Overlay error measurement method

Abstract

The application discloses a kind of overlay precision error measurement methods, comprising: step one, design a set of 2D OVL mark group including front layer and when layer 2D OVL mark.Step two, establish OCD model, comprising: step 21, after completing when layer photolithography process on the first wafer, using OCD measurement method is measured and obtains first spectral data.Step 22, using second OVL measurement method is measured and obtains first OVL result;Second OVL measurement method includes BSE OVL measurement method.Step 23, using first OVL result as reference and based on first spectral data, modeling is carried out and obtains OCD model.Step three, after completing when layer photolithography process on the second wafer, using OCD measurement method is measured and obtains second spectral data, second spectral data and the first spectral data of OCD model are matched and obtain second OVL result.The application can use OCD measurement to realize the OVL measurement of 2D OVL mark pattern, can improve the measurement speed and measurement accuracy simultaneously.

Description

Technical Field

[0001] This invention relates to a method for manufacturing semiconductor integrated circuits, and more particularly to a method for measuring overlay accuracy error (OVL). Background Technology

[0002] Overlay error, or OVL measurement, is one of the most important inspection methods in photolithography. OVL refers to the alignment requirements between the current layer pattern and the previous layer pattern. Only with layer-by-layer alignment can the functionality of the front-end devices be accurately transferred through the interconnect process; otherwise, the devices will not function, directly causing yield losses. Most OVL measurement methods are based on optical means, specifically divided into image-based overlay (IBO) and diffraction-based overlay (DBO) methods. However, from the essence of physical optics, both methods require obtaining high-order diffraction light signals from the measurement mark. The former is used to improve image quality, while the latter directly uses the relative intensity of + / -1 order light to calculate OVL. Since the diffraction angle of the high-order diffraction light signal is directly determined by the spatial period of the pattern, the smaller the period, the larger the diffraction angle. Therefore, in order to collect at least + / -1 order diffraction light, the period of the OVL mark is approximately 1 to 2 micrometers. With advancements in integrated circuit manufacturing technology, the critical dimension (CD) of semiconductor devices continues to shrink, with the minimum patterning period of critical photolithography layers at advanced nodes even less than 100nm. OVL (Optical Characteristic) marking designs with a patterning period size comparable to the device's dimensions cannot be applied to traditional OVL measurements. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method for measuring overlay accuracy error, which can use optical critical dimension (OCD) measurement to realize OVL measurement of 2D OVL marking graphics, which can not only improve the measurement speed, but also improve the measurement accuracy at the same time.

[0004] To solve the above-mentioned technical problems, the overlay accuracy error measurement method provided by the present invention includes the following steps:

[0005] Step 1: Design a set of 2D OVL labels, including the front layer 2D OVL labels and the current layer 2D OVL labels.

[0006] Step 2: Establish the OCD model, including:

[0007] Step 21: After completing the in-layer photolithography process on the first wafer, the 2D OVL marker group on the first wafer is measured using the OCD measurement method to obtain the first spectral data.

[0008] Step 22: Use the second OVL measurement method to measure the 2D OVL mark group on the first wafer and obtain the first OVL result; the second OVL measurement method includes the BSE OVL measurement method.

[0009] Step 23: Using the first OVL result as a reference, and modeling based on the first spectral data, the OCD model is obtained.

[0010] Step 3: After completing the in-layer photolithography process on the second wafer, the 2D OVL marker group on the second wafer is measured using the OCD measurement method to obtain the second spectral data. The second spectral data is then matched with the first spectral data of the OCD model to obtain the second OVL result.

[0011] A further improvement is that the front-layer 2D OVL label is designed based on the front-layer process requirements, and the front-layer 2D OVL label adopts a 2D grating structure.

[0012] The side lengths of the front 2D OVL markings are all greater than the longest axis of the incident light spot.

[0013] The 2D grating structure marked by the front layer 2D OVL is formed by a two-dimensional arrangement of the first block structure; the two-dimensional arrangement of the first block structure is either periodic or aperiodic.

[0014] A further improvement is that the current layer 2D OVL mark is designed based on the current layer process requirements, and the current layer 2D OVL mark adopts a 2D grating structure.

[0015] The side lengths of the 2D OVL markings in the current layer are all greater than the longest axis of the incident light spot.

[0016] The 2D grating structure marked by the current layer 2D OVL is formed by a two-dimensional arrangement of the second block structure; the two-dimensional arrangement of the second block structure is either periodic or aperiodic.

[0017] A further improvement is that the top view of the first block structure is circular or square; the first block structure is a raised island structure or a recessed hole structure.

[0018] A further improvement is that the top view of the second block structure is circular or square; the second block structure is a raised island structure or a recessed hole structure.

[0019] A further improvement is that, in the current layer lithography process, the placement rule for the current layer 2D OVL marker is to make the center coordinates of the current layer 2D OVL marker coincide with the center coordinates of the previous layer 2D OVL marker.

[0020] A further improvement is that, in the 2D OVL identifier group, each of the second block structures has an aligned first block structure.

[0021] Each of the first block structures has at least one aligned second block structure.

[0022] A further improvement is that, when the two-dimensional arrangement of the first block structure is a periodic arrangement, in the front layer 2D OVL identifier, the side length of the first block structure has a first side length value in the X direction, and the side length and spacing of the first block structure have a first periodic value.

[0023] In the Y direction, the side length of the first block structure has a second side length value, and the side length and spacing of the first block structure have a second periodic value.

[0024] A further improvement is that, when the two-dimensional arrangement of the second block structure is a periodic arrangement, in the current layer 2D OVL identifier, in the X direction, the side length of the second block structure has a third side length value, and the side length and spacing of the second block structure have a third periodic value.

[0025] In the Y direction, the side length of the second block structure has a fourth side length value, and the side length and spacing of the second block structure have a fourth periodic value.

[0026] A further improvement is that the first side length is an integer multiple of the third side length, and the first period is an integer multiple of the third period.

[0027] The second side length is an integer multiple of the fourth side length, and the second period is an integer multiple of the fourth period.

[0028] A further improvement is that the side lengths of the previous 2D OVL identifiers are all equal, and the side lengths of the current 2D OVL identifiers are all equal and equal to the side lengths of the previous 2D OVL identifiers.

[0029] A further improvement is that the first side length is half of the first period value, and the second side length is half of the second period value.

[0030] The third side length is 1 / 2 of the third period value, and the fourth side length is 1 / 2 of the fourth period value.

[0031] A further improvement is that, in step two, steps 21 to 23 need to be repeated multiple times to obtain multiple sets of the first spectral data and the corresponding first OVL results.

[0032] A further improvement is that in step 23, a rigorous coupled-wave algorithm or a machine learning-based method is used for modeling.

[0033] A further improvement is that the second OVL measurement method also includes an OVL measurement method based on secondary electrons and an OVL measurement method based on TEM slices.

[0034] The overlay accuracy error of this invention is achieved by combining a set of 2D OVL marking groups and OCD measurement, and is calibrated using the OVL results of a second OVL measurement method, such as the BSE OVL measurement method, wherein:

[0035] The 2D OVL marker group of the present invention can achieve a minimum effective period that is closer to the device pattern period than the minimum effective period of the existing 1D OVL marker. Therefore, the 2D OVL marker group of the present invention can better reflect the true condition of the device, thereby improving the accuracy of OVL measurement.

[0036] The 2D OVL label group can obtain OVL information in the x and y directions through a single graphic and measurement, thereby saving the graphic area of ​​the OVL label and the number of OVL measurements, and improving measurement efficiency.

[0037] OCD measurement uses optical methods and is much faster than BSE OVL measurement. Therefore, this invention can improve the measurement speed.

[0038] This invention uses a second OVL measurement method, such as the BSE OVL measurement method, for calibration. Since the second OVL measurement method has high accuracy, it can improve the accuracy of the first OVL result used in modeling, and further improve the accuracy of the second OVL result obtained by OCD measurement after modeling is completed. Therefore, it can ultimately improve the accuracy of OVL measurement.

[0039] Compared to existing methods that combine 2D OVL marker groups and BSE OVL measurement methods for OVL measurement, the OCD measurement speed of this invention is greater than that of BSE OVL measurement, thus improving measurement speed. Furthermore, the OCD measurement modeling of this invention uses the first OVL result obtained by the BSE OVL measurement method for calibration, so the measurement accuracy of the OCD measurement method of this invention is comparable to that of the BSE OVL measurement method. Therefore, this invention can achieve OVL measurement of 2D OVL marker graphics using OCD measurement, improving both measurement speed and accuracy. Attached Figure Description

[0040] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0041] Figure 1 This is a flowchart of the overlay accuracy error measurement method according to an embodiment of the present invention;

[0042] Figure 2A This is a schematic diagram of the structure of a front-layer 2D OVL mark in the overlay accuracy error measurement method of the present invention;

[0043] Figure 2B This is a schematic diagram of the structure of a 2D OVL mark in the overlay accuracy error measurement method of the present invention.

[0044] Figure 2C These are the graphics of the front layer 2D OVL mark and the current layer 2D OVL mark obtained by the BSE OVL measurement method in the overlay accuracy error measurement method of the present invention.

[0045] Figure 2D This is the first OVL result calculated from the graphic representations of the previous layer 2D OVL identifier and the current layer 2D OVL identifier obtained by the BSE OVL measurement method in the overlay accuracy error measurement method of this embodiment of the invention.

[0046] Figure 3A This is a schematic diagram of another front-layer 2D OVL marking structure in the overlay accuracy error measurement method of this invention;

[0047] Figure 3B This is a schematic diagram of another layer 2D OVL marking structure in the overlay accuracy error measurement method of the present invention;

[0048] Figure 3C This is a schematic diagram of the OCD measurement method used in the overlay accuracy error measurement method of the present invention.

[0049] Figure 4 This is a flowchart of the OCD model establishment process in the overlay accuracy error measurement method of this invention. Detailed Implementation

[0050] like Figure 1 The diagram shows a flowchart of the overlay accuracy error measurement method according to an embodiment of the present invention. The overlay accuracy error measurement method according to an embodiment of the present invention includes the following steps:

[0051] Step 1: Design a set of 2D OVL labels, including front layer 2D OVL label 101 and current layer 2D OVL label 103.

[0052] Embodiments of the present invention, such as Figure 2A As shown, the front-layer 2D OVL identifier 101 is designed based on the front-layer process requirements, and the front-layer 2D OVL identifier 101 adopts a 2D grating structure.

[0053] The side lengths of the front-layer 2D OVL marker 101 are all greater than the longest axis of the incident light spot 105. (Please refer to the image for the incident light spot 105.) Figure 3C . Figure 2A The image shows the side length L1 in the X direction of the front layer 2D OVL identifier 101.

[0054] The 2D grating structure of the front-layer 2D OVL identifier 101 is formed by a two-dimensional arrangement of first block structures 102; the two-dimensional arrangement of the first block structures 102 is periodic. In other embodiments, the two-dimensional arrangement of the first block structures 102 can also be non-periodic.

[0055] like Figure 2B As shown, the current layer 2D OVL mark 103 is designed based on the current layer process requirements, and the current layer 2D OVL mark 103 adopts a 2D grating structure.

[0056] The side lengths of the 2D OVL identifier 103 are all greater than the longest axis of the incident light spot 105. Figure 2B The image shows the side length L2 in the X direction of the 2D OVL identifier 103 of the current layer.

[0057] In this embodiment of the invention, the side lengths of the front layer 2D OVL identifier 101 are all equal, and the side lengths of the current layer 2D OVL identifier 103 are all equal and equal to the side lengths of the front layer 2D OVL identifier 101.

[0058] The 2D grating structure of the current layer 2D OVL identifier 103 is formed by a two-dimensional arrangement of the second block structure 104; the two-dimensional arrangement of the second block structure 104 is periodic. In other embodiments, the two-dimensional arrangement of the second block structure 104 can also be a non-periodic arrangement.

[0059] In embodiments of the present invention, such as Figure 2A As shown, the top view of the first block structure 102 is square. In some embodiments, the top view of the first block structure 102 may also be circular.

[0060] The first block structure 102 is a raised island structure or a recessed hole structure.

[0061] In embodiments of the present invention, such as Figure 2B As shown, the top view of the second block structure 104 is circular. In some embodiments, the top view of the second block structure 104 may also be square.

[0062] The second block structure 104 is a raised island structure or a recessed hole structure.

[0063] In this embodiment of the invention, in the current layer photolithography process, the placement rule of the current layer 2D OVL identifier 103 is that the center coordinates of the current layer 2D OVL identifier 103 coincide with the center coordinates of the previous layer 2D OVL identifier 101.

[0064] In this embodiment of the invention, in the 2D OVL identifier group, each of the second block structures 104 has an aligned first block structure 102. Each of the first block structures 102 has at least one aligned second block structure 104.

[0065] like Figure 2C As shown, the second block structure 104 and the first block structure 102 are in one-to-one correspondence.

[0066] like Figure 2A As shown, when the two-dimensional arrangement of the first block structure 102 is a periodic arrangement, in the front layer 2D OVL identifier 101, in the X direction, the side length of the first block structure 102 has a first side length value a1, and the side length and spacing of the first block structure 102 have a first periodic value b1.

[0067] In the Y direction, the side length of the first block structure 102 has a second side length value c1, and the side length and spacing of the first block structure 102 have a second period value d1.

[0068] like Figure 2B As shown, when the two-dimensional arrangement of the second block structure 104 is a periodic arrangement, in the current layer 2D OVL identifier 103, in the X direction, the side length of the second block structure 104 has a third side length value a2, and the side length and spacing of the second block structure 104 have a third periodic value b2.

[0069] In the Y direction, the side length of the second block structure 104 has a fourth side length value c2, and the side length and spacing of the second block structure 104 have a fourth period value d2.

[0070] In some embodiments, the first side length value a1 is half of the first period value b1, and the second side length value c1 is half of the second period value d1. Figure 2A In the above, the first side length value a1 and the second side length value c1 are equal.

[0071] The third side length value a2 is half of the third period value b2, and the fourth side length value c2 is half of the fourth period value d2. Figure 2B In this context, the third side length value a2 and the fourth side length value c2 are equal and are the diameter of the second block structure 104.

[0072] In this embodiment of the invention, the first side length value a1 is an integer multiple of 1 or more of the third side length value a2, and the first period value b1 is an integer multiple of 1 or more of the third period value b2. Figure 2A and Figure 2B In a corresponding 2DOVL identifier group, the first side length value a1 is equal to the third side length value a2, and the first period value b1 is equal to the third period value b2.

[0073] The second side length value c1 is an integer multiple of 1 or more of the fourth side length value c2, and the second period value d1 is an integer multiple of 1 or more of the fourth period value d2. Figure 2A and Figure 2B In a corresponding 2D OVL identifier group, the second side length value c1 and the fourth side length value c2 are equal, and the second period value d1 and the fourth period value d2 are equal.

[0074] The 2D OVL identifier group in the embodiments of the present invention will be further explained below with reference to specific numerical values:

[0075] like Figure 3A The diagram shown is a schematic diagram of another front-layer 2D OVL marking structure in the overlay accuracy error measurement method of this invention. Figure 3A The front-layer 2D OVL identifier 101a corresponds to a gate layer, i.e., a polysilicon gate layer (poly) process. The first block structure 102a of the front-layer 2D OVL identifier 101a is a polysilicon block. In some embodiments, the side lengths of the front-layer 2D OVL identifier are equal and equal to 40 μm.

[0076] In the X direction, the first side length is 50nm and the first period is 100nm.

[0077] In the Y direction, the polysilicon line end cut (LEC) 102b, i.e. the first block structure 102a in the Y direction, has a spacing of 50nm, and the second period value is 200nm, so the second side length value is 150nm.

[0078] like Figure 3B The diagram shown is a schematic diagram of another layer 2D OVL marking structure in the overlay accuracy error measurement method of this invention. Figure 3B The current layer process corresponding to the current layer 2D OVL identifier 103a is a contact hole layer process. The second block structure 104a of the current layer 2D OVL identifier 103a is a contact hole. In some embodiments, the side lengths of the current layer 2D OVL identifier 103a are equal and equal to 40 μm.

[0079] In the X direction, the third side length is 50nm and the third period is 100nm.

[0080] In the Y direction, the fourth side length is 50 nm, and the fourth period is 100 nm. It can be seen that the second period is twice the fourth period, and each of the first block structures 102a corresponds to the two second block structures 104a.

[0081] like Figure 3C The diagram shown is a schematic diagram of the measurement using the OCD measurement method in the overlay accuracy error measurement method of the present invention. For the OCD measurement method, please refer to the following steps 21 and 3. The OCD measurement method is to use an optical method for measurement. After the incident light spot 105 illuminates the 2DOVL mark group formed by the front layer 2D OVL mark 101a and the current layer 2D OVL mark 103a, the corresponding spectral data is obtained by collecting the 0th order light signals at different wavelengths.

[0082] Step 2: Establish the OCD model, including:

[0083] Step 21: After completing the in-layer photolithography process on the first wafer, the 2D OVL marker group on the first wafer is measured using the OCD measurement method to obtain the first spectral data.

[0084] Step 22: Use the second OVL measurement method to measure the 2D OVL marking group on the first wafer and obtain the first OVL result; the second OVL measurement method is the BSE OVL measurement method.

[0085] In some embodiments, the second OVL measurement method can also be replaced by a secondary electron-based OVL measurement method or a TEM slice-based OVL measurement method. The BSE OVL measurement method, the secondary electron-based OVL measurement method, and the TEM slice-based OVL measurement method can all obtain accurate first OVL results, which can be used as calibration data. In actual processes, one method can be selected as the second OVL measurement method as needed. Among them, the BSE OVL measurement method is more efficient and less costly.

[0086] Step 23: Using the first OVL result as a reference, and modeling based on the first spectral data, the OCD model is obtained.

[0087] In some embodiments, in step 23, a rigorous coupled-wave algorithm or a machine learning-based method is used for modeling.

[0088] In this embodiment of the invention, steps 21 to 23 need to be repeated multiple times to obtain multiple sets of the first spectral data and the corresponding first OVL results, so as to continuously optimize the OCD model.

[0089] like Figure 4 The diagram shown is a flowchart of establishing an OCD model in the overlay accuracy error measurement method of this invention, including:

[0090] Step S101: Obtain multiple first spectral data, each of which is obtained through step 21. Figure 4 The n first spectral data points are displayed, namely Spectrum1, Spectrum2, and so on up to Spectrum. n .

[0091] Step S102: Obtain multiple first OVL results. Each first OVL result is obtained through step 22. Figure 4 The n first OVL results are displayed, namely BSE OVL1, BSE OVL2, up to BSE OVL. n .

[0092] Step S103: Tuning the OCD library.

[0093] Step S104: Perform a linear correlation check, followed by judgment, such as determining whether the slope is within 1 ± 0.01 and the goodness of fit (R²). 2 Is it greater than 0.99, i.e., Slope within 1 ± 0.01 & R² > 0.99?

[0094] If the judgment result is negative (N), then return to step S103 to continue adjusting the OCD database.

[0095] If the judgment result is affirmative (Y), then proceed as follows:

[0096] Step S105: Output the OCD database, i.e., the OCD library.

[0097] Steps S103 to S105 above achieve step 23.

[0098] Step 3: After completing the in-layer photolithography process on the second wafer, the 2D OVL marker group on the second wafer is measured using the OCD measurement method to obtain the second spectral data. The second spectral data is then matched with the first spectral data of the OCD model to obtain the second OVL result.

[0099] The overlay accuracy error of this invention is achieved by combining a set of 2D OVL marking groups and OCD measurement, and is calibrated using the OVL results of a second OVL measurement method, such as the BSE OVL measurement method, wherein:

[0100] In the 2D OVL label group, the front layer 2D OVL label 101 and the current layer 2D OVL label 103 can achieve a minimum effective period that is closer to the device pattern period than the minimum effective period of the existing 1D OVL label. Therefore, the 2D OVL label group of the present invention is more likely to reflect the true condition of the device, thereby improving the accuracy of OVL measurement.

[0101] The 2D OVL label group can obtain OVL information in the x and y directions through a single graphic and measurement, thereby saving the graphic area of ​​the OVL label and the number of OVL measurements, and improving measurement efficiency.

[0102] OCD measurement uses optical methods and is much faster than BSE OVL measurement. Therefore, this invention can improve the measurement speed.

[0103] This invention uses a second OVL measurement method, such as the BSE OVL measurement method, for calibration. Since the second OVL measurement method has high accuracy, it can improve the accuracy of the first OVL result used in modeling, and further improve the accuracy of the second OVL result obtained by OCD measurement after modeling is completed. Therefore, it can ultimately improve the accuracy of OVL measurement.

[0104] Compared to existing methods that combine 2D OVL marker groups and BSE OVL measurement methods for OVL measurement, the OCD measurement speed of this invention is greater than that of BSE OVL measurement, thus improving measurement speed. Furthermore, the OCD measurement modeling of this invention uses the first OVL result obtained by the BSE OVL measurement method for calibration, so the measurement accuracy of the OCD measurement method of this invention is comparable to that of the BSE OVL measurement method. Therefore, this invention can achieve OVL measurement of 2D OVL marker graphics using OCD measurement, improving both measurement speed and accuracy.

[0105] Existing OVL measurement methods require a long OVL measurement marking period, resulting in a significant difference from the actual device dimensions. Furthermore, the accuracy of existing OVL measurements is also affected by the asymmetry of the OVL measurement markings. While BSE OVL can improve the accuracy of OVL measurements, its measurement speed is relatively slow.

[0106] Optical Critical Dimension (OCD) technology, based on scattering measurement, inversely calculates the required graphic structural information by characterizing the signal changes of 0th-order light at different wavelengths. Therefore, this technology can characterize OVL by measuring smaller periodic structures. However, two issues need to be addressed and optimized before applying this technology. First, traditional OVL markings are one-dimensional (1D) gratings, composed of lines and spaces. The diffraction distribution of a 1D grating is also 1D, meaning its 0th-order light signal only reflects OVL changes in one direction. To obtain OVL in two directions simultaneously, the OCD measurement grating should be modified to two-dimensional (2D). Second, OCD measurements require calibration with other measurement results, such as CD or height information obtained from transmission electron microscopy (TEM) images or data from scanning electron microscopy (SEM) images. OVL obtained through backscattered electron (BSE) imaging is unaffected by optical asymmetry effects and is considered a true OVL. Therefore, choosing BSE OVL for calibration can bring the accuracy of OCD OVL to the level of BSE OVL.

[0107] Based on the above principles, this invention implements a method for measuring overlay accuracy error based on scattering measurement. The method includes a 2D OVL marker design for OCD measurement and uses BSE OVL for scattering measurement OVL calibration. It has four advantages: First, the minimum effective period of the OVL marker is closer to the device pattern period, making it easier to reflect the true condition of the device; second, the 2D OVL marker can obtain OVL information in the x and y directions through a single pattern and measurement; third, using BSE OVL for calibration, the OVL accuracy of this method surpasses that of optical OVL measurement; fourth, since it still uses an optical method, its measurement speed is much higher than that of BSE OVL measurement.

[0108] The present invention has been described in detail above through specific embodiments, but these are not intended to limit the invention. Many modifications and improvements can be made by those skilled in the art without departing from the principles of the invention, and these should also be considered within the scope of protection of the present invention.

Claims

1. A method for measuring overlay accuracy error, characterized in that, Includes the following steps: Step 1: Design a set of 2D OVL labels, including the front layer 2D OVL labels and the current layer 2D OVL labels; Step 2: Establish the OCD model, including: Step 21: After completing the in-layer photolithography process on the first wafer, the 2D OVL marker group on the first wafer is measured using the OCD measurement method to obtain the first spectral data; Step 22: Measure the 2D OVL marker group on the first wafer using a second OVL measurement method and obtain a first OVL result; the second OVL measurement method includes BSE OVL measurement method, secondary electron-based OVL measurement method, or TEM slice-based OVL measurement method. Step 23: Using the first OVL result as a reference and modeling based on the first spectral data, the OCD model is obtained; Step 3: After completing the in-layer photolithography process on the second wafer, the 2D OVL marker group on the second wafer is measured using the OCD measurement method to obtain the second spectral data. The second spectral data is then matched with the first spectral data of the OCD model to obtain the second OVL result. The front-layer 2D OVL label is designed based on the front-layer process requirements, and the front-layer 2D OVL label adopts a 2D grating structure. The side lengths of the front 2D OVL markings are all greater than the longest axis of the incident light spot; The 2D grating structure marked by the front layer 2D OVL is formed by a two-dimensional arrangement of the first block structure; the two-dimensional arrangement of the first block structure is either periodic or aperiodic. The in-layer 2D OVL marking is designed based on the in-layer process requirements, and the in-layer 2D OVL marking adopts a 2D grating structure; The side lengths of the 2D OVL markings in the current layer are all greater than the longest axis of the incident light spot; The 2D grating structure marked by the current layer 2D OVL is formed by a two-dimensional arrangement of the second block structure; the two-dimensional arrangement of the second block structure is either periodic or aperiodic.

2. The method for measuring overlay accuracy error as described in claim 1, characterized in that: The top view of the first block structure is circular or square; the first block structure is a raised island structure or a recessed hole structure.

3. The method for measuring overlay accuracy error as described in claim 2, characterized in that: The top view of the second block structure is circular or square; the second block structure is a raised island structure or a recessed hole structure.

4. The method for measuring overlay accuracy error as described in claim 3, characterized in that: In the in-layer photolithography process, the placement rule for the in-layer 2D OVL marker is that the center coordinates of the in-layer 2D OVL marker coincide with the center coordinates of the previous layer 2D OVL marker.

5. The method for measuring overlay accuracy error as described in claim 4, characterized in that: In the 2D OVL identifier group, each of the second block structures has an aligned first block structure; Each of the first block structures has at least one aligned second block structure.

6. The method for measuring overlay accuracy error as described in claim 5, characterized in that: When the two-dimensional arrangement of the first block structure is a periodic arrangement, in the front layer 2D OVL identifier, in the X direction, the side length of the first block structure has a first side length value, and the side length and spacing of the first block structure have a first periodic value; In the Y direction, the side length of the first block structure has a second side length value, and the side length and spacing of the first block structure have a second periodic value.

7. The method for measuring overlay accuracy error as described in claim 6, characterized in that: When the two-dimensional arrangement of the second block structure is a periodic arrangement, in the current layer 2D OVL identifier, in the X direction, the side length of the second block structure has a third side length value, and the side length and spacing of the second block structure have a third periodic value; In the Y direction, the side length of the second block structure has a fourth side length value, and the side length and spacing of the second block structure have a fourth periodic value.

8. The method for measuring overlay accuracy error as described in claim 7, characterized in that: The first side length is an integer multiple of the third side length, and the first period is an integer multiple of the third period. The second side length is an integer multiple of the fourth side length, and the second period is an integer multiple of the fourth period.

9. The method for measuring overlay accuracy error as described in claim 1, characterized in that: All sides of the preceding 2D OVL identifier are of equal length, and all sides of the current 2D OVL identifier are of equal length and equal to the side lengths of the preceding 2D OVL identifier.

10. The method for measuring overlay accuracy error as described in claim 7, characterized in that: The first side length is half of the first period value, and the second side length is half of the second period value; The third side length is 1 / 2 of the third period value, and the fourth side length is 1 / 2 of the fourth period value.

11. The method for measuring overlay accuracy error as described in claim 1, characterized in that: In step two, steps 21 to 23 need to be repeated multiple times to obtain multiple sets of the first spectral data and the corresponding first OVL results.

12. The method for measuring overlay accuracy error as described in claim 11, characterized in that: In step 23, a rigorous coupled-wave algorithm or a machine learning-based method is used for modeling.

Images