Compensation Method for Alignment Deviation

Abstract

A method for compensating alignment deviation, comprising: obtaining alignment model deviation information according to a first alignment model and a second alignment model; after performing initial lithography alignment on an initial second photolithography layer, performing compensated lithography alignment on the initial second photolithography layer according to the alignment model deviation information to pattern the initial second photolithography layer and form a second photolithography layer. Thus, in the lithography process, the accuracy and reliability of lithography alignment are improved, and the stability of lithography alignment is also improved.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor manufacturing, and particularly to a method for compensating alignment deviation. Background Art

[0002] Lithography technology is a crucial technology in semiconductor manufacturing technology, which can transfer patterns from a mask to the surface of a silicon wafer to form a semiconductor product that meets the design requirements.

[0003] During the lithography process, before exposure, first, the wafer needs to be aligned, that is, the alignment layer used for alignment in the wafer is aligned with the mask.

[0004] During the process of wafer manufacturing, due to the influence of manufacturing processes, the wafer will deform. Therefore, the alignment marks of the alignment layer of the wafer are also affected by the deformation and change, resulting in a change in the actual position of the alignment marks. Usually, during the alignment process of the wafer, an alignment model is used to compensate the alignment process of the wafer according to the detected position information of the alignment marks, so as to improve the alignment degree between the actual position of the alignment marks and the mask.

[0005] In order to make alignment compensation for the minute deformation of the wafer to further improve the lithography alignment accuracy, a Radial Basis Function (RBF) alignment model is proposed. Since the RBF alignment model has high compensation sensitivity, it can also make sufficient alignment compensation for the minute deformation of the wafer.

[0006] However, when detecting the position information of the alignment marks, affected by some special process steps, such as some high-temperature steps, planarization steps, etc., it may lead to misdetection of the position information of the alignment marks, resulting in a large deviation between the position of the alignment marks reflected by the detected position information of the alignment marks and the actual position of the alignment marks. Since compensation alignment is performed according to the RBF alignment model for the wrong position information of the alignment marks, when misdetection occurs, wrong compensation will be generated accordingly during compensation alignment according to the RBF alignment model, resulting in low lithography alignment accuracy, poor reliability, and low overlay accuracy. Summary of the Invention

[0007] The technical problem solved by the present invention is to provide a method for compensating alignment deviation, so as to improve the accuracy and reliability of lithography alignment, and improve the stability of lithography alignment in the lithography process, so as to improve the overlay accuracy.

[0008] To solve the above technical problems, the technical solution of the present invention provides a compensation method for alignment deviation, including: providing a wafer to be compensated, which includes an initial first layer and an initial first photolithography layer located on the initial first layer; providing a first alignment model and a second alignment model; performing photolithography alignment on the initial first photolithography layer according to the first alignment model to pattern the initial first photolithography layer to form a first photolithography layer; patterning the initial first layer according to the first photolithography layer to form a first layer; forming an initial second photolithography layer on the first layer; performing initial photolithography alignment on the initial second photolithography layer according to the second alignment model; obtaining alignment model deviation information according to the first alignment model and the second alignment model; after performing initial photolithography alignment on the initial second photolithography layer, performing compensated photolithography alignment on the initial second photolithography layer according to the alignment model deviation information to pattern the initial second photolithography layer to form a second photolithography layer.

[0009] Optionally, the wafer to be compensated further includes: a first alignment layer, the initial first layer is located on the first alignment layer, and when performing photolithography alignment on the initial first photolithography layer, it aligns with the first alignment layer.

[0010] Optionally, the method for obtaining alignment model deviation information according to the first alignment model and the second alignment model includes: providing first preset alignment position information of the first alignment layer; obtaining first detected alignment position information according to the first preset alignment position information and the first alignment layer; obtaining first alignment offset information according to the first detected alignment position information and the first alignment model; obtaining second alignment offset information according to the first detected alignment position information and the second alignment model; obtaining alignment model deviation information according to the first alignment offset information and the second alignment offset information.

[0011] Optionally, when performing photolithography alignment on the initial first photolithography layer according to the first alignment model, the first alignment offset information is obtained.

[0012] Optionally, when performing photolithography alignment on the initial first photolithography layer according to the first alignment model, the second alignment offset information is obtained.

[0013] Optionally, before patterning the initial first photolithography layer to form a first photolithography layer, photolithography alignment is performed on the initial first photolithography layer according to the second alignment model to obtain the second alignment offset information.

[0014] Optionally, the first alignment layer includes a plurality of first alignment marks, and the first preset alignment position information includes the preset position information of each first alignment mark; the first photolithography layer includes a plurality of first photolithography alignment marks, and each first photolithography alignment mark corresponds to 1 first alignment mark.

[0015] Optionally, the first alignment offset information includes a plurality of first offset information, and each first lithography alignment mark corresponds to one first offset information; the second alignment offset information includes a plurality of second offset information, and each first lithography alignment mark also corresponds to one second offset information; the alignment model deviation information includes: the model offset information of each first lithography alignment mark, and the model offset information is the deviation information between one first offset information corresponding to the first lithography alignment mark and one second offset information.

[0016] Optionally, the method for obtaining the first detection alignment position information includes: according to the preset position information of each of the first alignment marks, detecting the position of each corresponding first alignment mark in the first alignment layer to obtain the detection position information of each first alignment mark, and the first detection alignment position information includes the detection position information of each first alignment mark.

[0017] Optionally, the method for obtaining the first offset information includes: according to the first alignment model, the preset position information of one first alignment mark, and the detection position information of the first alignment mark, obtaining one first offset information corresponding to one first lithography alignment mark corresponding to the first alignment mark.

[0018] Optionally, the method for obtaining the second offset information includes: according to the second alignment model, the preset position information of one first alignment mark, and the detection position information of the first alignment mark, obtaining one second offset information corresponding to the first lithography alignment mark corresponding to the first alignment mark.

[0019] Optionally, the method for compensating lithography alignment further includes: providing an overlay error compensation model; before patterning the initial second lithography layer, further performing compensating lithography alignment on the initial second lithography layer according to the overlay error compensation model.

[0020] Optionally, the method for compensating lithography alignment further includes: providing an overlay error compensation model; obtaining a compensation alignment model according to the alignment model deviation information and the overlay error compensation model; performing compensating lithography alignment on the initial second lithography layer according to the compensation alignment model.

[0021] Optionally, it further includes: forming an initial second layer on the first layer before forming the initial second lithography layer; after forming the second lithography layer, patterning the initial second layer according to the second lithography layer to form the second layer.

[0022] Optionally, it further includes: patterning the first layer according to the second lithography layer to form the second layer.

[0023] Optionally, it further includes: providing a target overlay accuracy range for the second layer; providing a preset overlay accuracy range; providing an overlay safety factor X; providing n batches of wafers to be compensated, where n is a natural number; obtaining the overlay accuracy Y of the second layer of each wafer to be compensated in the k-th batch to the (k + i)-th batch among the n batches, where k is a natural number, i is a natural number, and k + i is within the range of n; when the target overlay accuracy range of the second layer is within the preset overlay accuracy range, and any (Y ± X) is within the target overlay accuracy range of the second layer, obtaining batch alignment model deviation information according to one of the alignment model deviation information of at least one batch among the first batch to the (k + i)-th batch; compensating and aligning the initial second photolithography layer of the wafer to be compensated in the (k + i + 1)-th batch according to the batch alignment model deviation information, so as to pattern the initial second photolithography layer of the wafer to be compensated in the (k + i + 1)-th batch, and form the second photolithography layer of the wafer to be compensated in the (k + i + 1)-th batch.

[0024] Optionally, the first alignment model is a radial basis function alignment model.

[0025] Optionally, the second alignment model is a linear alignment model or a high-order alignment model.

[0026] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0027] In the compensation method for alignment deviation provided by the technical solution of the present invention, the second photolithography layer is used to form the second layer of the wafer to be compensated. Since the initial first photolithography layer is lithographically aligned according to the first alignment model, and the initial second photolithography layer is initially lithographically aligned according to the second alignment model, during the formation of the first layer and the second layer, different alignment models can be respectively adopted according to different situations of the wafer to be compensated. Thus, by switching the alignment models, the applicability of each alignment model to the wafer in different situations is improved in lithographic alignment, so as to improve the accuracy and reliability of lithographic alignment. At the same time, since the alignment model deviation information is obtained according to the first alignment model and the second alignment model, and after the initial lithographic alignment of the initial second photolithography layer, according to the alignment model deviation information, and after the initial lithographic alignment of the initial second photolithography layer, the initial second photolithography layer is subjected to compensated lithographic alignment according to the alignment model deviation information. Therefore, after the initial lithographic alignment of the initial second photolithography layer, the model deviation between the first alignment model and the second alignment model can be compensated in the compensated lithographic alignment. Thus, the stability of the lithography process between the first layer and the second layer is improved. In summary, through the compensation method for alignment deviation, in the lithography process, the accuracy and reliability of lithographic alignment are improved, and the stability of lithographic alignment is improved, so that the overlay accuracy of the wafer to be compensated is high.

[0028] Furthermore, since the target overlay accuracy range of the second layer is within the preset overlay accuracy range, and any (Y±X) is within the target overlay accuracy range of the second layer, the overlay accuracy stability of the second layer of the wafer to be compensated in the (k+i)-th batch is high. On this basis, since the batch alignment model deviation information is obtained according to one of the alignment model deviation information of at least one batch among the 1st batch to the (k+i)-th batch, and according to the batch alignment model deviation information, the initial second photolithography layer of the wafer to be compensated in the (k+i + 1)-th batch is subjected to compensated lithographic alignment to pattern the initial second photolithography layer of the wafer to be compensated in the (k+i + 1)-th batch, and form the second photolithography layer of the wafer to be compensated in the (k+i + 1)-th batch. Therefore, one alignment model deviation information (batch alignment model deviation information) obtained according to historical processing experience is used to replace the alignment model deviation information corresponding to each wafer to be compensated in the (k+i + 1)-th batch to perform compensated lithographic alignment on the wafer to be compensated in the (k+i + 1)-th batch. Thus, on the basis of ensuring that the overlay accuracy of the formed second photolithography layer is relatively stable after compensated lithographic alignment, the efficiency of compensating the alignment deviation of the wafers to be compensated in each batch after the (k+i)-th batch is improved. Description of the Drawings

[0029] Figure 1It is a schematic flow chart of a method for compensating alignment deviation according to an embodiment of the present invention;

[0030] Figures 2 to 7 It is a schematic structural diagram of each step in the method for compensating alignment deviation according to an embodiment of the present invention. Detailed implementation manners

[0031] As described in the background art, when detecting the position information of the alignment mark, affected by some special process steps, such as some high-temperature steps, planarization steps, etc., it may lead to misdetection of the position information of the alignment mark, resulting in a large deviation between the alignment mark position reflected by the detected position information of the alignment mark and the actual alignment mark position. Since when compensating and aligning according to the RBF alignment model, compensation is also made for the wrong position information of the alignment mark, therefore, when misdetection occurs, when compensating and aligning according to the RBF alignment model, corresponding wrong compensation will be generated, resulting in low lithography alignment accuracy and poor reliability, and low overlay accuracy.

[0032] To solve the above technical problems, the technical solution of the present invention provides a method for compensating alignment deviation. After initial lithography alignment of the initial second lithography layer, compensation lithography alignment is performed on the initial second lithography layer according to the alignment model deviation information to pattern the initial second lithography layer and form the second lithography layer, thereby improving the lithography alignment accuracy and reliability of the wafer.

[0033] To make the above objects, features, and beneficial effects of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be given with reference to the accompanying drawings.

[0034] Figure 1 It is a schematic flow chart of a method for compensating alignment deviation according to an embodiment of the present invention.

[0035] Please refer to Figure 1 , the method for compensating alignment deviation includes:

[0036] Step S100, providing a wafer to be compensated, the wafer to be compensated includes an initial first layer and an initial first lithography layer located on the initial first layer;

[0037] Step S110, providing a first alignment model and a second alignment model;

[0038] Step S120, performing lithography alignment on the initial first lithography layer according to the first alignment model to pattern the initial first lithography layer and form the first lithography layer;

[0039] Step S130, patterning the initial first layer according to the first lithography layer to form the first layer;

[0040] Step S140, form an initial second photolithography layer on the first layer;

[0041] Step S150, perform initial photolithography alignment on the initial second photolithography layer according to the second alignment model;

[0042] Step S160, obtain alignment model deviation information according to the first alignment model and the second alignment model;

[0043] Step S170, after performing initial photolithography alignment on the initial second photolithography layer, perform compensated photolithography alignment on the initial second photolithography layer according to the alignment model deviation information to pattern the initial second photolithography layer and form a second photolithography layer.

[0044] Wherein, the second photolithography layer is used to form the second layer of the wafer to be compensated.

[0045] Since the initial first photolithography layer is subjected to photolithography alignment according to the first alignment model, and the initial second photolithography layer is subjected to initial photolithography alignment according to the second alignment model, therefore, during the formation of the first layer and the second layer, different alignment models can be respectively adopted according to different situations of the wafer to be compensated. Thus, by switching the alignment model, the applicability of each alignment model to the wafer in different situations in photolithography alignment is improved, so as to improve the accuracy and reliability of photolithography alignment. At the same time, since the alignment model deviation information is obtained according to the first alignment model and the second alignment model, and after performing photolithography alignment on the initial second photolithography layer, according to the alignment model deviation information, and after performing initial photolithography alignment on the initial second photolithography layer, the initial second photolithography layer is subjected to compensated photolithography alignment according to the alignment model deviation information. Therefore, after performing initial photolithography alignment on the initial second photolithography layer, the model deviation between the first alignment model and the second alignment model can be compensated in the compensated photolithography alignment, thereby improving the stability of the photolithography process between the first layer and the second layer. In summary, through the alignment deviation compensation method, in the photolithography process, the accuracy and reliability of photolithography alignment are improved, and the stability of photolithography alignment is improved, making the overlay accuracy of the wafer to be compensated high.

[0046] It should be noted that the order between step S130 to step S150 and step S160 does not affect the technical effect of the embodiment of the present invention. Therefore, step S160 can be performed before, after, or between step S130 to step S150.

[0047] The following will be described in detail with reference to the accompanying drawings.

[0048] Figures 2 to 7 It is a structural schematic diagram of each step in the alignment deviation compensation method according to an embodiment of the present invention.

[0049] Please refer to Figure 2 to provide a wafer 200 to be compensated, where the wafer 200 to be compensated includes an initial first layer 210 and an initial first photolithography layer 220 located on the initial first layer 210.

[0050] The initial first layer 210 is used to form the first layer subsequently.

[0051] The initial first photolithography layer 220 is used to form the first photolithography layer subsequently.

[0052] In this embodiment, the wafer 200 to be compensated further includes: a first alignment layer 110, the initial first layer 210 is located on the first alignment layer 110, and subsequently, when the initial first photolithography layer 220 is subjected to photolithography alignment, it aligns with the first alignment layer 110.

[0053] The first alignment layer 110 includes a plurality of first alignment marks (not shown), and each exposure field (shot) includes 1 of the first alignment marks.

[0054] It should be noted that the exposure field refers to the area covered by the lithography machine for one imaging. In the process of fabricating an integrated circuit on a wafer, for the convenience of the process fabrication, the wafer is divided into a plurality of exposure fields as the basic units in production, and the exposure fields are arranged periodically and repetitively on the wafer.

[0055] Please refer to Figure 3 to provide a first alignment model and a second alignment model; perform photolithography alignment on the initial first photolithography layer 220 according to the first alignment model to pattern the initial first photolithography layer 220 and form a first photolithography layer 221.

[0056] The method for performing photolithography alignment on the initial first photolithography layer 220 according to the first alignment model includes: providing first preset alignment position information of the first alignment layer 110; obtaining first detected alignment position information according to the first preset alignment position information and the first alignment layer 110; obtaining first alignment offset information according to the first detected alignment position information and the first alignment model; performing photolithography alignment on the initial first photolithography layer 220 according to the first alignment offset information.

[0057] Specifically, in this embodiment, when performing photolithography alignment on the initial first photolithography layer 220 according to the first alignment model, the first alignment offset information is obtained.

[0058] The first preset alignment position information includes preset position information of each of the first alignment marks.

[0059] The method for obtaining the first detection alignment position information includes: detecting the positions of each corresponding first alignment mark in the first alignment layer 110 according to the preset position information of each first alignment mark, so as to obtain the detection position information of each first alignment mark, and the first detection alignment position information includes the detection position information of each first alignment mark.

[0060] Specifically, in order to align the actual positions of the first alignment marks in subsequent lithography alignment, it is necessary to detect the actual positions of each first alignment mark to adjust the lithography alignment process, so that the subsequent formed lithography layer can be aligned with the first alignment layer 110 to meet the design requirements.

[0061] The preset position information of the first alignment mark refers to the preset information that can reflect the preset position of the first alignment mark. The detection position information of the first alignment mark refers to the detection information that can reflect the actual position of the first alignment mark. It can be seen that the preset position information and the detection position information of each first alignment mark correspond to each other.

[0062] It should be noted that in order not to perform destructive detection on the wafer to be compensated, an optical method is usually used to detect the actual position of the first alignment mark. Therefore, the detection information of the actual position of the first alignment mark is a detection information obtained through an optical signal, rather than the actual position of the first alignment mark. When detecting the position information of the first alignment mark, it is affected by some special process steps, such as some high-temperature steps, planarization steps, etc., resulting in a risk of misdetection of the detection position information of the first alignment mark. Thus, when there is such misdetection, there is a large deviation between the detection position of the first alignment mark reflected by the detected detection position information of the first alignment mark and the actual position of the first alignment mark.

[0063] In this embodiment, the first lithography layer 221 includes a plurality of first lithography alignment marks (not shown), and each first lithography alignment mark corresponds to one first alignment mark.

[0064] Specifically, when performing lithography alignment on the initial first lithography layer 220, a first mask plate for exposure (not shown) is provided. The pattern of the first mask plate includes a plurality of first mask alignment marks, and each first mask alignment mark corresponds to one first alignment mark. During the lithography alignment process, by moving the wafer to be compensated or moving the first mask plate, the corresponding first mask alignment mark is aligned with the first alignment mark. Thus, after exposure and development, a plurality of first lithography alignment marks are formed in the first lithography layer 221, and each first lithography alignment mark corresponds to one first alignment mark.

[0065] In this embodiment, the first alignment offset information includes a plurality of first offset information, and each first photolithography alignment mark corresponds to one first offset information.

[0066] Specifically, during the photolithography alignment process, the alignment model calculates the offset information of each first mask alignment mark based on the detected position information of each first alignment mark, so that each first mask alignment mark is as close as possible to its corresponding first alignment mark, realizing the alignment between the corresponding first mask alignment mark and the first alignment mark.

[0067] The first alignment model and the second alignment model are different alignment models.

[0068] Specifically, in this embodiment, the first alignment model is a radial basis function alignment model. The second alignment model is a high-order alignment model (HOWA Model, High Order Wafer Alignment Model).

[0069] In another embodiment, the second alignment model is a linear alignment model (Liner Alignment Model).

[0070] In another alternative embodiment, the first alignment model is a high-order alignment model or a linear alignment model, and the second alignment model is a radial basis function alignment model.

[0071] Since the first alignment model and the second alignment model are different alignment models, when calculating the offset information of the first mask alignment mark based on the same first detected alignment information, different offset information may be calculated. Thus, the first alignment model and the second alignment model respectively obtain different alignment offset information according to the first detected alignment information.

[0072] Among them, the first alignment offset information is obtained according to the first alignment model and the first detected alignment information. And in the first alignment offset information, each first offset information corresponds to one first mask alignment mark. Thus, each first photolithography alignment mark corresponds to one first offset information.

[0073] Specifically, the method for obtaining the first offset information includes: obtaining one first offset information corresponding to one first photolithography alignment mark corresponding to a first alignment mark according to the first alignment model and the detected position information of one first alignment mark.

[0074] In this embodiment, the method for lithographic alignment of the initial first photolithographic layer 220 further includes: providing an overlay error compensation model for the first layer; before patterning the initial first photolithographic layer 220, performing lithographic alignment on the initial first photolithographic layer 220 according to the overlay error compensation model of the first layer.

[0075] It should be noted that the lithographic alignment performed on the initial first photolithographic layer 220 can be to separately perform actual lithographic alignment actions according to the first alignment offset information and the overlay error compensation model of the first layer, or to perform only 1 actual lithographic alignment action after obtaining the parameters for lithographic alignment according to the first alignment offset information and the overlay error compensation model of the first layer.

[0076] Please continue to refer to Figure 3 , and obtaining alignment model deviation information according to the first alignment model and the second alignment model.

[0077] In this embodiment, the method for obtaining alignment model deviation information according to the first alignment model and the second alignment model includes: providing first preset alignment position information of the first alignment layer 110; obtaining first detected alignment position information according to the first preset alignment position information and the first alignment layer 110; obtaining first alignment offset information according to the first detected alignment position information and the first alignment model; obtaining second alignment offset information according to the first detected alignment position information and the second alignment model; and obtaining alignment model deviation information according to the first alignment offset information and the second alignment offset information.

[0078] Among them, providing the first preset alignment position information of the first alignment layer 110, obtaining the first detected alignment position information according to the first preset alignment position information and the first alignment layer 110, and obtaining the first alignment offset information according to the first detected alignment position information and the first alignment model have been described in the process of performing lithographic alignment on the initial first photolithographic layer 220, and will not be elaborated here.

[0079] Moreover, since the first alignment model and the second alignment model respectively obtain different alignment offset information according to the first detected alignment information, therefore, second alignment offset information is obtained according to the first detected alignment position information and the second alignment model.

[0080] In this embodiment, before patterning the initial first photolithographic layer 220 to form the first photolithographic layer 221, lithographic alignment is performed on the initial first photolithographic layer 220 according to the second alignment model to obtain the second alignment offset information.

[0081] It should be noted that performing photolithographic alignment on the initial first photolithographic layer 220 according to the second alignment model means that, through the second alignment model and based on the first detection alignment position information, the offset information of the first mask alignment marks is calculated, that is, no actual photolithographic alignment action is performed.

[0082] The second alignment offset information includes a number of second offset information, and each first photolithographic alignment mark also corresponds to one second offset information.

[0083] Similar to the first alignment offset information, in the second alignment offset information, each second offset information corresponds to one first mask alignment mark. Thus, each first photolithographic alignment mark corresponds to one second offset information.

[0084] Moreover, similarly, the method for obtaining the second offset information includes: according to the second alignment model and the detection position information of one first alignment mark, obtaining one second offset information corresponding to the first photolithographic alignment mark corresponding to this first alignment mark.

[0085] In another embodiment, the second alignment model is a linear alignment model, and the radial basis function alignment model is an alignment model that performs quadratic compensation based on the linear alignment model. Therefore, when performing photolithographic alignment based on the radial basis function alignment model, the alignment offset information obtained when performing photolithographic alignment based on the linear alignment model can be directly obtained. That is, in another embodiment, when performing photolithographic alignment on the initial first photolithographic layer according to the first alignment model, the second alignment offset information can also be obtained.

[0086] In this embodiment, the alignment model deviation information includes: the model offset information of each first photolithographic alignment mark, and the model offset information is the deviation information between one first offset information corresponding to this first photolithographic alignment mark and one second offset information.

[0087] The alignment model deviation information reflects the model deviation between the first alignment model and the second alignment model, that is, it reflects the deviation between the offset information of each first mask alignment mark calculated by the first alignment model and the second alignment model based on the same first detection alignment information.

[0088] Please refer to Figure 4 , pattern the initial first layer 210 according to the first photolithographic layer 221 to form the first layer 211.

[0089] In this embodiment, after forming the first layer 211, the first photolithographic layer 221 is removed.

[0090] Please refer to Figure 5, an initial second layer 230 is formed on the first layer 211; after the initial second layer 230 is formed, an initial second photolithography layer 240 is formed on the initial second layer 230.

[0091] The initial second photolithography layer 240 is used to form a second photolithography layer subsequently, and the second photolithography layer is used to pattern the initial second layer 230 to form a second layer.

[0092] In other embodiments, the initial second layer is not formed. The second photolithography layer is used to pattern the first layer to form a second layer.

[0093] Please continue to refer to Figure 5 , according to the second alignment model, initial photolithography alignment is performed on the initial second photolithography layer 240.

[0094] Since photolithography alignment is performed on the initial first photolithography layer 220 according to the first alignment model, and initial photolithography alignment is performed on the initial second photolithography layer 240 according to the second alignment model, therefore, during the formation of the first layer 211 and the second layer, different alignment models can be respectively adopted according to the different situations of the wafer to be compensated before and after the formation of the first layer 211. Thus, by switching the alignment model, the applicability of each alignment model to the wafer in different situations in photolithography alignment is improved, so as to improve the accuracy and reliability of photolithography alignment.

[0095] Specifically, when the risk of misdetection of the detection position information of the first alignment mark is relatively small, a radial basis function alignment model can be adopted. Since the radial basis function alignment model has high sensitivity, the accuracy of photolithography alignment is better. After some special process steps, such as high-temperature steps, planarization steps, etc., since the risk of misdetection of the detection position information of the first alignment mark is relatively large, a high-order alignment model or a linear alignment model can be adopted to reduce the risk of incorrect compensation of the radial basis function alignment model in photolithography alignment, thereby improving the accuracy and reliability of photolithography alignment.

[0096] In this embodiment, in the initial photolithography alignment, the initial second photolithography layer 240 is aligned with the first alignment layer 110.

[0097] In other embodiments, the wafer to be compensated further includes a second alignment layer, and in the initial photolithography alignment, the initial second photolithography layer is aligned with the second alignment layer.

[0098] Please refer to Figure 6 , after initial photolithography alignment is performed on the initial second photolithography layer 240, compensation photolithography alignment is performed on the initial second photolithography layer 240 according to the alignment model deviation information to pattern the initial second photolithography layer 240 to form a second photolithography layer 241.

[0099] Since alignment model deviation information is obtained based on the first alignment model and the second alignment model, and after initial lithography alignment of the initial second photolithography layer 240, compensation lithography alignment is performed on the initial second photolithography layer 240 according to the alignment model deviation information, therefore, after the initial lithography alignment of the initial second photolithography layer 240, the model deviation between the first alignment model and the second alignment model can be compensated in the compensation lithography alignment, thereby improving the stability of the lithography process between the first layer 211 and the second layer.

[0100] It can be seen therefrom that through the compensation method of the alignment deviation, the accuracy and reliability of lithography alignment are improved in the lithography process, and the stability of lithography alignment is improved, thereby enabling high overlay accuracy of the wafer to be compensated.

[0101] In this embodiment, the method for the compensation lithography alignment further includes: providing an overlay deviation compensation model; obtaining a compensation alignment model according to the alignment model deviation information and the overlay deviation compensation model; and performing compensation lithography alignment on the initial second photolithography layer according to the compensation alignment model.

[0102] Specifically, in this embodiment, an overlay deviation compensation model for compensating the overlay deviation of the second layer is further provided. Then, a total optimized model (compensation alignment model) is formed according to the alignment model deviation information and the overlay deviation compensation model, and after the initial lithography alignment, compensation lithography alignment is performed on the initial second photolithography layer 240 according to the compensation alignment model.

[0103] Thus, according to the compensation alignment model, not only the overlay deviation is compensated, but also the model deviation is compensated.

[0104] In other embodiments, the method for the compensation lithography alignment further includes: providing an overlay deviation compensation model; and before patterning the initial second photolithography layer, performing compensation lithography alignment on the initial second photolithography layer according to the overlay deviation compensation model. Specifically, in other embodiments, parameters for lithography alignment are respectively obtained according to the alignment model deviation information and the overlay deviation compensation model, and compensation lithography alignment is performed according to the respectively obtained parameters. It should be noted that the initial lithography alignment and the compensation lithography alignment may be separate actual lithography alignment actions, or after obtaining the total parameters for lithography alignment according to the parameters for the initial lithography alignment and the parameters for the compensation lithography alignment, only one actual lithography alignment action is performed.

[0105] In this embodiment, the first layer 211 includes a plurality of first layer alignment marks (not shown), and each first layer alignment mark corresponds to one first alignment mark and one first photolithographic alignment mark. To meet the design requirements, the overlay accuracy between the second layer and the first layer 211 needs to be within a preset overlay accuracy range. Therefore, the first layer 211 and the first alignment layer 110 are also detected to obtain the alignment deviation information between the first layer 211 and the first alignment layer 110. The alignment deviation information between the first layer 211 and the first alignment layer 110 includes: a plurality of first alignment deviations, and one first alignment deviation is the position deviation information between the corresponding first layer alignment mark and the first alignment mark. Moreover, after the initial photolithographic alignment and before the feed-forward compensation photolithographic alignment, the initial second photolithographic layer 240 is subjected to feed-forward compensation photolithographic alignment according to the alignment deviation information between the first layer 211 and the first alignment layer 110, so as to achieve the alignment between the initial second photolithographic layer 240 and the first layer 211.

[0106] It should be noted that the initial photolithographic alignment, the feed-forward compensation photolithographic alignment, and the compensation photolithographic alignment can be actual photolithographic alignment actions performed separately, or after obtaining the total parameters for photolithographic alignment according to the parameters for the initial photolithographic alignment, the parameters for the feed-forward compensation photolithographic alignment, and the parameters for the compensation photolithographic alignment, only one actual photolithographic alignment action is performed.

[0107] Please refer to Figure 7 , after forming the second photolithographic layer 241, the initial second layer 230 is patterned according to the second photolithographic layer 241 to form the second layer 231.

[0108] In other embodiments, the initial second layer is not formed, and the first layer is patterned according to the second photolithographic layer to form the second layer.

[0109] In another embodiment, the compensation method for the alignment deviation is also adjusted according to wafers to be compensated in different batches. Specifically, the compensation method for the alignment deviation further includes: providing a target overlay accuracy range for the second layer; providing a preset overlay accuracy range; providing an overlay safety factor X; providing n batches of wafers to be compensated, where n is a natural number; obtaining the overlay accuracy Y of the second layer of each wafer to be compensated in the k-th batch to the (k+i)-th batch among the n batches, where k is a natural number, i is a natural number, and k+i is within the range of n; when the target overlay accuracy range of the second layer is within the preset overlay accuracy range, and any (Y±X) is within the target overlay accuracy range of the second layer, obtaining batch alignment model deviation information according to one of the several alignment model deviation information of at least one batch among the first batch to the (k+i)-th batch; compensating and lithographically aligning the initial second photolithography layer of the wafer to be compensated in the (k+i+1)-th batch according to the batch alignment model deviation information, so as to pattern the initial second photolithography layer of the wafer to be compensated in the (k+i+1)-th batch and form the second photolithography layer of the wafer to be compensated in the (k+i+1)-th batch.

[0110] Since the target overlay accuracy range of the second layer is within the preset overlay accuracy range, and any (Y±X) is within the target overlay accuracy range of the second layer, the overlay accuracy stability of the second layer of the wafer to be compensated in the (k+i)-th batch is high. On this basis, since batch alignment model deviation information is obtained according to one of the several alignment model deviation information of at least one batch among the first batch to the (k+i)-th batch, and the initial second photolithography layer of the wafer to be compensated in the (k+i+1)-th batch is compensated and lithographically aligned according to the batch alignment model deviation information, so as to pattern the initial second photolithography layer of the wafer to be compensated in the (k+i+1)-th batch and form the second photolithography layer of the wafer to be compensated in the (k+i+1)-th batch, one alignment model deviation information (batch alignment model deviation information) obtained according to historical processing experience is used to replace the alignment model deviation information corresponding to each wafer to be compensated in the (k+i+1)-th batch, so as to perform compensated lithographic alignment on the wafer to be compensated in the (k+i+1)-th batch. Thus, on the basis of ensuring that the overlay accuracy of the formed second photolithography layer is relatively stable after compensated lithographic alignment, the efficiency of compensating the alignment deviation of the wafers to be compensated in each batch after the (k+i)-th batch is improved.

[0111] Specifically, in another embodiment, the batch alignment model deviation information may be the average of the alignment model deviation information of multiple wafers to be compensated in any one or more batches among the first batch to the k+i-th batch. Alternatively, the batch alignment model deviation information may also be the largest one among the alignment model deviation information of multiple wafers to be compensated in any one or more batches among the first batch to the k+i-th batch.

[0112] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be subject to the scope defined by the claims.

Claims

1. A method for compensating alignment deviation, characterized in that, Including: Providing a wafer to be compensated, the wafer to be compensated including an initial first layer and an initial first photolithography layer located on the initial first layer; Providing a first alignment model and a second alignment model; Performing photolithography alignment on the initial first photolithography layer according to the first alignment model to pattern the initial first photolithography layer and form a first photolithography layer; Patterning the initial first layer according to the first photolithography layer to form a first layer; Forming an initial second photolithography layer on the first layer; Performing initial photolithography alignment on the initial second photolithography layer according to the second alignment model; Obtaining alignment model deviation information according to the first alignment model and the second alignment model; After performing initial photolithography alignment on the initial second photolithography layer, performing compensated photolithography alignment on the initial second photolithography layer according to the alignment model deviation information to pattern the initial second photolithography layer and form a second photolithography layer; Wherein, the wafer to be compensated further includes: a first alignment layer, the initial first layer is located on the first alignment layer, and when performing photolithography alignment on the initial first photolithography layer, it aligns with the first alignment layer; The method for obtaining alignment model deviation information according to the first alignment model and the second alignment model includes: providing first preset alignment position information of the first alignment layer; obtaining first detected alignment position information according to the first preset alignment position information and the first alignment layer; obtaining first alignment offset information according to the first detected alignment position information and the first alignment model; obtaining second alignment offset information according to the first detected alignment position information and the second alignment model; obtaining alignment model deviation information according to the first alignment offset information and the second alignment offset information.

2. The compensation method for alignment deviation according to claim 1, wherein When performing photolithography alignment on the initial first photolithography layer according to the first alignment model, obtaining the first alignment offset information.

3. The compensation method for alignment deviation according to claim 1, characterized in that When performing photolithography alignment on the initial first photolithography layer according to the first alignment model, obtaining the second alignment offset information.

4. The compensation method for alignment deviation according to claim 1, characterized in that, Before patterning the initial first photolithography layer to form a first photolithography layer, performing photolithography alignment on the initial first photolithography layer according to the second alignment model to obtain the second alignment offset information.

5. The compensation method for alignment deviation according to claim 1, characterized in that, The first alignment layer includes a plurality of first alignment marks, and the first preset alignment position information includes preset position information of each first alignment mark; the first photolithography layer includes a plurality of first photolithography alignment marks, and each first photolithography alignment mark corresponds to one first alignment mark.

6. The compensation method for alignment deviation according to claim 5, wherein The first alignment offset information includes a plurality of first offset information, and each first photolithography alignment mark corresponds to one first offset information; the second alignment offset information includes a plurality of second offset information, and each first photolithography alignment mark also corresponds to one second offset information; the alignment model deviation information includes: model offset information of each first photolithography alignment mark, and the model offset information is the deviation information between one first offset information and one second offset information corresponding to the first photolithography alignment mark.

7. The compensation method for alignment deviation according to claim 6, characterized in that, The method for obtaining the first detection alignment position information includes: detecting the positions of each corresponding first alignment mark in the first alignment layer according to the preset position information of each of the first alignment marks, so as to obtain the detection position information of each first alignment mark, and the first detection alignment position information includes the detection position information of each of the first alignment marks.

8. The compensation method for alignment deviation according to claim 7, characterized in that, The method for obtaining the first offset information includes: obtaining one first offset information corresponding to one first lithography alignment mark corresponding to the first alignment mark according to the first alignment model, the preset position information of one first alignment mark, and the detection position information of the first alignment mark.

9. The compensation method for alignment deviation according to claim 7, characterized in that, The method for obtaining the second offset information includes: obtaining one second offset information corresponding to the first lithography alignment mark corresponding to the first alignment mark according to the second alignment model, the preset position information of one first alignment mark, and the detection position information of the first alignment mark.

10. The compensation method for alignment deviation according to claim 1, characterized in that, The method for compensating lithography alignment further includes: providing an overlay error compensation model; before patterning the initial second lithography layer, further performing compensating lithography alignment on the initial second lithography layer according to the overlay error compensation model.

11. The compensation method for alignment deviation according to claim 1, characterized in that, The method for compensating lithography alignment further includes: providing an overlay error compensation model; obtaining a compensation alignment model according to the alignment model deviation information and the overlay error compensation model; performing compensating lithography alignment on the initial second lithography layer according to the compensation alignment model.

12. The compensation method for alignment deviation according to claim 1, characterized in that, It further includes: Before forming the initial second lithography layer, forming an initial second layer on the first layer; After forming the second lithography layer, patterning the initial second layer according to the second lithography layer to form the second layer.

13. The compensation method for alignment deviation according to claim 1, wherein It further includes: Patterning the first layer according to the second lithography layer to form the second layer.

14. The method for compensating alignment deviation according to claim 12 or 13, characterized in that, It further includes: Providing the target overlay accuracy range of the second layer; Providing a preset overlay accuracy range; Providing an overlay safety factor X; Providing n batches of wafers to be compensated, where n is a natural number; obtaining the overlay accuracy Y of the second layer of each wafer to be compensated in the k-th batch to the (k + i)-th batch among the n batches, where k is a natural number, i is a natural number, and k + i is within the range of n; when the target overlay accuracy range of the second layer is within the preset overlay accuracy range, and any (Y ± X) is within the target overlay accuracy range of the second layer, obtaining batch alignment model deviation information according to one of the alignment model deviation information of at least one batch among the first batch to the (k + i)-th batch; performing compensating lithography alignment on the initial second lithography layer of the wafer to be compensated in the (k + i + 1)-th batch according to the batch alignment model deviation information to pattern the initial second lithography layer of the wafer to be compensated in the (k + i + 1)-th batch and form the second lithography layer of the wafer to be compensated in the (k + i + 1)-th batch.

15. The compensation method for alignment deviation according to claim 1, wherein The first alignment model is a radial basis function alignment model.

16. The compensation method for alignment deviation according to claim 1, characterized in that, The second alignment model is a linear alignment model or a high-order alignment model.

Images