A method and system for detecting wafer parameters

Abstract

This application relates to a method and system for detecting wafer parameters. The detection method includes providing multi-wavelength sequential illumination; modulating the illumination light into multi-step phase-shifting structured light using a spatial light modulator and projecting it onto the wafer surface; acquiring the multi-step phase-shifting structured light reflected from the wafer surface using a broadband imaging unit to obtain a structured light image sequence, ensuring consistent pixel positions in the images under multi-wavelength sequential illumination; synchronously controlling the timing of wavelength switching of the illumination light, structured light modulation, and image acquisition to obtain a multi-wavelength structured light image sequence; performing multi-spectral film thickness inversion on the multi-wavelength structured light image sequence to obtain a film thickness distribution map; and performing multi-wavelength phase fusion three-dimensional reconstruction on the multi-wavelength structured light image sequence to obtain three-dimensional topography data. The detection system includes an illumination module, a spatial light modulator, a broadband imaging unit, and a data processing module. This invention can acquire the three-dimensional topography and surface film thickness of wafers, and is also suitable for topography detection of etching trenches with aspect ratios.

Description

Technical Field

[0001] This invention relates to the field of semiconductor testing technology, and in particular to a method and system for detecting wafer parameters. Background Technology

[0002] In advanced semiconductor processes (such as 3nm and below), the impact of process deviations on device performance is amplified exponentially. Therefore, online and full-field inspection during the process has become the core link in yield control.

[0003] In existing technologies, semiconductor process inspection technologies are mainly divided into two categories: The first category is offline high-precision inspection technology, represented by white light interferometers and critical dimension scanning electron microscopes (CD-SEM). Completing the inspection of an entire wafer often takes tens of minutes or even longer, which cannot meet the online monitoring requirements of "second-level" feedback during the process, and cannot provide three-dimensional morphology information. The second category is online optical inspection technology, mainly including laser interferometry endpoint detection and visible light-based machine vision systems. Laser interferometry endpoint detection can only achieve single-point or limited-point film thickness / endpoint monitoring, and cannot obtain morphology uniformity information of the entire wafer; for etching trenches with an aspect ratio greater than 5:1, multilayer film stacked structures, and microcracks under transparent films, it lacks sufficient penetration ability and phase sensitivity, resulting in signal loss at the bottom of deep trenches, and making it impossible to obtain complete three-dimensional morphology.

[0004] Therefore, it is necessary to provide a new method and system for detecting wafer parameters to solve the above-mentioned problems existing in the prior art. Summary of the Invention

[0005] The technical problem to be solved by this application is to provide a detection method and system capable of acquiring the three-dimensional morphology and surface film thickness parameters of a wafer.

[0006] To address the aforementioned technical problems, according to embodiments of this application, a method for detecting wafer parameters is provided, comprising the following steps: controlling an illumination module to sequentially switch illumination light of different wavelengths according to a preset timing sequence to provide multi-wavelength sequence illumination; modulating the illumination light into multi-step phase-shifting structured light using a spatial light modulator and projecting it onto the wafer surface; acquiring the multi-step phase-shifting structured light reflected from the wafer surface using a broadband imaging unit to obtain a structured light image sequence, ensuring consistent pixel positions in the images under the multi-wavelength sequence illumination; synchronously controlling the timing of wavelength switching of the illumination light, structured light modulation, and image acquisition to obtain a multi-wavelength structured light image sequence; performing multispectral film thickness inversion on the multi-wavelength structured light image sequence to obtain a film thickness distribution map; and performing multi-wavelength phase fusion three-dimensional reconstruction on the multi-wavelength structured light image sequence to obtain three-dimensional topography data.

[0007] According to an embodiment of this application, the control lighting module sequentially switches different wavelengths of illumination light according to a preset time sequence to provide multi-wavelength sequence illumination. This includes setting the wavelength of the illumination light to include multiple center wavelengths covering the visible light band and the short-wave infrared band; controlling the rotation of the filter wheel in the lighting module so that multiple filters disposed on the filter wheel sequentially enter the illumination light path, and loading the spatial light modulator with the corresponding wavelength encoding pattern.

[0008] According to an embodiment of this application, the step of modulating the illumination light into multi-step phase-shift structured light using a spatial light modulator and projecting it onto the wafer surface includes: obtaining a binary grating pattern based on the encoding pattern; changing the lateral position of the bright stripes in the binary grating pattern to obtain a multi-step phase-shift encoding pattern with different phase offsets; modulating the micromirror array of the spatial light modulator according to the multi-step phase-shift encoding pattern; spatially modulating the illumination light irradiated onto the spatial light modulator using the micromirror array, so that the illumination light is reflected at the micromirrors in the on state to form the multi-step phase-shift structured light; and controlling the multi-step phase-shift structured light to be projected onto the wafer surface after passing through an aperture.

[0009] According to an embodiment of this application, the broadband imaging unit is a quantum dot camera.

[0010] According to an embodiment of this application, the step of performing multi-wavelength phase fusion three-dimensional reconstruction on the multi-wavelength structured light image sequence to obtain three-dimensional topography data includes: obtaining a corresponding phase-shift image for each wavelength in the multi-wavelength structured light image sequence; calculating the wrapping phase value and average intensity value of each pixel in the phase-shift image to obtain a wrapping phase map and average intensity map corresponding to each wavelength; selecting two adjacent wavelengths to form a combined wavelength and calculating an equivalent composite wavelength; obtaining an absolute phase map corresponding to each wavelength based on the equivalent composite wavelength; converting the absolute phase map into a height coordinate map; and performing weighted average fusion on the height coordinate map to obtain the three-dimensional topography data.

[0011] According to an embodiment of this application, obtaining the absolute phase map corresponding to each wavelength based on the equivalent synthesized wavelength includes: selecting two adjacent wavelengths from multiple wavelengths in the multi-wavelength structured light image sequence and calculating the equivalent synthesized wavelength; performing phase unrolling using the wrapping phase map corresponding to the synthesized wavelength to obtain a relative phase map; and using the relative phase map as a guide to unroll the wrapping phase map corresponding to each wavelength to obtain the absolute phase map.

[0012] According to an embodiment of this application, the method further includes performing spatial statistical analysis on the three-dimensional topography data to extract global average deviation and radial gradient deviation; generating a first adjustment instruction based on the global average deviation to adjust the etching time or total gas flow rate of the etching process; and generating a second adjustment instruction based on the radial gradient deviation to adjust the spatial distribution of etching power or the spatial distribution of gas flow rate.

[0013] According to an embodiment of this application, the step of performing multispectral film thickness inversion on the multi-wavelength structured light image sequence to obtain a film thickness distribution map includes: extracting the light intensity values ​​of pixels at each wavelength from the multi-wavelength structured light image sequence to construct a measured reflectivity curve; establishing a thin-film optical interference model including film thickness, refractive index, and extinction coefficient; calculating the theoretical reflectivity based on the thin-film optical interference model and the current film thickness parameters to construct a theoretical reflectivity curve; calculating the deviation between the measured reflectivity curve and the theoretical reflectivity curve to construct a least squares error function; iteratively adjusting the film thickness parameters through a nonlinear optimization algorithm to minimize the value of the least squares error function and obtain the optimal film thickness estimate; and performing the above steps for each pixel to obtain the film thickness distribution map.

[0014] A wafer parameter detection system for implementing the above-described detection method, the detection system comprising: an illumination module including a light source, a filter wheel, and a drive unit, wherein the drive unit drives the filter wheel to rotate to switch the light source to illumination light of different wavelengths; a spatial light modulator disposed on one side of the illumination module to modulate the illumination light; a broadband imaging unit for acquiring structured light images reflected from the wafer surface; and a data processing module connected to the broadband imaging unit for processing the acquired image sequence to obtain three-dimensional topography data.

[0015] According to an embodiment of this application, the spatial light modulator includes a beam splitter cube, a modulation chip, and a projection optical component. The illumination light is guided by the beam splitter cube to the modulation chip, and after being reflected and modulated by the modulation chip, it is projected onto the wafer surface through the projection optical component.

[0016] By adopting the above technical solution, the illumination module, spatial light modulator, and broadband imaging unit work together to detect wafer parameters, including three-dimensional morphology parameters and film thickness parameters, while also being able to adapt to the morphology detection of etching trenches with aspect ratios; and by using a quantum dot camera to simultaneously capture visible light and short-wave infrared images on a single chip, ensuring pixel-level self-alignment of multispectral data and eliminating the registration error of traditional multi-camera systems. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the positional relationship of a wafer parameter detection system according to an embodiment of the present invention.

[0018] Figure 2 This is a step diagram of a wafer parameter detection method according to an embodiment of the present invention.

[0019] Figure label:

[0020] 100. Illumination module; 110. Light source; 120. Filter wheel; 130. Drive unit; 210. Beam splitter cube; 220. Modulation chip; 230. Projection optical assembly; 231. First lens; 232. Second lens; 233. Aperture; 234. Projection objective lens; 300. Broadband imaging unit. Detailed Implementation

[0021] 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.

[0022] The following is in conjunction with the appendix Figure 1-2 The specific embodiments of the present invention will be further described in detail below.

[0023] An embodiment of the present invention provides a wafer parameter detection system for detecting parameters of a wafer during the processing. The wafer parameter detection system includes an illumination module 100, a spatial light modulator, a broadband imaging unit 300, and a data processing module.

[0024] In some embodiments, the lighting module 100 includes a light source 110, a filter wheel 120, and a drive unit 130. The light source 110 provides broadband illumination light. The filter wheel 120 is disposed in the outgoing light path of the light source 110 and has multiple narrowband filters evenly spaced circumferentially on it. The drive unit 130 is mechanically connected to the filter wheel 120 and drives it to rotate, causing the multiple narrowband filters on the filter wheel 120 to sequentially enter the illumination light path, thereby switching the light source 110 to illumination light of different wavelengths and achieving multi-wavelength sequence illumination. In some specific embodiments, the drive unit 130 is a motor.

[0025] In some embodiments, a spatial light modulator is disposed on one side of the illumination module 100 to receive illumination light from the illumination module 100 and perform spatial modulation. Specifically, the spatial light modulator includes a beam splitter cube 210, a modulation chip 220, and a projection optics component 230. The beam splitter cube 210 is disposed between the emitting end of the illumination module 100 and the modulation chip 220 to guide the incident illumination light to the modulation chip 220; the modulation chip 220 is located in the transmission light path of the beam splitter cube 210 to modulate the illumination light transmitted through the beam splitter cube 210; the modulated illumination light is reflected by the beam splitter cube 210 to the projection optics component 230. Specifically, the modulation chip 220 is a DMD chip, which includes an independently flippable micromirror array. Spatial modulation of light intensity is achieved by controlling the flip angle of each micromirror, which is well known to those skilled in the art and will not be described in detail here. More specifically, the light source 110 is a xenon lamp. During the detection process, the light emitted by the illumination module 100 illuminates the beam splitter 210, passes through the beam splitter 210 and then illuminates the modulation chip 220. After being modulated by the modulation chip 220, the light is reflected back to the beam splitter 210 and then reflected by the beam splitter 210 to the projection optics component 230. After passing through the projection optics component 230, the light illuminates the wafer and is reflected by the wafer. The reflected light passes through the beam splitter 210 and is collected by the broadband imaging unit 300.

[0026] In some embodiments, the projection optics assembly 230 includes a first lens 231, a second lens 232, and a projection objective lens 234 arranged sequentially along the optical path. The first lens 231 is located between the exit end of the beam splitter 210 and the second lens 232, and the back focal plane of the first lens 231 coincides with the front focal plane of the second lens 232 to form a Fourier plane. An aperture 233 is provided at the Fourier plane to filter out diffraction stray light generated by the modulation chip 220, thereby improving the contrast of the projected pattern. The projection objective lens 234 is located between the exit end of the second lens 232 and the wafer, and is used to finally project the modulated structured light onto the wafer surface, thereby forming a structured light pattern on the wafer surface.

[0027] In some embodiments, a broadband imaging unit 300 is positioned above the wafer to acquire structured light images reflected from the wafer surface. The broadband imaging unit 300 is a quantum dot camera, with a focusing lens in front of it to facilitate image reception. The quantum dot camera has a single photosensitive chip covered with a quantum dot film, enabling simultaneous response to visible and short-wave infrared bands on a single chip, thereby achieving pixel-level alignment of multi-wavelength images. The broadband imaging unit 300 is electrically connected to a data processing module to transmit the acquired image sequence to the data processing module. The data processing module processes the acquired image sequence to obtain three-dimensional topographic data.

[0028] In some specific embodiments, the optical axis of the projection optical component 230 and the optical axis of the broadband imaging unit 300 are arranged in a near-coaxial manner, that is, the angle between the two is less than 15°, so that the structured light projected by the projection optical component 230 is received by the broadband imaging unit 300 after being reflected by the wafer surface.

[0029] This application also discloses a wafer parameter detection method, which is applicable to the aforementioned wafer parameter detection system. Specifically, the wafer parameter detection method includes the following steps:

[0030] S1. The control illumination module 100 switches different wavelengths of illumination light sequentially according to a preset time sequence to provide multi-wavelength sequential illumination. Specifically, through time-sequential spectral switching, it provides multi-dimensional optical information covering different penetration depths and phase sensitivities for subsequent three-dimensional reconstruction and film thickness inversion, so as to improve the high-precision measurement requirements of the surface morphology of high aspect ratio structures (such as the bottom of deep trenches) that are difficult to meet simultaneously with a single wavelength.

[0031] S2. The illumination light is modulated into multi-step phase-shift structured light by a spatial light modulator and projected onto the wafer surface. Specifically, compared with the traditional method of mechanically moving gratings, the multi-step phase-shift structured light modulated by the spatial light modulator can complete the pattern switching in a shorter time, avoid mechanical vibration errors, and thus improve the detection accuracy.

[0032] S3. The wide-spectrum imaging unit 300 acquires multi-step phase-shifting structured light reflected from the wafer surface to obtain a structured light image sequence, so that the pixel positions of the images under multi-wavelength sequence illumination are consistent. Specifically, the wide-spectrum imaging unit 300 can acquire structured light reflection signals under different wavelengths. More specifically, consistent pixel positions refer to pixel-level alignment.

[0033] S4. Synchronously control the timing of wavelength switching of illumination light, structured light modulation and image acquisition to obtain a multi-wavelength structured light image sequence; specifically, the wavelength switching of illumination light, structured light modulation and image acquisition within the same timing sequence can ensure that the multi-wavelength structured light image sequence has an accurate correspondence, thereby improving detection accuracy.

[0034] S5. Perform multispectral film thickness inversion on the multi-wavelength structured light image sequence to obtain the film thickness distribution map; specifically, by utilizing the different reflectivity information of the multi-wavelength structured light image sequence and performing inversion, the film thickness distribution on the wafer surface can be obtained. This achieves simultaneous online detection of morphology and film thickness.

[0035] S6. Perform multi-wavelength phase fusion three-dimensional reconstruction on the multi-wavelength structured light image sequence to obtain three-dimensional topography data; specifically, by performing multi-wavelength phase fusion three-dimensional reconstruction on the multi-wavelength structured light image sequence, more accurate wafer parameters can be obtained.

[0036] In some embodiments, the illumination module 100 is controlled to sequentially switch different wavelengths of illumination light according to a preset timing sequence to provide multi-wavelength sequence illumination. This includes setting the wavelength of the illumination light to include multiple center wavelengths covering the visible light band and the short-wave infrared band; for example, setting the visible light band to 450nm-650nm and the short-wave infrared band to 900nm-1700nm, selecting the number of wavelengths N≥3, such as three characteristic wavelengths of 635nm, 850nm, and 1550nm; controlling the rotation of the filter wheel 120 within the illumination module 100, the rotation speed of which can be set according to detection requirements, will not be elaborated here. The rotation of the filter wheel 120 causes multiple filters disposed on the filter wheel 120 to sequentially enter the illumination light path, enabling the spatial light modulator to load the encoding pattern of the corresponding wavelength. Specifically, the filters are narrowband filters with a bandwidth of 10nm-20nm. During the rotation of the filter wheel 120, a feedback signal is output after each wavelength stabilizes, so that the spatial light modulator can load the encoding pattern of the corresponding wavelength. Wavelength switching is achieved through the mechanical rotation of the filter wheel 120, and the switching speed can reach the millisecond level, thus meeting the real-time requirements of online detection.

[0037] In some embodiments, the illumination light is modulated into multi-step phase-shift structured light by a spatial light modulator and projected onto the wafer surface. This includes: obtaining a binary grating pattern according to a preset phase-shifting algorithm and an encoding pattern; obtaining a multi-step phase-shifting encoding pattern with different phase shifts by changing the lateral position of the bright stripes in the binary grating pattern; modulating the micromirror array of the spatial light modulator according to the multi-step phase-shifting encoding pattern; spatially modulating the illumination light irradiated onto the spatial light modulator by the micromirror array, so that the illumination light is reflected at the micromirrors in the on state to form multi-step phase-shift structured light; and controlling the multi-step phase-shift structured light to be projected onto the wafer surface after passing through aperture 233.

[0038] In some specific embodiments, the preset phase-shifting algorithm can employ three-step or four-step phase shifting, without limitation here; only four-step phase shifting is used as an example for explanation. The number of phase shifting steps M≥3, for example, when M=4, the phase step size is π / 2. The grating period P of the binary grating pattern is 32-128 pixels, and the lateral translation amount is P / M. Through this translation, multi-step phase shifting encoding patterns with phase offsets of 0, 2π / M, 4π / M, ..., 2π(M-1) / M are obtained. The spatial light modulator is specifically a DMD chip, which has a micromirror array. By adjusting the size and angle of the micromirrors in the micromirror array, spatial modulation of the illumination light is achieved. For example, the micromirror size is 5.4μm, and the micromirror flip angle is ±12°. The on / off state of the micromirrors is controlled by the binary grating pattern. The aperture 233 is specifically an adjustable pinhole spatial filter. Its aperture size can be dynamically adjusted according to the current illumination wavelength. For example, for a long wavelength of 1550nm, the aperture is adjusted to a diameter of 2mm-3mm to allow more low-frequency information to pass through; for a short wavelength of 635nm, the aperture is adjusted to a diameter of 1mm-1.5mm to filter out stray light from diffraction. This allows the multi-step phase-shifting structured light to be projected onto the wafer surface after passing through the aperture 233, ensuring that the bottom of the deep trench on the wafer surface can also receive high-contrast structured light illumination.

[0039] In some embodiments, the broadband imaging unit 300 is a quantum dot camera. Specifically, the quantum dot camera can respond to both visible light and short-wave infrared light simultaneously, thereby eliminating pixel mismatch caused by optical path difference and lens distortion in traditional multi-camera beam splitting schemes, thus ensuring that subsequent multi-wavelength image sequences can be directly fused at the pixel level.

[0040] In some embodiments, multi-wavelength phase fusion 3D reconstruction is performed on a multi-wavelength structured light image sequence to obtain 3D topography data. This includes: obtaining a phase-shifted image corresponding to each wavelength in the multi-wavelength structured light image sequence; calculating the wrapping phase value and average intensity value of each pixel in the phase-shifted image to obtain the wrapping phase map and average intensity map corresponding to each wavelength; selecting two adjacent wavelengths to form a combined wavelength and calculating the equivalent composite wavelength; obtaining the absolute phase map corresponding to each wavelength based on the equivalent composite wavelength; converting the absolute phase map into a height coordinate map; performing weighted average fusion on the height coordinate map to obtain 3D topography data, thereby obtaining relevant parameters of the 3D height / depth distribution.

[0041] In some specific embodiments, such as when a four-step phase shift is used, the formula is used to...

[0042] ;

[0043] Calculations are performed, in which, This represents the range of wrapper phase values ​​at pixel coordinates (x, y). π)-(+π); This represents the light intensity value at pixel (x,y) of the phase-shifted image at step n;

[0044] Equivalent composite wavelength for,

[0045] ;

[0046] in, and This indicates the selection of two different center wavelengths, such as selecting adjacent 635nm and 850nm;

[0047] The absolute phase map is obtained by phase unrolling using the wrapper phase map corresponding to the equivalent composite wavelength, and then converted into a height coordinate map according to the following mapping relationship;

[0048] ;

[0049] in, This represents the height value at pixel (x, y); Indicates the wavelength of the illumination light currently being used. or ; This represents the absolute phase value at pixel (x, y), i.e., the absolute phase value without a 2π transition;

[0050] The weights of the weighted average fusion can be dynamically allocated according to the modulation quality or signal-to-noise ratio of each wavelength at the current pixel point; thus, the short wavelength is used to ensure the longitudinal resolution, and the equivalent synthesized wavelength is used to solve the phase expansion ambiguity problem of deep trenches, thereby improving the detection accuracy.

[0051] In some embodiments, the absolute phase map corresponding to each wavelength is obtained based on the equivalent composite wavelength. This includes selecting two adjacent wavelengths from multiple wavelengths in a multi-wavelength structured light image sequence and calculating the equivalent composite wavelength. The phase transition period corresponding to the composite wavelength is greater than the maximum morphology height of the wafer under test. Phase expansion is performed using the wrapping phase map corresponding to the composite wavelength to obtain a relative phase map. Specifically, the wrapping phase map is obtained by subtracting two single-wavelength wrapping phase maps and taking the modulus of 2π (the remainder after dividing by 2π). At this time, the relative phase map has lower accuracy but the number of integer cycles is correct. Guided by the relative phase map, the integer number of cycles k that should be superimposed on each single-wavelength wrapping phase map is calculated, and the calculation result is rounded to obtain the integer number of cycles k.

[0052] ;

[0053] in, The phase value is obtained by expanding the equivalent composite wavelength, which has eliminated the 2π jump. The short wavelength, 635 nm, is used for the final accuracy measurement;

[0054] Unfold the wrapper phase diagrams corresponding to each wavelength to obtain the absolute phase diagram. Let the absolute phase diagram be... ;

[0055] ;

[0056] in, At pixel (x, y), wavelength The following package phase; At pixel (x, y), wavelength The corresponding stripe levels; At pixel (x, y), wavelength The absolute phase below.

[0057] The algorithm first uses a synthesized wavelength to resolve "integer ambiguity" and then uses a single wavelength to ensure accuracy, thus avoiding the problem that traditional single-wavelength unfolding algorithms are prone to errors on steep surfaces (such as the sidewalls of deep trenches).

[0058] In some embodiments, the method further includes performing spatial statistical analysis on the three-dimensional topography data to extract the global average deviation and radial gradient deviation; wherein, the global average deviation represents the average deviation of the entire topography data relative to the target value, and the radial gradient deviation represents the topography distribution trend along the wafer radius direction, such as the height difference between the center and the edge; specifically, the global average deviation is set as... ,

[0059] ;

[0060] Where N is the total number of pixels involved in the statistics, that is, the number of effective measurement points on the entire wafer; The etching depth value measured at a pixel (x, y) after multi-wavelength phase fusion 3D reconstruction; The target depth value at pixel (x, y);

[0061] A first adjustment command is generated based on the global average deviation to adjust the etching time or total gas flow rate of the etching process. A value greater than 0 indicates overall over-etching, which shortens the etching time or reduces the total gas flow rate; conversely, it extends the time or increases the flow rate. A second adjustment command is generated based on the radial gradient deviation to adjust the spatial distribution of etching power or gas flow rate. If the center is deep and the edge is shallow, the spatial distribution of etching power (such as reducing the edge ring power or increasing the center air intake flow rate) or gas flow rate is adjusted to improve etching uniformity.

[0062] In some embodiments, multispectral film thickness inversion is performed on multi-wavelength structured light image sequences to obtain film thickness distribution maps. This includes: extracting the light intensity values ​​of pixels at each wavelength from the multi-wavelength structured light image sequence to construct a measured reflectivity curve; establishing a thin-film optical interference model including film thickness, refractive index, and extinction coefficient; calculating the theoretical reflectivity based on the thin-film optical interference model and current film thickness parameters to construct a theoretical reflectivity curve; calculating the deviation between the measured reflectivity curve and the theoretical reflectivity curve to construct a least squares error function; iteratively adjusting the film thickness parameters through a nonlinear optimization algorithm to minimize the value of the least squares error function and obtain the optimal film thickness estimate; performing the above steps for each pixel to obtain a film thickness distribution map; wherein, the film thickness distribution map is based on the full-field film thickness spatial distribution data obtained by inversion from the optical interference model.

[0063] In some specific embodiments, the average intensity map at each wavelength is extracted from the multi-wavelength structured light image sequence, that is, the average reflected light intensity of each pixel at different phase shift steps at that wavelength, and a reflectivity curve is constructed for each pixel. The reflectivity curve is defined as R.

[0064] ;

[0065] in, The light intensity of the sample; For reference light intensity; It is dark current.

[0066] Simultaneously, the film thickness is defined as d, the refractive index as n, and the extinction coefficient as k; a thin-film optical interference model incorporating the film thickness d, refractive index n, and extinction coefficient b is established, which can be based on Fresnel equations, as is well known to those skilled in the art and will not be elaborated here; and the theoretical reflectivity R1 is calculated based on the current film thickness parameters. ,d,n,b), construct the theoretical reflectivity curve.

[0067] Calculate the deviation between the measured reflectance curve and the theoretical reflectance curve, construct the least squares error function, and define the least squares error function as follows: ,

[0068] ;

[0069] in, Indicates the wavelength of the illumination light currently being used. or .

[0070] The film thickness parameter d is iteratively adjusted using a nonlinear optimization algorithm, such as the Levenberg-Marquardt method (LM method), to minimize the value of the least squares error function, thus obtaining the optimal film thickness estimate. The above steps are then applied sequentially to each pixel to obtain a film thickness distribution map of the wafer surface, where the film thickness represents the thickness of the thin film material deposited or grown on the wafer surface. This enables online monitoring of film thickness uniformity during the deposition / etching process.

[0071] 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 method for detecting wafer parameters, characterized in that, Includes the following steps: The lighting control module switches different wavelengths of illumination light sequentially according to a preset time sequence to provide multi-wavelength sequential illumination; The illumination light is modulated into multi-step phase-shift structured light using a spatial light modulator and projected onto the wafer surface; The multi-step phase-shifting structured light reflected from the wafer surface is acquired by a broadband imaging unit to obtain a structured light image sequence, so that the pixel positions of the images under the multi-wavelength sequence illumination are consistent. The timing of wavelength switching, structured light modulation, and image acquisition of the illumination light is synchronously controlled to obtain a multi-wavelength structured light image sequence; Perform multispectral film thickness inversion on the multi-wavelength structured light image sequence to obtain a film thickness distribution map; Multi-wavelength phase fusion three-dimensional reconstruction is performed on the multi-wavelength structured light image sequence to obtain three-dimensional topography data.

2. The detection method according to claim 1, characterized in that, The lighting control module switches different wavelengths of illumination light sequentially according to a preset time sequence to provide multi-wavelength sequential illumination. include, The wavelength of the illumination light is set to include multiple center wavelengths covering the visible light band and the short-wave infrared band; The filter wheel inside the lighting module is controlled to rotate, so that multiple filters set on the filter wheel sequentially enter the lighting light path, and the spatial light modulator is loaded with the corresponding wavelength encoding pattern.

3. The detection method according to claim 2, characterized in that, The process of modulating the illumination light into multi-step phase-shift structured light using a spatial light modulator and projecting it onto the wafer surface includes: A binary grating pattern is obtained based on the encoding pattern. The lateral position of the bright stripes in the binary grating pattern is changed to obtain a multi-step phase shift encoding pattern with different phase offsets. The micromirror array of the spatial light modulator is modulated according to the multi-step phase-shift coding pattern; The illumination light irradiated onto the spatial light modulator is spatially modulated by the micromirror array, so that the illumination light is reflected at the micromirrors in the on state to form the multi-step phase-shifting structure light. The multi-step phase-shifting structured light is controlled to be projected onto the wafer surface after passing through the aperture.

4. The detection method according to claim 1, characterized in that, The broadband imaging unit is a quantum dot camera.

5. The detection method according to claim 1, characterized in that, The step of performing multi-wavelength phase fusion 3D reconstruction on the multi-wavelength structured light image sequence to obtain 3D topography data includes: Obtain the corresponding phase-shifted image based on each wavelength in the multi-wavelength structured light image sequence; The wrap-around phase value and average intensity value of each pixel in the phase-shifted image are calculated to obtain the wrap-around phase map and average intensity map corresponding to each wavelength; Select two adjacent wavelengths to form a combined wavelength, and calculate the equivalent combined wavelength; Obtain the absolute phase diagram corresponding to each wavelength based on the equivalent synthesized wavelength; The absolute phase map is converted into a height coordinate map, and the height coordinate map is weighted and averaged to obtain the three-dimensional topography data.

6. The detection method according to claim 5, characterized in that, The absolute phase diagram corresponding to each wavelength is obtained based on the equivalent synthesized wavelength. include, Two adjacent wavelengths are selected from the multiple wavelengths of the multi-wavelength structured light image sequence, and the equivalent composite wavelength is calculated. Phase unrolling is performed using the wrapped phase map corresponding to the synthesized wavelength to obtain a relative phase map; Guided by the relative phase diagram, the wrap-around phase diagrams corresponding to each wavelength are expanded to obtain the absolute phase diagram.

7. The detection method according to claim 1, characterized in that, It also includes, Spatial statistical analysis was performed on the three-dimensional topography data to extract the global average deviation and radial gradient deviation; A first adjustment command is generated based on the global average deviation to adjust the etching time or total gas flow rate of the etching process; A second adjustment command is generated based on the radial gradient deviation to adjust the spatial distribution of etching power or the spatial distribution of gas flow.

8. The detection method according to claim 1, characterized in that, The step of performing multispectral film thickness inversion on the multi-wavelength structured light image sequence to obtain a film thickness distribution map includes, The light intensity values ​​of pixels at each wavelength are extracted from the multi-wavelength structured light image sequence to construct a measurement reflectance curve; A thin-film optical interference model including film thickness, refractive index, and extinction coefficient is established. The theoretical reflectivity is calculated based on the thin-film optical interference model and the current film thickness parameters, and a theoretical reflectivity curve is constructed. Calculate the deviation between the measured reflectance curve and the theoretical reflectance curve, and construct a least squares error function; The film thickness parameters are iteratively adjusted using a nonlinear optimization algorithm to minimize the value of the least squares error function, thereby obtaining the optimal film thickness estimate. Perform the above steps for each pixel to obtain the film thickness distribution map.

9. A wafer parameter detection system, characterized in that, For implementing the detection method according to any one of claims 1-8, the detection system comprises: The lighting module includes a light source, a filter wheel, and a drive unit. The drive unit drives the filter wheel to rotate to switch the light source into different wavelengths of illumination light. A spatial light modulator is disposed on one side of the illumination module to modulate the illumination light; A broadband imaging unit is used to acquire structured light images reflected from the wafer surface; The data processing module, connected to the broadband imaging unit, is used to process the acquired image sequences to obtain three-dimensional topographic data.

10. The detection system according to claim 9, characterized in that, The spatial light modulator includes a beam splitter cube, a modulation chip, and a projection optical component. The illumination light is guided by the beam splitter cube to the modulation chip, and after being reflected and modulated by the modulation chip, it is projected onto the wafer surface through the projection optical component.

Images