Method for acquiring space image of pattern wafer based on strict vector diffraction integral

Abstract

The invention discloses a method for acquiring a pattern wafer space image based on strict vector diffraction integration, which belongs to the technical field of wafer defect detection, and comprises the following steps: discretizing a wafer pattern into a plurality of discrete points, discretizing the shape of an illumination light source into a plurality of point light sources according to the type of the illumination light source, and obtaining the coordinates of a single point light source; for each point light source, acquiring a space image of a wafer pattern corresponding to each discrete point during illumination of the point light source by using the coordinate of the point light source through a strict vector diffraction integral model; the space images of the wafer patterns corresponding to all the point light sources are superposed to obtain the space image corresponding to the wafer patterns during partial coherent illumination, so that the space images of the wafer patterns can be accurately obtained, and the reliability and the efficiency of wafer defect detection are effectively improved; the urgent demand of the semiconductor industry on a high-precision and high-adaptability detection technology is met, and the improvement of the manufacturing quality of an integrated circuit chip is promoted.

Description

Technical Field

[0001] The present invention belongs to the technical field of wafer defect detection, and in particular relates to a method for acquiring a pattern wafer spatial image based on strict vector diffraction integration. Background Art

[0002] In today's digital age, integrated circuit chips are the core components of various electronic devices. Their reliability and performance directly determine the quality and function of electronic products. Integrated circuit chip defect detection has become a key link to ensure high-quality chip production, which mainly covers two important areas: electrical performance testing and machine vision inspection.

[0003] In the field of machine vision inspection, it is further divided into macro-level chip appearance inspection (AOI) and micro-level nano-level defect inspection. Macro-AOI can quickly detect obvious defects on the chip surface such as scratches, stains, and pin deformation, providing strong support for the initial screening of chips. Micro-level nano-level defect inspection focuses on extremely small flaws inside the chip. These nano-level defects may have a profound impact on the performance of the chip, so the detection difficulty is also higher.

[0004] In mainstream nano-level defect detection equipment, the working principle is that the light emitted by the light source first passes through a precise shaping device to adjust the light into a specific shape, and then focuses on the wafer. The light diffracts on the surface of the wafer, and the diffracted light is accurately transmitted to the camera plane through a high-precision microscope system, thereby obtaining a spatial image of the wafer pattern. This process seems simple, but it actually involves cutting-edge technologies in multiple fields such as optics and precision machinery. Slight deviations in each link may lead to inaccurate detection results.

[0005] Taking patent applications CN 102495535 A and CN 102323722 A as examples, the methods disclosed in the patent applications provide new ideas for obtaining aerial images to a certain extent, but they are still limited when applied to wafer defect detection. In actual nano-level defect detection, the accuracy of the aerial images obtained is still insufficient for detecting tiny defects, and the calculation process is relatively complex and time-consuming, making it difficult to meet the high efficiency requirements of large-scale production detection.

[0006] The golden die detection method is one of the commonly used algorithms in semiconductor defect detection. The core of this algorithm is to rely on a reference image (Golden image) as a comparison benchmark. By subtracting the target image from the model database established by the calculation results, the defect contrast can be significantly improved, thereby accurately obtaining the defect detection results. However, in some actual semiconductor detection application scenarios, the reference image is either difficult to obtain or the image quality obtained is poor and cannot be used as a reliable comparison basis. Therefore, the main technical difficulty currently faced by the golden die detection method is how to establish a highly stable and high-precision reference image.

[0007] In order to overcome this problem and establish a high-precision and highly stable reference image, the industry is currently often faced with the problem of having to calculate the field in an extremely strict way. One way is to stay away from the object that causes scattering to reduce the interference of scattering on the calculation results; another way is to perform calculations on a surface whose lateral size is much larger than the wavelength, which can simplify the calculation process to a certain extent. In certain specific cases, through clever mathematical derivation and application of physical principles, it is possible to obtain an analytical solution, thereby successfully avoiding the occurrence of the above-mentioned complex problems. However, in numerical methods, such as the finite-difference time-domain method (FDTD) or the finite element method (FEM), even with the use of supercomputers, it is difficult to achieve accurate calculations due to memory and calculation time limitations. Therefore, for the task of accurately calculating the wafer diffraction field at any position in space, it is urgent to explore a new way that can simplify the calculation process while not reducing the calculation accuracy. This has also become a key technical bottleneck that needs to be broken through in the current field of integrated circuit chip defect detection. Summary of the invention

[0008] In view of the above, the purpose of the present invention is to provide a method for obtaining a pattern wafer aerial image based on strict vector diffraction integral, which realizes the accurate acquisition of the wafer pattern aerial image, thereby effectively improving the reliability and efficiency of wafer defect detection, meeting the semiconductor industry's urgent demand for high-precision and high-adaptability detection technology, and promoting the improvement of the manufacturing quality of integrated circuit chips.

[0009] To achieve the above-mentioned object of the invention, an embodiment provides a method for acquiring an aerial image of a patterned wafer based on strict vector diffraction integration, comprising the following steps: Discretize the wafer pattern into multiple discrete points, and discretize the shape of the illumination light source into multiple point light sources according to the type of illumination light source and obtain the coordinates of a single point light source; For each point light source, the coordinates of the point light source are used to obtain the spatial image of the wafer pattern corresponding to each discrete point when the point light source is illuminating by a strict vector diffraction integral model; The aerial images of the wafer pattern corresponding to all point light sources are superimposed to obtain the aerial image of the wafer pattern corresponding to the partially coherent illumination.

[0010] Preferably, for each point light source, using its coordinates by a strict vector diffraction integral model to obtain an aerial image of the wafer pattern corresponding to each discrete point when illuminated by the point light source, including: According to the coordinates of the point light source, calculate the electric field distribution of the near-field diffraction of the light emitted by the point light source at N*M discrete points on the surface of the wafer pattern after being reflected by the wafer pattern. and magnetic field distribution ; Calculation of electric field distribution using Fourier optical imaging theory and magnetic field distribution Fourier transform results on the spectrum plane and , where the coordinates of the spectrum surface are Obtained by transformation of the global coordinate system (x, y); Calculate the transfer function of each polarization component of the electromagnetic field based on a rigorous vector diffraction integral model; From the transfer function and Fourier transform results and Multiply and calculate the diffraction pattern produced by a given aperture; Using Fourier optical imaging theory, the diffraction pattern is inversely transformed to obtain the electric field distribution in the far field. and magnetic field distribution , according to the electric field distribution The spatial image I(x, y) of the wafer pattern at each discrete point corresponding to the point light source is obtained.

[0011] Preferably, the electric field distribution of the near-field diffraction at N*M discrete points on the wafer pattern surface is calculated and magnetic field distribution ,include: Wafer patterning Perform rigorous numerical calculations , and the electric field distribution is obtained and magnetic field distribution ; in, Indicates the coordinate position is The wafer pattern of discrete point areas, Denotes rigorous numerical computation using the finite element method, finite-difference time-domain method, finite integration method, or rigorous coupled-wave method.

[0012] Preferably, the transfer function of each polarization component of the electromagnetic field is calculated based on a strict vector diffraction integral model, including: According to the propagation of electromagnetic field strictly follows Maxwell's equations, the vector Helmholtz equation can be derived from Maxwell's equations, and according to the vector Green's theorem, after a series of vector operations, any point light sourcei Observe point in space The electromagnetic field distribution is expressed as a superposition integral form: ; ; Where z is the propagation distance, is the wavelength of illumination light, represents the wave vector, and Represents observation points in space The electric and magnetic field distributions at i Expressed as i Point light sources, represents the electromagnetic field of n different polarization components, The electric fields of different polarization components representing near-field diffraction, represents the magnetic field of different polarization components corresponding to the near-field diffraction, represents the coefficient of the electric field component, represents the coefficient of the magnetic field component, represents the global coordinate system, represents the position of any observation point in space, j represents a complex factor; The above and Rewritten in the form of convolution, we get: ; ; in, represents the spread function of different electric field polarization components, The spread function representing the different polarization components of the magnetic field; The diffusion function and Perform Fourier transform to obtain the transfer functions of different electric field polarization components and different magnetic field polarization components. and .

[0013] Preferably, the transfer function and the Fourier transform result and Multiply the diffraction pattern produced by a given aperture by: The transfer functions of different electric field polarization components are and Multiply to get the electric field diffraction pattern , the transfer functions of different magnetic field polarization components are and Multiply to get the magnetic field diffraction pattern .

[0014] Preferably, the method further comprises: when the partially coherent light source composed of point light sources illuminates the wafer pattern obliquely, the spectrum plane moves backward along the light propagation direction, and a frequency shift amount is added to the spectrum for each point light source, and the frequency shift amount provided by the oblique illumination of a single point light source is: ; or ; in, is the refractive index of air, is the incident angle of the point light source onto the wafer, is the diffraction angle of the light diffracted by the wafer, is the wavelength of illumination light, Represents the frequency shift, the positive and negative in the formula depends on the single point light source i coordinates, adding the electromagnetic field transfer function of the illumination angle frequency shift , becomes , .

[0015] Preferably, the method further comprises: when the partially coherent light source composed of point light sources illuminates the wafer pattern vertically, the spectrum plane remains unchanged, and the frequency shift range for all point light sources is calculated according to the internal and external coherence coefficients: ; in, is the refractive index of air, NA is the numerical aperture of the objective lens, is the internal coherence factor, is the external coherence factor, is the wavelength of illumination light, The electromagnetic field transfer function represents the frequency shift range of all point light sources i, adding the frequency shift of the lighting angle , becomes , .

[0016] Preferably, the method further comprises: transforming the spatial image of the wafer pattern corresponding to the partially coherent illumination to the image plane through a conversion matrix between global coordinates and image plane coordinates, so as to obtain an image surface result of the wafer pattern on the image plane.

[0017] Compared with the prior art, the present invention has the following beneficial effects: The present invention discretizes the illumination light source into multiple point light sources, and calculates the imaging in the corresponding space for different point light sources respectively, which has the advantage of high precision. This method can be applied to light sources of different shapes and sizes, and meets the wafer pattern simulation requirements of 28nm and below technology nodes.

[0018] The present invention establishes a matrix-form analytical expression of the spatial image under the vector diffraction imaging model, which does not require multiple grid divisions, is conducive to the rapid calculation of the strict vector diffraction integral model, and realizes rapid and accurate spatial imaging of wafer patterns at discrete points. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 is a flow chart of a method for acquiring a pattern wafer aerial image based on strict vector diffraction integration provided by an embodiment; Figure 2 It is a schematic diagram of an image formed at a position corresponding to the wafer pattern after the emitted light of the point light source provided in the embodiment is obliquely illuminated and reflected by the microscope objective lens; Figure 3 It is a schematic diagram of the emission light of the point light source provided in the embodiment vertically illuminating the wafer pattern and, after being reflected, being collected by the microscope objective lens and imaging at the position corresponding to the wafer pattern; Figure 4 is a schematic diagram of calculating the amount of frequency movement of a point light source added according to an embodiment; Figure 5 is a schematic diagram comparing transfer functions of various polarization components provided in the embodiments; Figure 6 is a schematic diagram of an initial binary wafer pattern provided in an embodiment; Figure 7 It is a schematic diagram of the total light intensity of the diffraction near field of the wafer pattern based on the strict vector diffraction integral under the illumination of the point light source provided in the embodiment; Figure 8 It is a schematic diagram of obtaining a spatial image of a pattern wafer by diffraction integral under illumination of a point light source provided in an embodiment. DETAILED DESCRIPTION

[0021] To make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation methods described herein are only used to explain the present invention and do not limit the scope of protection of the present invention.

[0022] The inventive concept of the present invention is: in the current mainstream nano-level defect detection, the traditional scalar approximate imaging model fails to fully consider the three-dimensional diffraction electric field generated by the wafer pattern and its transfer function in the system, resulting in deviations in the calculation of the far-field spatial image, making it difficult to accurately reflect the actual situation of the wafer, affecting the accuracy of defect detection. At the same time, faced with diverse wafer patterns and complex detection systems, traditional methods are difficult to adapt flexibly, and their application scope is relatively narrow. Based on this, in order to solve the problems of insufficient calculation accuracy and limited applicability in existing wafer defect detection technologies. An embodiment of the present invention provides a method for obtaining a pattern wafer spatial image based on a strict vector diffraction integral, which achieves accurate acquisition of a pattern wafer spatial image, and has wide adaptability and is not restricted.

[0023] Based on the above invention concept, the embodiment provides a method for obtaining a pattern wafer spatial image based on strict vector diffraction integral, which is used to perform spatial imaging of wafer patterns. The simple principle of the entire spatial imaging is as follows: Figure 2 and Figure 3 As shown, 201 represents an inclined illumination light source; 202 represents a spectrum plane, which can be the rear focal plane of the objective lens, or moved to other positions through a relay lens; 203 is a beam splitter, which is characterized by separating the beam intensity according to different proportions; 204 represents a partially coherent light source, which can be a ring illumination, a dipole illumination, a quadrupole illumination, etc.; 205 is a condenser. The patterned wafer surface is illuminated with inclined illumination and annular illumination light respectively, and the scattered light in a specific angle range of the wafer surface is received by the entrance pupil, propagated to the spectrum plane to complete Fourier transform, and then emitted through the exit pupil, and finally projected onto the image plane.

[0024] Assume that the optical axis direction is the z-axis, and establish a spatial global coordinate system (x, y, z) with the z-axis as the thumb direction according to the left-hand coordinate system principle. Assume that the global coordinates of any point on the illumination light source surface are (x 0 ,y 0 ,z 0 ), the light emitted by the point light source is incident on the wafer surface (x s ,y s ,0), the direction cosine of the plane wave emitted from the wafer surface and incident on the entrance pupil of the microscope objective is ( α s , β s , γ s ), then the spectrum surface coordinate system ( f x ，f y ) and the direction cosines are: ; Assume that the global coordinates of any point on the image plane are (x 1 ,y1 , z 1 ), the direction cosine of the plane wave incident from the spectrum plane to the image plane is ( α i ，β i , γ i ),in,( α i ，β i , γ i ) is the global coordinate system on the image plane (x 1 ,y 1 , z 1 ) is the coordinate system after Fourier transformation.

[0025] Based on the above wafer pattern wafer aerial image imaging principle, a method for obtaining a pattern wafer aerial image based on strict vector diffraction integral is implemented, comprising the following steps: S1, discretize the wafer pattern into multiple discrete points, and according to the type of the illumination light source, discretize the shape of the illumination light source into multiple point light sources and obtain the coordinates of a single point light source.

[0026] In the embodiment, the wafer pattern A is discretized into N*M discrete points, each of which corresponds to a sub-area, and the wafer pattern on it diffracts the received light. At the same time, since the illumination light sources corresponding to different light source types may have various shapes, it is necessary to discretize them according to the shape of the illumination light source. The shape of the partially coherent illumination light source is discretized into multiple point light sources and the light source coordinates (x 0 ,y 0 ,z 0 ).

[0027] S2, for each point light source, using its coordinates by a strict vector diffraction integral model to obtain a spatial image of the wafer pattern corresponding to each discrete point when the point light source is illuminating.

[0028] In the embodiment, the spatial image of the wafer pattern corresponding to each discrete point when illuminated by a point light source is obtained by a strict vector diffraction integral model, including: S2-1, according to the point light source coordinates (x 0 ,y 0 ,z 0 ), calculate the electric field distribution of the near-field diffraction of the light emitted by the point light source at N*M discrete points on the surface of the wafer pattern after being reflected by the wafer pattern and magnetic field distribution , including: Wafer patterning Perform rigorous numerical calculations , and the electric field distribution is obtained and magnetic field distribution ; in, Indicates the coordinate position is The wafer pattern of discrete point areas, Indicates strict numerical calculations, using the finite element method (FEM), finite difference time domain method (FDTD), finite integration technique (FIT), or rigorous coupled wave analysis (RCWA), where the electric field distribution and magnetic field distribution They are all vector matrices of N*M*3, and 3 represents the three polarization components in the x, y, and z directions in the diffraction field.

[0029] S2-2, Calculation of electric field distribution using Fourier optical imaging theory and magnetic field distribution Fourier transform results on the spectrum plane and , where the coordinates of the spectrum surface are Obtained by transformation of the global coordinate system (x, y).

[0030] In the embodiment, the electric field distribution and magnetic field distribution Perform Fourier transform and get the Fourier transform result and , the Fourier transform result and Used for subsequent multiplication with the transfer function to calculate the diffraction pattern, where the coordinates of the spectrum plane are Obtained by transformation of the global coordinate system (x, y).

[0031] S2-3, calculate the transfer function of each polarization component of the electromagnetic field based on the strict vector diffraction integral model.

[0032] Specifically, the propagation of electromagnetic fields strictly follows Maxwell's equations, from which the vector Helmholtz equations can be derived. According to the vector Green's theorem, after a series of vector operations, any point light source can be i Observe point in space The electromagnetic field distribution is expressed as a superposition integral form: ; ; Where z is the propagation distance, is the wavelength of illumination light, represents the wave vector, and Represents observation points in space The electric and magnetic field distributions at i Expressed as i Point light sources, Represents the electromagnetic field of n different polarization components. Note that the electromagnetic field is a vector and can be decomposed into the electromagnetic field of n different polarization components as needed. The electric fields of different polarization components representing near-field diffraction, represents the magnetic field of different polarization components corresponding to the near-field diffraction, represents the coefficient of the electric field component, Represents the coefficient of the magnetic field component. This coefficient can be a constant or a variable related to global coordinates, spectrum coordinates, etc. represents the global coordinate system, represents the position of any observation point in space, j represents a complex factor; Since the numerical aperture NA of the objective lens in the optical imaging system is small, generally less than 0.6, the distance approximation between the observation point and the object plane is introduced. These approximations are based on the binomial expansion. Considering the relevant distance in the electromagnetic field component coefficient, the direction cosine terms are binomially expanded and approximated. Specifically, these approximations are based on the mathematical formula Taylor expansion, as follows: ; in , is a vector; In this way, the above and Rewritten in the form of convolution, we get: ; ; in, represents the spread function of different electric field polarization components, The spread function representing the different polarization components of the magnetic field; The diffusion function and Perform Fourier transform to obtain the transfer functions of different electric field polarization components and different magnetic field polarization components. and .

[0033] The equation sources for calculating the transfer function include vector Rayleigh-Sommerfeld diffraction integrals, or Stratton-Chuintegrals, or Hertz vector diffraction integrals, photolithography Franz diffraction integrals. These equations can all derive different forms of calculation transfer functions, and these transfer functions can all be applied to the calculation framework proposed in the present invention. The selection of a specific transfer function depends on the actual application.

[0034] S2-4, from the transfer function and Fourier transform results and Multiply and calculate the diffraction pattern produced by a given aperture.

[0035] In the embodiment, the transfer function and the Fourier transform result are and Multiply the diffraction pattern produced by a given aperture by: and Multiply to get the electric field diffraction pattern , the transfer functions of different magnetic field polarization components are and Multiply to get the magnetic field diffraction pattern .

[0036] S2-5, using Fourier optical imaging theory, the diffraction pattern is inversely Fourier transformed to obtain the electric field distribution in the far field and magnetic field distribution , according to the electric field distribution The spatial image I(x, y) of the wafer pattern at each discrete point corresponding to the point light source is obtained.

[0037] In the embodiment, the diffraction pattern and Perform inverse Fourier transform to obtain the electric field distribution on the image plane and magnetic field distribution , according to the electric field distribution Obtain the spatial image I(x, y) of the wafer pattern at each discrete point corresponding to the point light source. and They are all vector matrices of N*M*3, where 3 represents the three polarization components in the x, y, and z directions in the diffraction field. The far-field distribution is characterized by being the electromagnetic field distribution at the focal plane of the optical system, at the defocused plane, or at the entrance pupil plane of the objective lens.

[0038] S3, superimposing the aerial images of the wafer pattern corresponding to all the point light sources to obtain the aerial image of the wafer pattern corresponding to the partially coherent illumination.

[0039] In the embodiment, it is determined whether the spatial images of the wafer pattern corresponding to all point light sources have been calculated. If so, step S3 is executed, otherwise, step S2 is returned. In S3, the spatial images I(x, y) corresponding to each point light source are superimposed according to the Abbe method to obtain the spatial image I corresponding to the pattern wafer position when the partially coherent light source is illuminated. total .

[0040] Based on the above analysis, a matrix-based analytical expression of the spatial image under the vector diffraction imaging model is established, and the overall imaging process can be described as: ; ; That is, the near-field electric field distribution and magnetic field distribution Perform Fourier Transform , and multiply it by the corresponding transfer function, and then perform inverse Fourier transform on the product of the two , and finally sum the different components and multiply them by a fixed value , and the electric field distribution on the image plane is obtained and magnetic field distribution , the amount of movement frequency caused by oblique lighting and the shape of the lighting source is reflected in the transfer function )and Specifically, each discrete point light source will affect the transfer function , , the range (f x , f y ) generates a frequency shift , The specific calculation methods include: like Figure 4 As shown, when the partially coherent light source is obliquely illuminated, the light source is discretized into multiple point light sources according to the shape of the partially coherent light source, and the spectrum surface moves backward along the light propagation direction, that is, from the initial position 301 to the position 302, and a frequency shift is added to each point light source on the spectrum. The frequency shift provided by the oblique illumination of a single point light source is: ; or ; in, is the refractive index of air, is the incident angle of the point light source onto the wafer, is the diffraction angle of the light diffracted by the wafer, is the wavelength of illumination light, Represents the frequency shift, the positive and negative in the formula depends on the single point light source i coordinates, adding the electromagnetic field transfer function of the illumination angle frequency shift , becomes , .

[0041] When the partially coherent light source is vertically illuminated, the light source is discretized into a plurality of point light sources according to the shape of the partially coherent light source, and the frequency shift of all the point light sources in the partially coherent light source is calculated based on the internal and external coherence coefficients 303 and 304. : ; in, is the refractive index of air, NA is the numerical aperture of the objective lens, is the internal coherence factor, is the external coherence factor, is the wavelength of illumination light, The electromagnetic field transfer function showing the frequency shift and adding the frequency shift of the lighting angle , becomes , .

[0042] like Figure 5 As shown, 401 is Quantity Correspondence Quantity, Quantity Correspondence The transfer function of the component, 402 is Quantity Correspondence The transfer function of the component, 403 is Quantity Correspondence The transfer function of the component, where 1 represents complete transmission and 0 represents complete absorption, can be seen that the transfer functions corresponding to different polarization components are significantly different.

[0043] Figure 6 Schematic diagram of the initial binary wafer pattern, with a critical dimension of 200nm, 1 represents a protrusion, 0 represents a groove, the wafer groove plane is located in the XY plane, and the lines are parallel to the Y axis. Figure 7 Schematic diagram of the total light intensity of the diffraction near field of the wafer pattern based on strict vector diffraction integration under point light source illumination. Figure 8 Schematic diagram of the spatial image of the pattern wafer obtained by diffraction integration under point light source illumination. Figure 7 and Figure 8 , Figure 7 The represented spatial image is more detailed, which illustrates the superiority of the calculation method of this patent.

[0044] The aerial image obtained above is located in the global coordinates, so the aerial image of the wafer pattern corresponding to the partially coherent illumination can also be transformed to the image plane through the conversion matrix between the global coordinates and the image plane coordinates to obtain the image surface result of the wafer pattern on the image plane.

[0045] The method of the present invention realizes the accurate acquisition of the wafer pattern spatial image, thereby effectively improving the reliability and efficiency of wafer defect detection, meeting the urgent demand of the semiconductor industry for high-precision and high-adaptability detection technology, and promoting the improvement of the manufacturing quality of integrated circuit chips.

[0046] The specific implementation methods described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for obtaining a pattern wafer aerial image based on strict vector diffraction integral, characterized in that: The following steps are involved: Discretize the wafer pattern into multiple discrete points, and discretize the shape of the illumination light source into multiple point light sources according to the type of illumination light source and obtain the coordinates of a single point light source; For each point light source, the coordinates of the point light source are used to obtain the spatial image of the wafer pattern corresponding to each discrete point when the point light source is illuminating by a strict vector diffraction integral model; The aerial images of the wafer pattern corresponding to all point light sources are superimposed to obtain the aerial image of the wafer pattern corresponding to the partially coherent illumination.

2. The method for obtaining a pattern wafer aerial image based on strict vector diffraction integral according to claim 1, characterized in that: For each point light source, the coordinates of the point light source are used to obtain the spatial image of the wafer pattern at each discrete point when illuminated by the strict vector diffraction integral model, including: According to the coordinates of the point light source, calculate the electric field distribution of the near-field diffraction of the light emitted by the point light source at N*M discrete points on the surface of the wafer pattern after being reflected by the wafer pattern. and magnetic field distribution ; Calculation of electric field distribution using Fourier optical imaging theory and magnetic field distribution Fourier transform results on the spectrum plane and , where the coordinates of the spectrum surface are Obtained by transformation of the global coordinate system (x, y); Calculate the transfer function of each polarization component of the electromagnetic field based on a rigorous vector diffraction integral model; From the transfer function and Fourier transform results and Multiply and calculate the diffraction pattern produced by a given aperture; Using Fourier optical imaging theory, the diffraction pattern is inversely transformed to obtain the electric field distribution in the far field. and magnetic field distribution , according to the electric field distribution The spatial image I(x, y) of the wafer pattern at each discrete point corresponding to the point light source is obtained.

3. The method for obtaining a pattern wafer aerial image based on strict vector diffraction integral according to claim 2, characterized in that: Calculate the electric field distribution of near-field diffraction at N*M discrete points on the wafer pattern surface and magnetic field distribution ,include: Wafer patterning Perform rigorous numerical calculations , and the electric field distribution is obtained and magnetic field distribution ; in, Indicates the coordinate position is The wafer pattern of discrete point areas, Denotes rigorous numerical computation using the finite element method, finite-difference time-domain method, finite integration method, or rigorous coupled-wave method.

4. The method for obtaining a pattern wafer aerial image based on strict vector diffraction integral according to claim 2, characterized in that: The transfer functions of each polarization component of the electromagnetic field are calculated based on a rigorous vector diffraction integral model, including: According to the propagation of electromagnetic field strictly follows Maxwell's equations, the vector Helmholtz equation can be derived from Maxwell's equations, and according to the vector Green's theorem, after a series of vector operations, any point light source i Observe point in space The electromagnetic field distribution is expressed as a superposition integral form: ； ； Where z is the propagation distance, is the wavelength of illumination light, represents the wave vector, and Represents observation points in space The electric and magnetic field distributions at i Expressed as i Point light sources, represents the electromagnetic field with n different polarization components, The electric fields of different polarization components representing near-field diffraction, represents the magnetic field of different polarization components corresponding to the near-field diffraction, represents the coefficient of the electric field component, represents the coefficient of the magnetic field component, represents the global coordinate system, represents the position of any observation point in space, j represents a complex factor; The above and Rewritten in the form of convolution, we get: ； ； in, represents the spread function of different electric field polarization components, The spread function representing the different polarization components of the magnetic field; The diffusion function and Perform Fourier transform to obtain the transfer functions of different electric field polarization components and different magnetic field polarization components. and .

5. The method for acquiring a pattern wafer aerial image based on strict vector diffraction integral according to claim 2, characterized in that: From the transfer function and Fourier transform results and Multiply the diffraction pattern produced by a given aperture by: The transfer functions of different electric field polarization components are and Multiply to get the electric field diffraction pattern , the transfer functions of different magnetic field polarization components are and Multiply to get the magnetic field diffraction pattern .

6. The method for acquiring a pattern wafer aerial image based on strict vector diffraction integral according to claim 2, characterized in that: Also includes: When the partially coherent light source composed of point light sources illuminates the wafer pattern obliquely, the spectrum plane will move backward along the direction of light propagation, and a frequency shift amount must be added to each point light source on the spectrum. The frequency shift amount provided by the oblique illumination of a single point light source is: ； or ； in, is the refractive index of air, is the incident angle of the point light source onto the wafer, is the diffraction angle of the light diffracted by the wafer, is the wavelength of illumination light, Represents the frequency shift, the positive and negative in the formula depends on the single point light source i coordinates, adding the electromagnetic field transfer function of the illumination angle frequency shift , becomes , .

7. The method for acquiring a pattern wafer aerial image based on strict vector diffraction integral according to claim 2, characterized in that: Also includes: When the partially coherent light source composed of point light sources illuminates the wafer pattern vertically, the spectrum surface remains unchanged. The frequency shift range calculated for all point light sources based on the internal and external coherence coefficients is: ； in, is the refractive index of air, NA is the numerical aperture of the objective lens, is the internal coherence factor, is the external coherence factor, is the wavelength of illumination light, The electromagnetic field transfer function represents the frequency shift range of all point light sources i, adding the frequency shift of the lighting angle , becomes , .

8. The method for acquiring a pattern wafer aerial image based on strict vector diffraction integral according to claim 2, characterized in that: Also includes: The spatial image of the wafer pattern corresponding to the partially coherent illumination is transformed to the image plane through the conversion matrix between the global coordinates and the image plane coordinates, and the image surface result of the wafer pattern on the image plane is obtained.

Images