Measuring method of overlay deviation

Abstract

A method for measuring overlay deviation includes: providing several batches of wafers to be tested; obtaining several sample wafers from the batches of wafers to be tested; performing a first overlay deviation measurement on the sample wafers using a first light dose to obtain a first measurement value group; performing a second overlay deviation measurement on the sample wafers using a second light dose to obtain a second measurement value group, wherein the first light dose is higher than the second light dose; obtaining an average deviation value group of the first and second measurement value groups; obtaining a measured wafer from the batches of wafers to be tested; performing a third overlay deviation measurement on all regions of the measured wafer using the second light dose or the first light dose to obtain a third measurement value group of the measured wafer; and calibrating the third measurement value group point-by-point using the average deviation value group to obtain an overlay deviation value group of the measured wafer. Thus, measurement efficiency can be improved while ensuring measurement accuracy.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor manufacturing, and in particular to a method for measuring overlay deviation. Background Art

[0002] Overlay (OVL) refers to the deviation in the alignment of the current layer pattern and the previous layer pattern during the photolithography manufacturing process. Since integrated circuit chips are manufactured by stacking multiple circuit layers, if the current layer and the previous layer are not aligned, the chip will not function properly. Therefore, during the formation of the current layer, it is extremely important to reduce the overlay deviation and ensure that the overlay deviation is within the deviation range. Because traditional overlay measurement technology based on imaging and image recognition has gradually failed to meet the requirements of new process nodes for overlay deviation measurement, diffraction-based overlay measurement technology (DBO, Diffraction-Based Overlay) is gradually becoming an important means of overlay measurement.

[0003] In the prior art, since some layers have low light transmittance, diffraction-based overlay metrology technology uses high-dose light for illumination to collect optical signals with sufficient contrast.

[0004] However, in the above method, high-dose light needs to be obtained by increasing the irradiation time, which will cause the measurement time to increase exponentially, causing great trouble for mass production. Summary of the Invention

[0005] The technical problem solved by the present invention is to provide a method for measuring overlay deviation, which can significantly reduce the measurement time and improve the measurement capacity while ensuring the measurement accuracy.

[0006] In order to solve the above technical problems, the technical solution of the present invention provides a method for measuring overlay deviation, including: providing several batches of wafers to be tested; obtaining several sample wafers from the several batches of wafers to be tested; using a first light dose to perform a first overlay deviation measurement on the sample wafer to obtain a first measurement value group; using a second light dose to perform a second overlay deviation measurement on the sample wafer to obtain a second measurement value group, wherein the first light dose is higher than the second light dose; obtaining an average deviation value group of the first measurement value group and the second measurement value group; obtaining a measured wafer from the several batches of wafers to be tested; using the second light dose or the first light dose to perform a third overlay deviation measurement on all regional points of the measured wafer to obtain a third measurement value group of the measured wafer; using the average deviation value group to calibrate the third measurement value group point-to-point to obtain an overlay deviation value group of the measured wafer.

[0007] Optionally, it also includes: obtaining the second light dose before performing the first overlay deviation measurement; the method for obtaining the second light dose includes: obtaining a test wafer; using multiple different test light doses to perform multiple overlay deviation measurements on all area points of the test wafer, and obtaining a plurality of discrete values ​​of optical signals that are less than or equal to a first preset value, and the discrete value of each optical signal corresponds to each test light dose; obtaining a light dose and optical signal discrete value relationship model based on the multiple test light doses and the corresponding multiple optical signal discrete values; obtaining a critical optical signal discrete value based on the light dose and optical signal discrete value relationship model; obtaining the second light dose through the light dose and optical signal discrete value relationship model and the discrete value of the critical optical signal, the second light dose being the lowest light dose when the discrete value of the optical signal is less than or equal to the discrete value of the critical optical signal.

[0008] Optionally, the multiple test light doses are greater than or equal to 500; the nth test light dose is 500 greater than the n-1th test light dose, and n is a natural number greater than or equal to 2; the discrete value of the first preset optical signal is 4.

[0009] Optionally, the method for obtaining the discrete value of the critical optical signal includes: obtaining the discrete value of the first optical signal corresponding to the first light dose according to the relationship model between the light dose and the discrete value of the optical signal; when the discrete value of the first optical signal is higher than 2, the discrete value of the critical optical signal is 4; when the discrete value of the first optical signal is lower than or equal to 2, the discrete value of the critical optical signal is 2.

[0010] Optionally, the method for obtaining the test wafer includes: obtaining any wafer from the wafers to be tested as the test wafer.

[0011] Optionally, the method for obtaining the third measurement value group includes: using the second light dose or the first light dose to measure all the regional points of the measured wafer to obtain the third measurement value group.

[0012] Optionally, the calibration method includes: adding the third measurement value group point-by-point to the average deviation value group.

[0013] Optionally, the method for obtaining the sample wafer includes: randomly selecting multiple wafers from the batches of wafers to be tested as sample wafers; the number of the sample wafers ranges from 4 to 6 wafers.

[0014] Optionally, the method for obtaining the first measurement value group includes: measuring all the regional points of the sample wafer using the first light dose to obtain the first measurement value group.

[0015] Optionally, the method for obtaining the second measurement value group includes: measuring all the regional points of the sample wafer using the second light dose to obtain the second measurement value group.

[0016] Optionally, the method for obtaining the average deviation value includes: using a high-order regression model to fit the first measurement value group to obtain a first fitting value group; using a high-order regression model to fit the second measurement value group to obtain a second fitting value group; filtering the first fitting value group to obtain a first filtered value group; filtering the second fitting value group to obtain a second filtered value group; compensating the first filtered value group to obtain a first compensated value group; compensating the second filtered value group to obtain a second compensated value group; and obtaining the average deviation value group by point-to-point subtracting the second compensated value group from the first compensated value group.

[0017] Optionally, each batch of wafers to be tested includes multiple wafers; the method for obtaining the sample wafer includes: obtaining the first wafer from each batch of wafers to be tested as the sample wafer.

[0018] Optionally, the sample wafer includes points in a first detection area; the percentage of the number of points in the first detection area to the number of points in all areas is less than or equal to 20%.

[0019] Optionally, the method for obtaining the first measurement value group includes: measuring the first detection area point of the sample wafer using the first light dose to obtain the first measurement value group.

[0020] Optionally, the method for obtaining the second measurement value group includes: measuring the first detection area point of the sample wafer using the second light dose to obtain the second measurement value group.

[0021] Optionally, before obtaining the average deviation value group, it also includes: obtaining a first local average deviation value group; the method for obtaining the first local average deviation value group includes: using a high-order regression model to fit the first measurement value group to obtain a first fitting value group; using a high-order regression model to fit the second measurement value group to obtain a second fitting value group; filtering the first fitting value group to obtain a first filtered value group; filtering the second fitting value group to obtain a second filtered value group; compensating the first filtered value group to obtain a first compensated value group; compensating the second filtered value group to obtain a second compensated value group; obtaining the first local average deviation value group by point-to-point subtracting the second compensated value group from the first compensated value group.

[0022] Optionally, the method for obtaining the average deviation value group includes: performing similarity calculation processing on the first local average deviation value group to obtain the average deviation value group.

[0023] Optionally, the similarity calculation method includes: performing similarity calculation using a Mahalanobis distance model.

[0024] Optionally, after obtaining the average deviation value group, the method further includes: performing error calibration processing on the average deviation value group using a method combining static and dynamic calibration to obtain a calibrated average deviation value group.

[0025] Optionally, the error calibration processing method includes: obtaining a calibration wafer, which is the first wafer in each batch of wafers to be tested; the calibration wafer includes a second detection area point, and the number of points in the second detection area point accounts for a percentage of less than or equal to 20% of the number of points in all areas; using the first light dose to measure the second detection area point on the calibration wafer to obtain the fourth measurement value group; using the second light dose to measure the second detection area point on the calibration wafer to obtain the fifth measurement value group; obtaining a second local average deviation value group of the fourth measurement value group and the fifth measurement value group at the second detection area point; using the Mahalanobis distance model to perform similarity calculation on the second local average deviation value group to obtain a similar average deviation value group; performing error analysis on the similar average deviation value group to obtain an error statistic; when the error statistic is greater than the specification critical value, the calibration average deviation value group is the similar average deviation value group; when the error statistic is less than or equal to the specification critical value, the calibration average deviation value group is the average deviation value group.

[0026] Optionally, the method for obtaining the second local average deviation value group includes: using a high-order regression model to fit the fourth measurement value group to obtain a fourth fitted value group; using a high-order regression model to fit the fifth measurement value group to obtain a fifth fitted value group; filtering the fourth fitted value group to obtain a fourth filtered value group; filtering the fifth fitted value group to obtain a fifth filtered value group; compensating the fourth filtered value group to obtain a fourth compensated value group; compensating the fifth filtered value group to obtain a fifth compensated value group; and obtaining the second local average deviation value group by point-to-point subtracting the fifth compensated value group from the fourth compensated value group.

[0027] Optionally, after obtaining the calibration average deviation value group, it also includes: using the method of adding the third measurement value group point-to-point to the calibration average deviation value group to obtain the calibration overlay deviation value group of the measured wafer.

[0028] Optionally, the specification critical value is 10% of the upper limit of the overlay deviation specification requirement.

[0029] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0030] In a method for measuring overlay deviation provided by the technical solution of the present invention, a second light dose is used to perform a third overlay deviation measurement on a measured wafer to obtain a third measurement value group of the measured wafer. The third measurement value group is calibrated point-to-point using the average deviation value group of the first measurement value group and the second measurement value group to obtain an overlay deviation value group of the measured wafer. Since the light dose and measurement time used in the overlay deviation measurement process have a linear relationship, the larger the light dose, the longer the measurement time. The second light dose is used to perform overlay deviation measurement on the measured wafer, and the second light dose is lower than the first light dose used in the prior art, thereby reducing the actual measurement time. In addition, the first measurement value group is obtained by performing a first overlay deviation measurement on the sample wafer using the first light dose, and the second measurement value group is obtained by performing a second overlay deviation measurement on the sample wafer using the second light dose. Since the point-to-point average deviation value group of the first measurement value group and the second measurement value group is very stable, using the average deviation value of the first measurement value group and the second measurement value group to perform point-to-point calibration on the third measurement value group can ensure that the overlay deviation value group of the measured wafer has a high accuracy. In summary, the overlay deviation measurement method provided by the technical solution of the present invention can improve measurement efficiency and measurement capacity while ensuring overlay measurement accuracy.

[0031] Furthermore, by setting a critical optical signal discreteness value, the conditions for obtaining the second light dose are defined: based on the relationship model between the discreteness of the light dose and the optical signal, the discreteness value of the first optical signal corresponding to the first light dose is obtained; when the discreteness value of the first optical signal is greater than 2, the discreteness value of the critical optical signal is 4; when the discreteness value of the first optical signal is less than or equal to 2, the discreteness value of the critical optical signal is 2. Typically, the discreteness value of the optical signal increases as the light dose decreases, and when the discreteness value of the optical signal is greater than 2, its rate of increase also increases rapidly, which means that the accuracy of overlay deviation measurement decreases rapidly. In general practical applications, the upper limit of the discreteness value of the optical signal is 4, and the discreteness value of the optical signal corresponding to the ideal light dose is less than 2. Therefore, by setting the discreteness value of the critical optical signal, the accuracy of overlay deviation measurement on the measured wafer using the second light dose is ensured to be within a reasonable range. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1It is a graph showing the relationship between the light dose and the measurement time on a certain sample layer;

[0033] Figure 2 1 is a flow chart of a method for measuring overlay deviation according to an embodiment of the present invention;

[0034] Figure 3 yes Figure 2 Flow chart of step S110;

[0035] Figure 4 This is a schematic diagram of a test wafer and detection position according to an embodiment of the present invention;

[0036] Figure 5 It is a relationship diagram between the light dose and the discrete value of the optical signal on a certain sample layer;

[0037] Figure 6 This is a schematic diagram of a sample wafer and a detection position according to an embodiment of the present invention;

[0038] Figure 7 This is a schematic diagram of a measured wafer and detection position according to an embodiment of the present invention;

[0039] Figure 8 is a flow chart of a method for measuring overlay deviation according to another embodiment of the present invention;

[0040] Figure 9 is a schematic diagram of a sample wafer and a detection position according to another embodiment of the present invention;

[0041] Figure 10 This is a schematic diagram of a measured wafer and detection position according to another embodiment of the present invention;

[0042] Figure 11 1 is a flow chart of a method for measuring overlay deviation according to another embodiment of the present invention;

[0043] Figure 12 yes Figure 11 Flow chart of step S360;

[0044] Figure 13 This is a schematic diagram of a calibration wafer and a detection position according to another embodiment of the present invention;

[0045] Figure 14 This is a schematic diagram of a measured wafer and detection position according to another embodiment of the present invention. DETAILED DESCRIPTION

[0046] As described in the background art, high light doses need to be obtained by increasing the irradiation time, which will cause the measurement time to increase exponentially and bring great trouble to mass production. This will be explained in detail below.

[0047] In the measurement method of overlay deviation, since the material transmittance of some layers is low, the diffraction-based overlay measurement technology uses high-dose light for irradiation. In actual measurement, the measurement time under different light doses increases linearly with the increase of light dose. The relationship between light dose and measurement time is y=ax+b, y is the light dose, x is the measurement time, a and b are constants, and a is greater than 0. It should be noted that the light dose refers to the time required for standard light intensity to reach the standard grayscale on the wafer. The grayscale refers to the logarithmic relationship between white and black. For details, please refer to Figure 1 .

[0048] Figure 1 This is a graph showing the relationship between the light dose and the measurement time on a certain sample layer. Figure 1 The relationship between light dose and measurement time is y = 0.0284x + 72.294, where y is the light dose and x is the measurement time. Figure 1 It can be seen that when the light dose increases from 2000 to 4000, the measurement time increases by 45%.

[0049] In the above-mentioned method for measuring overlay deviation, the light dose is increased by increasing the exposure time, which causes the measurement time to increase exponentially. This will cause great trouble to mass production, especially when the measurement module and the lithography machine are integrated together. The use of high light dose measurement will become a bottleneck of the entire process.

[0050] To solve the technical problem, an embodiment of the present invention provides a method for measuring overlay deviation, comprising: performing a third overlay deviation measurement on a measured wafer using a second light dose to obtain a third measurement value group of the measured wafer. Performing a point-to-point calibration on the third measurement value group using the average deviation value group of the first measurement value group and the second measurement value group to obtain an overlay deviation value group of the measured wafer. Since the light dose and measurement time used in the overlay deviation measurement process have a linear relationship, the greater the light dose, the longer the measurement time. The technical solution of the present invention uses the second light dose to measure the overlay deviation of the measured wafer, and the second light dose is lower than the first light dose used in the prior art, thereby reducing the actual measurement time. In addition, the first measurement value group is obtained by performing a first overlay deviation measurement on the sample wafer using the first light dose, and the second measurement value group is obtained by performing a second overlay deviation measurement on the sample wafer using the second light dose. Since the point-to-point average deviation value group of the first measurement value group and the second measurement value group is very stable, the method of using the average deviation value of the first measurement value group and the second measurement value group to perform point-to-point calibration on the third measurement value group does not affect the accuracy of the overlay deviation value group of the measured wafer.

[0051] In order to make the above-mentioned objects, features and beneficial effects of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0052] Figure 2 FIG. 4 is a flow chart of a method for measuring overlay deviation according to an embodiment of the present invention.

[0053] Please refer to Figure 2 , the method for measuring the overlay deviation includes:

[0054] Step S100, providing several batches of wafers to be tested;

[0055] Step S110, obtaining the second light dose;

[0056] Step S120, randomly selecting a plurality of wafers from the batches of wafers to be tested as sample wafers;

[0057] Step S130, measuring all the regional points of the sample wafer using the first light dose to obtain the first measurement value group;

[0058] Step S140, measuring all the regional points of the sample wafer using the second light dose to obtain the second measurement value group;

[0059] Step S150, obtaining an average deviation value group of the first measurement value group and the second measurement value group;

[0060] Step S160, obtaining a measured wafer from the batches of wafers to be measured;

[0061] Step S170, using the second light dose or the first light dose to perform overlay deviation measurement on all the regional points of the measured wafer to obtain the third measurement value group;

[0062] In step S180 , the third measurement value group is calibrated point-to-point using the average deviation value to obtain the overlay deviation value group of the measured wafer.

[0063] The following is a detailed description with reference to the accompanying drawings.

[0064] Step S100 is executed to provide several batches of wafers to be tested.

[0065] It should be noted that the wafer to be tested is used to obtain the subsequent sample wafer, the test wafer and the measured wafer.

[0066] Execute step S110 to obtain the second light dose.

[0067] Figure 3 yes Figure 2Flow chart of step S110 in FIG.

[0068] Please refer to Figure 3 The step of obtaining a second light dose includes:

[0069] Step S111, obtaining any wafer from the wafers to be tested as the test wafer;

[0070] Step S112, performing multiple overlay deviation measurements on all regions of the test wafer using multiple different test light doses to obtain discreteness values ​​of multiple optical signals that are less than or equal to a first preset discreteness value of the optical signal;

[0071] Step S113, obtaining a relationship model between the light dose and the dispersion value of the optical signal according to the plurality of test light doses and the corresponding plurality of dispersion values ​​of the optical signals;

[0072] Step S114, obtaining a discrete value of a critical optical signal according to the relationship model between the light dose and the discrete value of the optical signal;

[0073] Step S115 : obtaining the second light dose through the light dose and optical signal discreteness value relationship model and the critical optical signal discreteness value.

[0074] The following is combined with Figures 4 and 5 Steps S111 to S115 are described in detail.

[0075] Figure 4 Schematic diagram of a test wafer and detection position according to an embodiment of the present invention.

[0076] Please refer to Figure 4 , Figure 4 1 is a schematic diagram of the detection position distribution of the test wafer 100 , where any wafer among the wafers to be tested is the test wafer 100 .

[0077] The test wafer 100 includes all-region point detection positions 101. The all-region point detection positions 101 are detection positions for overlay deviation measurement for obtaining the second light dose.

[0078] It should be noted that the test wafer 100 is a wafer used to obtain the second light dose.

[0079] It should be noted that the light dose refers to the time required for a standard light intensity to reach a standard grayscale on the wafer. The grayscale refers to the number of levels between white and black divided according to a logarithmic relationship.

[0080] Step S112 is executed to perform multiple overlay deviation measurements on the test wafer 100 at all regional point detection positions 101 using multiple different test light doses to obtain discrete values ​​of multiple optical signals that are less than or equal to a first preset discrete value of the optical signal.

[0081] In this embodiment, each of the test light doses is greater than or equal to 500; the nth test light dose is 500 greater than the n-1th test light dose, where n is a natural number greater than or equal to 2; and the discrete value of the first preset optical signal is 4.

[0082] It should be noted that the overlay deviation measurement includes diffraction-based overlay deviation measurement. Diffraction-based overlay deviation measurement refers to a method in the integrated circuit manufacturing process that uses optical diffraction technology to compare the position deviation of overlay marks on different process layers to determine the relative displacement between process layers.

[0083] The diffraction-based overlay deviation measurement method includes: injecting a uniform light beam into the overlay mark of the current layer, diffracting the light beam when it passes through the overlay mark of the current layer, and then reflecting the diffracted light beam after reaching the previous layer. The overlay deviation is calculated by analyzing the reflected diffraction spot. When the current layer and the current layer are not aligned, the diffraction efficiency of the +1st order light and the -1st order light also changes with the overlay deviation. Within a certain range, the overlay deviation and the difference between the diffraction efficiency of the +1st order light and the -1st order light satisfy a linear relationship, that is, I +1 -I -1 | d+ε =k(d+ε), I +1 -I -1 | -d+ε =k(-d+ε), the I +1 is the diffraction efficiency of +1 order light, the I -1 is the diffraction efficiency of -1st-order light, ε is the overlay deviation, and +d and -d are the preset deviations between the two marks in the overlay mark. k is a constant related only to the overlay mark production process. The specified window is much larger than the current overlay deviation control range in wafer fabs. k can be measured experimentally, and the overlay deviation ε can be calculated based on the linear relationship.

[0084] It should be noted that the discrete value of the optical signal is an optical signal indicator in the overlay deviation measurement, and is included in the measurement data returned by the overlay deviation measurement. A large discrete value of the optical signal indicates a large measurement noise, which indicates that the accuracy of the measurement data is low. The process of light dose from high to low is a process in which the discrete value of the optical signal continuously increases. In the overlay deviation measurement process, the upper limit of the discrete value of the optical signal is 4. If the discrete value of the optical signal is higher than 4, the measurement accuracy will not meet the measurement requirements. In this embodiment, the discrete value of the first preset optical signal is set to 4.

[0085] It should be noted that when the test light dose is less than 500, the accuracy of the overlay deviation measurement data will not meet the measurement requirements. Therefore, in this embodiment, each test light dose is greater than or equal to 500.

[0086] Figure 5 It is a relationship diagram between the light dose and the discrete value of the optical signal on a certain sample layer.

[0087] Please refer to Figure 5 , through the plurality of test light doses and the corresponding plurality of discrete values ​​of optical signals, according to y=ax 2 -bx+c obtains the relationship model 110 between the light dose and the discrete value of the optical signal, where y is the discrete value of the optical signal, x is the light dose, and a, b, and c are constants; Figure 5 The relationship model 110 between the light dose and the discrete value of the optical signal is y=6E -8 x 2 -0.0004x+2.3421. A critical optical signal dispersion value 120 is obtained based on the light dose-optical signal dispersion value relationship model 110. A second light dose 150 is obtained based on the light dose-optical signal dispersion value relationship model 110 and the critical optical signal dispersion value 120. The second light dose 150 is the minimum light dose when the optical signal dispersion value is less than or equal to the critical optical signal dispersion value 120.

[0088] The method for obtaining the critical optical signal dispersion value 120 includes: obtaining a first optical signal dispersion value 140 corresponding to the first optical dose 130 according to the light dose-optical signal dispersion value relationship model 110; when the first optical signal dispersion value 140 is greater than 2, the critical optical signal dispersion value 120 is 4; and when the first optical signal dispersion value 140 is less than or equal to 2, the critical optical signal dispersion value 120 is 2. In this embodiment, the first optical dose 130 is 4000, and the corresponding first optical signal dispersion value 140 is less than 2, so the critical optical signal dispersion value 120 is 2.

[0089] It should be noted that the first light dose 130 is the optimal overlay deviation measurement dose value. In the prior art, the first light dose 130 is used to measure the overlay deviation. The first light dose 130 is greater than 2000. Under the optimal overlay deviation measurement condition, the discrete value 140 of the first optical signal corresponding to the first light dose 130 in the relationship model 110 is less than 2. The most ideal measurement condition is that the discrete value 140 of the first optical signal is less than 1. The upper limit of the discrete value 140 of the first optical signal is 4. When the discrete value 140 of the first optical signal is greater than 4, the measurement accuracy will not meet the measurement requirements, and it is not the optimal overlay deviation measurement condition.

[0090] It should be noted that the second light dose 150 is lower than the first light dose 130. In actual products, the measurement time under different light doses increases linearly with the increase of light dose. Figure 1 In overlay deviation measurement, the light dose is increased by increasing the exposure time. Although the first light dose 130 is the optimal dose value for overlay deviation measurement, the measurement time using the first light dose 130 is longer than the measurement time using the second light dose 150. Therefore, using the second light dose 150 for overlay deviation measurement can improve measurement efficiency.

[0091] It should be noted that the theoretical basis for determining the second light dose 150 is the relationship between the light dose and the optical signal's dispersion value. Since the progression of light dose from high to low is a process in which the optical signal's dispersion value continuously increases, when selecting the optimal second light dose 150, in addition to ensuring that the second light dose 150 is as low as possible, the accuracy of the measured data must also be considered. Therefore, a critical optical signal dispersion value 120 is set to ensure the accuracy of the measured data.

[0092] Figure 6 Schematic diagram of a sample wafer and detection position according to an embodiment of the present invention.

[0093] Please refer to Figure 6 , Figure 6 2 is a schematic diagram of the distribution of detection positions of the sample wafer 200 , where a plurality of wafers are randomly selected from the batches of wafers to be tested as the sample wafer 200 .

[0094] The sample wafer 200 includes all-region point detection positions 201. The all-region point detection positions 201 are detection positions for overlay deviation measurement to obtain the average deviation value group. The all-region point detection positions 201 are the same as the all-region point detection positions 101.

[0095] It should be noted that the sample wafer 200 is a wafer used to obtain the average deviation value group.

[0096] It should be noted that, in one embodiment, the number of the sample wafers 200 ranges from 4 to 6 pieces.

[0097] Step S130 is executed to measure all the regional points 201 of the sample wafer 200 using the first light dose 130 to obtain the first measurement value group.

[0098] It should be noted that the first measurement value group is used to obtain the average deviation value group.

[0099] Step S140 is executed to measure all the regional points 201 of the sample wafer 200 using the second light dose 150 to obtain the second measurement value group.

[0100] It should be noted that the second measurement value group is used to obtain the average deviation value group.

[0101] Step S150 is executed to obtain an average deviation value group of the first measurement value group and the second measurement value group.

[0102] In this embodiment, the method for obtaining an average deviation value group of the first measurement value group and the second measurement value group includes: fitting the first measurement value group using a high-order regression model to obtain a first fitted value group; fitting the second measurement value group using a high-order regression model to obtain a second fitted value group; filtering the first fitted value group based on a residual analysis method to filter out values ​​with residuals outside the range of -N times the standard deviation of the residual to N times the standard deviation of the residual, where N ranges from 4 to 6, to obtain a first filtered value group; filtering the second fitted value group based on a residual analysis method to filter out values ​​with residuals outside the range of -N times the standard deviation of the residual to N times the standard deviation of the residual, to obtain a second filtered value group; compensating the first filtered value group using a polar coordinate model to obtain a first compensated value group; compensating the second filtered value group using a polar coordinate model to obtain a second compensated value group; and obtaining an average deviation value group of the first measurement value group and the second measurement value group by point-by-point subtraction of the second compensated value group from the first compensated value group.

[0103] It should be noted that the average deviation value group is used to perform point-to-point calibration on the third measurement value group. Since the point-to-point average deviation value group of the first measurement value group and the second measurement value group is very stable, the method of using the average deviation value of the first measurement value group and the second measurement value group to perform point-to-point calibration on the third measurement value group ensures the accuracy of the overlay deviation value group of the measured wafer.

[0104] Figure 7 It is a schematic diagram of a measured wafer and detection position according to an embodiment of the present invention.

[0105] Please refer to Figure 7 , Figure 7 3 is a schematic diagram of the detection position distribution of the measured wafer 300, and the measured wafer 300 is obtained from the batches of wafers to be measured.

[0106] The measured wafer 300 includes all-region point detection positions 301. The all-region point detection positions 301 are used to obtain detection positions of an overlay deviation value group of the measured wafer 300. The all-region point detection positions 301 are the same as the all-region point detection positions 101.

[0107] Step S170 is executed to perform overlay deviation measurement on all the regional points 301 of the measured wafer 300 using the second light dose 150 or the first light dose 130 to obtain the third measurement value group.

[0108] This embodiment mainly uses the second light dose 150 to measure the overlay deviation.

[0109] It should be noted that, since the second light dose 150 is lower than the first light dose 130, according to the relationship between light dose and measurement time, y=ax+b, y is the light dose, x is the measurement time, a and b are constants, and a is greater than 0. For details, please refer to Figure 1 The measurement time for obtaining the third measurement value group is shorter than the measurement time using the first light dose 130. Therefore, the measurement efficiency is improved by using the second light dose 150 for measurement.

[0110] Step S180 is executed to calibrate the third measurement value group point-to-point using the average deviation value to obtain the overlay deviation value group of the measured wafer 300.

[0111] The calibration method includes: adding the third measurement value group point-by-point to the average deviation value group.

[0112] It should be noted that since the point-to-point average deviation value group of the first measurement value group and the second measurement value group is very stable, the method of using the average deviation value of the first measurement value group and the second measurement value group to perform point-to-point calibration on the third measurement value group ensures the accuracy of the overlay deviation value group of the measured wafer.

[0113] Figure 8 FIG. 4 is a flow chart of a method for measuring overlay deviation according to another embodiment of the present invention.

[0114] Please refer to Figure 8 , the method for measuring the overlay deviation includes:

[0115] Step S200, providing several batches of wafers to be tested;

[0116] Step S210, obtaining the second light dose;

[0117] Step S220, obtaining the first wafer from each batch of wafers to be tested as the sample wafer;

[0118] Step S230 , measuring the first detection area point of the sample wafer using the first light dose to obtain the first measurement value group;

[0119] Step S240 , measuring the first detection area point of the sample wafer using the second light dose to obtain the second measurement value group;

[0120] Step S250, obtaining a first local average deviation value group;

[0121] Step S260, performing similarity calculation on the first local average deviation value group using a Mahalanobis distance model to obtain the average deviation value group;

[0122] Step S270, obtaining a measured wafer from the batches of wafers to be measured;

[0123] Step S280, measuring all the regional points of the measured wafer using the second light dose or the first light dose to obtain the third measurement value group;

[0124] Step S290 , using the average deviation value to perform point-to-point calibration on the third measurement value group to obtain the overlay deviation value group of the measured wafer.

[0125] The following is a detailed description with reference to the accompanying drawings.

[0126] Step S200 is executed to provide several batches of wafers to be tested.

[0127] It should be noted that the wafer to be tested is used to obtain the subsequent sample wafer, the test wafer and the measured wafer.

[0128] Step S210 is executed to obtain the second light dose 150 .

[0129] The method of obtaining the second light dose 150 in step S210 of this embodiment is the same as that in step S110 of the first embodiment.

[0130] Figure 9 Schematic diagram of a sample wafer and detection position according to another embodiment of the present invention.

[0131] Please refer to Figure 9 , Figure 9 4 is a schematic diagram of the distribution of detection positions of the sample wafer 400 . The first wafer in each batch of wafers to be tested is taken as the sample wafer 400 .

[0132] It should be noted that, since the sample wafer 400 is the first wafer in each batch of wafers to be tested, this embodiment performs a calibration for each batch to obtain the overlay deviation of the measured wafer, thereby increasing the accuracy of the calibration overlay deviation value group of the measured wafer 500.

[0133] The sample wafer 400 includes a first inspection region point inspection position 401. The first inspection region point inspection position 401 is an inspection position for performing overlay deviation measurement to obtain a first local average deviation value set.

[0134] It should be noted that since the more points in the first detection area point 401, the more time is required for measurement, in order to improve measurement efficiency, the percentage of the number of points in the first detection area point 401 to the number of points in all area point detection positions 101 is less than or equal to 20%.

[0135] Step S230 is executed to measure the first detection area point 401 of the sample wafer 400 using the first light dose 130 to obtain the first measurement value group.

[0136] It should be noted that the first measurement value group is used to obtain the first local average deviation value group.

[0137] Executing step S240 , measuring the first detection area point 401 of the sample wafer 400 using the second light dose 150 to obtain a second measurement value group;

[0138] It should be noted that the second measurement value group is used to obtain the first local average deviation value group.

[0139] Execute step S250 to obtain a first local average deviation value group;

[0140] In this embodiment, the method for obtaining the first local average deviation value group includes: fitting the first measurement value group using a high-order regression model to obtain a first fitted value group; fitting the second measurement value group using a high-order regression model to obtain a second fitted value group; filtering the first fitted value group based on a residual analysis method to filter out values ​​whose residuals are outside the range of -N times the standard deviation of the residual to N times the standard deviation of the residual, where N ranges from 4 to 6, to obtain a first filtered value group; filtering the second fitted value group based on a residual analysis method to filter out values ​​whose residuals are outside the range of -N times the standard deviation of the residual to N times the standard deviation of the residual, to obtain a second filtered value group; compensating the first filtered value group using a polar coordinate model to obtain a first compensated value group; compensating the second filtered value group using a polar coordinate model to obtain a second compensated value group; and obtaining the first local average deviation value group by point-by-point subtraction of the second compensated value group from the first compensated value group.

[0141] It should be noted that the first local average deviation value group is used to obtain the average deviation value group.

[0142] Executing step S260, performing similarity calculation on the first local average deviation value group using a Mahalanobis distance model to obtain the average deviation value group;

[0143] It should be noted that the Mahalanobis distance model represents the distance between a point and a distribution and is an effective method for calculating the similarity between two unknown sample sets. In this embodiment, the Mahalanobis distance model is used to convert the first local average deviation value group at the first detection area point 401 into the average deviation value group for all area points.

[0144] It should be noted that the average deviation value group is used to perform point-to-point calibration on the third measurement value group. Since the point-to-point average deviation value group of the first measurement value group and the second measurement value group is very stable, the method of using the average deviation value of the first measurement value group and the second measurement value group to perform point-to-point calibration on the third measurement value group ensures the accuracy of the overlay deviation value group of the measured wafer.

[0145] Figure 10 It is a schematic diagram of a measured wafer and detection position according to another embodiment of the present invention.

[0146] Please refer to Figure 10 , Figure 10 Schematic diagram of the distribution of detection positions of the measured wafer 500 , where the measured wafer 500 is obtained from the batches of wafers to be measured.

[0147] The measured wafer 500 includes all-region point detection positions 501. The all-region point detection positions 501 are used to obtain detection positions of an overlay deviation value group of the measured wafer 500. The all-region point detection positions 501 are the same as the all-region point detection positions 101.

[0148] It should be noted that, in this embodiment, the measured wafer 500 needs to be from the same batch as the sample wafer 400 .

[0149] Executing step S280 , measuring all the regional points 501 of the measured wafer 500 using the second light dose 150 or the first light dose 130 to obtain the third measurement value group;

[0150] This embodiment mainly uses the second light dose 150 to measure the overlay deviation.

[0151] It should be noted that, since the second light dose 150 is lower than the first light dose 130, according to the relationship between light dose and measurement time, y=ax+b, y is the light dose, x is the measurement time, a and b are constants, and a is greater than 0. For details, please refer to Figure 1 The measurement time for obtaining the third measurement value group is shorter than the measurement time using the first light dose 130 in the prior art. Therefore, using the second light dose 150 for measurement improves measurement efficiency.

[0152] Executing step S290, using the average deviation value to perform point-to-point calibration on the third measurement value group to obtain the overlay deviation value group of the measured wafer;

[0153] The calibration method includes: adding the third measurement value group point-by-point to the average deviation value group.

[0154] It should be noted that, since the average deviation value group is very stable, the method of performing point-to-point calibration on the third measurement value group using the average deviation value of the first measurement value group and the second measurement value group ensures the accuracy of the overlay deviation value group of the measured wafer 500.

[0155] Figure 11 FIG. 4 is a flow chart of a method for measuring overlay deviation according to another embodiment of the present invention.

[0156] Please refer to Figure 11 , the method for measuring the overlay deviation includes:

[0157] Steps S300 to S350 are the same as steps S100 to S150 of an embodiment of the present invention;

[0158] Step S360, performing error calibration on the average deviation value group using a method combining static and dynamic calibration to obtain a calibration overlay deviation value group of the measured wafer;

[0159] Step S370, obtaining a measured wafer from the batches of wafers to be measured;

[0160] Step S380, using the second light dose or the first light dose to perform overlay deviation measurement on all the regional points of the measured wafer to obtain the third measurement value group;

[0161] Step S390 , using the method of adding the calibration average deviation value group point-by-point to the third measurement value group, to obtain the calibration overlay deviation value group of the measured wafer.

[0162] The following is a detailed description with reference to the accompanying drawings.

[0163] Steps S300 to S350 are executed. Steps S300 to S350 are the same as steps S100 to S150 in an embodiment of the present invention.

[0164] Execute step S360 to perform error calibration on the average deviation value group using a method combining static and dynamic calibration to obtain a calibrated average deviation value group.

[0165] Figure 12 yes Figure 11 Flow chart of step S360 in FIG.

[0166] In this embodiment, the steps of obtaining the calibration mean deviation value group are as follows: Figure 12 Shown, including:

[0167] Step S361, obtaining the first wafer from each batch of wafers to be tested as a calibration wafer;

[0168] Step S362 , measuring the second detection area points on the calibration wafer using the first light dose to obtain the fourth measurement value group;

[0169] Step S363, measuring the second detection area points on the calibration wafer using the second light dose to obtain the fifth measurement value group;

[0170] Step S364, obtaining a second local average deviation value group of the fourth measurement value group and the fifth measurement value group at points in the second detection area;

[0171] Step S365: performing similarity calculation on the second local average deviation value group using a Mahalanobis distance model to obtain a similar average deviation value group;

[0172] Step S366, performing error analysis on the similar average deviation value group to obtain error statistics;

[0173] Step S367, determining whether the error statistic value is less than or equal to the specification critical value;

[0174] Step S368, when the condition of step S367 is met, the calibration average deviation value group is the average deviation value group;

[0175] Step S369: When the condition of step S367 is not met, the calibration mean deviation value group is the similarity mean deviation value group.

[0176] The following is combined with Figure 13 Steps S361 to S369 are described in detail.

[0177] Figure 13 It is a schematic diagram of calibrating wafers and detecting positions according to another embodiment of the present invention.

[0178] Please refer to Figure 13 , Figure 13 Schematic diagram of the distribution of detection positions of the calibration wafer 600. The first wafer in each batch of wafers to be tested is taken as the calibration wafer 600.

[0179] It should be noted that, since the calibration wafer 600 is the first wafer in each batch of wafers to be tested, this embodiment performs a calibration for each batch to obtain the overlay deviation of the measured wafer, thereby increasing the accuracy of the calibration overlay deviation value group of the measured wafer 700.

[0180] The calibration wafer 600 includes a second inspection area point inspection position 601. The second inspection area point inspection position 601 is an inspection position for performing overlay deviation measurement to obtain a second local average deviation value set.

[0181] It should be noted that since the more points in the second detection area 601, the more time is required for measurement, in order to improve measurement efficiency, the percentage of the number of points in the second detection area 601 to the number of points in all area point detection positions 101 is less than or equal to 20%.

[0182] Executing step S362 , measuring the second detection area point 601 on the calibration wafer 600 using the first light dose 130 to obtain the fourth measurement value group;

[0183] It should be noted that the fourth measurement value group is used to obtain the second local average deviation value group.

[0184] Executing step S363 , measuring the second detection area point 601 on the calibration wafer 600 using the second light dose 150 to obtain the fifth measurement value group;

[0185] It should be noted that the fifth measurement value group is used to obtain the second local average deviation value group.

[0186] Executing step S364 to obtain a second local average deviation value group of the fourth measurement value group and the fifth measurement value group at the second detection area point 601;

[0187] In this embodiment, the method for obtaining the second local average deviation value group includes: fitting the fourth measurement value group using a high-order regression model to obtain a fourth fitted value group; fitting the fifth measurement value group using a high-order regression model to obtain a fifth fitted value group; filtering the fourth fitted value group based on a residual analysis method to filter out values ​​whose residuals are outside the range of -N times the standard deviation of the residual to N times the standard deviation of the residual, where N is in a range of 4 to 6, to obtain a fourth filtered value group; filtering the fifth fitted value group based on a residual analysis method to filter out values ​​whose residuals are outside the range of -N times the standard deviation of the residual to N times the standard deviation of the residual, to obtain a fifth filtered value group; compensating the fourth filtered value group using a polar coordinate model to obtain a fourth compensated value group; compensating the fifth filtered value group using a polar coordinate model to obtain a fifth compensated value group; and obtaining the second local average deviation value group by point-by-point subtraction of the fifth compensated value group from the fourth compensated value group.

[0188] It should be noted that the second local average deviation value group is used to obtain the calibration average deviation value group.

[0189] Execute step S365 to perform similarity calculation on the second local average deviation value group using the Mahalanobis distance model to obtain a similar average deviation value group.

[0190] It should be noted that the Mahalanobis distance model represents the distance between a point and a distribution and is an effective method for calculating the similarity between two unknown sample sets. In this embodiment, the Mahalanobis distance model is used to convert the second local average deviation value group at the second detection area point 601 into the similar average deviation value group at all area points.

[0191] Execute step S366 to perform error analysis on the similar average deviation value group to obtain error statistics.

[0192] The error statistic is the average of the second local average deviation value group plus three times the standard deviation of the second local average deviation value group.

[0193] It should be noted that since the similarity average deviation value group is obtained by calculating the similarity of the second local average deviation value using the Mahalanobis distance model, a certain error will occur. When the error statistic is less than or equal to the specification threshold, it indicates that the error of the similarity average deviation value group is too large.

[0194] The specification threshold is 10% of the upper limit of the overlay deviation specification requirement.

[0195] Execute step S367 to determine whether the error statistic value is less than or equal to the specification critical value.

[0196] The specification threshold is 10% of the upper limit of the overlay deviation specification requirement.

[0197] Execute step S368. When the condition of step S367 is met, the calibration average deviation value group is the average deviation value group.

[0198] It should be noted that when the error statistic is less than or equal to the specification critical value, it means that the error of the similar average deviation value group is too large, and only the previously obtained average deviation value group can be used as the calibration average deviation value group.

[0199] Execute step S369. When the condition of step S367 is not met, the calibration mean deviation value group is the similar mean deviation value group.

[0200] It should be noted that when the error statistic value is greater than the specification critical value, it means that the error of the similar average deviation value group meets the requirements, and the similar average deviation value group is used as the calibration average deviation value group.

[0201] Figure 14 This is a schematic diagram of a measured wafer and detection position according to another embodiment of the present invention.

[0202] Please refer to Figure 14 , Figure 14 Schematic diagram of the detection position distribution of the measured wafer 700, where the measured wafer 700 is obtained from the batches of wafers to be measured.

[0203] The measured wafer 700 includes all-region point detection positions 701. The all-region point detection positions 701 are used to obtain detection positions of an overlay deviation value group of the measured wafer 700. The all-region point detection positions 701 are the same as the all-region point detection positions 101.

[0204] It should be noted that the all-region point detection positions 701 are the same as the all-region point detection positions 101 .

[0205] It should be noted that, in this embodiment, when the condition of step S367 is not met, the measurement wafer 700 needs to be from the same batch as the calibration wafer 600 .

[0206] Step S380 is executed to perform overlay deviation measurement on all the regional points 701 of the measured wafer 700 using the second light dose 150 or the first light dose 130 to obtain the third measurement value group.

[0207] This embodiment mainly uses the second light dose 150 to measure the overlay deviation.

[0208] It should be noted that, since the second light dose 150 is lower than the first light dose 130, according to the relationship between light dose and measurement time, y=ax+b, y is the light dose, x is the measurement time, a and b are constants, and a is greater than 0. For details, please refer to Figure 1 The measurement time for obtaining the third measurement value group is shorter than the measurement time using the first light dose 130. Therefore, the measurement efficiency is improved by using the second light dose 150 for measurement.

[0209] Step S390 is executed to obtain the calibration overlay deviation value group of the measured wafer 700 by adding the calibration average deviation value group point-by-point to the third measurement value group.

[0210] It should be noted that, since the calibration average deviation value group is very stable, the method of adding the third measurement value group point-to-point to the calibration average deviation value group ensures the accuracy of the measured overlay deviation value group of the wafer 700.

[0211] It should be noted that the calibration mean deviation value group is obtained by judging the error statistical value, which avoids the problem of excessive error in the similarity mean deviation value group obtained by similarity calculation, and increases the accuracy of the calibration overlay deviation value group of the measured wafer 700.

[0212] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be based on the scope defined by the claims.

Claims

1. A method for measuring overlay deviation, characterized in that: include: Provide several batches of wafers to be tested; Obtaining a plurality of sample wafers from the plurality of batches of wafers to be tested; Performing a first overlay deviation measurement on the sample wafer using a first light dose to obtain a first measurement value group; performing a second overlay deviation measurement on the sample wafer using a second light dose to obtain a second measurement value group, wherein the first light dose is higher than the second light dose; Obtaining an average deviation value group of the first measurement value group and the second measurement value group; Obtaining a measured wafer from the batches of wafers to be measured; Performing a third overlay deviation measurement on all regional points of the measured wafer using the second light dose to obtain a third measurement value group of the measured wafer; The third measurement value group is calibrated point-to-point using the average deviation value group to obtain the overlay deviation value group of the measured wafer.

2. The method for measuring overlay deviation according to claim 1, wherein: Also includes: Before performing the first overlay deviation measurement, obtaining the second light dose; The method for obtaining the second light dose includes: obtaining a test wafer; A plurality of different test light doses are used to perform multiple overlay deviation measurements on all area points of the test wafer, respectively, to obtain discrete values ​​of a plurality of optical signals that are less than or equal to the discrete value of a first preset optical signal, and the discrete value of each optical signal corresponds to each test light dose; a light dose and optical signal discrete value relationship model is obtained based on the plurality of test light doses and the corresponding plurality of optical signal discrete values; a critical optical signal discrete value is obtained based on the light dose and optical signal discrete value relationship model; the second light dose is obtained through the light dose and optical signal discrete value relationship model and the discrete value of the critical optical signal, and the second light dose is obtained, which is the minimum light dose when the discrete value of the optical signal is less than or equal to the discrete value of the critical optical signal.

3. The method for measuring overlay deviation according to claim 2, wherein: The multiple test light doses are greater than or equal to 500; the nth test light dose is 500 greater than the n-1th test light dose, where n is a natural number greater than or equal to 2; The discreteness value of the first preset optical signal is 4.

4. The method for measuring overlay deviation according to claim 2, wherein: The method for obtaining the discrete value of the critical optical signal includes: obtaining the discrete value of the first optical signal corresponding to the first light dose according to the discrete value relationship model between the light dose and the optical signal; when the discrete value of the first optical signal is higher than 2, the discrete value of the critical optical signal is 4; when the discrete value of the first optical signal is lower than or equal to 2, the discrete value of the critical optical signal is 2.

5. The method for measuring overlay deviation according to claim 2, wherein: The method for obtaining the test wafer includes: obtaining any wafer from the wafers to be tested as the test wafer.

6. The method for measuring overlay deviation according to claim 1, wherein: The calibration method includes: adding the third measurement value group point-by-point to the average deviation value group.

7. The method for measuring overlay deviation according to claim 1, wherein: The method for obtaining the sample wafers includes: randomly selecting multiple wafers from the batches of wafers to be tested as sample wafers; the number of the sample wafers ranges from 4 to 6.

8. The method for measuring overlay deviation according to claim 7, wherein: The method for obtaining the first measurement value group includes: measuring all the regional points of the sample wafer using the first light dose to obtain the first measurement value group.

9. The method for measuring overlay deviation according to claim 7, wherein: The method for obtaining the second measurement value group includes: measuring all the regional points of the sample wafer using the second light dose to obtain the second measurement value group.

10. The method for measuring overlay deviation according to claim 7, wherein: The method for obtaining the average deviation value includes: using a high-order regression model to fit the first measurement value group to obtain a first fitting value group; using a high-order regression model to fit the second measurement value group to obtain a second fitting value group; filtering the first fitting value group to obtain a first filtered value group; filtering the second fitting value group to obtain a second filtered value group; compensating the first filtered value group to obtain a first compensated value group; compensating the second filtered value group to obtain a second compensated value group; and obtaining the average deviation value group by point-by-point subtraction of the second compensated value group from the first compensated value group.

11. The method for measuring overlay deviation according to claim 1, wherein: Each batch of wafers to be tested includes multiple wafers; the method for obtaining the sample wafer includes: obtaining the first wafer from each batch of wafers to be tested as the sample wafer.

12. The method for measuring overlay deviation according to claim 11, wherein: The sample wafer includes points in a first detection area; the percentage of the number of points in the first detection area to the number of points in all areas is less than or equal to 20%.

13. The method for measuring overlay deviation according to claim 12, wherein: The method for obtaining the first measurement value group includes: measuring the first detection area point of the sample wafer using the first light dose to obtain the first measurement value group.

14. The method for measuring overlay deviation according to claim 12, wherein: The method for obtaining the second measurement value group includes: measuring the first detection area point of the sample wafer using the second light dose to obtain the second measurement value group.

15. The method for measuring overlay deviation according to claim 12, wherein: Before obtaining the average deviation value group, it also includes: obtaining a first local average deviation value group; the method for obtaining the first local average deviation value group includes: using a high-order regression model to fit the first measurement value group to obtain a first fitting value group; using a high-order regression model to fit the second measurement value group to obtain a second fitting value group; filtering the first fitting value group to obtain a first filtered value group; filtering the second fitting value group to obtain a second filtered value group; compensating the first filtered value group to obtain a first compensated value group; compensating the second filtered value group to obtain a second compensated value group; obtaining the first local average deviation value group by point-to-point subtracting the second compensated value group from the first compensated value group.

16. The method for measuring overlay deviation according to claim 15, wherein: The method for obtaining the average deviation value group includes: performing similarity calculation processing on the first local average deviation value group to obtain the average deviation value group.

17. The method for measuring overlay deviation according to claim 16, wherein: The similarity calculation method includes: performing similarity calculation using a Mahalanobis distance model.

18. The method for measuring overlay deviation according to claim 7, wherein: Also includes: After obtaining the average deviation value group, the method further includes: performing error calibration processing on the average deviation value group by adopting a method combining static and dynamic calibration to obtain a calibrated average deviation value group.

19. The method for measuring overlay deviation according to claim 18, wherein: The error calibration method includes: obtaining a calibration wafer, wherein the calibration wafer is the first wafer in each batch of wafers to be tested; the calibration wafer includes a second detection area point, and the percentage of the number of points in the second detection area points to the number of points in all areas is less than or equal to 20%; using the first light dose to measure the second detection area points on the calibration wafer to obtain a fourth measurement value group; using the second light dose to measure the second detection area points on the calibration wafer to obtain a fifth measurement value group; obtaining a second local average deviation value group of the fourth measurement value group and the fifth measurement value group at the second detection area points; using a Mahalanobis distance model to perform similarity calculation on the second local average deviation value group to obtain a similar average deviation value group; performing error analysis on the similar average deviation value group to obtain an error statistic; when the error statistic is greater than a specification critical value, the calibration average deviation value group is the similar average deviation value group; when the error statistic is less than or equal to the specification critical value, the calibration average deviation value group is the average deviation value group.

20. The method for measuring overlay deviation according to claim 19, wherein: The method for obtaining the second local average deviation value group includes: using a high-order regression model to fit the fourth measurement value group to obtain a fourth fitted value group; using a high-order regression model to fit the fifth measurement value group to obtain a fifth fitted value group; filtering the fourth fitted value group to obtain a fourth filtered value group; filtering the fifth fitted value group to obtain a fifth filtered value group; compensating the fourth filtered value group to obtain a fourth compensated value group; compensating the fifth filtered value group to obtain a fifth compensated value group; and obtaining the second local average deviation value group by point-by-point subtracting the fifth compensated value group from the fourth compensated value group.

21. The method for measuring overlay deviation according to claim 19, wherein: Also includes: After obtaining the calibration average deviation value group, the method further includes: obtaining the calibration overlay deviation value group of the measured wafer by adding the third measurement value group point-to-point to the calibration average deviation value group.

22. The method for measuring overlay deviation according to claim 19, wherein: The specification critical value is 10% of the upper limit of the overlay deviation specification requirement.

Images