End point detection device and dynamic adjustment method for preventing free jet, entrainment

Abstract

The application relates to the technical field of semiconductor detection, and discloses an end-point detection device capable of preventing free jet flow and entrainment and a dynamic adjustment method, wherein the detection device comprises a light source module, a gas chamber module, a variable-diameter module, a segmented temperature control module and a detector module; the gas chamber module comprises coaxially nested first, second and third cylinder bodies, a first chamber is formed between the first and second cylinder bodies, the inner cavity of the third cylinder body is a second chamber, a third chamber is formed between the third and second cylinder bodies, the inner wall of the first cylinder body and the outer wall of the second cylinder body are provided with wall-attached flow guide structures, and the third cylinder body is densely provided with exhaust holes; the variable-diameter module is used for adjusting the flow area of an exhaust pipe; and the segmented temperature control module is used for forming a temperature gradient which decreases along the gas flow direction. Through the cooperation of the anti-entrainment gas chamber structure, the variable-diameter module and the segmented temperature control module, the jet flow entrainment effect can be inhibited, and the accuracy of gas concentration measurement and the reliability of clean process monitoring can be improved.

Description

Technical Field

[0001] This invention relates to the field of semiconductor testing technology, specifically to an endpoint detection device and dynamic adjustment method that prevents free jetting and entrainment. Background Technology

[0002] NDIR (non-dispersive infrared) gas sensors are widely used in monitoring the cleaning process of semiconductor thin-film chambers. They determine the degree of cleanliness by detecting changes in gas concentration within the chamber. During operation, the gas to be measured enters the sensor's detection chamber through the inlet, is illuminated by an infrared light source, and the signal is received by the detector module, thereby enabling the measurement of gas concentration.

[0003] In a typical NDIR sensor structure, the inlet aperture is relatively small, while the detection chamber volume is relatively large. When gas is injected into the detection chamber at high speed from the inlet at a certain flow rate, a free jet is formed. According to fluid dynamics principles, the boundary layer of this jet generates a strong velocity gradient, which entrains surrounding gas into the jet body through viscous shear force. Due to the dynamic development of the cleaning process, this is equivalent to instantaneously disrupting the stable flow range of the gas being measured, causing the concentration measured by the NDIR sensor to differ from the actual concentration generated by the process in the chamber, thus failing to reflect the actual cleaning progress of the chamber.

[0004] Simulation models are typically based on ideal flow and temperature fields. However, in practical applications, improper shape, location, and temperature control of the outlet can lead to gas diffusion, condensation, or uneven mixing with the background gas. For example, firstly, when different types of background gases are present (such as purging nitrogen or residual gas from the previous cycle), entrainment can cause direct dilution and cross-contamination of the sample gas, or cross-contamination can occur between different process cycles, affecting the accuracy of the next detection cycle. Secondly, even if the surrounding gas is the same type, self-entrainment of the jet boundary layer can reduce the concentration in the jet core region and expand the jet cross-section, resulting in uneven concentration distribution in the optical path. This causes the local concentration detected by the sensor to deviate from the average concentration throughout the chamber, and the reading cannot represent the true concentration. All of these situations can cause a significant deviation between the actual flow field and the simulated flow field. This deviation directly leads to discrepancies in time and amplitude between the simulated predicted "theoretical response curve" and the actual detected "measured response curve."

[0005] Furthermore, etching gases (such as NF3, CF4, C2, F6, etc.) or their byproducts (such as SiF4, HF) used in the cleaning process may condense or adsorb onto the pipe walls when flowing through pipes and outlets at lower temperatures. This not only causes detection signal attenuation and response delay, but also contaminates the equipment, leading to baseline drift and decreased long-term stability.

[0006] The aforementioned entrainment effect is one of the fundamental physical reasons for the systematic deviation between the simulated and actual NDIR detection curves. Because the local gas concentration detected by the sensor is not the average concentration throughout the detection chamber, but rather the gas concentration in a specific area affected by the jet entrainment effect, the measurement results cannot accurately reflect the actual gas concentration generated within the process chamber. This deviation makes it difficult for the control system to accurately report gas concentration and determine cleaning progress, thus affecting the accuracy and reliability of the cleaning endpoint. Summary of the Invention

[0007] To address the problems in the prior art where uneven airflow distribution and local concentration deviation from the average concentration in the detection chamber caused by free jetting and entrainment of gas, thus affecting the accuracy of cleaning progress monitoring, this invention provides an endpoint detection device and dynamic adjustment method to prevent free jetting and entrainment. This device can suppress the jet entrainment effect generated during gas injection, making the airflow distribution in the detection chamber more uniform and stable, thereby improving the accuracy of gas concentration measurement and the reliability of cleaning process monitoring.

[0008] To address the aforementioned problems, this invention provides an endpoint detection device to prevent free jetting and entrainment, used for detecting the endpoint of the cleaning process in thin film growth equipment. The device includes a light source module, a gas chamber module, a variable diameter module, a segmented temperature control module, and a detector module. The gas chamber module comprises a first cylinder, a second cylinder, and a third cylinder. The first cylinder has an air inlet on its side wall and a first wall-mounted flow guide structure on its inner wall. An air inlet pipe is connected to the air inlet. The second cylinder is disposed within the first cylinder and coaxial with it. A second wall-mounted flow guide structure is provided on the outer wall of the second cylinder. A first opening and an exhaust port are provided on the side wall of the second cylinder. An exhaust pipe is connected to the exhaust port, with one end of the exhaust pipe extending outwards from the exhaust port to the outside of the first cylinder. The third cylinder is disposed within the second cylinder and... Coaxial with the second cylinder, the third cylinder has a second opening and multiple exhaust holes on its wall, with a channel between the second opening and the first opening; the second cylinder and the first cylinder together form a first chamber, which is used to prevent free jets and form a wall-attached circulation; the inner cavity of the third cylinder is a second chamber, which is connected to the first chamber through the channel to form a stable detection area; the third cylinder and the second cylinder together form a third chamber, which is connected to the second chamber through the exhaust holes for uniform multi-point exhaust.

[0009] Furthermore, the variable diameter module is used to adjust the flow area of ​​the exhaust pipe. The segmented temperature control module includes an integrated partitioned heating jacket, a first PID controller, a second PID controller, and a linkage control module. The integrated partitioned heating jacket is fitted onto the outside of the gas chamber module and is divided into a first temperature zone and a second temperature zone. The first temperature zone covers the area where the intake pipe and the first cylinder are located and is independently controlled by the first PID controller. The second temperature zone covers the exhaust pipe and is independently controlled by the second PID controller. The linkage control module is connected to the first PID controller and the second PID controller respectively and is configured to coordinate the heating power of the first temperature zone and the second temperature zone based on the temperature information of the first temperature zone and the second temperature zone to maintain a preset target temperature difference, so that the temperature of the intake pipe, the first cylinder, the second cylinder, the third cylinder, and the exhaust pipe are all higher than the condensation point of the detected process gas, and the temperature of the intake pipe is the same as the temperature of the first cylinder, the second cylinder, and the third cylinder, and higher than the temperature of the exhaust pipe, so as to form a temperature gradient that decreases along the gas flow direction. The detector module includes a first detector and a second detector. The first detector is used for real-time concentration signal detection, and the second detector is used for light attenuation monitoring of the light source module.

[0010] The endpoint detection device for preventing free jet and entrainment provided by this invention allows the gas to be tested to enter the first chamber through an inlet on the side wall of the first cylinder. After entering the first chamber, the gas first changes from a beam jet to a ring flow that diffuses uniformly in the circumferential direction, initially eliminating the directional momentum of the jet. At the same time, the first wall-adhering flow guiding structure on the inner wall of the first cylinder and the second wall-adhering flow guiding structure on the outer wall of the second cylinder cooperate to guide the gas to adhere to the wall surface, so that the gas maintains a stable wall-adhering state while diffusing in the circumferential direction, and generates local micro-disturbances in the near-wall region to promote uniform gas mixing and create conditions for subsequent flow uniformity.

[0011] Subsequently, the gas enters the second chamber sequentially through the first inlet, the channel, and the second inlet. Once inside the second chamber, the gas no longer forms a jet or flows violently; instead, driven by the pressure difference, it seeps into the third chamber through multiple vent holes in the wall of the third cylinder. Because the vent holes are evenly distributed along the wall of the third cylinder, the gas can seep out synchronously and uniformly, avoiding localized high-speed jets or reverse disturbances in the second chamber, thus helping to maintain the uniformity and stability of the gas within the chamber.

[0012] Meanwhile, the segmented temperature control module maintains the first temperature zone, where the intake pipe and the first cylinder are located, at the first target temperature. It also adjusts the heating temperature of the exhaust pipe in the second temperature zone in real time based on temperature changes in the first zone, maintaining the temperature difference between the exhaust pipe, intake pipe, and first cylinder at a set value, thus creating a decreasing temperature gradient along the gas flow direction. Specifically, the intake pipe and the first cylinder have the same and highest temperature, rapidly preheating the low-temperature gas to above its condensation point, preventing condensation before entering the first chamber. The second and third cylinders have the same temperature as the first cylinder, ensuring the gas remains gaseous within the gas chamber module, preventing blockage of the exhaust port. The exhaust pipe has the lowest temperature but still above its condensation point, preventing condensation and accumulation of gas during exhaust. This temperature gradient ensures a smooth transition in gas temperature, avoiding thermally induced flow disturbances caused by sudden temperature changes. Furthermore, according to stratified fluid dynamics theory, when fluid moves from a low-density region to a high-density region (density increases along the flow direction), stable density stratification is formed. Under these conditions, turbulent vortices need to overcome buoyancy to achieve vertical mixing, resulting in a significant decrease in mixing efficiency. This manifests as a reduction in the entrainment rate of the jet boundary layer, which can be mitigated by stable stratification. In this invention, the inlet pipe and the first cylinder have the same and highest temperatures, the second and third cylinders have the same temperatures as the first cylinder, and the exhaust pipe has the lowest temperature. This creates a stable stratification condition with increasing density along the gas flow direction, fundamentally suppressing the entrainment of surrounding gas by the jet from a hydrodynamic perspective. Simultaneously, the stable temperature inside the gas chamber module ensures the consistency of the infrared absorption coefficient, further improving the repeatability and accuracy of concentration measurements.

[0013] The variable diameter module dynamically adjusts the flow area of ​​the exhaust pipe according to changes in gas concentration. When the gas concentration is high, the flow area is increased to accelerate the exhaust speed, improve the response speed of the detection device, and promptly track the concentration decrease trend. When the gas concentration is low, the flow area is reduced to decrease the gas flow rate, prolong the residence time of the gas in the second chamber, enhance the infrared absorption signal, and improve the signal-to-noise ratio and low-concentration measurement accuracy. Through variable diameter adjustment, the detection device can achieve optimal measurement performance at different concentration stages.

[0014] Because the wall-mounted flow guiding structure eliminates the directional momentum of the jet, the uniformly distributed exhaust holes achieve synchronous and uniform gas permeation, the segmented temperature control module maintains a stable temperature gradient and forms density stratification to suppress entrainment, and the variable diameter adjustment optimizes the measurement conditions at different concentration stages, the gas in the second chamber has almost no directional flow. The entire inner cavity is free of high-speed jets, entrainment vortices, and concentration gradients, forming a highly uniform and stable gas environment. Finally, the gas in the third chamber is discharged outside the device through the exhaust port and exhaust pipe.

[0015] Simultaneously with the gas flow, the light source module emits infrared light into the second chamber. This infrared light passes through the uniform and stable gas in the second chamber and is received by the first detector. The first detector outputs a corresponding gas concentration detection signal based on the degree of infrared light absorption. The second detector periodically monitors the light attenuation of the light source module to ensure that the light intensity is within a reasonable attenuation range (e.g., setting 95%~100% of the initial signal as the normal range) to guarantee high-accuracy detection.

[0016] The endpoint detection device for preventing free jet and entrainment provided by this invention transforms the beam jet into a wall-mounted circulating flow through the wall-mounted guiding structure of the first chamber. Combined with the densely distributed exhaust holes on the third cylinder, this achieves uniform seepage, fundamentally eliminating high-speed jets and their entrainment vortices. This creates a highly uniform and stable gas region in the second chamber with no directional flow and no concentration gradient, significantly reducing measurement errors. Furthermore, a segmented temperature control module maintains the first temperature zone at a target temperature and adjusts the heating temperature of the second temperature zone in real time based on temperature changes in the first zone. This maintains the set temperature difference between the exhaust pipe, the inlet pipe, and the first cylinder, forming a decreasing temperature gradient along the gas flow direction. This not only prevents gas condensation and blockage of the exhaust holes during transmission but also avoids thermal flow disturbances caused by sudden temperature changes. Simultaneously, according to stratified fluid dynamics theory, the stable density stratification formed by this temperature gradient suppresses the entrainment of surrounding gas by the jet. The stable temperature of the first zone further ensures the consistency of the infrared absorption coefficient, thereby improving the repeatability and accuracy of concentration measurements. Meanwhile, the variable diameter module dynamically adjusts the exhaust pipe flow area according to changes in gas concentration. At high concentrations, the area is increased to accelerate exhaust and improve response speed; at low concentrations, the area is reduced to extend gas residence time and improve signal-to-noise ratio, achieving optimal measurement performance across the entire concentration range. The synergistic effect of wall-mounted flow guidance, uniform exhaust, gradient temperature control, and variable diameter adjustment enables the detection device to accurately reflect the actual gas concentration changes within the process chamber, reliably capture the inflection point at the cleaning endpoint, and avoid incomplete or excessive cleaning. Furthermore, the detector module performs concentration detection and light source attenuation monitoring functions, promptly detecting light source aging or malfunctions and ensuring the long-term measurement reliability of the device.

[0017] Since the gas, after being decelerated in the first chamber, may still experience velocity fluctuations and pressure shocks, directly entering the inner cavity would interfere with the flow uniformity and stability of the exhaust port. Therefore, preferably, the inner wall of the channel is arc-shaped, the cross-sectional area in the middle of the channel is smaller than the cross-sectional areas at both ends, and the cross-sectional area of ​​the channel continuously decreases from both ends to the middle.

[0018] Understandably, if a conventional straight-line fixed-diameter channel is used, it can only serve a guiding function and cannot suppress residual velocity pulsations and pressure shocks in the upstream airflow. The disturbances will be directly transmitted to the second chamber, interfering with the uniform flow stability of the exhaust port. A stepped variable-section channel would introduce new vortices. Therefore, an arc-shaped variable-section channel is chosen, allowing the channel's flow area to first smoothly contract and then expand. This enables the airflow to achieve a smooth conversion between kinetic and pressure potential energy during the continuous change of its cross-sectional area. In this process, the velocity pulsations and pressure shocks of the airflow are gradually smoothed out by the flow channel, and the fluctuating energy is converted into internal energy dissipation. Ultimately, this allows the gas to enter the second chamber more smoothly, forming a highly uniform and stable gas environment within the second chamber.

[0019] Preferably, the air inlet and the exhaust outlet are both located on the first side of the central axis of the first cylinder and are located in the same radial position, and the channel is located on the second side of the central axis of the first cylinder. The first side and the second side are opposite sides of the first cylinder in the radial direction.

[0020] This layout requires gas to travel circumferentially through the inlet on the first side to the channel on the second side before entering the second chamber after entering the first chamber. This extends the circumferential flow path and residence time of the gas in the first chamber, promoting thorough diffusion and mixing, and also makes the circumferential distribution of airflow within the third cylinder more uniform. Ultimately, this results in a more uniform and stable gas flow within the third cylinder.

[0021] Preferably, the first wall-mounted flow guiding structure includes a plurality of first arc-shaped protrusions, which are evenly distributed on the inner circumferential wall of the first cylinder; the second wall-mounted flow guiding structure includes a plurality of second arc-shaped protrusions, which are evenly distributed on the outer circumferential wall of the second cylinder.

[0022] As the gas flows circumferentially and axially within the first chamber, it sequentially passes through a first arc-shaped protrusion uniformly distributed on the inner circumferential wall of the first cylinder and a second arc-shaped protrusion uniformly distributed on the outer circumferential wall of the second cylinder. When the gas flows past the arc-shaped protrusions, the Coanda effect occurs on the arc-shaped surface, allowing the gas to naturally adhere to the surface and flow without separation. Simultaneously, due to local variations in the flow channel cross-section, the gas is accelerated by compression on the windward side of the protrusion and decelerated by streamline expansion on the leeward side. This velocity difference forms tiny local vortices on both sides of the protrusion. These micro-vortices are small in scale and low in energy, and will not develop into large-scale turbulence or disrupt the overall wall-attached flow, but are sufficient to promote gas exchange between adjacent flow layers, achieving gentle mixing in the near-wall region. The uniform distribution of multiple protrusions circumferentially and axially ensures that the gas is continuously subjected to this local micro-disturbance throughout the flow path, maintaining its wall-attached state and gradually achieving homogenization of concentration and temperature, resulting in a more uniform state of the gas before entering the channel.

[0023] Preferably, the first arc-shaped protrusion and the second arc-shaped protrusion are alternately distributed in the circumferential and axial directions of the first chamber.

[0024] As the gas flows along the wall in both the circumferential and axial directions, it passes through the first and second arc-shaped protrusions. Because these protrusions are staggered both circumferentially and axially, the gas maintains a continuous, free-mixing flow, gradually transitioning from near the outer wall to near the inner wall and then back, forming multiple back-and-forth flow segments. This radial migration allows for sufficient exchange of gas between the inner and outer layers of the annular channel without disrupting the overall wall-hugging flow or generating large-scale eddies, resulting in a more uniform gas concentration. The staggered distribution, combined with the Coanda effect of the arc-shaped protrusions, ensures that the gas maintains both wall-hugging flow and continuous, free-mixing flow and radial exchange, avoiding measurement deviations caused by uneven concentrations between the inner and outer layers, and further improving the reliability of determining the cleanliness endpoint.

[0025] Preferably, the vent is a frustum-shaped hole, and the diameter of the end of the vent facing the interior of the third cylinder is smaller than that of the other end.

[0026] Because the exhaust port adopts a gradually expanding structure with a small inlet and a large outlet, when the gas seeps out of the second chamber through the exhaust port, as the flow area gradually increases, the gas velocity and kinetic energy gradually decrease. This allows the gas to enter the third chamber more gently, avoiding reverse disturbance in the second chamber and improving the uniformity and stability of the gas in the second chamber.

[0027] Preferably, the ratio of the central diameter of the exhaust hole to the depth of the exhaust hole is in the range of 0.3 to 0.5, and the outward guide angle of the exhaust hole is in the range of 15° to 30°.

[0028] When the ratio of the central diameter to the depth of the vent is less than 0.3, the vent is too long and narrow, resulting in high flow resistance and a sharp increase in pressure drop as gas flows through it. This leads to a decrease in gas percolation velocity and a prolonged response time of the detection device to changes in the concentration of the process chamber, making it unable to reflect the cleaning progress in real time. Simultaneously, the long and narrow channel is easily blocked by tiny particles carried in the gas. When this ratio is greater than 0.5, the vent is too short and thick, resulting in insufficient gas deceleration and diffusion, and the percolation velocity remains high, failing to effectively suppress reverse disturbances to the second chamber. When the outward guide angle of the vent is less than 15°, the gradual diffusion effect is not significant, and the reduction in gas velocity is limited. When the outward guide angle is greater than 30°, the orifice wall expands too drastically, and the airflow is prone to separation, which can generate new disturbances. An angle range of 15° to 30° maintains a pressure gradient within the orifice, guiding the airflow to diffuse naturally without separation. The coupling of these two parameters achieves a better balance between low-disturbance percolation, appropriate pressure drop, and fast response, ensuring the long-term stable and reliable operation of the detection device.

[0029] Preferably, the integrated partitioned heating jacket includes a first heating jacket section and a second heating jacket section. The first heating jacket section is sleeved on the outside of the air intake pipe and the first cylinder. The first heating jacket section is provided with a first heating component and a first temperature measuring component. The first heating component and the first temperature measuring component are respectively connected to the first PID controller. The second heating jacket section is sleeved on the outside of the exhaust pipe. The second heating jacket section is provided with a second heating component and a second temperature measuring component. The second heating component and the second temperature measuring component are respectively connected to the second PID controller.

[0030] The integrated zone heating jacket is divided into a first heating section and a second heating section, which are respectively fitted onto the outer side of the air inlet pipe and the first cylinder, and the outer side of the exhaust pipe. This ensures that the first and second temperature zones are physically independent, avoiding thermal interference between them and facilitating precise temperature control. Each heating section integrates independent heating and temperature sensing components, which can be connected to the corresponding PID controller signals. Based on the real-time temperature feedback signals from each temperature sensing component, the PID controller uses a proportional unit to accelerate the response speed, an integral unit to eliminate steady-state errors, and a derivative unit to suppress temperature fluctuations, independently adjusting the heating power of each heating component. This allows the temperature in each zone to quickly converge and stabilize at its preset target temperature. This achieves real-time feedback adjustment of the heating power, improving the response speed and stability of temperature control.

[0031] Optionally, the linkage control module is configured to coordinate the first PID controller and the second PID controller in a master-slave follow mode. In the master-slave follow mode, the second PID controller dynamically adjusts the heating power of the second heating component according to the measured temperature fed back by the first temperature measuring component, so that the temperature fed back by the second temperature measuring component is 10°C to 30°C lower than the temperature fed back by the first temperature measuring component.

[0032] A fixed temperature difference of 10℃ to 30℃ ensures a stable temperature gradient along the gas flow direction, causing the gas density to increase along the flow direction. This fundamentally suppresses the entrainment of surrounding gas by the jet, improving the uniformity and stability of the gas in the second chamber. Simultaneously, dynamic tracking control can respond in real-time to temperature fluctuations in the first temperature zone, adjusting the heating power of the second temperature zone to maintain the set temperature difference and ensuring that the temperature gradient always meets process requirements.

[0033] Preferably, the system further includes an angle adjustment module. The light source module is connected to the power output terminal of the angle adjustment module. The angle adjustment module is used to adjust the optical axis direction of the light source module to minimize the eccentricity error of the optical path and the flow field. It can be understood that the optical axis is the center line of the main direction of infrared light emission, that is, the central axis of the infrared beam emitted by the light source module.

[0034] Before using the detection device, the optical path can be calibrated using an angle adjustment module. First, pure nitrogen gas is introduced into the detection device. Nitrogen does not absorb infrared light. Then, the angle adjustment module is driven to change the angle of the light source module, while simultaneously detecting the infrared light intensity received by the first detector, obtaining a curve showing the relationship between light intensity and angle. When the light source module and the first detector are in optimal coaxial alignment, the infrared light propagates along the axis of the second chamber, and most of the light energy directly reaches the first detector, resulting in maximum received light intensity. When the angle of the light source module deviates, some light is absorbed or scattered by the inner wall of the cylinder, reducing the received light intensity. The greater the deviation angle, the more severe the light intensity loss. Since pure nitrogen does not absorb infrared light, the change in light intensity is entirely determined by the degree of optical path alignment. Therefore, the angle of the light source module corresponding to the maximum light intensity on the curve is the optimal optical axis alignment position. Adjusting the light source module to this optimal position and locking it eliminates the eccentricity error between the center of the optical path and the center of the flow field in the second chamber caused by assembly tolerances or structural deformation, ensuring that the infrared light always passes through the area corresponding to the optimal detection angle position, thereby improving the accuracy of concentration measurement.

[0035] Optionally, the angle adjustment module can adopt various structural forms to adjust the optical axis direction of the light source module, such as: a universal ball joint structure, where the light source module is mounted on the universal ball joint to achieve tilting and locking in any direction; a dual-axis tilting stage, where the tilt angles of the X and Y axes are independently controlled by two orthogonal adjustment screws; a micro motor combined with an eccentric cam or lead screw, where the light source module is connected to the eccentric cam or lead screw, and the motor drives the eccentric cam or lead screw to actively adjust the angle of the light source module; an electromagnetic drive structure, where the current of the electromagnetic coil is controlled to generate magnetic force to tilt the base of the light source module, achieving non-contact adjustment; and a turntable structure, where the light source module is mounted on a rotatable turntable, and the rotation of the turntable changes the optical axis direction of the light source module, which, in conjunction with the tilt adjustment mechanism, enables spatial angle adjustment. All of the above solutions are existing technologies and are clear to those skilled in the art, therefore they will not be described in detail.

[0036] Besides the above solutions, any structure that can achieve the angle adjustment effect of the light source module can be used as an angle adjustment module.

[0037] Preferably, the exhaust pipe has a constriction section and an expansion section. The constriction section is disposed between the inlet end of the exhaust pipe and the expansion section. The diameter of the constriction section facing the expansion section is smaller than the diameter of the expansion section away from the constriction section.

[0038] The structure consisting of a contraction section and an expansion section can create a local negative pressure zone within the exhaust pipe, thereby accelerating gas discharge, reducing gas residence time, and increasing gas renewal rate. Increased exhaust velocity effectively prevents backflow of external ambient gases into the exhaust pipe due to diffusion or pressure fluctuations.

[0039] Preferably, the exhaust pipe includes an inner pipe and an outer pipe. One end of the inner pipe is connected to the third cylinder. The contraction section and the expansion section are both located on the inner pipe. The expansion section is located at the end of the inner pipe facing away from the third cylinder. The outer pipe is sleeved on the outer periphery of the other end of the inner pipe, and an annular gap is formed between the outer pipe and the inner pipe. The outer pipe is used to connect to an external inert gas source, so that the inert gas is ejected through the annular gap along the exhaust direction. The exhaust direction is the same as the gas flow direction in the inner pipe.

[0040] The protective gas ejected from the annular gap forms a laminar protective gas sleeve, enveloping the gas jet exiting the inner tube from all sides and preventing external gas from backflowing into the inner tube due to diffusion or pressure fluctuations. This protective gas can be a dry, inert sheath gas (such as nitrogen), and its flow rate ratio to the gas flow rate in the inner tube can be controlled between 5:1 and 20:1. This high flow rate ratio ensures that the protective gas sleeve has sufficient momentum and thickness to effectively resist interference from external airflow while maintaining laminar flow.

[0041] Optionally, the variable diameter module can adopt various structural forms to adjust the flow area of ​​the exhaust pipe, such as: a needle valve structure, which changes the annular gap between the valve core and the valve seat by rotating the valve core to achieve continuous and precise flow regulation; a butterfly valve structure, which sets a rotatable butterfly plate in the exhaust pipe and adjusts the flow area by changing the angle between the butterfly plate and the axis of the exhaust pipe; and a ball valve structure, which changes the relative angle or the size of the through hole between the ball core and the pipe axis by rotating the perforated ball core to achieve flow on / off and regulation.

[0042] In addition to the above-mentioned structural forms, any structure that can adjust the flow area of ​​the exhaust pipe can be used, and the specific configuration can be selected according to actual needs.

[0043] Preferably, the variable diameter module includes a movable component and a driving component. The movable component is disposed inside the exhaust pipe, and the driving component is used to drive the movable component to move, thereby changing the flow area of ​​the exhaust pipe.

[0044] The actuator can employ a piezoelectric actuator or a shape memory alloy driven mechanical structure. The piezoelectric actuator utilizes the inverse piezoelectric effect to generate displacement, and its output characteristics are less affected by temperature, maintaining stable driving force and displacement accuracy within the normal operating temperature range of the exhaust pipe. While the shape memory alloy drive utilizes temperature changes to generate phase transition restoring force, its phase transition temperature can be designed and selected according to the actual operating temperature range of the exhaust pipe, ensuring reliable execution of the drive action within the stable temperature environment established by the temperature control module. Therefore, temperature changes in the exhaust pipe will not interfere with the normal operation of the diameter-changing module. Compared to traditional solenoid valves or stepper motors, piezoelectric actuators and shape memory alloy drivers are smaller, consume less power, and have higher positioning accuracy, which is beneficial for the miniaturization and low-power design of the detection device. Simultaneously, neither drive method generates significant electromagnetic interference, avoiding influence on the signal of the first detector and improving the signal-to-noise ratio and stability of concentration measurement.

[0045] Preferably, the system further includes a control unit configured to: receive a gas concentration signal output by the detector module, and control the drive component to change the flow area of ​​the exhaust pipe according to a preset relationship; wherein the preset relationship is a functional relationship between gas concentration and flow area, and satisfies the following: when the gas concentration is higher than a first threshold, the flow area increases with the increase of gas concentration; when the gas concentration is lower than a second threshold, the flow area decreases with the decrease of gas concentration, and the flow area is greater than zero; when the gas concentration is between the first threshold and the second threshold, the flow area is a constant value, and the first threshold is greater than the second threshold.

[0046] When the gas concentration is higher than the first threshold, the flow area increases with the increase of concentration, which can accelerate the exhaust to improve the response speed and track the decreasing trend of gas concentration in a timely manner. When the gas concentration is lower than the second threshold, the flow area decreases with the decrease of concentration, which can reduce the exhaust speed to prolong the residence time of gas in the second chamber, enhance the infrared absorption signal, and improve the signal-to-noise ratio and accuracy of low concentration measurement.

[0047] Preferably, the control unit is further configured to: perform feedforward-feedback composite control based on the theoretical gas concentration predicted by the prediction model and the measured concentration of the detector module, and control the action of the drive component to dynamically adjust the flow area of ​​the exhaust pipe; wherein, the prediction model is a model pre-established based on the law of gas concentration change over time in the cleaning process before the start of the cleaning process, and the prediction model is used to predict the theoretical gas concentration at different times during the operation of the cleaning process.

[0048] Feedforward control, based on a predictive model, anticipates gas concentration changes and adjusts the flow area before actual concentration changes occur. This overcomes the inherent lag in pure feedback control, allowing the detection device to respond promptly to rapidly decreasing concentration trends and preventing delays in cleaning progress assessment. Feedback control, on the other hand, corrects for deviations between the measured concentration and the predicted value from the detector module. This compensates for model prediction errors and random disturbances in the process, such as temperature fluctuations or deviations in gas composition and flow rate, thereby improving control accuracy. Even at low concentration levels with low signal-to-noise ratios, feedforward control maintains stable adjustment without relying on the measured signal. Through this combined feedforward and feedback control, the detection device actively compensates for deviations between the actual flow field and ideal simulation conditions, consistently approaching the optimal operating state throughout the cleaning process and accurately capturing the cleaning endpoint.

[0049] On the other hand, the present invention also provides a dynamic adjustment method for the aforementioned infrared endpoint detection device, comprising the following steps: predicting the gas concentration at the current moment based on a prediction model; when the theoretical gas concentration predicted by the prediction model or the gas concentration measured by the detector module is higher than a first threshold, controlling the diameter-changing module to increase the flow area of ​​the exhaust pipe; when the theoretical gas concentration predicted by the prediction model or the gas concentration measured by the detector module is lower than a second threshold, controlling the diameter-changing module to reduce the flow area of ​​the exhaust pipe, wherein the flow area is greater than zero, and the first threshold is greater than the second threshold; through a PID control algorithm, independently adjusting and stabilizing the actual temperatures of the intake pipe, the first cylinder and / or the second cylinder and / or the third cylinder, and the exhaust pipe to their respective target temperatures, so that a decreasing temperature gradient is formed along the gas flow direction, and each target temperature is higher than the condensation point of the detected process gas; wherein, the prediction model is an exponential decay model pre-established based on reaction kinetics theory before the start of the cleaning process, and its expression is: C(t)=C0·e -kt In the formula, C(t) is the predicted gas concentration at time t; C0 is the initial gas concentration at the start of the cleaning process; k is the reaction rate constant, which is predetermined based on the type of cleaning gas, the volume of the process chamber, the cleaning temperature, and the initial concentration of the substance being cleaned; and t is the time during which the cleaning process is carried out.

[0050] In the early stages of the cleaning process, the gas concentration is high and there may be residual background gas in the gas chamber module. At this time, the variable diameter module increases the flow area of ​​the exhaust pipe to accelerate the discharge of background gas from the gas chamber module and improve the gas renewal rate within the gas chamber module, enabling the detection device to respond quickly to concentration changes. In the middle stages of the cleaning process, as the gas concentration gradually decreases, the variable diameter module uses a feedforward-feedback composite control strategy for dynamic adjustment: on the one hand, it predicts the concentration decrease trend in advance based on the prediction model and pre-adjusts the flow area to overcome the inherent lag of feedback control, allowing the detection device to keep up with the pace of concentration changes; on the other hand, it performs feedback correction based on the deviation between the measured concentration and the predicted value from the detector module, compensating for model prediction errors and process disturbances such as temperature fluctuations and changes in gas composition, thereby achieving a balance between response speed and control accuracy. Simultaneously, the segmented temperature control module uses a PID control algorithm to stabilize the temperatures of the inlet pipe, gas chamber, and exhaust pipe at their respective preset target temperatures, forming a decreasing temperature gradient along the gas flow direction. This ensures that the gas remains in a gaseous state during transmission, preventing condensation that could clog the exhaust port or contaminate the optical window. The stable gas chamber temperature also keeps the infrared absorption coefficient consistent, providing stable environmental conditions for concentration measurement. As the cleaning process nears its expected endpoint, the gas concentration drops to a low level. At this point, reducing the flow area of ​​the exhaust pipe decreases the gas flow rate and extends the gas residence time in the second chamber. This allows the detection device to perform more photon absorption integrations on the low-concentration residual gas per unit time, thereby improving the signal-to-noise ratio and accurately capturing the inflection point of the cleaning endpoint. This avoids misjudgments of the endpoint due to insufficient signal-to-noise ratio. Through the synergistic effect of variable diameter adjustment and segmented temperature control, the detection device maintains optimal measurement performance throughout the entire cleaning process.

[0051] Preferably, the flow area adjustment step of the exhaust pipe specifically includes: predicting the gas concentration at the current moment according to the prediction model, and calculating the feedforward flow area according to a preset relationship; comparing the concentration fed back by the detector module in real time with the predicted concentration of the prediction model, and calculating the area correction amount based on the deviation between the two; superimposing the feedforward flow area and the area correction amount to obtain the target flow area, and controlling the variable diameter module to adjust the flow area of ​​the exhaust pipe to the target flow area; determining whether the gas concentration predicted by the prediction model or measured by the detector module is lower than a preset endpoint threshold, if yes, the adjustment ends, otherwise returning to the calculation of the feedforward flow area and subsequent steps, until the gas concentration predicted by the prediction model or measured by the detector module is lower than the endpoint threshold; wherein, the prediction model is a model pre-established before the start of the cleaning process based on the law of gas concentration changing with time in the cleaning process; the preset relationship includes: when the gas concentration is higher than the first threshold, the flow area increases with the increase of gas concentration, when the gas concentration is lower than the second threshold, the flow area decreases with the decrease of gas concentration, and the flow area is greater than zero, the first threshold is greater than the second threshold, and the endpoint threshold is less than the second threshold.

[0052] The beneficial effects of this invention are: (1) A three-stage anti-free jet and entrainment structure is constructed by coaxially nested first, second and third cylinders. The first chamber converts the high-speed beam jet into a low-speed annular wall-attached flow to eliminate the initial directional momentum required for jet entrainment. The densely distributed and uniformly distributed exhaust holes on the third cylinder divide the airflow entering the second chamber into multiple micro-jet streams, which simultaneously seep into the third chamber, so that there is no dominant flow direction in the second chamber, thus eliminating the physical basis for the jet entrainment effect. On this basis, a segmented temperature control module is set along the gas transmission path to maintain the set temperature difference between the exhaust pipe, the inlet pipe and the first cylinder, forming a decreasing temperature gradient with the inlet pipe and the first cylinder having the same and highest temperature, and the exhaust pipe having the lowest temperature. On the one hand, the gas temperature transitions smoothly to prevent condensation and blockage of the exhaust holes. On the other hand, a stable density stratification is formed from the low density area to the high density area, further suppressing the entrainment of the jet on the surrounding gas, while keeping the infrared absorption coefficient consistent. Furthermore, a variable-diameter module connected to the exhaust pipe is installed to dynamically adjust the flow area according to changes in gas concentration. Combined with a feedforward-feedback composite control strategy, the area is increased to accelerate exhaust and improve response speed at high concentrations, while the area is reduced to extend gas residence time and improve signal-to-noise ratio at low concentrations. Simultaneously, the concentration change trend is predicted in advance based on a reaction kinetic model to overcome the inherent lag of pure feedback control, and feedback correction is performed based on the deviation between the measured concentration and the predicted value to compensate for model errors and process disturbances. The aforementioned anti-free-jet and anti-entrainment structure, segmented temperature control module, and variable-diameter adjustment work synergistically. The anti-entrainment structure provides a stable flow field basis for variable-diameter adjustment; the segmented temperature control module further suppresses entrainment through stable density stratification and creates a stable temperature environment for variable-diameter adjustment; and the variable-diameter adjustment achieves the optimal balance between response speed and measurement accuracy across the entire concentration range. This ensures that the gas concentration along the infrared measurement path accurately reflects the average concentration of the entire cavity, reliably captures the cleaning endpoint inflection point, and avoids incomplete or over-cleaning.

[0053] (2) Uniformly distributed and circumferentially alternating arc-shaped protrusions are provided on the inner wall of the first cylinder and the outer wall of the second cylinder. The first arc-shaped protrusion and the second arc-shaped protrusion utilize the Coanda effect to guide the gas to flow along the wall. At the same time, the alternating distribution in the circumferential and axial directions forces the gas to continuously and freely mix and flow in the axial and radial directions while maintaining overall wall adhesion. This can promote the full exchange and integration of gas, and further improve the uniformity of gas concentration in the first chamber while maintaining the prevention of free jetting and entrainment.

[0054] (3) The multi-point distribution of exhaust holes can eliminate the gas entrainment of traditional single outlet. The exhaust holes adopt a frustum-shaped structure with a small inlet and a large outlet. With the optimized parameters of a diameter-to-depth ratio of 0.3~0.5 and an outward guide angle of 15°~30°, the flow velocity of gas can be gradually reduced when it seeps out, avoiding reverse disturbance to the second chamber. At the same time, the pressure gradient inside the hole is maintained, guiding the airflow to diffuse naturally without separation. It can also prevent the small particles that may be generated during chamber cleaning from entering the pipeline with the airflow and accumulating at the outlet, affecting the gas flow field distribution or even blocking the outlet.

[0055] (4) Set up an angle adjustment module. The angle adjustment module can find and lock the optimal optical axis position by passing in the light intensity change curve, eliminate the eccentricity error between the center of the optical path and the center of the flow field, and ensure that the infrared light is always adjusted to pass through the area corresponding to the optimal detection angle position, thereby improving the accuracy of gas concentration measurement.

[0056] (5) The exhaust pipe is equipped with a contraction section and an expansion section, which can form a local negative pressure zone to accelerate exhaust and prevent external gas backflow. At the same time, the exhaust pipe adopts a double-layer structure composed of an inner pipe and an outer pipe. The annular gap of the double-layer structure can be used to spray out protective gas to form a protective gas sleeve in a laminar flow state, which wraps the gas jet discharged from the inner pipe from all sides, preventing external gas from backflowing into the inner pipe due to diffusion or pressure fluctuation, thereby ensuring the purity of the gas to be tested and the accuracy of the measurement results during the detection process. Attached Figure Description

[0057] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0058] Figure 1 A schematic diagram of the end point detection device for preventing free jet and entrainment provided in an embodiment of the present invention; Figure 2 This is a radial sectional view of the air chamber module; Figure 3 This is a schematic diagram showing the distribution of the first and second arc-shaped protrusions in the circumferential direction of the first chamber; Figure 4 This is a schematic diagram showing the distribution of the first and second arc-shaped protrusions along the axial direction of the first chamber; Figure 5 This is a cross-sectional view of the vent. Figure 6 This is a structural diagram of the angle adjustment module; Figure 7This is a schematic diagram of the contraction and expansion sections; Figure 8 This is a schematic diagram showing the connection between the inner and outer pipes.

[0059] Explanation of reference numerals in the attached figures: 1. Light source module; 2. First cylinder; 201. Air inlet; 202. First chamber; 203. Exhaust port; 3. Second cylinder; 301. Third chamber; 302. First opening; 4. Exhaust pipe; 401. Inner pipe; 402. Outer pipe; 4021. Annular gap; 403. Contraction section; 404. Expansion section; 5. Third cylinder; 501. Exhaust hole; 502. Second opening; 503. Second chamber; 6. Channel; 7. Detector module; 8. First collimating lens; 9. Second collimating lens; 10. First arc-shaped protrusion; 11. Second arc-shaped protrusion; 12. Angle adjustment module; 13. Variable diameter module; 14. Air inlet pipe. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] The following is combined Figures 1 to 8 The following describes embodiments of the present invention.

[0062] On one hand, embodiments of the present invention provide an endpoint detection device to prevent free jetting and entrainment, used to detect the endpoint of the cleaning process in thin film growth equipment, combined with... Figure 1 and Figure 2As shown, it includes a light source module 1, a gas chamber module, a variable diameter module 13, a segmented temperature control module, and a detector module 7. The gas chamber module includes a first cylinder 2, a second cylinder 3, and a third cylinder 5. The first cylinder 2 has an air inlet 201 on its side wall and a first wall-mounted flow guide structure on its inner wall. An air inlet pipe 14 is connected to the air inlet 201. The second cylinder 3 is located inside the first cylinder 2 and is coaxial with it. The second cylinder 3 has a second wall-mounted flow guide structure on its outer wall and a first opening 302 and an exhaust port 203 on its side wall. An exhaust pipe 4 is connected to the exhaust port 203, with one end of the exhaust pipe 4 extending outwards from the exhaust port 203 to the outside of the first cylinder 2. The third cylinder 5 is located inside the second cylinder 3 and is coaxial with the second cylinder 2. The body 3 is coaxial, and the wall of the third cylinder 5 is provided with a second opening 502 and multiple exhaust holes 501. A channel 6 is provided between the second opening 502 and the first opening 302. The second cylinder 3 and the first cylinder 2 enclose a first chamber 202, which is used to prevent free jets and form a wall-attached circulation. The inner cavity of the third cylinder 5 is a second chamber 503, which is connected to the first chamber 202 through the channel 6 to form a stable detection area. The third cylinder 5 and the second cylinder 3 enclose a third chamber 301, which is connected to the second chamber 503 through the exhaust holes 501 for uniform multi-point exhaust.

[0063] Furthermore, the variable diameter module 13 is used to adjust the flow area of ​​the exhaust pipe 4. The segmented temperature control module includes an integrated partitioned heating jacket, a first PID controller, a second PID controller, and a linkage control module. The integrated partitioned heating jacket is fitted on the outside of the gas chamber module and is divided into a first temperature zone and a second temperature zone. The first temperature zone covers the area where the intake pipe 14 and the first cylinder 2 are located and is independently controlled by the first PID controller. The second temperature zone covers the exhaust pipe 4 and is independently controlled by the second PID controller. The linkage control module is connected to the first PID controller and the second PID controller respectively and is configured to coordinate the heating power of the first temperature zone and the second temperature zone based on the temperature information of the first temperature zone and the second temperature zone to maintain the preset target temperature difference, so that the temperature of the intake pipe 14, the first cylinder 2, the second cylinder 3, the third cylinder 5, and the exhaust pipe 4 are all higher than the condensation point of the detected process gas, and the temperature of the intake pipe 14 is the same as the temperature of the first cylinder 2, the second cylinder 3, and the third cylinder 5, and higher than the temperature of the exhaust pipe 4, so as to form a temperature gradient that decreases along the gas flow direction. The detector module 7 includes a first detector and a second detector. The position of the first detector corresponds to the position of the second chamber 503 and is used for real-time concentration signal detection. The position of the second detector corresponds to the position of the third chamber 301 and is used for light attenuation monitoring of the light source module 1.

[0064] In this embodiment, the endpoint detection device for preventing free jet and entrainment is used such that the gas to be tested enters the first chamber 202 through the air inlet 201 on the side wall of the first cylinder 2. After entering the first chamber 202, the gas first changes from a beam jet to a ring flow that diffuses uniformly in the circumferential direction, initially eliminating the directional momentum of the jet. At the same time, the first wall-adhering flow guiding structure on the inner wall of the first cylinder 2 and the second wall-adhering flow guiding structure on the outer wall of the second cylinder 3 cooperate to guide the gas to adhere to the wall surface and flow, so that the gas maintains a stable wall-adhering state while diffusing in the circumferential direction, and generates local micro-disturbances in the near-wall region to promote uniform gas mixing and create conditions for subsequent flow uniformity.

[0065] Subsequently, the gas enters the second chamber 503 sequentially through the first port 302, channel 6, and second port 502. Once inside the second chamber 503, the gas no longer forms a jet or flows violently; instead, driven by the pressure difference, it seeps out into the third chamber 301 through multiple vent holes 501 opened in the wall of the third cylinder 5. Because the vent holes 501 are evenly distributed on the wall of the third cylinder 5, the gas can seep out synchronously and uniformly, avoiding localized high-speed jets or reverse disturbances in the second chamber 503, thus helping to maintain the uniformity and stability of the gas inside the chamber.

[0066] Meanwhile, the segmented temperature control module controls the first temperature zone, where the intake pipe 14 and the first cylinder 2 are located, to the first target temperature. It also adjusts the heating temperature of the exhaust pipe 4 in the second temperature zone in real time according to temperature changes in the first temperature zone, maintaining the temperature difference between the exhaust pipe 4, the intake pipe 14, and the first cylinder 2 at a set value, thus forming a decreasing temperature gradient along the gas flow direction. Specifically, the intake pipe 14 and the first cylinder 2 have the same and highest temperature, rapidly preheating the low-temperature gas to above its condensation point, preventing condensation before entering the first chamber 202. The second cylinder 3 and the third cylinder 5 have the same temperature as the first cylinder 2, ensuring the gas remains gaseous within the gas chamber module, preventing blockage of the exhaust port 501. The exhaust pipe 4 has the lowest temperature but still above its condensation point, preventing condensation and accumulation of gas during exhaust. This temperature gradient ensures a smooth transition in gas temperature, avoiding thermally induced flow disturbances caused by sudden temperature changes. Furthermore, according to stratified fluid dynamics theory, when fluid moves from a low-density region to a high-density region (density increases along the flow direction), a stable density stratification is formed. Under these conditions, turbulent vortices need to overcome buoyancy to achieve vertical mixing, significantly reducing mixing efficiency. This manifests as a decrease in the entrainment rate of the jet boundary layer, which can be further reduced by stable stratification. The inlet pipe 14 and the first cylinder 2 have the same and highest temperatures, the second cylinder 3 and the third cylinder 5 have the same temperatures as the first cylinder 2, and the exhaust pipe 4 has the lowest temperature. This creates a stable stratification condition with increasing density along the gas flow direction, fundamentally suppressing the entrainment of the jet on the surrounding gas from a hydrodynamic perspective. Simultaneously, the stable temperature within the detection area ensures the consistency of the infrared absorption coefficient, further improving the repeatability and accuracy of concentration measurements.

[0067] The variable diameter module 13 dynamically adjusts the flow area of ​​the exhaust pipe 4 according to changes in gas concentration. When the gas concentration is high, the flow area is increased to accelerate the exhaust speed, improve the response speed of the detection device, and promptly track the concentration decrease trend. When the gas concentration is low, the flow area is reduced to decrease the gas flow rate, prolong the residence time of the gas in the second chamber 503, enhance the infrared absorption signal, and improve the signal-to-noise ratio and low-concentration measurement accuracy. Through variable diameter adjustment, the detection device can obtain optimal measurement performance at different concentration stages.

[0068] Because the wall-mounted flow guiding structure eliminates the directional momentum of the jet, the uniformly distributed exhaust holes 501 achieve synchronous and uniform gas permeation, the segmented temperature control module maintains a stable temperature gradient and forms density stratification to suppress entrainment, and the variable diameter adjustment optimizes the measurement conditions at different concentration stages, so that the gas in the second chamber 503 has almost no directional flow. The entire inner cavity is free of high-speed jets, entrainment vortices, and concentration gradients, forming a highly uniform and stable gas environment. Finally, the gas in the third chamber 301 is discharged to the outside of the device through the exhaust port 203 and the exhaust pipe 4.

[0069] Simultaneously with the gas flow, the light source module 1 emits infrared light into the second chamber 503. The infrared light passes through the uniform and stable gas in the second chamber 503 and is received by the first detector. The first detector outputs a corresponding gas concentration detection signal based on the degree of infrared light absorption. The second detector periodically monitors the light attenuation of the light source module 1 to ensure that the light intensity is within a reasonable attenuation range (e.g., setting 95%~100% of the initial signal as the normal range) to guarantee high-accuracy detection.

[0070] The endpoint detection device for preventing free jet and entrainment provided in this embodiment converts the beam jet into a wall-mounted circulating flow through the wall-mounted guiding structure of the first chamber 202. Combined with the densely distributed exhaust holes 501 on the third cylinder 5, it achieves uniform seepage, fundamentally eliminating the high-speed jet and its entrainment vortices. This allows the second chamber 503 to form a highly uniform and stable gas region with no directional flow and no concentration gradient, significantly reducing measurement errors. Furthermore, the segmented temperature control module controls the first temperature zone at the first target temperature and adjusts the heating temperature of the second temperature zone in real time according to the temperature changes in the first temperature zone. This maintains the set temperature difference between the exhaust pipe 4, the inlet pipe 14, and the first cylinder 2, forming a decreasing temperature gradient along the gas flow direction. This not only prevents gas condensation and blockage of the exhaust holes 501 during transmission but also avoids thermal flow disturbances caused by sudden temperature changes. Simultaneously, according to stratified fluid dynamics theory, the stable density stratification formed by this decreasing temperature gradient suppresses the entrainment of the jet into the surrounding gas. The stable temperature of the first temperature zone further ensures the consistency of the infrared absorption coefficient, thereby improving the repeatability and accuracy of concentration measurements. Meanwhile, the variable diameter module 13 dynamically adjusts the flow area of ​​the exhaust pipe 4 according to changes in gas concentration. At high concentrations, the area is increased to accelerate exhaust and improve response speed; at low concentrations, the area is reduced to prolong gas residence time and improve signal-to-noise ratio, achieving optimal measurement performance across the entire concentration range. The synergistic effect of wall-mounted flow guidance, uniform exhaust, gradient temperature control, and variable diameter adjustment enables the detection device to accurately reflect the actual gas concentration changes within the process chamber, reliably capture the inflection point at the cleaning endpoint, and avoid incomplete or excessive cleaning. Furthermore, the detector module performs concentration detection and light source attenuation monitoring functions, promptly detecting light source aging or malfunctions and ensuring the long-term measurement reliability of the device.

[0071] In one optional embodiment, a first collimating lens 8 is provided between the light source module 1 and the first cylinder 2, and a second collimating lens 9 is provided between the detector module 7 and the first cylinder 2.

[0072] The function of the first collimating lens 8 is to convert the divergent infrared light emitted by the light source module 1 into parallel light, so that the infrared light propagates along the axis of the second chamber 503, avoiding the light from being absorbed or scattered by obliquely hitting the inner wall of the cylinder, and reducing light energy loss. The function of the second collimating lens 9 is to refocus the parallel light after passing through the gas chamber module onto the photosensitive surface of the detector module 7, ensuring that the light is effectively received. The two collimating lenses work together to ensure that the path length of the infrared light is consistent when passing through the gas to be measured, which can improve the accuracy and linearity of gas concentration measurement; on the other hand, it can improve the utilization efficiency of light energy and enhance the signal strength received by the detector module 7, thereby improving the signal-to-noise ratio and low concentration measurement capability.

[0073] Since the gas, after being decelerated in the first chamber 202, may still experience velocity fluctuations and pressure shocks, directly entering the inner cavity would interfere with the flow uniformity stability of the exhaust port 501. Therefore, in an optional embodiment, the inner wall of the channel 6 is arc-shaped, the cross-sectional area of ​​the middle part of the channel 6 is smaller than the cross-sectional areas of its two ends, and the cross-sectional area of ​​the channel 6 continuously decreases from both ends to the middle.

[0074] Understandably, if a conventional straight-line fixed-diameter channel 6 is used, it can only serve a guiding function and cannot suppress residual velocity pulsations and pressure shocks in the upstream airflow. The disturbances will be directly transmitted to the second chamber 503, interfering with the flow uniformity and stability of the exhaust port 501. Furthermore, a stepped variable-section channel 6 would introduce new vortices. Therefore, an arc-shaped variable-section channel 6 is chosen, allowing the flow area of ​​channel 6 to first smoothly contract and then expand. This enables the airflow to achieve a smooth conversion between kinetic energy and pressure potential energy during the continuous change of cross-sectional area. In this process, the velocity pulsations and pressure shocks of the airflow are gradually smoothed out by the flow channel, and the wave energy is converted into internal energy dissipation. Ultimately, this allows the gas to enter the second chamber 503 more smoothly, forming a highly uniform and stable gas environment within the second chamber 503.

[0075] In one optional embodiment, the air inlet 201 and the exhaust outlet 203 are both located on the first side of the central axis of the first cylinder 2 and are located in the same radial position. The channel 6 is located on the second side of the central axis of the first cylinder 2. The first side and the second side are opposite sides of the first cylinder 2 in the radial direction.

[0076] This arrangement requires gas to travel circumferentially through the inlet 201 into the first chamber 202 before entering the second chamber 503 via the channel 6 on the second side. This extends the circumferential flow path and residence time of the gas in the first chamber 202, promoting thorough diffusion and mixing, and resulting in a more uniform circumferential distribution of airflow within the third cylinder 5. Ultimately, this leads to a more uniform and stable gas flow within the third cylinder 5.

[0077] In one alternative implementation, combined with Figures 1 to 4 As shown, the first wall-mounted flow guiding structure includes multiple first arc-shaped protrusions 10, which are evenly distributed on the inner circumferential wall of the first cylinder 2; the second wall-mounted flow guiding structure includes multiple second arc-shaped protrusions 11, which are evenly distributed on the outer circumferential wall of the second cylinder 3.

[0078] As the gas flows circumferentially and axially within the first chamber 202, it sequentially passes through a first arc-shaped protrusion 10 uniformly distributed on the inner circumferential wall of the first cylinder 2 and a second arc-shaped protrusion 11 uniformly distributed on the outer circumferential wall of the second cylinder 3. When the gas flows past the arc-shaped protrusions, the arc-shaped surface generates a Coanda effect, allowing the gas to naturally adhere to the surface and flow without separation. Simultaneously, due to local variations in the flow channel cross-section, the gas is accelerated by compression on the windward side of the protrusion and decelerated by streamline expansion on the leeward side. This velocity difference forms tiny local vortices on both sides of the protrusion. These micro-vortices are small in scale and low in energy, and will not develop into large-scale turbulence or disrupt the overall wall-attached flow, but are sufficient to promote gas exchange between adjacent flow layers, achieving gentle mixing in the near-wall region. The uniform distribution of multiple protrusions along the circumferential and axial directions ensures that the gas is continuously subjected to this local micro-disturbance throughout the entire flow path, maintaining a wall-attached state and gradually achieving homogenization of concentration and temperature, resulting in a more uniform state before entering the channel 6.

[0079] In one alternative embodiment, the first arc-shaped protrusion 10 and the second arc-shaped protrusion 11 are alternately distributed in the circumferential and axial directions of the first chamber 202.

[0080] As the gas flows along the wall in both the circumferential and axial directions, it passes through the first arc-shaped protrusion 10 and the second arc-shaped protrusion 11. Because the first arc-shaped protrusion 10 and the second arc-shaped protrusion 11 are staggered in both the circumferential and axial directions, the gas can maintain a continuous, free-mixing flow state, exhibiting a gentle positional migration in the radial direction along a smooth trajectory, gradually transitioning from near the outer wall to near the inner wall, and then gradually transitioning back, forming multiple back-and-forth flow segments. This radial migration allows for sufficient exchange of gas between the inner and outer layers of the annular channel without disrupting the overall wall-hugging flow or generating large-scale eddies, resulting in a more uniform gas concentration. The staggered distribution, combined with the Coanda effect of the arc-shaped protrusions, ensures that the gas maintains wall-hugging flow while achieving continuous, free-mixing flow and radial exchange, avoiding measurement deviations caused by uneven concentrations between the inner and outer layers, and further improving the reliability of the cleanliness endpoint determination.

[0081] In one alternative implementation, such as Figure 5 As shown, the exhaust port 501 is a frustum-shaped hole, and the diameter of the end of the exhaust port 501 facing the inside of the third cylinder 5 is smaller than that of the other end.

[0082] Because the exhaust port 501 adopts a gradually expanding structure with a small inlet and a large outlet, when the gas seeps out of the second chamber 503 through the exhaust port 501, as the flow area gradually increases, the gas velocity and kinetic energy gradually decrease. In this way, the gas can enter the third chamber 301 more gently, avoiding reverse disturbance to the second chamber 503, which is conducive to improving the uniformity and stability of the gas in the second chamber 503.

[0083] In one optional embodiment, the ratio of the central diameter to the depth of the vent hole 501 ranges from 0.3 to 0.5, where the units for both the central diameter and depth of the vent hole 501 are mm; the outward guide angle of the vent hole 501 (i.e., Figure 5 The α angle in the equation ranges from 15° to 30°.

[0084] For example, the ratio of the central diameter of the vent hole 501 to the depth of the vent hole 501 can be selected as 0.3, 0.35, 0.4, 0.45, 0.5, etc., or any value between the two values.

[0085] For example, the outward guide angle of the exhaust port 501 can be selected as 15°, 17°, 19°, 21°, 23°, 25°, 27°, 29° or 30°, etc., as well as any value between two values.

[0086] When the ratio of the central diameter to the depth of the exhaust port 501 is less than 0.3, the exhaust port 501 is too long and thin, resulting in high flow resistance and a sharp increase in pressure drop when gas flows through it. This leads to a decrease in gas percolation velocity and a prolonged response time of the detection device to changes in the concentration of the process chamber, making it unable to reflect the cleaning progress in real time. Simultaneously, the long and thin channel is easily blocked by tiny particles carried in the gas. When this ratio is greater than 0.5, the exhaust port 501 is too short and thick, resulting in insufficient gas deceleration and diffusion effects, and the percolation velocity remains high, failing to effectively suppress the reverse disturbance to the second chamber 503. When the outward guide angle of the exhaust port 501 is less than 15°, the gradual diffusion effect is not obvious, and the effect of reducing gas flow velocity is limited. When the outward guide angle of the exhaust port 501 is greater than 30°, the orifice wall expands too violently, and the airflow is prone to flow separation, which can easily generate new disturbances. An angle range of 15° to 30° maintains a pressure gradient within the orifice, guiding the airflow to diffuse naturally without separation. The two parameters mentioned above are coupled together, which can achieve a better balance between low disturbance leakage, appropriate pressure drop and fast response, ensuring that the detection device can work stably and reliably for a long time.

[0087] In one alternative implementation, such as Figure 1 As shown, the integrated partitioned heating jacket includes a first heating jacket section and a second heating jacket section. The first heating jacket section is fitted onto the outside of the air inlet pipe 14 and the first cylinder 2. The first heating jacket section is equipped with a first heating component and a first temperature measuring component. The first heating component and the first temperature measuring component are respectively connected to the first PID controller signal. The second heating jacket section is fitted onto the outside of the exhaust pipe 4. The second heating jacket section is equipped with a second heating component and a second temperature measuring component. The second heating component and the second temperature measuring component are respectively connected to the second PID controller signal.

[0088] The integrated zone heating jacket is divided into a first heating section and a second heating section, which are respectively fitted onto the outer sides of the air inlet pipe 14 and the first cylinder 2, and the outer side of the exhaust pipe 4. This makes the first and second temperature zones physically independent, avoiding thermal interference between the two zones and facilitating precise temperature control. Each heating section integrates independent heating and temperature sensing components, which can be connected to the corresponding PID controller signals. Based on the real-time temperature feedback signals from each temperature sensing component, the PID controller uses a proportional unit to accelerate the response speed, an integral unit to eliminate steady-state errors, and a derivative unit to suppress temperature fluctuations, independently adjusting the heating power of each heating component. This allows the temperature in each zone to quickly converge and stabilize at its preset target temperature. This achieves real-time feedback adjustment of the heating power, which improves the response speed and stability of temperature control.

[0089] In one optional embodiment, the first heating sleeve and the second heating sleeve each adopt a multi-layer composite structure, comprising, from the inside out, a heat-conducting inner layer, an electric heating layer, a heat insulation layer, and a protective outer shell. The heat-conducting inner layer is made of copper or aluminum with a thickness of 0.5 mm to 1.5 mm, tightly fitted to the outer walls of the air inlet pipe 14, the first cylinder 2, and the exhaust pipe 4 to ensure uniform heat conduction. The electric heating layer is an etched metal heating film with a polyimide substrate. The heating wires in the first heating sleeve and the second heating sleeve are arranged independently and have different power densities per unit length. The heat insulation layer is made of glass fiber or aerogel felt with a thickness of 5 mm to 15 mm to reduce heat loss to the environment. The protective outer shell is made of stainless steel or aluminum alloy to provide mechanical protection and electrical insulation.

[0090] Furthermore, the heating element in the first heating section is a heating wire, employing high-density serpentine wiring with a power per unit length of 8W / cm to 12W / cm; the heating element in the second heating section is also a heating wire, employing low-density straight wiring with a power per unit length of 2W / cm to 4W / cm. A 2mm to 5mm insulating gap is maintained between the heating wires of the first and second heating sections to prevent thermal short circuits.

[0091] In one optional embodiment, the first temperature sensing component is located in the middle of the outer wall of the first cylinder 2, using a PT100 platinum resistance thermometer with an accuracy class of A and a measurement accuracy of ±0.15℃, and is in close contact with the outer wall of the first cylinder 2 via thermally conductive silicone grease; the second temperature sensing component is installed on the outer wall of the exhaust pipe 4 near the end, also using a PT100 platinum resistance thermometer. The signals from the two temperature sensing components are transmitted to the first PID controller and the second PID controller respectively via shielded cables to achieve precise temperature feedback control.

[0092] In one optional implementation, the linkage control module is configured to coordinate the first PID controller and the second PID controller in a master-slave follow mode. In the master-slave follow mode, the second PID controller dynamically adjusts the heating power of the second heating element according to the measured temperature fed back by the first temperature measuring element, so that the temperature fed back by the second temperature measuring element is 10°C to 30°C lower than the temperature fed back by the first temperature measuring element.

[0093] A fixed temperature difference of 10℃ to 30℃ ensures a stable temperature gradient along the gas flow direction, causing the gas density to increase along the flow direction. This fundamentally suppresses the entrainment of surrounding gas by the jet, improving the uniformity and stability of the gas in the second chamber 503. Simultaneously, dynamic tracking control can respond in real-time to temperature fluctuations in the first temperature zone, adjusting the heating power of the second temperature zone promptly to maintain the set temperature difference, ensuring that the temperature gradient always meets process requirements.

[0094] In one optional implementation, the temperatures of the first and second temperature zones are set according to process requirements. For example, when the radial length of the chamber is 200 mm, the first target temperature can be set to 200°C and the second target temperature to 180°C, i.e., a temperature difference of 20°C; or the first target temperature can be set to 150°C and the second target temperature to 140°C, i.e., a temperature difference of 10°C. The temperature control accuracy of both the first and second PID controllers is ±0.05°C to ensure the precise stability of the temperature gradient.

[0095] In one alternative implementation, combined with Figure 1 and Figure 6 As shown, it also includes an angle adjustment module 12. The power output terminal of the light source module 1 is connected to the angle adjustment module 12. The angle adjustment module 12 is used to adjust the optical axis direction of the light source module 1 to minimize the eccentricity error of the optical path and the flow field. It can be understood that the optical axis is the center line of the main direction of infrared light emission, that is, the central axis of the infrared beam emitted by the light source module 1.

[0096] Before using the detection device, the optical path can be calibrated using the angle adjustment module 12. First, pure nitrogen gas is introduced into the detection device. Nitrogen gas does not absorb infrared light. Then, the angle adjustment module 12 is driven to change the angle of the light source module 1, while simultaneously detecting the infrared light intensity received by the first detector, obtaining a curve showing the relationship between light intensity and angle. When the light source module 1 and the first detector are in optimal coaxial alignment, infrared light propagates along the axis of the second chamber 503, and most of the light energy directly reaches the first detector, resulting in maximum received light intensity. When the angle of the light source module 1 shifts, some light is absorbed or scattered by the inner wall of the cylinder, reducing the received light intensity. The greater the shift angle, the more severe the light intensity loss. Since pure nitrogen gas does not absorb infrared light, the change in light intensity is entirely determined by the degree of optical path alignment. Therefore, the angle of the light source module 1 corresponding to the maximum light intensity on the curve is the optimal optical axis alignment position. By adjusting and locking the light source module 1 to the optimal position, the eccentricity error between the center of the optical path and the center of the flow field of the second chamber 503 caused by assembly tolerances or structural deformation can be eliminated, ensuring that the infrared light is always adjusted to pass through the area corresponding to the optimal detection angle position, thereby improving the accuracy of concentration measurement.

[0097] Optionally, the angle adjustment module 12 can adopt various structural forms to achieve the adjustment of the optical axis direction of the light source module 1, such as: a universal ball joint structure, where the light source module 1 is mounted on the universal ball joint to achieve tilting and locking in any direction; a dual-axis tilting stage, where the tilt angles of the X and Y axes are independently controlled by two orthogonal adjustment screws; a micro motor in conjunction with an eccentric cam or lead screw, where the light source module 1 is connected to the eccentric cam or lead screw, and the motor drives the eccentric cam or lead screw to actively adjust the angle of the light source module 1; an electromagnetic drive structure, where the current of the electromagnetic coil is controlled to generate magnetic force to tilt the base of the light source module 1, achieving non-contact adjustment; and a turntable structure, where the light source module 1 is mounted on a rotatable turntable, and the optical axis direction of the light source module 1 is changed by the rotation of the turntable, which, in conjunction with the tilt adjustment mechanism, can achieve spatial angle adjustment. The above solutions are all prior art and are clear to those skilled in the art, therefore they will not be described in detail.

[0098] In addition to the above solutions, any structure that can achieve the angle adjustment effect of the light source module 1 can be used as the angle adjustment module 12.

[0099] In one alternative implementation, such as Figure 7 As shown, the exhaust pipe 4 has a constriction section 403 and an expansion section 404. The constriction section 403 is disposed between the inlet end of the exhaust pipe 4 and the expansion section 404. The diameter of the end of the constriction section 403 facing the expansion section 404 is smaller than the diameter of the end of the expansion section 404 facing away from the constriction section 403. Both the constriction section 403 and the expansion section 404 can be tapered surfaces formed on the inner wall of the exhaust pipe 4.

[0100] The structure consisting of the contraction section 403 and the expansion section 404 can create a local negative pressure zone within the exhaust pipe 4, thereby accelerating gas discharge, reducing gas residence time, and increasing gas renewal speed. With the increased exhaust speed, it can effectively prevent backflow of external ambient gas into the exhaust pipe 4 due to diffusion or pressure fluctuations.

[0101] In one alternative implementation, such as Figure 8 As shown, the exhaust pipe 4 includes an inner pipe 401 and an outer pipe 402. One end of the inner pipe 401 is connected to the third cylinder 5. The contraction part 403 and the expansion part 404 are both located on the inner pipe 401. The expansion part 404 is located at the end of the inner pipe 401 facing away from the third cylinder 5. The outer pipe 402 is sleeved on the outer periphery of the other end of the inner pipe 401, and an annular gap 4021 is formed between the outer pipe 402 and the inner pipe 401. The outer pipe 402 is used to connect to an external inert gas source, so that the inert gas is ejected through the annular gap 4021 in the exhaust direction. The exhaust direction is the same as the gas flow direction in the inner pipe 401.

[0102] The protective gas ejected from the annular gap 4021 can form a laminar flow protective gas sleeve, which surrounds the gas jet discharged from the inner tube 401 from all sides, preventing external gas from backflowing into the inner tube 401 due to diffusion or pressure fluctuations. This protective gas can be a dry, inert sheath gas (such as nitrogen), and the ratio of its flow rate to the gas flow rate of the inner tube 401 can be controlled between 5:1 and 20:1 to surround the jet from the inner tube 401 and form a collimated sheath flow with aerodynamic focusing.

[0103] In one alternative embodiment, the temperature of the inert sheath gas can be controlled within a range of 10°C to 50°C above the gas temperature inside the gas chamber module, so as to form a stable thermal barrier around the jet of the inner tube 401.

[0104] In an optional embodiment, a self-cleaning system is also included for removing particulate matter adhering to the air chamber module. This system includes a high-pressure inert gas backflushing device and a piezoelectric ultrasonic transducer. The backflushing device is detachably connected to the inlet pipe 14 via a solenoid valve. It can introduce high-pressure pulsed inert gas into the air chamber module after each process cycle or periodically, using a high-speed airflow to flush the inner wall of the air chamber module and the exhaust port 501, thus blowing away the adhering particulate matter. The piezoelectric ultrasonic transducer is disposed on the outer wall of the first cylinder 2, the second cylinder 3, or the third cylinder 5, and is activated synchronously during the backflushing process. It uses high-frequency ultrasonic vibration to loosen and peel off the adhering particulate matter, enhancing the cleaning effect of the backflushing. The backflushing device and the ultrasonic transducer work together to effectively remove particulate matter accumulation within the air chamber module, prevent pore blockage and optical window contamination, and reduce the frequency of manual maintenance.

[0105] In one alternative implementation, combined with Figure 1 and Figure 2As shown, the variable diameter module 13 includes a movable component and a driving component. The movable component is disposed inside the exhaust pipe 4, and the driving component is used to drive the movable component to move, so as to change the flow area of ​​the exhaust pipe 4.

[0106] The actuator can employ a piezoelectric actuator or a shape memory alloy driven mechanical structure. The piezoelectric actuator utilizes the inverse piezoelectric effect to generate displacement, and its output characteristics are less affected by temperature, maintaining stable driving force and displacement accuracy within the normal operating temperature range of the exhaust pipe 4. While the shape memory alloy drive utilizes temperature changes to generate phase transition restoring force, its phase transition temperature can be designed and selected based on the actual operating temperature range of the exhaust pipe 4, ensuring reliable execution of the drive action within the stable temperature environment established by the temperature control module. Therefore, temperature changes in the exhaust pipe 4 will not interfere with the normal operation of the diameter-changing module 13. Compared to traditional solenoid valves or stepper motors, piezoelectric actuators and shape memory alloy drivers are smaller, consume less power, and have higher positioning accuracy, which is beneficial for the miniaturization and low-power design of the detection device. Simultaneously, neither drive method generates significant electromagnetic interference, avoiding influence on the signal of the first detector and improving the signal-to-noise ratio and stability of concentration measurement.

[0107] In one optional embodiment, a control unit is further included. The control unit is configured to receive a gas concentration signal output by the first detector and control the drive unit to change the flow area of ​​the exhaust pipe 4 according to a preset relationship. The preset relationship includes: when the gas concentration is higher than a first threshold, the flow area increases with the increase of the gas concentration; when the gas concentration is lower than a second threshold, the flow area decreases with the decrease of the gas concentration and the flow area is greater than zero; when the gas concentration is between the first threshold and the second threshold, the flow area is a constant value and the first threshold is greater than the second threshold.

[0108] When the gas concentration is higher than the first threshold, the flow area increases with the increase of concentration, which can accelerate the exhaust to improve the response speed and track the decreasing trend of gas concentration in a timely manner. When the gas concentration is lower than the second threshold, the flow area decreases with the decrease of concentration, which can reduce the exhaust speed to prolong the residence time of gas in the second chamber 503, enhance the infrared absorption signal, and improve the signal-to-noise ratio and accuracy of low concentration measurement.

[0109] In a specific embodiment, the variable diameter module 13 adjusts the flow area to a maximum range of 90% and a minimum range of 10% of the maximum flow area. During the initial concentration rise phase of the cleaning process, the control unit controls the variable diameter module 13 to increase the flow area, up to a maximum of 90% of the maximum flow area, to accelerate the discharge of background gas from the gas chamber module, enabling the detection device to respond quickly to concentration changes. When the gas concentration enters the stable range between the first and second thresholds, the flow area remains constant, allowing the detection device to achieve stable measurement performance during the stable phase. During the low-concentration phase near the end of the cleaning process, the control unit controls the variable diameter module 13 to decrease the flow area, down to a minimum of 10% of the maximum flow area, to extend the gas residence time in the second chamber 503, enabling the detection device to maintain high-precision detection even under low-concentration conditions and accurately capture the inflection point at the cleaning endpoint.

[0110] In an optional implementation, the control unit is further configured to perform feedforward-feedback composite control based on the theoretical gas concentration predicted by the prediction model and the measured concentration of the first detector, and control the action of the drive component to dynamically adjust the flow area of ​​the exhaust pipe 4; wherein, the prediction model is a model pre-established based on the law of gas concentration change over time in the cleaning process before the start of the cleaning process, and the prediction model is used to predict the theoretical gas concentration at different times during the operation of the cleaning process.

[0111] Feedforward control, based on a predictive model, anticipates gas concentration changes and adjusts the flow area before actual concentration changes occur. This overcomes the inherent lag in pure feedback control, allowing the detection device to respond promptly to rapidly decreasing concentration trends and preventing delays in cleaning progress assessment. Feedback control, on the other hand, corrects for deviations between the measured concentration and the predicted value from the first detector. This compensates for model prediction errors and random disturbances in the process, such as temperature fluctuations or deviations in gas composition and flow rate, thereby improving control accuracy. Even at low concentration levels with low signal-to-noise ratios, feedforward control maintains stable adjustment without relying on the measured signal. Through this combined feedforward and feedback control, the detection device actively compensates for deviations between the actual flow field and ideal simulation conditions, consistently approaching the optimal operating state throughout the cleaning process and accurately capturing the cleaning endpoint.

[0112] On the other hand, the present invention also provides a dynamic adjustment method for the infrared endpoint detection device of the previous embodiment, which includes the following steps: predicting the gas concentration at the current moment based on a prediction model; when the theoretical gas concentration predicted by the prediction model or the gas concentration measured by the first detector is higher than a first threshold, controlling the diameter-changing module 13 to increase the flow area of ​​the exhaust pipe 4; when the theoretical gas concentration predicted by the prediction model or the gas concentration measured by the first detector is lower than a second threshold, controlling the diameter-changing module 13 to reduce the flow area of ​​the exhaust pipe 4, and the flow area is greater than zero, and the first threshold is greater than the second threshold; through a PID control algorithm, independently adjusting and stabilizing the actual temperatures of the intake pipe 14, the first cylinder 2 and / or the second cylinder 3 and / or the third cylinder 5, and the exhaust pipe 4 to their respective target temperatures, so that a decreasing temperature gradient is formed along the gas flow direction, and each target temperature is higher than the condensation point of the detected process gas; wherein, the prediction model is an exponential decay model pre-established based on reaction kinetics theory before the start of the cleaning process, and its expression is: C(t)=C0·e -kt In the formula, C(t) is the predicted gas concentration at time t; C0 is the initial gas concentration at the start of the cleaning process; k is the reaction rate constant, which is predetermined based on the type of cleaning gas, the volume of the process chamber, the cleaning temperature, and the initial concentration of the substance being cleaned; and t is the time during which the cleaning process is carried out.

[0113] In the early stages of the cleaning process, the gas concentration is high and there may be residual background gas in the gas chamber module. At this time, the variable diameter module 13 increases the flow area of ​​the exhaust pipe 4 to accelerate the discharge of background gas from the gas chamber module and improve the gas renewal rate in the gas chamber module, enabling the detection device to respond quickly to concentration changes. In the middle stages of the cleaning process, the gas concentration gradually decreases. The variable diameter module 13 adopts a feedforward-feedback composite control strategy for dynamic adjustment: on the one hand, it predicts the concentration decrease trend in advance based on the prediction model and adjusts the flow area in advance to overcome the inherent lag of feedback control, so that the detection device can keep up with the concentration change rhythm in a timely manner; on the other hand, it performs feedback correction based on the deviation between the measured concentration and the predicted value of the first detector to compensate for model prediction errors and process disturbances such as temperature fluctuations and gas composition changes, thereby achieving a balance between response speed and control accuracy. Meanwhile, the segmented temperature control module uses a PID control algorithm to stabilize the temperatures of the inlet pipe 14, the gas chamber, and the exhaust pipe 4 at their respective preset target temperatures, forming a decreasing temperature gradient along the gas flow direction. This ensures that the gas remains in a gaseous state throughout the transmission process, preventing condensation and blockage of the exhaust port 501 or contamination of the optical window. The stable gas chamber temperature also keeps the infrared absorption coefficient consistent, providing stable environmental conditions for concentration measurement. As the cleaning process approaches its expected endpoint, the gas concentration drops to a low level. At this point, the flow area of ​​the exhaust pipe 4 is reduced, the gas flow rate is decreased, and the residence time of the gas in the second chamber 503 is extended. This allows the detection device to perform more photon absorption integrations on the low-concentration residual gas per unit time, thereby improving the detection signal-to-noise ratio and accurately capturing the inflection point of the cleaning endpoint, avoiding endpoint misjudgment due to insufficient signal-to-noise ratio. Through the synergistic effect of variable diameter adjustment and segmented temperature control, the detection device maintains optimal measurement performance throughout the entire cleaning process. Furthermore, in practical process applications, light attenuation can be monitored based on specific time intervals, process cycles, or real-time detection curve anomalies, avoiding inaccurate detection due to light attenuation baseline drift, reduced signal-to-noise ratio, etc., so that the detection device of this embodiment can continuously maintain a high-precision detection state.

[0114] In one possible implementation, the flow area adjustment step of the exhaust pipe 4 specifically includes: predicting the gas concentration at the current moment according to the prediction model, and calculating the feedforward flow area according to the preset relationship; comparing the concentration fed back by the first detector in real time with the predicted concentration of the prediction model, and calculating the area correction amount based on the deviation between the two; superimposing the feedforward flow area and the area correction amount to obtain the target flow area, and controlling the variable diameter module 13 to adjust the flow area of ​​the exhaust pipe 4 to the target flow area; determining whether the gas concentration predicted by the prediction model or measured by the first detector is lower than the preset endpoint threshold, if yes, the adjustment ends, otherwise returning to the calculation of the feedforward flow area and subsequent steps, until the gas concentration predicted by the prediction model or measured by the detector module is lower than the endpoint threshold; wherein, the prediction model is a model pre-established before the start of the cleaning process based on the law of gas concentration changing with time in the cleaning process; the preset relationship includes: when the gas concentration is higher than the first threshold, the flow area increases with the increase of gas concentration, when the gas concentration is lower than the second threshold, the flow area decreases with the decrease of gas concentration, and the flow area is greater than zero, the first threshold is greater than the second threshold, and the endpoint threshold is less than the second threshold.

[0115] Regarding the determination of the cleaning process endpoint, the percentage of the maximum falling edge value or the preset fixed endpoint concentration value method can be used to determine the endpoint of the cleaning process. Taking the process of cleaning silicon-containing residues with NF3 as an example, after cleaning the process chamber of the gas NF3 thin film growth equipment and exciting the plasma, the silicon-containing residues on the inner wall of the process chamber begin to react with the plasma active species to generate gaseous SiF4. As the reaction proceeds, the SiF4 generation rate gradually increases, and after being sampled and sent into the gas chamber module, the concentration rises rapidly from zero to a peak value. When the cleaning reaction enters the steady-state stage, the SiF4 generation rate and the NF3 supply rate reach a dynamic equilibrium. As the residues on the inner wall of the process chamber are gradually consumed, the SiF4 generation rate begins to decrease, and the concentration signal falls back from the peak value. The rate of signal decline in this decline stage directly reflects the progress of the cleaning reaction.

[0116] Once the residue on the inner wall of the process chamber is completely removed, the SiF4 formation rate approaches zero, and the concentration signal returns to a baseline level close to zero.

[0117] The specific judgment logic of the percentage method for the maximum value of the falling edge is as follows: The endpoint detection device monitors the SiF4 concentration signal in real time. When the signal reaches its highest value from the initial rising stage, the system automatically records the maximum concentration value at that moment, or locks a stable maximum value through a sliding window algorithm. The system can set a rising threshold to trigger the peak tracking function, for example, the signal rising amplitude exceeds 3 to 5 times the standard deviation of the background noise. Then, falling edge trigger tracking is performed. When the signal starts to fall continuously from the peak and the falling amplitude exceeds the preset hysteresis threshold, the system determines that it has entered the falling edge stage and starts the endpoint monitoring program. The starting point of the falling edge can be set to the moment when the signal falls from the peak to a certain percentage threshold, such as 95% of the peak value. After finding the maximum value, the system enters the real-time percentage calculation and endpoint determination stage. In the falling edge stage, the system continuously calculates the percentage of the current signal value relative to the peak value. When this percentage first falls to the preset threshold, such as 10%, 5%, or 3% of the peak value, the system determines that the cleaning process has reached the endpoint.

[0118] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A device for detecting the endpoint of a cleaning process in a thin film growth equipment, characterized in that, include: Light source module (1); The air chamber module includes: A first cylindrical body (2) is provided with an air inlet (201) on its side wall and a first wall-mounted flow guiding structure on its inner wall. An air inlet pipe (14) is connected to the air inlet (201). A second cylinder (3) is disposed inside the first cylinder (2) and coaxial with the first cylinder (2). The outer wall of the second cylinder (3) is provided with a second wall-mounted flow guiding structure, and the side wall of the second cylinder (3) is provided with a first through-hole (302) and an exhaust port (203). An exhaust pipe (4) is connected to the exhaust port (203), and one end of the exhaust pipe (4) facing away from the exhaust port (203) extends to the outside of the first cylinder (2). A third cylindrical body (5) is disposed inside the second cylindrical body (3) and is coaxial with the second cylindrical body (3). The wall of the third cylindrical body (5) is provided with a second opening (502) and a plurality of evenly distributed exhaust holes (501). A channel (6) is provided between the second opening (502) and the first opening (302). The first chamber (202), formed by the second cylinder (3) and the first cylinder (2), is used to prevent free jets and form a wall-attached circulation. The second chamber (503) is the inner cavity of the third cylinder (5), and is connected to the first chamber (202) through the channel (6) to form a stable detection area. The third chamber (301), formed by the third cylinder (5) and the second cylinder (3), is connected to the second chamber (503) through the exhaust port (501) and is used for uniform multi-point exhaust. A variable diameter module (13) is used to adjust the flow area of ​​the exhaust pipe (4); The segmented temperature control module includes an integrated partitioned heating jacket, a first PID controller, a second PID controller, and a linkage control module. The integrated partitioned heating jacket is fitted on the outside of the gas chamber module and is divided into a first temperature zone and a second temperature zone. The first temperature zone covers the area where the air inlet pipe (14) and the first cylinder (2) are located and is independently controlled by the first PID controller. The second temperature zone covers the exhaust pipe (4) and is independently controlled by the second PID controller. The linkage control module is connected to the first PID controller and the second PID controller respectively and is configured to coordinate the heating power of the first temperature zone and the second temperature zone based on the temperature information of the first temperature zone and the second temperature zone to maintain a preset target temperature difference, so that the temperature of the air inlet pipe (14), the first cylinder (2), the second cylinder (3), the third cylinder (5), and the exhaust pipe (4) are all higher than the condensation point of the detected process gas, and the temperature of the air inlet pipe (14) is the same as the temperature of the first cylinder (2), the second cylinder (3), and the third cylinder (5) and higher than the temperature of the exhaust pipe (4) to form a temperature gradient that decreases along the gas flow direction. The detector module (7) includes a first detector and a second detector. The first detector is used for real-time concentration signal detection, and the second detector is used for light attenuation monitoring of the light source module (1).

2. The endpoint detection device for preventing free jetting and entrainment according to claim 1, characterized in that, The inner wall of the channel (6) is arc-shaped, the cross-sectional area of ​​the middle part of the channel (6) is smaller than the cross-sectional area of ​​its two ends, and the cross-sectional area of ​​the channel (6) decreases continuously from the two ends to the middle.

3. The endpoint detection device for preventing free jet and entrainment as described in claim 1, characterized in that, The air inlet (201) and the exhaust outlet (203) are both located on the first side of the central axis of the first cylinder (2) and are located in the same radial position. The channel (6) is located on the second side of the central axis of the first cylinder (2). The first side and the second side are opposite sides of the first cylinder (2) in the radial direction.

4. The endpoint detection device for preventing free jetting and entrainment according to any one of claims 1 to 3, characterized in that, The first wall-mounted flow guiding structure includes a plurality of first arc-shaped protrusions (10), which are evenly distributed on the inner circumferential wall of the first cylinder (2); the second wall-mounted flow guiding structure includes a plurality of second arc-shaped protrusions (11), which are evenly distributed on the outer circumferential wall of the second cylinder (3).

5. The endpoint detection device for preventing free jet and entrainment as described in claim 4, characterized in that, The first arc-shaped protrusion (10) and the second arc-shaped protrusion (11) are alternately distributed in the circumferential and axial directions of the first chamber (202).

6. The endpoint detection device for preventing free jetting and entrainment according to claim 1, characterized in that, The exhaust port (501) is a frustum-shaped hole, and the diameter of the end of the exhaust port (501) facing the interior of the third cylinder (5) is smaller than that of the other end.

7. The endpoint detection device for preventing free jet and entrainment as described in claim 6, characterized in that, The ratio of the central diameter of the exhaust hole (501) to the depth of the exhaust hole (501) is in the range of 0.3 to 0.5, and the outward guide angle of the exhaust hole (501) is in the range of 15° to 30°.

8. The endpoint detection device for preventing free jet and entrainment as described in claim 1, characterized in that, The integrated partitioned heating jacket includes: The first heating sleeve is sleeved on the outside of the air inlet pipe (14) and the first cylinder (2). The first heating sleeve is provided with a first heating component and a first temperature measuring component. The first heating component and the first temperature measuring component are respectively connected to the first PID controller signal. The second heating sleeve is fitted on the outside of the exhaust pipe (4). The second heating sleeve is provided with a second heating component and a second temperature measuring component. The second heating component and the second temperature measuring component are respectively connected to the second PID controller signal.

9. The endpoint detection device for preventing free jet and entrainment as described in claim 8, characterized in that, The linkage control module is configured to coordinate the first PID controller and the second PID controller in a master-slave follow mode. In the master-slave follow mode, the second PID controller dynamically adjusts the heating power of the second heating component according to the measured temperature fed back by the first temperature measuring component, so that the temperature fed back by the second temperature measuring component is 10°C to 30°C lower than the temperature fed back by the first temperature measuring component.

10. The endpoint detection device for preventing free jetting and entrainment according to any one of claims 1 to 3, characterized in that, It also includes an angle adjustment module (12), the light source module (1) is connected to the power output end of the angle adjustment module (12), and the angle adjustment module (12) is used to adjust the optical axis direction of the light source module (1) to minimize the eccentricity error of the optical path and the flow field.

11. The endpoint detection device for preventing free jetting and entrainment according to any one of claims 1 to 3, characterized in that, The exhaust pipe (4) is provided with a constriction section (403) and an expansion section (404). The constriction section (403) is located between the inlet end of the exhaust pipe (4) and the expansion section (404). The diameter of the constriction section (403) facing the expansion section (404) is smaller than the diameter of the expansion section (404) away from the constriction section (403).

12. The endpoint detection device for preventing free jetting and entrainment according to claim 11, characterized in that, The exhaust pipe (4) includes an inner pipe (401) and an outer pipe (402). One end of the inner pipe (401) is connected to the third cylinder (5). The contraction part (403) and the expansion part (404) are both located on the inner pipe (401). The expansion part (404) is located at the end of the inner pipe (401) facing away from the third cylinder (5). The outer pipe (402) is sleeved on the outer periphery of the other end of the inner pipe (401), and an annular gap (4021) is formed between the outer pipe (402) and the inner pipe (401). The outer pipe (402) is used to connect to an external inert gas source, so that the inert gas is ejected through the annular gap (4021) in the exhaust direction. The exhaust direction is the same as the gas flow direction in the inner pipe (401).

13. The endpoint detection device for preventing free jetting and entrainment according to any one of claims 1 to 3, characterized in that, The variable diameter module (13) includes a movable component and a driving component. The movable component is disposed inside the exhaust pipe (4), and the driving component is used to drive the movable component to move so as to change the flow area of ​​the exhaust pipe (4).

14. The endpoint detection device for preventing free jet and entrainment as described in claim 13, characterized in that, It also includes a control unit, which is configured to receive the gas concentration signal output by the first detector and control the drive to change the flow area of ​​the exhaust pipe (4) according to a preset relationship. The preset relationships include: when the gas concentration is higher than the first threshold, the flow area increases with the increase of gas concentration; when the gas concentration is lower than the second threshold, the flow area decreases with the decrease of gas concentration, and the flow area is greater than zero; when the gas concentration is between the first threshold and the second threshold, the flow area is a constant value, and the first threshold is greater than the second threshold.

15. The endpoint detection device for preventing free jet and entrainment as described in claim 14, characterized in that, The control unit is further configured to perform feedforward-feedback composite control based on the theoretical gas concentration predicted by the prediction model and the measured concentration of the first detector, and control the action of the drive component to dynamically adjust the flow area of ​​the exhaust pipe (4); wherein, the prediction model is a model pre-established based on the law of gas concentration changing with time in the cleaning process before the start of the cleaning process, and the prediction model is used to predict the theoretical gas concentration at different times during the operation of the cleaning process.

16. A dynamic adjustment method for use in the endpoint detection device for preventing free jetting and entrainment as described in any one of claims 1 to 15, characterized in that, Includes the following steps: Based on the prediction model, the theoretical gas concentration at the current moment is predicted. When the theoretical gas concentration predicted by the prediction model or the gas concentration measured by the first detector is higher than the first threshold, the variable diameter module (13) is controlled to increase the flow area of ​​the exhaust pipe (4). When the theoretical gas concentration predicted by the prediction model or the gas concentration measured by the first detector is lower than the second threshold, the variable diameter module (13) is controlled to reduce the flow area of ​​the exhaust pipe (4), and the flow area is greater than zero. The first threshold is greater than the second threshold. The temperatures of the first and second temperature zones are controlled by a PID control algorithm. The first PID controller controls the temperature of the first temperature zone to a first target temperature, and the second PID controller controls the temperature of the second temperature zone to a second target temperature. The linkage control module adopts a master-slave follow mode, so that the second PID controller adjusts the second target temperature in real time according to the temperature change of the first temperature zone to maintain the temperature difference between the exhaust pipe (4), the intake pipe (14), and the first cylinder (2) as a set value. The temperatures of the intake pipe (14), the first cylinder (2), and the exhaust pipe (4) are all higher than the condensation point of the detected process gas. The prediction model is an exponential decay model pre-established based on reaction kinetics theory before the start of the cleaning process, and its expression is: C(t)=C0·e -kt , In the formula, C(t) is the predicted gas concentration at time t; C0 is the initial gas concentration at the start of the cleaning process; k is the reaction rate constant, which is predetermined based on the type of cleaning gas, the volume of the process chamber, the cleaning temperature, and the initial concentration of the substance being cleaned; and t is the time during which the cleaning process is carried out.

17. The dynamic adjustment method according to claim 16, characterized in that, The flow area adjustment steps of the exhaust pipe (4) specifically include: The gas concentration at the current moment is predicted based on the prediction model, and the feedforward flow area is calculated according to the preset relationship. The concentration fed back by the first detector in real time is compared with the concentration predicted by the prediction model, and the area correction amount is calculated based on the deviation between the two. The target flow area is obtained by superimposing the feedforward flow area and the area correction amount, and the variable diameter module (13) is controlled to adjust the flow area of ​​the exhaust pipe (4) to the target flow area. If the gas concentration predicted by the prediction model or measured by the first detector is lower than the preset endpoint threshold, the adjustment ends; otherwise, the process returns to calculate the feedforward flow area and subsequent steps until the gas concentration predicted by the prediction model or measured by the first detector is lower than the endpoint threshold. The preset relationships include: when the gas concentration is higher than the first threshold, the flow area increases with the increase of gas concentration; when the gas concentration is lower than the second threshold, the flow area decreases with the decrease of gas concentration, and the flow area is greater than zero; when the gas concentration is between the first threshold and the second threshold, the flow area is a constant value, the first threshold is greater than the second threshold, and the endpoint threshold is less than the second threshold.

Images