and energy loss quantification characterization value calculation formula; A303. Determine the candidate feature data set corresponding to the minimum energy loss quantitative characterization value as the optimal feature data set.

[0014] By screening similar cases from historical data, the search efficiency is improved; by using the energy loss model for quantitative evaluation, the scientific nature of the solution selection is ensured; by selecting the optimal solution and applying it to actual control, new data is accumulated for the historical database, providing a basis for future optimization. This method not only improves the efficiency of a single cooling, but also continuously improves the control effect by continuously accumulating experience.

[0015] Preferably, step A302 includes: Input the candidate characteristic data set into the established energy loss model to obtain the energy loss factor of the cooling process output by the established energy loss model, which is recorded as the estimated energy loss factor; The energy loss quantification characterization value of the candidate feature data set is calculated according to the following energy loss quantification characterization value calculation formula: ; in, is a quantitative characterization value of the energy loss, is the cooling time in the candidate feature data set, is the temperature threshold deviation in the candidate feature data set, is the input power in the candidate feature data set, is the heat energy conversion coefficient, is the estimated energy loss factor.

[0016] Preferably, step A4 comprises: Obtaining a maximum temperature threshold and a minimum temperature threshold corresponding to the temperature threshold deviation in the optimal feature data group, which are recorded as a first maximum temperature threshold and a first minimum temperature threshold respectively; Adjusting the input power of the remote plasma source device to the input power in the optimal characteristic data set; When a temperature sensor detects that the local temperature reaches the first maximum temperature threshold, the cooling devices on both sides of the corresponding temperature sensor are started, and the started cooling devices are cooled at the cooling speed in the optimal characteristic data group until the local temperature drops to the first minimum temperature threshold, then the cooling devices on both sides of the corresponding temperature sensor are turned off, and the working time of the cooling devices is recorded; The cooling time in the optimal characteristic data set is replaced by the working time to form a new characteristic data set, and is added to the historical database.

[0017] In a second aspect, the present application provides a local cooling system for a rubber ring of a remote plasma source device, comprising a controller and a plurality of cooling devices uniformly arranged circumferentially on the rubber ring of the remote plasma source device, wherein a temperature sensor is arranged between any two adjacent cooling devices; The temperature sensor is used to detect the local temperature of the rubber ring of the remote plasma source equipment; The cooling device is used to cool the rubber ring of the remote plasma source equipment; The controller is used to: Extracting a plurality of historical characteristic data sets from a historical database, wherein the historical characteristic data sets include historical input power of the remote plasma source device, historical temperature threshold deviation, historical cooling time, and historical cooling speed of the cooling device; Establishing an energy loss model using the extracted historical feature data set; According to the determined characteristic data, the optimal characteristic data group is determined by using the energy loss model; the characteristic data group includes the input power of the remote plasma source device, the temperature threshold deviation, the cooling time and the cooling speed of the cooling device, and the determined characteristic data is at least one characteristic data in the characteristic data group; Based on the local temperature measured in real time by the temperature sensor, the local temperature reduction control is performed on the rubber ring of the remote plasma source equipment according to the optimal characteristic data group, and the characteristic data group is collected during the control process and added to the historical database.

[0018] Preferably, the remote plasma source equipment rubber ring local cooling system further includes a plurality of indicating devices, each of the indicating devices is electrically connected to each of the temperature sensors in a one-to-one correspondence, and the indicating device is used to issue an indication signal according to the local temperature measured by the corresponding temperature sensor.

[0019] Beneficial effect: The method and system for local cooling of the rubber ring of a remote plasma source device provided in the present application are based on cooling devices and temperature sensors evenly arranged circumferentially on the rubber ring of the remote plasma source device. By extracting characteristic data groups from a historical database, an energy loss model is established, the optimal characteristic data group is determined, and cooling control is performed based on the local temperature measured in real time, thereby achieving precise local temperature control, improving cooling efficiency, reducing energy waste, and optimizing cooling effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart of a method for local cooling of a rubber ring of a remote plasma source device provided in an embodiment of the present application.

[0021] Figure 2 A schematic diagram of the structure of the local cooling system of the rubber ring of the remote plasma source equipment provided in the embodiment of the present application.

[0022] Explanation of reference numerals: 1. Cooling device; 2. Temperature sensor; 3. Control module; 4. Indication device; 90. Rubber ring of remote plasma source equipment. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0024] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0025] Remote plasma source equipment is widely used in fields such as semiconductor manufacturing, and its rubber ring is one of the key components. During the operation of the equipment, the temperature of the rubber ring tends to rise, affecting the performance and service life of the equipment. Traditional rubber ring cooling methods usually adopt an overall cooling method, which is difficult to accurately control local high-temperature areas and has low cooling efficiency. In addition, the cooling control strategy with fixed parameters is difficult to adapt to temperature changes under different working conditions, and is prone to insufficient or excessive cooling. Therefore, how to achieve efficient local cooling control of the rubber ring of remote plasma source equipment has become a technical problem that needs to be solved urgently.

[0026] If the problem of local high temperature of the rubber ring of the remote plasma source equipment cannot be effectively solved, it will directly affect the yield rate of semiconductor manufacturing and the service life of the equipment. Continuous high temperature will accelerate the aging of the rubber ring material, causing its physical and chemical properties to deteriorate, and increase the risk of particle shedding, thereby contaminating the wafer surface. In addition, the uneven distribution of temperature will cause thermal deformation of the rubber ring, affecting the uniformity and stability of the plasma, and then affecting the accuracy and consistency of the etching or deposition process. In extreme cases, local overheating may cause the rubber ring to rupture or damage, resulting in equipment downtime and high maintenance costs. In the long run, if this problem is not solved, it will limit the application of remote plasma source equipment in advanced processes and hinder the advancement of semiconductor manufacturing technology. Therefore, developing a method that can achieve efficient and precise local cooling control is of great significance to improving the performance and reliability of remote plasma source equipment.

[0027] To this end, the present application proposes a method for local cooling of the rubber ring of a remote plasma source device.

[0028] Please refer to Figure 1 In some embodiments of the present application, a method for local cooling of a remote plasma source device rubber ring is based on a local cooling system (e.g. Figure 2 The local cooling system for the rubber ring of the remote plasma source device shown in the figure) comprises a plurality of cooling devices 1 uniformly arranged along the circumference on the rubber ring 90 of the remote plasma source device, and a temperature sensor 2 is arranged between any two adjacent cooling devices 1; The method for local cooling of the rubber ring of a remote plasma source device comprises the following steps: A1. Extracting multiple historical feature data sets from the historical database, the historical feature data sets include the historical input power of the remote plasma source device, the historical temperature threshold deviation, the historical cooling time and the historical cooling speed of the cooling device 1; A2. Establishing an energy loss model using the extracted historical feature data set; A3. According to the determined characteristic data, the energy loss model is used to determine the optimal characteristic data group; the characteristic data group includes the input power of the remote plasma source device, the temperature threshold deviation, the cooling time and the cooling speed of the cooling device 1, and the determined characteristic data is at least one characteristic data in the characteristic data group; A4. Based on the local temperature measured in real time by the temperature sensor 2, the local temperature reduction control is performed on the rubber ring 90 of the remote plasma source device according to the optimal characteristic data set, and the characteristic data set is collected during the control process and added to the historical database.

[0029] This method extracts characteristic data sets from the historical database, establishes an energy loss model, determines the optimal characteristic data set, and performs cooling control based on the local temperature measured in real time, thereby achieving precise local temperature control, improving cooling efficiency, reducing energy waste, and optimizing cooling effects. In addition, by accumulating new data for the historical database, it provides a basis for future optimization, thereby continuously improving the control effect.

[0030] Specifically, the historical temperature threshold deviation is the deviation between the historical highest temperature threshold and the historical lowest temperature threshold; the historical cooling time is the time taken for the local temperature of the rubber ring 90 of the remote plasma source device to drop from the historical highest temperature threshold to the historical lowest temperature threshold under the conditions of historical input power and historical cooling speed; Correspondingly, the temperature threshold deviation is the deviation between the highest temperature threshold and the lowest temperature threshold; the cooling time is the time it takes for the local temperature of the rubber ring 90 of the remote plasma source device to drop from the highest temperature threshold to the lowest temperature threshold under the conditions of input power and cooling speed; The local temperature of the rubber ring 90 of the remote plasma source device is measured by the temperature sensor 2 .

[0031] In step A1, part or all of the historical feature data groups in the historical database may be extracted; for example, based on the determined feature data, a historical feature data group having the same and / or similar historical feature data as the determined feature data (the four data in the historical feature data group are collectively referred to as historical feature data) may be extracted from the historical database. When the deviation between each historical feature data corresponding to the determined feature data in a historical feature data group and the corresponding determined feature data does not exceed a preset deviation range, it is considered that the historical feature data group has historical feature data similar to the determined feature data; if the number of the extracted historical feature data groups is insufficient (for example, less than a preset number threshold), other historical feature data groups are extracted to supplement it.

[0032] Specifically, step A2 includes: A201. Establish an initial energy loss model; A202. Use the extracted historical feature data set to train the initial energy loss model to obtain a confirmed energy loss model.

[0033] This method combines theoretical models with actual data, and can more accurately describe and predict the energy loss of the rubber ring 90 of the remote plasma source equipment, providing a reliable theoretical basis for local cooling control.

[0034] In practical applications, the establishment of the initial energy loss model can be done in a variety of ways. One common method is to establish a corresponding mathematical model based on theoretical analysis by studying the thermodynamic properties and energy transfer mechanism of the rubber ring 90 of the remote plasma source equipment. Another method is to construct a preliminary energy loss calculation formula based on empirical formulas using expert experience and existing engineering practice data. These two methods can be used in combination to obtain a more comprehensive and accurate initial model.

[0035] For example, in some embodiments, in step A201, an initial energy loss model is established according to the following formula: (1); (2); in, is the cooling time, is the temperature threshold deviation, is the input power of the remote plasma source device, is the heat energy conversion coefficient, is the energy loss factor before cooling (a constant that can be determined through experiments), is the energy loss factor of the cooling process, is the cooling rate of the cooling device 1, is the energy loss factor function, is the error factor.

[0036] Formula (1) describes the relationship between the cooling time and the temperature threshold deviation, input power and energy loss factor. Formula (2) introduces the energy loss factor function and error factor to more accurately describe the change of energy loss in the cooling process with cooling rate.

[0037] Among them, the initial energy loss model can be established using a variety of machine learning algorithms, such as support vector regression (SVR), random forest or deep neural network.

[0038] In fact, there is the following relationship between the input power of the remote plasma source device, the working time of the remote plasma source device and the temperature of the rubber ring 90 of the remote plasma source device: (3); in, is the temperature of the rubber ring 90 of the remote plasma source equipment, The working time of the remote plasma source equipment, is the energy loss factor (the faster the temperature drops, the larger its value).

[0039] When the temperature of the rubber ring 90 of the remote plasma source device reaches the maximum temperature threshold and the temperature does not start to drop (hereinafter referred to as the first moment), according to formula (3): (4); in, is the maximum temperature threshold, This is the working time of the remote plasma source device at the first moment.

[0040] When the temperature of the rubber ring 90 of the remote plasma source device is cooled to the minimum temperature threshold (hereinafter referred to as the second moment), according to formula (3): (5); in, is the minimum temperature threshold, It is the working time of the remote plasma source device at the second moment.

[0041] Subtracting formula (4) from formula (5) yields: ; Right now, (6); Formula (6) can be converted into formula (1).

[0042] In formula (1) and formula (2), only and is unknown, therefore, in step A202, the process of training the initial energy loss model is actually to determine and process.

[0043] Specifically, step A202 includes: The extracted historical characteristic data group is used as sample data, and the reference value of the energy loss factor of the corresponding cooling process is calculated according to each sample data; The sample data are input into the initial energy loss model in turn, and the loss function is calculated using the estimated value of the energy loss factor of the cooling process output by the initial energy loss model and the corresponding reference value, so as to iteratively optimize the model parameters of the initial energy loss model and obtain the established energy loss model.

[0044] Among them, the sample data can be , and Substitute into formula (1) and solve to get the corresponding , which is the reference value of the energy loss factor in the cooling process.

[0045] The input data of the energy loss model is , , and , the output data is The estimated value of .

[0046] The loss function used in the iterative optimization process can be set according to actual needs, such as using mean square error (MSE) or mean absolute error (MAE) as the loss function, but not limited to this. Based on the calculated loss function, this application uses optimization algorithms such as gradient descent to iteratively optimize the parameters of the initial model. The optimization process may require multiple iterations, and each iteration will adjust the model parameters to reduce the loss function value. Iterative optimization can set stop conditions, such as reaching a predetermined number of iterations or the loss function value is lower than a certain threshold.

[0047] In this way, the initial energy loss model is gradually optimized, and finally a model that can more accurately predict the energy loss factor of the cooling process is obtained. Compared with the initial model, this established model can better adapt to the energy loss prediction under different working conditions and improve the accuracy and efficiency of the local cooling control of the rubber ring 90 of the remote plasma source equipment.

[0048] Preferably, step A3 comprises: A301. Filter historical characteristic data groups having the same and / or similar historical characteristic data as the determined characteristic data from the historical database as candidate characteristic data groups; A302. Calculate the energy loss quantification characterization value corresponding to each candidate feature data set according to the established energy loss model and energy loss quantification characterization value calculation formula; A303. Determine the candidate feature data set corresponding to the minimum energy loss quantization characterization value as the optimal feature data set.

[0049] These steps are closely related to each other, forming a closed-loop optimization process. By screening similar cases from historical data, the search efficiency is improved; by using the energy loss model for quantitative evaluation, the scientific nature of the solution selection is ensured; the optimal solution is selected and applied to actual control, and new data is accumulated for the historical database, providing a basis for future optimization. This method not only improves the efficiency of a single cooling, but also continuously improves the control effect by continuously accumulating experience.

[0050] Among them, the determined characteristic data may be one or more of input power, temperature threshold deviation, cooling time and cooling speed, but not all. For example, due to the working requirements of the remote plasma source device, it may be necessary to keep the input power of the remote plasma source device at a preset power, then the determined characteristic data includes the input power, and its value is the preset power; if the cooling speed of the cooling device 1 is not adjustable, the cooling speed is a fixed value, then the determined characteristic data includes the cooling speed, and its value is the fixed value; if the maximum temperature threshold and the minimum temperature threshold of the rubber ring 90 of the remote plasma source device are pre-defined, then the determined characteristic data includes the temperature threshold deviation, and its value is the deviation of the pre-defined maximum temperature threshold and the minimum temperature threshold; if the cooling time is pre-defined, then the determined characteristic data includes the cooling time, and its value is the preset value.

[0051] Among them, in step A301, the historical feature data group having the same historical feature data as the determined feature data is first screened from the historical database. If there is no historical feature data group having the same historical feature data as the determined feature data or the number of the screened historical feature data groups is insufficient, the historical feature data group having historical feature data similar to the determined feature data is screened to supplement it.

[0052] Further, step A302 includes: Input the candidate characteristic data set into the established energy loss model to obtain the energy loss factor of the cooling process output by the established energy loss model, which is recorded as the estimated energy loss factor; The energy loss quantitative characterization value of the candidate characteristic data set is calculated according to the following energy loss quantitative characterization value calculation formula: ; in, is the quantitative representation value of energy loss, is the cooling time in the alternative characteristic data set, is the temperature threshold deviation in the candidate feature data set, is the input power in the candidate characteristic data set, is the heat energy conversion coefficient, To estimate the energy loss factor.

[0053] This calculation method takes into account multiple influencing factors, including cooling time, temperature change, input power, thermal energy conversion efficiency, and energy loss. By comprehensively considering these factors, the energy loss of different candidate feature data sets can be more accurately evaluated.

[0054] Among them, the smaller the energy loss quantification characterization value is, the more conducive the corresponding candidate feature data group is to reducing energy waste. Therefore, in step A303, the candidate feature data group corresponding to the minimum energy loss quantification characterization value is determined as the optimal feature data group. Subsequently, local cooling control is performed according to the optimal feature data group, which can effectively reduce energy waste.

[0055] In this embodiment, step A4 includes: A401. Obtain the highest temperature threshold and the lowest temperature threshold corresponding to the temperature threshold deviation in the optimal feature data set, which are recorded as the first highest temperature threshold and the first lowest temperature threshold respectively; A402. The input power of the remote plasma source device is adjusted to the input power in the optimal characteristic data set; A403. When a temperature sensor 2 detects that the local temperature reaches the first maximum temperature threshold, the cooling device 1 on both sides of the corresponding temperature sensor 2 is started, and the started cooling device 1 is cooled at the cooling speed in the optimal characteristic data group until the local temperature drops to the first minimum temperature threshold, then the cooling device 1 on both sides of the corresponding temperature sensor 2 is turned off, and the working time of the cooling device 1 is recorded; A404. Replace the cooling time in the optimal characteristic data group with the working time to form a new characteristic data group and add it to the historical database.

[0056] The scheme solves the problem of precise control and dynamic optimization through the following technical features: using the temperature threshold deviation in the optimal feature data group to determine the specific maximum temperature threshold and minimum temperature threshold, providing a benchmark for precise temperature control; adjusting the input power to the value in the optimal feature data group to ensure that the equipment works in the best state; based on the real-time detection results of the temperature sensor 2, the cooling device 1 is started when the local temperature reaches the maximum temperature threshold, realizing precise control of the cooling timing; using the cooling speed in the optimal feature data group for cooling, ensuring the efficiency and stability of the cooling process; shutting down the cooling device 1 when the temperature drops to the minimum temperature threshold, avoiding overcooling and improving energy utilization; recording the actual working time and updating the feature data group, realizing dynamic optimization and continuous improvement of the cooling process. The scheme realizes efficient and precise control of the local cooling of the rubber ring 90 of the remote plasma source equipment by accurately controlling the start and stop timing of cooling, optimizing the cooling speed and input power. At the same time, by updating the feature data group in real time, the scheme can continuously optimize the cooling strategy, adapt to changes in the operating status of the equipment, and improve the overall cooling efficiency and energy utilization. This closed-loop optimization mechanism makes the cooling process more intelligent and adaptive, and effectively solves the problem of local cooling control of the rubber ring 90 of the remote plasma source equipment.

[0057] In step A401, obtaining the highest temperature threshold and the lowest temperature threshold corresponding to the temperature threshold deviation in the optimal feature data group is the key to achieving accurate temperature control. This can be achieved in a variety of ways, such as: calculated based on the temperature threshold deviation and the preset core temperature; or dynamically determined through historical data analysis and machine learning algorithms. The choice of these methods depends on the specific application scenario and device characteristics.

[0058] refer to Figure 2 , the present application provides a local cooling system for a rubber ring of a remote plasma source device, comprising a controller and a plurality of cooling devices 1 uniformly arranged along the circumference on the rubber ring 90 of the remote plasma source device, and a temperature sensor 2 is arranged between any two adjacent cooling devices 1; The temperature sensor 2 is used to detect the local temperature of the rubber ring 90 of the remote plasma source device; The cooling device 1 is used to cool the rubber ring 90 of the remote plasma source equipment; The controller is used to: Extracting a plurality of historical characteristic data groups from the historical database, the historical characteristic data groups including the historical input power of the remote plasma source device, the historical temperature threshold deviation, the historical cooling time and the historical cooling speed of the cooling device 1 (refer to step A1 for the specific process); Establishing an energy loss model using the extracted historical feature data set (refer to step A2 for the specific process); According to the determined characteristic data, an optimal characteristic data group is determined using an energy loss model; the characteristic data group includes input power of the remote plasma source device, temperature threshold deviation, cooling time and cooling speed of the cooling device 1, and the determined characteristic data is at least one characteristic data in the characteristic data group (refer to step A3 for the specific process); Based on the local temperature measured in real time by the temperature sensor 2, the remote plasma source device rubber ring 90 is locally cooled according to the optimal characteristic data set, and the characteristic data set is collected during the control process and added to the historical database (refer to step A4 for the specific process).

[0059] exist Figure 2 In the figure, s is a hollow cylindrical cavity (usually a quartz tube), and the rubber ring 90 of the remote plasma source device is set on the hollow cylindrical cavity.

[0060] It should be noted that the cooling device 1 and the temperature sensor 2 are both arranged on one end face of the remote plasma source device rubber ring 90, and the end face of the remote plasma source device rubber ring 90 is a circular ring-shaped side face located between the inner and outer peripheral faces. On the one hand, the inner and outer peripheral faces of the remote plasma source device rubber ring 90 are respectively connected to the hollow cylindrical cavity and the fixing seat of the hollow cylindrical cavity, so there is no space for the cooling device 1 and the temperature sensor 2 to be installed. On the other hand, being arranged on the end face is closer to the hollow cylindrical cavity than being arranged on the outer peripheral face, so that the situation of local temperature exceeding the standard can be discovered more timely and the location where the local temperature is too high can be cooled more quickly.

[0061] The selection of temperature sensor 2 is crucial to achieve accurate control of the cooling timing. A highly sensitive thermocouple or thermistor can be considered, with a temperature measurement accuracy of ±0.1°C.

[0062] The cooling device 1 may be a liquid cooling device or a semiconductor cooling device.

[0063] The controller may be arranged outside the remote plasma source device rubber ring 90 and electrically connected to each cooling device 1 and each temperature sensor 2, or as Figure 2 As shown, the controller includes multiple control modules 3, and a control module 3 is connected between each temperature sensor 2 and the adjacent cooling device 1. When any temperature sensor 2 detects that the local temperature reaches the maximum temperature threshold, the control module 3 connected thereto will control the cooling device 1 connected thereto to start, until the local temperature drops to the minimum temperature threshold, that is, the cooling device 1 connected thereto is controlled to shut down; for the cooling device 1, as long as one of the two control modules 3 connected thereto requires the cooling device 1 to work, the cooling device 1 remains in working state.

[0064] In some embodiments, see Figure 2The local cooling system of the rubber ring of the remote plasma source equipment also includes a plurality of indicating devices 4, each indicating device 4 is electrically connected to each temperature sensor 2 in a one-to-one correspondence, and the indicating device 4 is used to send an indicating signal according to the local temperature measured by the corresponding temperature sensor 2.

[0065] By adding the indicating device 4, the technical solution achieves the following effects: providing visual feedback of the temperature status, so that the operator can intuitively understand the temperature conditions of various parts of the rubber ring; helping to detect abnormal temperature changes in a timely manner, improving the safety and reliability of the system; facilitating the operator to monitor and adjust the cooling process, improving the accuracy and efficiency of the cooling control. The solution enhances the functionality and practicality of the local cooling system of the rubber ring of the remote plasma source equipment in a simple and effective way. Compared with relying solely on the controller for automatic control, the system operation status is more transparent and controllable. This improvement has a certain degree of creativity and provides a new technical idea for solving the problem of intuitive indication of temperature status.

[0066] The indicating device 4 can adopt various forms to realize the intuitive indication of the temperature status. For example, the indicating device 4 can include an LED light, and different colors represent different temperature intervals. When the temperature is lower than the minimum temperature threshold, the LED light can display green; when it is within the range of the minimum temperature threshold to the maximum temperature threshold, it can display yellow; when the temperature exceeds the maximum temperature threshold, it displays red. Another implementation method is to use a digital display screen (that is, the indicating device 4 includes a digital display screen) to directly display the specific temperature value measured by the temperature sensor 2. An audible and visual alarm can also be used to emit an alarm sound and a flashing signal when the temperature exceeds the set threshold.

[0067] Furthermore, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0068] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0069] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for local cooling of rubber rings of remote plasma source equipment, characterized in that: Based on a local cooling system, the local cooling system includes a plurality of cooling devices uniformly arranged circumferentially on the rubber ring of the remote plasma source device, and a temperature sensor is arranged between any two adjacent cooling devices; The method for local cooling of the rubber ring of the remote plasma source equipment comprises the following steps: A1. extracting a plurality of historical characteristic data sets from a historical database, wherein the historical characteristic data sets include historical input power of a remote plasma source device, historical temperature threshold deviation, historical cooling time, and historical cooling speed of a cooling device; A2. Establishing an energy loss model using the extracted historical feature data set; A3. According to the determined characteristic data, using the energy loss model, determine the optimal characteristic data group; the characteristic data group includes the input power of the remote plasma source device, the temperature threshold deviation, the cooling time and the cooling speed of the cooling device, and the determined characteristic data is at least one characteristic data in the characteristic data group; A4. Based on the local temperature measured in real time by the temperature sensor, the local temperature reduction control is performed on the rubber ring of the remote plasma source device according to the optimal characteristic data set, and the characteristic data set is collected during the control process and added to the historical database.

2. The method for local cooling of the rubber ring of a remote plasma source device according to claim 1, characterized in that: The historical temperature threshold deviation is the deviation between the historical highest temperature threshold and the historical lowest temperature threshold; the historical cooling time is the time taken for the local temperature of the rubber ring of the remote plasma source device to drop from the historical highest temperature threshold to the historical lowest temperature threshold under the conditions of the historical input power and the historical cooling speed; Correspondingly, the temperature threshold deviation is the deviation between the highest temperature threshold and the lowest temperature threshold; the cooling time is the time it takes for the local temperature of the rubber ring of the remote plasma source device to drop from the highest temperature threshold to the lowest temperature threshold under the conditions of the input power and the cooling speed; The local temperature of the rubber ring of the remote plasma source equipment is measured by a temperature sensor.

3. The method for local cooling of the rubber ring of a remote plasma source device according to claim 2, characterized in that: Step A2 includes: A201. Establish an initial energy loss model; A202. Use the extracted historical feature data set to train the initial energy loss model to obtain a confirmed energy loss model.

4. The method for local cooling of the rubber ring of a remote plasma source device according to claim 3, characterized in that: In step A201, an initial energy loss model is established according to the following formula: ； ； in, is the cooling time, is the temperature threshold deviation, is the input power of the remote plasma source device, is the heat energy conversion coefficient, is the energy loss factor before cooling, is the energy loss factor of the cooling process, is the cooling speed of the cooling device, is the energy loss factor function, is the error factor.

5. The method for local cooling of the rubber ring of a remote plasma source device according to claim 4, characterized in that: Step A202 includes: The extracted historical characteristic data group is used as sample data, and a reference value of the energy loss factor of the corresponding cooling process is calculated according to each sample data; Input each of the sample data into the initial energy loss model in turn, and use the estimated value of the energy loss factor of the cooling process output by the initial energy loss model and the corresponding reference value to calculate the loss function, so as to iteratively optimize the model parameters of the initial energy loss model to obtain the established energy loss model.

6. The method for local cooling of the rubber ring of a remote plasma source device according to claim 1, characterized in that: Step A3 includes: A301. Filtering a historical feature data group having the same and / or similar historical feature data as the determined feature data from the historical database as an alternative feature data group; A302. Calculate the energy loss quantification characterization value corresponding to each of the candidate feature data sets according to the established energy loss model and energy loss quantification characterization value calculation formula; A303. Determine the candidate feature data set corresponding to the minimum energy loss quantitative characterization value as the optimal feature data set.

7. The method for local cooling of the rubber ring of a remote plasma source device according to claim 6, characterized in that: Step A302 includes: Input the candidate characteristic data set into the established energy loss model to obtain the energy loss factor of the cooling process output by the established energy loss model, which is recorded as the estimated energy loss factor; The energy loss quantification characterization value of the candidate feature data set is calculated according to the following energy loss quantification characterization value calculation formula: ； in, is a quantitative characterization value of the energy loss, is the cooling time in the candidate feature data set, is the temperature threshold deviation in the candidate feature data set, is the input power in the candidate feature data set, is the heat energy conversion coefficient, is the estimated energy loss factor.

8. The method for local cooling of the rubber ring of a remote plasma source device according to claim 1, characterized in that: Step A4 includes: Obtaining a maximum temperature threshold and a minimum temperature threshold corresponding to the temperature threshold deviation in the optimal feature data group, which are recorded as a first maximum temperature threshold and a first minimum temperature threshold respectively; Adjusting the input power of the remote plasma source device to the input power in the optimal characteristic data set; When a temperature sensor detects that the local temperature reaches the first maximum temperature threshold, the cooling devices on both sides of the corresponding temperature sensor are started, and the started cooling devices are cooled at the cooling speed in the optimal characteristic data group until the local temperature drops to the first minimum temperature threshold, then the cooling devices on both sides of the corresponding temperature sensor are turned off, and the working time of the cooling devices is recorded; The cooling time in the optimal characteristic data set is replaced by the working time to form a new characteristic data set, and is added to the historical database.

9. A local cooling system for rubber rings of remote plasma source equipment, characterized in that: It includes a controller and a plurality of cooling devices evenly arranged along the circumference on the rubber ring of the remote plasma source equipment, and a temperature sensor is arranged between any two adjacent cooling devices; The temperature sensor is used to detect the local temperature of the rubber ring of the remote plasma source equipment; The cooling device is used to cool the rubber ring of the remote plasma source equipment; The controller is used to: Extracting a plurality of historical characteristic data sets from a historical database, wherein the historical characteristic data sets include historical input power of the remote plasma source device, historical temperature threshold deviation, historical cooling time, and historical cooling speed of the cooling device; Establishing an energy loss model using the extracted historical feature data set; According to the determined characteristic data, the optimal characteristic data group is determined by using the energy loss model; the characteristic data group includes the input power of the remote plasma source device, the temperature threshold deviation, the cooling time and the cooling speed of the cooling device, and the determined characteristic data is at least one characteristic data in the characteristic data group; Based on the local temperature measured in real time by the temperature sensor, the local temperature reduction control is performed on the rubber ring of the remote plasma source equipment according to the optimal characteristic data group, and the characteristic data group is collected during the control process and added to the historical database.

10. The local cooling system of the rubber ring of the remote plasma source equipment according to claim 9, characterized in that: It also includes a plurality of indicating devices, each of which is electrically connected to each of the temperature sensors in a one-to-one correspondence, and is used to send an indicating signal according to the local temperature measured by the corresponding temperature sensor.

Images