Semiconductor material defect detection method and device and computer storage medium

Abstract

The invention relates to the technical field of semiconductor material defect detection, in particular to a semiconductor material defect detection method and device and a computer storage medium, and the method comprises the following steps: carrying out three-dimensional scanning on a to-be-detected sample by using detection light, and generating a multi-photon signal by the to-be-detected sample under the excitation action of the detection light, the multiphoton signal is reversely transmitted along the same path as the detection light; arranging at least one detection node on a reverse transmission path of the multiphoton signal; and based on the at least one detection node, performing one-to-one correspondence on the multi-photon signal obtained by sparse sampling and the spatial position of the to-be-detected sample, and performing three-dimensional reconstruction to obtain a three-dimensional defect distribution image of the to-be-detected sample. According to the invention, by virtue of the accurate detection capability of the device on the three-dimensional structure of the semiconductor material, the defect detection result of the material can be efficiently and accurately obtained, and powerful technical support is provided for quality control and performance optimization of the semiconductor material.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor material defect detection, and in particular to a semiconductor material defect detection method, device and computer storage medium. Background Art

[0002] As the core components of modern electronic devices, semiconductor materials are widely used in many fields such as computers, communication equipment, and energy systems. However, their complex crystal structure and rigorous production process often lead to the generation of various defects, such as dislocations, voids, impurities, etc. These defects pose a serious threat to the performance and yield of semiconductor devices, thereby increasing production costs. Therefore, the development of an efficient and accurate defect detection technology is crucial to improving the quality of semiconductor devices and reducing production costs.

[0003] In the field of traditional defect detection, photoluminescence imaging technology has been widely integrated into practical applications due to its non-destructive detection characteristics, especially showing unique advantages in dislocation detection. However, facing the increasingly complex dislocation structure detection needs, the existing photoluminescence imaging technology has shown obvious limitations.

[0004] Taking silicon carbide (SiC) as an example, the common screw dislocations and edge dislocations in its crystal structure have different effects on material properties, but existing detection technologies cannot accurately distinguish between these two types of dislocations, which undoubtedly poses a challenge to the accurate evaluation of material properties.

[0005] Furthermore, the current photoluminescence imaging technology mainly focuses on the detection of two-dimensional planes, and its ability to analyze three-dimensional dislocation structures is insufficient, making it difficult to meet the growing demand for high-precision detection. In the existing three-dimensional detection technology system, wide-field imaging or laser scanning imaging technology occupies a dominant position, but these technologies often need to rely on mechanical drive devices such as piezoelectric objective lens drivers and Z-axis translation stages when performing Z-axis scanning. The Z-axis response time of such equipment is usually in the order of ten milliseconds. If 1,000 points need to be scanned on the Z-axis, the movement of the Z-axis alone will take at least ten seconds. In addition to the scanning time of the X-axis and Y-axis, the overall detection cycle is significantly extended, making it impossible to achieve fast three-dimensional scanning imaging. This problem seriously restricts the efficient capture of defect information during the detection process, thereby limiting the improvement of detection efficiency.

[0006] In addition, traditional photoluminescence imaging technology mainly relies on single-photon imaging. Due to the short wavelength and limited penetration depth of the excitation light, this technology can only image the surface of the sample, and the Z-axis resolution is poor. For semiconductor materials that require deep detection, such as silicon carbide (SiC) and gallium nitride (GaN), the bandgap width of these materials is usually greater than 3eV, and a light source in the ultraviolet band is required for direct excitation. However, the production capacity of ultraviolet light sources is limited, and the cost of related optical components is high, which further limits the widespread application of traditional single-photon photoluminescence imaging technology in the field of semiconductor defect detection.

[0007] Based on the limitations of the above existing technologies, there is an urgent need for a detection method that can achieve three-dimensional imaging and has both high efficiency and high resolution to meet the challenges in current semiconductor material defect detection. Summary of the invention

[0008] In order to solve the above technical problems, the present invention provides a semiconductor material defect detection method, comprising the following steps: S1: Use detection light to perform three-dimensional scanning on the sample to be tested, and the sample to be tested generates a multi-photon signal under the excitation of the detection light, and the multi-photon signal is transmitted in the reverse direction along the same path as the detection light; S2: Setting at least one detection node on the reverse transmission path of the multi-photon signal; S3: Based on the at least one detection node, the multi-photon signal obtained by sparse sampling is matched one-to-one with the spatial position of the sample to be tested, and three-dimensional reconstruction is performed to obtain a three-dimensional defect distribution image of the sample to be tested.

[0009] In one embodiment of the present invention, in S1, the step of using the detection light to perform three-dimensional scanning on the sample to be tested is as follows: S11: adjusting the light intensity and spot size of the detection light to obtain modulated detection light; S12: separating the detection light of a specified wavelength based on the modulated detection light, and adjusting the deflection angle of the detection light of the specified wavelength in the X-axis and Y-axis dimensions respectively to obtain a two-dimensional scanning beam to perform a two-dimensional scanning on the cross section of the sample to be tested; S13: adjusting the focus of the two-dimensional scanning light beam, and focusing the adjusted scanning light beam at different depths of the sample to be tested for scanning.

[0010] In one embodiment of the present invention, in S3, the step of obtaining the three-dimensional defect distribution image of the sample to be tested is as follows: S31: construct a sparse basis matrix Ψ, and select a random matrix Φ that is irrelevant to the sparse basis matrix Ψ as a measurement matrix; S32: using the measurement matrix Φ to randomly sample the multi-photon signal x to obtain a measurement value y; S33: Since y=Φx, the multi-photon signal x is sparsely represented as x=Ψs, where s is a sparse vector, and the sparsity constraint of s is minimized as the optimization goal, and the estimated value of the sparse vector s is obtained by solving ; S34: Based on the estimated value and the sparse basis matrix Ψ, the reconstructed signal is obtained ; S35: All reconstructed signals are combined to form the three-dimensional defect distribution image.

[0011] In one embodiment of the present invention, the measurement matrix Φ satisfies the following conditions: For the measurement matrix Φ∈ R m×n , there exists a constant δ∈(0,1) such that for any sparse vector s of any order, the following inequality holds: , Among them, m is the number of rows of the measurement matrix Φ, n is the number of columns of the measurement matrix Φ, and m <n, represents the L2 norm of s, that is, the Euclidean norm.

[0012] Based on the same inventive concept, the present invention also provides a semiconductor material defect detection device, comprising: A light source unit, the light source unit is used to provide detection light; A three-dimensional scanning unit, the three-dimensional scanning unit is used to process the detection light on its transmission path, and use the processed detection light to scan the sample to be tested; A signal acquisition system, wherein the signal acquisition system comprises at least one photoelectric detector, any one of which corresponds to one detection node, and is used to collect multi-photon signals generated by the sample to be tested and convert them into electrical signals; And a control and signal processing system, which controls the connection between the three-dimensional scanning unit and the signal acquisition system, and is used to make a one-to-one correspondence between the multi-photon signal obtained by sparse sampling and the spatial position of the sample to be tested, perform three-dimensional reconstruction, and obtain a three-dimensional defect distribution image of the sample to be tested.

[0013] In one embodiment of the present invention, the three-dimensional scanning unit includes an incident light modulation unit, an incident light scanning unit, an incident light high-speed zoom unit and a sample carrier which are sequentially arranged on the incident path of the detection light; Wherein, the incident light modulation unit is used to adjust the light intensity and spot size of the detection light to obtain the modulated detection light; The incident light scanning unit is used to separate the detection light of a specified wavelength based on the modulated detection light, and adjust the deflection angle of the detection light of the specified wavelength in the X and Y axis dimensions respectively to obtain a two-dimensional scanning light beam to perform a two-dimensional scanning on the cross section of the sample to be tested; The incident light high-speed zoom unit is used to focus the two-dimensional scanning light beam on different depths of the sample to be tested for scanning.

[0014] In one embodiment of the present invention, the incident light scanning unit includes a first beam splitter and a two-dimensional galvanometer, wherein the first beam splitter separates the detection light of a specified wavelength from the modulated detection light, and the two-dimensional galvanometer adjusts the deflection angle of the detection light of the specified wavelength in different dimensions of the X and Y axes respectively.

[0015] In one embodiment of the present invention, the incident light scanning unit further includes a field lens and a tube lens, and the field lens and the tube lens are arranged between the two-dimensional galvanometer and the incident light high-speed zoom unit.

[0016] In one embodiment of the present invention, the incident light high-speed zoom unit includes a second beam splitter and a zoom objective lens, the second beam splitter causes the two-dimensional scanning beam to be incident on the zoom objective lens, and the zoom objective lens causes the two-dimensional scanning beam to be focused on different depths of the sample to be tested for scanning.

[0017] The present invention also provides a semiconductor material defect detection device, which includes the above-mentioned semiconductor material defect detection apparatus.

[0018] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the instructions of the semiconductor material defect detection method are implemented.

[0019] The above technical solution of the present invention has the following advantages compared with the prior art: 1. Three-dimensional structure and deep defect detection capability: Through three-dimensional scanning and multi-photon signal monitoring, different depth cross-section detection of semiconductor material structure can be achieved; further, by observing and analyzing the three-dimensional defect distribution image, different types of defects can be clearly distinguished.

[0020] 2. High resolution and high efficiency: Through sparse sampling and signal reconstruction technology, the complexity of data processing is significantly reduced while maintaining the high resolution of the image. The entire detection process is highly automated and intelligent. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein: Figure 1 is a flow chart of a semiconductor material defect detection method provided in the first embodiment of the present invention; Figure 2 is a flow chart of a method for performing three-dimensional scanning of a sample to be tested using detection light provided in the first embodiment of the present invention; Figure 3 It is a schematic structural diagram of a semiconductor material defect detection device provided in the second embodiment of the present invention; Figure 4 is a schematic diagram of the specific structure of a semiconductor material defect detection device provided in the second embodiment of the present invention; Figure 5 is a schematic diagram of piezoelectric objective lens zoom in Embodiment 2 of the present invention; Figure 6 is a schematic diagram of zooming of an ultrasonic zoom objective lens in Embodiment 2 of the present invention; Figure 7 The scanning signals and trajectory diagrams of sparse sampling of electrical signals using Lissajous scanning method, where (a) is a sine wave scanning signal, (b) is a trajectory diagram obtained by two-dimensional scanning using a sine wave, (c) is a triangle wave scanning signal, and (d) is a trajectory diagram obtained by two-dimensional scanning using a triangle wave of the same frequency; Figure 8 Schematic diagram of the sparse sampling and reconstruction process of the image after the resolution plate is scanned line by line by laser, wherein (a) is the laser scanning image of the resolution plate, (b) is the sparse sampling image obtained by sparse sampling the laser scanning image by Lissajous scanning method, and (c) is the image obtained by processing the sparse sampling image by the reconstruction algorithm; Fig. 9 This is a three-dimensional dislocation defect image taking SiC as an example; Explanation of the reference numerals in the specification: 1. light source unit; 2. three-dimensional scanning unit; 21. incident light modulation unit; 22. incident light scanning unit; 221. first beam splitter; 222. two-dimensional galvanometer; 223. field lens; 224. tube lens; 23. incident light high-speed zoom unit; 231. second beam splitter; 232. zoom objective lens; 24. sample carrier; 3. signal acquisition system; 31. first photodetector; 32. second photodetector; 4. control and signal processing system. DETAILED DESCRIPTION

[0022] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention. Embodiment 1

[0023] Reference Figure 1 As shown, the present invention provides a semiconductor material defect detection method, comprising the following steps: S1: Use detection light to perform three-dimensional scanning on the sample to be tested, and the sample to be tested generates a multi-photon signal under the excitation of the detection light, and the multi-photon signal is transmitted in the reverse direction along the same path as the detection light; S2: Setting at least one detection node on the reverse transmission path of the multi-photon signal; S3: Based on the at least one detection node, the multi-photon signal obtained by sparse sampling is matched one-to-one with the spatial position of the sample to be tested, and three-dimensional reconstruction is performed to obtain a three-dimensional defect distribution image of the sample to be tested.

[0024] In this embodiment, in S1, in order to solve the problem that the traditional two-dimensional detection technology is difficult to detect deep defects, such as Figure 2 As shown, the steps of using detection light to perform three-dimensional scanning on the sample to be tested are as follows: S11: Adjust the intensity and spot size of the detection light to obtain modulated detection light. The modulated detection light is more suitable for the characteristics of the sample to be tested, can reduce sample damage caused by excessive light intensity, and improve the uniformity and accuracy of the spot, laying a good foundation for subsequent scanning and signal acquisition; S12: Based on the modulated detection light, separate the detection light of the specified wavelength to detect the specific absorption characteristics of the sample to be tested, so as to improve the sensitivity and accuracy of the detection; wherein the value of the specified wavelength is not fixed, but needs to be flexibly adjusted according to the corresponding excited luminescence peak position of the sample to be tested. This is because if the corresponding adjustment is not made, the originally expected two-photon or multi-photon excitation process may be converted into single-photon excitation, thereby affecting the accuracy and sensitivity of the detection; Then, the deflection angle of the detection light of the specified wavelength is adjusted in different dimensions of the X and Y axes respectively to obtain a two-dimensional scanning beam, so as to perform a two-dimensional scanning on the cross section of the sample to be tested, covering various areas of the cross section of the sample to be tested; S13: In order to obtain deep defect information of the sample to be tested, the two-dimensional scanning light beam is focused and adjusted. The focusing adjustment can ensure that the scanning light beam forms a clear focus at different depths (i.e., Z-axis direction) of the sample to be tested, and can accurately detect the three-dimensional structure of the sample.

[0025] Furthermore, in S3, the step of obtaining the three-dimensional defect distribution image of the sample to be tested is as follows: S31: construct a sparse basis matrix Ψ, and select a random matrix Φ that is irrelevant to the sparse basis matrix Ψ as a measurement matrix; S32: using the measurement matrix Φ to randomly sample the multi-photon signal x to obtain a measurement value y; S33: Since y=Φx, the multi-photon signal x is sparsely represented as x=Ψs, where s is a sparse vector, so as to minimize the L1 norm of s (i.e., sparsity constraint) is the optimization goal, and the estimated value of the sparse vector s is obtained by solving , while suppressing noise and interference; S34: Based on the estimated value and the sparse basis matrix Ψ, the reconstructed signal is obtained ; S35: By combining the reconstructed electrical signals, a clear and intuitive three-dimensional defect distribution image is generated. This step realizes the conversion of complex electrical signal data into image information that is easy to understand and analyze, providing strong support for subsequent defect analysis and processing.

[0026] In addition to minimizing the L1 norm to effectively reconstruct sparse signals, matching pursuit (MP), orthogonal matching pursuit (OMP) and basis pursuit (BP) can also be used. Specifically, the MP algorithm is an iterative algorithm that approximates the original signal by gradually finding the basis function that is most relevant to the measurement value y. In each iteration, the algorithm selects an atom that is most relevant to the current residual from a given dictionary, and then updates the residual to continue to approximate the original signal in the next iteration. The OMP algorithm is an improved matching pursuit algorithm that improves the accuracy of reconstruction through the orthogonalization step. The BP algorithm is a reconstruction method based on linear programming. It transforms the sparse signal reconstruction problem into a linear programming problem and obtains the sparsest solution by solving the linear programming problem. Specifically, the BP algorithm attempts to find a sparse signal that satisfies a given measurement value so that the L1 norm (i.e., the sum of the absolute values ​​of the signal) of the signal is minimized.

[0027] It is important to note that for successful signal reconstruction, the following two key conditions must be met: 1. Sparsity: The signal must be sparse under a certain basis function or dictionary, that is, there are only a few non-zero or approximately non-zero components in the transform domain.

[0028] 2. The measurement matrix Φ satisfies the restricted isometry property (RIP) condition. The RIP condition requires that the measurement matrix Φ maintains a small distortion for any sparse vector s. In simple terms, the measurement matrix Φ should be able to approximately retain the core information of the sparse signal with a small error. Specifically, the RIP condition is: For the measurement matrix Φ∈ R m×n , there exists a constant δ∈(0,1) such that for any sparse vector s of any order, the following inequality holds: , Among them, m is the number of rows of the measurement matrix Φ, n is the number of columns of the measurement matrix Φ, and m <n, represents the L2 norm of s, that is, the Euclidean norm. Embodiment 2

[0029] Based on the same inventive concept as the first embodiment, the present invention also provides a semiconductor material defect detection device, such as Figure 3 and Figure 4 As shown, the device includes a light source unit 1, a three-dimensional scanning unit 2, a signal acquisition system 3 and a control and signal processing system 4, wherein the light source unit 1 is used to provide detection light; The three-dimensional scanning unit 2 is used to process the detection light on its transmission path, and use the processed detection light to scan the sample to be tested; The signal acquisition system 3 includes at least one photoelectric detector, any one of which corresponds to a detection node, and is used to collect multi-photon signals generated by the sample to be tested and convert them into electrical signals; The control and signal processing system 4 controls the connection between the three-dimensional scanning unit 2 and the signal acquisition system 3, and is used to make a one-to-one correspondence between the multi-photon signal obtained by sparse sampling and the spatial position of the sample to be tested, perform three-dimensional reconstruction, and obtain a three-dimensional defect distribution image of the sample to be tested.

[0030] Specifically, the three-dimensional scanning unit 2 includes an incident light modulation unit 21, an incident light scanning unit 22, an incident light high-speed zoom unit 23 and a sample carrier 24 which are sequentially arranged on the incident path of the detection light.

[0031] The incident light modulation unit 21 is used to adjust the light intensity and spot size of the detection light to obtain the modulated detection light; The incident light scanning unit 22 is used to separate the detection light of the specified wavelength based on the modulated detection light, and adjust the deflection angle of the detection light of the specified wavelength in different dimensions of the X and Y axes to obtain a two-dimensional scanning beam to perform a two-dimensional scanning of the cross section of the sample to be tested; wherein the value of the specified wavelength is not fixed, but needs to be flexibly adjusted according to the corresponding excited luminescence peak position of the sample to be tested. This is because if the corresponding adjustment is not made, the originally expected two-photon or multi-photon excitation process may be converted into single-photon excitation, thereby affecting the accuracy and sensitivity of the detection; The incident light high-speed zoom unit 23 is used to focus the two-dimensional scanning light beam on different depths of the sample to be tested for scanning.

[0032] In this embodiment, the incident light scanning unit 22 includes a first beam splitter 221 and a two-dimensional galvanometer 222, wherein the first beam splitter 221 separates the detection light of a specified wavelength from the modulated detection light, and the two-dimensional galvanometer 222 is composed of two independent axes, which adjust the deflection angle of the detection light of the specified wavelength in different dimensions, thereby realizing two-dimensional scanning of the sample.

[0033] Furthermore, the incident light scanning unit 22 further includes a field lens 223 and a tube lens 224 , and the field lens 223 and the tube lens 224 are arranged between the two-dimensional galvanometer 222 and the incident light high-speed zoom unit 23 .

[0034] In this embodiment, the incident light high-speed zoom unit 23 includes a second beam splitter 231 and a zoom objective lens 232. The second beam splitter 231 causes the two-dimensional scanning beam to be incident on the zoom objective lens, and the zoom objective lens 232 focuses the two-dimensional scanning beam on different depths of the sample to be tested for scanning.

[0035] Specifically, the control and signal processing system 4 controls the zoom lens 232, and the zoom lens 232 can adopt two different technical solutions: a piezoelectric lens or an ultrasonic zoom lens, each of which has its own unique advantages. The following are the specific implementation details of these two zoom methods: See also Figure 5 When the zoom lens 232 is a piezoelectric lens, piezoelectric ceramics are used as the core component of the piezoelectric lens zoom device. The control and signal processing system 4 can quickly respond and cause the displacement of the zoom lens 232 in the Z-axis direction by accurately controlling the voltage applied to the piezoelectric ceramics. The adjustment of this displacement can be completed in a very short time (usually within ten milliseconds), thereby achieving accurate adjustment of the focal length of the zoom lens 232. Since the core component of the piezoelectric lens zoom device is piezoelectric ceramics, its volume is relatively small, making the entire zoom system more compact, which is conducive to reducing the volume and weight of the optical system and improving the portability and flexibility of the system.

[0036] See also Figure 6When the zoom objective lens 232 is an ultrasonic zoom objective lens, the control and signal processing system 4 can accurately control the refractive index change in the medium by adjusting the frequency, amplitude or waveform of the incident ultrasonic wave. These changes will cause the effective focal length of the lens to change, thereby realizing dynamic focal length adjustment. For example, when the density change caused by ultrasonic waves forms a gradient refractive index region in the lens material, the curvature of the lens will change, thereby adjusting the focal length. In some complex applications, the ultrasonic variable focal length objective lens will also be combined with phased array technology to further enhance the accuracy and response speed of focal length adjustment by generating multiple interference waves in the lens material. Due to the high-frequency characteristics of ultrasonic waves, the focal length adjustment of the lens can be completed in a very short time (usually in the order of microseconds), which makes the ultrasonic variable focal length objective lens very suitable for application scenarios that require fast zooming or continuous imaging.

[0037] Furthermore, the signal acquisition system 3 includes a first photodetector 31 and a second photodetector 32 , and can acquire multi-photon signals through the first photodetector 31 or the second photodetector 32 , and convert the acquired multi-photon signals into electrical signals.

[0038] The first photodetector 31 is disposed near the first beam splitter 221 to capture the multi-photon signal reflected by the first beam splitter 221 . The acquisition efficiency at this position is higher, but the target surface of the first photodetector 31 is required to be larger.

[0039] The second photodetector 32 is disposed near the second beam splitter 231 to capture the designated light signal transmitted through the second beam splitter 231 . The collection efficiency at this position is relatively low, but the target surface of the second photodetector 32 is required to be relatively small.

[0040] Preferably, both the first beam splitter 221 and the second beam splitter 231 are dichroic mirrors.

[0041] In order to further improve the scanning rate, the signal acquisition system 3 is set by the control and signal processing system 4 to perform sparse sampling on the multi-photon signal. The scanning method is preferably a Lissajous scanning method. For Lissajous scanning, the fill rate is the core indicator to measure the quality of its scanning performance, which is specifically defined as the ratio of the area actually covered by the scan to the total area of ​​the entire scanning area. The level of this fill rate is closely related to the number of lobes of the Lissajous scan. The number of lobes, that is, the total number of intersections between the scanning trajectory and the X-axis and the Y-axis, is calculated by adding the scanning frequencies in the two directions and then dividing by the greatest common divisor of the two frequencies.

[0042] By setting a suitable scanning frequency, two-dimensional scanning or three-dimensional scanning is performed. In an example of this embodiment, the frequency can be set to , and . Reference Figure 7 As shown in (a) to (d) in the figure, compared with the sine wave scanning method, the use of the triangle wave scanning method can significantly improve the fill rate, thereby obtaining a more ideal scanning effect.

[0043] Reference Figure 8 As shown in (a) to (c) in the figure, the core goal of image reconstruction is to use limited sampling data to restore a complete image that is as close to the original image quality as possible. Sparse sampling technology is used to collect sampling points between 1 / 5 and 1 / 10 of the original number of samples, thereby increasing the imaging speed by 5 to 10 times. Since the amount of data to be processed is significantly reduced, sparse sampling technology has shown great advantages in application scenarios that require fast imaging or have limited resources. In addition, when the application dimension of sparse sampling is expanded from two dimensions to three dimensions, its advantages are further magnified. Since three-dimensional space provides more degrees of freedom to select sample points, its sparse sampling rate can be further reduced.

[0044] See also Fig. 9 As shown, the collected multi-photon signals are matched one by one with the spatial position of SiC, and three-dimensional reconstruction and processing are performed to obtain a three-dimensional dislocation defect image of SiC. Compared with the single-photon photoimaging depth of only 1 μm, the Z-axis depth of the multi-photon photoimaging in the present invention can reach 30 μm. Through the three-dimensional image, different dislocation types of SiC, screw dislocations and edge dislocations, can be clearly distinguished. Embodiment 3

[0045] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the instructions of the semiconductor material defect detection method described in the above-mentioned embodiment 1 are implemented.

[0046] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt 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 codes.

[0047] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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 generate 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.

[0048] 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.

[0049] 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.

[0050] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.

Claims

1. A semiconductor material defect detection method, characterized in that: The following steps are involved: S1: Use detection light to perform three-dimensional scanning on the sample to be tested, and the sample to be tested generates a multi-photon signal under the excitation of the detection light, and the multi-photon signal is transmitted in the reverse direction along the same path as the detection light; S2: Setting at least one detection node on the reverse transmission path of the multi-photon signal; S3: Based on the at least one detection node, the multi-photon signal obtained by sparse sampling is matched one-to-one with the spatial position of the sample to be tested, and three-dimensional reconstruction is performed to obtain a three-dimensional defect distribution image of the sample to be tested.

2. The semiconductor material defect detection method according to claim 1, characterized in that: In S1, the steps of using the detection light to perform three-dimensional scanning on the sample to be tested are as follows: S11: adjusting the light intensity and spot size of the detection light to obtain modulated detection light; S12: separating the detection light of a specified wavelength based on the modulated detection light, and adjusting the deflection angle of the detection light of the specified wavelength in the X-axis and Y-axis dimensions respectively to obtain a two-dimensional scanning beam to perform a two-dimensional scanning on the cross section of the sample to be tested; S13: focusing the two-dimensional scanning light beam, and focusing the adjusted scanning light beam on different depths of the sample to be tested for scanning.

3. The semiconductor material defect detection method according to claim 1, characterized in that: In S3, the steps of obtaining the three-dimensional defect distribution image of the sample to be tested are as follows: S31: construct a sparse basis matrix Ψ, and select a random matrix Φ that is irrelevant to the sparse basis matrix Ψ as a measurement matrix; S32: using the measurement matrix Φ to randomly sample the multi-photon signal x to obtain a measurement value y; S33: Since y=Φx, the multi-photon signal x is sparsely represented as x=Ψs, where s is a sparse vector, and the sparsity constraint of s is minimized as the optimization goal, and the estimated value of the sparse vector s is obtained by solving ; S34: Based on the estimated value and the sparse basis matrix Ψ, the reconstructed signal is obtained ; S35: All reconstructed signals are combined to form the three-dimensional defect distribution image.

4. The semiconductor material defect detection method according to claim 3, characterized in that: The measurement matrix Φ satisfies the following conditions: For the measurement matrix Φ∈ R m×n , there exists a constant δ∈(0,1) such that for any sparse vector s of any order, the following inequality holds: , where m is the number of rows of the measurement matrix Φ, n is the number of columns of the measurement matrix Φ, and m <n, represents the L2 norm of s, that is, the Euclidean norm.

5. A semiconductor material defect detection device, characterized in that: include: A light source unit, the light source unit is used to provide detection light; A three-dimensional scanning unit, the three-dimensional scanning unit is used to process the detection light on its transmission path, and use the processed detection light to scan the sample to be tested; A signal acquisition system, wherein the signal acquisition system comprises at least one photoelectric detector, any one of which corresponds to one detection node, and is used to collect multi-photon signals generated by the sample to be tested and convert them into electrical signals; And a control and signal processing system, which controls the connection between the three-dimensional scanning unit and the signal acquisition system, and is used to make a one-to-one correspondence between the multi-photon signal obtained by sparse sampling and the spatial position of the sample to be tested, perform three-dimensional reconstruction, and obtain a three-dimensional defect distribution image of the sample to be tested.

6. The semiconductor material defect detection device according to claim 5, characterized in that: The three-dimensional scanning unit includes an incident light modulation unit, an incident light scanning unit, an incident light high-speed zoom unit and a sample carrier which are sequentially arranged on the incident path of the detection light; Wherein, the incident light modulation unit is used to adjust the light intensity and spot size of the detection light to obtain the modulated detection light; The incident light scanning unit is used to separate the detection light of a specified wavelength based on the modulated detection light, and adjust the deflection angle of the detection light of the specified wavelength in the X and Y axis dimensions respectively to obtain a two-dimensional scanning light beam to perform a two-dimensional scanning on the cross section of the sample to be tested; The incident light high-speed zoom unit is used to focus the two-dimensional scanning light beam on different depths of the sample to be tested for scanning.

7. The semiconductor material defect detection device according to claim 6, characterized in that: The incident light scanning unit includes a first beam splitter and a two-dimensional galvanometer, wherein the first beam splitter separates the detection light of a specified wavelength from the modulated detection light, and the two-dimensional galvanometer adjusts the deflection angle of the detection light of the specified wavelength in different dimensions of the X and Y axes respectively.

8. The semiconductor material defect detection device according to claim 7, characterized in that: The incident light scanning unit further comprises a field lens and a tube lens, and the field lens and the tube lens are arranged between the two-dimensional galvanometer and the incident light high-speed zoom unit.

9. The semiconductor material defect detection device according to claim 8, characterized in that: The incident light high-speed zoom unit comprises a second beam splitter and a zoom objective lens. The second beam splitter causes the two-dimensional scanning beam to be incident on the zoom objective lens. The zoom objective lens causes the two-dimensional scanning beam to be focused on different depths of the sample to be tested for scanning.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the computer program implements the instructions of the semiconductor material defect detection method according to any one of claims 1 to 4.

Images