Method and system for improving defect detection sensitivity of semiconductor wafer

Abstract

The invention discloses a method and system for improving the defect detection sensitivity of a semiconductor wafer, and belongs to the technical field of wafer defect detection.The method is characterized by comprising the steps that an under-sampling Fourier coefficient vector of the semiconductor wafer to be detected is obtained based on a single-pixel sensing method; reconstructing an undersampled Fourier coefficient vector according to a compressed sensing algorithm to obtain a target Fourier coefficient vector; obtaining the frequency spectrum of the semiconductor wafer to be detected according to the target Fourier coefficient vector; according to the method, defect detection is carried out based on a single-pixel sensing method, and compared with a method for collecting optical signals through array detection devices such as a CCD in a traditional optical imaging technology, the method is wide in spectral response range, high in sampling speed, high in accuracy and high in accuracy. Moreover, high detection sensitivity can be kept in a weak light scene and a long-distance scene, and meanwhile, defect detection is carried out on the semiconductor wafer to be detected by utilizing the frequency spectrum characteristics, so that the detection speed is increased.

Description

Technical Field

[0001] The present invention relates to the technical field of wafer defect detection, and more particularly to a method and system for improving the sensitivity of semiconductor wafer defect detection. Background Art

[0002] Semiconductor wafer surface defect detection technology is one of the core technologies in chip manufacturing. Initially, wafer surface defect detection relied on human vision. With the development of methods and theories such as image processing and machine learning, semiconductor wafer surface defect detection methods based on machine vision and artificial intelligence are gradually replacing the traditional technology of human eye detection. However, most of these methods based on artificial intelligence and machine vision are based on the method and mode of imaging first and then processing, that is, using array imaging equipment to image the wafer surface first, and then using relevant algorithms or models to process the acquired images.

[0003] For example, a patent application with publication number CN117423639A provides a wafer surface modeling defect detection system, which uses a CCD camera to photograph the wafer surface to be inspected, determines the wafer surface image signal, and determines the wafer surface feature data based on machine vision through the collected signal information. The wafer surface feature data is then input into the wafer surface detection model to perform defect detection on the wafer surface.

[0004] However, the system uses a large amount of data, which puts a lot of storage and transmission bandwidth pressure on the observation system, and easily causes the loss of high-dimensional information such as frequency, phase, and polarization of the optical signal. In addition, the CCD camera has strict requirements on the environment and is not suitable for collecting images in low-light scenes and long-distance scenes. Therefore, the existing technology has shortcomings. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method and system for improving the sensitivity of semiconductor wafer defect detection. Based on the characteristics of the single-pixel sensing method, the present invention can quickly and accurately collect information in low-light scenes and long-distance scenes, and combine the spectral characteristics to quickly detect defects in the semiconductor wafer to be inspected, thereby improving detection efficiency and sensitivity.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides a method for improving the sensitivity of semiconductor wafer defect detection, comprising:

[0008] Obtaining an under-sampled Fourier coefficient vector of a semiconductor wafer to be inspected based on a single pixel sensing method;

[0009] Reconstruct the under-sampled Fourier coefficient vector according to a compressed sensing algorithm to obtain a target Fourier coefficient vector;

[0010] Obtaining a frequency spectrum of the semiconductor wafer to be inspected according to the target Fourier coefficient vector;

[0011] Defect detection is performed on the semiconductor wafer to be inspected according to the frequency spectrum.

[0012] As a further improvement of the present invention, the method for obtaining the under-sampled Fourier coefficient vector of the semiconductor wafer to be inspected based on the single-pixel sensing method includes:

[0013] Constructing color-coded patterns based on Bayer filters and Fourier transforms;

[0014] Based on the circular sampling method and the spectrum energy parameter, a sampling point is determined on the surface of the semiconductor wafer to be inspected;

[0015] Projecting the color coding pattern onto each of the sampling points using a four-step phase shift method to obtain a Fourier coefficient corresponding to each of the sampling points;

[0016] The Fourier coefficients corresponding to each of the sampling points are combined to obtain an under-sampled Fourier coefficient vector of the semiconductor wafer to be inspected.

[0017] As a further improvement of the present invention, the method of constructing a color coding pattern based on a Bayer filter and a Fourier transform includes:

[0018] Decomposing the Bayer filter into three colors: red, green and blue, to obtain red, green and blue coding templates;

[0019] generating a grayscale base pattern according to Fourier spectrum information;

[0020] Multiply the red, green and blue coding templates by the grayscale basis pattern to obtain three single-spectrum Fourier basis patterns;

[0021] The three single-spectrum Fourier basis patterns are combined to obtain the color-coded pattern.

[0022] As a further improvement of the present invention, the method of determining sampling points on the surface of the semiconductor wafer to be inspected based on the circular sampling method and the spectrum energy parameter includes:

[0023] Based on the circular sampling method, taking the surface center of the semiconductor wafer to be inspected as a starting point, a first sampling path is determined according to a preset sampling angle and sampling radius, wherein the first sampling path includes a plurality of sections;

[0024] For two adjacent sections in the first sampling path, the second section is updated according to the spectrum energy parameter of the first section, and the first sampling path is traversed in sequence with one section as a step length to obtain a second sampling path, wherein the first section is the section closer to the starting point among the two adjacent sections, and the second section is the section farther from the starting point among the two adjacent sections;

[0025] A plurality of sampling points are determined on the second sampling path based on equidistant sampling.

[0026] As a further improvement of the present invention, the updating of the second section according to the spectrum energy parameter of the first section includes:

[0027] Select N seed points in the first section and calculate the spectrum energy E of each seed point i , i=1, ..., N, N is a positive integer greater than 1;

[0028] According to the spectrum energy E i , obtain the spectrum energy parameter of the first section Where μ is the compensation coefficient;

[0029] The spectrum energy parameter ξ is compared with a preset threshold, and the second road section is updated according to the comparison result.

[0030] As a further improvement of the present invention, the step of reconstructing the undersampled Fourier coefficient vector according to a compressed sensing algorithm to obtain a target Fourier coefficient vector includes:

[0031] Selecting a measurement matrix and a sparse matrix based on the characteristics of the under-sampled Fourier coefficient vector;

[0032] Calculate a perception matrix according to the measurement matrix and the sparse matrix;

[0033] The undersampled Fourier coefficient vector is reconstructed based on the perception matrix to obtain the target Fourier coefficient vector.

[0034] As a further improvement of the present invention, the reconstructing the undersampled Fourier coefficient vector based on the perception matrix to obtain the target Fourier coefficient vector includes:

[0035] According to the smoothing function F σ (u) Construct the reconstruction model: Wherein, parameter σ represents the approximation degree of the smoothing function to the L0 norm, φ is the perception matrix, u is the undersampled Fourier coefficient vector, and v is the target Fourier coefficient vector;

[0036] Set the value of the parameter σ to σ∈(σ1, σ2, ..., σz ), where σ1>σ2>…>σ z , where z is the number of values ​​of σ;

[0037] Optimizing and solving the smoothing function corresponding to each of the values ​​to obtain a solution corresponding to each value;

[0038] A global optimal solution is selected from the solutions corresponding to each of the values, and the under-sampled Fourier coefficient vector is reconstructed according to the global optimal solution to obtain the target Fourier coefficient vector.

[0039] As a further improvement of the present invention, the defect detection of the semiconductor wafer to be detected according to the frequency spectrum includes:

[0040] Calculating a spectrum similarity coefficient between the spectrum and a preset spectrum, wherein the preset spectrum is obtained based on a defect-free wafer template;

[0041] Selecting a plurality of feature points in the frequency spectrum, and calculating a correlation coefficient between each of the feature points and a corresponding preset feature point, wherein the preset feature point is obtained according to the preset frequency spectrum;

[0042] Defect detection is performed on the semiconductor wafer to be inspected according to the frequency spectrum similarity coefficient and the correlation coefficient.

[0043] As a further improvement of the present invention, the spectrum similarity coefficient Wherein f(n) represents the frequency spectrum, g(n) represents the preset frequency spectrum, and n represents the frequency.

[0044] The present invention provides a system for improving the sensitivity of semiconductor wafer defect detection, comprising:

[0045] Acquisition module: Acquires the under-sampled Fourier coefficient vector of the semiconductor wafer to be inspected based on the single-pixel perception method;

[0046] Reconstruction module: reconstruct the under-sampled Fourier coefficient vector according to the compressed sensing algorithm to obtain the target Fourier coefficient vector;

[0047] Calculation module: obtaining the frequency spectrum of the semiconductor wafer to be inspected according to the target Fourier coefficient vector;

[0048] Detection module: performs defect detection on the semiconductor wafer to be detected according to the frequency spectrum.

[0049] The present invention is based on a single-pixel sensing method and a compressed sensing algorithm, uses fewer sampling points to obtain the spectrum of a semiconductor wafer to be inspected, and performs defect detection on the wafer based on the spectrum characteristics. The present invention has a wide range of application scenarios, requires less data, and can quickly and accurately implement defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flowchart of the steps of a method for improving the sensitivity of semiconductor wafer defect detection according to the present invention;

[0051] Figure 2 It is a schematic diagram of the existing circular sampling method;

[0052] Figure 3 A flowchart of the steps of updating a road section in the present invention;

[0053] Figure 4 The present invention is a schematic structural diagram of a system for improving the sensitivity of semiconductor wafer defect detection. DETAILED DESCRIPTION

[0054] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations of the technical solution of the present invention.

[0055] The term "and / or" in the following text is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0056] like Figure 1 As shown, the embodiment of the present application provides a method for improving the sensitivity of semiconductor wafer defect detection, including:

[0057] Obtaining an under-sampled Fourier coefficient vector of a semiconductor wafer to be inspected based on a single pixel sensing method;

[0058] Reconstruct the under-sampled Fourier coefficient vector according to the compressed sensing algorithm to obtain the target Fourier coefficient vector;

[0059] Obtaining a frequency spectrum of the semiconductor wafer to be inspected according to the target Fourier coefficient vector;

[0060] Defect detection is performed on the semiconductor wafer to be inspected based on the frequency spectrum.

[0061] Traditional methods based on artificial intelligence and machine vision mostly use array imaging devices (such as CCD and CMOS) to image the wafer surface, and then use relevant algorithms or models to process the acquired images. However, this method has high requirements on imaging quality, and the imaging quality of array imaging devices is greatly affected by the environment. When the image quality is poor, the sensitivity and accuracy of detection will be greatly affected.

[0062] Therefore, compared with the traditional method, this embodiment combines the single-pixel sensing method to collect data through a single-pixel detector. Compared with array imaging devices, the single-pixel detector has a fast sampling speed and high sensitivity, and can be applied to a variety of scenarios. Then, combined with the compressed sensing algorithm, the number of sampling times is reduced, and finally the spectrum characteristics are used for defect detection. Compared with the existing technology, this embodiment can improve the sensitivity and efficiency of semiconductor wafer defect detection in low-light scenes or long-distance scenes.

[0063] Furthermore, this embodiment provides a method for obtaining an undersampled Fourier coefficient vector of a semiconductor wafer to be inspected based on a single pixel sensing method, comprising:

[0064] Constructing color-coded patterns based on Bayer filters and Fourier transforms;

[0065] Based on the circular sampling method and spectrum energy parameters, the sampling points are determined on the surface of the semiconductor wafer to be inspected;

[0066] The color coding pattern is projected onto each sampling point using a four-step phase shift method to obtain the Fourier coefficients corresponding to each sampling point;

[0067] The Fourier coefficients corresponding to each sampling point are combined to obtain an under-sampled Fourier coefficient vector of the semiconductor wafer to be inspected.

[0068] The core of the single-pixel sensing method is to modulate the coding pattern with the target detection object, measure the Fourier spectrum of the target detection object, and perform analysis based on the measurement results.

[0069] Compared with the traditional grayscale coding pattern, the color coding pattern provided in this embodiment can alleviate the reflection or interference of light by the thin film structure on the surface of the semiconductor wafer to be detected, thereby improving the detection accuracy.

[0070] Furthermore, this embodiment provides a step of constructing a color coding pattern based on a Bayer filter and a Fourier transform, the step comprising:

[0071] Decompose the Bayer filter into three colors: red, green and blue, and obtain red, green and blue coding templates;

[0072] generating a grayscale base pattern according to Fourier spectrum information;

[0073] The red, green and blue coding templates are dot-multiplied with the grayscale basis pattern to obtain three single-spectrum Fourier basis patterns;

[0074] The three single-spectral Fourier basis patterns are merged to obtain a color-coded pattern.

[0075] Specifically, the Bayer filter refers to adding a filter with periodic color distribution to the surface of a monochrome image sensor, and the filter is composed of smaller filters. Each period of the Bayer filter consists of 2×2 small filters, including a total of three colors of small filters, namely red, green and blue.

[0076] If the Bayer filter is decomposed into three small filters of different colors, we can get the red, green and blue coding templates, which are denoted as M R 、M G and M B , the expression is:

[0077]

[0078] Then generate a grayscale basis pattern containing Fourier spectrum information for:

[0079]

[0080] Where a represents the DC component, b represents the coefficient of modulation contrast, x and y represent two-dimensional Cartesian coordinates, and f x and f y represents the frequency in the x and y directions, represents the phase. Since this embodiment adopts the four-step phase shift method, Four grayscale base patterns are obtained.

[0081] The red, green and blue coding templates are respectively multiplied with the grayscale basis pattern to obtain three single-spectrum Fourier basis patterns, and the three single-spectrum Fourier basis patterns are combined to obtain the color coding pattern. for:

[0082]

[0083] Where c represents color, R, G, and B represent red, green, and blue respectively. Therefore there are four color-coded patterns.

[0084] Furthermore, this embodiment provides a step of projecting the color coding pattern to each sampling point using a four-step phase shift method to obtain the Fourier coefficient corresponding to each sampling point, specifically including modulating the color coding pattern with the semiconductor wafer to be detected to obtain the expected light intensity response value collected for:

[0085]

[0086] Where S is the field of view, that is, the size of the color-coded pattern projection, and R(x, y) represents the intensity of the reflected light on the surface of the semiconductor wafer to be detected. However, considering the influence of detector noise and environmental noise, the actual light intensity response value collected is Should be expressed as:

[0087]

[0088] Where D n represents the ambient noise, and γ represents the gain of the single-pixel detector.

[0089] According to the light intensity response value at different phases, the Fourier coefficient C(f x , f y )for:

[0090]

[0091] Where j is an imaginary unit, D0(f x , f y )、D π (f x , f y ), Respectively The light intensity response value at .

[0092] By performing the above steps at each sampling point, the Fourier coefficient corresponding to each sampling point can be obtained.

[0093] Furthermore, this embodiment provides a method for determining sampling points on the surface of a semiconductor wafer to be detected based on a circular sampling method and spectrum energy parameters, which can ensure detection accuracy on the basis of reducing sampling points, specifically including:

[0094] Based on the circular sampling method, taking the surface center of the semiconductor wafer to be inspected as the starting point, determining a first sampling path according to a preset sampling angle and sampling radius, wherein the first sampling path includes a plurality of sections;

[0095] For two adjacent sections in the first sampling path, the second section is updated according to the spectrum energy parameter of the first section, and the first sampling path is traversed in sequence with one section as the step length to obtain the second sampling path, where the first section is the section closer to the starting point among the two adjacent sections, and the second section is the section farther from the starting point among the two adjacent sections;

[0096] A plurality of sampling points are determined in the second sampling path based on equidistant sampling.

[0097] like Figure 2As shown, the existing circular sampling method expands outward in a spiral trajectory in the direction of the arrow according to a certain radius growth amount, and gradually increases the sampling radius. This spiral line method ensures progressive coverage from low frequency to high frequency, and only collects points on the spiral path. However, this method is usually based on a preset radius growth amount for step-by-step collection, which is highly subjective. If the radius growth amount is large, it will lead to too few sampling points, affecting the detection results. If the radius growth amount is small, it will lead to too many sampling points, and a large amount of data needs to be processed. Therefore, this embodiment further provides a method that can adaptively adjust the radius growth amount based on the spectrum energy parameter to ensure sampling efficiency.

[0098] Specifically, the first sampling path is determined based on the circular sampling method, that is, Figure 2 The path shown in , and each ring in the first sampling path is called a segment, and then the update step is performed, such as Figure 3 As shown, the step includes:

[0099] Select N seed points in the first section and calculate the spectrum energy E of each seed point i , i=1, ..., N, N is a positive integer greater than 1;

[0100] According to the spectrum energy E i , get the spectrum energy parameters of the first section Where μ is the compensation coefficient;

[0101] The spectrum energy parameter ξ is compared with a preset threshold, and the second section is updated according to the comparison result.

[0102] The larger the value of the spectrum energy parameter ξ, the more concentrated the energy distribution of the road section is, and μ is used to compensate for the sampling error.

[0103] For example, assuming that the preset threshold is k1, the preset radius growth amount is r1, and the ratio obtained by comparing the spectrum energy parameter ξ with the preset threshold is The adjusted radius increase is The second section is then updated according to the adjusted radius increase.

[0104] This embodiment determines the sampling points based on the spectrum energy parameter, that is, if the value of the spectrum energy parameter ξ of the first path is high, the radius growth amount of the second path is reduced, and if the value of the spectrum energy parameter ξ of the first path is low, the radius growth amount of the second path is increased, and the sampling points are equidistantly allocated in the second sampling path according to the preset sampling rate. The above method can ensure the collection efficiency of effective information while reducing the amount of data processing.

[0105] Furthermore, after combining the Fourier coefficients corresponding to each sampling point to obtain an under-sampled Fourier coefficient vector, this embodiment provides a method for reconstructing the under-sampled Fourier coefficient vector, including:

[0106] Selecting a measurement matrix and a sparse matrix based on the features of the undersampled Fourier coefficient vector;

[0107] Calculate the perception matrix based on the measurement matrix and the sparse matrix;

[0108] The undersampled Fourier coefficient vector is reconstructed based on the perception matrix to obtain the target Fourier coefficient vector.

[0109] Furthermore, the step of reconstructing the undersampled Fourier coefficient vector based on the perception matrix to obtain the target Fourier coefficient vector includes:

[0110] According to the smoothing function F σ (u) Construct the reconstruction model: Among them, the parameter σ represents the approximation degree of the smoothing function to the L0 norm, φ is the perception matrix, u is the undersampled Fourier coefficient vector, and v is the target Fourier coefficient vector;

[0111] Set the value of parameter σ to σ∈(σ1,σ2,...,σ z ), where σ1>σ2>…>σ z , where z is the number of values ​​of σ;

[0112] Optimize and solve the smooth function corresponding to each value to obtain the solution corresponding to each value;

[0113] The global optimal solution is selected from the solutions corresponding to each value, and the undersampled Fourier coefficient vector is reconstructed according to the global optimal solution to obtain the target Fourier coefficient vector.

[0114] Specifically, the target Fourier coefficient vector v is sparsely represented as s=ψv, s is the target Fourier coefficient vector after sparse representation, ψ is a sparse matrix, and a matrix Φ that is unrelated to the sparse matrix is ​​designed as the measurement matrix. The undersampled Fourier coefficient vector u can be obtained by reducing the dimension of v, denoted as u=Φs=Φψv. Let Φψ=φ, φ is the perception matrix, and u=φv is obtained.

[0115] Since the dimension of u is smaller than that of v, it is impossible to find a definite solution when directly using u to solve v. Therefore, optimization theory is usually used to reconstruct and solve it:

[0116]

[0117] Wherein ||v||0 represents the L0 norm of v, but the sparse solution based on the L0 norm is a non-convex optimization problem, and the solution is complex. Moreover, since the L0 norm is a non-smooth function, the algorithm directly based on the L0 norm requires combinatorial optimization, and the amount of calculation increases exponentially. Therefore, this embodiment solves through a smooth function, that is, the smooth function is made close to the L0 norm, and the solution of the L0 norm problem is converted into an optimization solution problem of a smooth function, which speeds up the operation rate and reduces the computational complexity.

[0118] Specifically, since the L0 norm of v is the non-zero coefficient in v, we define a function h(v k ):

[0119]

[0120] Among them, v k is the kth element in the target Fourier coefficient vector v;

[0121] The L0 norm of v can be expressed as: Where K is the dimension of v, it is found that the discontinuous function h(v k ) will lead to discontinuity of the L0 norm of v. Therefore, we use an approximate smooth function f σ (v k ) instead of h(v k ) can achieve smooth estimation of L0 norm.

[0122] If the continuous function satisfies:

[0123]

[0124] Then this function can replace h(v k ) to achieve the purpose of approximating the L0 norm. At the same time, combined with h(v k ), we can get:

[0125]

[0126] Let function but:

[0127]

[0128] According to the definition of L0 norm, the L0 norm of v can be expressed as:

[0129]

[0130] You can solve Convert to solve

[0131] As σ is smaller, the corresponding function approximates the L0 norm better. However, as σ decreases, the function curve becomes less and less smooth. Therefore, in order to obtain the global optimal solution, a set of descending sequences (σ1, σ2, ..., σ2) can be set for σ. z ), by solving the solution corresponding to each value, and selecting a global optimal solution from all the solutions as the final solution, so that it can simultaneously satisfy the smoothness of the function and the approximation of σ to 0 to the greatest extent.

[0132] This embodiment reduces the difficulty of calculation while speeding up the calculation by setting a smoothing function, and finally obtains the reconstructed target Fourier coefficient vector, that is, obtains all the Fourier coefficients of the semiconductor wafer to be inspected.

[0133] Furthermore, according to all Fourier coefficients, the complete spectrum T(f x , f y )for:

[0134]

[0135] where F -1 represents the inverse Fourier transform.

[0136] Furthermore, this embodiment provides a method for detecting wafer defects according to the above spectrum, including:

[0137] Calculating a spectrum similarity coefficient between the spectrum and a preset spectrum, the preset spectrum being obtained based on a defect-free wafer template;

[0138] Selecting multiple feature points in the frequency spectrum and calculating the correlation coefficient between each feature point and a corresponding preset feature point, where the preset feature point is obtained according to the preset frequency spectrum;

[0139] Defect detection is performed on the semiconductor wafer to be inspected based on the spectrum similarity coefficient and the correlation coefficient.

[0140] The spectral similarity coefficient Wherein f(n) represents the frequency spectrum, g(n) represents the preset frequency spectrum, and n represents the frequency.

[0141] When defects appear on the surface of the semiconductor wafer to be inspected, the corresponding frequency spectrum will also change. Therefore, it is possible to preliminarily determine whether the wafer contains defects based on the similarity coefficient between the frequency spectrum and the preset frequency spectrum.

[0142] For wafers with a high probability of containing defects, feature points can be further selected in the spectrum for detection. The low-frequency part of the spectrum corresponds to feature information such as contours and edges, and the high-frequency part corresponds to detail information. Feature points can be set in different parts for analysis according to detection needs, and the correlation coefficient can be selected as the Pearson correlation coefficient. Based on the results obtained from the spectrum similarity coefficient and the correlation coefficient, the general type of wafer defects can be further analyzed.

[0143] The embodiment of the present application provides a method for improving the sensitivity of semiconductor wafer defect detection. The method expands the detection scenario and improves the detection sensitivity through a single-pixel sensing method, and selects sampling points in combination with an adaptive sampling method and a compressed sensing algorithm. The complete data is reconstructed through less sampling information, the sampling steps are optimized, and the acquisition efficiency is improved. Finally, defect detection is performed through spectral features, which speeds up the detection speed compared to the detection method based on image features.

[0144] Further, such as Figure 4 As shown, an embodiment of the present application provides a system for improving the sensitivity of semiconductor wafer defect detection, including:

[0145] Acquisition module: Acquires the under-sampled Fourier coefficient vector of the semiconductor wafer to be inspected based on the single-pixel perception method;

[0146] Reconstruction module: reconstruct the undersampled Fourier coefficient vector according to the compressed sensing algorithm to obtain the target Fourier coefficient vector;

[0147] Calculation module: obtains the frequency spectrum of the semiconductor wafer to be inspected according to the target Fourier coefficient vector;

[0148] Detection module: performs defect detection on the semiconductor wafer to be inspected based on the frequency spectrum.

[0149] The acquisition module is specifically a single-pixel detector, the reconstruction module, the calculation module and the detection module are located in the data processing device, and the single-pixel detector is communicatively connected with the data processing device.

[0150] The embodiments of the present application provide a method and system for improving the sensitivity of semiconductor wafer defect detection. Based on the single-pixel sensing method, the method and system combine adaptive sampling and compressed sensing steps, and detect wafer defects in combination with spectral characteristics. The method has a wide range of application scenarios, fast detection speed and high sensitivity.

[0151] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0152] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0153] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0154] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0155] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for improving the sensitivity of semiconductor wafer defect detection, characterized in that: include: Obtaining an under-sampled Fourier coefficient vector of a semiconductor wafer to be inspected based on a single pixel sensing method; Reconstruct the under-sampled Fourier coefficient vector according to a compressed sensing algorithm to obtain a target Fourier coefficient vector; Obtaining a frequency spectrum of the semiconductor wafer to be inspected according to the target Fourier coefficient vector; Defect detection is performed on the semiconductor wafer to be inspected according to the frequency spectrum.

2. A method for improving semiconductor wafer defect detection sensitivity according to claim 1, characterized in that: The method of obtaining an under-sampled Fourier coefficient vector of a semiconductor wafer to be inspected based on a single pixel sensing method includes: Constructing color-coded patterns based on Bayer filters and Fourier transforms; Based on the circular sampling method and the spectrum energy parameter, a sampling point is determined on the surface of the semiconductor wafer to be inspected; Projecting the color coding pattern onto each of the sampling points using a four-step phase shift method to obtain a Fourier coefficient corresponding to each of the sampling points; The Fourier coefficients corresponding to each of the sampling points are combined to obtain an under-sampled Fourier coefficient vector of the semiconductor wafer to be inspected.

3. The method for improving the sensitivity of semiconductor wafer defect detection according to claim 2, characterized in that: The method of constructing a color coding pattern based on a Bayer filter and a Fourier transform comprises: Decomposing the Bayer filter into three colors: red, green and blue, to obtain red, green and blue coding templates; generating a grayscale base pattern according to Fourier spectrum information; Multiply the red, green and blue coding templates by the grayscale basis pattern to obtain three single-spectrum Fourier basis patterns; The three single-spectrum Fourier basis patterns are combined to obtain the color-coded pattern.

4. The method for improving the sensitivity of semiconductor wafer defect detection according to claim 2, characterized in that: The method of determining sampling points on the surface of the semiconductor wafer to be inspected based on the circular sampling method and the spectrum energy parameter includes: Based on the circular sampling method, taking the surface center of the semiconductor wafer to be inspected as a starting point, a first sampling path is determined according to a preset sampling angle and sampling radius, wherein the first sampling path includes a plurality of sections; For two adjacent sections in the first sampling path, the second section is updated according to the spectrum energy parameter of the first section, and the first sampling path is traversed in sequence with one section as a step length to obtain a second sampling path, wherein the first section is the section closer to the starting point among the two adjacent sections, and the second section is the section farther from the starting point among the two adjacent sections; A plurality of sampling points are determined on the second sampling path based on equidistant sampling.

5. The method for improving the sensitivity of semiconductor wafer defect detection according to claim 4, characterized in that: The updating of the second section according to the spectrum energy parameter of the first section comprises: Select N seed points in the first section and calculate the spectrum energy E of each seed point i , i=1,…,N, N is a positive integer greater than 1; According to the spectrum energy E i , obtain the spectrum energy parameter of the first section Where μ is the compensation coefficient; The spectrum energy parameter ξ is compared with a preset threshold, and the second road section is updated according to the comparison result.

6. The method for improving the sensitivity of semiconductor wafer defect detection according to claim 1, characterized in that: The step of reconstructing the under-sampled Fourier coefficient vector according to the compressed sensing algorithm to obtain the target Fourier coefficient vector includes: Selecting a measurement matrix and a sparse matrix based on the characteristics of the under-sampled Fourier coefficient vector; Calculate a perception matrix according to the measurement matrix and the sparse matrix; The undersampled Fourier coefficient vector is reconstructed based on the perception matrix to obtain the target Fourier coefficient vector.

7. The method for improving the sensitivity of semiconductor wafer defect detection according to claim 6, characterized in that: The reconstructing the under-sampled Fourier coefficient vector based on the perception matrix to obtain the target Fourier coefficient vector includes: According to the smoothing function F σ (u) Construct the reconstruction model: Wherein, parameter σ represents the approximation degree of the smoothing function to the L0 norm, φ is the perception matrix, u is the undersampled Fourier coefficient vector, and v is the target Fourier coefficient vector; Set the value of the parameter σ to σ∈(σ1,σ2,…,σ z ), where σ1>σ2>…>σ z , where z is the number of values ​​of σ; Optimizing and solving the smoothing function corresponding to each of the values ​​to obtain a solution corresponding to each value; A global optimal solution is selected from the solutions corresponding to each of the values, and the under-sampled Fourier coefficient vector is reconstructed according to the global optimal solution to obtain the target Fourier coefficient vector.

8. The method for improving the sensitivity of semiconductor wafer defect detection according to claim 1, characterized in that: The performing defect detection on the semiconductor wafer to be detected according to the frequency spectrum comprises: Calculating a spectrum similarity coefficient between the spectrum and a preset spectrum, wherein the preset spectrum is obtained based on a defect-free wafer template; Selecting a plurality of feature points in the frequency spectrum, and calculating a correlation coefficient between each of the feature points and a corresponding preset feature point, wherein the preset feature point is obtained according to the preset frequency spectrum; Defect detection is performed on the semiconductor wafer to be inspected according to the frequency spectrum similarity coefficient and the correlation coefficient.

9. The method for improving the sensitivity of semiconductor wafer defect detection according to claim 8, characterized in that: The spectrum similarity coefficient Wherein f(n) represents the frequency spectrum, g(n) represents the preset frequency spectrum, and n represents the frequency.

10. A system for improving the sensitivity of semiconductor wafer defect detection, used to implement a method for improving the sensitivity of semiconductor wafer defect detection as claimed in any one of claims 1 to 9, characterized in that: include: Acquisition module: Acquires the under-sampled Fourier coefficient vector of the semiconductor wafer to be inspected based on the single-pixel perception method; Reconstruction module: reconstruct the under-sampled Fourier coefficient vector according to the compressed sensing algorithm to obtain the target Fourier coefficient vector; Calculation module: obtaining the frequency spectrum of the semiconductor wafer to be inspected according to the target Fourier coefficient vector; Detection module: performs defect detection on the semiconductor wafer to be detected according to the frequency spectrum.

Images