Method for quantitatively representing surface layer stress of wafer, and standard method for measuring surface layer stress of wafer
Through resistivity testing and fitting function prediction of Young's modulus, combined with XRD testing, the accuracy of surface stress detection of silicon carbide wafers in the existing technology is solved, and rapid, economical and high-precision surface stress detection is achieved, suitable for Industrial production.
Patent Information
- Application Number
- PCT/CN2023/143076
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-31
- Filing Date
- 2023-12-29
- Publication Date
- 2024-12-05
AI Technical Summary
Existing test methods are difficult to accurately detect surface stress of silicon carbide wafers, especially because the Raman spectroscopy wavelength is long, the body stress calculation is not suitable for surface stress analysis. Neutron scattering method is not common in industrial production, and the optical interference method has low accuracy. Theoretically The calculation is difficult to simulate non-uniform doping and complex surface defects, resulting in low accuracy of detection results.
By testing the resistivity value of the wafer, the Young's modulus is predicted using the fitting function, combining the stress, Young's modulus and strain relationships, the surface stress distribution is calculated, and the XRD test is used to obtain in-plane and interlayer strains. It can be divided into stress directions, providing quantitative characterization methods and standard detection procedures.
It greatly shortens the detection time, saves costs, improves detection accuracy, is suitable for large-scale inspection, can accurately obtain the surface stress distribution of silicon carbide wafers, and optimizes the production and processing process.
Smart Images

Figure CN2023143076_05122024_PF_FP_ABST
Abstract
Description
A method for quantitatively characterizing wafer surface stress and a standard method for detecting wafer surface stress Technical Field
[0001] The present application relates to a method for quantitatively characterizing wafer surface stress and a standard method for detecting wafer surface stress, belonging to the technical field of wafer stress detection. Background Art
[0002] Silicon carbide (SiC) possesses excellent electrical and physical properties and has attracted considerable attention as a substrate material for high-temperature, high-frequency, and high-efficiency power devices. To date, most commercial SiC wafers still contain a certain number of crystal defects, such as micropipes, subgrain boundaries, and dislocations, which prevent them from fully realizing their potential. Furthermore, surface defects are introduced during wafer processing. These defects, along with intrinsic crystal defects, can lead to residual stress on the wafer surface, causing deformation and failure, seriously impacting the material's quality and service life. Therefore, studying the residual stress state on the SiC wafer surface is crucial.
[0003] Currently, there are multiple methods for testing the surface stress of silicon carbide wafers. Micro-Raman spectroscopy uses a biaxial stress model and phonon deformation energy to calculate stress values. However, due to the long wavelength of Raman spectroscopy, the calculated stress value is the bulk stress of the wafer, which cannot effectively analyze the surface stress of the wafer. Neutron scattering is suitable for analyzing stress with thicker samples, such as ingots of a certain thickness. Testing wafer stress requires a small-angle neutron scatterer, which is not common in industrial production. Optical interferometry has relatively low test accuracy and is not suitable for scenarios requiring precise stress resolution. Although stress can be obtained by comparing the unit cell parameters obtained through theoretical calculation simulation with the actual measured unit cell parameters, theoretical calculations cannot effectively simulate non-uniform doping and complex surface defect states, resulting in low test accuracy. Summary of the Invention
[0004] In order to solve the above problems, a method for quantitatively characterizing the surface stress of a wafer and a standard method for detecting the surface stress of a wafer are provided. This method predicts the Young's modulus of different regions of the entire wafer through the resistivity value of the wafer and a fitting function. Compared with directly testing the Young's modulus of the entire wafer, this method saves detection time and can quickly and accurately predict the Young's modulus distribution of the wafer. The relationship between stress, Young's modulus and strain is used to obtain the surface stress distribution of the entire wafer. This method can be widely used for standardized detection of wafer surface stress.
[0005] According to one aspect of the present application, a method for quantitatively characterizing wafer surface stress is provided, comprising the following steps:
[0006] (1) performing a resistivity test on a wafer, and obtaining the Young's modulus of the wafer based on the measured resistivity value and a fitting function;
[0007] (2) Testing the strain of the surface layer of the wafer, and calculating the surface stress of the wafer based on the strain and Young's modulus.
[0008] Directly testing the Young's modulus of the entire wafer takes a long time, while testing the resistivity of the wafer only takes 2-3 minutes. This method uses a constructed fitting function to predict the Young's modulus of different areas of the entire wafer by detecting the resistivity of the wafer, greatly shortening the detection time and saving detection costs. It is suitable for industrial large-scale surface stress detection of wafers. The surface stress refers to the absolute stress of the wafer, that is, the residual stress contained in the wafer compared to a stress-free wafer.
[0009] Optionally, in step (1), the Young's modulus and resistivity at several test points on the wafer are tested, and the Young's modulus and resistivity at each test point are fitted to obtain the fitting function.
[0010] Optionally, the Young's modulus of the wafer is tested using a nanoindentation continuous stiffness method, with the depth of the wafer pressed into the wafer being 0-1000 nm. Since the load force at different depths of the wafer is different, the Young's modulus at the test point can be calculated accordingly.
[0011] Preferably, there are no less than 10 test points.
[0012] Optionally, the fitting function is y = a*x + b*x 2 +c, where a ranges from 100 to 300, b ranges from -100 to 100, and c ranges from -2000 to 2000.
[0013] By matching the resistivity value and Young's modulus at the same test point on the wafer one by one, an accurate functional relationship can be fitted, and the Young's modulus at any point on the wafer can be obtained through the resistivity. This fitting function is obtained by fitting the Young's modulus and resistivity of a limited number of test points, rather than fitting the Young's modulus and resistivity of the entire wafer area. The above fitting method can reduce the amount of testing in the fitting process and save fitting time. Even with limited test points, the reliability of the fitting function can still be improved. The above fitting function improves detection accuracy and provides a basis for quantitative characterization of wafer surface stress.
[0014] Optionally, the surface stress includes in-plane stress and interlayer stress, and the calculation formulas of the in-plane stress and interlayer stress are:
[0015] F a =G*△a,F c =G*△c,
[0016] Among them F a is the in-plane stress, △a is the in-plane strain, F c is the in-plane stress, △c is the interlaminar strain, and G is the Young's modulus.
[0017] Optionally, the in-plane strain and interlayer strain are obtained by XRD testing. The Omega-Rel scanning mode is used in the XRD test. The lattice parameters a and c of the wafer are obtained according to the Bragg formula and the lattice calculation formula of the hexagonal crystal system. The in-plane strain and the interlayer strain are calculated according to the following formula based on the standard lattice parameters a1 and c1 of the silicon carbide crystal:
[0018] ,
[0019] Where △a is the in-plane strain and △c is the interlaminar strain.
[0020] Optionally, the lattice calculation formula of the hexagonal crystal system is:
[0021] ,
[0022] Among them, d is the interplanar spacing, and h, k, and l are all crystal plane indices.
[0023] This application can calculate the in-plane strain and interlayer strain based on the changes in the lattice parameters of the wafer within the plane and between layers, and combine the Young's modulus at the test point to obtain the in-plane stress and interlayer stress at this test point. The above method subdivides the stress of the test point according to direction, and can quantitatively characterize the stress distribution in different directions of the wafer surface, thereby obtaining more accurate surface stress data of the wafer. Based on the stress distribution in different directions, the above stress distribution can be fed back into the wafer production and processing process, thereby optimizing the production and processing of the wafer.
[0024] Optionally, the test crystal planes of the XRD test are (004) plane and (105) plane.
[0025] Optionally, the half-height widths of the (004) and (105) planes are less than 40 arcsec.
[0026] When XRD tests were performed on the two crystal planes mentioned above, the measured diffraction peaks were strong and relatively reasonable in peak width, making it easier to scan the signal. The obtained parameters a and c were also more accurate, improving the detection accuracy. Moreover, when testing the (105) and (004) crystal planes at the same test point, when the equipment sample holder was lifted, the lifting angle was small, the change in the coordinate point was negligible, and the half-height width was not likely to exceed the standard, which can further improve the accuracy of the method. The half-height width of the two crystal planes mentioned above was less than 40 arcsec in the overall detection. When other diffraction crystal planes were used, if the half-height width was greater than 40 arcsec, the detection accuracy would be reduced, thereby reducing the accuracy of the method.
[0027] Optionally, the symmetrical scattering method is used to test the (004) surface, and the skew-symmetrical scattering method is used to test the (105) surface.
[0028] Optionally, the wafer is a silicon carbide wafer, and the silicon carbide wafer includes a semi-insulating type and a doped type.
[0029] Optionally, the wafer has a diameter of 6 inches or more;
[0030] The vertical depth of the surface layer is 3-60 μm.
[0031] Whether for semi-insulating silicon carbide or doped silicon carbide, the Young's modulus at different positions of the wafer is different. Since doped silicon carbide may be non-uniformly doped, the variation of the Young's modulus of doped silicon carbide will be greater and the detection difficulty will be higher. The method of the present application can quantitatively characterize the surface stress of large-size doped silicon carbide wafers through fitting functions, and the detection accuracy is higher than other detection methods.
[0032] According to another aspect of the present application, a standard method for detecting wafer surface stress is provided, comprising the following steps:
[0033] Step 1: Select several test points on the wafer, perform Young's modulus tests on the test points, and perform resistivity tests on the entire wafer. Fit the resistivity and Young's modulus at the test points to obtain a fitting function.
[0034] Step 2: Based on the resistivity of the entire wafer and the fitting function, the Young's modulus size and distribution on the entire wafer are obtained;
[0035] Step 3: Perform XRD test on the surface of the wafer to obtain the strain of the wafer surface, and calculate the surface stress of the wafer based on the strain and the Young's modulus obtained in the second step.
[0036] In this standard method for detecting wafer surface stress, the interlayer strain and in-plane strain calculated by fitting the function and the lattice parameters after XRD testing of the wafer are the same as the method steps for quantitatively characterizing the wafer surface stress. The operator can calculate the in-plane stress and interlayer stress of the wafer according to the above standard method steps and process for detecting wafer surface stress.
[0037] The beneficial effects of this application include but are not limited to:
[0038] 1. According to the method for quantitatively characterizing wafer surface stress in this application, the Young's modulus value at any location on the wafer can be obtained by utilizing the wafer resistivity distribution and fitting function. Compared with testing the Young's modulus of the entire wafer, this method greatly shortens the detection time and saves detection costs.
[0039] 2. According to the method for quantitatively characterizing wafer surface stress described in this application, the wafer is tested using X-ray diffraction, and specific diffraction crystal planes are selected to obtain accurate strain values. Based on the Young's modulus and strain values at any position, the overall surface stress distribution of the wafer is quantitatively calculated, thereby improving test accuracy.
[0040] 3. According to the method for quantitatively characterizing wafer surface stress in this application, by fitting the resistivity values and Young's modulus at the same location on the wafer, an accurate fitting function is obtained. Furthermore, the Young's modulus at any point on the wafer can be obtained from the resistivity, providing a basis for quantitatively characterizing wafer surface stress.
[0041] 4. According to the method for quantitatively characterizing wafer surface stress of the present application, XRD ray diffraction is performed on the (004) and (105) planes of the wafer to obtain the interlayer strain and in-plane strain at different test points of the wafer. This can quantitatively characterize the stress distribution of this test point in different directions, thereby obtaining more accurate wafer surface stress data.
[0042] 5. The method for quantitatively characterizing wafer surface stress according to the present application can measure the stress distribution of large-scale doped silicon carbide wafers and feed the stress distribution back into the wafer production and processing, thereby optimizing wafer production and processing.
[0043] 6. The method for quantitatively characterizing wafer surface stress according to the present application saves testing costs, shortens testing time, and can obtain accurate wafer surface stress data. It is suitable for surface stress testing of silicon carbide wafers of different sizes and thicknesses. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0045] FIG1 is a Young's modulus test process diagram and distribution diagram of a 6-inch doped silicon carbide wafer involved in Example 1 of the present application.
[0046] FIG2 is a resistivity distribution diagram of the wafer involved in Example 1 of the present application.
[0047] FIG3 is a fitting diagram of the Young's modulus and resistivity of the chip involved in Example 1 of the present application.
[0048] FIG4 is an XRD test diagram of the wafer involved in Example 1 of the present application on the (004) plane and the (105) plane.
[0049] FIG5 is an XRD test point and coordinate conversion diagram of the chip involved in Example 1 of the present application. Implementation Method
[0050] The present application is described in detail below with reference to embodiments, but the present application is not limited to these embodiments.
[0051] Unless otherwise specified, the raw materials in the examples of this application were purchased through commercial channels. Example
[0052] This embodiment relates to a method for quantitatively characterizing wafer surface stress, comprising the following steps:
[0053] (1) Obtaining the fitting function and calculating the Young's modulus
[0054] The first step is to test the Young's modulus at two or three test points at the center, above, below, and to the left and right of the polished 6-inch doped silicon carbide wafer. The Young's modulus was measured using the nanoindentation continuous stiffness method, with a depth of 0 to 1000 nm. The test process is shown in Figure 1(a), and the Young's modulus distribution of the final wafer is shown in Figure 1(b).
[0055] The second step is to test the resistivity of the entire wafer, and obtain the resistivity distribution diagram of the wafer as shown in Figure 2. The Young's modulus and resistivity at each test point are fitted to obtain a fitting function, as shown in Figure 3;
[0056] The third step is to obtain the size and distribution of the Young's modulus of the entire wafer based on the resistivity distribution of the entire wafer and the fitting function;
[0057] (2) Test the surface strain of the chip and calculate the surface stress of the chip
[0058] In the first step, XRD was used to measure the 2θ angles at different points on the wafer. The mode used was the Omega-Rel scanning mode. The crystal planes tested were the (004) and (105) planes. The (004) plane was measured using the symmetric scattering method, with the test results shown in Figure 4 (b). The (105) plane was measured using the oblique symmetric scattering method, with the test results shown in Figure 4 (a).
[0059] In the second step, the lattice parameters a and c of the wafer are obtained according to the Bragg formula and the lattice calculation formula of the hexagonal crystal system. Based on the standard lattice parameters a1 and c1 of the wafer, the in-plane strain and the interlayer strain are calculated according to the following formula:
[0060] ,
[0061] Where △a is the in-plane strain and △c is the interlaminar strain;
[0062] The lattice calculation formula of the hexagonal crystal system is:
[0063] ,
[0064] Where d is the interplanar spacing, h, k, and l are all crystal plane indices;
[0065] The Bragg formula is: 2dsinθ=nλ, where d is the interplanar spacing, θ is the diffraction half-angle, n is the diffraction order, and λ is the wavelength of the incident wave.
[0066] In the third step, based on the above-mentioned in-plane strain, interlayer strain and Young's modulus, the in-plane stress and interlayer stress of the surface layer of the 6-inch doped silicon carbide wafer are calculated according to the following formula:
[0067] F a =G*△a,F c =G*△c,
[0068] Among them F a is the in-plane stress, △a is the in-plane strain, F c is the in-plane stress, △c is the interlaminar strain, G is the Young's modulus, and according to the scanning path of the XRD ray in Figure 5 (a), the coordinates are converted into Figure 5 (b) with the center of the chip as the origin, and the test data of the in-plane stress and interlaminar stress of the chip under the above coordinates are obtained. The results are shown in Table 1.
[0069] Table 1
[0070]
[0071] Example 2
[0072] In this example, a quantitative characterization of a ground 8-inch semi-insulating silicon carbide wafer was performed using the same steps as in Example 1 to obtain the in-plane stress and interlayer stress of the surface layer of the 8-inch semi-insulating silicon carbide wafer. The test data are shown in Table 2.
[0073] Table 2
[0074]
[0075] Reliability Verification
[0076] The Young's modulus of the silicon carbide wafer tested by the method of Examples 1 and 2 above was measured comprehensively, and based on the comparison of the actually measured Young's modulus and the Young's modulus predicted in Examples 1 and 2, the prediction error of the Young's modulus was calculated using the following formula: E=[(G-G0) / G0]*100%, where G0 is the Young's modulus at the test point actually measured by the nanoindentation continuous stiffness method, and G is the Young's modulus calculated based on the resistivity and the fitting function. The error at different test points in Example 1 is in the range of 0.4%-0.6%, and the error at different test points in Example 2 is in the range of 0.3%-0.5%.
[0077] The foregoing is merely an embodiment of the present application, and the scope of protection of the present application is not limited by these specific embodiments, but is determined by the claims of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc. made within the technical ideas and principles of the present application should be included in the scope of protection of the present application.
Claims
1. A method for quantitatively characterizing the surface stress of a wafer, characterized in that: The steps include: (1) performing a resistivity test on a wafer, and obtaining a Young's modulus of the wafer according to the measured resistivity value and a fitting function; (2) Testing the strain of the surface layer of the wafer, and calculating the surface stress of the wafer based on the strain and Young's modulus.
2. The method according to claim 1, characterized in that In step (1), the Young's modulus and resistivity at a plurality of test points on the wafer are tested, and the Young's modulus and resistivity at each test point are correspondingly fitted to obtain the fitting function.
3. The method according to claim 2, characterized in that The fitting function is y = a*x + b*x 2 +c, where a ranges from 100 to 300, b ranges from -100 to 100, and c ranges from -2000 to 2000.
4. The method according to claim 1, characterized in that The surface stress includes in-plane stress and interlaminar stress, and the calculation formulas of the in-plane stress and interlaminar stress are: F a =G*△a,F c =G*△c, Among them, F a is the in-plane stress, △a is the in-plane strain, F c is the in-plane stress, △c is the interlaminar strain, and G is the Young's modulus.
5. The method according to claim 4, characterized in that The in-plane strain and interlayer strain are obtained by XRD testing. According to the Bragg formula and the lattice calculation formula of the hexagonal crystal system, the lattice parameter a and parameter c of the wafer are obtained. According to the standard lattice parameters a1 and c1 of the wafer, the in-plane strain and the interlayer strain are calculated according to the following formula: , Where △a is the in-plane strain and △c is the interlaminar strain.
6. The method according to claim 5, characterized in that The lattice calculation formula of the hexagonal crystal system is: , Among them, d is the interplanar spacing, and h, k, and l are all crystal plane indices.
7. The method according to claim 6, characterized in that The test crystal planes of the XRD test are the (004) plane and the (105) plane.
8. The method according to claim 7, characterized in that The half-widths of the (004) and (105) planes are less than 40 arcsec.
9. The method according to any one of claims 1 to 8, characterized in that: The wafer is a silicon carbide wafer, and the silicon carbide wafer includes a semi-insulating type and a doped type.
10. A standard method for detecting surface stress of a wafer, characterized in that: The steps include: Step 1: Select several detection points on the wafer, perform Young's modulus test on the detection points, and perform resistivity test on the entire wafer, and fit the resistivity and Young's modulus at the detection points to obtain a fitting function; Step 2: According to the resistivity of the whole wafer and the fitting function, the size and distribution of Young's modulus on the whole wafer are obtained; Step 3: Perform XRD test on the surface of the wafer to obtain the strain of the surface of the wafer, and calculate the surface stress of the wafer based on the strain and the Young's modulus obtained in the second step.
Citation Information
Patent Citations
Test method of residual stress of silicon single crystal piece
CN102435361A
Residual stress testing method for large-size ground wafer
CN104075830A
Stress detection method and stress detection system for dielectric film layers
CN109119351A
Method for optimizing laser nondestructive characterized Youngs modulus of membrane
CN109521090A
Rapid simulating method of resistance spot welding residual stress and deformation of complicated metal plate component
CN111745276A
