A multi-focal plane imaging defect detection method and a defect detection system

Abstract

This application relates to a defect detection method and system using multifocal planar imaging. The defect detection method includes the following steps: illuminating different regions of a beam splitter with a first polarization optical path and a second polarization optical path to obtain multiple sub-beams; focusing the multiple sub-beams at different axial depths of a wafer; collecting reflected light to form multiple non-overlapping detection beams; illuminating different photosensitive areas of a detector with the multiple detection beams, so that images of different focal planes are simultaneously imaged at different spatial positions of the detector to obtain focal plane images corresponding to different focal planes; performing geometric correction and pixel alignment on the multiple focal plane images, and stacking them along the axial direction to form a data volume; and obtaining the three-dimensional spatial coordinates (xc, yc, zc) of the defect through connected component analysis. The defect detection system includes an illumination module, a polarization module, a beam splitter, a micromirror array, and a detector arranged sequentially along the optical path. This invention can determine the location of defects.

Description

Technical Field

[0001] This invention relates to the field of semiconductor inspection technology, and in particular to a defect detection method and system using multifocal planar imaging. Background Technology

[0002] In the semiconductor manufacturing field, the wafer serves as the core substrate for integrated circuit manufacturing, and its surface quality directly determines the final yield, electrical performance, and long-term reliability of the chip. As process nodes continue to shrink to the nanometer level, any tiny surface defects (such as particle contamination, scratches, residues, or crystal defects) can lead to short circuits, leakage, or parameter drift in devices, thereby causing chip malfunction.

[0003] In existing technologies, defect identification typically relies on single intensity imaging or simple grayscale threshold analysis. However, this method has significant limitations when dealing with complex samples: on the one hand, different types of defects (such as surface specular reflection defects and internal volume scattering defects) may exhibit similar grayscale characteristics, leading to classification difficulties; on the other hand, simple two-dimensional image analysis is difficult to effectively distinguish the physical depth level of the defect, and it is very easy to misjudge strong surface reflections as internal impurity particles.

[0004] Therefore, it is necessary to provide a novel defect detection method and system for multifocal plane imaging to solve the above-mentioned problems in the prior art. Summary of the Invention

[0005] The technical problem to be solved by this application is to provide a defect detection method and system that can determine the location of defects through multifocal plane imaging.

[0006] To address the aforementioned technical problems, according to embodiments of this application, a defect detection method using multifocal planar imaging is provided, comprising the following steps: separating illumination light into a first polarization path and a second polarization path; illuminating the first polarization path and the second polarization path to different regions of a beam splitter to obtain multiple sub-beams; focusing the multiple sub-beams at different axial depths of a wafer; collecting reflected light from each sub-beam reflected by the wafer, and redirecting the multiple reflected lights to form multiple non-overlapping detection lights; illuminating different photosensitive areas of a detector to simultaneously image images of different focal planes at different spatial positions of the detector, thereby obtaining focal plane images corresponding to different focal planes; performing geometric correction and pixel alignment on the multiple focal plane images, and stacking them along the axial direction to form a data volume; and obtaining the three-dimensional spatial coordinates (xc, yc, zc) of the defect through connected component analysis.

[0007] According to an embodiment of this application, the step of irradiating different regions of a beam splitter with the first polarizing optical path and the second polarizing optical path respectively to obtain multiple sub-beams includes: dividing the beam splitter into regions, such that the beam splitter has a first region, a second region, and a third region; the first region is located on one side of the beam splitter, with an angle of 6-10° between it and the optical axis of the illumination light, and a transmittance of 50-70%; the second region is located in the middle of the beam splitter, with an angle of 1-5° between it and the optical axis of the illumination light, and a transmittance of 20-40%; the third region is located on the other side of the beam splitter, with an angle of 6-10° between it and the optical axis of the illumination light, and a transmittance of 5-15%; controlling the first polarizing optical path to irradiate the first region and the second region to form a first beam and a second beam; and controlling the second polarizing optical path to irradiate the third region to form a third beam.

[0008] According to an embodiment of this application, focusing the plurality of sub-beams onto different axial depths of the wafer includes: arranging micro-mirrors and micro-objectives corresponding to the first region, the second region, and the third region along the optical path; controlling the first beam, the second beam, and the third beam to be redirected to form parallel light after passing through the corresponding micro-mirrors, and then focusing them onto the wafer after passing through the micro-objectives; the first beam, the second beam, and the third beam having different optical path differences. , , where t is the thickness of the beam splitter; The angle between the corresponding region and the optical axis of the illumination light; based on the optical path difference Determine the axial focusing position of each beam. , , where n is the refractive index of the corresponding region; 1 is the focal length of each micro objective; WD is the working distance.

[0009] According to an embodiment of this application, the step of illuminating different photosensitive areas of a detector with multiple detection lights, so that images of different focal planes are simultaneously imaged on different spatial positions of the detector to obtain focal plane images corresponding to different focal planes, includes: dividing the detection surface of the detector to give the detector a first photosensitive area, a second photosensitive area, and a third photosensitive area; setting relay mirrors along the optical path corresponding to the first area, the second area, and the third area; controlling the reflected light formed by the first beam to illuminate the first photosensitive area after passing through the corresponding relay mirror to form a first focal plane image; controlling the reflected light formed by the second beam to illuminate the second photosensitive area after passing through the corresponding relay mirror to form a second focal plane image; and controlling the reflected light formed by the third beam to illuminate the third photosensitive area after passing through the corresponding relay mirror to form a third focal plane image.

[0010] According to embodiments of this application, the geometric correction and pixel alignment of the plurality of focal plane images includes, based on the focal length of the relay lens... The focal length of the corresponding micro objective To address the differences in magnification caused by variations in focal planes, digital scaling and resampling are performed on each focal plane image to ensure a uniform pixel resolution across the first, second, and third focal plane images. Using the field-of-view coordinates (x, y) of each focal plane image as variables, and based on the angle between each region and the optical axis of the illumination light... Calculate the lateral misalignment of pixels at the edge of the field of view caused by the fan-shaped optical path formed by the beam splitter. , , where y is the vertical pixel coordinate; The tangent of the angle between each region and the optical axis of the illumination light characterizes the degree of fan-shaped expansion; The magnification factor; based on the stated lateral misalignment. Construct a transformation matrix that includes lateral translation and shearing components; perform a geometric transformation on the field coordinates (x, y) according to the transformation matrix to obtain the corrected coordinates (x1, y1) that eliminate the pixel misalignment at the edge of the field of view.

[0011] According to an embodiment of this application, the axial stacking to form a data volume includes using the first polarization optical path and the second polarization optical path as polarization state dimensions, combined with correction coordinates (x1, y1) and axial focusing position. Obtain the data volume V(x1,y1,z,φ), where φ represents the polarization state of the first polarization optical path or the polarization state of the second polarization optical path.

[0012] According to an embodiment of this application, obtaining the three-dimensional spatial coordinates of the defect through connected component analysis includes calculating a polarization difference image D(x1,y1,z) based on the data volume V(x1,y1,z,φ). Threshold segmentation is performed based on the grayscale features of the polarization difference image D(x1,y1,z). Pixels that meet the threshold are labeled with three-dimensional connected components. The lateral coordinates (xc,yc) of the defect and its depth layer zc are identified, and the three-dimensional spatial coordinates (xc,yc,zc) of the defect are obtained.

[0013] According to an embodiment of this application, the method further includes calculating the polarization degree M of each pixel based on the data volume V(x1,y1,z,φ). According to the degree of polarization M and the axial focusing position Defect type determination is performed; when the polarization degree M is greater than or equal to the first threshold and the focusing position is... When corresponding to the third region, the defect is determined to be a specular reflection type defect; when the polarization degree M is less than or equal to the second threshold and the focusing position When corresponding to the first region, the defect is determined to be a volume scattering type defect; when the polarization degree M is between the first threshold and the second threshold, and the focusing position... When it corresponds to the second region, it is determined to be a mid-layer interface defect.

[0014] A multifocal plane imaging defect detection system is provided for implementing the aforementioned defect detection method. The defect detection system includes, sequentially arranged along the optical path: an illumination module for emitting illumination light; a polarization module having a polarization beam-splitting interface for separating the illumination light into a first polarization optical path and a second polarization optical path; a beam splitter having a first region, a second region, and a third region, each region having a different tilt angle with the optical axis and a different transmittance; the first polarization optical path is incident on the first region and the second region, and the second polarization optical path is incident on the third region; a micromirror array, respectively arranged corresponding to the first region, the second region, and the third region, to focus the illumination light onto the wafer; and a detector for receiving signals reflected from the wafer.

[0015] According to an embodiment of this application, the micromirror array includes micromirrors and microobjectives arranged sequentially along the optical path; the micromirrors are used to redirect the illumination light to form parallel light, and the microobjectives are used to focus the parallel light onto the wafer.

[0016] By employing the above technical solution, using a single-piece, multi-zone tilting beam splitter, the illumination light is separated into multiple sub-beams with different tilt angles, different energy distribution ratios, and corresponding to different polarization paths. These sub-beams are then focused onto different axial depths of the wafer by a micromirror array, and the reflected light signals are received by a detector. This achieves multi-focal plane synchronous imaging, reducing errors that occur when using multiple cameras for detection. Furthermore, based on the degree of polarization and axial focusing position... The joint analysis enables the identification of defect types, improving the speed and accuracy of wafer defect detection. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the optical path of a defect detection device according to an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of the optical path received by a detector according to an embodiment of the present invention.

[0019] Figure 3 This is a flowchart illustrating the steps of a defect detection method according to an embodiment of the present invention.

[0020] Figure label:

[0021] 100. Illumination module; 110. Collimating lens; 200. Polarization module; 300. Beam splitter; 310. First region; 320. Second region; 330. Third region; 410. Micromirror; 420. Micro objective; 500. Beam splitter cube; 600. Relay mirror; 700. Detector. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art. The terms "comprising" and similar expressions used herein mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, but does not exclude other elements or objects.

[0023] The following is in conjunction with the appendix Figure 1-3 The specific embodiments of the present invention will be further described in detail below.

[0024] The present invention provides a defect detection method and system for multifocal plane imaging. The defect detection method is applied to the defect detection system to detect defects in wafers. Specifically, the defect detection system includes an illumination module 100, a polarization module 200, a beam splitter 300, a micromirror array, and a detector 700 arranged sequentially along the optical path.

[0025] In some embodiments, the illumination module 100 is used to emit illumination light; specifically, the illumination module 100 includes a light source and a collimating lens 110, with a gap between the collimating lens 110 and the light source; the light source is used to provide illumination light with a continuous spectrum in the wavelength range of 400-700nm, and the illumination light is divergent at this time, and the collimating lens 110 is used to convert the divergent light into a parallel beam, thereby making the illumination light parallel.

[0026] In some embodiments, the polarization module 200 has a polarization beam splitting interface for separating the illumination light into a first polarization optical path and a second polarization optical path. Specifically, the polarization module 200 is a birefringent crystal or a fiber beam splitter. In this embodiment, the polarization module 200 is selected as a birefringent crystal, and the polarization beam splitting interface is the surface of the birefringent crystal. The illumination light emitted by the light source enters the polarization module 200 and is separated into two beams after passing through the polarization beam splitting interface, forming the first polarization optical path and the second polarization optical path respectively, and the angle between the first polarization optical path and the second polarization optical path is less than 45°.

[0027] In some embodiments, the surface of the beam splitter 300 is laterally divided into at least three regions, each region forming a different first tilt angle, a second tilt angle, and a third tilt angle with respect to the optical axis of the illumination light, thus dividing the beam splitter 300 into a first region 310, a second region 320, and a third region 330 with different transmittances; a first polarized light path is incident on the first region 310 and the second region 320, and a second polarized light path is incident on the third region 330; specifically, the first region 310 is located on the left side of the beam splitter 300 and has a first angle with the optical axis of the illumination light, the first angle ranging from 6-10°. It is worth noting that the first included angle here refers to the angle between the normal of the first region 310 and the optical axis. In this embodiment, the first included angle is preferably 8°. At the same time, the surface of the first region 310 is coated. For example, a neutral density filter film or a selective absorption film can be coated on the surface of the first region 310. By adjusting the thickness of the film, the transmittance of the first region 310 is made up to 50-70%, preferably 60%. After the first polarized light path passes through the first region 310, it forms a first beam. The first beam carries 60% of the total energy of the illumination light so that it can be focused on the deep region of the wafer. The second region 320 is located in the middle of the beam splitter 300 and has a second angle with the optical axis of the illumination light. The second angle is in the range of 1-5°, preferably 3°. It is worth noting that the second angle here refers to the angle between the normal of the second region 320 and the optical axis, that is, perpendicular to the optical axis. At the same time, the surface of the second region 320 is coated. For example, a neutral density filter film or a selective absorption film can be coated on the surface of the second region 320. By adjusting the thickness of the film, the transmittance of the second region 320 is 20-40%, preferably 30%. After the first polarized light path passes through the second region 320, a second beam is formed. The second beam carries 30% of the total energy of the illumination light so that it can be focused onto the middle layer region of the wafer. The third region 330 is located to the right of the beam splitter 300 and has a third angle with the optical axis of the illumination light. The third angle ranges from -6° to -10°, which is opposite to the tilt direction of the first region 310. It is worth noting that the third angle here refers to the angle between the normal of the third region 330 and the optical axis. In this embodiment, the third angle is preferably -8°. At the same time, the surface of the third region 330 is coated. For example, a high transmittance anti-reflection film can be coated on the surface of the third region 330 to make the transmittance of the third region 330 5-15%, preferably 10%. After the second polarized light path passes through the third region 330, a third beam is formed. The third beam carries 10% of the total energy of the illumination light so as to focus on the wafer surface to avoid the situation where the surface mirror reflection signal is too strong and the detector 700 is saturated and unable to detect.

[0028] In some embodiments, the micromirror array is configured corresponding to the first region 310, the second region 320, and the third region 330, respectively, to focus the illumination light onto the wafer. Specifically, the micromirror array includes a micromirror 410 and a micro objective 420 arranged sequentially along the optical path; the micromirror 410 is used to redirect the illumination light to form parallel light, and the micro objective 420 is used to focus the parallel light onto the wafer. More specifically, in this embodiment, the micromirror array is provided in three groups, corresponding to the first region 310, the second region 320, and the third region 330, respectively. Taking the first region 310 as an example, the illumination light becomes a first beam after passing through the first region 310. At this time, the first beam has an angle with the optical axis. If it is directly irradiated onto the wafer, scattering will occur, affecting the focusing process. Therefore, a micro-reflector 410 is provided. The angle of the micro-reflector 410 is adapted to the first region 310 so that the angle of the first beam changes after irradiating the micro-reflector 410, that is, there is no angle with the optical axis. At this time, the first beam is perpendicular to the wafer surface, forming parallel light parallel to the optical axis. At the same time, the micro-objective 420 can focus the parallel light parallel to the optical axis onto the wafer. In addition, the micro-objective 420 has different focal lengths, so that the light is focused to different axial depths of the wafer, realizing multi-focal synchronous illumination detection.

[0029] In some specific embodiments, the central axis of the wafer is divided into three parts: the uppermost part is defined as the surface layer of the wafer, the lowermost part is defined as the deep layer of the wafer, and the middle part is defined as the middle layer of the wafer. By moving the wafer in a direction perpendicular to the optical axis, the three focal points sequentially cover the same lateral position on the wafer surface.

[0030] In some embodiments, the defect detection system further includes a detector 700, multiple beam splitters 500, and multiple relay mirrors 600. The beam splitters 500 are positioned between the micromirror 410 and the microobjective 420, the detector 700 is positioned on one side of the beam splitter 500, and the relay mirrors 600 are positioned between the detector 700 and the beam splitter 500. Taking a first beam as an example, the first beam, after passing through the micromirror 410, forms parallel light parallel to the optical axis, which illuminates the beam splitter 500 and passes through it. The first beam is reflected by the wafer and returned along the original path. After passing through the micro-objective 420, it is reflected by the beam splitter 500 and then reflected by the beam splitter 500 to the relay mirror 600, where it is then focused by the relay mirror 600 onto the detector 700. The detector 700 has a detection surface, which is divided from top to bottom into a first photosensitive area, a second photosensitive area, and a third photosensitive area. Simultaneously, the beam splitter 500 corresponding to the first beam and the beam splitter 500 corresponding to the second beam are also included. The beam cube 500 and the beam split cube 500 corresponding to the third beam are arranged sequentially from top to bottom, and all three beam split cubes are within the range of the detection surface, so that the reflected light formed by the first beam can illuminate the first photosensitive area, the reflected light formed by the second beam can illuminate the second photosensitive area, and the reflected light formed by the third beam can illuminate the third photosensitive area. For example, the detector 700 has an M×N pixel array (such as 2048×2048), and its photosensitive surface is divided into a first photosensitive area, a second photosensitive area, and a third photosensitive area. The pixel range of the first photosensitive area is from the first row of pixels to the 640th row of pixels to receive the first beam; the pixel range of the second photosensitive area is from the 705th row of pixels to the 1344th row of pixels to receive the second beam; and the pixel range of the third photosensitive area is from the 1409th row of pixels to the 2048th row of pixels to receive the third beam. The pixels between adjacent areas are used as isolation, for example, the pixels from the 641st row of pixels to the 704th row of pixels are used as an isolation band to isolate the first photosensitive area and the second photosensitive area.

[0031] In some more specific embodiments, the detector 700 is electrically connected to a signal processing device, which may be a field-programmable gate array (FPGA), a digital signal processor (DSP), or a device capable of processing images. No specific limitation is made here. The appropriate signal processing device shall be selected according to the actual needs during the actual detection process.

[0032] The embodiments of this application also disclose a defect detection method for focal plane imaging. Specifically, the defect detection method includes the following steps: S1. Separate the illumination light into a first polarization optical path and a second polarization optical path; specifically, by setting a polarization module 200 with a polarization beam splitting interface in the optical path, the incident illumination light is decomposed into a first polarization optical path and a second polarization optical path by utilizing the difference in refractive index of the material for polarized light of different polarization optical paths or the interface reflection characteristics, thereby realizing the separation of the illumination light and providing a basis for the subsequent detection process.

[0033] S2. The first polarization optical path and the second polarization optical path are respectively irradiated onto different regions of the beam splitter 300 to obtain multiple sub-beams; specifically, the optical path after polarization separation is irradiated onto different regions of the beam splitter 300 to obtain multiple sub-beams, so that the first polarization optical path and the second polarization optical path can be irradiated to different axial depths of the wafer through the beam splitter 300; that is, by utilizing the physical position and film characteristics of different regions of the beam splitter 300, a single beam is converted into multiple spatially separated sub-beams, realizing the construction of a fan-shaped optical path, thereby eliminating the energy attenuation caused by multi-stage beam splitting.

[0034] S3. Focus multiple sub-beams at different axial depths of the wafer.

[0035] S4. Collect the reflected light formed by each sub-beam being reflected from the wafer, and redirect the multiple reflected beams to form multiple non-overlapping detection beams. Specifically, collect the light signals reflected from different depths of the wafer, and utilize the principle of optical path reversibility to allow the reflected light to return along the original path. By using a combination of beam splitter cube 500 and repeater mirror 600, the reflected light is converted into detection light, and the formed detection beams do not interfere with each other. This physically isolates the different detection beams on detector 700, avoiding image aliasing and ensuring image clarity and signal-to-noise ratio.

[0036] S5. Multiple detection lights are respectively irradiated onto different photosensitive areas of the detector 700, so that images of different focal planes are simultaneously imaged on different spatial positions of the detector 700 to obtain focal plane images corresponding to different focal planes; specifically, the detection lights are controlled to irradiate different photosensitive areas on the detector 700, thereby realizing synchronous exposure using multiple photosensitive areas of a single detector 700, thus realizing the simultaneous acquisition of images of different axial depths of the wafer, eliminating synchronization errors and calibration complexity in multi-camera systems, and reducing hardware costs.

[0037] S6. Perform geometric correction and pixel alignment on multiple focal plane images, and stack them along the axis to form a data volume. Obtain the three-dimensional spatial coordinates (xc, yc, zc) of the defect through connected component analysis. Specifically, due to the differences in field of view and magnification between different beams, it is necessary to correct each focal plane image and form a data volume based on multiple focal plane images. Finally, analyze the data volume using connected component analysis to restore the two-dimensional image to three-dimensional spatial information, namely the three-dimensional spatial coordinates (xc, yc, zc) of the defect.

[0038] In some embodiments, the first polarization path and the second polarization path are respectively irradiated onto different regions of the beam splitter 300 to obtain multiple sub-beams. This includes dividing the beam splitter 300 into regions, such that the beam splitter 300 has a first region 310, a second region 320, and a third region 330. Specifically, this can be achieved by combining three beam splitters 300 with different angles between them, where one beam splitter 300 forms the first region 310, another beam splitter 300 forms the second region 320, and the last beam splitter 300 forms the third region 330. Alternatively, the surface of the same beam splitter 300 can be processed to have the first region 310, the second region 320, and the third region 330. Specific settings are not limited here; in actual use, they can be configured according to different usage requirements. Specifically, the first region 310 is located on one side of the beam splitter 300, with an angle of 6-10° between its normal and the optical axis of the illumination light, and a transmittance of 50-70%. The second region 320 is located in the middle of the beam splitter 300, with an angle of 1-5° between its normal and the optical axis of the illumination light, and a transmittance of 20-40%. The third region 330 is located on the other side of the beam splitter 300, with an angle of 6-10° between its normal and the optical axis of the illumination light, and a transmittance of 5-15%. The first polarized light path is controlled to illuminate the first region 310 and the second region 320, forming the first beam and the second beam. The second polarized light path is controlled to illuminate the third region 330, forming the third beam.

[0039] In some embodiments, multiple sub-beams are focused at different axial depths of the wafer, including: setting micro-mirrors 410 and micro-objectives 420 along the optical path corresponding to the first region 310, the second region 320, and the third region 330; controlling the first beam, the second beam, and the third beam to be redirected to form parallel light after passing through the corresponding micro-mirrors 410, and then focused onto the wafer after passing through the micro-objectives 420; the first beam, the second beam, and the third beam have different optical path differences. And based on the optical path difference Determine the axial focusing position of each beam. ,

[0040]

[0041] Where t is the thickness of the beam splitter 300; θ is the angle between the corresponding region and the optical axis of the illumination light; n is the refractive index of the corresponding region; 420 represents the focal length of each micro objective; WD represents the working distance.

[0042] In some specific embodiments, since the first region 310, the second region 320, and the third region 330 have different tilt angles (i.e., different included angles) relative to the optical axis of the illumination light, the path length of the beam propagating within the medium changes when the beam passes through these regions. Specifically, the larger the tilt angle, the longer the refraction path of the beam inside the beam splitter 300, resulting in an additional optical path delay relative to the reference optical axis. Therefore, beams passing through different regions have different optical path differences. Optical path difference After being focused by the micro objective 420, it will be directly converted into an axial focusing position. Displacement along the optical axis. Therefore, based on the optical path difference... It can obtain the focusing position of the corresponding beam on the wafer. This allows the first, second, and third beams to be focused simultaneously on different depth layers inside the wafer, enabling the simultaneous acquisition of images of different depths of the wafer during the inspection process without the need for mechanical movement of the Z-axis, thus improving inspection efficiency.

[0043] In some embodiments, multiple detection beams are respectively irradiated onto different photosensitive areas of the detector 700, so that images of different focal planes are simultaneously imaged at different spatial positions of the detector 700 to obtain focal plane images corresponding to different focal planes. This includes dividing the detection surface of the detector 700 to have a first photosensitive area, a second photosensitive area, and a third photosensitive area; setting relay mirrors 600 along the optical path corresponding to the first area 310, the second area 320, and the third area 330; controlling the reflected light formed by the first beam to irradiate the first photosensitive area after passing through the corresponding relay mirror 600 to form a first focal plane image; controlling the reflected light formed by the second beam to irradiate the second photosensitive area after passing through the corresponding relay mirror 600 to form a second focal plane image; controlling the reflected light formed by the third beam to irradiate the third photosensitive area after passing through the corresponding relay mirror 600 to form a third focal plane image; specifically, since a beam splitter 500 is also set in the optical path... Therefore, the light beam passes through the beam splitter cube 500. Taking the first light beam as an example, the first light beam passes through the micro-reflector 410 and then illuminates the beam splitter cube 500. After passing through the beam splitter cube 500, it illuminates the micro-objective lens 420 and is focused on the deep layer of the wafer. After being reflected by the wafer, the first light beam returns along the original path and is refracted at the beam splitter cube 500, illuminating the first photosensitive area of ​​the detector 700, thereby enabling the detector 700 to output a first focal plane image. The first focal plane image, the second focal plane image, and the third focal plane image represent optical slicing information of the wafer at different axial depths. Specifically, the first focal plane image reflects the imaging information of the deep structure of the wafer and is used to detect deep voids, buried defects, etc.; the second focal plane image reflects the imaging information of the intermediate layer of the wafer and is used to detect interlayer peeling or internal cracks, etc.; the third focal plane image reflects the imaging information of the outermost surface of the wafer and is used to detect defects such as surface particles or scratches.

[0044] In some embodiments, geometric correction and pixel alignment are performed on multiple focal plane images, including based on the focal length of the relay lens 600. The focal length of the corresponding micro objective 420 The differences in magnification caused by these variations are addressed by digitally scaling and resampling the images on each focal plane to ensure a uniform pixel resolution across the first, second, and third focal plane images. Specifically, in the actual optical path design, the repeater lens has a focal length of 600... With the corresponding micro objective 420 The inherent differences in focal lengths lead to varying magnifications in the focal plane images across different channels. Without correction, directly stacking or stitching these images results in blurred edges and misaligned details. Therefore, digital scaling and resampling of each focal plane image are necessary to ensure that the first, second, and third focal plane images have the same physical pixel resolution, thus eliminating the overall size deviation caused by differences in hardware optical magnification.

[0045] In some specific embodiments, the field-of-view coordinates (x, y) of each focal plane image are used as variables, and the angle between each region and the optical axis of the illumination light is used as the basis. Calculate the lateral misalignment of pixels at the edge of the field of view caused by the fan-shaped optical path formed by the beam splitter 300. ,

[0046] Where y is the vertical pixel coordinate; The tangent of the angle between each region and the optical axis of the illumination light characterizes the degree of sector expansion; This refers to the magnification factor; specifically, to address the fan-shaped optical path distortion introduced by the beam splitter 300, precise calculation and correction of the field-of-view coordinates are performed. This embodiment introduces the field-of-view coordinates (x, y) as variables, utilizing the tangent of the angle between different regions and the optical axis. This value characterizes the degree of fan-shaped expansion; a larger value indicates more significant edge distortion. Additionally, it can be expressed using a magnification factor. This comprehensively reflects the scale changes during optical path transmission. Based on the above parameters, the lateral misalignment caused by the fan-shaped optical path formed by the beam splitter 300 at different longitudinal pixel coordinates y is calculated. This quantifies the degree of pixel phase offset at the edge of the field of view, solving the edge misalignment problem that traditional translation alignment cannot handle.

[0047] In some specific embodiments, based on the lateral misalignment amount A transformation matrix incorporating lateral translation and shearing components is constructed. Specifically, by introducing the shearing component, the divergence or convergence characteristics of light rays under the action of the beam splitter 300 are effectively simulated, thereby working in conjunction with lateral translation to eliminate edge misalignment. More specifically, based on the lateral misalignment amount... Build 3 The transformation matrix F of 3,

[0048]

[0049] Where s is the shear component; This is the lateral translation component; This represents the longitudinal translation component. A geometric transformation is performed on the field-of-view coordinates (x, y) based on the transformation matrix to obtain the corrected coordinates (x1, y1) to eliminate pixel misalignment at the field-of-view edges. Specifically... ; Thus, the corrected coordinates (x1, y1) for eliminating pixel misalignment at the edge of the field of view are obtained, which not only solves the alignment problem of the center of the field of view, but also corrects the nonlinear misalignment at the edge of the field of view, improving the accuracy of wide field of view imaging; and aligns the field of view boundaries of the first focal plane image, the second focal plane image and the third focal plane image to the same rectangular coordinate system.

[0050] In some embodiments, the data volume is formed by stacking along the axial direction, including using the first polarization optical path and the second polarization optical path as polarization state dimensions, combined with the correction coordinates (x1, y1) and the axial focusing position. Obtain the data volume V(x1,y1,z,φ), where φ represents the polarization state of the first polarization path or the second polarization path. Define the polarization state of the first polarization path as P and the polarization state of the second polarization path as S. The axial focusing position... The curing tilt angle of each region of the beam splitter 300 Focal length of the micro objective lens 420 This unique determination establishes the deep coordinates z1 corresponding to the first region 310, the mid-layer coordinates z2 corresponding to the second region 320, and the surface coordinates z3 corresponding to the third region 330. The deep image obtained by the first polarized light path through the first region 310 is used as the V(x1,y1,z1,P) data layer, the mid-layer image obtained through the second region 320 is used as the V(x1,y1,z2,P) data layer, and the surface image obtained by the second polarized light path through the third region 330 is used as the V(x1,y1,z3,S) data layer. These data layers are stacked along the z-axis and the polarization state to construct the required data volume. Specifically, the polarization state of the first polarized light path corresponds to the deep position of the first region 310 and the mid-layer position of the second region 320, and the polarization state of the second polarized light path corresponds to the surface position of the third region 330. The polarization state identifier distinguishes the scattering characteristics at different depths. In subsequent data processing, images under different polarization states can be analyzed independently or fused. For example, for defects with birefringence, the transmission intensity of light with different polarizations varies greatly. By comparing the polarization states of the first polarization path and the second polarization path, background noise can be effectively suppressed, thereby improving detection accuracy.

[0051] In some embodiments, the three-dimensional spatial coordinates of the defect are obtained through connected component analysis, including: Based on the data volume V(x1,y1,z,φ), calculate the polarization difference image D(x1,y1,z).

[0052] By utilizing the difference in depolarization characteristics between the first and second polarization optical paths, isotropic background scattering is suppressed, thereby enhancing the contrast of defects. A threshold for pixel grayscale is set to 0-225; in this embodiment, a threshold of 50 is selected. Threshold segmentation is performed based on the grayscale features of the polarization difference image D(x1,y1,z). Pixels with grayscale values ​​greater than or equal to the threshold are marked as foreground, and the rest are marked as background, resulting in a binarized three-dimensional defect mask. Pixels meeting the threshold are then labeled with three-dimensional connected components, i.e., six-neighbor connected component analysis is performed on the three-dimensional defect mask, grouping spatially adjacent pixels into the same defect connected component. The centroid coordinates (xc,yc) of each connected component and its axial distribution range are calculated; thereby identifying the lateral coordinates (xc,yc) of the defect and its depth layer zc, and obtaining the three-dimensional spatial coordinates (xc,yc,zc) of the defect; more specifically, based on the axial focusing position of the connected component... Determine the axial coordinate zc of the defect; if the connected domain spans multiple depth layers (deep, middle or surface), it is determined to be a cross-layer penetration defect; if it exists only in a single depth layer, it is determined to be an intra-layer defect; combine the lateral coordinate (xc, yc) and the depth layer zc in which it is located to obtain the three-dimensional spatial coordinates (xc, yc, zc) of the defect.

[0053] In some embodiments, the method further includes calculating the polarization degree M of each pixel based on the data volume V(x1,y1,z,φ).

[0054]

[0055] The polarization degree M characterizes the ability of this position to maintain the polarization state, with a value range of [0,1]. When the polarization degree M approaches 1, the polarization difference is large, the polarization retention is high, and it exhibits specular reflection characteristics. When the polarization degree M approaches 0, the polarization difference is small, depolarization occurs, and it exhibits volume scattering characteristics. The first threshold is set to 0.7, and the second threshold is set to 0.3. Based on the polarization degree M and the focusing position... Determine the type of defect; When the polarization degree M is greater than or equal to the first threshold and the focal position When corresponding to region 330, the defect is determined to be a specular reflection type defect, corresponding to scratches, metal residues, or contamination on the wafer surface. Its physical mechanism is manifested in the complete preservation of the polarization state by surface specular reflection; when the polarization degree M is less than or equal to the second threshold and the focusing position... When corresponding to the first region 310, the defect is determined to be a volume scattering type defect, corresponding to through-silicon via cracks, bulk voids, or deep structural fractures. Its physical mechanism is manifested as a depolarization effect caused by multiple scattering of light within the defect. When the polarization degree M is between the first and second thresholds, and the focusing position... When corresponding to the second region 320, it is determined to be a mid-layer interface defect, corresponding to interlayer peeling, dielectric layer fracture or interface debonding. Its physical mechanism is a mixture of partial specular reflection and partial volume scattering.

[0056] While embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations fall within the scope and spirit of the invention as set forth in the claims. Furthermore, the invention described herein may have other embodiments and can be implemented or carried out in various ways.

Claims

1. A defect detection method using multifocal plane imaging, characterized in that, Includes the following steps: The illumination light is separated into a first polarization path and a second polarization path; The first polarization optical path and the second polarization optical path are respectively irradiated onto different tilted regions of a monolithic beam splitter to obtain multiple sub-beams; The multiple sub-beams are focused at different axial depths of the wafer; The reflected light formed by the reflection of each sub-beam by the wafer is collected, and the multiple reflected lights are respectively redirected to form multiple non-overlapping detection lights; Multiple detection lights are respectively irradiated onto different photosensitive areas of the detector, so that images of different focal planes are simultaneously imaged on different spatial positions of the detector to obtain focal plane images corresponding to different focal planes; Multiple focal plane images are geometrically corrected and pixel aligned, and stacked along the axis to form a data volume. The three-dimensional spatial coordinates (xc, yc, zc) of the defect are obtained through connected component analysis.

2. The defect detection method according to claim 1, characterized in that, The step of illuminating different regions of a beam splitter with the first polarizing optical path and the second polarizing optical path respectively to obtain multiple sub-beams includes, The beam splitter is divided into three regions: a first region, a second region, and a third region. The first region is located on one side of the beam splitter, with an angle of 6-10° to the optical axis of the illumination light and a transmittance of 50-70%. The second region is located in the middle of the beam splitter, with an angle of 1-5° to the optical axis of the illumination light and a transmittance of 20-40%. The third region is located on the other side of the beam splitter, with an angle of 6-10° to the optical axis of the illumination light and a transmittance of 5-15%. The first polarization optical path is controlled to irradiate the first region and the second region, forming a first beam and a second beam; The second polarization path is controlled to illuminate the third region, forming a third beam.

3. The defect detection method according to claim 2, characterized in that, The step of focusing the multiple sub-beams onto different axial depths of the wafer includes, Micromirrors and microobjectives corresponding to the first region, the second region, and the third region are arranged along the optical path; The first beam, the second beam, and the third beam are controlled to be redirected to form parallel light after passing through the corresponding micro-reflectors, and then focused onto the wafer after passing through the micro-objective. The first beam, the second beam, and the third beam have different optical path differences. , Where t is the thickness of the beam splitter; The angle between the corresponding region and the optical axis of the illumination light; According to optical path difference Determine the axial focusing position of each beam. , Where n is the refractive index of the corresponding region; 1 is the focal length of each micro objective; WD is the working distance.

4. The defect detection method according to claim 3, characterized in that, The step of illuminating different photosensitive areas of the detector with multiple detection lights, so that images of different focal planes are simultaneously imaged on different spatial positions of the detector, to obtain focal plane images corresponding to different focal planes, includes: The detection surface of the detector is divided into a first photosensitive area, a second photosensitive area, and a third photosensitive area. A relay mirror corresponding to the first region, the second region, and the third region is set along the optical path; The reflected light formed by the first beam is controlled to illuminate the first photosensitive area after passing through the corresponding relay mirror, forming a first focal plane image; The reflected light formed by controlling the second beam is irradiated onto the second photosensitive area after passing through the corresponding relay mirror, forming a second focal plane image; The reflected light formed by the control of the third beam is irradiated onto the third photosensitive area after passing through the corresponding relay mirror, forming a third focal plane image.

5. The defect detection method according to claim 4, characterized in that, The step of performing geometric correction and pixel alignment on the multiple focal plane images includes, Based on the focal length of the relay mirror The focal length of the corresponding micro objective The difference in magnification caused by the difference in focal plane is addressed by digital scaling and resampling of each focal plane image to ensure that the first focal plane image, the second focal plane image, and the third focal plane image have a uniform pixel resolution. Using the field-of-view coordinates (x, y) of each focal plane image as variables, and based on the angle between each region and the optical axis of the illumination light... Calculate the lateral misalignment of pixels at the edge of the field of view caused by the fan-shaped optical path formed by the beam splitter. , Where y is the vertical pixel coordinate; The tangent of the angle between each region and the optical axis of the illumination light characterizes the degree of fan-shaped expansion; This is the magnification factor; According to the lateral misalignment Construct a transformation matrix that includes lateral translation and shearing components; Geometric transformation is performed on the field coordinates (x, y) according to the transformation matrix to obtain the corrected coordinates (x1, y1) to eliminate the pixel misalignment at the edge of the field of view.

6. The defect detection method according to claim 3, characterized in that, The data volume formed by stacking along the axial direction includes, Using the first polarization optical path and the second polarization optical path as polarization state dimensions, combined with the correction coordinates (x1, y1) and axial focusing position... Obtain the data volume V(x1,y1,z,φ), where φ represents the polarization state of the first polarization optical path or the polarization state of the second polarization optical path.

7. The defect detection method according to claim 6, characterized in that, The process of obtaining the three-dimensional spatial coordinates of the defect through connected component analysis includes, Based on the data volume V(x1,y1,z,φ), the polarization difference image D(x1,y1,z) is calculated. Threshold segmentation is performed based on the grayscale features of the polarization difference image D(x1,y1,z). Pixels that meet the threshold are labeled with three-dimensional connected components. The lateral coordinates (xc,yc) of the defect and its depth layer zc are identified, and the three-dimensional spatial coordinates (xc,yc,zc) of the defect are obtained.

8. The defect detection method according to claim 6, characterized in that, It also includes, Based on the data volume V(x1,y1,z,φ), calculate the polarization degree M of each pixel; Based on the degree of polarization M and the axial focusing position Determine the type of defect; When the polarization degree M is greater than or equal to the first threshold and the focusing position When the defect corresponds to the third region, it is determined to be a specular reflection type defect; When the polarization degree M is less than or equal to the second threshold and the focusing position When corresponding to the first region, the defect is determined to be a volume scattering type defect; When the degree of polarization M is between the first threshold and the second threshold, and the focusing position When it corresponds to the second region, it is determined to be a mid-layer interface defect.

9. A defect detection system using multifocal planar imaging, characterized in that, For implementing the defect detection method according to any one of claims 1-8, the defect detection system comprises, sequentially arranged along the optical path: The lighting module is used to emit illumination light; The polarization module has a polarization beam splitter interface for separating the illumination light into a first polarization path and a second polarization path; The beam splitter has a first region, a second region, and a third region, each region having a different tilt angle with the optical axis and a different transmittance; the first polarized light path is incident on the first region and the second region, and the second polarized light path is incident on the third region. A micromirror array includes micromirrors and microobjectives arranged sequentially along the optical path, with a beam splitter cubic arrangement between the micromirrors and the microobjectives. The micromirror array is respectively arranged corresponding to the first region, the second region, and the third region to focus the illumination light onto the wafer. A relay mirror is positioned between the detector and the beam splitter. as well as A detector used to receive signals reflected from a wafer.

10. The defect detection system according to claim 9, characterized in that, The micromirror is used to redirect the illumination light to form parallel light, and the micro objective is used to focus the parallel light onto the wafer.

Images