Electrostatic chuck leak rate detection system

Abstract

This application discloses an electrostatic chuck leak rate detection system, comprising: multiple channels connecting a gas source to the vacuum chamber of the electrostatic chuck under test, transmitting gas supplied from the gas source to the vacuum chamber; multiple universal pressure controller modules disposed outside the vacuum chamber and connected to the multiple channels one by one, each universal pressure controller module including a universal pressure controller, a pressure sensor, a flow sensor, and an electric regulating valve; a main control unit module communicating with the multiple universal pressure controller modules, used to: input a configured preset voltage and preset back pressure into a preset three-dimensional dynamic correlation model of voltage-back pressure-leak rate, so as to output the corresponding leak rate prediction value; when it is determined that there is a leak in the electrostatic chuck under test based on the leak rate prediction value and the leak rate detection value detected by the gas flow rate collected through each channel, the corresponding electric regulating valve is controlled by each universal pressure controller to adjust the gas flow rate transmitted to the vacuum chamber to form a preset back pressure.

Description

Technical Field

[0001] This application relates to the field of semiconductor manufacturing technology, and in particular to an electrostatic chuck leakage rate detection system. Background Technology

[0002] An electrostatic chuck (ESC) is an adsorption-type fixture used to hold wafers and other objects in place. It is widely used in semiconductor manufacturing, panel displays, and precision optics. The ESC uses electrostatic force generated by a preset voltage to hold the wafer to be processed on its surface. The vacuum chamber between the ESC and the wafer uses a back-blown gas, such as an inert gas including helium, to control the wafer surface temperature. By setting the required back-blown gas pressure (referred to as "back pressure"), the wafer's heat dissipation efficiency can be improved, effectively reducing the wafer temperature.

[0003] Electrostatic chucks are used in high-vacuum processes such as semiconductor manufacturing, where sealing and leakage rates are critical. Leakage rate is a key indicator for evaluating the sealing performance and process stability of electrostatic chuck equipment. By collecting the actual gas flow rate in the vacuum chamber under set back pressure, the leakage rate of the electrostatic chuck can be detected, thereby identifying any leaks.

[0004] Leakage detection of electrostatic chucks involves parameters including the voltage used to generate the electrostatic force for adsorption and the back pressure used to control the wafer temperature. Leakage detection requires switching the same electrostatic chuck under test at different voltage and / or back pressure combinations. Existing electrostatic chuck leakage detection technology uses a static, manual method to configure and switch the corresponding parameters, with manual intervention in adjusting the voltage and back pressure to ensure that the actual voltage-back pressure generated by the electrostatic chuck under test matches the required voltage-back pressure.

[0005] Thus, the static, manual configuration and switching of parameters results in a long time consumption for a single leak rate test, typically exceeding 30 minutes in practice. This is incompatible with the precision testing requirements of mass production of electrostatic chucks and makes it difficult to adapt to the mass production needs of electrostatic chucks in the semiconductor industry. Furthermore, manual debugging leads to significant errors in the leak rate test results, severely impacting the accuracy and reliability of the test. Therefore, the existing electrostatic chuck leak rate detection is inefficient and has significant limitations in practical industrial applications. Summary of the Invention

[0006] The embodiments of this application aim to provide an electrostatic chuck leak rate detection system to solve the problems of low efficiency, insufficient accuracy and reliability in the existing electrostatic chuck leak rate detection technology.

[0007] To achieve the above objectives, embodiments of this application provide an electrostatic chuck leak rate detection system. This system revolutionizes the traditional manual detection process by integrating automated control and intelligent prediction models.

[0008] In a first aspect, the electrostatic chuck leakage rate detection system provided in the embodiments of this application includes: Multiple pathways connect the gas source to the vacuum chamber of the electrostatic chuck under test, used to transfer gas supplied from the gas source to the vacuum chamber; Multiple universal pressure controller modules are located outside the vacuum chamber and are connected to multiple corresponding passages. Each universal pressure controller module includes a universal pressure controller, a pressure sensor, a flow sensor, and an electric regulating valve. The pressure sensor is used to collect the back pressure in the vacuum chamber corresponding to the passage it is connected to, the flow sensor is used to collect the gas flow in the vacuum chamber corresponding to the passage it is connected to, and the electric regulating valve is connected to the gas source and the gas inlet of the connected passage to regulate the gas flow transmitted to the vacuum chamber by each passage. The main control unit module communicates with multiple general-purpose pressure controller modules and is used to: control each general-purpose pressure controller to configure the preset voltage and preset back pressure corresponding to the electrostatic chuck under test based on instructions carrying different voltage-back pressure combinations required for each leak rate test of the electrostatic chuck under test; input the configured preset voltage and preset back pressure into a preset three-dimensional dynamic correlation model of voltage-back pressure-leakage rate, so as to output the corresponding leak rate prediction value through the three-dimensional dynamic correlation model; when it is determined that there is a leak in the electrostatic chuck under test based on the leak rate prediction value and the leak rate detection value detected by the gas flow rate collected through each channel, control the corresponding electric regulating valve through each general-purpose pressure controller to adjust the gas flow rate transmitted to the vacuum chamber, so as to form a preset back pressure in the vacuum chamber.

[0009] The system mainly consists of a gas delivery pathway module, a universal pressure controller module, and a main control unit module. The gas delivery pathway module connects the gas source to the vacuum chamber of the electrostatic chuck under test, forming a gas transmission channel. The universal pressure controller module, as the core execution and sensing unit, is connected to each pathway. Each module includes a universal pressure controller (UPC), a pressure sensor, a flow sensor, and an electric regulating valve, enabling precise measurement of the vacuum chamber back pressure and monitoring and regulation of the gas flow. The main control unit module acts as the system's brain, communicating with each universal pressure controller module and coordinating the entire testing process.

[0010] The main control unit module operates using a dynamic and continuous intelligent detection method: First, it receives detection commands containing a series of different voltage-backpressure combinations and automatically controls each general-purpose pressure controller to configure the preset voltage and backpressure corresponding to the electrostatic chuck under test. Second, it inputs the configured preset voltage and preset backpressure into a preset three-dimensional dynamic correlation model of voltage-backpressure-leakage rate, outputting the corresponding leakage rate prediction value through the three-dimensional dynamic correlation model. Third, under stable backpressure conditions, based on the gas flow data collected by the flow sensors in each channel, it automatically calculates and judges the leakage rate detection value of the electrostatic chuck under test, and combines it with the leakage rate prediction value to determine whether the leakage rate detection value is qualified, i.e., whether the electrostatic chuck under test has a leak. When a leak is determined to exist in the electrostatic chuck under test, based on the backpressure data fed back in real time by the pressure sensors in each channel, it drives the corresponding electric regulating valve through the general-purpose pressure controller to dynamically adjust the input gas flow rate, so that the preset backpressure is quickly and stably formed in the vacuum chamber. Finally, under stable backpressure conditions, based on the gas flow data collected by the flow sensors in each channel, it automatically calculates and judges whether the leakage rate of the electrostatic chuck under test is qualified.

[0011] The electrostatic chuck leak detection system of this application, through the aforementioned modular design and automated process, incorporates a three-dimensional dynamic correlation model into the detection system. When leakage occurs in the electrostatic chuck, the detection system can directly determine the leakage based on the data predicted by the three-dimensional dynamic correlation model and the current actual detection data. It then performs rapid gas compensation and pressure adjustment, ensuring that the actual back pressure formed in the vacuum chamber of the electrostatic chuck accurately meets the preset back pressure requirements. This achieves automatic distribution, configuration, and free switching of detection parameters, as well as closed-loop automatic adjustment of back pressure. This significantly shortens the time required for a single detection, enables dynamic, continuous, and high-precision leak detection, better meets the urgent need for efficient and precise detection in the mass production of electrostatic chucks in the semiconductor industry, greatly improves detection efficiency and result reliability, and expands the industrial application scope of this technology. Attached Figure Description

[0012] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a structural block diagram of the electrostatic chuck leakage rate detection system according to an embodiment of this application.

[0013] Figure 2 This is a structural block diagram of a general pressure controller module according to an embodiment of this application.

[0014] Figure 3 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0015] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0016] To address the problems existing in the prior art, this application provides an electrostatic chuck leak rate detection system, comprising: multiple channels connecting a gas source to the vacuum chamber of the electrostatic chuck under test, used for transmitting gas supplied from the gas source to the vacuum chamber; multiple universal pressure controller modules disposed outside the vacuum chamber and connected to the multiple channels one by one, each universal pressure controller module including a universal pressure controller, a pressure sensor, a flow sensor, and an electric regulating valve; wherein, the pressure sensor is used to collect the back pressure in the vacuum chamber corresponding to the channel it is connected to, the flow sensor is used to collect the gas flow rate in the vacuum chamber corresponding to the channel it is connected to, and the electric regulating valve is connected to the gas source and the gas inlet of the connected channel, used to regulate the gas flow rate transmitted to the vacuum chamber by each channel; main The control unit module communicates with multiple general-purpose pressure controller modules and is used to: control each general-purpose pressure controller to configure the preset voltage and preset back pressure corresponding to the electrostatic chuck under test based on instructions carrying different voltage-back pressure combinations required for each leak rate detection of the electrostatic chuck under test; input the configured preset voltage and preset back pressure into a preset three-dimensional dynamic correlation model of voltage-back pressure-leakage rate, so as to output the corresponding leak rate prediction value through the three-dimensional dynamic correlation model; when it is determined that there is a leak in the electrostatic chuck under test based on the leak rate prediction value and the leak rate detection value detected by the gas flow rate collected through each channel, based on the back pressure collected by each channel, control the corresponding electric regulating valve through each general-purpose pressure controller to adjust the gas flow rate transmitted to the vacuum chamber, so as to form a preset back pressure in the vacuum chamber.

[0017] like Figure 1 As shown, the electrostatic chuck leak detection system includes multiple channels 302 and multiple universal pressure controller (UPC) modules that are connected one-to-one with each of the multiple channels 302. Figure 1 The UPC module 304 shown.

[0018] Electrostatic chuck leakage rate detection mainly involves placing the electrostatic chuck 200 under test in a vacuum chamber 400 environment, with the wafer (not shown in the figure) located above the electrostatic chuck 200 under test in the vacuum chamber environment. The electrostatic chuck 200 under test uses electrostatic force to adsorb the lower surface of the wafer. Figure 1The circle in the illustrated embodiment is a schematic diagram of the electrostatic chuck 200 under test from a top view angle, and the corresponding vacuum chamber 400 is a closed cylindrical chamber. It can be understood that the projection of the wafer above the electrostatic chuck 200 under test in the vacuum chamber 400 coincides with the circle of the electrostatic chuck 200 under test.

[0019] Passage 302, serving as a gas delivery module, connects the gas source to the vacuum chamber of the electrostatic chuck under test (ESC2), forming a gas transmission channel. Each passage 302 connects the gas source 100 to the vacuum chamber 400 of the ESC2 200 under test. The gas source 100 provides the inert gas, such as helium, required for leak rate detection of the ESC2 200. Passage 302 transmits the gas provided by the gas source 100 to the vacuum chamber 400. The gas provided by the gas source 100 is uniformly released into the vacuum chamber 400 through micropores on the ESC2 200 and permeates to the lower surface of the wafer above the ESC2 200, which is electrostatically adsorbed, forming back pressure as a back-blown gas.

[0020] Combination Figure 1 Multiple UPC modules 304 are disposed around the outside of the vacuum chamber 400 and are connected to multiple passages 302 respectively. The UPC module 304 can be connected to the corresponding passage 302 through a pipe mechanically connected to the passage 302 for transmitting gas. Each UPC module 304 includes functional units such as a general pressure controller, a pressure sensor, a flow sensor and an electric regulating valve.

[0021] The general-purpose pressure controller is a high-precision pressure control device widely used in semiconductor manufacturing, laboratory equipment, and industrial production lines. Here, it is used to control the pressure of the gas supplied to the vacuum chamber at 400°C.

[0022] Combination Figure 2 The pressure sensor 3044 is used to collect the back pressure in the connected vacuum chamber 400 corresponding to the access path, i.e., the path 302 connected to the corresponding UPC module 304. The pressure sensor 3044 obtains the back pressure in the connected vacuum chamber 400 corresponding to the access path 302 by collecting the pressure in the access path 302.

[0023] The flow sensor 3046 is used to collect the gas flow rate in the connected vacuum chamber 400 corresponding to the access path, i.e., the path 302 connected to the corresponding UPC module 304. By collecting the gas flow rate in the access path 302, the flow sensor 3046 obtains the gas flow rate in the connected vacuum chamber 400 corresponding to the path 302.

[0024] The electric regulating valve 3048 is physically connected to the gas source 100 and the gas inlet 3022 of the access passage 302. The gas inlet 3022 is the port for inputting gas into the vacuum chamber 400 corresponding to the access passage 302. Correspondingly, the gas outlet 3024 of the access passage 302 is located inside the vacuum chamber 400 and is the port for outputting or injecting gas from the gas inlet 3022 into the vacuum chamber 400.

[0025] The gas supplied by the gas source 100 is transmitted to the vacuum chamber 400 through the valves of the electrically controlled regulating valves 3048 connected to each passage 302 and the gas inlet 3022 and gas outlet 3024 of each passage 302. By adjusting the opening degree of the electrically controlled regulating valves 3048, the gas flow rate transmitted to the vacuum chamber through each passage 302 can be adjusted. The larger the valve opening degree, the larger the gas flow rate transmitted to the vacuum chamber; conversely, the smaller the valve opening degree, the smaller the gas flow rate transmitted to the vacuum chamber.

[0026] Optionally, the UPC module 304 can be an integrated UPC module including a pressure sensor, a flow sensor and an electric regulating valve. Alternatively, a combination of an integrated pressure / flow sensor of the same accuracy level and an independent high-precision electric regulating valve can be used to replace the integrated UPC module to achieve the same gas flow control and back pressure and gas flow data acquisition functions.

[0027] The main control unit module 306 communicates with multiple UPC modules 304 in the same manner. Figure 2 A block diagram showing the communication connection structure between the main control unit module 306 and one of the UPC modules 304 is shown.

[0028] When performing leak rate testing on the same electrostatic chuck, it is necessary to consider the voltage and back pressure that the electrostatic chuck may use during actual operation. This means testing whether gas leakage exists under different voltage-back pressure combinations. Each leak rate test is performed for one voltage-back pressure combination, and subsequent tests require switching to a different combination. By combining leak rate tests under different voltage-back pressure combinations, the sealing performance of the electrostatic chuck can be checked to see if it meets the requirements.

[0029] The main control unit module 306 can send the voltage-back pressure combination required for each test of the leakage rate of the electrostatic chuck 200 to the general pressure controller 3042 included in each UPC module 304 in the form of an instruction, thereby controlling each general pressure controller 3042 to automatically configure the preset voltage and preset back pressure corresponding to the electrostatic chuck 200 according to the currently received instruction.

[0030] Specifically, each general-purpose pressure controller 3042 is configured with the following parameters according to the received instructions: pressure measurement range, for example, a range of 0~50 Torr, with an accuracy of ±0.3% of full scale (FS); flow measurement range, for example, a range of 0.01~100 μL / s, with a response time <10 ms; electric regulating valve, for example, an electric needle valve with an adjustment accuracy of 0.1 mm³ / min; electrostatic adsorption electrode: adsorption range for dual-electrode version or adsorption voltage range for single-electrode version; and power supply parameters for electrostatic adsorption power supply.

[0031] By automatically sending instructions carrying the aforementioned configuration parameters to each general-purpose pressure controller 3042 via the main control unit module 306, the automatic configuration of the voltage-back pressure combination required for detecting the leakage rate of the electrostatic chuck 200 under test can be achieved. Correspondingly, after the current test is completed, the main control unit module 306 can automatically send instructions carrying the configuration parameters required for the next test, and automatically switch the voltage-back pressure combination through the configuration of each general-purpose pressure controller 3042. The different voltage-back pressure combinations required for leak rate detection of the electrostatic chuck 200 under test can be recorded in a queue and input to the main control unit module 306. The main control unit module 306 reads the corresponding voltage-back pressure combination in the queue during each test and sends it to each general-purpose pressure controller 3042 in the form of an instruction.

[0032] After configuring the preset voltage and preset back pressure required for this leak rate detection by the electrostatic chuck 200 under test in each general pressure controller 3042, the pressure sensor 3044, flow sensor 3046 and electric regulating valve 3048 included in each UPC module 304 also work at the same time.

[0033] Under normal circumstances, after initially configuring the preset voltage and preset back pressure, the main control unit module 306 needs to control the electric regulating valves corresponding to each channel to adjust the gas flow rate transmitted to the vacuum chamber in order to form a preset back pressure in the region within the vacuum chamber. When the main control unit module 306 controls the formation of the preset back pressure, it needs to combine the deviation between the back pressure collected by the pressure sensors connected to each channel and the preset back pressure to determine whether the current gas flow rate transmitted to the vacuum chamber can form a stable preset back pressure.

[0034] Here, the difference between the collected back pressure and the preset back pressure includes cases where the collected back pressure is greater than or less than the preset back pressure, cases where the difference between the collected back pressure and the preset back pressure is very small with slight fluctuations, and cases where the difference exceeds the preset back pressure threshold. When the difference is very small, the collected back pressure is considered equal to the preset back pressure and no adjustment is needed. When the difference exceeds the preset back pressure threshold, the electric regulating valve in the corresponding path needs to be controlled to adjust the gas flow rate transmitted to the vacuum chamber 400 to establish a stable preset back pressure.

[0035] The main control unit module 306, based on the back pressure collected by the pressure sensors 3044 corresponding to each passage 302, determines whether the back pressure in the vacuum chamber 400 has reached the preset back pressure, and controls the electric regulating valve 3048 to precisely inject the gas required to form the preset back pressure. The gas provided by the gas source 100 is injected into the back of the electrostatic chuck 200 under test through each passage 302, and uniformly permeates through the micropores on the electrostatic chuck 200 to the lower surface of the wafer adsorbed by electrostatic force, so as to form the required stable preset back pressure. When the stable preset back pressure is reached, the main control unit module 306 can calculate the actual leakage rate (i.e., leakage rate detection value) of the electrostatic chuck 200 under the current preset voltage-back pressure combination based on the gas flow rate collected by the flow sensors 3046 of each passage 302, and detect whether the leakage rate of the electrostatic chuck under test meets the sealing requirements.

[0036] If the leakage rate of the electrostatic chuck under test does not meet the sealing requirements, i.e., a leak exists, even if the main control unit module 306 controls the vacuum chamber 400 to initially reach a stable preset back pressure, subsequent gas will escape from the leak and be released into the vacuum chamber 400, causing a localized drop in back pressure near the leak within the vacuum chamber 400. At this time, the main control unit module 306 also needs to control the electric regulating valves corresponding to each passage to adjust the gas flow rate transmitted to the vacuum chamber in order to form a stable preset back pressure in the area within the vacuum chamber.

[0037] Specifically, the main control unit module is used to: when the leak rate prediction value is different from the leak rate detection value of the gas flow detection collected in the target path, control the electric regulating valve corresponding to the target path to adjust the gas flow transmitted to the vacuum chamber, so that a preset back pressure is formed in the vacuum chamber of the area corresponding to the target path.

[0038] When the main control unit module 306 controls the formation of the preset back pressure, it needs to combine the back pressure collected by the pressure sensors connected to each channel with the deviation of the preset back pressure to determine whether the current gas flow rate transmitted to the vacuum chamber can form a stable preset back pressure.

[0039] In this embodiment, the main control unit module 306 determines whether the electrostatic chuck under test has a leak based on the leak rate prediction value and the leak rate detection value. Specifically, the main control unit module 306 uses the preset voltage and preset back pressure corresponding to the current leak rate detection as input data, inputs them into a preset three-dimensional dynamic correlation model of voltage-back pressure-leak rate, and predicts and outputs the corresponding leak rate prediction value through the three-dimensional dynamic correlation model. Furthermore, it detects the actual leak rate detection value by collecting gas flow rates from each channel. Then, by comparing the leak rate detection value and the leak rate prediction value, it determines whether the electrostatic chuck under test has a leak. For example, if the difference between the leak rate detection value and the leak rate prediction value does not fall within the preset leak rate range corresponding to the sealing requirements, it is determined that the electrostatic chuck under test has a leak.

[0040] When calculating the leak rate based on the collected gas flow rate, the main control unit module 306 can pre-process the collected gas flow rate data and filter abnormal data, thereby effectively suppressing noise data caused by environmental vibration interference in the vacuum chamber and improving the accuracy of leak rate detection.

[0041] When the electrostatic chuck 200 under test is in normal working condition and has good sealing performance, the gas leakage of the electrostatic chuck 200 is extremely low. The main control unit module 306 can make fine adjustments through the electric regulating valves 3048 included in each UPC module 304 to maintain a constant back pressure. The gas flow detected by the flow sensors 3046 connected to each passage 302 is low and stable, and the gas flow is lower than the preset flow threshold, for example, 0.5 standard cubic centimeters per minute (SCCM / MIN), indicating that the electrostatic adsorption and sealing performance of the electrostatic chuck 200 under test is qualified.

[0042] Standard cubic centimeters per minute (SCCM / MIN) is a unit of gas flow rate that indicates the volume of gas flowing through per minute under standard temperature and pressure conditions in cubic centimeters.

[0043] The main control unit module 306 can use high-precision dynamic pressure control algorithms such as proportional-integral-derivative (PID) feedforward control algorithm, fuzzy PID or adaptive PID to realize real-time dynamic compensation of back pressure in the vacuum chamber and maintain constant back pressure.

[0044] In this embodiment, the main control unit module operates as a dynamic and continuous intelligent detection method: First, it receives detection commands containing a series of different voltage-backpressure combinations and automatically controls each general-purpose pressure controller to configure the preset voltage and backpressure corresponding to the electrostatic chuck under test. Second, it inputs the configured preset voltage and preset backpressure into a preset three-dimensional dynamic correlation model of voltage-backpressure-leakage rate, so as to output the corresponding leakage rate prediction value through the three-dimensional dynamic correlation model. Third, under stable backpressure conditions, based on the gas flow data collected by each channel flow sensor, it automatically calculates and judges the leakage rate detection value of the electrostatic chuck under test, and combines it with the leakage rate prediction value to determine whether the leakage rate detection value is qualified, that is, whether the electrostatic chuck under test has a leak. When it is determined that the electrostatic chuck under test has a leak, based on the backpressure data fed back in real time by each channel pressure sensor, it drives the corresponding electric regulating valve through the general-purpose pressure controller to dynamically adjust the input gas flow rate, so that the preset backpressure is quickly and stably formed in the vacuum chamber. Finally, under stable backpressure conditions, based on the gas flow data collected by each channel flow sensor, it automatically calculates and judges whether the leakage rate of the electrostatic chuck under test is qualified.

[0045] In this embodiment, the parameters involved in the leak rate detection of the electrostatic chuck can be automatically issued and configured through the main control unit module and the general-purpose pressure controller during each leak rate detection. This enables automatic and free switching between different voltage and / or back pressure combinations, and automatic back pressure adjustment through the main control unit module in conjunction with pressure sensors and electric regulating valves, ensuring that the actual back pressure formed by the electrostatic chuck under test accurately matches the configured preset back pressure. The automatic parameter configuration and switching significantly shortens the time required for a single leak rate detection, enabling dynamic and continuous detection of the electrostatic chuck leak rate; for example, in practice, a single detection time is less than 5 minutes. Furthermore, automatic adjustment improves the accuracy and reliability of leak rate detection. The electrostatic chuck leak rate detection system of this embodiment is adapted to the precision detection requirements of mass production of electrostatic chucks, meets the mass production needs of electrostatic chucks in the semiconductor industry, improves the efficiency of electrostatic chuck leak rate detection, and expands the scope of practical industrial applications.

[0046] If the leak rate prediction and leak rate detection values ​​indicate that the sealing performance of the electrostatic chuck 200 under test does not meet the requirements, it is determined that there is a leak in the electrostatic chuck under test. To effectively locate the leak and reduce troubleshooting costs and time, the back pressure corresponding to the leak location can be quickly and accurately adjusted and compensated. Optionally, in one specific embodiment, the gas outlets of multiple channels are symmetrically distributed relative to the vacuum chamber. The main control unit module is also used to: locate the leak based on the gas propagation speed within the vacuum chamber, and the back pressure and gas flow data collected before and after adjustment by the electric regulating valves of multiple channels; control the electric regulating valve corresponding to the channel closest to the leak among the multiple channels to adjust the gas flow transmitted to the vacuum chamber, so that a preset back pressure is formed in the vacuum chamber corresponding to the area of ​​the closest channel.

[0047] like Figure 1 As shown, the system includes three passages 302, each with a gas outlet 3024 symmetrically distributed relative to the vacuum chamber 400. Thus, the flow sensor 3046 of each passage collects the gas flow rate in the area near the corresponding gas outlet 3024. It should be noted that this application is not limited to the number of passages in this embodiment; two passages, or four or more passages, may also be used in the embodiment.

[0048] As described above, when a leak exists in the electrostatic chuck 200 under test, gas will escape from the leak and be released into the vacuum chamber 400, causing a localized drop in back pressure near the leak within the vacuum chamber 400. The difference between the back pressure drop caused by the leak and the preset back pressure will exceed the preset back pressure threshold. Correspondingly, the pressure sensor 3044, which collects the back pressure in the area near the leak, will collect a back pressure that is less than the preset back pressure but exceeds the preset back pressure threshold.

[0049] To maintain the preset back pressure, the main control unit module 306 controls the electric regulating valve in the corresponding passage to increase the gas flow rate injected into the vacuum chamber 400 for compensation, thereby forming the preset back pressure. Before and after the electric regulating valve 3048 increases the injected gas flow rate, i.e., at the corresponding moments before and after the electric regulating valve adjusts, the gas flow rate collected by the flow sensor 3046 will show a significant increase. Similarly, the back pressure collected by the pressure sensor 3044 will also show a significant pressure change, rising from a locally decreasing back pressure before adjustment to the preset back pressure after adjustment and compensation.

[0050] The main control unit module 306 can locate the leak point by combining back pressure and gas flow data collected before and after the electric regulating valve is adjusted along the pathway near the leak point, along with the gas propagation speed within the vacuum chamber 400 (e.g., the propagation speed of helium cHe≈972 m / s). For the same leak point in the electrostatic chuck 200 under test, there may be two or more pathways where the gas flow and back pressure collected before and after the electric regulating valve adjustment show significant changes.

[0051] Specifically, the main control unit module locates the leak by: acquiring the first back pressure collected by the pressure sensor corresponding to the first path in multiple paths before its electric regulating valve is adjusted, wherein the first back pressure is less than a preset back pressure and the difference between the first and preset back pressures exceeds a preset back pressure threshold; acquiring the second back pressure collected by the pressure sensor corresponding to the second path adjacent to the first path in multiple paths before its electric regulating valve is adjusted, wherein the second back pressure is less than a preset back pressure and the difference between the second and preset back pressures exceeds a preset back pressure threshold; determining a first moment, which is the first acquisition moment of the pressure sensor and flow sensor after the electric regulating valve corresponding to the first path is adjusted; determining a second moment, which is the first acquisition moment of the pressure sensor and flow sensor after the electric regulating valve corresponding to the second path is adjusted, wherein the second moment is later than the first moment; and calculating and locating the leak location based on the gas propagation speed in the vacuum chamber, the first moment, the second moment, the first back pressure, the second back pressure, and the distance between the gas outlet of the first path and the gas outlet of the second path.

[0052] This embodiment addresses the scenario where the electrostatic chuck 200 under test has a leak, causing the deviation between the back pressure collected by the pressure sensor corresponding to the first path and the preset back pressure to exceed the preset back pressure threshold for the absence of a leak (i.e., meeting the sealing requirements). The first path is any one of multiple paths, and the second path is the path adjacent to the first path. If there are three or more paths, the second path includes two paths adjacent to both sides of the first path. When a gas leak occurs, the gas flow rate is higher in the path closer to the leak. By using multiple paths at fixed positions, the location of the leak can be determined.

[0053] For the case where there is no leak in the electrostatic chuck 200 under test, the flow rate of multiple channels is consistent and close to 0, and the deviation between the back pressure collected by the pressure sensor corresponding to each channel and the preset back pressure does not exceed the preset back pressure threshold.

[0054] After receiving the back pressure data collected by the pressure sensors of each channel, the main control unit module 306 determines that if the back pressure collected by the pressure sensor of the first channel at the target time is less than the preset back pressure and the difference between the two exceeds the preset back pressure threshold, then a leak point exists near the first channel. That is, the target time is the moment when the pressure sensor of the first channel collects abnormal data (first back pressure) indicating a leak point before the electric regulating valve is adjusted. Similarly, if the main control unit module 306 determines that the back pressure collected by the pressure sensor of the second channel at the target time is less than the preset back pressure and the difference between the two exceeds the preset back pressure threshold, then a leak point exists near the second channel. That is, the target time is the moment when the pressure sensor of the second channel collects abnormal data (second back pressure) indicating a leak point before the electric regulating valve is adjusted.

[0055] The data for the first and second back pressures are collected before the electric regulating valves in the corresponding channels are adjusted. Back pressure compensation is needed after the electric regulating valves are adjusted by increasing the gas flow rate injection to maintain the preset back pressure. When the gas flow rate injection is increased after the electric regulating valves are adjusted, the gas flow rate collected by the flow sensor in the corresponding channel will show a significant increase compared to before the adjustment.

[0056] For the same leak point, if the first path is closer to the leak point than the second path, the first path will collect the first back pressure earlier than the second path will collect the second back pressure. Therefore, the first back pressure collected by the first path will be lower than the back pressure collected by the second path. Correspondingly, the first moment when the gas flow rate is increased after the electric regulating valve of the first path is adjusted will be earlier than the second moment when the gas flow rate is increased after the electric regulating valve of the second path is adjusted. That is, based on the back pressure collected by the path closer to the leak point, the main control unit module 306 will detect the leak in the electrostatic chuck 200 earlier and will first control the electric regulating valve of the closer path to increase the gas flow rate. Therefore, there is a time delay between the first and second moments.

[0057] The first moment is the initial data collection time of the pressure and flow sensors after the electric regulating valve in the first passage is adjusted. This initial data collection time is closer to the moment when the gas flow rate is increased after the electric regulating valve is adjusted, and closer to the gas flow rate required for the preset back pressure. Similarly, the second moment is the initial data collection time of the pressure and flow sensors after the electric regulating valve in the second passage is adjusted. Therefore, based on the first and second moments, which more accurately reflect the gas flow rate required for the preset back pressure, the leak location can be accurately pinpointed.

[0058] Therefore, based on the gas propagation speed in the vacuum chamber 400, the time delay between the first and second moments, the first back pressure, the second back pressure, and the distance between the gas outlet of the first passage and the gas outlet of the second passage, the location of the leak can be calculated and located.

[0059] Optionally, based on the gas propagation speed within the vacuum chamber, the first moment, the second moment, the first back pressure, the second back pressure, and the distance between the gas outlets of the first and second passages, the location of the leak is calculated and located, including: calculating the time difference between the second moment and the first moment; determining the direction of the leak relative to the first and second passages based on the time difference and the propagation speed; calculating the pressure gradient weight of the first back pressure and the pressure gradient weight of the second back pressure based on the first back pressure, the second back pressure, and the differences between the first and second back pressures and preset back pressures; wherein, the larger the pressure gradient weight, the closer the corresponding passage is to the leak; determining the area adjacent to both sides of the passage with the maximum pressure gradient weight corresponding to the electrostatic chuck under test as the area where the leak is located; and determining the specific location of the leak based on the direction of the leak and the area where the leak is located.

[0060] In this embodiment, the main control unit module 306 adopts a leak location mechanism. By analyzing the time difference of gas flow surge and the gradient difference of pressure change among multiple channels 302, and using the multi-point location principle of multiple symmetrically distributed channels, such as the triangular location principle of three symmetrically distributed channels, the spatial location of the leak can be calculated.

[0061] Specifically, the time difference between the gas flow rate that increases after the electric regulating valve is adjusted and collected successively in the first and second adjacent channels is first calculated, that is, the difference between the second moment that is later than the first moment and the first moment.

[0062] Then, combining the known propagation speed of gas in the vacuum chamber and the known distance between the gas outlets of the first and second passages, the direction θ of the leak point relative to the first and second passages is initially estimated.

[0063] θ=arccos (cHe * Δt / dsensor); Where cHe represents the propagation speed of gas in the vacuum chamber, Δt represents the time difference, and dsensor represents the distance between the gas outlet of the first passage and the gas outlet of the second passage.

[0064] After determining the direction of the leak, the corresponding pressure gradient weight is calculated by combining the pressure gradient changes of the first and second passages before and after the corresponding electric regulating valve is adjusted.

[0065] Based on the pressure field distribution law caused by gas leakage, that is, the pressure is inversely proportional to the square of the distance, the pressure gradient weights of the first and second paths are calculated.

[0066] The pressure gradient weights of the pathway are shown in the following equation:

[0067] in, Represents the pressure gradient weight of the i-th path, for Figure 1 In the example, i is 1 to 3, for example, representing the first pathway, the second pathway, and the third pathway respectively; P i This represents the back pressure collected before the electric regulating valve adjusts the i-th path; This represents the difference between the preset back pressure before and after adjustment by the electric regulating valve corresponding to the j-th channel; n represents the number of channels. Figure 1 For example, n=3.

[0068] Using the above formula, the pressure gradient weights corresponding to multiple pathways can be calculated separately. The larger the pressure gradient weight, the closer the corresponding pathway is to the leak point. This allows us to determine that the leak point is located in the area where the electrostatic chuck under test intersects with the pathway on both sides with the largest pressure gradient weight.

[0069] Finally, based on the initially estimated direction of the leak and the adjacent regions on both sides of the path with the maximum pressure gradient weight where the leak is located, the location of the leak on the electrostatic chuck under test is determined within the adjacent region on which side of the path it is located.

[0070] Therefore, the above-mentioned leak location mechanism can effectively locate the specific spatial position of the leak and has high positioning accuracy, meeting the precision detection requirements of electrostatic chuck. For example, the positioning accuracy is maintained within 10 mm (φ10 mm) in diameter, which significantly reduces the cost and time of leak troubleshooting.

[0071] Furthermore, the main control unit module is also used to: acquire the first gas flow rate collected by the corresponding flow sensor in the first channel before its electric regulating valve is adjusted, and the second gas flow rate collected at the first moment, and calculate the flow difference between the second gas flow rate and the first gas flow rate; acquire the third gas flow rate collected by the corresponding flow sensor in the second channel before its electric regulating valve is adjusted, and the fourth gas flow rate collected at the second moment, and calculate the flow difference between the fourth gas flow rate and the third gas flow rate; input the multi-dimensional feature data including time difference, pressure gradient weight and flow difference into a preset machine learning model, and output a distribution heatmap containing the leak location and corresponding leak probability of the electrostatic chuck under test through the machine learning model.

[0072] The flow difference or flow amplitude is the change in flow rate before and after the main control unit module 306 controls the corresponding electric regulating valve of each channel to adjust the flow rate of the injected gas in order to maintain the preset back pressure after a leak occurs in the electrostatic chuck under test.

[0073] In this embodiment, the main control unit module 306 inputs multi-dimensional feature data, including time difference, pressure gradient weight, and flow difference between adjacent channels, into a preset machine learning model, such as a Support Vector Machine (SVM), random forest, or neural network. The machine learning model then outputs a two-dimensional heatmap of the probability distribution of leaks. The two-dimensional data includes the location of the leak in the electrostatic chuck under test and the corresponding leak rate, which is based on the flow difference. This heatmap visually displays the possible locations of leaks and the corresponding leak rates.

[0074] In addition, this machine learning model can not only record the location and rate of leaks in a single test, but also generate a two-dimensional heat map of the probability distribution of leaks for all different voltage-back pressure combinations of the electrostatic chuck under test under all operating conditions. At the same time, it generates boundary curves for qualified and non-compliant leak rates, intuitively showing the performance of the electrostatic chuck under test within the safe operating window.

[0075] In one specific embodiment, optionally, the main control unit module is further configured to: before detecting the leak rate of the electrostatic chuck under test based on the gas flow rate collected from each channel, obtain different voltage-back pressure combinations required for leak rate detection of the electrostatic chuck under test based on the model and / or process requirements of the electrostatic chuck under test; input the different voltage-back pressure combinations into a preset three-dimensional dynamic correlation model of voltage-back pressure-leak rate, and output the leak rate prediction value corresponding to each voltage-back pressure combination through the three-dimensional dynamic correlation model, wherein the three-dimensional dynamic correlation model is trained based on historical detection data including voltage-back pressure combinations and corresponding leak rates; plan the configuration sequence of different voltage-back pressure combinations required for each leak rate detection of the electrostatic chuck under test according to the leak rate prediction values ​​from largest to smallest; and based on the planned configuration sequence, sequentially control each general pressure controller to configure the preset voltage and preset back pressure corresponding to the electrostatic chuck under test, and perform leak rate detection.

[0076] In this embodiment, based on Figure 1 The electrostatic chuck leak rate detection system with the hardware architecture shown combines the voltage-backpressure combination for leak rate detection of the electrostatic chuck under test. Before detecting the actual leak rate of the electrostatic chuck under test, the main control unit module 306 first calls a pre-trained three-dimensional dynamic correlation model corresponding to the voltage-backpressure-leak rate, based on the gas flow rate collected by the flow sensors in each channel. This three-dimensional dynamic correlation model is used to output the predicted leak rate value corresponding to the input voltage-backpressure combination.

[0077] The three-dimensional dynamic correlation model can be initially constructed based on advanced data fitting algorithms, such as multinomial regression and Gaussian process regression, and trained using historical data including voltage-backpressure combinations and corresponding leakage rates as sample data. This three-dimensional dynamic correlation model also has online self-updating capabilities, continuously absorbing new data samples as the number of detected electrostatic chucks increases. Through incremental learning algorithms, it optimizes its parameters, enabling the model's prediction accuracy and generalization ability to continuously improve over time, gradually accumulating into an electrostatic chuck performance knowledge base.

[0078] Before starting the actual leak rate detection of the electrostatic chuck under test, the main control unit module 306 calls the three-dimensional dynamic correlation model to quickly predict the leak rate for multiple different voltage-back pressure combinations required for leak rate detection, so as to predict the leak rate value corresponding to each voltage-back pressure combination. The larger the leak rate value, the more likely a leak point will occur under the corresponding voltage-back pressure combination.

[0079] Based on this prediction, the main control unit module 306 can dynamically plan the configuration sequence of the voltage-back pressure combination corresponding to the leakage rate of the electrostatic chuck under test for each detection. Priority is given to detecting the operating condition with the highest risk predicted by the three-dimensional dynamic correlation model, that is, the voltage-back pressure combination most likely to have a leakage point.

[0080] Specifically, based on the leak rate prediction values ​​in descending order, the sequence of voltage-back pressure combinations configured or switched for each leak rate detection is planned. This enables leak rate prediction under different operating conditions, significantly reducing the number of invalid or low-risk leak rate detection steps, improving the reliability of electrostatic chuck performance evaluation and screening, and increasing leak rate detection efficiency. Experimental verification shows that the leak rate detection optimization strategy driven by the prediction results of the three-dimensional dynamic correlation model can further reduce the single detection time of the electrostatic chuck under test by more than 10%, building upon the gas flow rate detection based on each channel.

[0081] In addition to being used for leak rate prediction before detection, the embodiments of this application can also upgrade the three-dimensional dynamic correlation model of voltage-back pressure-leakage rate from a static data analysis tool to a core decision engine that drives the entire electrostatic chuck leak rate detection system to achieve intelligent detection, predictive optimization and adaptive calibration. The role of the three-dimensional dynamic correlation model runs through the entire process before, during and after detection.

[0082] In one specific embodiment, optionally, the main control unit module is further configured to: before inputting different voltage-backpressure combinations into the preset three-dimensional dynamic correlation model of voltage-backpressure-leakage rate, obtain a three-dimensional dynamic correlation model adapted to the electrostatic chuck under test from a cross-model model library based on the model and / or process requirements of the electrostatic chuck under test, wherein the cross-model model library stores three-dimensional dynamic correlation models of electrostatic chucks with different models and / or process requirements; or before inputting different voltage-backpressure combinations into the preset three-dimensional dynamic correlation model of voltage-backpressure-leakage rate, use the voltage-backpressure corresponding to the electrostatic chuck under test as sample data, and fine-tune the model parameters of the target three-dimensional dynamic correlation model through a transfer learning algorithm so that the fine-tuned three-dimensional dynamic correlation model is adapted to the electrostatic chuck under test; wherein the model and / or process requirements of the electrostatic chuck adapted to the target three-dimensional dynamic correlation model are different from those of the electrostatic chuck under test.

[0083] In one of the above embodiments, the main control unit module, based on the model and / or process requirements of the electrostatic chuck under test, calls a 3D dynamic correlation model adapted to the current electrostatic chuck under test from a cross-model model library, and then uses it for corresponding leak rate prediction. The cross-model model library stores benchmark 3D dynamic correlation models applicable to electrostatic chucks of different models and / or process requirements. Knowing the model of the current electrostatic chuck under test, it can automatically match the benchmark 3D dynamic correlation model with the same or closest features in the cross-model model library for leak rate prediction before testing, achieving "out-of-the-box" usability.

[0084] In another embodiment, if a target 3D dynamic correlation model exists, but the model corresponds to an electrostatic chuck whose model type and / or process requirements differ from the current electrostatic chuck under test, the main control unit module can acquire a small amount of sample data from the electrostatic chuck under test, specifically the voltage-back pressure-leakage rate obtained from actual testing. It then uses a transfer learning algorithm to fine-tune the model parameters of the target 3D dynamic correlation model, ensuring the fine-tuned model is suitable for the current electrostatic chuck under test. This fine-tuned model is then used to predict the leakage rate of different voltage-back pressure combinations required by the current electrostatic chuck under test.

[0085] Thus, based on the existing three-dimensional dynamic correlation model, only a small amount of sample data under the new working conditions of the electrostatic chuck under test is needed to quickly fine-tune the parameters of the existing three-dimensional dynamic correlation model to adapt it to the new voltage-back pressure combination range or space of the electrostatic chuck under test, realize the transfer learning adaptation of the three-dimensional dynamic correlation model, and greatly shorten the development cycle of the detection scheme of the new process.

[0086] In the above embodiments, when faced with electrostatic chucks of different models or process requirements, which cause changes in the voltage-back pressure combination required for leak rate detection, it is not necessary to completely rebuild the three-dimensional dynamic correlation model. Through cross-model model library and / or transfer learning adaptation, the three-dimensional dynamic correlation model can achieve self-adaptation and maintain prediction accuracy under new operating conditions.

[0087] In one specific embodiment, the three-dimensional dynamic correlation model also outputs a prediction confidence level while outputting the leak rate prediction value. The main control unit module is also used to: when performing a target detection based on the gas flow rate collected from each channel, if the difference between the leak rate detection value and the leak rate prediction value of the electrostatic chuck under test is less than a preset threshold, and the prediction confidence level output by the three-dimensional dynamic correlation model is less than a preset confidence threshold, then increase the acquisition density of the pressure sensor and flow sensor corresponding to the target detection, and perform leak rate detection of the electrostatic chuck under test based on the increased acquisition density.

[0088] In this embodiment, the three-dimensional dynamic correlation model outputs the corresponding prediction confidence level while providing the leak rate prediction value. When detection is performed at the edge of a new operating condition of the electrostatic chuck under test based on the target voltage-back pressure combination, i.e., the difference between the leak rate detected by the flow sensor and the leak rate prediction value is less than a preset threshold, but the detected leak rate is close to the leak rate prediction value of normal sealing. If the prediction confidence level of the leak rate prediction value output by the three-dimensional dynamic correlation model based on the target voltage-back pressure combination is less than the preset confidence threshold, the prediction result may be inaccurate. At this time, based on the current detection of the target voltage-back pressure combination, the acquisition density of the corresponding pressure sensor and flow sensor can be increased, and more data can be collected by shortening the required sensor acquisition time interval. This realizes the automatic triggering of the accurate verification mode, which combines the leak rate detection results with the high acquisition density to verify the low confidence leak rate prediction value output by the three-dimensional dynamic correlation model to ensure the reliability of the final leak rate detection.

[0089] In one specific embodiment, the main control unit module is further configured to: cross-validate the leak rate prediction value corresponding to each voltage-back pressure combination output by the three-dimensional dynamic correlation model with the leak rate detection value detected based on the gas flow data collected from each channel; and when the deviation between the verification leak rate prediction value and the leak rate detection value exceeds the preset leak rate threshold, send an alarm to indicate that the electrostatic chuck under test has sensor drift or an unforeseen fault mode.

[0090] In this embodiment, during the process of obtaining the actual detection value by detecting the real-time leak rate using gas flow rate collected from each channel based on the target voltage-back pressure combination (i.e., any combination of different voltage-back pressure combinations required for leak rate detection of the electrostatic chuck under test), the leak rate prediction value corresponding to the target voltage-back pressure combination output by the three-dimensional dynamic correlation model is simultaneously combined to perform cross-validation between the model leak rate prediction value and the real-time detection value. If the deviation between the verified leak rate prediction value and the leak rate detection value exceeds a preset leak rate threshold, an alarm is immediately sent to indicate that the electrostatic chuck under test has sensor drift or an unforeseen fault mode. Thus, a closed-loop quality control system of self-validation and alarm is achieved for the leak rate detection process.

[0091] In this embodiment, the electrostatic chuck leak detection system includes: multiple channels connecting a gas source to the vacuum chamber of the electrostatic chuck under test, used to transmit gas supplied from the gas source to the vacuum chamber; multiple universal pressure controller modules located outside the vacuum chamber and connected to each of the multiple channels, each universal pressure controller module including a universal pressure controller, a pressure sensor, a flow sensor, and an electric regulating valve; wherein, the pressure sensor is used to collect the back pressure in the vacuum chamber corresponding to the connected channel, the flow sensor is used to collect the gas flow rate in the vacuum chamber corresponding to the connected channel, and the electric regulating valve is connected to the gas source and the gas inlet of the connected channel, used to regulate the gas flow rate transmitted to the vacuum chamber by each channel; a main control unit module, connected to multiple... Each universal pressure controller module is interconnected and used for: configuring the preset voltage and preset back pressure corresponding to the electrostatic chuck under test based on instructions carrying different voltage-back pressure combinations required for each leak rate test of the electrostatic chuck; inputting the configured preset voltage and preset back pressure into a preset three-dimensional dynamic correlation model of voltage-back pressure-leakage rate, so as to output the corresponding leak rate prediction value through the three-dimensional dynamic correlation model; when a leak is determined to exist in the electrostatic chuck under test based on the leak rate prediction value and the leak rate detection value detected by the gas flow rate collected through each channel, the universal pressure controller controls the corresponding electric regulating valve to adjust the gas flow rate transmitted to the vacuum chamber based on the back pressure collected through each channel, so as to form the preset back pressure in the vacuum chamber. Therefore, by embedding the three-dimensional dynamic correlation model into the detection system, when a leak occurs in the electrostatic chuck, the detection system can directly determine that the electrostatic chuck is leaking based on the data predicted by the three-dimensional dynamic correlation model and the current actual detection data, and perform rapid gas compensation to adjust the pressure, so that the actual back pressure formed in the vacuum chamber of the electrostatic chuck accurately meets the preset back pressure requirements. It enables automatic distribution, configuration, and free switching of detection parameters, as well as closed-loop automatic adjustment of back pressure. This significantly shortens the time required for a single test, achieves dynamic, continuous, and high-precision detection of leak rates, better meets the urgent needs of the semiconductor industry for efficient and precise detection in the mass production of electrostatic chucks, greatly improves detection efficiency and result reliability, and expands the industrial application scope of this technology.

[0092] Optionally, such as Figure 3 As shown, this application embodiment also provides an electronic device 2000, including a processor 2400 and a memory 2200. The memory 2200 stores a program or instructions that can run on the processor 2400. When the program or instructions are executed by the processor 2400, they implement the various steps of the main control unit module of the above embodiment and can achieve the same technical effect. To avoid repetition, they will not be described again here.

[0093] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the main control unit module in any of the above embodiments and achieve the same technical effects. To avoid repetition, further details are omitted here. The readable storage medium includes computer-readable storage media, such as read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0094] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program. The computer program is operable to enable a computer to execute various processes of the main control unit module in any of the above embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0095] It should be noted that, in this document, 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. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0096] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0097] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. An electrostatic chuck leak rate detection system, characterized in that, include: Multiple pathways connect the gas source to the vacuum chamber of the electrostatic chuck under test, for transmitting gas supplied from the gas source to the vacuum chamber; Multiple universal pressure controller modules are disposed outside the vacuum chamber and connected to the multiple channels one by one. Each universal pressure controller module includes a universal pressure controller, a pressure sensor, a flow sensor, and an electric regulating valve. The pressure sensor is used to collect the back pressure in the vacuum chamber corresponding to the channel it is connected to. The flow sensor is used to collect the gas flow rate in the vacuum chamber corresponding to the channel it is connected to. The electric regulating valve is connected to the gas source and the gas inlet of the channel it is connected to, and is used to regulate the gas flow rate transmitted to the vacuum chamber by each channel. The main control unit module, which is communicatively connected to the plurality of general-purpose pressure controller modules, is used to: control each of the general-purpose pressure controllers to configure the preset voltage and preset back pressure corresponding to the electrostatic chuck under test based on instructions carrying different voltage-back pressure combinations required for each detection of the leak rate of the electrostatic chuck under test; input the configured preset voltage and preset back pressure into a preset three-dimensional dynamic correlation model of voltage-back pressure-leakage rate, so as to output the corresponding leak rate prediction value through the three-dimensional dynamic correlation model; when it is determined that there is a leak in the electrostatic chuck under test based on the leak rate prediction value and the leak rate detection value detected by the gas flow rate collected through each channel, control the corresponding electric regulating valve through each general-purpose pressure controller to adjust the gas flow rate transmitted to the vacuum chamber based on the back pressure collected by each channel, so as to form the preset back pressure in the vacuum chamber.

2. The system according to claim 1, characterized in that, The main control unit module is specifically used for: When the predicted leak rate value differs from the leak rate detection value of the gas flow rate detection corresponding to the target path, the electric regulating valve corresponding to the target path is controlled to adjust the gas flow rate transmitted to the vacuum chamber, so that the preset back pressure is formed in the vacuum chamber of the area corresponding to the target path.

3. The system according to claim 2, characterized in that, The gas outlets of the multiple passages are symmetrically distributed relative to the vacuum chamber. The main control unit module is further used for: Based on the propagation speed of the gas in the vacuum chamber, as well as the back pressure and gas flow data collected before and after the adjustment of the electric regulating valve in the multiple channels, the leak point is located. The electric regulating valve corresponding to the passage closest to the leak point among the multiple passages is controlled to regulate the gas flow rate transmitted to the vacuum chamber, so that the preset back pressure is formed in the vacuum chamber corresponding to the area of ​​the closest passage.

4. The system according to claim 3, characterized in that, The main control unit module locates the leak, including: The pressure sensor corresponding to the first path in the plurality of paths acquires the first back pressure before its electric regulating valve is adjusted. The first back pressure is less than the preset back pressure, and the difference between the first back pressure and the preset back pressure exceeds the preset back pressure threshold. The pressure sensor corresponding to the second path adjacent to the first path in the plurality of paths acquires the second back pressure before its electric regulating valve is adjusted. The second back pressure is less than the preset back pressure, and the difference between the second and the preset back pressure exceeds the preset back pressure threshold. The first moment is determined as the moment when the pressure sensor and flow sensor first collect data after the electric regulating valve corresponding to the first passage is adjusted. The second moment is determined as the first data collection moment of the pressure sensor and flow sensor after the electric regulating valve corresponding to the second passage is adjusted, wherein the second moment is later than the first moment; The location of the leak is calculated and located based on the gas propagation speed in the vacuum chamber, the first moment, the second moment, the first back pressure, the second back pressure, and the distance between the gas outlet of the first passage and the gas outlet of the second passage.

5. The system according to claim 4, characterized in that, Based on the gas propagation speed within the vacuum chamber, the first moment, the second moment, the first back pressure, the second back pressure, and the distance between the gas outlet of the first passage and the gas outlet of the second passage, the location of the leak is calculated and located, including: Calculate the time difference between the second time point and the first time point; Based on the time difference and the propagation speed, the direction of the leak point relative to the first path and the second path is determined; Based on the first back pressure, the second back pressure, and the differences between the first back pressure, the second back pressure, and the preset back pressure, the pressure gradient weight of the first back pressure and the pressure gradient weight of the second back pressure are calculated; wherein, the larger the pressure gradient weight, the closer the corresponding path is to the leak point; The regions adjacent to both sides of the path corresponding to the maximum pressure gradient weight of the electrostatic chuck under test are determined as the regions where the leak point is located. The specific location of the leak is determined based on the direction of the leak and the area where the leak is located.

6. The system according to claim 5, characterized in that, The main control unit module is also used for: The first gas flow rate collected by the corresponding flow sensor before its electric regulating valve is adjusted in the first channel, and the second gas flow rate collected at the first moment are obtained, and the flow rate difference between the second gas flow rate and the first gas flow rate is calculated. The third gas flow rate collected by the corresponding flow sensor before its electric regulating valve is adjusted in the second channel, and the fourth gas flow rate collected at the second moment are obtained, and the flow rate difference between the fourth gas flow rate and the third gas flow rate is calculated. The multidimensional feature data, including the time difference, the pressure gradient weight, and the flow rate difference, is input into a preset machine learning model, and the machine learning model outputs a distribution heatmap containing the leak location and corresponding leak probability of the electrostatic chuck under test.

7. The system according to any one of claims 3-5, characterized in that, The main control unit module is also used for: Before detecting the leak rate of the electrostatic chuck under test based on the gas flow rate collected from each channel, different voltage-back pressure combinations required for leak rate detection of the electrostatic chuck under test are obtained based on the model and / or process requirements of the electrostatic chuck under test. Different voltage-back pressure combinations are input into the preset three-dimensional dynamic correlation model of voltage-back pressure-leakage rate. The three-dimensional dynamic correlation model outputs the leakage rate prediction value corresponding to each voltage-back pressure combination. The three-dimensional dynamic correlation model is trained based on historical detection data including voltage-back pressure combinations and corresponding leakage rates. Based on the leak rate prediction values ​​in descending order, plan the configuration sequence of different voltage-back pressure combinations required for each detection of the leak rate of the electrostatic chuck under test; Based on the planned configuration sequence, each general pressure controller is sequentially controlled to configure the preset voltage and preset back pressure corresponding to the electrostatic chuck under test, and leakage rate detection is performed.

8. The system according to claim 7, characterized in that, The main control unit module is also used for: Before inputting different voltage-back pressure combinations into the preset three-dimensional dynamic correlation model of voltage-back pressure-leakage rate, based on the model and / or process requirements of the electrostatic chuck under test, a three-dimensional dynamic correlation model adapted to the electrostatic chuck under test is obtained from a cross-model model library, wherein the cross-model model library stores three-dimensional dynamic correlation models of electrostatic chucks with different models and / or process requirements; or Before inputting different voltage-back pressure combinations into the preset three-dimensional dynamic correlation model of voltage-back pressure-leakage rate, the voltage-back pressure corresponding to the electrostatic chuck under test is used as sample data, and the model parameters of the target three-dimensional dynamic correlation model are fine-tuned through a transfer learning algorithm so that the fine-tuned three-dimensional dynamic correlation model is adapted to the electrostatic chuck under test; wherein, the model and / or process requirements of the electrostatic chuck adapted to by the target three-dimensional dynamic correlation model are different from those of the electrostatic chuck under test.

9. The system according to claim 7, characterized in that, The three-dimensional dynamic correlation model also outputs the prediction confidence level along with the leak rate prediction value. The main control unit module is also used for: When performing a target detection based on the gas flow rate collected from each channel, if the difference between the detected leak rate value and the predicted leak rate value of the electrostatic chuck under test is less than a preset threshold, and the prediction confidence level output by the three-dimensional dynamic correlation model is less than a preset confidence threshold, then the acquisition density of the pressure sensor and flow sensor corresponding to the target detection is increased, and the leak rate detection of the electrostatic chuck under test is performed based on the increased acquisition density.

10. The system according to claim 7, characterized in that, The main control unit module is also used for: The predicted leakage rate values ​​corresponding to each voltage-back pressure combination output by the three-dimensional dynamic correlation model are cross-validated with the leakage rate detection values ​​detected based on the gas flow data collected from each channel. When the deviation between the predicted leak rate and the detected leak rate exceeds a preset leak rate threshold, an alarm is sent to indicate that the electrostatic chuck under test has sensor drift or an unforeseen fault mode.

Images