Method for amplifying TDs defects of silicon carbide epitaxial wafer based on time resolution technology
Through the fluorescence imaging detection method based on time resolution technology, the accuracy and non-destructive problems of dislocation defect detection of silicon carbide epitaxial layer in the prior art are solved, and defect detection with high sensitivity and high accuracy is achieved, supporting the sustainable utilization of materials.
Patent Information
- Application Number
- CN202311698837.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-12
- Publication Date
- 2025-06-13
AI Technical Summary
The prior art is difficult to accurately and non-destructively detect dislocation defects in the silicon carbide epitaxial layer, and commonly used methods can damage the surface morphology of the material, limiting the reprocessing and utilization of the material.
The fluorescence imaging detection method based on time resolution technology is adopted to analyze fluorescence attenuation kinetic information, eliminate interference from background and impurities, and amplify the dislocation defects of the silicon carbide epitaxial layer.
Accurate detection of dislocation defects in the epitaxial layer of silicon carbide is achieved, avoiding material surface damage, improving detection sensitivity and accuracy, and supporting the sustainable use of materials.
Smart Images
Figure CN120142252A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of silicon carbide defect detection, and specifically relates to a method for amplifying TDs defects in silicon carbide epitaxial wafers based on time-resolved technology. Background Art
[0002] Silicon carbide is a wide-bandgap semiconductor. Due to its excellent properties, it has broad application prospects in high-power, high-temperature, and high-frequency devices. However, the extended defects that appear in the epitaxial layer and their adverse effects on devices limit the commercialization of 4H-SiC power devices. These extended defects, especially non-surface morphology defects such as dislocations, have a great impact on the performance of bipolar devices and Schottky devices. Dislocations are a common type of electroactive defect existing in silicon carbide crystals, and their typical density can reach 10 4 cm -2 . They introduce a large number of carrier recombination centers in the silicon carbide material, reducing the local density of free carriers under external excitation; dislocations can also act as carrier generation centers, leading to an increase in leakage current; in addition, impurity atoms are more likely to aggregate or separate around dislocations. Therefore, to solve these problems, it is crucial to develop or upgrade technologies to accurately visualize these extended defects. If there are continuously updated detection technologies to assist in continuously reducing defects, the market for silicon carbide devices will be further expanded. Currently, many technologies have been developed to detect dislocation defects in SiC. Among them, identifying defects through the etch pits after KOH etching is the most commonly used method. However, this will seriously damage the surface morphology of the material and easily lose a lot of surface layer information. In addition, this method is extremely unfavorable for the reprocessing and utilization of the material after detection. Therefore, from the perspective of sustainable development, non-destructive detection methods are more friendly, including X-ray imaging, fluorescence imaging, and cathodoluminescence imaging, etc. However, these methods are only limited to the detection of defect positions and cannot provide more information about the defects. Therefore, we have developed a fluorescence imaging detection technology with time-resolved function, which can detect the fluorescence decay kinetic information at different positions. Using this technology, we can exclude the interference of background and impurities, thereby obtaining more information about the nature of the defects themselves. Based on this time-resolved fluorescence detection, we can amplify the dislocation defects in the silicon carbide epitaxial layer, making the defect detection technology more accurate and sensitive. Summary of the Invention
[0003] In view of the above-mentioned problems in the prior art, the present invention proposes a method for amplifying silicon carbide epitaxial TDs defects based on time-resolved technology: first, build a fluorescence detection device, then collect the sample information with silicon carbide TDs defects, and then extract the fluorescence intensity imaging maps at different times, and the silicon carbide defects that expand with the increase of the extraction time can be obtained.
[0004] The technical solution adopted by the present invention to achieve the above object is as follows:
[0005] A device for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology, comprising an excitation light source module and a detection device module arranged in sequence on the optical path. Among them,
[0006] The excitation light source module includes a femtosecond laser, a beam shrinking component, a third harmonic generation component, a beam expanding component, and an attenuation sheet component. The laser emitted by the femtosecond laser sequentially passes through the mirror component 1, the beam shrinking component, the third harmonic generation component, the beam expanding component, the mirror component 2, and the attenuation sheet component and then enters the detection device module;
[0007] The detection device module includes a microscope system, an electric displacement stage, a spectrometer system, a photon detector, and an industrial control computer. The light source output by the excitation light source module enters the microscope system and is reflected into the spectrometer system through a preset optical path. The photon detector, the spectrometer system, and the electric displacement stage are all connected to the industrial control computer. The silicon carbide epitaxial wafer to be measured is placed on the electric displacement stage located below the microscope system.
[0008] A filter and a beam booster are sequentially arranged between the third harmonic generation component and the beam expanding component.
[0009] The beam shrinking component is composed of lens groups with focal lengths of f = 400 mm and f = -50 mm.
[0010] The third harmonic generation component is composed of a second harmonic generation BBO crystal and a sum frequency BBO crystal. A calcite is arranged between the second harmonic generation BBO crystal and the sum frequency BBO crystal.
[0011] The beam expanding component includes lens groups with focal lengths of f = 30 mm and f = 100 mm.
[0012] A method for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology includes the following steps:
[0013] 1) Control the electric displacement stage through the industrial control computer to reach a preset pixel position within a preset scanning range and step size. Output fluorescence spectrograms at different times at the preset pixel position through the spectrometer system to obtain the corresponding fluorescence spectrograms at each preset pixel position; obtain the fluorescence decay kinetic curve diagram corresponding to the preset pixel position at different times through the photon detector;
[0014] 2) Calculate the total number of photons in all time channels in the fluorescence decay kinetic curve at each preset pixel position to obtain the total fluorescence intensity at this pixel position;
[0015] 3) Control the electric displacement stage to move, repeat step 2), traverse and record the total fluorescence intensity of all pixel positions to obtain a fluorescence intensity imaging map with time-resolved information;
[0016] 4) Extract the fluorescence decay kinetic curve graphs at different pixel positions in the fluorescence intensity imaging graph, compare the fluorescence decay conditions in the defective area and the non-defective area, and determine the area of the defect and the defect coordinates;
[0017] 5) Extract the fluorescence intensity imaging graphs of the defect at different decay times, determine the state of the defect at different decay times and the fluorescence intensity contour graphs related to any row and position at different decay times at the defect position, so as to show the contrast degree between the defective area and the surrounding non-defective area.
[0018] The said step 2) includes the following steps:
[0019] 2.1) Mobilize the spectrometer system to collect fluorescence spectra;
[0020] 2.2) Mobilize the single-photon counter to collect the fluorescence decay kinetic curve graph at the wavelength where the fluorescence peak value of the sample is located;
[0021] 2.3) Mobilize the electric displacement stage to collect the fluorescence decay kinetic curve graphs at different pixel positions;
[0022] 2.4) Perform row summation on the fluorescence intensities at each time point in the fluorescence decay kinetic curve graph of each pixel position to obtain the fluorescence total intensity imaging graph of the preset scanning area.
[0023] The said step 4) is specifically as follows:
[0024] In the fluorescence intensity imaging graph with time-resolved information, take the positions with a contrast of 20%-30% as the defect positions, and extract the defect position coordinates, where the contrast = (fluorescence intensity of the material body - fluorescence intensity of the defect position) / fluorescence intensity of the material body;
[0025] The said step 5) includes the following steps:
[0026] 5.1) Customize the starting time on the decay curve corresponding to each pixel point, and take the time corresponding to the last time channel in the decay curve of each pixel point as the cut-off time. The starting time and cut-off time of all pixel points are kept consistent. Extract the total number of photons in all time channels within the starting time and cut-off time to obtain the fluorescence intensity imaging graph of the defect position at different decay times;
[0027] 5.2) According to the coordinates of the defect position, extract the intensity data of the row or column where the defect is located after different decay times and perform normalization processing;
[0028] 5.3) Compare all the normalized data to obtain the fluorescence intensity contour graph of the defect at different decay times, and obtain the contrast value of the time-related defect and the surrounding area.
[0029] The present invention has the following beneficial effects and advantages:
[0030] The present invention can detect the fluorescence decay kinetic information at different positions, eliminate the interference of background and impurities, and thus obtain more information about the nature of the defects themselves. Based on the time-resolved fluorescence detection method, the dislocation defects in the silicon carbide epitaxial layer can be amplified, making the defect detection technology more accurate and sensitive. Description of the Drawings
[0031] Figure 1 It is a technical roadmap of the self-built device;
[0032] Figure 2 It is a detailed view of the microscope in the self-built device;
[0033] Figure 3 It is an intrinsic fluorescence spectrum diagram of silicon carbide;
[0034] Figure 4 It is a fluorescence intensity distribution diagram of the defect area;
[0035] Figure 5 It is a flowchart of the operation of the data acquisition program and the data processing program;
[0036] Figure 6 It is a comparison diagram of the fluorescence kinetic decay between the defect area and the defect-free area around;
[0037] Figure 7 It is a comparison diagram of the fluorescence intensity imaging diagrams of the defects after different moments;
[0038] Figure 8 It is a comparison diagram of the intensities of the row where the defect is located after different decay times;
[0039] Figure 9 It is a specific operation flowchart of the whole set of technologies. Detailed Embodiments
[0040] The present invention will be further described in detail below with reference to the drawings and embodiments.
[0041] As Figure 9 shown, a method for amplifying TDs defects in silicon carbide epitaxy based on time-resolved technology is characterized by including the following steps in sequence:
[0042] 1) Set up a fluorescence imaging detection device with time-resolved function. The specific operation steps are as follows: Extract the femtosecond laser at 1030 nm (230 fs, 1 MHz, Light Conversion), and obtain the third harmonic of 1030 nm, which is 343 nm, through a series of operations: The fundamental frequency light emitted from the laser first passes through a lens group (f = 400 mm & f = -50 mm) for beam shrinking. This is to prevent the laser emitted from the laser from diverging and also to better perform the SHG frequency doubling operation on the fundamental frequency light in the future. Inject the 1030 nm light into its second harmonic BBO crystal, and femtosecond light at 515 nm can be obtained. The second harmonic generation efficiency can reach 20% - 30%; Send the 515 nm light and the remaining 1030 nm light after the second harmonic generation into the sum frequency BBO crystal together. The SFG sum frequency function of the crystal can sum the above two beams of light to produce 343 nm light. Still, adjust the position and angle of the crystal to adjust the 343 nm light to the best state under this condition. In addition, a 1000 nm short-pass filter and a 343 - 10 nm band-pass filter need to be added to obtain relatively clean 343 nm femtosecond light.
[0043] 2) The 343 nm femtosecond light obtained in the above steps is introduced into the microscope system as the excitation light source of the whole device: The laser first passes through an optical path elevator and is lifted to a position equal to the height of the microscope light inlet; Then it passes through a lens group (f = 30 mm & f = 100 mm), and the purpose is to expand and collimate the laser; There is a diaphragm before and after the lens group (the purpose of the diaphragm is to mark the position of the laser to ensure that the laser injected each time is correct); After passing through three mirrors and two attenuation sheets, the laser can be introduced into the microscope system.
[0044] 3) Optical path of the microscope system: The laser is introduced into the objective lens through a dichroic mirror (FF376-Di01-25x36, semrock), and the objective lens can focus the laser on the sample to excite the sample; the fluorescence signal generated by the sample will return from the objective lens to the aforementioned dichroic mirror together with the excitation light. This dichroic mirror can filter out the laser and only allow the fluorescence to pass through; the passed fluorescence passes through a beam splitter. Most of the fluorescence will be reflected by the beam splitter, and a small part of the fluorescence will pass through the beam splitter and enter the eyepiece at the upper end of the microscope or an external CCD system (268M-PH, QHYCCD). The CCD can sense the fluorescence signal for subsequent observation and adjustment of the excitation light state. In addition, for fluorescence imaging, we place the sample on a motorized displacement stage, and the range and step size of the displacement stage movement can be controlled through a program. The range and step size that the displacement stage can move are limited by its own production technical standards. We use two types of displacement stages. One is a piezoelectric ceramic nano-displacement stage (P-545.3C7&M-545, Physik Instrumente), which has a moving range of 200x200x200 μm and a minimum step size that can reach 20 nm; the other is a displacement stage controlled by a linear motor (RGW-120X100-O0.1, Sanying Motion Control Instruments Ltd), which has a moving range of 100x100 mm and a minimum step size that can reach 1 μm. For the sic dislocation defects to be detected, the scanning range only needs to be larger than the imaging size of a single defect (5-20 μm). Generally speaking, the scanning range is set at 20-60 μm and the step size is set at 0.2 - 0.5 μm for better imaging effects.
[0045] Most of the fluorescence reflected by the beam splitter described in Step 3 will be collected by the light collection system. The specific light collection method is as follows: The fluorescence passing through the beam splitter will be divided into two paths. One path passes through a lens (f = +100 mm) and is focused into a monochromator (Omni-λ3007i, Zolix), and the other path passes through a lens (f = +100 mm) and is focused into a single-photon counter (PMA Hybrid 06, Pico Quant). A PMT (PMTH-S1-R928, Zolix) is connected behind the monochromator and can be used to collect fluorescence spectra and also test fluorescence intensity imaging at different wavelengths (without time-resolved function). The ultraviolet single-photon counter can be used to collect fluorescence kinetic information and also perform fluorescence intensity imaging with time-resolved function.
[0046] The above-mentioned technical roadmap and microscope details are as Figure 1 and Figure 2 shown.
[0047] 2. Data acquisition: Use the self-developed Labview program in the laboratory to collect data. This program realizes functions such as laser scanning samples, collecting fluorescence spectrum information and fluorescence kinetics information by calling the original programs of equipment such as electric displacement stages, spectrometers, and single-photon counters. It can also further process the collected information to present a fluorescence intensity imaging map. The specific operation steps are described as follows:
[0048] 1) Place the silicon carbide sample to be tested on the electric displacement stage. After the excitation light power is adjusted to an appropriate range (for this equipment, an excitation light power of 0.1 - 0.5 mw is more appropriate), it is introduced into the microscope system to excite the silicon carbide sample. After setting the relevant parameters (scanning range 20x20μm, 80x80 pixel points, step size 0.25μm), the sample signal can be collected. Collect the fluorescence spectrum and fluorescence kinetics curve at the center of the scanning area. As Figure 3 shown, after the information is confirmed to be correct, scanning imaging can be performed to obtain a fluorescence spectrum map ( Figure 3 a) and a fluorescence decay kinetics curve map ( Figure 3 b).
[0049] 2) The fluorescence imaging detection equipment with time-resolved function can collect the fluorescence kinetics decay curve of each pixel point among 80x80 pixel points and calculate the total number of photons in all time channels in the curve. Through the movement of the displacement stage, the laser can scan the specified area and finally obtain a fluorescence total intensity imaging map with time-resolved information about this area, such as Figure 4 .
[0050] The working steps of the above acquisition program are shown in Figure 5 (a).
[0051] 3. Data analysis and processing: Use the self-developed Labview program in the laboratory to analyze data. This program is used to analyze the fluorescence total intensity imaging map with time-resolved function described in step 2: It can separately extract the kinetics curve of each pixel point of this map; it can extract the fluorescence intensity imaging map at different times or the intensity sum imaging map in all time channels between two specified times; it can also extract the intensity data of the row and column where the specified pixel point is located. The specific operation steps are described as follows:
[0052] Import the fluorescence total intensity imaging map collected in step 2 into the processing program and perform the following operations:
[0053] (1) Extract the fluorescence kinetics of the defective pixel points (the position marked as 2 in the fluorescence intensity map, with coordinates X12Y11 and a gray value of 4119) and the normal region pixel points (the positions marked as 1 and 3, with coordinates X2Y11 and X18Y11 respectively, and gray values of 5842 and 5921), and compare the two. Generally, the kinetic decay of the defective pixel points is faster than that of the defect-free region. As Figure 6 shown;
[0054] (2) Extract the intensity distribution maps after different decay times for comparison: The starting time can be customized from the decay curve, and the end moment of the decay curve is used as the cut-off time. Extract the total number of photons in all time channels within these two moments. The later the extraction time, the larger the size of the defect in the fluorescence imaging map. As Figure 7 shown.
[0055] (3) To more clearly distinguish the defective region, define a concept of contrast, that is, (intrinsic intensity - defective intensity) / intrinsic intensity. The larger the contrast, the more obvious the defect. Generally, the fluorescence intensity contrast between the dislocation defect and the surrounding defect-free region in the full-time imaging map is 20%-30%: Extract the intensity data of the row or column where the defect is located after different decay times, and perform a normalization operation on them (average the data of the first 10 pixels and the last 10 pixels in the row where the defect is located, and divide all the data in this row by this average value, so that the original data can be normalized), and compare the data at each moment after processing. The later the extraction time, the larger the contrast between the defective region and the surrounding region, and the more obvious the defect. As Figure 8 shown.
[0056] The working steps of the above processing program are shown in Figure 5 (b).
[0057] Example:
[0058] A method for amplifying TDs defects in a silicon carbide epitaxial wafer based on time-resolved mode described in this example specifically includes the following steps:
[0059] 1. Self-build a set of fluorescence imaging equipment with time-resolved function. The specific technical roadmap and detailed drawings are shown in Figure 1 and Figure 2 .
[0060] 2. Select a silicon carbide epitaxial wafer with an epitaxial thickness of 13 μm, set the excitation light power of 343 nm (500 KHz) to 0.2 mw, and collect the fluorescence spectrum of the central pixel point of the scanning area. As Figure 3(a). The collected central wavelength is set near 390 nm, and the fluorescence decay kinetics of the central pixel in the scanning area are collected by a single-photon counter, as Figure 3 (b).
[0061] 3. Set the acquisition parameters: range 20 μm x 20 μm, step size 0.25 μm, with a total of 80 x 80 pixels. Start scanning the entire area. The acquisition program will automatically calculate the total number of photons in all time channels under the fluorescence decay kinetics curve. By scanning different pixels, a two-dimensional matrix related to position and fluorescence intensity can be obtained and plotted as a graph, which is the fluorescence intensity imaging map with time-resolved information, as Figure 4 .
[0062] 4. Import the raw data obtained in step 3 into the analysis data analysis program to obtain more time-related information.
[0063] (1) Compare the fluorescence lifetimes of the defective and defect-free regions of silicon carbide by extracting the fluorescence decay kinetics curves at different positions. After comparison, it is found that the lifetime of the defective region is significantly shorter than that of the defect-free region, as Figure 5 .
[0064] (2) Obtain more time information related to TD defects by extracting the intensity distribution maps after different decay times, as Figure 6 . After comparison, it is found that the later the extraction time, the larger the imaging area occupied by the defect, which is related to the energy band structure of the defect and the diffusion properties of the material itself.
[0065] (3) To better describe the phenomenon described in step (2), extract the fluorescence intensity profile related to the position of the curve (at the white dotted line), put the profiles after different decay times together and perform normalized comparison, as Figure 7 . The specific normalization steps are as follows: regard the fluorescence intensity far from the defective region as 1, calculate the average value of the first ten pixels and the last ten pixels, and divide all the data in this row by the calculated average value. After comparison, it is found that the later the extraction time, the greater the contrast between the defective region and the surrounding region.
Claims
1. A device for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology, characterized in that, it includes an excitation light source module and a detection device module arranged in sequence on the optical path, wherein, the excitation light source module includes a femtosecond laser, a beam shrinking component, a third-harmonic generation component, a beam expanding component, and an attenuation sheet component. The laser emitted by the femtosecond laser sequentially passes through a mirror component 1, the beam shrinking component, the third-harmonic generation component, the beam expanding component, a mirror component 2, and the attenuation sheet component and then enters the detection device module; the detection device module includes a microscope system, a motorized displacement stage, a spectrometer system, a photon detector, and an industrial control computer. The light source output by the excitation light source module enters the microscope system, is reflected into the spectrometer system through a preset optical path, and the photon detector, the spectrometer system, and the motorized displacement stage are all connected to the industrial control computer. The silicon carbide epitaxial wafer to be tested is placed on the motorized displacement stage located below the microscope system.
2. The device for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology according to claim 1, characterized in that, a filter and a beam booster are sequentially arranged between the third-harmonic generation component and the beam expanding component.
3. The device for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology according to claim 1, characterized in that, the beam shrinking component is composed of lens groups with focal lengths of f = 400 mm and f = -50 mm.
4. The device for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology according to claim 1, characterized in that, the third-harmonic generation component is composed of a second-harmonic generation BBO crystal and a sum-frequency generation BBO crystal, and calcite is arranged between the second-harmonic generation BBO crystal and the sum-frequency generation BBO crystal.
5. The device for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology according to claim 1, characterized in that, the beam expanding component includes lens groups with focal lengths of f = 30 mm and f = 100 mm.
6. A method for amplifying TDs defects of a silicon carbide epitaxial wafer based on time-resolved technology, characterized in that, it includes the following steps: 1) Control the motorized displacement stage through the industrial control computer to reach a preset pixel position within a preset scanning range and step size, output fluorescence spectrograms at different times at the preset pixel position through the spectrometer system, and obtain the corresponding fluorescence spectrograms at each preset pixel position; Obtain the fluorescence decay kinetic curve diagram corresponding to the preset pixel position at different times through the photon detector; 2) Calculate the sum of the number of photons in all time channels in the fluorescence decay kinetic curve at each preset pixel position to obtain the total fluorescence intensity of this pixel position; 3) Control the motorized displacement stage to move, repeat step 2), traverse and record the total fluorescence intensity of all pixel positions, and obtain a fluorescence intensity imaging diagram with time-resolved information; 4) Extract the fluorescence decay kinetic curve diagrams at different pixel positions in the fluorescence intensity imaging diagram, compare the fluorescence decay conditions in the defect area and the defect-free area, and determine the area and coordinates of the defect. 5) Extract the fluorescence intensity imaging maps after different decay times of the defects, determine the states of the defects after different decay times, and the fluorescence intensity contour maps related to any row and position at different decay times at the defect positions, so as to show the contrast degree between the defect area and the defect-free area around.
7. A method for amplifying TDs defects in a silicon carbide epitaxial wafer based on time-resolved technology according to claim 6, characterized in that, the step 2) includes the following steps: 2.1) Mobilize the spectrometer system to collect fluorescence spectra; 2.2) Mobilize the single-photon counter to collect the fluorescence decay kinetic curve graph at the wavelength where the fluorescence peak value of the sample is located; 2.3) Mobilize the electric displacement stage to collect the fluorescence decay kinetic curve graphs at different pixel positions; 2.4) Sum the fluorescence intensities at each time point in the fluorescence decay kinetic curve graph at each pixel position row by row to obtain the total fluorescence intensity imaging map of the preset scanning area.
8. A method for amplifying TDs defects in a silicon carbide epitaxial wafer based on time-resolved technology according to claim 6, characterized in that, the step 4) is specifically: In the fluorescence intensity imaging map with time-resolved information, take the position with a contrast of 20%-30% as the defect position, and extract the coordinates of the defect position, where the contrast = (fluorescence intensity of the material body - fluorescence intensity of the defect position) / fluorescence intensity of the material body.
9. A method for amplifying TDs defects in a silicon carbide epitaxial wafer based on time-resolved technology according to claim 6, characterized in that, the step 5) includes the following steps: 5.1) Customize the starting time on the decay curve corresponding to each pixel point, and take the time corresponding to the last time channel in the decay curve of each pixel point as the cut-off time. The starting time and cut-off time of all pixel points are kept consistent, and extract the total number of photons in all time channels within the starting time and cut-off time to obtain the fluorescence intensity imaging map of the defect position after different decay times; 5.2) According to the coordinates of the defect position, extract the intensity data of the row or column where the defect is located after different decay times and perform normalization processing; 5.3) Compare all the normalized data to obtain the fluorescence intensity contour map of the defect after different decay times, and obtain the contrast value between the time-related defect and the surrounding area.