A precision coating uniformity intelligent regulation and detection system

Abstract

The application discloses a precision coating uniformity intelligent regulation and detection system, and relates to the technical field of non-contact thickness measurement.The system comprises a spectral confocal tomography detection module, a non-contact multi-wavelength full-width detection, a light flow field dynamic measurement module and the like.A data processing unit generates an evaluation model in combination with a model, and an actuator realizes high-precision regulation and control.The system highlights full-dimension detection and intelligent linkage.The application has the following advantages: a linear array sensor of the spectral confocal tomography detection module acquires multi-layer coating structure data in a full-width manner, a light flow field dynamic measurement module analyzes a slurry full-width flow velocity vector field, a data processing unit fuses multi-modal data to generate an evaluation model, a regulation strategy generation module outputs an instruction containing edge air pressure adjustment, precise coating uniformity control is realized, substrate damage is avoided, material waste and manual debugging time are reduced, and production efficiency and product yield are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of non-contact thickness measurement technology, specifically to a precision coating uniformity intelligent control and detection system. Background Technology

[0002] Precision coating technology, as a core process in the field of advanced manufacturing, is widely used in high-precision scenarios such as lithium battery electrode coating, semiconductor photoresist coating, optical thin film preparation, and electronic circuit board printing. Its coating uniformity directly determines product performance and yield.

[0003] Existing technologies have certain drawbacks. First, existing technologies use single-point or narrow-area detection, which cannot cover the entire area and is prone to missing overall uniformity and edge area anomalies. Second, traditional contact detection is prone to damaging flexible substrates and is limited in detecting transparent / reflective / multilayer materials, and the control lags behind real-time changes. To address these issues, we propose a precision coating uniformity intelligent control and detection system. Summary of the Invention

[0004] The purpose of this invention is to provide a precision coating uniformity intelligent control and detection system.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a precision coating uniformity intelligent control and detection system, the intelligent control and detection system comprising:

[0006] Spectroscopic confocal tomography detection module: used to emit multi-wavelength light beams and acquire layer thickness distribution and three-dimensional surface morphology data of multi-layer coating structures. The data includes thickness information of substrate layer, functional coating and interface layer, and is used for non-contact detection of the uniformity of each layer of transparent, reflective or multi-layer composite coating materials.

[0007] The optical flow field dynamic measurement module is used to acquire the velocity vector field data of the slurry surface using a dual-light source backlight system and a high-speed image acquisition device. The data includes the magnitude and direction of the velocity at each point on the slurry surface, and is used to analyze the dynamic characteristics of the slurry flow.

[0008] Data processing unit: electrically connected to the spectral confocal tomography detection module and the optical flow field dynamic measurement module respectively, used to receive layer thickness distribution data, three-dimensional surface morphology data and flow velocity vector field data, and generate a coating uniformity evaluation model containing flow field stability parameters based on the preset flow field-thickness correlation model;

[0009] Control strategy generation module: electrically connected to the data processing unit, used to generate control instructions for coating head position, slurry flow rate and drying parameters according to the coating uniformity evaluation model. The control instructions include independent adjustment instructions for the slurry supply pressure in the edge area.

[0010] Actuator control module: electrically connected to the regulation strategy generation module, used to convert regulation commands into control signals for the magnetic levitation coating head, piezoelectric ceramic slurry valve and drying unit, to realize micron-level adjustment of coating head position, high-frequency dynamic adjustment of slurry flow rate and real-time compensation of drying temperature;

[0011] Storage module: electrically connected to the data processing unit, used to store historical detection data, preset algorithm parameters and historical control strategies. The preset algorithm parameters include spectral reflectance thresholds, flow field stability criterion thresholds and control strategy priority rules corresponding to different material types.

[0012] Human-computer interaction module: electrically connected to the data processing unit and storage module, used to input process parameters, display real-time detection results and query historical data. The real-time detection results include a full-width thickness distribution heat map, a flow velocity vector field streamline map and defect location marking information.

[0013] As a further aspect of the present invention: the spectral confocal tomography detection module includes a linear array of spectral confocal sensors uniformly arranged along the coating width direction. The sensors emit light beams in the wavelength range of 400nm-1000nm and use the dispersive confocal principle to detect the layer thickness of the multilayer coating structure, separating the reflection spectral signals of the substrate layer, functional coating layer and interface layer.

[0014] As a further aspect of the present invention: the dual-source backlight system of the optical flow field dynamic measurement module includes a blue light source and an infrared light source, and the high-speed image acquisition device is a linear CCD camera with a frame rate of not less than 10kHz. Based on particle image velocimetry technology, the tracer particles in the slurry are tracked, and the flow velocity vector field on the slurry surface is calculated.

[0015] As a further aspect of the present invention: the flow field-thickness correlation model of the data processing unit generates the predicted value of coating thickness fluctuation through the following formula:

[0016] ;

[0017] in: For the future Predicted thickness fluctuation values ​​at any given time. For the current moment The shear rate gradient on the slurry surface reflects the flow field velocity at the current moment. The degree of drastic spatial change For the current moment The Reynolds number of the slurry flow, through calculate,( For the current moment Slurry density, For the current moment Average flow rate of slurry For the current moment Feature length, For the current moment (Slurry dynamic viscosity), used to quantify the current moment. The transition state from laminar to turbulent flow. For the current moment The surface tension of the slurry was detected by a confocal spectral chromatography module in conjunction with the current time. The interface reflectance spectral data is obtained through inversion and affects the current time. Uniformity of coating edge spreading For the past Historical values ​​of thickness fluctuation at time ( ), The historical data window length is used to introduce time series data. The memory effect of moments in time , , and To use a deep neural network for the current time step The adaptive weight coefficients obtained by training with previous historical data satisfy... + + + =1, where To be in relation to the current moment Sensitivity coefficient related to slurry temperature.

[0018] As a further aspect of the present invention: the control strategy generation module includes an edge compensation submodule, which is configured as follows:

[0019] Edge area definition: within 50mm on each of the left and right edges of the coating width;

[0020] Triggering condition: When the coating uniformity assessment model identifies that the standard deviation of the thickness in the edge region exceeds a preset threshold, the threshold is 0.5%-2%, which is dynamically adjusted according to the material type;

[0021] Adjustment method: Generate air pressure adjustment commands for the micro air pump array in the edge area, wherein the micro air pump array is arranged at equal intervals along the edge of the width (the interval does not exceed 10mm), and the air pressure adjustment range is 0-5kPa;

[0022] Compensation logic: Correct the slurry flow path through local airflow to ensure that the slurry flow velocity in the edge area deviates from that in the center area by no more than 5%.

[0023] As a further aspect of the present invention: the actuator control module includes a magnetic levitation coating head drive unit and a piezoelectric ceramic slurry valve drive unit.

[0024] The magnetic levitation coating head drive unit adopts electromagnetic levitation technology. Based on a closed-loop PID controller, it adjusts the strength of the levitation magnetic field and dynamically adjusts the position of the coating head. The position adjustment resolution of the coating head is not less than ±1μm, and the maximum moving speed is not less than 50mm / s.

[0025] The piezoelectric ceramic slurry valve drive unit has a response time of no more than 1μs, a flow rate adjustment range of 0-50mL / min, and an adjustment accuracy of no less than ±0.05mL / min. The piezoelectric ceramic slurry valve drive unit converts the control signal into the deformation displacement of the piezoelectric ceramic based on a high-voltage amplifier, thereby realizing high-frequency adjustment of the slurry flow rate at the kilohertz level.

[0026] The drying unit is an infrared heating array. The drying unit control signal includes power adjustment instructions for each heating unit (adjustment range 0-100%), which is used to compensate for local drying rate differences caused by slurry flow fluctuations.

[0027] As a further aspect of the present invention: the historical control strategies stored in the storage module include parameter configuration schemes for different coating processes (blade coating, slot coating, and inkjet printing), and support the rapid switching of different process types by calling historical schemes. The historical detection data is indexed by timestamps and stored in CSV and JSON formats, including thickness distribution data (unit: μm), flow velocity vector field data (unit: mm / s), and defect location coordinates (unit: mm).

[0028] As a further aspect of the present invention, the specific configuration of the human-computer interaction module is as follows:

[0029] Interface display content:

[0030] Real-time detection results: full-width thickness distribution heat map (color mapping range ±10%), flow velocity vector field streamline map (vector length represents flow velocity magnitude, color represents flow direction) and defect location marking (marking defects with diameter ≥0.5mm).

[0031] Key parameter dashboard: Current coating speed (unit: m / min), slurry temperature (unit: ℃), and standard deviation of edge area thickness (unit: %).

[0032] Hierarchical access control:

[0033] Administrator-level users have the highest privileges and can perform algorithm parameter calibration (such as adjusting the weight coefficients of the flow field-thickness correlation model), hardware system calibration (such as zero-point calibration of the spectral confocal sensor), and user permission allocation.

[0034] Engineer-level users: can view historical test data and log files (including system alarm records and control command execution records), and modify process parameters (such as the initial position of the coating head and the reference value of slurry flow rate).

[0035] Operator-level users: can only perform equipment start / stop operations and view real-time detection results, but do not have parameter modification permissions;

[0036] Interactive functions: Supports inputting process parameters via touch screen and keyboard. Input content includes, but is not limited to, coating width (unit: mm), target thickness (unit: μm) and slurry type (selected from drop-down menu). Input data is automatically synchronized to the data processing unit.

[0037] Compared with the prior art, the beneficial effects of the present invention by adopting the above technical solution are as follows:

[0038] 1. This invention acquires multi-layer coating structure data across the entire width of the system through a linear array sensor in a spectral confocal tomography detection module. Combined with a dynamic optical flow field measurement module, it analyzes the full-width velocity vector field of the slurry, solving the problems of missed detection of full-width and edge anomalies in existing single-point detection. The data processing unit integrates multimodal data to generate an evaluation model, and the control strategy generation module outputs instructions including edge air pressure adjustment to drive the magnetic levitation coating head, piezoelectric ceramic slurry valve, and drying unit to work together. The edge compensation submodule uses a micro air pump array to correct the slurry path. The entire system achieves precise control of coating uniformity through full-width detection, multi-data fusion, and intelligent linkage control, avoiding substrate damage, reducing material waste and manual debugging time, and significantly improving production efficiency and product yield.

[0039] 2. This invention utilizes non-contact multi-wavelength beam technology in a spectral confocal tomography detection module to achieve non-destructive testing of the full thickness and surface morphology of transparent, reflective, and multilayer composite coated materials. This solves the problem of traditional contact testing scratching flexible substrates. The optical flow field dynamic measurement module uses particle image velocimetry technology to capture the slurry flow velocity vector field. The data processing unit integrates fluid parameters and historical data through a flow field-thickness correlation model to proactively predict thickness fluctuation trends. The control strategy generation module drives actuators such as the magnetic levitation coating head and piezoelectric ceramic slurry valve to perform micron-level position adjustments and high-frequency flow control, while simultaneously compensating for drying parameters. The entire system, through non-contact full-layer detection, dynamic flow field modeling, and proactive control, overcomes the limitations of complex material detection and hysteresis adjustment bottlenecks, significantly improves the uniformity and stability of nanoscale coatings, reduces waste of high-value-added materials, shortens the debugging cycle of new processes, and enhances adaptability to precision scenarios such as semiconductors and solid-state batteries. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the system flow in an embodiment of the present invention;

[0041] Figure 2 This is a schematic diagram of the workflow of the data processing unit in an embodiment of the present invention. Detailed Implementation

[0042] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0043] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0044] Please see the appendix Figure 1 -Appendix Figure 2 This invention discloses a precision coating uniformity intelligent control and detection system, the intelligent control and detection system comprising:

[0045] Spectroscopic confocal tomography module: used to emit multi-wavelength beams and acquire layer thickness distribution and three-dimensional surface morphology data of multi-layer coating structures. The data includes thickness information of substrate layer, functional coating and interface layer, and is used for non-contact detection of the uniformity of each layer of transparent, reflective or multi-layer composite coating materials.

[0046] The optical flow field dynamic measurement module is used to acquire the velocity vector field data of the slurry surface using a dual-light source backlight system and high-speed image acquisition equipment. The data includes the magnitude and direction of the velocity at each point on the slurry surface, which is used to analyze the dynamic characteristics of the slurry flow.

[0047] Data processing unit: Electrically connected to the spectral confocal tomography detection module and the optical flow field dynamic measurement module respectively, used to receive layer thickness distribution data, three-dimensional surface morphology data and flow velocity vector field data, and generate a coating uniformity evaluation model containing flow field stability parameters based on the preset flow field-thickness correlation model.

[0048] Control strategy generation module: electrically connected to the data processing unit, used to generate control instructions for coating head position, slurry flow rate and drying parameters based on the coating uniformity evaluation model. The control instructions include independent adjustment instructions for the slurry supply pressure in the edge area.

[0049] Actuator control module: Electrically connected to the control strategy generation module, it is used to convert control commands into control signals for the magnetic levitation coating head, piezoelectric ceramic slurry valve and drying unit, so as to realize micron-level adjustment of coating head position, high-frequency dynamic adjustment of slurry flow rate and real-time compensation of drying temperature.

[0050] Storage module: Electrically connected to the data processing unit, used to store historical detection data, preset algorithm parameters and historical control strategies. The preset algorithm parameters include spectral reflectance thresholds, flow field stability criterion thresholds and control strategy priority rules corresponding to different material types.

[0051] Human-machine interaction module: electrically connected to the data processing unit and storage module, used to input process parameters, display real-time detection results and query historical data. Real-time detection results include full-width thickness distribution heat map, flow velocity vector field streamline map and defect location marking information.

[0052] In one embodiment of the present invention: the spectral confocal tomography detection module includes a linear array of spectral confocal sensors uniformly arranged along the coating width direction. The sensors emit light beams in the wavelength range of 400nm-1000nm and use the dispersive confocal principle to detect the layer thickness of the multilayer coating structure, separating the reflection spectral signals of the substrate layer, functional coating layer and interface layer.

[0053] In one embodiment of the present invention: the dual-source backlight system of the optical flow field dynamic measurement module includes a blue light source and an infrared light source, and the high-speed image acquisition device is a linear CCD camera with a frame rate of not less than 10kHz. Based on particle image velocimetry technology, the tracer particles in the slurry are tracked, and the flow velocity vector field on the surface of the slurry is calculated.

[0054] In one embodiment of the present invention: the flow field-thickness correlation model of the data processing unit generates the predicted value of coating thickness fluctuation using the following formula:

[0055] ;

[0056] in: For the future Predicted thickness fluctuation values ​​at any given time. For the current moment The shear rate gradient on the slurry surface reflects the flow field velocity at the current moment. The degree of drastic spatial change For the current moment The Reynolds number of the slurry flow, through calculate,( For the current moment Slurry density, For the current moment Average flow rate of slurry For the current moment Feature length, For the current moment (Slurry dynamic viscosity), used to quantify the current moment. The transition state from laminar to turbulent flow. For the current moment The surface tension of the slurry was detected by a confocal spectral chromatography module in conjunction with the current time. The interface reflectance spectral data is obtained through inversion and affects the current time. Uniformity of coating edge spreading For the past Historical values ​​of thickness fluctuation at time ( ), The historical data window length is used to introduce time series data. The memory effect of moments in time , , and To use a deep neural network for the current time step The adaptive weight coefficients obtained by training with previous historical data satisfy... + + + =1, where To be in relation to the current moment Sensitivity coefficient related to slurry temperature.

[0057] In one embodiment of the present invention: the control strategy generation module includes an edge compensation submodule, which is configured as follows:

[0058] Edge area definition: within 50mm on each of the left and right edges of the coating width;

[0059] Triggering condition: When the coating uniformity assessment model identifies that the standard deviation of the thickness in the edge region exceeds a preset threshold, the threshold is 0.5%-2%, which is dynamically adjusted according to the material type;

[0060] Adjustment method: Generate air pressure adjustment commands for the micro air pump array in the edge area. The micro air pump array is arranged at equal intervals along the edge of the width (the interval does not exceed 10mm), and the air pressure adjustment range is 0-5kPa.

[0061] Compensation logic: Correct the slurry flow path through local airflow to ensure that the slurry flow velocity in the edge area deviates from that in the center area by no more than 5%.

[0062] In one embodiment of the present invention: the actuator control module includes a magnetic levitation coating head drive unit and a piezoelectric ceramic slurry valve drive unit.

[0063] The magnetic levitation coating head drive unit adopts electromagnetic levitation technology. Based on a closed-loop PID controller, it adjusts the strength of the levitation magnetic field and dynamically adjusts the position of the coating head. The position adjustment resolution of the coating head is not less than ±1μm, and the maximum moving speed is not less than 50mm / s.

[0064] The piezoelectric ceramic slurry valve drive unit has a response time of no more than 1μs, a flow rate adjustment range of 0-50mL / min, and an adjustment accuracy of no less than ±0.05mL / min. The piezoelectric ceramic slurry valve drive unit converts the control signal into the deformation displacement of the piezoelectric ceramic based on a high-voltage amplifier, thereby realizing high-frequency adjustment of the slurry flow rate at the kilohertz level.

[0065] The drying unit is an infrared heating array. The control signal of the drying unit includes power adjustment instructions for each heating unit (adjustment range 0-100%), which is used to compensate for local drying rate differences caused by slurry flow fluctuations.

[0066] In one embodiment of the present invention: the historical control strategy stored in the storage module includes parameter configuration schemes for different coating processes (blade coating, slot coating and inkjet printing), and supports the rapid switching of different process types by calling historical schemes. The historical detection data is indexed by timestamps and stored in CSV and JSON formats, including thickness distribution data (unit: μm), flow velocity vector field data (unit: mm / s) and defect location coordinates (unit: mm).

[0067] In one embodiment of the present invention, the specific configuration of the human-computer interaction module is as follows:

[0068] Interface display content:

[0069] Real-time detection results: full-width thickness distribution heat map (color mapping range ±10%), flow velocity vector field streamline map (vector length represents flow velocity magnitude, color represents flow direction) and defect location marking (marking defects with diameter ≥0.5mm).

[0070] Key parameter dashboard: Current coating speed (unit: m / min), slurry temperature (unit: ℃), and standard deviation of edge area thickness (unit: %).

[0071] Hierarchical access control:

[0072] Administrator-level users have the highest privileges and can perform algorithm parameter calibration (such as adjusting the weight coefficients of the flow field-thickness correlation model), hardware system calibration (such as zero-point calibration of the spectral confocal sensor), and user permission allocation.

[0073] Engineer-level users: can view historical test data and log files (including system alarm records and control command execution records), and modify process parameters (such as the initial position of the coating head and the reference value of slurry flow rate).

[0074] Operator-level users: can only perform equipment start / stop operations and view real-time detection results, but do not have parameter modification permissions;

[0075] Interactive functions: Supports inputting process parameters via touch screen and keyboard. Input content includes, but is not limited to, coating width (unit: mm), target thickness (unit: μm) and slurry type (selected from drop-down menu). Input data is automatically synchronized to the data processing unit.

[0076] Example 1, please refer to the appendix. Figure 1 -Appendix Figure 2 Lithium-ion battery electrode three-layer composite structure coating scenario:

[0077] Application Background: This paper addresses the challenges of traditional testing methods for the three-layer composite coating of lithium battery electrodes, consisting of a copper foil substrate (8μm), an active material layer (100μm), and a carbon coating (5μm). These methods aim to solve the problems of uneven edge thickness (error ±8%) and surface density fluctuations caused by turbulent slurry flow.

[0078] System configuration and parameters

[0079] Spectroscopic confocal chromatography detection module

[0080] Sensor layout: 24,000 linear array spectral confocal sensors (50μm spacing) are arranged at equal intervals along the coating width (1200mm), emitting dual-wavelength beams of 450nm (blue light) and 850nm (infrared light);

[0081] Detection parameters:

[0082] The thickness detection resolution for copper foil substrate is ±50nm, for active material layer ±1μm, and for carbon coating ±200nm.

[0083] Real-time output of full-width thickness distribution data, with a lateral detection interval of 50μm and a longitudinal detection frequency of 10kHz (synchronized with the substrate running speed of 60m / min).

[0084] Optical flow field dynamic measurement module

[0085] Light source and camera: Blue light source (wavelength 450nm, power 100mW) + infrared backlight (wavelength 850nm, power 200mW), linear CCD camera with frame rate of 10kHz and resolution of 2048×1 pixel;

[0086] Tracer particles: 2μm fluorescent microspheres (density 1.2g / cm³) were uniformly mixed in the slurry, and the flow velocity vector field was calculated using PIV technology with an accuracy of ±0.05mm / s;

[0087] Feature extraction: Real-time calculation of shear rate gradient (Threshold setting:) >500s -1 Triggering turbulence warning), Reynolds number (critical value) =2000).

[0088] Data processing and regulation

[0089] Flow field-thickness correlation model:

[0090] Initial values ​​of weighting coefficients: =0.4 (shear rate) =0.3 (Reynolds number) =0.2 (surface tension) =0.1 (historical data);

[0091] An LSTM network was trained using historical lithium battery slurry data (5000 samples), and the weight coefficients were dynamically adjusted to achieve a prediction error of ≤±0.5%.

[0092] Actuator parameters:

[0093] Magnetic levitation coating head: Position adjustment resolution ±0.5μm, the gap in the edge area is 1μm larger than that in the center to compensate for slurry accumulation;

[0094] Piezoelectric ceramic slurry valve: response time 0.8μs, flow rate adjustment accuracy ±0.03mL / min, slurry flow rate at the edge is 3% higher than at the center;

[0095] Infrared drying unit: heating power in the edge area is increased by 15% to compensate for differences in solvent evaporation.

[0096] Human-computer interaction and storage

[0097] Parameter input: Set the target thickness (copper foil + active material + carbon coating = 113±1μm) and slurry type (NCM811 cathode slurry) via the touch screen.

[0098] Real-time display: In the full-width thickness heat map, the color threshold for the 50mm edge area is set to ±1.5% (±1% for the center). If the limit is exceeded, an automatic red alarm will be triggered.

[0099] Workflow

[0100] Testing phase:

[0101] A spectral confocal sensor simultaneously acquires the reflection spectrum of the three-layer structure, separating the thickness data of the copper foil (60% reflectivity), the active material (30% reflectivity), and the carbon coating (10% reflectivity).

[0102] The optical flow field module detected the coating head outlet. =1800 (laminar flow), but in the edge region ∇u=600s -1 (Above the threshold), the predicted edge thickness will exceed the upper limit.

[0103] Regulation phase:

[0104] The control strategy generation module triggers the edge compensation submodule, which sends a 0.5 kPa air pressure command to 120 micro air pumps (spaced 10 mm) in the edge area to form a lateral airflow to suppress slurry overflow;

[0105] The piezoelectric ceramic valve simultaneously increases the edge slurry flow rate by 2%, and reduces the edge gap of the magnetic levitation coating head by 0.8μm.

[0106] Example 2, please refer to the appendix. Figure 1 -Appendix Figure 2 Nanoscale coating scenario for semiconductor EUV photoresist:

[0107] Application Background

[0108] For EUV photoresist coating (thickness 150±1.5nm) in 7nm node semiconductor chip manufacturing, this invention solves the problems of traditional laser thickness measurement not being able to penetrate the transparent adhesive layer, uneven distribution of nanoparticles (particle size deviation ±10%), and abrupt changes in edge thickness caused by turbulence.

[0109] System configuration and parameters

[0110] Spectroscopic confocal chromatography detection module

[0111] Sensor layout: 6000 linear array spectral confocal sensors (50μm spacing) are arranged along the coating width (300mm) to emit dual-wavelength beams of 405nm (ultraviolet light) and 980nm (near-infrared light);

[0112] Detection parameters:

[0113] The detection resolution for the SiO2 insulating layer (200nm thickness) is ±20nm, and for the photoresist layer it is ±5nm.

[0114] Real-time generation of 3D surface topography point cloud maps to identify particle defects with a height of ≥50nm.

[0115] Optical flow field dynamic measurement module

[0116] Light source and camera: ultraviolet light source (wavelength 365nm, power 50mW) + infrared backlight (wavelength 940nm, power 150mW), linear CCD camera with frame rate of 20kHz and resolution of 4096×1 pixel;

[0117] Tracer particles: 1 μm fluorescent nanospheres (density 1.5 g / cm³) were added to the photoresist, and the flow velocity vector field detection accuracy was ±0.02 mm / s;

[0118] Feature extraction: Focus on monitoring flow velocity vector deviation (threshold ±3%) and surface tension in edge regions. (Critical value 65mN / m).

[0119] Data processing and regulation

[0120] Flow field-thickness correlation model:

[0121] Dynamic adjustment of weighting coefficients: during photoresist coating (Reynolds number weight) is automatically reduced to 0.1 (low Reynolds number scenario). (Surface tension weight) increased to 0.4;

[0122] Historical data window length =10, predicting thickness fluctuations in the next 5ms, with a root mean square error ≤±0.3%.

[0123] Actuator parameters:

[0124] Magnetic levitation coating head: Position adjustment resolution ±0.2μm, adopts "edge pre-compression" mode (edge ​​gap is 0.5μm smaller than center);

[0125] Piezoelectric ceramic slurry valve: response time 0.5μs, flow rate adjustment accuracy ±0.01mL / min, pulse frequency 2000Hz;

[0126] Cleanroom-specific drying unit: temperature control accuracy ±0.1℃, and 20% increased airflow velocity in edge areas to reduce solvent residue.

[0127] Human-computer interaction and storage

[0128] Parameter input: Set the photoresist type (chemically amplified CAR photoresist), target thickness 150nm, and cleanliness level ISO3 through the hierarchical permission interface (administrator level);

[0129] Real-time display: When the difference in vector length in the edge region of the velocity vector field streamline diagram is greater than 5%, a red warning is automatically triggered.

[0130] Workflow

[0131] Testing phase:

[0132] The spectral confocal sensor detected an average thickness of 150.2 nm across the entire photoresist layer, but there were 3 particle defects with a height of ≥80 nm in the 50 mm edge region.

[0133] The optical flow field module detected an edge velocity vector deviation of 4.5% (exceeding the threshold of 3%), predicting that the edge thickness will be 2% thinner.

[0134] Regulation phase:

[0135] The control strategy generation module initiates the "edge flow rate compensation" command, and the edge area of ​​the magnetic levitation coating head moves 0.3μm towards the substrate. The piezoelectric ceramic valve increases the supply of edge slurry in a pulsed manner at a frequency of 2000Hz.

[0136] Simultaneously trigger the defect repair procedure: reapply photoresist at the defect location of the particle with an accuracy of ±100nm.

[0137] As described in the embodiments, the system acquires data across the entire area through the spectral confocal tomography detection module and analyzes the flow velocity vector field through the optical flow field dynamic measurement module. After data processing and control strategies are implemented, the system drives the actuators to coordinate their actions. This solves problems such as missed detections and damage to the substrate, improves coating uniformity and efficiency, and is suitable for precision scenarios.

[0138] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Any variations and modifications can be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention, without departing from the scope of the invention, fall within the protection scope defined by the claims of the present invention.

Claims

1. A precision coating uniformity intelligent regulation and detection system, comprising an intelligent regulation and detection system, characterized in that, The intelligent regulation and detection system comprises the following modules: A spectral confocal tomography detection module: for emitting a multi-wavelength light beam and acquiring a multi-layer coating structure's layered thickness distribution and three-dimensional surface topography data, the data containing the thickness information of the substrate layer, the functional coating layer and the interface layer; A light flow field dynamic measurement module: for acquiring the flow velocity vector field data of the slurry surface by using a double light source backlight system and a high-speed image acquisition device, the data containing the flow velocity size and direction of each point on the slurry surface; A data processing unit: for receiving the layered thickness distribution data, the three-dimensional surface topography data and the flow velocity vector field data, and generating a coating uniformity evaluation model containing a flow field stability parameter based on a preset flow field-thickness correlation model; A regulation strategy generation module: for generating regulation instructions for the coating head position, the slurry flow and the drying parameters according to the coating uniformity evaluation model; An actuator control module: for converting the regulation instructions into control signals for the magnetic suspension coating head, the piezoelectric ceramic slurry valve and the drying unit; A storage module: for storing historical detection data, preset algorithm parameters and historical regulation strategies; A human-computer interaction module: for inputting process parameters, displaying real-time detection results and querying historical data; The flow field-thickness correlation model of the data processing unit generates a coating thickness fluctuation prediction value through the following formula: ； wherein: is the future thickness fluctuation prediction value at the current time instant, is the current time instant, shear rate gradient at the surface of the slurry, is the current time instant, Reynolds number of the slurry flow, is the current time instant, surface tension of the slurry, is the past thickness fluctuation history value at the current time instant, , is the history data window length for introducing the memory effect of the time series before the current time instant, , , , and are adaptive weight coefficients trained by a deep neural network for the history data before the current time instant , satisfying + + + = 1, where is a sensitive coefficient related to the current time instant slurry temperature.

2. The precision coating uniformity intelligent regulation and detection system according to claim 1, characterized in that: The spectral confocal tomography detection module comprises linear array spectral confocal sensors arranged uniformly along the coating width direction, the sensors emit light beams in the wavelength range of 400 nm-1000 nm, and use the dispersion confocal principle to detect the layered thickness of the multi-layer coating structure, and separate the reflected spectrum signals of the substrate layer, the functional coating layer and the interface layer.

3. The precision coating uniformity intelligent regulation and detection system according to claim 1, characterized in that: The double light source backlight system of the light flow field dynamic measurement module comprises a blue light source and an infrared light source, the high-speed image acquisition device is a linear array CCD camera with a frame frequency not less than 10 kHz, and the flow velocity vector field of the slurry surface is calculated based on the particle image velocimetry technology to track the tracer particles in the slurry.

4. The precision coating uniformity intelligent regulation and detection system according to claim 1, characterized in that: The regulation strategy generation module comprises an edge compensation submodule, which is configured as follows: Edge area definition: 50 mm range on the left and right edges of the coating width; Trigger condition: when the coating uniformity evaluation model identifies that the thickness standard deviation of the edge area exceeds the preset threshold value, the threshold value is 0.5%-2%; Adjustment mode: generate air pressure adjustment instructions for the micro air pump array along the width edge, the micro air pump array is arranged at equal intervals along the width edge, and the air pressure adjustment range is 0-5 kPa; Compensation logic: correct the slurry flow path by local air flow to make the deviation of the slurry flow velocity in the edge area from that in the center area not more than 5%.

5. The precision coating uniformity intelligent regulation and detection system according to claim 1, characterized in that: The actuator control module comprises a magnetic suspension coating head driving unit and a piezoelectric ceramic slurry valve driving unit: The magnetic suspension coating head driving unit adopts electromagnetic suspension technology, adjusts the suspension magnetic field strength based on a closed-loop PID controller, dynamically adjusts the position of the coating head, and the position adjustment resolution of the coating head is not less than ±1 μm, and the maximum moving speed is not less than 50 mm / s; The response time of the piezoelectric ceramic slurry valve of the piezoelectric ceramic slurry valve driving unit is not more than 1 μs, the flow regulation range is 0-50 mL / min, and the regulation accuracy is not less than ± 0.05 mL / min; The drying unit is an infrared heating array, and the drying unit control signal comprises power regulation instructions for each heating unit.

6. The precision coating uniformity intelligent regulation and detection system of claim 1, wherein: The historical regulation strategy stored in the storage module comprises parameter configuration schemes for different coating processes, supports rapid switching of different process types by calling historical schemes, historical detection data is indexed by time stamp, stored in CSV and JSON formats, and comprises thickness distribution data, flow velocity vector field data and defect position coordinates.

7. The precision coating uniformity intelligent regulation and detection system of claim 1, wherein The specific configuration of the human-computer interaction module is as follows: Interface display content: Real-time detection results: full-width thickness distribution thermal map, flow velocity vector field streamline diagram and defect position marking; Key parameter instrument panel: current coating speed, slurry temperature and edge area thickness standard deviation; Permission hierarchical management: Administrator-level user: has the highest permission, can calibrate algorithm parameters, calibrate hardware system and assign user permissions; Engineer-level user: can view historical detection data and log files, and modify process parameters; Operator-level user: can only start / stop the device, view real-time detection results, and has no parameter modification permission; Interaction function: support inputting process parameters through touch screen and keyboard, input content includes but is not limited to coating width, target thickness and slurry type, and input data is automatically synchronized to the data processing unit.

Images