Method of measuring overlay error
By using light sources of different wavelengths to obtain measurement values and deviation value sets on the first batch of wafers, and performing data fitting and compensation, the shortcomings of existing overlay measurement methods in terms of accuracy and efficiency are solved, and efficient and high-precision overlay deviation measurement is achieved.
Patent Information
- Application Number
- CN202011424765.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-12-08
- Publication Date
- 2025-12-19
- Estimated Expiration
- 2040-12-08
AI Technical Summary
Existing optical diffraction-based overlay measurement methods have shortcomings in both measurement accuracy and efficiency, especially in the measurement of second batches of wafers, where it is difficult to improve both simultaneously.
The first batch of wafers is measured using a first wavelength light source and a second wavelength light source to obtain a set of measured values and a set of deviation values. These values are then added together to compensate for the measurement of the second batch of wafers. Combined with data fitting and outlier handling, the measurement accuracy and efficiency are optimized.
This improved the measurement accuracy and efficiency of the second batch of wafers, ensured stable measurement of overlay deviations, and enhanced production quality.
Smart Images

Figure CN114609090B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of semiconductor manufacturing, and particularly relates to a method for measuring overlay deviation. BACKGROUND
[0002] Overlay deviation (OVL) refers to the deviation of the position of a layer pattern and a previous layer pattern in a photolithography manufacturing process. Since an integrated circuit chip is manufactured by adding multiple circuit layers, if the current layer and the previous layer are not aligned, the chip will not work properly. Therefore, in the process of forming the current layer, it is extremely important to reduce the overlay deviation and ensure that the overlay deviation is within a deviation range.
[0003] In the prior art, a diffraction-based overlay (DBO) is usually used instead of a traditional image-based overlay (IBO). The diffraction-based overlay uses periodic structures as overlay marks, and the marks are respectively located on a reference layer and a current photoresist layer of a wafer. If the two layers of marks are completely aligned, then the +1 and -1 order diffraction light intensities under illumination light should be completely equal. If the overlay deviation is not zero, then the +1 and -1 order light intensities will have a difference. Through calculation, the relationship between the overlay deviation and the difference between the +1 and -1 order light intensities can be quantitatively obtained. The intensity of the reflection spectrum is a function of the wavelength λ of the illumination light and the grating position X. When the overlay deviation exists, the difference in the first order diffraction light intensity and the overlay deviation value have a good linear relationship. By obtaining the light intensity difference, the measurement device can obtain the overlay condition between different process layers.
[0004] However, the diffraction-based overlay in the prior art still has many problems. SUMMARY
[0005] The technical problem solved by the present application is to provide a method for measuring overlay deviation, which can effectively improve the measurement accuracy of the second batch of wafers and improve the measurement efficiency.
[0006] To solve the above problems, the application provides a method for measuring overlay deviation, comprising: providing a first batch of wafers, the first batch of wafers comprising a plurality of first measurement points; measuring the first batch of wafers by using a first wavelength light source to obtain a first measurement value group, the first measurement value group comprising a plurality of first measurement values corresponding to the first measurement points; measuring the first batch of wafers by using a second wavelength light source to obtain a second measurement value group, the second measurement value group comprising a plurality of second measurement values corresponding to the first measurement points; obtaining a deviation value group between the second measurement value group and the first measurement value group, the deviation value group comprising a plurality of deviation values corresponding to the first measurement points; providing a second batch of wafers, the second batch of wafers comprising a plurality of second measurement points corresponding to the first measurement points; measuring the second batch of wafers by using the first wavelength light source to obtain a third measurement value group, the third measurement value group comprising a plurality of third measurement values corresponding to the second measurement points; adding the third measurement value group and the deviation value group to obtain a fourth measurement value group, the fourth measurement value group comprising a plurality of fourth measurement values corresponding to the second measurement points.
[0007] Optionally, the first wavelength light source comprises a single-wavelength light source or a dual-wavelength light source.
[0008] Optionally, the second wavelength light source comprises a multi-wavelength light source, the number of wavelengths in the multi-wavelength light source is n, and n is greater than or equal to 3.
[0009] Optionally, the number of wavelengths n in the multi-wavelength light source is 4 or 5.
[0010] Optionally, the method for obtaining the multi-wavelength light source comprises: obtaining a plurality of first-level wavelengths in a wavelength range of a preset light wave, the first-level wavelengths being wavelengths at wave peaks and wave troughs of the preset light wave.
[0011] Optionally, the method for obtaining the multi-wavelength light source further comprises: when the number of wavelengths required by the multi-wavelength light source is greater than the number of the first-level wavelengths, obtaining a plurality of second-level wavelengths in the wavelength range of the preset light wave, the second-level wavelengths being wavelengths at ±1 / 2 amplitudes of the preset light wave.
[0012] Optionally, the method for obtaining the multi-wavelength light source further comprises: when the number of wavelengths required by the multi-wavelength light source is greater than the sum of the number of the first-level wavelengths and the number of the second-level wavelengths, obtaining a plurality of third-level wavelengths in the wavelength range of the preset light wave, the third-level wavelengths being wavelengths at ±3 / 4 amplitudes of the preset light wave.
[0013] Optionally, the method for obtaining the first measurement value set comprises: measuring the first batch of wafers by using a first wavelength light source to obtain a first measurement value set, wherein the first measurement value set comprises a plurality of initial first measurement values corresponding to the first measurement points; removing a plurality of first abnormal measurement values in the first measurement value set to obtain an initial first measurement value set; providing a data fitting model; performing data fitting operation on the initial first measurement value set according to the data fitting model to obtain a first fitting value corresponding to the first abnormal measurement value, and composing the first measurement value set by using the first fitting value and the initial first measurement value set.
[0014] Optionally, the method for obtaining the first fitting value comprises: obtaining a first measurement point corresponding to the first abnormal measurement value as a first abnormal point; obtaining a first measurement point adjacent to the first abnormal point as a first fitting reference point; obtaining an initial first measurement value corresponding to the first fitting reference point as a first fitting reference value; and performing the data fitting operation on a plurality of first fitting reference values to obtain the first fitting value.
[0015] Optionally, the method for obtaining the first abnormal measurement value comprises: obtaining a first standard deviation sigma1 of the first measurement value set; obtaining a first value reference range, wherein the first value reference range is 4 times the first standard deviation sigma1 to 6 times the first standard deviation sigma1; and regarding the initial first measurement value outside the first value reference range as the first abnormal measurement value.
[0016] Optionally, the method for obtaining the second measurement value set comprises: measuring the first batch of wafers by using a second wavelength light source to obtain a second measurement value set, wherein the second measurement value set comprises a plurality of initial second measurement values corresponding to the first measurement points; removing a plurality of second abnormal measurement values in the second measurement value set to obtain an initial second measurement value set; providing a data fitting model; performing data fitting operation on the initial second measurement value set according to the data fitting model to obtain a second fitting value corresponding to the second abnormal measurement value, and composing the second measurement value set by using the second fitting value and the initial second measurement value set.
[0017] Optionally, the method for obtaining the second fitting value comprises: obtaining a first measurement point corresponding to the second abnormal measurement value as a second abnormal point; obtaining a first measurement point adjacent to the second abnormal point as a second fitting reference point; obtaining an initial second measurement value corresponding to the second fitting reference point as a second fitting reference value; and performing the data fitting operation on a plurality of second fitting reference values to obtain the second fitting value.
[0018] Optionally, the method for obtaining the second abnormal measurement value comprises: obtaining a second standard deviation sigma1 of the second measurement value set; obtaining a second value reference range, the second value reference range being 4 times the second standard deviation sigma2~6 times the second standard deviation sigma2; and the initial second measurement value outside the second value reference range being the second abnormal measurement value.
[0019] Optionally, the method for obtaining the deviation value set between the second measurement value set and the first measurement value set comprises: subtracting the second measurement value from the corresponding first measurement value to obtain a plurality of deviation values, and the deviation value set being composed of the plurality of deviation values.
[0020] Optionally, the method for obtaining the second measurement value set comprises: measuring the first batch of wafers by using a second wavelength light source to obtain a measurement value set, the measurement value set comprising a plurality of initial second measurement values; and performing operation processing on the measurement value set to obtain an ideal measurement value corresponding to each first measurement point, and taking the ideal measurement value as the second measurement value, and the second measurement value set being composed of a plurality of second measurement values.
[0021] Optionally, the method for obtaining the measurement value set comprises: obtaining the measurement value set by using N multi-wavelength light sources with different wavelength numbers to measure P first measurement points respectively, the wavelength number of each multi-wavelength light source being i, and 1≤i≤N, obtaining N initial second measurement values of each first measurement point, and P*N initial second measurement values of P first measurement points being obtained in total, and the measurement value set being composed of P*N initial second measurement values.
[0022] Optionally, the method for obtaining an ideal measurement value corresponding to each first measurement point comprises: performing linear regression on P*N initial second measurement values, obtaining a linear regression value corresponding to each first measurement point on the linear regression, and taking the linear regression value as the ideal measurement value.
[0023] Optionally, after obtaining the ideal measurement value of each first measurement point, the method further comprises: obtaining the optimal wavelength number of each first measurement point during measurement according to the ideal measurement value.
[0024] Optionally, the method for obtaining the optimal wavelength number of each first measurement point during measurement comprises: obtaining an optimal measurement value in N initial second measurement values of each first measurement point, and taking the wavelength number corresponding to the optimal measurement value as the optimal wavelength number.
[0025] Optionally, the method for obtaining the optimal measurement value comprises: taking the initial second measurement value with the smallest difference from the ideal measurement value as the optimal measurement value.
[0026] Optionally, after the fourth measurement value is acquired, the method further comprises: detecting the deviation value to obtain a detection result; and determining whether to re-measure the overlay deviation according to the detection result.
[0027] Optionally, the method of determining whether to re-measure the overlay deviation according to the detection result comprises: providing a detection threshold; comparing the deviation value with the detection threshold; and when the deviation threshold is greater than the detection threshold, re-measuring the overlay deviation is needed.
[0028] Optionally, when the overlay deviation is re-measured, the number of wavelengths in the multi-wavelength light source is the optimal number of wavelengths.
[0029] Compared with the prior art, the technical scheme of the present application has the following advantages:
[0030] In the forming method of the technical scheme of the present application, a first batch of wafers is provided, and the first batch of wafers includes a plurality of first measurement points; a first wavelength light source is used to measure the first batch of wafers to obtain a first measurement value group, and the first measurement value group includes a plurality of first measurement values corresponding to the first measurement points; and a second wavelength light source is used to measure the first batch of wafers to obtain a second measurement value group, and the second measurement value group includes a plurality of second measurement values corresponding to the first measurement points. Since the first wavelength light source has high measurement efficiency but low measurement accuracy, and the second wavelength light source has high measurement accuracy but low measurement efficiency, a deviation value group between the second measurement value group and the first measurement value group is obtained, the deviation value group includes a plurality of deviation values corresponding to the first measurement points, and the obtained deviation value group is used for additive compensation when a second batch of wafers is measured by the first wavelength light source in the future, so that the second batch of wafers has high measurement efficiency while the measurement accuracy is effectively improved.
[0031] Further, the method for obtaining the second measurement value set comprises: measuring the first batch of wafers by using a second wavelength light source to obtain a measurement value set, wherein the measurement value set comprises a plurality of initial second measurement values; performing operation processing on the measurement value set to obtain an ideal measurement value corresponding to each first measurement point, and taking the ideal measurement value as the second measurement value, and the second measurement value set is composed of a plurality of second measurement values. Since the first batch of wafers comprises a center region and an edge region, there is a large difference between the structures of the center region and the edge region, and it is not the most reasonable to obtain the measurement values of the center region and the edge region by using a unified measurement standard. Therefore, by performing operation processing on the measurement value set, the ideal measurement value corresponding to each first measurement point is obtained according to the difference between the center region and the edge region, and the ideal value is taken as the second measurement value, so that the deviation value set between the second measurement value set and the first measurement value set is more accurate, and the measurement accuracy can be further improved.
[0032] Further, after obtaining the fourth measurement value, the method further comprises: detecting the deviation value to obtain a detection result; and determining whether to re-measure the overlay deviation according to the detection result. By detecting in a timely manner, stable measurement accuracy can be ensured, and product quality can be improved.
[0033] Further, when re-measuring the overlay deviation, the number of wavelengths in the multi-wavelength light source is the optimal number of wavelengths. By using the optimal number of wavelengths, it is avoided to use a multi-wavelength light source with 1 to N wavelengths to measure P first measurement points respectively, and the measurement efficiency is effectively improved. BRIEF DESCRIPTION OF DRAWINGS
[0034] Figure 1 is a flow chart of an overlay deviation measurement method according to an embodiment of the present application;
[0035] Figures 2 to 11 is a structure schematic diagram of each step of an embodiment of the overlay deviation measurement method according to the present application;
[0036] Figures 12 to 13 is a structure schematic diagram of each step of another embodiment of the overlay deviation measurement method according to the present application. DETAILED DESCRIPTION
[0037] As described in the background, the overlay measurement based on optical diffraction in the prior art still has many problems. The following will be described in detail.
[0038] Research shows that in the process of overlay measurement based on optical diffraction, the measurement accuracy of dual-wavelength (hereinafter referred to as DW) is better than that of single-wavelength (hereinafter referred to as SW), and the measurement accuracy of multi-wavelength (hereinafter referred to as MW) is better than that of dual-wavelength. Therefore, DW / MW is gradually popularized in smaller nodes.
[0039] The single wavelength selected by SW is the best condition selected based on a certain film layer state. This wavelength can be applied to the case where the film layer pattern changes little, but SW is difficult to cope with the case where the film layer pattern is large asymmetric, the film layer uniformity difference in the wafer is large, and the film layer stacking difference between wafers is large. Especially in the new process research and development stage, the process change usually exceeds the effective interval of SW, and the edge region of the wafer in the mass production stage usually also exceeds the effective interval of SW. MW is the most effective solution at present.
[0040] However, with the increase of the number of wavelengths, the measurement of MW means that the wafer is measured at all selected wavelengths, and the time is multiplied, which has a great impact on the production capacity and cannot meet the demand of mass production.
[0041] On this basis, the present application provides a kind of overlay deviation measurement method, using the first wavelength light source is measured to the second batch wafer, obtains third measurement value group, third measurement value group includes a plurality of third measurement values corresponding to the second measurement point;The third measurement value group and the deviation value group are added to obtain the fourth measurement value group, and the fourth measurement value group includes a plurality of fourth measurement values corresponding to the second measurement point. Although the measurement efficiency is high, the measurement accuracy is low by using the first wavelength light source, the third measurement value group is added and compensated by the deviation value group obtained before, so that the measurement accuracy of the third measurement value group is effectively improved, and the measurement accuracy of the second batch wafer is effectively improved, and the measurement efficiency of the second batch wafer is improved.
[0042] In order to make the above-mentioned purpose, characteristics and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings.
[0043] Figure 1 It is the flow chart of the overlay deviation measurement method of the embodiment of the present application, comprising:
[0044] Step S11, providing a first batch of wafers, the first batch of wafers includes a plurality of first measurement points;
[0045] Step S12, using a first wavelength light source to measure the first batch of wafers to obtain a first measurement value group, the first measurement value group including a plurality of first measurement values corresponding to the first measurement points;
[0046] Step S13, using a second wavelength light source to measure the first batch of wafers to obtain a second measurement value group, the second measurement value group including a plurality of second measurement values corresponding to the first measurement points;
[0047] Step S14, obtaining a deviation value group between the second measurement value group and the first measurement value group, the deviation value group including a plurality of deviation values corresponding to the first measurement points;
[0048] Step S15, providing a second batch of wafers, the second batch of wafers including a plurality of second measurement points corresponding to the first measurement points;
[0049] Step S16, using the first wavelength light source to measure the second batch of wafers to obtain a third measurement value group, the third measurement value group including a plurality of third measurement values corresponding to the second measurement points;
[0050] Step S17, adding the third measurement value group and the deviation value group to obtain a fourth measurement value group, the fourth measurement value group including a plurality of fourth measurement values corresponding to the second measurement points.
[0051] The steps of the measurement method will be described in detail below with reference to the accompanying drawings.
[0052] Figures 2 to 10 is a structural schematic diagram of the steps of the overlay deviation measurement method of the embodiment of the present application.
[0053] Please refer to Figure 2 , providing a first batch of wafers 100, the first batch of wafers 100 including a plurality of first measurement points.
[0054] In this embodiment, there is an overlay deviation problem on each of the first measurement points in the first batch of wafers 100, and these overlay deviations are formed due to the misalignment of the overlay position of the layer pattern and the previous layer pattern in the photolithography manufacturing process. Since the integrated circuit chip is manufactured by adding multiple circuit layers, if the layer pattern and the previous layer pattern are not aligned, the chip will not work properly.
[0055] Therefore, the size of the overlay deviation is measured accurately, and then the overlay deviation is compensated in the subsequent process to ensure that the final compensated overlay deviation is within the deviation range.
[0056] Please refer to Figure 3The first batch of wafers 100 is measured by using a first wavelength light source to obtain a first measurement value group, wherein the first measurement value group includes a plurality of first measurement values 101 corresponding to the first measurement points.
[0057] In the embodiment, the first wavelength light source includes a single-wavelength light source or a dual-wavelength light source.
[0058] Although the single-wavelength light source or the dual-wavelength light source has high measurement efficiency, the measurement accuracy of the single-wavelength light source or the dual-wavelength light source is relatively low, and if only the single-wavelength light source or the dual-wavelength light source is used for measurement, the final wafer yield after deviation compensation will be reduced.
[0059] In the embodiment, the method for obtaining the first measurement value group includes: measuring the first batch of wafers 100 by using a first wavelength light source to obtain a first measurement value set, wherein the first measurement value set includes a plurality of initial first measurement values corresponding to the first measurement points; removing a plurality of first abnormal measurement values in the first measurement value set to obtain an initial first measurement value group; providing a data fitting model; performing data fitting operation on the initial first measurement value group according to the data fitting model to obtain first fitting values corresponding to the first abnormal measurement values, and the first fitting values and the initial first measurement value group form the first measurement value group.
[0060] In the embodiment, the method for obtaining the first fitting values includes: obtaining the first measurement points corresponding to the first abnormal measurement values as first abnormal points; obtaining the first measurement points adjacent to the first abnormal points as first fitting reference points; obtaining the initial first measurement values corresponding to the first fitting reference points as first fitting reference values; and performing the data fitting operation on a plurality of the first fitting reference values to obtain the first fitting values.
[0061] In the embodiment, the method for obtaining the first abnormal measurement values includes: obtaining a first standard deviation sigma1 of the first measurement value set; obtaining a first value reference range, wherein the first value reference range is 4 times the first standard deviation sigma1 to 6 times the first standard deviation sigma1; and regarding the initial first measurement values outside the first value reference range as the first abnormal measurement values.
[0062] Please refer to Figure 4 The first batch of wafers 100 is measured by using a second wavelength light source to obtain a second measurement value group, wherein the second measurement value group includes a plurality of second measurement values 102 corresponding to the first measurement points.
[0063] The second wavelength light source includes a multi-wavelength light source, wherein the number of wavelengths in the multi-wavelength light source is n, and n is greater than or equal to 3.
[0064] Although the measurement accuracy of the multi-wavelength light source is high, the measurement efficiency of the multi-wavelength light source is low, and if only the multi-wavelength light source is used for measurement, the production efficiency will be affected.
[0065] In the embodiment, the method for obtaining the second measurement value set comprises: measuring the first batch of wafers 100 by using the second wavelength light source to obtain a second measurement value set, wherein the second measurement value set comprises a plurality of initial second measurement values corresponding to the first measurement points; removing a plurality of second abnormal measurement values in the second measurement value set to obtain an initial second measurement value set; providing a data fitting model; performing data fitting operation on the initial second measurement value set according to the data fitting model to obtain a second fitting value corresponding to the second abnormal measurement value, and the second fitting value and the initial second measurement value set form the second measurement value set.
[0066] In the embodiment, the method for obtaining the second fitting value comprises: obtaining the first measurement point corresponding to the second abnormal measurement value as a second abnormal point; obtaining the first measurement point adjacent to the second abnormal point as a second fitting reference point; obtaining the initial second measurement value corresponding to the second fitting reference point as a second fitting reference value; and performing the data fitting operation on a plurality of second fitting reference values to obtain the second fitting value.
[0067] In the embodiment, the method for obtaining the second abnormal measurement value comprises: obtaining a second standard deviation sigma1 of the second measurement value set; obtaining a second value reference range, wherein the second value reference range is 4 times the second standard deviation sigma2 to 6 times the second standard deviation sigma2; and the initial second measurement value outside the second value reference range is the second abnormal measurement value.
[0068] Please refer to Figure 5 , obtain a deviation value set between the second measurement value set and the first measurement value set, wherein the deviation value set comprises a plurality of deviation values 103 corresponding to the first measurement points.
[0069] In the embodiment, the method for obtaining the deviation value set between the second measurement value set and the first measurement value set comprises: performing difference processing on the second measurement value 102 and the corresponding first measurement value 101 to obtain a plurality of deviation values 103, and the deviation value set is composed of a plurality of deviation values 103.
[0070] The first batch of wafers 100 are measured by a multi-wavelength light source, and a single-wavelength light source or a dual-wavelength light source, respectively, and then a deviation value group of two measurement results is obtained, and the deviation value group is used for adding compensation to a third measurement value group obtained by subsequently measuring a second batch of wafers 200 only by a single-wavelength light source or a dual-wavelength light source, so that the measurement of the second batch of wafers 200 can have both higher measurement efficiency and higher measurement accuracy.
[0071] Please refer to Figure 6 Through a large amount of measurement data, it is known that when the number of wavelengths in the multi-wavelength light source reaches a certain value, the increase in the accuracy of the second measurement value 102 obtained by increasing the number of wavelengths in the multi-wavelength light source for measurement is very small, and the increase in the deviation value 103 of each first measurement point is also very small. However, each increase in the number of wavelengths for measurement consumes a large amount of measurement time. Therefore, in order to ensure higher measurement efficiency and not to reduce the measurement accuracy too much, in the embodiment, the number n of wavelengths in the multi-wavelength light source is 4 or 5.
[0072] Please refer to Figure 7 In the embodiment, the acquisition method of the multi-wavelength light source includes: in a wavelength range of a preset light wave, a plurality of first-level wavelengths are acquired, and the first-level wavelengths are wavelengths at wave peaks and wave troughs of the preset light wave.
[0073] The purpose of first selecting the wavelengths at the wave peaks and the wave troughs in the preset light wave is that the change rate of the wavelengths at the wave peaks and the wave troughs is the lowest, and the measurement result is more stable.
[0074] In the embodiment, the acquisition method of the multi-wavelength light source further includes: when the number of wavelengths required by the multi-wavelength light source is greater than the number of the first-level wavelengths, a plurality of second-level wavelengths are acquired in the wavelength range of the preset light wave, and the second-level wavelengths are wavelengths at ±1 / 2 amplitudes of the preset light wave.
[0075] In the embodiment, the acquisition method of the multi-wavelength light source further includes: when the number of wavelengths required by the multi-wavelength light source is greater than the sum of the number of the first-level wavelengths and the number of the second-level wavelengths, a plurality of third-level wavelengths are acquired in the wavelength range of the preset light wave, and the third-level wavelengths are wavelengths at ±3 / 4 amplitudes of the preset light wave.
[0076] Please refer to Figure 8 The second batch of wafers 200 is provided, and the second batch of wafers 200 includes a plurality of second measurement points corresponding to the first measurement points.
[0077] In the embodiment, the second batch of wafers 200 also has the problem of overlay deviation at each second measurement point. The deviation value group obtained by measuring the first batch of wafers 100 is used for addition compensation of the measurement of the second batch of wafers 200, so that the measurement of the second batch of wafers 200 has higher measurement accuracy while obtaining higher measurement efficiency.
[0078] Please refer to Figure 9 , the first wavelength light source is used to measure the second batch of wafers 200 to obtain a third measurement value group, and the third measurement value group includes a plurality of third measurement values 201 corresponding to the second measurement points.
[0079] The process of measuring the second batch of wafers 200 by using the first wavelength light source is the same as the process of measuring the first batch of wafers 100 by using the first wavelength light source. For details, please refer to Figure 3 and related descriptions, which will not be repeated here.
[0080] Please refer to Figure 10 , the third measurement value group and the deviation value group are added to obtain a fourth measurement value group, and the fourth measurement value group includes a plurality of fourth measurement values 202 corresponding to the second measurement points.
[0081] In the embodiment, the first batch of wafers 100 includes a plurality of first measurement points. The first wavelength light source is used to measure the first batch of wafers 100 to obtain a first measurement value group, and the first measurement value group includes a plurality of first measurement values 101 corresponding to the first measurement points. The second wavelength light source is used to measure the first batch of wafers 100 to obtain a second measurement value group, and the second measurement value group includes a plurality of second measurement values 102 corresponding to the first measurement points. Since the measurement efficiency of the first wavelength light source is high but the measurement accuracy is low, and the measurement accuracy of the second wavelength light source is high but the measurement efficiency is low, a deviation value group between the second measurement value group and the first measurement value group is obtained, and the deviation value group includes a plurality of deviation values 103 corresponding to the first measurement points. The deviation value group is used for addition compensation when the second batch of wafers 200 is measured by using the first wavelength light source to obtain the fourth measurement value group 202, so that the second batch of wafers 200 has high measurement efficiency and the measurement accuracy is also effectively improved.
[0082] Please refer to Figure 11 , after obtaining the fourth measurement value 202, the deviation value 103 is detected to obtain a detection result, and whether to re-measure the overlay deviation is determined according to the detection result.
[0083] In the embodiment, the method for determining whether to re-measure the overlay deviation according to the detection result comprises: providing a detection threshold; comparing the deviation value 103 with the detection threshold; and when the deviation threshold is greater than the detection threshold, re-measuring the overlay deviation is needed. The stable measurement accuracy can be ensured by the timing detection, and the product quality is improved.
[0084] However, in the embodiment, the number of wavelengths used by the second wavelength light source in measurement is fixed, but the first batch of wafers 100 includes a center region and an edge region, and the structures of the center region and the edge region have large differences. If the number of wavelengths is fixed to obtain the measurement values of the center region and the edge region, it is not the most reasonable, and therefore, in order to further improve the measurement accuracy of the overlay deviation, the application also provides another embodiment of the measurement method of the overlay deviation.
[0085] Figures 12 to 13 is a schematic structural diagram of each step of the measurement method of the overlay deviation of another embodiment of the application.
[0086] The embodiment is based on the above-mentioned embodiment and continues to describe the measurement method of the overlay deviation. The difference between the embodiment and the above-mentioned embodiment is that the method for obtaining the second measurement value set comprises: measuring the first batch of wafers 100 by using a second wavelength light source to obtain a measurement value set, the measurement value set comprising a plurality of initial second measurement values; and performing operation processing on the measurement value set to obtain an ideal measurement value corresponding to each first measurement point, taking the ideal measurement value as the second measurement value 102, and comprising the second measurement value set by a plurality of second measurement values. For specific processes, please refer to Figures 12 to 13 .
[0087] Please refer to Figure 12 , a plurality of multi-wavelength light sources with different numbers of wavelengths are used to measure P first measurement points, the number of wavelengths of each multi-wavelength light source is i, and 1≤i≤N, N initial second measurement values 104 of each first measurement point are obtained, and P*N initial second measurement values 104 of P first measurement points are obtained, and the measurement value set is composed of P*N initial second measurement values 104.
[0088] In the embodiment, the plurality of initial second measurement values obtained by using the multi-wavelength light source with the same number of wavelengths each time also need to be removed from the abnormal points and subjected to data fitting operation. For processes, please refer to FIG. xx and related descriptions, which will not be repeated here.
[0089] In the embodiment, the measurement value set composed of P*N initial second measurement values 104 is represented by a matrix A, that is:
[0090]
[0091] Please refer to Figure 13 , the linear regression value corresponding to each of the first measurement points is obtained on the linear regression, and the linear regression value is taken as the ideal measurement value.
[0092] In this embodiment, the method for obtaining the linear regression value comprises:
[0093] A coordinate coefficient model x is provided, and the x is a parameter matrix, so that A·x can always obtain the most accurate value. Since the surface of the first batch of wafers 100 is affected by the regularity during measurement, each item in x is obviously an equation about coordinates, that is:
[0094]
[0095] Wherein, (L x , L y ) represents the coordinates of each of the first measurement points of the first batch of wafers 100;
[0096] The Taylor expansion is performed on the equation in the parameter matrix x, that is:
[0097] f N (L x , L y )=K N00 +K N10 ·L x +K N01 ·L y +…+R
[0098] Since the coordinate range of the first batch of wafers 100 is always [-150, 150] on the wafer, it can always be expanded as a polynomial in this range.
[0099] If A·x is expanded as a five-order polynomial to obtain A’·x’, then A’ will be a matrix of P*21N size, that is:
[0100]
[0101] Since the accuracy of the best measurement value 105 is always closest to the ideal measurement value in the N initial second measurement values 104 of each of the first measurement points, the regression is performed on A·x=[OV mw ], wherein OV mw is the best measurement value 105 corresponding to each of the first measurement points, [OV mw ]=[OV]+R, wherein OV is the ideal measurement value, and R is the residual between the ideal measurement value and the best measurement value 105.
[0102] In the embodiment, the method for obtaining the optimal measurement value 105 includes: taking the initial second measurement value 104 with the smallest difference from the ideal measurement value as the optimal measurement value, that is, in:
[0103] f N (L x , L y ) = K N00 + K N10 · L x + K N01 · L y + … + R
[0104] In the formula, the initial second measurement value 104 corresponding to the smallest R value is the optimal measurement value 105.
[0105] Therefore, in the embodiment, by performing operation processing on the measurement value set, the ideal measurement value corresponding to each first measurement point is obtained according to the difference between the center area and the edge area, and the ideal value is taken as the second measurement value 102, so that the deviation value set between the second measurement value set and the first measurement value set is more accurate, and the measurement accuracy can be further improved.
[0106] In the embodiment, after obtaining the ideal measurement value of each first measurement point, the method further includes: obtaining the optimal wavelength number of each first measurement point during measurement according to the ideal measurement value.
[0107] In the embodiment, the method for obtaining the optimal wavelength number of each first measurement point during measurement includes: obtaining an optimal measurement value in N initial second measurement values in each first measurement point, and taking the wavelength number corresponding to the optimal measurement value as the optimal wavelength number.
[0108] Please continue to refer to Figure 11 After obtaining the fourth measurement value 202, the deviation value 103 is detected to obtain a detection result; and whether to re-measure the overlay deviation is determined according to the detection result.
[0109] In the embodiment, the method for determining whether to re-measure the overlay deviation according to the detection result includes: providing a detection threshold; comparing the deviation value 103 with the detection threshold; and when the deviation threshold is greater than the detection threshold, the overlay deviation needs to be re-measured. By detecting at a fixed time, the stable measurement accuracy can be ensured, and the product quality can be improved.
[0110] In the embodiment, the number of wavelengths in the multi-wavelength light source is the optimal number of wavelengths when the overlay deviation is re-measured. By using the optimal number of wavelengths, the measurement efficiency is effectively improved by avoiding the use of the multi-wavelength light source with 1 to N number of wavelengths to measure P first measurement points respectively.
[0111] Although the present application is disclosed by the above, the present application is not limited to this. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application, and the protection scope of the present application should be limited by the scope defined by the claims.
Claims
1. A method for measuring overlay deviation, characterized in that, include: Provide a first batch of wafers, which includes a number of first measurement points; The first batch of wafers is measured using a first wavelength light source to obtain a first set of measurement values, which includes several first measurement values corresponding to the first measurement point. The first batch of wafers is measured using a second wavelength light source to obtain a second set of measurement values, which includes several second measurement values corresponding to the first measurement point. Obtain a deviation value group between the second measurement value group and the first measurement value group, wherein the deviation value group includes several deviation values corresponding to the first measurement point; A second batch of wafers is provided, the second batch of wafers including a plurality of second measurement points corresponding to the first measurement point; The second batch of wafers is measured using the first wavelength light source to obtain a third set of measurement values, which includes several third measurement values corresponding to the second measurement points. The third set of measured values is added to the set of deviation values to obtain a fourth set of measured values, which includes several fourth measured values corresponding to the second measurement points; wherein... The first wavelength light source includes: a single-wavelength light source or a dual-wavelength light source; The second wavelength light source includes: a multi-wavelength light source, wherein the number of wavelengths in the multi-wavelength light source is n, and n≥3.
2. The method for measuring overlay deviation as described in claim 1, characterized in that, The number of wavelengths n in the multi-wavelength light source is 4 or 5.
3. The method for measuring overlay deviation as described in claim 1, characterized in that, The method for acquiring the multi-wavelength light source includes: acquiring several first-level wavelengths within a certain wavelength range, wherein the first-level wavelengths are the wavelengths at the peaks and troughs within the certain wavelength range.
4. The method for measuring overprinting deviation as described in claim 3, characterized in that, The method for obtaining a multi-wavelength light source further includes: when the number of wavelengths required by the multi-wavelength light source is greater than the number of first-level wavelengths, obtaining a number of second-level wavelengths within a certain wavelength range, wherein the second-level wavelengths are wavelengths at ±1 / 2 amplitude within a certain wavelength range.
5. The method for measuring overlay deviation as described in claim 4, characterized in that, The method for obtaining a multi-wavelength light source further includes: when the number of wavelengths required by the multi-wavelength light source is greater than the sum of the number of first-level wavelengths and the number of second-level wavelengths, obtaining a number of third-level wavelengths within a certain wavelength range, wherein the third-level wavelengths are wavelengths at ±3 / 4 of the amplitude within a certain wavelength range.
6. The method for measuring overlay deviation as described in claim 1, characterized in that, The method for obtaining the first set of measurement values includes: measuring the first batch of wafers using a first wavelength light source to obtain a first set of measurement values, the first set of measurement values including several initial first measurement values corresponding to the first measurement points; removing several first abnormal measurement values from the first set of measurement values to obtain an initial set of measurement values; providing a data fitting model; performing data fitting operations on the initial set of measurement values according to the data fitting model to obtain first fitted values corresponding to the first abnormal measurement values, and forming the first set of measurement values by the first fitted values and the initial first measurement values.
7. The method for measuring overlay deviation as described in claim 6, characterized in that, The method for obtaining the first fitted value includes: obtaining a first measurement point corresponding to the first abnormal measurement value as a first abnormal point; obtaining a first measurement point adjacent to the first abnormal point as a first fitting reference point; obtaining an initial first measurement value corresponding to the first fitting reference point as a first fitting reference value; and performing the data fitting operation on a plurality of first fitting reference values to obtain the first fitted value.
8. The method for measuring overlay deviation as described in claim 6, characterized in that, The method for obtaining the first abnormal measurement value includes: obtaining the first standard deviation sigma1 of the first measurement value set; obtaining a first value reference range, wherein the first value reference range is 4 times the first standard deviation sigma1 to 6 times the first standard deviation sigma1; and the initial first measurement value outside the first value reference range is the first abnormal measurement value.
9. The method for measuring overlay deviation as described in claim 1, characterized in that, The method for obtaining the second set of measurement values includes: measuring the first batch of wafers using a second wavelength light source to obtain a second set of measurement values, the second set of measurement values including several initial second measurement values corresponding to the first measurement points; removing several second abnormal measurement values from the second set of measurement values to obtain an initial set of second measurement values; providing a data fitting model; performing data fitting operations on the initial set of second measurement values according to the data fitting model to obtain second fitted values corresponding to the second abnormal measurement values, and forming the second set of measurement values by the second fitted values and the initial second measurement values.
10. The method for measuring overlay deviation as described in claim 9, characterized in that, The method for obtaining the second fitted value includes: obtaining a first measurement point corresponding to the second abnormal measurement value as a second abnormal point; obtaining a first measurement point adjacent to the second abnormal point as a second fitting reference point; obtaining an initial second measurement value corresponding to the second fitting reference point as a second fitting reference value; and performing the data fitting operation on a plurality of second fitting reference values to obtain the second fitted value.
11. The method for measuring overlay deviation as described in claim 9, characterized in that, The method for obtaining the second abnormal measurement value includes: obtaining the second standard deviation sigma1 of the second measurement value set; obtaining a second reference range of values, wherein the second reference range is 4 times the second standard deviation sigma2 to 6 times the second standard deviation sigma2; and the initial second measurement value outside the second reference range is the second abnormal measurement value.
12. The method for measuring overprinting deviation as described in claim 1, characterized in that, The method for obtaining the deviation value group between the second measurement value group and the first measurement value group includes: subtracting the second measurement value from the corresponding first measurement value to obtain a plurality of deviation values, and forming the deviation value group from the plurality of deviation values.
13. The method for measuring overprinting deviation as described in claim 1, characterized in that, The method for obtaining the second set of measurement values includes: measuring the first batch of wafers using a second wavelength light source to obtain a set of measurement values, the set of measurement values including several initial second measurement values; performing calculation processing on the set of measurement values to obtain an ideal measurement value corresponding to each first measurement point, using the ideal measurement value as the second measurement value, and forming the second set of measurement values from several second measurement values.
14. The method for measuring overlay deviation as described in claim 13, characterized in that, The method for obtaining the measurement value set includes: using N multi-wavelength light sources with different wavelength numbers to measure P first measurement points respectively, where the wavelength number of each multi-wavelength light source is i, and 1≤i≤N, obtaining N initial second measurement values for each first measurement point, and obtaining a total of P*N initial second measurement values for P first measurement points, and forming the measurement value set from the P*N initial second measurement values.
15. The method for measuring overlay deviation as described in claim 14, characterized in that, The method for obtaining the ideal measurement value corresponding to each first measurement point includes: performing a linear regression on P*N initial second measurement values, obtaining the linear regression value corresponding to each first measurement point on the linear regression, and using the linear regression value as the ideal measurement value.
16. The method for measuring overlay deviation as described in claim 15, characterized in that, After obtaining the ideal measurement value for each of the first measurement points, the method further includes: obtaining the optimal number of wavelengths for each of the first measurement points during measurement based on the ideal measurement value.
17. The method for measuring overlay deviation as described in claim 16, characterized in that, The method for obtaining the optimal number of wavelengths for each first measurement point during measurement includes: obtaining the optimal measurement value from the N initial second measurement values at each first measurement point, and taking the number of wavelengths corresponding to the optimal measurement value as the optimal number of wavelengths.
18. The method for measuring overlay deviation as described in claim 17, characterized in that, The method for obtaining the optimal measurement value includes: taking the initial second measurement value, which has the smallest difference from the ideal measurement value, as the optimal measurement value.
19. The method for measuring overlay deviation as described in claim 16, characterized in that, After obtaining the fourth measurement value, the method further includes: detecting the fourth measurement value and obtaining the detection result; and determining whether to remeasure the overlay deviation based on the detection result.
20. The method for measuring overlay deviation as described in claim 19, characterized in that, The method for determining whether to remeasure the overprinting deviation based on the detection results includes: obtaining the difference between the fourth measurement value and the standard deviation value and using it as a deviation threshold; providing a detection threshold; comparing the deviation threshold with the detection threshold; and when the deviation threshold is greater than the detection threshold, the overprinting deviation needs to be remeasured.
21. The method for measuring overprinting deviation as described in claim 20, characterized in that, When re-measuring the overlay deviation, the number of wavelengths in the multi-wavelength light source is the optimal number of wavelengths.
Citation Information
Patent Citations
Bi-directional beam divider, aligning system using same and lithography device using the system
CN101165597A
Method to determine a patterning process parameter
CN111133384A