An angular momentum-based wafer defect detection method and system

Abstract

The application relates to a wafer defect detection method and system based on angular momentum. The wafer defect detection method comprises the following steps: acquiring a multilayer detection light beam; calculating an angular position coordinate; dividing a plurality of concentric annular zones on a wafer surface; focusing the multilayer detection light beam on different concentric annular zones according to a topological charge value I and rotating the wafer to enable the multilayer detection light beam to circumferentially scan the corresponding concentric annular zones; collecting the detection light beam reflected by the wafer surface as signal light; performing interference on the signal light and laser to obtain a frequency domain modulation image; performing phase-locked demodulation on the frequency domain modulation image to obtain a phase signal corresponding to each topological charge value I; inversely calculating the axial depth Z of a defect according to the phase signal; and determining the three-dimensional space position of the defect on the wafer according to the axial depth Z. The wafer defect detection system comprises an illumination module, a lens module, a wafer stage and a detection module. The application can meet the imaging requirements of a large field of view and high resolution and can determine the defect position.

Description

Technical Field

[0001] This invention relates to the field of semiconductor inspection technology, and in particular to a wafer defect detection method and system based on angular momentum. Background Technology

[0002] In semiconductor manufacturing processes, online detection of surface and subsurface defects (such as particles, scratches, and voids) on wafers is a key step in ensuring yield.

[0003] In existing technologies, due to the limitation of the conservation of optical invariants, a single objective lens cannot simultaneously meet the imaging requirements of a large field of view (inspection of the entire wafer) and high resolution. In addition, when determining the location of defects, especially the depth of defects, it is necessary to perform three-dimensional scanning of the wafer, which requires moving the objective lens or the wafer to different focal planes multiple times. This is time-consuming, sensitive to vibration, and has low detection efficiency.

[0004] Therefore, it is necessary to provide a novel wafer defect detection method and system based on angular momentum 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 how to provide a wafer defect detection method and system based on angular momentum that meets the imaging requirements of large field of view and high resolution and can determine the defect location.

[0006] To address the aforementioned technical problems, according to embodiments of this application, a wafer defect detection method based on angular momentum is provided, comprising the following steps: converting a laser into a vortex beam carrying different topological charge values ​​I; obtaining a multilayer detection beam based on the vortex beam; acquiring the wafer's rotational state information; and calculating the angular position coordinates. Multiple concentric rings are divided on the wafer surface. Based on the topological charge value I, multiple layers of detection beams are focused onto different concentric rings, and the wafer is rotated to circumferentially scan the corresponding concentric rings. The detection beams reflected from the wafer surface are collected as signal beams. The signal beams are interfered with the laser to obtain interference signals. A frequency-domain modulated image is obtained based on the interference signals. The frequency-domain modulated image is phase-locked and demodulated to obtain phase signals corresponding to each topological charge value I. The axial depth Z of the defect is inverted based on the phase signals. The three-dimensional spatial position of the defect on the wafer is determined based on the axial depth Z.

[0007] According to an embodiment of this application, the step of converting the laser into a vortex beam carrying different topological charge values ​​I, and obtaining a multilayer detection beam based on the vortex beam, includes setting a spiral phase plate in the optical path of the laser, wherein the surface height of the spiral phase plate is... for, Where n is the refractive index of the spiral phase plate; The wavelength of the laser; The azimuth angle of the spiral phase plate is given; the spiral phase plate is rotated to generate a sequence of vortex beams with different degrees of spirality in sequence, and the topological charge value I of each vortex beam is marked in the frequency domain to obtain the multilayer detection beams with different topological charge values ​​I.

[0008] According to an embodiment of this application, focusing the multiple layers of detection beams onto different concentric rings based on the topological charge value I includes: setting an objective lens in the optical path of the detection beam; setting a central blocking aperture at the entrance pupil position of the objective lens to form an annular pupil on the objective lens; obtaining the focusing radius r based on the topological charge value I, and focusing the detection beam onto different concentric rings after passing through the annular pupil.

[0009] According to an embodiment of this application, acquiring a frequency-domain modulated image based on the interference signal includes converting the interference signal into a time-domain electrical signal using a photodetector, wherein the time-domain electrical signal contains characteristic frequencies corresponding to different topological charge values ​​I. The time-domain electrical signal is subjected to frequency domain transformation processing, based on the characteristic frequency. The time-domain electrical signal is separated into multiple independent signals corresponding to different topological charge values ​​I; the frequency-domain modulation image of the concentric ring band is constructed based on the independent signals.

[0010] According to an embodiment of this application, the step of performing phase-locked demodulation on the frequency-domain modulated image to obtain a phase signal with respect to each of the topological charge values ​​I includes generating a phase signal with respect to the characteristic frequency. The sinusoidal reference signal and cosine reference signal of the same frequency are mixed and multiplied with the independent signal to obtain a mixed signal; the mixed signal is filtered out by a low-pass filter to remove the high-frequency components and retain the DC component; the in-phase component X and the quadrature component Y are obtained from the DC component; the phase signal corresponding to the topological charge value I is calculated from the in-phase component X and the quadrature component Y.

[0011] According to an embodiment of this application, the step of inverting the axial depth Z of the defect based on the phase signal includes sequentially selecting two different topological charge values ​​I, defined as a first topological charge value I1 and a second topological charge value I2, and obtaining the corresponding first phase signal. Second phase signal ; ; ;in, Phase distortion caused by defective surface morphology; This is due to the inherent phase delay; Given the Rayleigh length; calculate the first phase signal. and the second phase signal phase difference , According to the phase difference The axial depth Z is inverted.

[0012] According to an embodiment of this application, determining the three-dimensional spatial position of the defect on the wafer based on the axial depth Z includes integrating the focusing radius r of the defect and the angular position coordinates. And the axial depth Z, to obtain the three-dimensional spatial coordinates. .

[0013] According to embodiments of this application, the wafer defect detection method further includes performing a Fourier transform on the independent signal to obtain the corresponding characteristic frequency. The main peak of the spectral line is obtained; the half-width of the spectral line and the sideband intensity ratio are obtained based on the main peak of the spectral line. If the half-width of the spectral line is less than 100 Hz and the sideband intensity ratio is less than 0.3, the defect is determined to be a spherical particle; if the half-width of the spectral line is greater than 100 Hz and the sideband intensity ratio is greater than 0.3, the defect is determined to be an orientation scratch; if the spectrum shows a multi-peak stray distribution, the defect is determined to be an internal cavity.

[0014] A wafer defect detection system based on angular momentum is provided for implementing the aforementioned wafer defect detection method. The wafer defect detection system includes: an illumination module for providing a vortex beam; a lens module including an objective lens and a central blocking aperture, wherein the central blocking aperture is located at the entrance pupil position of the objective lens so that the vortex beam forms a detection beam after passing through the objective lens and is focused onto the corresponding concentric ring; a wafer stage for carrying the wafer and rotating it at a constant angular velocity; and a detection module for acquiring and processing signals reflected from the wafer surface.

[0015] According to an embodiment of this application, the illumination module includes a light source and a spiral phase plate arranged sequentially along the optical path; the light source is used to emit laser light; the spiral phase plate is placed between the light source and the objective lens to convert the laser light into the vortex beam.

[0016] By adopting the above technical solution, introducing orbital angular momentum, and utilizing detection beams with different topological charge values ​​I, combined with the annular pupil formed by the central blocking aperture, it is possible to simultaneously satisfy the requirement of covering the wafer in the field of view and achieving high resolution during the detection process. In addition, based on the difference in phase response of the same defect to two detection beams with different topological charge values ​​I, the depth of the defect can be inverted, and the type of defect can be determined, thereby improving the detection efficiency. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the optical path of the detection beam in wafer imaging of a wafer defect detection system according to an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of the optical path of the optical signal received by the detection module of a wafer defect detection system according to an embodiment of the present invention.

[0019] Figure 3 This is a step diagram of a wafer defect detection method according to an embodiment of the present invention.

[0020] Figure label:

[0021] 100. Illumination module; 110. Light source; 120. Spiral phase plate; 130. RF modulator; 200. Lens module; 210. Objective lens; 220. Central blocking aperture; 300. Wafer stage; 400. Detection module; 500. Beam splitter; 600. Mirror. 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 embodiments of the present invention provide a wafer defect detection method and a wafer defect detection system based on angular momentum. The wafer defect detection system is used to implement the wafer defect detection method. Specifically, the wafer defect detection system includes an illumination module 100, a lens module 200, a wafer stage 300, and a detection module 400.

[0025] In some embodiments, the illumination module 100 is used to provide a vortex beam; specifically, the illumination module 100 includes a light source 110 and a spiral phase plate 120 arranged sequentially along the optical path; the light source 110 is a laser for emitting laser light; the spiral phase plate 120 is placed between the light source 110 and the lens module 200 to convert the laser light into a vortex beam. More specifically, the illumination module 100 also includes an RF modulator 130 and a driver; wherein, the spiral phase plate 120 is disposed between the light source 110 and the RF modulator 130, and the RF modulator 130 is used to load different RF carrier frequencies F onto the vortex beam for frequency domain marking, so that vortex beams with different topological charge values ​​carry different characteristic frequencies. The driving component is connected to the spiral phase plate 120 to drive the spiral phase plate 120 to rotate. Specifically, the driving component can be a motor or other device that can drive the spiral phase plate 120 to rotate.

[0026] In some embodiments, the lens module 200 includes an objective lens 210 and a central blocking aperture 220. The central blocking aperture 220 is located at the entrance pupil of the objective lens 210, so that the vortex beam, after passing through the objective lens 210, forms a detection beam and is focused onto the corresponding concentric ring. Specifically, the central blocking aperture 220 is a circular light-shielding plate, thereby blocking the center of the objective lens 210 to form an annular pupil on the objective lens 210. More specifically, the objective lens 210 is located between the central blocking aperture 220 and the wafer stage 300. The vortex beam, after passing through the objective lens 210, forms a detection beam, and the detection beam can be focused onto the wafer surface to facilitate the detection process. More specifically, the central blocking aperture 220 has a hole in its center so that the detection beam can illuminate the center of the wafer, thereby facilitating the wafer detection process.

[0027] In some embodiments, a beam splitter 500 is provided to facilitate the detection process. The beam splitter 500 is positioned between the spiral phase plate 120 and the light source 110, with its beam-splitting surface set at a 45° angle to the principal optical axis. After the laser light from the light source 110 illuminates the beam splitter 500, part of it passes through the beam splitter 500 and illuminates the spiral phase plate 120, forming a vortex beam, while the other part is reflected by the beam splitter 500 to form a reference beam.

[0028] In some embodiments, the wafer stage 300 is used to carry the wafer and rotate it at a constant angular velocity. Specifically, the wafer stage 300 is located below the objective lens 210 and arranged coaxially with the objective lens 210. A drive device and a rotary encoder are provided on the wafer stage 300. The drive device can be a motor or other device capable of driving the wafer stage 300 to rotate at a constant angular velocity. The rotary encoder is mounted on the rotating shaft of the wafer stage 300 and is used to output angular position coordinates in real time. .

[0029] In some embodiments, the detection module 400 is used to collect and process signals reflected from the wafer surface. Specifically, the detection module 400 includes a reflector 600 and a photodetector. The side of the beam splitter 500 illuminated by the laser emitted from the light source 110 is defined as the first side, and the side of the beam splitter 500 illuminated by the light reflected from the wafer is defined as the second side. The reflector 600 is disposed on the reflected light path of the first side, spaced apart from the first side. The photodetector is disposed on the reflected light path of the second side, spaced apart from the second side. The photodetector is used to receive the interference signal formed by the mixing of laser light and signal light. Specifically, the signal light formed by the reflection from the wafer surface returns along the original path, passing sequentially through the objective lens 210, the central blocking aperture 220, the radio frequency modulator 130, and the spiral phase plate 120 before entering the beam splitter 500. After being reflected by the beam splitter 500, it enters the photodetector. Simultaneously, the laser light reflected by the reflector 600 also enters the photodetector, thereby enabling the photodetector to receive both laser light and signal light. More specifically, the detection module 400 also includes a signal processing device, which is connected to the photodetector to receive and process the signal from the photodetector; more specifically, the photodetector can be an avalanche photodiode; the signal processing device can be a field-programmable gate array (FPGA) or a digital signal processor (DSP), and there are no restrictions here. The appropriate signal processing device will be selected according to the actual needs during the actual detection process.

[0030] Embodiments of the present invention also provide a wafer defect detection method based on angular momentum. Specifically, the wafer defect detection method includes the following steps: S1. The laser is converted into a vortex beam carrying different topological charge values ​​I, and a multi-layer detection beam is obtained based on the vortex beam. Specifically, the laser is processed to convert it into a vortex beam. At this time, the vortex beam carries different topological charge values ​​I, that is, the multi-layer detection beam formed by it also carries different topological charge values ​​I. The inherent correlation between the topological charge value and the optical parameters is used to perform axial detection of the wafer with a large field of view, so as to improve the problem that a large field of view and high resolution cannot be obtained at the same time in the traditional single illumination mode, and it is difficult to simultaneously cover the detection needs of the surface and the interior.

[0031] S2. Obtain the wafer's rotational state information and calculate the angular position coordinates. Specifically, this involves converting the mechanical rotational motion of the wafer into angular position coordinates. This establishes a benchmark for circumferential defect localization; and enables circumferential scanning using the rotation of the wafer in conjunction with the detection beam, thereby improving detection efficiency while ensuring a large field of view.

[0032] S3. Divide multiple concentric rings on the wafer surface. Based on the topological charge value I, focus the multilayer detection beam onto different concentric rings and rotate the wafer to make the multilayer detection beam scan the corresponding concentric rings circumferentially. Specifically, by utilizing the inherent relationship between the topological charge value and the focusing radius (described in detail later), multiple optically divided concentric rings are formed on the wafer surface. At the same time, the wafer rotation forms a spiral scanning trajectory to achieve full wafer surface coverage under radial scanning.

[0033] S4. Collect the detection beam reflected by the wafer surface as the signal light, interfere with the laser to obtain the interference signal, and obtain the frequency domain modulation image based on the interference signal; specifically, the detection beam reflected by the wafer carries defect information, this part is used as the signal light, and it is mixed with the laser for interference, thereby realizing parallel detection of multiple rings in the full field of view with a single detection channel.

[0034] S5. Perform phase-locked demodulation on the frequency domain modulated image to obtain the phase signal with each topological charge value I; specifically, by performing phase-locked demodulation on the frequency domain modulated image, extract the defect phase information with high signal-to-noise ratio and containing axial depth information to eliminate noise and provide a basis for defect axial analysis.

[0035] S6. Determine the axial depth Z of the defect based on the phase signal; S7. Determine the three-dimensional spatial position of the defect on the wafer based on the axial depth Z; specifically, by integrating the already determined radial and angular position coordinates. The axial depth determined by phase inversion is used to construct the three-dimensional Cartesian coordinate positioning of the defect, thereby determining the specific location of the defect.

[0036] In some embodiments, the laser is converted into vortex beams carrying different topological charge values ​​I, and a multi-layer detection beam is obtained based on the vortex beams. This includes setting a spiral phase plate 120 in the optical path of the laser, rotating the spiral phase plate 120 to sequentially generate a sequence of vortex beams with different degrees of spirality, and performing frequency domain marking on the topological charge value I of each vortex beam to obtain a multi-layer detection beam with different topological charge values ​​I.

[0037] Specifically, the surface height of the spiral phase plate 120 for,

[0038] Where n is the refractive index of the spiral phase plate 120; The wavelength of the laser; The azimuth angle of the spiral phase plate 120 is the polar coordinate angle measured in the plane of the plate with the optical axis as the center, ranging from 0 to 2π. The spiral phase plate 120 is a transparent optical element with a spiral stepped surface microstructure; the step height of its spiral stepped surface varies with the azimuth angle. The increase, that is, the optical path difference introduced for each revolution of the spiral phase plate 120, is The corresponding phase change is 2πI. By rotating the spiral phase plate 120 to different positions, the laser is irradiated onto steps of different heights on its surface. Each step of different height corresponds to a different topological charge value I, thus converting the laser into a sequence of vortex beams carrying different topological charge values ​​I. The radio frequency modulator 130 uses a frequency domain marker to mark the topological charge value I of each vortex beam in the sequence, resulting in multilayer detection beams with different topological charge values ​​I. More specifically, the radio frequency modulator 130 can be an acousto-optic modulator or an electro-optic modulator; the specific choice is not limited here but selected according to different needs in actual use. The radio frequency modulator 130 loads different radio frequency carrier frequencies F onto the vortex beams with different topological charge values ​​I, thus enabling each layer of detection beam to carry a unique frequency domain marker, facilitating subsequent identification and separation of signals from different concentric rings through spectral analysis. In other words, without increasing the physical aperture of the objective lens 210 or introducing complex freeform surface correction, high-resolution detection of the entire wafer can be achieved simply by adjusting the topological charge value of the beam, thus meeting the requirements of a large field of view and high resolution.

[0039] In some embodiments, the rotational state information of the wafer is obtained, and the angular position coordinates are calculated. This includes setting up a drive unit and a rotary encoder, for example, setting the rotary encoder resolution to 0.001° to accurately provide angular position coordinates. Feedback, while controlling the drive unit to maintain a constant angular velocity. The wafer stage is driven to rotate 300 degrees. The angular position coordinates are calculated based on the signal acquisition time t during constant-speed rotation. ,get ,in Take the equivalent angle within the range of 0-2π, that is, the total angle. Divide by 2π (i.e., 360 degrees) and take the remainder so that the final result always falls within the range of 0-2π.

[0040] In some embodiments, multiple concentric rings are divided on the wafer surface. Multilayer detection beams are focused onto different concentric rings according to the topological charge value I, allowing the multilayer detection beams to scan the corresponding concentric rings circumferentially. This includes setting an objective lens 210 in the optical path of the detection beam; setting a central blocking aperture 220 at the entrance pupil position of the objective lens 210 to form an annular pupil on the objective lens 210. Specifically, the central blocking aperture 220 is located at the entrance pupil position near the beam splitter 500 at the front end of the objective lens 210, thereby blocking the middle of the objective lens 210 and allowing only beams of a specific radius to pass through, forming the annular pupil; obtaining the focusing radius r according to the topological charge value I, and focusing the detection beam onto different concentric rings after passing through the annular pupil. The number and distribution of concentric rings on the wafer surface are determined according to different wafers; for example, for a 300mm diameter wafer, 6 concentric rings are set to cover the detection area from the center to the edge; simultaneously, the mapping relationship between the topological charge value I and the focusing radius r is...

[0041] Where f is the focal length of the objective lens (210). The wavelength of the laser. To detect the beam waist radius.

[0042] Different topological charge values ​​I are selected based on the spiral phase plate 120, for example, I = 0, 10, 20, 30, 40, 50, which correspond to focusing radii r of approximately 0 mm, 2.5 mm, 5.0 mm, 7.5 mm, 10.0 mm, and 12.5 mm, respectively. The spiral phase plate 120 is controlled to allow the laser to pass through steps of different heights on its surface, thereby forming vortex beams of different radii, thus generating a sequence of vortex beams corresponding to the topological charge value I. The blocking ratio of the central blocking aperture 220 is adjusted to 0.3. The value is 0.5, preferably 0.4, so that the detection beams carrying different topological charge values ​​I are focused onto different concentric rings, so as to achieve detection with a large field of view while maintaining high resolution.

[0043] In some embodiments, the detection beam reflected from the wafer surface is collected as the signal light, and the signal light is interfered with the laser to obtain an interference signal. A frequency-domain modulated image is then obtained based on the interference signal. This includes converting the interference signal into a time-domain electrical signal using a photodetector, the time-domain electrical signal containing characteristic frequencies corresponding to different topological charge values. ; Perform frequency domain transformation on the time-domain electrical signal, based on the characteristic frequency The time-domain electrical signal is separated into multiple independent signals corresponding to different topological charge values; a frequency-domain modulation image of concentric rings is constructed based on these independent signals. Specifically, the signal light reflected from the wafer surface carries the corresponding topological charge value I, and its contact with defects causes phase distortion. It is the phase distortion caused by the surface morphology of the defect. The signal light returns along its original path and interferes with the unconverted portion of the laser emitted by the light source 110—that is, the portion that did not pass through the spiral phase plate 120—at the second surface of the beam splitter 500, forming an interference signal. The interference signal is then converted into a time-domain electrical signal by a photodetector. Since different radio frequency carrier frequencies F have been applied to beams with different topological charge values ​​I in step S1 using the radio frequency modulator 130, and the wafer introduces a rotational Doppler shift proportional to the topological charge value I during rotation, this interference signal is generated. ,in, Specifically, after the laser emitted by the light source 110 passes through the spiral phase plate 120, its geometric phase modulation causes the laser to carry orbital angular momentum. Therefore, when the wafer rotates, its defects sweep across the detection beam, generating a rotational Doppler effect, introducing a rotational Doppler frequency shift proportional to the topological charge value I. Therefore, at this time, the radio frequency carrier frequency F and the rotating Doppler frequency shift Together, they form the characteristic frequency of the time-domain electrical signal. Frequency domain transformation processing, such as Fast Fourier Transform, is performed on time-domain electrical signals based on characteristic frequencies. The time-domain electrical signal is separated into multiple independent signals corresponding to different topological charge values ​​I. Specifically, the time-domain electrical signal is defined as V(t), and the independent signals are defined as... .

[0044]

[0045] in, For amplitude, The total phase includes the Gouy phase and the defective phase distortion.

[0046] The spectrum is obtained by performing a Fast Fourier Transform on the time-domain electrical signal V(t). Since each topological charge value I corresponds to a different characteristic frequency... The resulting spectrum exhibits multiple separate peaks, each peak corresponding to a topological charge value I. Finally, bandpass filtering is used to extract the frequency component corresponding to the specific topological charge value I from the spectrum, filtering out other topological charge values ​​I and noise components to obtain the independent signal. ,

[0047] For each topological charge value I, based on the amplitude and phase of the corresponding independent signal, combined with the angular position coordinates... The phase values ​​are mapped to the corresponding image grayscale values ​​of the concentric rings, thereby constructing a frequency-domain modulated image of the concentric rings. In other words, by converting the weak phase modulation caused by defects, which is difficult to detect directly, into an intensity modulation signal that can be photodetected, i.e., an interference signal, the subsequent detection process is facilitated.

[0048] In some embodiments, phase-locked demodulation is performed on the frequency-domain modulated image to obtain the phase signal with respect to each topological charge value I, including generating a phase signal with respect to the characteristic frequency. The sinusoidal and cosine reference signals of the same frequency are multiplied with independent signals to obtain a mixed signal. The mixed signal is then filtered by a low-pass filter to remove high-frequency components, retaining the DC component. The in-phase component X and quadrature component Y are obtained from the DC component. The phase signal corresponding to the topological charge value I is calculated based on the in-phase component X and the quadrature component Y. Specifically, the sinusoidal reference signal is set as V1(t), and the cosine reference signal is set as V2(t). The generated reference signals V1(t) and V2(t) are respectively...

[0049]

[0050] Each signal is multiplied by a frequency mixing operation with an independent signal to obtain a mixed signal. Specifically, this yields a first mixed signal M1(t) corresponding to the reference signal V1(t) and a second mixed signal M2(t) corresponding to the cosine reference signal V2(t).

[0051]

[0052] It includes a phase-dependent DC component. and and frequency of The high-frequency AC component is removed. The first mixed signal M1(t) and the second mixed signal M2(t) are passed through a low-pass filter to remove the high-frequency AC component and retain the DC component. The in-phase component X and the quadrature component Y are obtained based on the DC component.

[0053]

[0054] The phase signal corresponding to the topological charge value I is calculated based on the in-phase component X and the quadrature component Y. The phase signal corresponding to the topological charge value I is the total phase including the Gouye phase and the defect phase distortion. ,

[0055] in, It contains the Gouye phase component K, and the phase distortion caused by the surface morphology of the defect. and inherent phase delay .

[0056]

[0057]

[0058] Where Z is the axial depth of the defect. Where p is the Rayleigh length and p is the number of radial rings. The wavelength of the laser. The beam waist radius is used to detect the beam. Since the Gouye phase is only related to the topological charge I and the axial depth Z, the radial ring number p does not change with these parameters and is set to 0 here.

[0059] In some embodiments, the axial depth Z of the defect is inverted based on the phase signal, including: Two different topological charge values ​​I are selected sequentially, defined as the first topological charge value I1 and the second topological charge value I2, and the corresponding first phase signals are obtained. Second phase signal ;

[0060]

[0061] in, Phase distortion caused by defective surface morphology; This is due to the inherent phase delay; Rayleigh length; Calculate the first phase signal Second phase signal phase difference ,

[0062] Based on phase difference Inversion axial depth Z,

[0063] This allows the determination of the axial depth Z of the defect relative to the wafer surface. Specifically, by utilizing the dependence of the Gouye phase component K on the topological charge value I and the axial depth Z, the depth of the defect can be determined by comparing the phase response differences of the detection beams for different topological charge values ​​I without moving the objective lens 210, thus eliminating the need for three-dimensional scanning of the wafer.

[0064] In some embodiments, determining the three-dimensional spatial position of the defect on the wafer based on the axial depth Z includes, after determining the focusing radius r and the angular position coordinates... And after the axial depth Z, the focusing radius r and angular position coordinates of the integrated defect. And the axial depth Z, to obtain the three-dimensional spatial coordinates. .

[0065] In some embodiments, the wafer defect detection method further includes performing a Fourier transform on the independent signals to obtain the corresponding characteristic frequencies. The main peak of the spectral line is identified. Based on this main peak, the half-width of the spectral line and the sideband intensity ratio are obtained. The half-width is the frequency bandwidth corresponding to when the intensity of the main peak drops to half its maximum value, reflecting the frequency purity of the signal. The sideband intensity ratio is the ratio of the integral of the sideband intensity within ±50Hz of the center frequency of the main peak to the intensity of the main peak, reflecting the frequency modulation depth of the defect scattering. If the half-width is less than 100Hz and the sideband intensity ratio is less than 0.3, it indicates that the defect is a point-like, isotropic scatterer with a narrow-band single-peak characteristic in the spectrum, meaning the defect passes through the detection beam rapidly without significant velocity. If the spectral gradient is positive, the defect is identified as a spherical particle. If the half-width of the spectral line is greater than 100 Hz and the sideband intensity ratio is greater than 0.3, or if a symmetrical double-sideband structure appears, it indicates that the defect is linear with a significant velocity gradient, meaning that the tangential velocity of different parts of the scratch relative to the detection beam is different, resulting in frequency shift broadening, and the defect is identified as an oriented scratch. If the spectrum shows a multi-peak stray distribution, meaning that the mode purity is significantly reduced and non-integer harmonics or continuous broadening appear, it indicates that the defect causes complex multiple scattering and phase random walks. This is due to optical path disorder caused by multiple reflections inside the cavity, and therefore the defect is identified as an internal cavity. In other words, based on spectral analysis of independent signals, defect classification can be completed directly based on the time-domain signal of the photodetector without the need for additional imaging hardware.

[0066] 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 wafer defect detection method based on angular momentum, characterized in that, Includes the following steps: The laser is converted into a vortex beam carrying different topological charge values ​​I, and a multi-layer detection beam is obtained based on the vortex beam; Obtain the wafer's rotational state information and calculate its angular position coordinates. ; Multiple concentric rings are divided on the wafer surface. The detection beams of the multiple layers are focused onto different concentric rings according to the topological charge value I, and the wafer is rotated so that the detection beams of the multiple layers scan the corresponding concentric rings circumferentially. The detection beam reflected from the wafer surface is collected as a signal beam. The signal beam is then interfered with the laser to obtain an interference signal. A frequency-domain modulation image is obtained based on the interference signal. The frequency-domain modulation image is obtained by combining the amplitude and phase of the corresponding independent signal with the angular position coordinates. An image constructed by mapping phase values ​​to the corresponding image grayscale values ​​of concentric rings; The frequency domain modulated image is phase-locked demodulated to obtain the phase signal with respect to each of the topological charge values ​​I; The axial depth Z of the defect is inverted based on the phase signal; The three-dimensional spatial location of the defect on the wafer is determined based on the axial depth Z.

2. The wafer defect detection method according to claim 1, characterized in that, The process of converting laser light into vortex beams carrying different topological charge values ​​I, and obtaining multi-layer detection beams based on the vortex beams, includes: A spiral phase plate is disposed in the optical path of the laser, and the surface height of the spiral phase plate is... for, Where n is the refractive index of the spiral phase plate; The wavelength of the laser is [wavelength]. The azimuth angle of the spiral phase plate; Rotating the spiral phase plate sequentially generates a sequence of vortex beams with different degrees of spirality. The topological charge value I of each vortex beam is frequency-domain labeled to obtain multilayer detection beams with different topological charge values ​​I.

3. The wafer defect detection method according to claim 1, characterized in that, The step of focusing the multilayer detection beams onto different concentric rings according to the topological charge value I includes, An objective lens is placed in the optical path of the detection beam; A central blocking aperture is provided at the entrance pupil position of the objective lens to form an annular pupil on the objective lens; The focusing radius r is obtained based on the topological charge value I, and the detection beam is focused on different concentric rings after passing through the annular pupil.

4. The wafer defect detection method according to claim 1, characterized in that, The step of obtaining the frequency domain modulated image based on the interference signal includes, The interference signal is converted into a time-domain electrical signal using a photodetector. The time-domain electrical signal contains characteristic frequencies corresponding to different topological charge values ​​I. ; The time-domain electrical signal is subjected to frequency-domain transformation processing, based on the characteristic frequency. The time-domain electrical signal is separated into multiple independent signals corresponding to different topological charge values ​​I; The frequency domain modulation image of the concentric ring band is constructed based on the independent signals.

5. The wafer defect detection method according to claim 4, characterized in that, The step of performing phase-locked demodulation on the frequency-domain modulated image to obtain the phase signal with respect to each of the topological charge values ​​I includes, Generate with the characteristic frequency The sinusoidal reference signal and the cosine reference signal of the same frequency are respectively mixed with the independent signal to obtain a mixed signal; The mixed signal is filtered out of high-frequency components by a low-pass filter, while retaining the DC component; The in-phase component X and the quadrature component Y are obtained based on the DC component; The phase signal corresponding to the topological charge value I is calculated based on the in-phase component X and the quadrature component Y.

6. The wafer defect detection method according to claim 1, characterized in that, The axial depth Z of the defect is inverted based on the phase signal. include, Two different topological charge values ​​I are selected sequentially, defined as the first topological charge value I1 and the second topological charge value I2, and the corresponding first phase signals are obtained. Second phase signal ; in, Phase distortion caused by defective surface morphology; This is due to the inherent phase delay; Rayleigh length; Calculate the first phase signal and the second phase signal phase difference , According to the phase difference The axial depth Z is inverted.

7. The wafer defect detection method according to claim 3, characterized in that, The step of determining the three-dimensional spatial location of the defect on the wafer based on the axial depth Z includes, The focusing radius r and the angular position coordinates of the integrated defect And the axial depth Z, to obtain the three-dimensional spatial coordinates. .

8. The wafer defect detection method according to claim 4, characterized in that, It also includes, Perform a Fourier transform on the independent signals to obtain the corresponding characteristic frequencies. The main peak of the spectrum; The spectral line half-width and sideband intensity ratio are obtained based on the main peak of the spectral line. The spectral line half-width is the frequency bandwidth corresponding to when the intensity of the main peak of the spectral line drops to half of its maximum value. The sideband intensity ratio is the ratio of the integral of the sideband intensity within ±50Hz of the center frequency of the main peak of the spectral line to the intensity of the main peak of the spectral line. If the half-width of the spectral line is less than 100 Hz and the sideband intensity ratio is less than 0.3, then the defect is determined to be a spherical particle. If the half-width of the spectral line is greater than 100 Hz and the sideband intensity ratio is greater than 0.3, then the defect is determined to be an orientation scratch. If the spectrum shows a multi-peaked stray distribution, the defect is determined to be an internal void.

9. A wafer defect detection system based on angular momentum, characterized in that, For implementing the wafer defect detection method according to any one of claims 1-8, the wafer defect detection system comprises: Illumination module for providing vortex beams; The lens module includes an objective lens and a central blocking aperture. The central blocking aperture is located at the entrance pupil of the objective lens so that the vortex beam forms a detection beam after passing through the objective lens and is focused onto the corresponding concentric ring. A wafer stage is used to support a wafer and rotate it at a constant angular velocity. The detection module is used to collect and process signals reflected from the wafer surface.

10. The wafer defect detection system according to claim 9, characterized in that, The lighting module includes a light source and a spiral phase plate arranged sequentially along the optical path; The light source is used to emit laser light; The spiral phase plate is placed between the light source and the objective lens to convert the laser into the vortex beam.

Images