Method and system for detecting fault type of 4H-SiC single crystal wafer

Abstract

The invention discloses a 4H-SiC single crystal wafer fault type detection method and system. The method comprises the following steps: irradiating a to-be-detected surface of a to-be-detected 4H-SiC single crystal wafer by using a laser; acquiring a plurality of fluorescence images of the to-be-detected surface irradiated by the laser based on the CCD array camera; the plurality of fluorescence images are respectively obtained by CCD cameras carrying different wavelength pass filters; splicing the plurality of fluorescent images to obtain a wafer fluorescent panoramic image; and determining the fault type of the to-be-detected 4H-SiC single crystal wafer based on the bright and dark stripes in the fluorescent panoramic image of the wafer and the epitaxial growth result of the to-be-detected 4H-SiC single crystal wafer. According to the method provided by the invention, the sample is not contacted and damaged in the whole detection process, the two-dimensional laser scanning mode is rapid, efficient and high in sensitivity, information such as morphology, distribution and density of fault can be directly obtained, and the result is visual and reliable. Meanwhile, the detection result obtained by the method provided by the invention can provide guidance for the optimization of the subsequent generation process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor material manufacturing and detection, and particularly relates to a 4H-SiC single crystal wafer layer defect type detection method and system. BACKGROUND

[0002] Silicon carbide, as a representative of the third generation semiconductor material, has a broad application prospect in the field of high-temperature, high-frequency and high-power devices due to its excellent physical and chemical properties, and 4H-SiC is one of the most main crystal forms. However, in the process of crystal growth and wafer processing, various crystal defects will inevitably be introduced, and stacking faults are a common extended defect.

[0003] Stacking faults will significantly affect the mobility of carriers, leading to an increase in device leakage current and a decrease in breakdown voltage, which seriously restricts the performance and reliability of the device. Therefore, it is crucial to quickly and accurately detect stacking faults in substrate preparation and epitaxial processes.

[0004] At present, the methods for detecting stacking faults mainly include chemical etching, transmission electron microscopy and X-ray topography, etc. The molten KOH etching method is simple to operate and easy to implement, but it will damage the sample and cannot be used for online detection, and the interpretation of the etching morphology depends on experience. Although the transmission electron microscopy method has extremely high resolution, the sample preparation is complex and time-consuming, and it is destructive, so it can only be used for sampling inspection and cannot be used for large-area and full-slice detection. The X-ray topography device is expensive, the detection speed is slow, and the spatial resolution is relatively low. SUMMARY

[0005] The embodiments of the present application provide a 4H-SiC single crystal wafer layer defect type detection method and system to solve the problem that the prior art cannot quickly and accurately identify the type of stacking faults at low cost.

[0006] In a first aspect, the present application provides a 4H-SiC single crystal wafer layer defect type detection method, comprising: irradiating a to-be-detected surface of a to-be-detected 4H-SiC single crystal wafer with a laser; acquiring a plurality of fluorescence images of the to-be-detected surface after being irradiated by the laser based on a CCD array camera; the plurality of fluorescence images are acquired by CCD cameras carrying different wavelength pass filters; stitching the plurality of fluorescence images to obtain a wafer fluorescence panoramic image; determining the type of layer defects of the to-be-detected 4H-SiC single crystal wafer based on bright and dark stripes in the wafer fluorescence panoramic image and an epitaxial growth result of the to-be-detected 4H-SiC single crystal wafer.

[0007] In some embodiments, the type of stacking faults includes a Shockley type or a Frank type; the epitaxial growth result includes inheritability or non-inheritability. determining the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer, including: when the epitaxial growth result of the to-be-tested 4H-SiC single crystal wafer is non-inheritability, determining the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer as the Frank type; when the epitaxial growth result of the to-be-tested 4H-SiC single crystal wafer is inheritability, determining the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer as the Shockley type.

[0008] Based on the scheme, by establishing a simple and reliable classification logic, the ambiguity of relying on experience is avoided, and the accuracy and efficiency of the determination of the type of stacking faults are improved.

[0009] In some embodiments, the CCD array camera includes four groups of CCD cameras, and the four groups of CCD cameras are respectively equipped with 430nm short-wave pass filter, 460nm short-wave pass filter, 500nm long-wave pass filter and 660nm long-wave pass filter.

[0010] Based on the scheme, by collecting multiple wavelength fluorescence signals, rich feature information basis is provided for subsequent differentiation of stacking fault subtypes based on bright and dark stripe differences, and multiple wavelength combinations can capture specific differences in fluorescence response of different stacking fault types, ensuring the effectiveness and accuracy of subsequent stripe analysis.

[0011] In some embodiments, based on the CCD array camera, a plurality of fluorescence images of the to-be-tested surface after being irradiated by the laser are acquired, including: The laser and the CCD array camera are used to perform two-dimensional scanning synchronously at a step distance of 5-50μm to obtain the plurality of fluorescence images.

[0012] Based on the scheme, the spatial continuity and resolution of image acquisition are ensured, and the step distance of 5-50μm balances the detection speed and image detail retention, avoids missing small stacking faults or excessive redundant data, provides a reliable image source for splicing high-precision wafer fluorescence panoramic images, and improves the integrity of the panoramic image and the accuracy of subsequent analysis.

[0013] In some embodiments, the Shockley type includes a first Shockley type, a second Shockley type, a third Shockley type or a fourth Shockley type; determining the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer, including: When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are bright stripes, bright stripes, dark stripes and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be a type of Shockley-type wafer. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, bright stripes and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be a type II Shockley-type wafer. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, dark stripes and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be of type three Shockley type. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters, and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, bright stripes, dark stripes, and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be of type four Shockley type.

[0014] In some embodiments, the Frank type includes an internal Frank type, an external Frank type, or a multi-layer stacked type; Determining the stacking fault type of the 4H-SiC single crystal wafer under test includes: When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, dark stripes and dark stripes in sequence, the 4H-SiC single crystal wafer under test is determined to be of intrinsic Frank type. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are bright stripes, bright stripes, dark stripes and dark stripes in sequence, the 4H-SiC single crystal wafer under test of the Frank type is determined to be of external Frank type. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters, and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, bright stripes, dark stripes, and dark stripes in sequence, it is determined that the Frank-type 4H-SiC single crystal wafer under test is a multilayer stacked type.

[0015] In some embodiments, before irradiating the test surface of the 4H-SiC single crystal wafer with a laser, the method further includes: The test surface of the 4H-SiC single crystal wafer to be tested is polished and cleaned.

[0016] In some embodiments, irradiating the test surface of a 4H-SiC single crystal wafer with a laser includes: A laser with a power density of 10mW / cm² to 1000mW / cm² is used to vertically irradiate the test surface of the 4H-SiC single crystal wafer with an ultraviolet laser in the wavelength range of 300nm to 360nm.

[0017] In some embodiments, the method further includes: The stacking fault type, stacking fault number, and stacking fault parameters in the wafer fluorescence panoramic image are obtained, and a defect density distribution map is generated.

[0018] Based on this solution, the stacking fault distribution pattern on the wafer surface is presented intuitively through a visualization distribution map, providing direct defect distribution data support for process improvement and yield enhancement, and enhancing the engineering application value of the test results.

[0019] Secondly, this application also provides a 4H-SiC single-crystal wafer stacking fault type detection system, applied to the detection method described in the first aspect, comprising: A scanning platform is provided, which is equipped with a laser and a CCD camera array; the laser is used to emit laser light to illuminate the surface of the 4H-SiC single crystal wafer under test; the CCD camera array is used to acquire multiple fluorescence images of the surface under test after being illuminated by the laser. The processor is used to stitch together the multiple fluorescence images to obtain a panoramic fluorescence image of the wafer; based on the bright and dark stripes in the panoramic fluorescence image of the wafer and the epitaxial growth results of the 4H-SiC single crystal wafer under test, the stacking fault type of the 4H-SiC single crystal wafer under test is determined.

[0020] Compared with the prior art, the beneficial effects of the present invention are: 1. This application acquires fluorescence images of the 4H-SiC single crystal wafer under test using a laser and a CCD camera array. Based on the analysis of the fluorescence images, the stacking fault type detection results are obtained. The entire detection process is non-contact and does not damage the wafer. The detected wafer can continue to be used for subsequent processes.

[0021] 2. This application uses two-dimensional scanning technology, which can automatically scan the entire wafer in a few minutes and is applicable to wafers of various specifications.

[0022] 3. This application has high sensitivity and can detect stacking faults at the micrometer or even submicrometer scale.

[0023] 4. Intuitive and accurate: It can directly obtain information on the morphology, distribution, and density of stacking faults, and the results are intuitive and reliable. Furthermore, based on imaging and further observation of crystal inheritance, the specific types of stacking faults can be classified, thereby guiding the optimization of production processes.

[0024] 5. Easy to integrate: This method is easy to automate and can be integrated into the production line for online quality monitoring. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the structure of a 4H-SiC single crystal wafer stacking fault type detection system provided in an embodiment of this application; Figure 2 A flowchart illustrating a method for detecting stacking fault types in a 4H-SiC single crystal wafer, as provided in this application embodiment; Figure 3 This is a schematic diagram of a panoramic fluorescence image of a 4H-SiC single-crystal wafer in an embodiment of this application for detecting stacking fault types. Figure 4 The defect density distribution map generated in a method for detecting stacking faults in a 4H-SiC single crystal wafer provided in this application embodiment. Detailed Implementation

[0026] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of this application.

[0027] Stacking faults can be divided into two main categories based on their formation mechanism: Shockley-type stacking faults (SSF) and Frank-type stacking faults (FSF). SSF originates from partial atomic plane slip. This type of stacking fault does not change the number of basal planes; its partial dislocations (PDs) have boundaries parallel to the basal planes. Based on the number of partial dislocations, Shockley-type stacking faults can be further classified into 1SSF, 2SSF, 3SSF, and 4SSF, etc., with the main differences being the number of atomic plane slips and the degree of crystal structure change caused. One atomic plane slip forms 1SSF, two atomic plane slips form 2SSF, and so on; three or four atomic plane slips form 3SSF and 4SSF, respectively.

[0028] Frank-type stacking faults are formed by the insertion or deletion of atomic layers on the basal plane, and some of their dislocation boundaries are perpendicular to the basal plane. There are three main types of stacking faults: intrinsic FSFs, extrinsic FSFs, and multilayer FSFs (also known as multilayer stacking faults) with multiple additional (missing) layers on the basal plane. The formation mechanisms of different types of stacking faults are different, and their impact on subsequent epitaxy and device performance varies.

[0029] The presence of stacking faults alters the local band structure of 4H-SiC single crystals, causing their photoluminescence spectra to exhibit characteristic peaks distinct from those of perfectly crystalline regions. Furthermore, in fluorescence imaging, stacking fault regions exhibit different contrasts with the surrounding matrix due to varying recombination luminescence efficiencies, and different types of stacking faults also display different signal imaging within different fluorescence imaging ranges.

[0030] To achieve accurate detection of stacking fault types in 4H-SiC single crystals, embodiments of this application provide a system for detecting stacking fault types in 4H-SiC single crystal wafers.

[0031] See Figure 1 This is a schematic diagram of the structure of a 4H-SiC single crystal wafer stacking fault type detection system provided in an embodiment of this application.

[0032] like Figure 1 As shown, the detection system provided in this application embodiment may include: A scanning platform (not shown in the figure) is provided with a laser and a CCD camera array. The laser is used to emit laser light to illuminate the test surface of the 4H-SiC single crystal wafer to be tested, which can be placed on a silicon carbide substrate. When the laser is incident on the test surface of the 4H-SiC single crystal wafer, a fluorescence signal will be generated. The CCD camera array can generate multiple fluorescence images based on the captured fluorescence signal.

[0033] The detection system is also equipped with a processor (not shown in the figure). The processor can be set in the scanning platform or in a remote control platform. The processor is used to stitch together the multiple fluorescence images to obtain a panoramic fluorescence image of the wafer. Based on the bright and dark stripes in the panoramic fluorescence image of the wafer and the epitaxial growth results of the 4H-SiC single crystal wafer to be tested, the stacking fault type of the 4H-SiC single crystal wafer to be tested is determined.

[0034] In some embodiments, the detection system also includes a stage for holding the sample and a vacuum system for providing a vacuum environment.

[0035] In some embodiments, the scanning platform may include a track that works in conjunction with a laser and a CCD camera array, enabling the laser and CCD camera to move synchronously and acquire fluorescence images of every point on the entire wafer surface, making it suitable for wafer inspection of different sizes.

[0036] Specifically, in practical applications, the above system can utilize the following detection methods to achieve rapid and accurate detection of stacking fault types.

[0037] See Figure 2 The flowchart below shows a method for detecting stacking fault types in a 4H-SiC single crystal wafer, as provided in an embodiment of this application.

[0038] like Figure 2 As shown, based on the above system, this application provides a method for detecting stacking fault types in a 4H-SiC single crystal wafer, comprising: S100: The test surface of the 4H-SiC single crystal wafer is illuminated by a laser.

[0039] In some embodiments, preparatory operations may be performed before executing step S100, such as: Constructing a detection system: Select an appropriate laser emitter (laser). In some examples, wavelengths below 300nm have too shallow a penetration depth in 4H-SiC, and laser energy above 360nm is too weak, both of which will result in unclear stacking fault imaging. Ultraviolet lasers with a wavelength range of 300nm to 360nm can be selected as the excitation source, and the power density of the laser should be 10mW / cm² to 1000mW / cm² to ensure appropriate penetration depth and laser energy.

[0040] In some embodiments, the laser emitter can be positioned directly above the 4H-SiC single crystal wafer under test, illuminating the surface under test with a vertical or oblique incident direction.

[0041] Selection of the sample to be tested: In this embodiment of the application, a 4H-SiC single crystal substrate wafer with a nitrogen doping concentration of 1E15~1E20 can be selected as the 4H-SiC single crystal wafer to be tested.

[0042] Sample pretreatment: If the surface of the 4H-SiC single crystal wafer to be tested is uneven or contains impurities, a pretreatment process is required. For example, the surface to be tested on the wafer can be polished first, and then ultrasonically cleaned with acetone, ethanol and deionized water in sequence, and then dried with high-purity nitrogen to obtain a clean and flat surface to be tested, thereby improving the accuracy of subsequent detection.

[0043] Sample placement: The polished 4H-SiC single crystal wafer to be tested is placed on the stage, and a vacuum environment is created around the stage using a vacuum system.

[0044] Once the above preparations are completed, step S100 can be executed. Specifically, the laser can be turned on when the laser emitted from the laser is irradiated vertically or at an angle onto the test surface of the 4H-SiC single crystal wafer and the test surface is facing the objective lens of the CCD array camera. The laser can use an ultraviolet laser with a wavelength range of 300nm to 360nm as the excitation source.

[0045] S200: Acquire multiple fluorescence images of the surface under test after it has been irradiated by the laser using a CCD array camera; the multiple fluorescence images are acquired by CCD cameras equipped with different wavelength pass filters.

[0046] Specifically, in some embodiments, the CCD array camera may include four sets of CCD cameras, each equipped with a 430nm short-pass filter, a 460nm short-pass filter, a 500nm long-pass filter, and a 660nm long-pass filter, so that four sets of fluorescence images can be acquired simultaneously through different CCD cameras.

[0047] Furthermore, when acquiring fluorescence images using the CCD array camera, the movable structure of the scanning platform can be utilized to achieve automatic scanning of the wafer surface. For example, the laser and the CCD array camera can be used to simultaneously perform two-dimensional scanning at a step distance of 5-50 μm to obtain the multiple fluorescence images. The step distance is preferably 20 μm, and the integration time for the CCD camera array to simultaneously acquire the fluorescence image of each point is 100 ms-1000 ms, preferably 200 ms.

[0048] In some embodiments, the scanning platform can move automatically to achieve global wafer testing and is equipped with a CCD camera array to collect fluorescence imaging of the sample generated under laser excitation. Since the electrons in the excited state are insufficient to transition back to the stable ground state when the integration time is less than 100ms, and the electron transition has become stable when the integration time is greater than 1000ms, neither of which can form a clear photoluminescence image. Therefore, the integration time in this embodiment can be 100-1000ms.

[0049] After performing the image signal acquisition step in step S200, the acquired image can be further processed by the processor, see steps S300-S400.

[0050] S300: The multiple fluorescence images are stitched together to obtain a panoramic image of the wafer fluorescence.

[0051] See Figure 3 This is a schematic diagram of a panoramic fluorescence image of a 4H-SiC single crystal wafer in an embodiment of this application for detecting stacking fault types.

[0052] like Figure 3 As shown, multiple fluorescence images acquired at different wavelengths can be stitched together in an array to form a single image, and the corresponding information can be displayed for each image.

[0053] In some embodiments, before stitching the fluorescence images, image processing operations can be performed on each fluorescence image, such as adjusting parameters like image size, resolution, sharpness, aspect ratio, and grayscale to ensure consistency, so as to facilitate the identification of differences between the images. The image processing methods should be applicable to existing technologies and are not limited in the embodiments of this application.

[0054] S400: Based on the bright and dark stripes in the wafer fluorescence panoramic image and the epitaxial growth results of the 4H-SiC single crystal wafer under test, determine the stacking fault type of the 4H-SiC single crystal wafer under test.

[0055] The characteristic photoluminescence emission wavelengths of each stacking fault, determined through photoluminescence studies, are used to classify stacking faults as shown in the table below:

[0056] Frank-type stacking faults (FSFs) are formed by removing or inserting a single atom layer on the close-packed plane (basal plane) (in 4H-SiC, this means changing a period of c / 2). This results in its Burgers vector b = ±1 / 3. <0001> (c-axis direction), perpendicular to the basal plane. Because the Burgers vector of the FSF has a c-axis component, it cannot be eliminated or moved by slip on the basal plane. Once formed, it becomes part of the crystal structure, like a "marker," and is perfectly inherited by subsequent crystal growth. That is, stacking faults will extend along the c-axis from the substrate into the epitaxial layer. Formed by the slip of Shockley partial dislocations, it does not increase or decrease the number of atomic layers, but only changes the stacking order. Its Burgers vector is within the basal plane and will not be completely replicated during continuous epitaxial growth. Based on these characteristics, the type of stacking fault, whether it is Shockley-type or Frank-type, can be determined by whether the epitaxial growth result of a 4H-SiC single crystal wafer is inheritable.

[0057] Specifically, step S400 may include: S410: When the epitaxial growth result of the 4H-SiC single crystal wafer under test is non-inheritable, the stacking fault type of the 4H-SiC single crystal wafer under test is determined to be the Frank type; S420: When the epitaxial growth result of the 4H-SiC single crystal wafer under test is inheritable, the stacking fault type of the 4H-SiC single crystal wafer under test is determined to be Shockley type.

[0058] Furthermore, considering the different display states of various types of stacking faults in the combined images of four filter types (430nm short-pass filter, 460nm short-pass filter, 500nm long-pass filter, and 660nm long-pass filter), the specific type of stacking fault can be further determined by combining the properties of FSF and SSF with the bright and dark stripes in the wafer fluorescence panoramic image.

[0059] For example, the specific type of stacking fault can be determined based on the following table:

[0060] As shown in the table above, the following process can be used to determine the specific type of stacking fault: When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are bright stripes, bright stripes, dark stripes and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be a type of Shockley-type wafer. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, bright stripes and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be a type II Shockley-type wafer. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, dark stripes and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be of type three Shockley type. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters, and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, bright stripes, dark stripes, and dark stripes in sequence, the Shockley-type 4H-SiC single crystal wafer under test is determined to be of type four Shockley type.

[0061] When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, dark stripes and dark stripes in sequence, the 4H-SiC single crystal wafer under test is determined to be of intrinsic Frank type. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters and 660nm long-pass filters in the wafer fluorescence panoramic image are bright stripes, bright stripes, dark stripes and dark stripes in sequence, the 4H-SiC single crystal wafer under test of the Frank type is determined to be of external Frank type. When the bright and dark stripes in the images corresponding to the CCD cameras equipped with 430nm short-pass filters, 460nm short-pass filters, 500nm long-pass filters, and 660nm long-pass filters in the wafer fluorescence panoramic image are dark stripes, bright stripes, dark stripes, and dark stripes in sequence, it is determined that the Frank-type 4H-SiC single crystal wafer under test is a multilayer stacked type.

[0062] In the embodiments of this application, by identifying the bright and dark stripes in the panoramic fluorescent image of the wafer and combining them with the epitaxial growth results, the specific type of stacking fault can be accurately identified, providing guidance and optimization reference for subsequent processes.

[0063] It should be noted that, in the embodiments of this application, when performing step S400, the two judgment criteria—the bright and dark stripes in the wafer fluorescence panoramic image and the epitaxial growth results—can be used separately or in combination in different scenarios. For example, in a scenario where it is only necessary to determine whether the stacking fault is Shockley-type or Frank-type, only the epitaxial growth results are needed; or, if the stacking fault has already been determined to be Shockley-type or Frank-type, only the bright and dark stripes are needed for judgment.

[0064] In some embodiments, if it is necessary to use two criteria simultaneously to determine the type of stacking fault, in addition to the methods provided in the aforementioned embodiments, it is also possible to first use bright and dark stripes to narrow down the range, and then use the epitaxial growth results to make the determination, for example: See Figure 3 In the illustrated embodiment, several bright and dark bands can be observed in different wafer fluorescence images based on the panoramic wafer fluorescence image. Figure 3 Medium stacking faults exhibit more pronounced bright fringes in 430nm short-pass imaging, weaker bright fringes in 460nm short-pass imaging, and dark fringes in 500nm and 660nm long-pass imaging. Referring to the correspondence in the previous table, we can first obtain results for stacking fault types 1SSF or Extrinsic FSF. Then, homoepitaxial growth can be performed on the wafer. If the stacking fault is observed to be completely replicated into the epitaxial layer, the stacking fault type can be accurately determined to be Extrinsic FSF.

[0065] In the embodiments of this application, there is no restriction on whether to determine the light and dark stripes first or to determine the epitaxial growth results first, and it should be considered that the same technical effect can be achieved.

[0066] Furthermore, in some embodiments, after the specific type of stacking fault is determined by step S400, image analysis software can be used to further analyze the bright and dark stripes in the image, such as measuring the size of the stacking fault, calculating and counting the number of stacking faults, etc.

[0067] See Figure 4 This is a defect density distribution map generated in a method for detecting stacking fault types in a 4H-SiC single crystal wafer provided in an embodiment of this application.

[0068] In some embodiments, after acquiring the stacking fault type, number, and parameters from the wafer fluorescence panoramic image, a defect density distribution map can be generated to visually display the distribution of stacking faults in the wafer. Specifically, specialized image analysis software can be used to mark the location, size, and other information of each detected stacking fault, and based on this information, a two-dimensional or three-dimensional defect density distribution map can be generated. See alsoFigure 4 In a defect density distribution map, different types of stacking faults can be distinguished by icons of different shapes or colors. The density and range of the icons represent different stacking fault densities, making the distribution of stacking faults in the wafer immediately clear. Such a defect density distribution map is of great reference value for evaluating wafer quality, guiding subsequent process optimization, and improving product yield.

[0069] As can be seen from the above technical solution, the detection method provided in this application involves irradiating the test surface of a 4H-SiC single crystal wafer with a laser and acquiring a panoramic fluorescence image of the wafer using a CCD camera array. The type of stacking fault is quickly and accurately determined using the bright and dark fringes in the panoramic fluorescence image and the epitaxial growth results. The method provided in this application does not contact or damage the sample during the entire detection process. Furthermore, it is fast, efficient, and highly sensitive through two-dimensional laser scanning, directly obtaining information such as the morphology, distribution, and density of stacking faults, resulting in intuitive and reliable results. Simultaneously, the detection results obtained through the method provided in this application can provide guidance for optimizing subsequent production processes. It is easy to implement and can be applied to online quality monitoring of production lines.

[0070] The above detailed embodiments further illustrate the purpose, technical solution, and beneficial effects of the embodiments of this application. It should be understood that the above are merely specific embodiments of the embodiments of this application and are not intended to limit the protection scope of the embodiments of this application. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solutions of the embodiments of this application should be included within the protection scope of the embodiments of this application.

Claims

1. A method for detecting a type of a layer fault of a 4H-SiC single crystal wafer, characterized by, The method comprises: A laser is used to irradiate a to-be-tested surface of a to-be-tested 4H-SiC single crystal wafer; A plurality of fluorescence images of the to-be-tested surface after being irradiated by the laser are acquired based on a CCD array camera; the plurality of fluorescence images are acquired by the CCD camera carrying different wavelength pass filters respectively; The plurality of fluorescence images are spliced to obtain a wafer fluorescence panoramic image; Based on bright and dark stripes in the wafer fluorescence panoramic image and epitaxial growth results of the to-be-tested 4H-SiC single crystal wafer, a type of stacking faults of the to-be-tested 4H-SiC single crystal wafer is determined.

2. The 4H-SiC single crystal wafer dislocation type detection method according to claim 1, characterized by, The type of stacking faults comprises a Shockley type or a Frank type; the epitaxial growth results comprise inheritability or non-inheritability; Determining the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer comprises: When the epitaxial growth results of the to-be-tested 4H-SiC single crystal wafer are non-inheritability, it is determined that the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer is the Frank type; When the epitaxial growth results of the to-be-tested 4H-SiC single crystal wafer are inheritability, it is determined that the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer is the Shockley type.

3. The 4H-SiC single crystal wafer dislocation type detection method according to claim 2, characterized by, The CCD array camera comprises four groups of CCD cameras, and the four groups of CCD cameras carry 430 nm short-wave pass filters, 460 nm short-wave pass filters, 500 nm long-wave pass filters and 660 nm long-wave pass filters respectively.

4. The 4H-SiC single crystal wafer dislocation type detection method according to claim 3, characterized by, Acquiring the plurality of fluorescence images of the to-be-tested surface after being irradiated by the laser based on the CCD array camera comprises: The laser and the CCD array camera are used to perform two-dimensional scanning synchronously at a step distance of 5-50 μm to obtain the plurality of fluorescence images.

5. The 4H-SiC single crystal wafer dislocation type detection method according to claim 4, characterized by, The Shockley type comprises a first Shockley type, a second Shockley type, a third Shockley type or a fourth Shockley type; Determining the type of stacking faults of the to-be-tested 4H-SiC single crystal wafer comprises: When the bright and dark stripes in the images corresponding to the CCD cameras carrying the 430 nm short-wave pass filters, the 460 nm short-wave pass filters, the 500 nm long-wave pass filters and the 660 nm long-wave pass filters in the wafer fluorescence panoramic image are bright stripes, bright stripes, dark stripes and dark stripes respectively, it is determined that the to-be-tested 4H-SiC single crystal wafer of the Shockley type is the first Shockley type; When the bright and dark stripes in the images corresponding to the CCD cameras carrying the 430 nm short-wave pass filters, the 460 nm short-wave pass filters, the 500 nm long-wave pass filters and the 660 nm long-wave pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, bright stripes and dark stripes respectively, it is determined that the to-be-tested 4H-SiC single crystal wafer of the Shockley type is the second Shockley type; When the bright and dark stripes in the images corresponding to the CCD cameras carrying the 430 nm short-wave pass filters, the 460 nm short-wave pass filters, the 500 nm long-wave pass filters and the 660 nm long-wave pass filters in the wafer fluorescence panoramic image are dark stripes, dark stripes, dark stripes and dark stripes respectively, it is determined that the to-be-tested 4H-SiC single crystal wafer of the Shockley type is the third Shockley type; When the bright and dark stripes in the image corresponding to the CCD camera with the 430nm short wave pass filter, the 460nm short wave pass filter, the 500nm long wave pass filter and the 660nm long wave pass filter in the wafer fluorescence panoramic image are dark stripe, bright stripe, dark stripe, dark stripe in turn, the 4H-SiC single crystal wafer to be measured of the Shokley type is determined to be the fourth Shokley type.

6. The 4H-SiC single crystal wafer dislocation type detection method according to claim 4, characterized by, The Frank type includes an intrinsic Frank type, an extrinsic Frank type or a multi-layer stacking type; The method further comprises: When the bright and dark stripes in the image corresponding to the CCD camera with the 430nm short wave pass filter, the 460nm short wave pass filter, the 500nm long wave pass filter and the 660nm long wave pass filter in the wafer fluorescence panoramic image are dark stripe, dark stripe, dark stripe, dark stripe in turn, the 4H-SiC single crystal wafer to be measured of the Frank type is determined to be the intrinsic Frank type; When the bright and dark stripes in the image corresponding to the CCD camera with the 430nm short wave pass filter, the 460nm short wave pass filter, the 500nm long wave pass filter and the 660nm long wave pass filter in the wafer fluorescence panoramic image are bright stripe, bright stripe, dark stripe, dark stripe in turn, the 4H-SiC single crystal wafer to be measured of the Frank type is determined to be the extrinsic Frank type; When the bright and dark stripes in the image corresponding to the CCD camera with the 430nm short wave pass filter, the 460nm short wave pass filter, the 500nm long wave pass filter and the 660nm long wave pass filter in the wafer fluorescence panoramic image are dark stripe, bright stripe, dark stripe, dark stripe in turn, the 4H-SiC single crystal wafer to be measured of the Frank type is determined to be the multi-layer stacking type.

7. The 4H-SiC single crystal wafer dislocation type detection method according to claim 1, characterized by, Before the laser is used to irradiate the to-be-measured surface of the 4H-SiC single crystal wafer to be measured, the method further comprises: The to-be-measured surface of the 4H-SiC single crystal wafer to be measured is polished and cleaned.

8. The 4H-SiC single crystal wafer dislocation type detection method according to claim 1, characterized by, The laser is used to irradiate the to-be-measured surface of the 4H-SiC single crystal wafer to be measured, comprising: The laser with a power density of 10mW / cm² to 1000mW / cm² is used to vertically irradiate the to-be-measured surface of the 4H-SiC single crystal wafer to be measured with ultraviolet laser in a wavelength range of 300nm to 360nm.

9. The 4H-SiC single crystal wafer dislocation type detection method according to claim 1, characterized by, The method further comprises: Obtaining the dislocation type, the dislocation number and the dislocation parameter in the wafer fluorescence panoramic image, and generating a defect density distribution map.

10. A 4H-SiC single crystal wafer dislocation type detection system applied to the detection method of any one of claims 1 to 9, characterized in that, Comprising: A scanning platform, the scanning platform is provided with a laser and a CCD camera array; the laser is used to emit laser to irradiate the to-be-measured surface of the 4H-SiC single crystal wafer to be measured; the CCD camera array is used to acquire a plurality of fluorescence images of the to-be-measured surface after being irradiated by the laser; A processor is used to splice the plurality of fluorescence images to obtain a wafer fluorescence panoramic image; based on the bright and dark stripes in the wafer fluorescence panoramic image and the epitaxial growth result of the 4H-SiC single crystal wafer to be measured, the dislocation type of the 4H-SiC single crystal wafer to be measured is determined.

Images