Edge searching machine for semiconductor processing and edge searching method
By combining area scan cameras and line scan cameras for edge finding, along with data processing and control modules, the problem of balancing positioning accuracy and efficiency in PLP substrate edge finding equipment has been solved. This achieves high-precision, low-cost substrate positioning, meeting the high cleanliness requirements of semiconductor manufacturing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-06
- Publication Date
- 2026-04-07
AI Technical Summary
Existing PLP substrate edge-finding equipment suffers from problems such as poor positioning accuracy and efficiency, low stability, and high cost.
A positioning method combining area scan and line scan cameras is adopted. Combined with data processing and control modules, high-precision alignment of the substrate is achieved through preliminary edge finding algorithm and high-precision edge fitting algorithm. Vacuum adsorption and high-rigidity base structure are used to ensure substrate stability and optimize the heat dissipation and dust prevention performance of the equipment.
It achieves a significant reduction in hardware costs while ensuring accuracy, balances speed and precision, improves equipment stability and positioning accuracy, and meets the cleanliness requirements of semiconductor manufacturing.
Smart Images
Figure CN121816003A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of semiconductor manufacturing equipment technology, and in particular to an edge finding machine and edge finding method for semiconductor processing. Background Technology
[0002] In fields such as semiconductor packaging and display panel manufacturing, the edge positioning accuracy of PLP substrates directly affects the processing quality of subsequent processes such as bonding, etching, and cutting, and is a key link in ensuring product yield. As electronic devices develop towards higher precision and miniaturization, the accuracy requirements for edge positioning of PLP substrates have increased to the micrometer level, while simultaneously balancing production efficiency and equipment costs.
[0003] Existing PLP substrate edge-finding equipment suffers from the following technical defects: First, most equipment uses a single-vision imaging system. While a high-resolution area array camera can meet the accuracy requirements, the hardware cost is high, and the field of view is limited in a single shot, resulting in low positioning efficiency. If a standard-resolution camera is used, it cannot achieve micron-level positioning accuracy, making it unsuitable for high-precision processing needs. Second, some equipment uses a structure design where the camera moves while the substrate is fixed. Vibration during camera movement can cause image blurring, affecting positioning accuracy. Furthermore, the moving cables are prone to damage and tangling, reducing the long-term stability of the equipment. Third, existing equipment lacks adequate heat dissipation, dustproofing, and cleanroom protection. Exposed cables and non-standard pipe connections easily generate dust or contaminants, failing to meet the cleanliness requirements of the semiconductor and packaging industries. Fourth, the frame structure of some edge-finding equipment is made of a single material, resulting in either insufficient rigidity, leading to deformation during movement, or excessive weight, hindering installation and handling. Additionally, the adsorption stability of the load-bearing unit is poor, causing substrate displacement during rotation, further affecting positioning accuracy. Summary of the Invention
[0004] The purpose of this invention is to provide an edge finding machine and method for semiconductor processing, which aims to solve the technical problems of existing PLP substrate edge finding equipment, such as the inability to balance positioning accuracy and efficiency, poor stability, and high cost.
[0005] To achieve the above objectives, in a first aspect, the present invention provides an edge finding machine for semiconductor processing, comprising a base, a fixing component, a rotating component, an adjusting component, and a shooting and positioning component. The adjusting component is disposed on the base, the rotating component is disposed on the adjusting component, the adjusting component is used to drive the rotating component to move, and the fixing component is disposed on the rotating component for adsorbing a substrate. The imaging and positioning component includes a line scan camera and an area scan camera, a data processing module and a control module. The area scan camera captures partial images of the substrate. The data processing module acquires a local image of the substrate and runs a preliminary edge-finding algorithm to calculate the current angular deviation value of the substrate. The control module is used to control the rotation of the rotating component based on the angle deviation value, so that the substrate rotates in the opposite direction to complete one alignment. After one alignment, the adjustment component is used to drive the rotating component and the substrate to move, and the line scan camera synchronously and continuously acquires substrate images and stitches them into substrate edge images. The data processing module is also used to calculate the residual angular deviation and center offset of the substrate edge based on the substrate edge image. The control module is also used to control the rotation component and the adjustment component to make adjustments based on the residual angular deviation and the center offset.
[0006] The base includes a base body, a housing, a cover plate, and a camera support rod. The housing is fixed to the base body, the cover plate is fixed to the top of the housing, and the camera support rod is fixed to one side of the housing for mounting the line scan camera and the area scan camera.
[0007] The adjustment component includes an X-axis moving component and a Y-axis moving component. The X-axis moving component is used to drive the rotating component to move along the X direction, and the Y-axis moving component is disposed on the X-axis moving component to drive the rotating component to move along the Y direction.
[0008] The X-axis moving component includes a servo motor, a gearbox, a lead screw module, and an X-axis moving block. The X-axis moving block is slidably mounted on the base. The lead screw module is threadedly connected to the X-axis moving block. The gearbox is connected to the lead screw module. The output end of the servo motor is connected to the gearbox.
[0009] The rotating assembly includes a rotating motor and a coupling. The rotating motor is fixed on the Y-axis moving assembly, and the coupling is connected to the output end of the rotating motor and the fixed assembly.
[0010] The fixing component includes a vacuum suction cup, a support base, a negative pressure valve, and a baffle. The support base is fixed to the coupling, the vacuum suction cup is disposed on the support base, the negative pressure valve is connected to the vacuum suction cup, and the baffle is disposed on the support base.
[0011] Secondly, the present invention also provides an edge-finding method for semiconductor processing, comprising: Take partial images of the substrate; Acquire a local image of the substrate and run a preliminary edge-finding algorithm to calculate the current angular deviation value of the substrate; The rotation of the rotating component is controlled by the angle deviation value, which drives the substrate to rotate in the opposite direction to complete one alignment. After one alignment, the adjustment component is used to drive the rotating component and the substrate to move, and the line scan camera synchronously and continuously acquires substrate images and stitches them into substrate edge images. The residual angular deviation and center offset of the substrate edge are calculated based on the substrate edge image. The control rotation component and adjustment component are adjusted based on the residual angular deviation and center offset.
[0012] The method further includes, after the control rotation component and adjustment component are adjusted based on the residual angle deviation and center offset, the system records the final pose data after positioning is completed and sends it to the downstream device.
[0013] This invention relates to an edge-finding machine and method for semiconductor processing. The adjustment component 104 is mounted on the base 101, serving as the motion platform for the entire system. It has two-dimensional translation capabilities (typically in the X and Y axes) and is used to drive the structure above it to perform precise positioning and movement. The rotation component 103 is disposed on the adjustment component 104 and is supported and driven by the adjustment component 104. It can be adjusted in position in the horizontal plane along with the adjustment component 104 as a whole. The fixing component 102 is integrated on the top of the rotation component 103 and is used to reliably adsorb and fix the semiconductor substrate (such as silicon wafers, glass substrates, etc.) to be processed. It typically uses vacuum adsorption to ensure that the substrate remains stable and does not slip during rotation and movement.
[0014] The area scan camera 107 is used to acquire high-resolution images of a local area of the substrate in the initial stage, and obtain local image information containing the edge features of the substrate. After receiving the image, the data processing module 108 runs a preset preliminary edge finding algorithm (such as image processing techniques based on edge detection, Hough transform or template matching) to quickly calculate the angular deviation value of the current substrate relative to the ideal coordinate system.
[0015] The control module 109 generates corresponding control commands based on the angular deviation value, driving the rotating component 103 to rotate the substrate in the opposite direction, completing the first coarse alignment operation. Based on this, the adjustment component 104 is activated, causing the entire rotating component 103, along with the fixed substrate, to move smoothly along a predetermined path (usually a straight scanning trajectory). Simultaneously, the line scan camera 106 continuously acquires strip-shaped images of the substrate edge, and seamlessly stitches multiple frames into a complete substrate edge contour image using an image stitching algorithm.
[0016] The data processing module 108 further calculates the residual angular deviation of the substrate edge and the offset of the center position relative to the target coordinates (i.e., the center offset) based on the stitched edge image using a high-precision edge fitting algorithm (such as least squares fitting of a circle or rectangle). Finally, the control module 109 combines the residual angular deviation and the center offset and sends fine adjustment commands to the rotation component 103 and the adjustment component 104 respectively: the rotation component 103 performs a small angle correction to eliminate the residual angular error, and the adjustment component 104 performs displacement compensation in the X / Y directions to correct the center position, thereby achieving high-precision secondary alignment of the substrate.
[0017] Replacing multiple high-resolution area scan cameras 107 with a single area scan camera 107 and a line scan camera 106 significantly reduces hardware costs while maintaining final accuracy. The area scan camera 107 enables millisecond-level rapid coarse adjustment, greatly reducing the deviation range; the line scan camera 106 obtains extremely high-resolution global images through motion scanning, achieving micrometer-level fine adjustment. The two work together to balance speed and accuracy in a single positioning operation. Attached Figure Description
[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 This is a structural diagram of an edge-finding machine for semiconductor processing according to the present invention.
[0020] Figure 2 This is a first cross-sectional view of an edge-finding machine for semiconductor processing according to the present invention.
[0021] Figure 3 This is a second cross-sectional view of an edge-finding machine for semiconductor processing according to the present invention.
[0022] Figure 4 This is a top structural diagram of an edge-finding machine for semiconductor processing according to the present invention.
[0023] Figure 5 This is a structural diagram of the imaging and positioning component of the present invention.
[0024] Figure 6 This is a flowchart of an edge-finding method for semiconductor processing according to the present invention.
[0025] Base 101, fixing component 102, rotating component 103, adjusting component 104, shooting and positioning component 105, line scan camera 106, area scan camera 107, data processing module 108, control module 109, base body 110, outer shell 111, cover plate 112, camera support rod 113, X-axis moving component 114, Y-axis moving component 115, servo motor 116, gearbox 117, lead screw module 118, X-axis moving block 119, rotating motor 120, coupling 121, vacuum suction cup 122, support base 123, negative pressure valve 124, baffle 125. Detailed Implementation
[0026] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.
[0027] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0028] First Embodiment Please see Figures 1-5 This invention provides an edge-finding machine for semiconductor processing, including a base 101, a fixing component 102, a rotating component 103, an adjusting component 104, and an image positioning component 105. The adjusting component 104 is disposed on the base 101, and the rotating component 103 is disposed on the adjusting component 104. The adjusting component 104 is used to drive the rotating component 103 to move. The fixing component 102 is disposed on the rotating component 103 and is used to adsorb a substrate. The image positioning component 105 includes a line scan camera 106 and an area scan camera 107, a data processing module 108, and a control module 109. The area scan camera 107 captures a partial image of the substrate. The data processing module 108 acquires the partial image of the substrate and runs a preliminary edge-finding algorithm to calculate the current angular deviation value of the substrate. The control module 109 is used to control the rotation component 103 to rotate based on the angle deviation value, thereby causing the substrate to rotate in the opposite direction to complete one alignment. After one alignment, the adjustment component 104 is used to drive the rotation component 103 and the substrate to move. The line scan camera 106 synchronously and continuously acquires substrate images and stitches them into a substrate edge image. The data processing module 108 is also used to calculate the residual angle deviation and center offset of the substrate edge based on the substrate edge image. The control module 109 is also used to control the rotation component 103 and the adjustment component 104 to adjust based on the residual angle deviation and center offset.
[0029] The adjustment component 104 is mounted on the base 101, serving as the motion platform for the entire system. It has two-dimensional translation capabilities (typically in the X and Y axes) and is used to drive the structure above it to perform precise positioning and movement. The rotation component 103 is disposed on the adjustment component 104 and is supported and driven by the adjustment component 104. It can be adjusted in position in the horizontal plane along with the adjustment component 104 as a whole. The fixing component 102 is integrated on the top of the rotation component 103 and is used to reliably adsorb and fix the semiconductor substrate (such as silicon wafers, glass substrates, etc.) to be processed. It typically uses vacuum adsorption to ensure that the substrate remains stable and does not slip during rotation and movement.
[0030] The area scan camera 107 is used to acquire high-resolution images of a local area of the substrate in the initial stage, and obtain local image information containing the edge features of the substrate. After receiving the image, the data processing module 108 runs a preset preliminary edge finding algorithm (such as image processing techniques based on edge detection, Hough transform or template matching) to quickly calculate the angular deviation value of the current substrate relative to the ideal coordinate system.
[0031] The control module 109 generates corresponding control commands based on the angular deviation value, driving the rotating component 103 to rotate the substrate in the opposite direction, completing the first coarse alignment operation. Based on this, the adjustment component 104 is activated, causing the entire rotating component 103, along with the fixed substrate, to move smoothly along a predetermined path (usually a straight scanning trajectory). Simultaneously, the line scan camera 106 continuously acquires strip-shaped images of the substrate edge, and seamlessly stitches multiple frames into a complete substrate edge contour image using an image stitching algorithm.
[0032] The data processing module 108 further calculates the residual angular deviation of the substrate edge and the offset of the center position relative to the target coordinates (i.e., the center offset) based on the stitched edge image using a high-precision edge fitting algorithm (such as least squares fitting of a circle or rectangle). Finally, the control module 109 combines the residual angular deviation and the center offset and sends fine adjustment commands to the rotation component 103 and the adjustment component 104 respectively: the rotation component 103 performs a small angle correction to eliminate the residual angular error, and the adjustment component 104 performs displacement compensation in the X / Y directions to correct the center position, thereby achieving high-precision secondary alignment of the substrate.
[0033] Replacing multiple high-resolution area scan cameras 107 with a single area scan camera 107 and a line scan camera 106 significantly reduces hardware costs while maintaining final accuracy. The area scan camera 107 enables millisecond-level rapid coarse adjustment, greatly reducing the deviation range; the line scan camera 106 obtains extremely high-resolution global images through motion scanning, achieving micrometer-level fine adjustment. The two work together to balance speed and accuracy in a single positioning operation.
[0034] The base 101 includes a base body 110, a housing 111, a cover plate 112, and a camera support rod 113. The housing 111 is fixed on the base body 110, the cover plate 112 is fixed on the top of the housing 111, and the camera support rod 113 is fixed on one side of the housing 111 for mounting the line scan camera 106 and the area scan camera 107.
[0035] The base body 110 is made of high-rigidity material, possessing excellent mechanical strength and dimensional stability, effectively suppressing the impact of external vibrations and temperature changes on equipment accuracy. The base body 110 is equipped with multiple mounting holes and guide rail interfaces for fixing the adjustment assembly 104 and other moving mechanisms, ensuring their positional consistency during long-term operation.
[0036] The outer casing 111 is fixedly mounted on the base body 110 and is typically made of sheet metal or engineering plastics. It serves to protect internal precision mechanisms (such as guide rails, lead screws, and motors) from dust, moisture, and external interference, while also providing excellent electromagnetic shielding to prevent external electromagnetic noise from interfering with the image acquisition and control system. The casing 111 is designed to meet industrial ergonomic requirements, with a smooth surface and rounded edges for easy cleaning and maintenance. It also includes cable channels and ventilation holes to ensure long-term stable operation of the equipment.
[0037] The cover plate 112 is fixed to the top of the housing 111 and is typically designed to be detachable, facilitating equipment debugging, maintenance, or replacement of internal components. The cover plate 112 and the housing 111 can be connected by screws, clips, or magnetic attraction. Sealing strips or dustproof pads are installed at the joints to further enhance the overall dustproof and waterproof rating of the machine, meeting the requirements of semiconductor cleanroom environments (such as ISO Class 5 or higher).
[0038] The camera support rod 113 is securely mounted on one side (typically the front or side) of the housing 111. It is made of a high-rigidity metal material (such as stainless steel or hard aluminum alloy) and precision-machined to ensure its straightness and parallelism. Extending vertically or at an angle, the support rod provides a stable and adjustable mounting platform for the line scan camera 106 and the area scan camera 107. The support rod is equipped with standard optical rails, sliders, or quick-release interfaces, allowing for fine-tuning of the camera's height, pitch, or horizontal position to accommodate substrates of different thicknesses, sizes, or materials, ensuring optimal relative positioning between the imaging optical axis and the substrate surface. Furthermore, the support rod can integrate cable management channels to systematically route the camera's power cables, data cables, etc., to the control unit, preventing exposed cables from interfering with moving parts or affecting visual acquisition.
[0039] The adjustment component 104 includes an X-axis moving component 114 and a Y-axis moving component 115. The X-axis moving component 114 is used to drive the rotation component 103 to move along the X direction, and the Y-axis moving component 115 is disposed on the X-axis moving component 114 and is used to drive the rotation component 103 to move along the Y direction.
[0040] The X-axis moving component 114 drives the entire upper structure to move linearly along the X direction (usually defined as the left-right direction) of the device, while the Y-axis moving component 115 is mounted on the X-axis moving component 114 and drives the rotating component 103 to move along the Y direction (usually defined as the front-back direction). Through independent and coordinated movements in the X and Y directions, precise positioning of the substrate at any position on the two-dimensional plane can be achieved, meeting the trajectory control requirements of the line scan camera 106 for continuous scanning of the substrate edge.
[0041] The X-axis moving assembly 114 includes a servo motor 116, a reduction gearbox 117, a lead screw module 118, and an X-axis moving block 119. The X-axis moving block 119 is slidably disposed on the base 101. The lead screw module 118 is threadedly connected to the X-axis moving block 119. The reduction gearbox 117 is connected to the lead screw module 118. The output end of the servo motor 116 is connected to the reduction gearbox 117.
[0042] The X-axis moving block 119 is slidably mounted on the base body 110 via a high-precision linear guide rail, ensuring smooth and wobbly movement. The lead screw module 118 consists of a precision ball screw and a matching nut, with the nut portion fixedly connected to the X-axis moving block 119, achieving efficient conversion from lead screw rotation to linear displacement. The reduction gearbox 117 is connected between the servo motor 116 and the lead screw module 118, used to match the motor output speed with the torque required by the lead screw, while improving system rigidity and control resolution. The servo motor 116 serves as the power source, with its output end connected to the input shaft of the reduction gearbox 117 via a coupling 121 or directly, receiving commands from the control module 109 to achieve high-response, high-precision closed-loop position control. The entire X-axis moving assembly 114 possesses nanometer-level resolution potential, with repeatability accuracy within ±1 μm, meeting the stringent requirements of semiconductor processes for motion accuracy.
[0043] The Y-axis moving component 115 adopts a similar design concept to the X-axis, typically including a servo motor 116, a reduction mechanism, a lead screw module 118, and a Y-axis slide, but is integrated and mounted on the X-axis moving block 119, forming a "stacked" XY platform structure. This layout not only saves space but also ensures the orthogonality and synchronization of the two-axis movements, avoiding positioning errors caused by structural interference.
[0044] The rotating assembly 103 includes a rotating motor 120 and a coupling 121. The rotating motor 120 is fixed on the Y-axis moving assembly 115, and the coupling 121 is connected to the output end of the rotating motor 120 and the fixed assembly 102.
[0045] The rotary motor 120 is preferably a high-resolution stepper motor or a servo motor 116 with encoder feedback, and is fixedly mounted on the slide of the Y-axis moving assembly 115. Its output shaft is rigidly connected to the fixed assembly 102 below through a high-concentricity, zero-backlash coupling 121, ensuring accurate transmission of rotation commands without angular lag or jitter. The rotary assembly 103 can be finely adjusted with high precision within a range of ±10°, with an angular resolution of up to 0.001°, which is sufficient to compensate for angular deviations during the initial placement of the substrate.
[0046] The fixing assembly 102 includes a vacuum suction cup 122, a support base 123, a negative pressure valve 124, and a baffle 125. The support base 123 is fixed on the coupling 121, the vacuum suction cup 122 is disposed on the support base 123, the negative pressure valve 124 is connected to the vacuum suction cup 122, and the baffle 125 is disposed on the support base 123.
[0047] The fixing component 102 is used to reliably adsorb and support the substrate under test. Its structure includes a vacuum suction cup 122, a support base 123, a negative pressure valve 124, and a baffle 125. The support base 123 is firmly installed at the output end of the coupling 121 through a flange or threaded structure, serving as the mechanical foundation of the entire adsorption platform. The vacuum suction cup 122 is embedded in the upper surface of the support base 123 and is usually made of porous ceramic, polymer, or sintered metal material. It has uniform adsorption force, high temperature resistance, and antistatic properties, and is suitable for various semiconductor substrates such as silicon wafers, glass, and sapphire. The negative pressure valve 124 is integrated inside or on the side wall of the support base 123 and is connected to an external vacuum source (such as a vacuum pump) through a gas path. It is used to control the negative pressure state inside the suction cup cavity to achieve rapid adsorption and release of the substrate. The baffle 125 is arranged around the edge of the support base 123 for dust prevention.
[0048] Second Embodiment Please see Figure 6 The present invention also provides an edge-finding method for semiconductor processing, comprising: S201 captures partial images of the substrate; After the substrate is attracted and fixed to the rotating assembly 103 by the vacuum chuck 122, the system activates the area array camera 107 to perform high-resolution static imaging on a preset area of the substrate edge (usually a location with obvious straight edges or arc features). To improve imaging quality, a ring LED light source or coaxial illumination can be activated simultaneously to ensure clear edge contours and high contrast, while avoiding reflections or shadows.
[0049] S202 acquires a local image of the substrate and runs a preliminary edge-finding algorithm to calculate the current angular deviation value of the substrate; The data processing module 108 receives local images acquired by the area scan camera 107. It first performs image preprocessing (including denoising, grayscale conversion, and edge enhancement). Then, it calls a preliminary edge-finding algorithm (such as Canny edge detection combined with Hough linear transform, or a fast positioning algorithm based on template matching) to extract local geometric features of the substrate edges. These features are then compared with a standard ideal position (such as a reference direction in the equipment coordinate system) to quickly estimate the initial angular deviation (θ0) of the substrate relative to the target coordinate system. This step emphasizes speed and robustness, and can be completed within hundreds of milliseconds, making it suitable for high-speed production line requirements.
[0050] S203 controls the rotation of the rotating component 103 based on the angular deviation value, which drives the substrate to rotate in the opposite direction to complete one alignment. Based on the angle deviation value θ0 output in step S202, the control module 109 generates a corresponding rotation command, driving the rotation motor 120 to rotate the coupling 121 and the fixing component 102 in the opposite direction by an angle −θ0 (or a compensation amount close to this value), so that the edge of the substrate is approximately parallel to the X / Y axis direction of the device coordinate system. This alignment is a coarse alignment, designed to reduce the substrate angle error to within the effective field of view of the line scan camera 106 and the tolerance of the stitching algorithm (usually controlled within ±0.5°), laying the foundation for subsequent high-precision edge reconstruction.
[0051] After S204 is aligned once, the adjustment component 104 is used to drive the rotation component 103 and the substrate to move, and the line scan camera 106 synchronously and continuously acquires substrate images and stitches them into substrate edge images. After coarse alignment, the adjustment component 104 is activated, controlling the X-axis and Y-axis movement components 115 to move in tandem, causing the substrate to translate at a constant speed along a preset scanning trajectory (such as a concentric path around the outer edge of the substrate or a straight reciprocating path). During this process, the line scan camera 106 continuously acquires strip-shaped image data of the substrate edge at a high frame rate. Due to its high resolution (up to several micrometers per pixel) and high sampling frequency, the line scan camera 106 can capture continuous, distortion-free edge details. The system synchronously records the displacement encoder position information corresponding to each frame of image and uses image registration and stitching algorithms (such as stitching techniques based on feature point matching or phase correlation methods) to seamlessly fuse thousands of line scan images, reconstructing a complete, high-precision image of the entire perimeter edge contour of the substrate.
[0052] S205 calculates the residual angular deviation and center offset of the substrate edge based on the substrate edge image; The data processing module 108 performs high-precision geometric analysis on the stitched full-edge image. For circular substrates (such as standard silicon wafers), the least squares circle fitting algorithm is used to fit the edge point cloud, calculating the offset (Δx, Δy) between the actual center coordinates and the theoretical center, as well as the residual angular deviation (θ1) of the fitted circle relative to the ideal direction. For rectangular or irregularly shaped substrates, the direction of each side and the intersection point are extracted through multi-segment straight line fitting or contour matching, and then the overall pose deviation is calculated. The accuracy of this step can reach the sub-pixel level, with an angular deviation resolution of 0.001° and a center offset accuracy better than ±1 μm.
[0053] S206 controls the rotation component 103 and the adjustment component 104 to make adjustments based on the residual angular deviation and the center offset.
[0054] The control module 109, based on the residual angular deviation θ1 and center offset (Δx, Δy) output from step S205, sends fine adjustment commands to the rotation component 103 and the adjustment component 104, respectively: the rotation motor 120 performs a small angular correction (rotation −θ1) to eliminate the remaining angular error; simultaneously, the X-axis and Y-axis movement components 115 move in tandem to translate the substrate along the (−Δx, −Δy) direction, ensuring its geometric center is precisely aligned with the origin of the process coordinate system. This process can employ closed-loop feedback control, performing multiple iterative fine adjustments as needed until the pose error converges to a preset tolerance range (e.g., angular error <0.005°, center offset <0.5μm).
[0055] The S207 system records the final pose data and sends it to downstream devices.
[0056] After final alignment, the system writes the precise pose parameters of the substrate (including final center coordinates, rotation angle, edge integrity assessment results, etc.) into the log database and transmits them in real time to downstream process equipment (such as lithography machines, etching machines, and placement machines) through industrial communication interfaces (such as SECS / GEM, EtherCAT, TCP / IP, etc.) as the benchmark for their station alignment, exposure positioning, or process compensation, realizing fully automated collaborative operation.
[0057] The above description discloses only one preferred embodiment of the present invention, and should not be construed as limiting the scope of the present invention. Those skilled in the art will understand that all or part of the processes of the above embodiments can be implemented, and equivalent changes made in accordance with the claims of the present invention are still within the scope of the invention.
Claims
1. An edge-finding machine for semiconductor processing, characterized in that, It includes a base, a fixing component, a rotating component, an adjusting component, and a shooting positioning component. The adjusting component is disposed on the base, the rotating component is disposed on the adjusting component, and the adjusting component is used to drive the rotating component to move. The fixing component is disposed on the rotating component and is used to adsorb the substrate. The imaging and positioning component includes a line scan camera and an area scan camera, a data processing module and a control module. The area scan camera captures partial images of the substrate. The data processing module acquires a local image of the substrate and runs a preliminary edge-finding algorithm to calculate the current angular deviation value of the substrate. The control module is used to control the rotation of the rotating component based on the angle deviation value, so that the substrate rotates in the opposite direction to complete one alignment. After one alignment, the adjustment component is used to drive the rotating component and the substrate to move, and the line scan camera synchronously and continuously acquires substrate images and stitches them into substrate edge images. The data processing module is also used to calculate the residual angular deviation and center offset of the substrate edge based on the substrate edge image. The control module is also used to control the rotation component and the adjustment component to make adjustments based on the residual angular deviation and the center offset.
2. The edge-finding machine for semiconductor processing as described in claim 1, characterized in that, The base includes a base body, a housing, a cover plate, and a camera support rod. The housing is fixed to the base body, the cover plate is fixed to the top of the housing, and the camera support rod is fixed to one side of the housing for mounting the line scan camera and the area scan camera.
3. The edge-finding machine for semiconductor processing as described in claim 2, characterized in that, The adjustment assembly includes an X-axis moving assembly and a Y-axis moving assembly. The X-axis moving assembly is used to drive the rotating assembly to move along the X direction, and the Y-axis moving assembly is disposed on the X-axis moving assembly and is used to drive the rotating assembly to move along the Y direction.
4. The edge-finding machine for semiconductor processing as described in claim 3, characterized in that, The X-axis moving assembly includes a servo motor, a gearbox, a lead screw module, and an X-axis moving block. The X-axis moving block is slidably mounted on the base. The lead screw module is threadedly connected to the X-axis moving block. The gearbox is connected to the lead screw module. The output end of the servo motor is connected to the gearbox.
5. The edge-finding machine for semiconductor processing as described in claim 4, characterized in that, The rotating assembly includes a rotating motor and a coupling. The rotating motor is fixed on the Y-axis moving assembly, and the coupling is connected to the output end of the rotating motor and the fixed assembly.
6. The edge-finding machine for semiconductor processing as described in claim 5, characterized in that, The fixing assembly includes a vacuum suction cup, a support base, a negative pressure valve, and a baffle. The support base is fixed to the coupling, the vacuum suction cup is disposed on the support base, the negative pressure valve is connected to the vacuum suction cup, and the baffle is disposed on the support base.
7. A semiconductor processing edge finding method, employing the semiconductor processing edge finding machine according to any one of claims 1 to 6, characterized in that, include: Take partial images of the substrate; Acquire a local image of the substrate and run a preliminary edge-finding algorithm to calculate the current angular deviation value of the substrate; The rotation of the rotating component is controlled by the angle deviation value, which drives the substrate to rotate in the opposite direction to complete one alignment. After one alignment, the adjustment component is used to drive the rotating component and the substrate to move, and the line scan camera synchronously and continuously acquires substrate images and stitches them into substrate edge images. The residual angular deviation and center offset of the substrate edge are calculated based on the substrate edge image. The control rotation component and adjustment component are adjusted based on the residual angular deviation and center offset.
8. The edge-finding method for semiconductor processing as described in claim 7, characterized in that, After the control rotation component and adjustment component are adjusted based on the residual angular deviation and center offset, the method further includes, after positioning is completed, the system records the final pose data and sends it to the downstream device.