Quartz wafer defect detection equipment with multispectral imaging detection function

Abstract

The invention discloses quartz wafer defect detection equipment with a multispectral imaging detection function, and relates to the technical field of quartz wafer defect detection, and the quartz wafer defect detection equipment comprises a rack which is used as a mounting and positioning carrier of each component of the equipment; and the wafer conveying mechanism is arranged on the rack and comprises a vacuum adsorption platform for fixing a quartz wafer and a linear motor driving guide rail for driving the vacuum adsorption platform to do linear motion. In the multi-spectral imaging module, a photonic crystal metasurface light splitting module eliminates multi-band optical path difference and coaxial deviation, a multi-spectral coplanar integrated detector array solves the problems of response and transmission delay through a shared trigger circuit and a parallel reading channel, and high-quality acquisition of multi-band signals is ensured in cooperation with a wide-spectrum light source. Space-time synchronization errors are eradicated from optical and hardware levels; the closed-loop control system collects motion parameters in a high-frequency mode through a laser Doppler interferometer, precise time sequence matching under dynamic scanning is achieved, and detection precision and efficiency are both considered.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of quartz wafer defect detection, in particular to a quartz wafer defect detection device with multi-spectral imaging detection function. BACKGROUND

[0002] As a key basic material for semiconductor lithography, etching and other core processes, the surface scratches, particle contamination and subsurface bubbles, lattice defects of quartz wafers directly affect the yield of chips. Multi-spectral imaging detection technology has become the core technology path for quartz wafer defect detection because it can capture the unique optical response of different defects.

[0003] For example, a wafer detection scattering light collection objective lens and wafer detection equipment with publication number CN119620341A is applied to wafer defect detection. The scattering light collection lens includes a first lens group, a second lens group, a third lens group, and a fourth lens group along the optical axis direction from the object plane to the image plane, a total of twelve lenses and an aperture stop, and the twelve lenses and the aperture stop are coaxially arranged. The first lens group is a catadioptric lens group with positive focal power, which is used to image the object plane to the intermediate image plane by twice reflection.

[0004] However, in the prior art, the quartz wafer defect detection device with multi-spectral imaging detection function has inherent differences in the response speed of the detector, the circuit trigger delay of different spectral channels at the hardware level, the switching delay of the mechanical light splitting structure such as the filter wheel at the optical level, the different propagation times of different wavebands caused by the prism and other light splitting elements, and the further amplification of errors caused by the mismatch between the trigger signal transmission deviation and the movement speed at the control level. In order to avoid this error, the existing equipment mostly adopts an intermittent process of "wafer static collection - moving positioning", which prolongs the detection time of a single wafer to tens of seconds, making it difficult to adapt to the production capacity demand of the production line. SUMMARY

[0005] The purpose of the present application is to provide a quartz wafer defect detection device with multi-spectral imaging detection function to solve the problems raised in the background art.

[0006] To achieve the above purpose, the present application provides the following technical scheme: a quartz wafer defect detection device with multi-spectral imaging detection function, comprising a rack as the mounting and positioning carrier of various components of the device; a wafer conveying mechanism provided on the rack, comprising a vacuum adsorption platform for fixing the quartz wafer and a linear motor driven guide rail for driving the vacuum adsorption platform to move linearly, for realizing the feeding, accurate positioning, continuous conveying and discharging of the quartz wafer; A multispectral imaging module, mounted on the rack and located directly above the detection path of the wafer transport mechanism, is used to simultaneously acquire ultraviolet, visible, and near-infrared multi-band image data of the quartz wafer. It includes a light source assembly, a photonic crystal metasurface spectral dispersive module, and a multispectral coplanar integrated detector array connected in sequence by optical paths. The light source assembly is a broadband surface light source with a spectrum covering the ultraviolet to near-infrared bands. The closed-loop control system is electrically connected to the wafer conveying mechanism and the multispectral imaging module, respectively, and is used to realize the spatiotemporal synchronization of multispectral data by real-time acquisition of motion parameters and dynamic adjustment of acquisition timing. It includes a laser Doppler interferometer, an FPGA main controller and a drive module. The data processing module is electrically connected to the multispectral imaging module and the closed-loop control system. It is used to perform real-time registration and defect analysis on multispectral image data and output defect location, type and size parameters. It includes an ASIC on-chip computing unit and a back-end analysis unit.

[0007] Preferably, the photonic crystal metasurface beam-splitting module includes a quartz glass substrate and a silicon nitride nanopillar array disposed on the substrate; the nanopillar array is composed of a number of nanopillars arranged according to a preset periodic rule, and the height and spacing of the nanopillars are customized according to the optical path difference of the target spectral band, so that the optical path difference of the incident light in the ultraviolet, visible and near-infrared bands is compensated to 0±5ps after being modulated by the nanopillar array, and the propagation direction of the outgoing light in each band is coaxial, with a coaxiality deviation of <0.1°.

[0008] Preferably, the target spectral band includes the 260nm ultraviolet band, the 550nm visible light band, and the 940nm near-infrared band; the period of the nanopillar array is 200-400nm, and the height-to-diameter ratio of the nanopillars is 2-5:1.

[0009] Preferably, the multispectral coplanar integrated detector array is fabricated using a CMOS substrate epitaxial heterojunction process, with AlGaN-based ultraviolet photosensitive units, Si-based visible light photosensitive units, and InGaAs-based near-infrared photosensitive units arranged pixel-level alternately on the same chip substrate; the detector array also integrates a shared trigger circuit and a parallel charge readout channel, the shared trigger circuit receiving the synchronous trigger signal from the FPGA main controller to drive each photosensitive unit to be exposed synchronously, with an exposure delay deviation of <10ps; the parallel charge readout channel synchronously outputs image data of each band, with a data transmission delay difference of <50ns.

[0010] Preferably, the laser Doppler interferometer is mounted on the frame and parallel to the motion trajectory of the linear motor drive rail, with a sampling frequency ≥1MHz, used to collect the moving speed, acceleration and instantaneous position parameters of the vacuum adsorption platform in real time, with a position detection accuracy of ±0.01μm.

[0011] Preferably, the FPGA main controller is built-in motion-trigger timing mapping model, which dynamically calculates and adjusts the exposure trigger time of the multi-spectral coplanar integrated detector array according to the real-time motion parameters collected by the laser Doppler interferometer, with an adjustment accuracy of 10 ps and a trigger parameter update of once per 1 ms.

[0012] Preferably, the driving module is electrically connected with the linear motor driving guide rail, and can adjust the motion speed of the vacuum adsorption platform to adaptively change in the range of 50-300 mm / s, with a speed fluctuation error of less than 0.5%, and can realize millisecond-level speed correction in response to the instruction of the FPGA main controller.

[0013] Preferably, the ASIC on-chip computing unit is directly connected with the output end of the multi-spectral coplanar integrated detector array, and is built-in with a preset multi-spectral pixel coordinate mapping table and a dynamic feature anchor point matching algorithm; the dynamic feature anchor point matching algorithm takes the lattice periodic texture on the surface of the quartz wafer as an anchor point, and adjusts the pixel position of different spectral images in real time, with a registration error of less than 0.05 pixels and a single-frame image registration time consumption of less than 100 μs.

[0014] Preferably, the light source assembly has a light intensity uniformity of greater than or equal to 90%, an ultraviolet waveband radiation intensity of greater than or equal to 10 mW / cm², and a near-infrared waveband radiation intensity of greater than or equal to 20 mW / cm²; the adsorption force of the vacuum adsorption platform can be adjusted in the range of 0.1-0.5 MPa, and is suitable for quartz wafers with a diameter of 150-300 mm.

[0015] Preferably, the back-end analysis unit adopts a deep learning model based on an attention mechanism to perform defect identification and classification on the registered multi-spectral images.

[0016] Compared with the prior art, the present application has the following advantages: 1、In the present application, the wafer conveying mechanism realizes stable fixing and high-speed smooth motion of the wafer by means of the multi-hole ceramic adsorption platform and the high-precision linear motor guide rail, which is suitable for the continuous conveying demand of mass production line; in the multi-spectral imaging module, the photonic crystal superstructure surface light splitting module eliminates the multi-waveband optical path difference and coaxial deviation, the multi-spectral coplanar integrated detector array solves the response and transmission delay problem through a shared trigger circuit and a parallel reading channel, and cooperates with a wide-spectrum light source to ensure high-quality acquisition of multi-waveband signals, thereby eliminating the space-time synchronization error from the optical and hardware levels; the closed-loop control system relies on the high-frequency acquisition of motion parameters by the laser Doppler interferometer to form a "perception-computation-driving" closed loop, realizes accurate timing matching under dynamic scanning, and balances the detection accuracy and efficiency; the defect positioning accuracy is high, the detection rate of common defects is high, and the quality control demand of "high efficiency + high precision" of the mass production line is perfectly met, while the operation and maintenance calibration frequency and cost are reduced; 2、In the application, through the stable fixing and ultra-low fluctuation movement of the wafer, the influence of wafer posture deviation on defect positioning is avoided, in the multi-spectral imaging module, the wide-spectrum light source is matched with the optical fiber coupling system to ensure the stability of light intensity, the four-waveband photonic crystal superstructure eliminates the dispersion delay and coaxial deviation, the high-resolution detector array accurately captures the near-infrared weak scattering signal with low trigger delay and high quantum efficiency, and the synchronous high-quality collection of subsurface defect signals is realized from the optical and hardware levels; the closed-loop control system relies on the 2MHz high-frequency laser interferometer and the FPGA model optimized by Kalman filtering, combined with the EtherCAT real-time communication drive module, to form a "motion prediction-time sequence regulation-speed correction" closed loop, to ensure the defect positioning accuracy under dynamic scanning; the data processing module fuses multi-band features through the ASIC unit and the Spectral-Transformer model, and accurately identifies defects to perfectly adapt to the strict control requirements of high-end wafer factories for "high-precision identification, deep detection and full-process traceability" of hidden defects. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The flow chart of the quartz wafer defect detection equipment with multi-spectral imaging detection function of the application; Figure 2 The first perspective structural schematic diagram of the quartz wafer defect detection equipment with multi-spectral imaging detection function of the application; Figure 3 The second perspective structural schematic diagram of the quartz wafer defect detection equipment with multi-spectral imaging detection function of the application; Figure 4 The top view structural schematic diagram of the quartz wafer defect detection equipment with multi-spectral imaging detection function of the application; Figure 5 The perspective structural schematic diagram of the wafer conveying mechanism in the quartz wafer defect detection equipment with multi-spectral imaging detection function of the application; Figure 6 The perspective structural schematic diagram of the multi-spectral imaging module in the quartz wafer defect detection equipment with multi-spectral imaging detection function of the application.

[0018] In the figure: 1, rack; 2, wafer conveying mechanism; 21, vacuum adsorption platform; 22, linear motor driven guide rail; 3, multi-spectral imaging module; 4, closed-loop control system. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the scope of protection of the present application.

[0020] Embodiment one: refer to Figures 1-6 As shown in the figure: quartz wafer defect detection equipment with multispectral imaging detection function, for the conventional defect detection demand of semiconductor mass production line quartz wafer, need to balance the detection efficiency and basic accuracy, the equipment needs to meet the continuous detection rhythm of "second level / piece" of mass production line, at the same time cover the common defects such as surface scratch, particle pollution.

[0021] The rack 1 adopts aluminum alloy integral casting structure, equipped with damping foot pad at the bottom, the horizontal error of the table top is less than 0.02mm / m, which provides rigid installation reference for each component.

[0022] The wafer conveying mechanism 2 includes a vacuum adsorption platform 21 and a linear motor driven guide rail 22; the vacuum adsorption platform adopts a porous ceramic panel, the adsorption force can be adjusted in the range of 0.2-0.5MPa, which is suitable for a quartz wafer with a diameter of 300mm, and the flatness deviation of the wafer after adsorption is less than 0.01mm; the linear motor driven guide rail selects Yaskawa SGLF series, the movement stroke is 600mm, the positioning accuracy is ±0.1μm, and the maximum movement speed is 300mm / s.

[0023] The light source assembly in the multispectral imaging module 3: the spectral range is 200-1100nm, the light intensity uniformity is 92%, the ultraviolet band (260nm) radiation intensity is 12mW / cm², the near-infrared band (940nm) radiation intensity is 25mW / cm², and the distance between the light source light outlet and the wafer surface is 50mm.

[0024] The photonic crystal superstructure surface light splitting module: the substrate is quartz glass (thickness 2mm), the surface grows silicon nitride nanocolumn array, the period is 300nm, the nanocolumn diameter is 100nm, and the height is 300nm (height to diameter ratio 3:1); it is designed for 260nm ultraviolet, 550nm visible light and 940nm near-infrared band, and after modulation, the optical path difference of three wave bands is compensated to 0±4ps, and the deviation of the outgoing light coaxiality is 0.08°.

[0025] Multi-spectral coplanar integrated detector array: prepared by CMOS-based epitaxial heterojunction process, pixel size 5 pm, resolution 2048x2048; AlGaN-based ultraviolet light sensing unit quantum efficiency 65%, Si-based visible light sensing unit quantum efficiency 75%, InGaAs-based near-infrared light sensing unit response range 900-1700 nm; common trigger circuit delay deviation 8 ps, parallel charge reading channel data transmission delay difference 40 ns.

[0026] The laser Doppler interferometer in the closed-loop control system 4 has a sampling frequency of 1 MHz and a position detection accuracy of ±0.01 pm. It is installed on the side of the rack and parallel to the motion trajectory of the linear motor driven guide rail. The laser beam is incident on the reflective target on the side of the vacuum suction platform. The FPGA main controller has a motion-trigger timing mapping model built-in and is based on C language algorithm logic. The trigger time can be updated every 1 ms according to the motion parameters, and the adjustment accuracy is 10 ps.

[0027] The driving module is electrically connected with the linear motor driven guide rail. The speed regulation range is 50-300 mm / s, and the speed fluctuation error is 0.4%. The speed correction delay of the FPGA command response is less than 2 ms.

[0028] The ASIC on-chip computing unit in the data processing module uses an integrated multi-spectral pixel coordinate mapping table and a dynamic feature anchor point matching algorithm with a quartz lattice 1 pm period texture as an anchor point. The single-frame image registration takes 80 ps, and the registration error is 0.04 pixels.

[0029] The deep learning model of the back-end analysis unit combined with the attention mechanism has a training data set containing 100,000 labeled defect images, which can identify six common defects such as surface scratches and particle contamination.

[0030] The quartz wafer is transported to the vacuum suction platform 21 by the mechanical hand for loading and pre-positioning. The platform starts to fix the wafer with a suction force of 0.3 MPa. The laser positioning system (Keensys LK-H020) completes the alignment of the wafer center and the detection reference. The alignment error is less than 0.02 mm.

[0031] Dynamic scanning and synchronous acquisition: the driving module regulates the linear motor driven guide rail to move at a constant speed of 200 mm / s. The laser Doppler interferometer collects real-time platform moving speed, acceleration and instantaneous position data, which are transmitted to the FPGA main controller. The FPGA generates a synchronous trigger signal according to the motion parameters to drive the multi-spectral coplanar integrated detector array to expose. The light source assembly irradiates the wafer surface, and the reflected light is incident into three types of light sensing units after being split by the photonic crystal superstructure surface, and the original data are output in parallel channels.

[0032] Real-time registration and defect analysis: The ASIC on-chip computing unit receives the raw data, completes preliminary registration through the coordinate mapping table, and fine-tunes the deviation through the dynamic feature anchor point algorithm; the registered data is transmitted to the backend analysis unit, and the model completes defect identification and classification within 50 ms, calculates the defect position (accuracy ±0.05 μm), size, and other parameters.

[0033] Blanking and distribution: The guide rail drives the platform to the blanking position, according to the test results, the manipulator moves the qualified wafers to the finished product box, and the defective wafers to the rework area, the vacuum adsorption force is released, and the single detection is completed.

[0034] In this embodiment, the rigid structure and vibration reduction design of the rack 1 provide a stable installation reference for the equipment and ensure the running accuracy of each component; the wafer conveying mechanism 2 realizes stable fixation and high-speed smooth movement of the wafer through the multi-hole ceramic adsorption platform and high-precision linear motor guide rail, which adapts to the continuous conveying demand of the production line; in the multi-spectral imaging module 3, the photonic crystal superstructure surface light splitting module eliminates the multi-band optical path difference and coaxial deviation, the multi-spectral coplanar integrated detector array solves the response and transmission delay problem through the shared trigger circuit and parallel reading channel, and cooperates with the wide-spectrum light source to ensure high-quality acquisition of multi-band signals, and eliminates the space-time synchronization error from the optical and hardware levels; the laser Doppler interferometer of the closed-loop control system 4 collects motion parameters at high frequency, the FPGA main controller updates the trigger time (10 ps accuracy) at 1 ms level, and the driving module accurately controls the speed (fluctuation error 0.4%) to form a "perception-computation-driving" closed loop, realize the timing accurate matching under dynamic scanning; the ASIC on-chip computing unit of the data processing module completes high-precision registration (error 0.04 pixels) at a speed of 80 μs / frame, the backend model quickly identifies 6 types of defects (within 50 ms), and ensures the detection accuracy and efficiency; the whole process realizes "second / plate" continuous detection of 300 mm wafers, the defect positioning accuracy is ±0.05 μm, the detection rate of common defects is high, and it perfectly adapts to the quality control demand of "high efficiency + high precision" of the production line, while reducing the operation and calibration frequency and cost.

[0035] Embodiment two: according to Figures 1-6 The embodiment focuses on detecting hidden defects such as subsurface bubbles and small lattice defects, requiring defect detection accuracy of 0.05 μm level, subsurface defect detection depth ≥50 μm, and adapting to the strict quality control standards of high-end wafer factories.

[0036] The rack 1 adopts granite base + aluminum alloy frame structure, the horizontal error of the table top is <0.01 mm / m, and the built-in constant temperature control system (temperature control accuracy ±0.5℃) reduces the influence of temperature on the optical system.

[0037] Wafer conveying mechanism 2: The vacuum adsorption platform adopts a porous sapphire panel, with an adsorption force adjustment range of 0.3-0.5 MPa, suitable for 300 mm wafers, and an adsorption plane flatness deviation of <0.005 mm. The linear motor driven guide rail is selected from the THK SSR series, with a positioning accuracy of ±0.05 μm, a movement speed of 50-250 mm / s, and an air bearing vibration reduction structure.

[0038] Light source assembly of multi-spectral imaging module 3: A wide-spectrum light source is used, with a spectral range of 200-1700 nm, a light intensity uniformity of 95%, a 260 nm ultraviolet radiation intensity of 15 mW / cm², and a 940 nm near-infrared radiation intensity of 30 mW / cm². An optical fiber coupling light output system is equipped, with a light intensity stability of ±1% / h.

[0039] Photonic crystal superstructure surface light splitting module: The quartz glass substrate is 3 mm thick, the silicon nitride nanocolumn array period is 250 nm, the nanocolumn diameter is 80 nm, and the height is 240 nm (height to diameter ratio 3:1). It is designed for four wavebands of 260 nm, 550 nm, 940 nm, and 1310 nm, with an optical path difference compensation of 0±3 ps, and an exit light coaxiality deviation of 0.05°.

[0040] Multi-spectral coplanar integrated detector array: Pixel size 3.5 μm, resolution 4096×4096; AlGaN-based ultraviolet photosensitive unit quantum efficiency 70%, InGaAs-based near-infrared photosensitive unit response range 900-1700 nm, quantum efficiency 72%; common trigger circuit delay deviation 5 ps, parallel read channel transmission delay difference 30 ns.

[0041] Laser Doppler interferometer in closed-loop control system 4: Sampling frequency 2 MHz, position detection accuracy ±0.005 μm, laser beam incident to high-precision mirror of vacuum adsorption platform, reflected signal signal-to-noise ratio ≥50 dB.

[0042] FPGA main controller built-in optimized motion-trigger timing mapping model, introduces Kalman filter algorithm to predict motion trajectory, adjustment accuracy 5 ps, parameter update frequency 2 ms.

[0043] Drive module uses a servo driver, with a speed fluctuation error of 0.2%, a response delay of <1 ms, and supports EtherCAT real-time communication with FPGA.

[0044] ASIC on-chip computing unit registration algorithm in data processing module 5 introduces a lightweight deep learning model, with a single-frame registration time of 60 μs and a registration error of 0.03 pixels; dynamic feature anchor point matching algorithm supports lattice texture and artificial marker dual-anchor calibration.

[0045] Backend analysis unit: Spectral-Transformer deep learning model is adopted to fuse multi-band spectral-spatial features. The training dataset contains 200,000 labeled images, covering 8 types of defects such as surface scratches and subsurface bubbles, among which 50,000 samples are labeled for subsurface defects.

[0046] Loading and pre-positioning: Wafer is transported by vacuum chuck manipulator, and fixed by vacuum adsorption platform with 0.4 MPa adsorption force. Laser positioning system completes alignment with an error less than 0.01 mm. Constant temperature system is started to stabilize the internal temperature of the equipment at 23±0.5℃.

[0047] Dynamic scanning and synchronous acquisition: Linear motor drives the guide rail at a constant speed of 150 mm / s. Laser Doppler interferometer collects motion parameters every 0.5 ms and transmits them to the FPGA main controller. FPGA generates a synchronous trigger signal by combining Kalman filter to predict the trajectory. The exposure time of the detector array is dynamically adjusted according to the characteristics of the waveband. After the reflected light is split by the metasurface, the four wavebands of light are synchronously incident into the detector, and the data are output in parallel.

[0048] Real-time registration and defect analysis: ASIC unit first registers through coordinate mapping table, and then fine-tunes with lattice texture as anchor point. It strengthens the near-infrared waveband feature extraction for subsurface defects. Backend analysis unit completes defect recognition within 80 ms, calculates the depth of subsurface defects through multi-waveband light intensity attenuation model, and the depth detection range is 1-50 μm.

[0049] Unloading and diversion: After detection, the wafer is moved to the sorting machine by the manipulator, and according to the defect type (such as repairable scratch / non-repairable bubble), it is diverted to the rework area or scrap area. The detection data are automatically uploaded to the MES system.

[0050] In this embodiment, the gantry 1 adopts a granite base and a constant temperature control system to reduce the interference of temperature and vibration on the optical system from the environmental source, and to provide a stable reference for high-precision detection, with a table levelness of <0.01 mm / m and a temperature control accuracy of ±0.5℃; the wafer conveying mechanism 2 realizes stable fixation and ultra-low fluctuation movement of a 300mm wafer through a multi-hole sapphire adsorption platform (flatness deviation <0.005mm) and a high-precision guide rail with air bearing (positioning accuracy ±0.05μm), avoiding the influence of wafer posture deviation on defect positioning; in the multispectral imaging module 3, a wide-spectrum light source (200-1700nm) is matched with a fiber coupling system to ensure light intensity stability, a four-waveband photonic crystal superstructure (optical path difference compensation 0±3ps) eliminates the delay and coaxial deviation of light splitting, and a high-resolution detector array (pixel 3.5μm) accurately captures near-infrared weak scattering signals with low trigger delay (5ps) and high quantum efficiency, realizing synchronous high-quality collection of subsurface defect signals from the optical and hardware levels; the closed-loop control system 4 relies on a 2MHz high-frequency laser interferometer (accuracy ±0.005μm) and a Kalman filter optimized FPGA model (trigger accuracy 5ps), combined with an EtherCAT real-time communication drive module (response delay <1ms) to form a "motion prediction-time sequence regulation-speed correction" closed loop, ensuring that the defect positioning accuracy under dynamic scanning reaches the level of 0.05μm; the data processing module combines an ASIC unit (error 0.03 pixels) with a Spectral-Transformer model through deep learning registration, accurately identifies 8 types of defects by fusing multi-band features, realizes 1-50μm depth detection based on a model trained by 50,000 subsurface samples, and automatically uploads data to the MES system, perfectly adapting to the strict control requirements of high-end wafer factories for "high-precision identification, depth detection, and full-process traceability" of hidden defects.

[0051] Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can modify the technical solutions described in the foregoing embodiments, or make equivalent replacements to part of the technical features, and any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A quartz wafer defect detection device with multispectral imaging detection function, characterized in that: include: The frame (1) serves as the mounting and positioning carrier for the various components of the equipment; The wafer conveying mechanism (2) is located on the frame (1) and includes a vacuum adsorption platform (21) for fixing the quartz wafer and a linear motor drive rail (22) for driving the vacuum adsorption platform (21) to move linearly, for realizing the loading, precise positioning, continuous conveying and unloading of the quartz wafer; The multispectral imaging module (3) is mounted on the rack (1) and located directly above the detection path of the wafer transport mechanism (2). It is used to simultaneously acquire ultraviolet, visible, and near-infrared multi-band image data of the quartz wafer. It includes a light source component, a photonic crystal metasurface beam splitting module, and a multispectral coplanar integrated detector array connected in sequence by optical paths. The light source component is a broadband surface light source with a spectrum covering the ultraviolet to near-infrared bands. The closed-loop control system (4) is electrically connected to the wafer conveying mechanism (2) and the multispectral imaging module (3) respectively. It is used to realize the spatiotemporal synchronization of multispectral data by real-time acquisition of motion parameters and dynamic adjustment of acquisition timing. It includes a laser Doppler interferometer, an FPGA main controller and a drive module. The data processing module is electrically connected to the multispectral imaging module (3) and the closed-loop control system (4) for real-time registration and defect analysis of multispectral image data, and outputs defect location, type and size parameters. It includes an ASIC on-chip computing unit and a back-end analysis unit.

2. The quartz wafer defect detection equipment with multispectral imaging detection function according to claim 1, characterized in that: The photonic crystal metasurface beam-splitting module includes a quartz glass substrate and a silicon nitride nanopillar array disposed on the substrate. The nanopillar array is composed of several nanopillars arranged according to a preset periodic rule. The height and spacing of the nanopillars are customized according to the optical path difference of the target spectral band, so that the optical path difference of the incident light in the ultraviolet, visible and near-infrared bands is compensated to 0±5ps after being modulated by the nanopillar array, and the propagation direction of the outgoing light in each band is coaxial with a coaxiality deviation of <0.1°.

3. The quartz wafer defect detection equipment with multispectral imaging detection function according to claim 1, characterized in that: The target spectral bands include the 260nm ultraviolet band, the 550nm visible light band, and the 940nm near-infrared band; the period of the nanopillar array is 200-400nm, and the height-to-diameter ratio of the nanopillars is 2-5:

1.

4. The quartz wafer defect detection equipment with multispectral imaging detection function according to claim 1, characterized in that: The multispectral coplanar integrated detector array is fabricated using a CMOS substrate epitaxial heterojunction process. AlGaN-based ultraviolet photosensitive units, Si-based visible light photosensitive units, and InGaAs-based near-infrared photosensitive units are arranged pixel-level alternately on the same chip substrate. The detector array also integrates a shared trigger circuit and a parallel charge readout channel. The shared trigger circuit receives the synchronous trigger signal from the FPGA main controller and drives each photosensitive unit to be exposed synchronously with an exposure delay deviation of <10ps. The parallel charge readout channel synchronously outputs image data for each band with a data transmission delay difference of <50ns.

5. The quartz wafer defect detection equipment with multispectral imaging detection function according to claim 1, characterized in that: The laser Doppler interferometer is mounted on the frame (1) and parallel to the motion trajectory of the linear motor drive rail (22). The sampling frequency is ≥1MHz. It is used to collect the moving speed, acceleration and instantaneous position parameters of the vacuum adsorption platform (21) in real time, with a position detection accuracy of ±0.01μm.

6. The quartz wafer defect detection equipment with multispectral imaging detection function according to claim 1, characterized in that: The FPGA main controller has a built-in motion-trigger timing mapping model. The model dynamically calculates and adjusts the exposure triggering time of the multispectral coplanar integrated detector array based on the real-time motion parameters collected by the laser Doppler interferometer. The adjustment accuracy is up to 10 ps, ​​and the triggering parameters are updated every 1 ms.

7. The quartz wafer defect detection device with multispectral imaging detection function according to claim 1, characterized in that: The drive module is electrically connected to the linear motor drive rail (22), and can adjust the movement speed of the vacuum adsorption platform (21) to adapt within the range of 50-300mm / s with a speed fluctuation error of <0.5%, and respond to the instructions of the FPGA main controller to achieve millisecond-level speed correction.

8. The quartz wafer defect detection device with multispectral imaging detection function according to claim 1, characterized in that: The on-chip computing unit of the ASIC is directly connected to the output of the multispectral coplanar integrated detector array, and has a built-in preset multispectral pixel coordinate mapping table and dynamic feature anchor point matching algorithm. The dynamic feature anchor point matching algorithm uses the lattice periodic texture of the quartz wafer surface as anchor points to finely adjust the pixel position of different spectral images in real time, with a registration error of <0.05 pixels and a single frame image registration time of <100μs.

9. The quartz wafer defect detection device with multispectral imaging detection function according to claim 1, characterized in that: The light intensity uniformity of the light source component is ≥90%, the ultraviolet band radiation intensity is ≥10mW / cm², and the near-infrared band radiation intensity is ≥20mW / cm²; the adsorption force of the vacuum adsorption platform (21) can be adjusted within the range of 0.1-0.5MPa, and it is suitable for quartz wafers with a diameter of 150-300mm.

10. The quartz wafer defect detection device with multispectral imaging detection function according to claim 1, characterized in that: The back-end analysis unit uses a deep learning model based on an attention mechanism to identify and classify defects in the registered multispectral images.

Images