Methods for compensating high-order overlay errors

Abstract

The present invention discloses a method for compensating for high-order overlay errors, comprising the following steps: Step 1, calculating a first overlay error introduced by each process step of an exposed wafer; Step 2, stripping off the wafer-level linear overlay error in the first overlay error and forming a second overlay error at the exposure area level; Step 3, correcting the mask writing according to the second overlay error and performing mask writing to obtain a corrected mask; Step 4, using the corrected mask to expose the wafer to compensate for the high-order overlay error. The present invention can compensate for high-order overlay errors and can ensure that the overlay errors caused during the process are fully compensated.

Description

Technical Field

[0001] The present invention relates to a method for manufacturing a semiconductor integrated circuit, and in particular to a method for compensating for high-order overlay error (OVL). Background Art

[0002] As the size of semiconductor devices decreases, the requirements for the accuracy of overlay error are becoming increasingly higher. The size of overlay error plays a key role in device performance. Changes in overlay error will significantly affect the electrical performance and product yield of the device.

[0003] To improve device performance and yield, a series of efforts have been made to compensate for overlay accuracy. Significant progress has been made in improving overlay accuracy. However, due to the limitations of existing technology, it is not yet possible to compensate for all parameters, especially high-order OVL parameters.

[0004] At advanced technology nodes, overlay error requirements are becoming increasingly stringent. Currently, the industry's common approach is to use advanced compensation software provided by exposure machine manufacturers like ASML, or to employ more advanced exposure machines or even EUV to reduce overlay error. However, even advanced compensation methods cannot guarantee that overlay errors caused during the process can be fully compensated.

[0005] like Figure 1 , which are various parameters in the overlay error compensation process in the existing photolithography process; Figure 1 The figure shows 20 parameters of overlay error from k1 to k20, which are also 20 categories of overlay error. The arrow line 101 represents the overlay error, including the offset direction and magnitude corresponding to the overlay error. k1 and k2 are linear parameters, and k3 to k20 are nonlinear parameters. Figure 1 All overlay error parameters in are compensated.

[0006] like Figure 2 Figure 2 shows the variation in overlay error after various process steps in conventional photolithography. Pre-layer overlay error graph 201a forms a modified overlay error graph 201b after process step 202. Process step 202 includes etching, thin film (TF) growth, chemical mechanical polishing (CMP), wet etching, and annealing. Comparing overlay error graphs 201a and 201b reveals that process step 202 causes unique variations in both the direction and magnitude of the overlay error. Existing overlay error compensation methods cannot guarantee complete compensation for the overlay error introduced during the process. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method for compensating high-order overlay errors, which can compensate for high-order overlay errors and ensure that the overlay errors caused in the process are fully compensated.

[0008] To solve the above technical problems, the present invention provides a method for compensating high-order overlay errors, comprising the following steps:

[0009] Step 1: Calculate the first overlay error introduced by each process step of the exposed wafer.

[0010] Step 2: Stripping the wafer-level linear overlay error from the first overlay error to form a second overlay error at the shot level.

[0011] Step three: correcting the mask writing according to the second overlay error and performing the mask writing to obtain a corrected mask.

[0012] Step 4: Expose the wafer using the corrected mask to compensate for high-order overlay errors.

[0013] A further improvement is that step one includes the following sub-steps:

[0014] The wafer is exposed using the photomask before correction.

[0015] The exposed wafer is measured to obtain overlay error data of each layer.

[0016] The first overlay error introduced by each process step of the wafer is calculated by performing software fitting on the overlay error data of each layer.

[0017] A further improvement is that step 2 further includes: calculating a characteristic distribution value (fingerprint) according to the second overlay error.

[0018] A further improvement is that step three includes the following sub-steps:

[0019] The feature distribution values ​​are imported into direct writing software.

[0020] The mask writing is performed under the control of the direct writing software.

[0021] A further improvement is that after step 4 is completed, iterative operations of steps 1 to 4 are repeated.

[0022] Further improvements include:

[0023] A first database is established through the iterative operation.

[0024] A relationship model between each of the process steps and the overlay error variation is established through the first database.

[0025] Further improvements include:

[0026] The relational model is introduced into the direct write software.

[0027] The direct writing software obtains the correction amount in the mask writing according to the relationship model and the process and performs the mask writing.

[0028] A further improvement is that, in processes with the same process platform and stable process steps, the same relationship model is used to perform mask correction.

[0029] A further improvement is that the data in the first database includes the first overlay error, the second overlay error or the characteristic distribution value of each section of the process obtained in each iterative operation.

[0030] The present invention pre-measures the first overlay error introduced in each process section of the wafer after exposure, and then eliminates the linear overlay error at the wafer level from the first overlay error. The nonlinear overlay error of the remaining exposure area constitutes the second overlay error. Since the exposure area is the area where the mask is exposed once, the second overlay error can be used to directly rewrite the mask to achieve correction of the mask. For example, an accurate fingerprint can be obtained through the second overlay error, and the fingerprint can be directly introduced into the mask direct writing machine to achieve correction of the mask during the mask production process. The corrected mask can be used to correct high-order terms of the overlay error. Finally, various compensation parameters of the overlay error can be corrected from the source, and the overlay error introduced by the process can be eliminated. Therefore, the present invention can compensate for high-order overlay errors and ensure that the overlay errors caused in the process are fully compensated. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0032] Figure 1 These are the various parameters in the overlay error compensation process in the existing photolithography process;

[0033] Figure 2 It is the change of overlay error after each process in the existing photolithography process;

[0034] Figure 3 is a flow chart of a method for compensating for high-order overlay errors according to an embodiment of the present invention;

[0035] Figure 4 2 is a schematic diagram of the implementation steps of the method for compensating for high-order overlay errors according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] like Figure 3 FIG. 1 is a flow chart of a method for compensating for high-order overlay errors according to an embodiment of the present invention; FIG. Figure 4 FIG. 1 is a schematic diagram of the implementation of each step of the method for compensating for high-order overlay errors according to an embodiment of the present invention. The method for compensating for high-order overlay errors according to an embodiment of the present invention includes the following steps:

[0037] Step 1: Calculate the first overlay error introduced by each process step of the exposed wafer.

[0038] In the embodiment of the present invention, step 1 includes the following sub-steps:

[0039] The wafer is exposed using the photomask before correction.

[0040] like Figure 4 As shown, the wafer after exposure is measured to obtain the overlay error data of each layer.

[0041] The method includes measuring the pre-layer 401 of the wafer after exposure to obtain an overlay error graph 301 a corresponding to the overlay error data.

[0042] Then, the process 402 is performed. To help understand the process 402 in a more vivid way, the photo corresponding to the mark 302 shows the pattern structure on the wafer and the various processes it has undergone.

[0043] After the process 402 is completed, an overlay error data (OVL Data) measurement step 403 is performed. The measurement step 403 generates a corresponding overlay error map 301b.

[0044] The first overlay error introduced by each process step of the wafer is calculated by software fitting the overlay error data of each layer. The first overlay error corresponding to each characteristic process step can be fitted through the overlay error graphs 301a and 301b.

[0045] Step 2: Stripping the wafer-level linear overlay error from the first overlay error to form a second overlay error at the exposure area level.

[0046] The wafer-level overlay error is usually a linear overlay error, such as Figure 1 The errors corresponding to the parameters k1 and k2 in .

[0047] The exposure area level error is usually a nonlinear error, which forms the second overlay error. The exposure area, or shot, is the area exposed once by the mask, so the pattern of the exposure area corresponds to the pattern on the mask.

[0048] In an embodiment of the present invention, the method further includes: calculating a characteristic distribution value according to the second overlay error.

[0049] Step 3: Figure 4 As shown, the mask writing 404 is corrected according to the second overlay error and the mask writing 404 is performed to obtain a corrected mask.

[0050] In the embodiment of the present invention, step three includes the following sub-steps:

[0051] The feature distribution values ​​are imported into direct writing software.

[0052] The mask writing 404 is performed under the control of the direct writing software.

[0053] like Figure 4 As shown, the mask writing 404 is implemented by a mask direct writing machine 303. The mask direct writing machine 303 is controlled by a computer 304. The direct writing software of the computer 304 has the mask layout. After the feature distribution is imported into the direct writing software, the mask layout will be changed.

[0054] Step 4: Expose the wafer using the corrected mask 405 to compensate for high-order overlay errors.

[0055] like Figure 4 As shown, the exposure 405 is achieved using a photolithography machine 306 .

[0056] After the exposure 405 , a final overlay error (Final OVL) measurement is performed to obtain an overlay error graph 301 c . Compared with the overlay error graph 301 b , the embodiment of the present invention can substantially eliminate the influence of the process 402 on the overlay error.

[0057] In some embodiments, the following also applies:

[0058] After step 4 is completed, the iterative operation of steps 1 to 4 is repeated.

[0059] A first database is established through the iterative operation. Preferably, the data in the first database includes the first overlay error, the second overlay error or the characteristic distribution value of each section of the process obtained in each iterative operation.

[0060] A relationship model between each of the process steps and the overlay error variation is established through the first database.

[0061] The relational model is introduced into the direct write software.

[0062] The direct writing software obtains the correction amount in the mask writing 404 according to the relationship model and the process flow and performs the mask writing 404 .

[0063] In the process of the same process platform and the process is stable, the same relationship model is used to perform mask correction. That is, under the condition that the relationship model is known, when performing the overlay error compensation related to the process, there is no need to repeat steps 1 to 2. Figure 4 Instead of steps 401 to 403 in the above method, the characteristic distribution value related to the process is directly obtained according to the relationship model and the process, and then steps 3 and 4 are directly performed. Figure 4 Therefore, the embodiment of the present invention has a certain universality for the same process platform and stable processes.

[0064] The embodiment of the present invention pre-measures the first overlay error introduced in each process section of the wafer after exposure, and then eliminates the wafer-level linear overlay error from the first overlay error. The nonlinear overlay error of the remaining exposure area constitutes the second overlay error. Since the exposure area is the area where the mask is exposed once, the second overlay error can be used to directly rewrite the mask to achieve correction of the mask. For example, an accurate fingerprint can be obtained through the second overlay error, and the fingerprint can be directly introduced into the mask direct writing machine to achieve correction of the mask during the mask production process. The corrected mask can be used to correct the high-order terms of the overlay error. Finally, the various compensation parameters of the overlay error can be corrected from the source, and the overlay error introduced by the process can be eliminated. Therefore, the embodiment of the present invention can compensate for high-order overlay errors and ensure that the overlay errors caused in the process are fully compensated.

[0065] The present invention has been described in detail above by means of specific embodiments, but these do not constitute limitations of the present invention. Without departing from the principles of the present invention, those skilled in the art may make many variations and improvements, which should also be considered as the scope of protection of the present invention.

Claims

1. A method for compensating high-order overlay errors, characterized in that: The steps include: Step 1: Calculate the first overlay error introduced by each process step of the exposed wafer; Step 2: removing the wafer-level linear overlay error from the first overlay error and forming a second overlay error at the exposure area level; Step 3, correcting the mask writing according to the second overlay error and performing the mask writing to obtain a corrected mask; Step 4: Expose the wafer using the corrected mask to compensate for high-order overlay errors.

2. The method for compensating high-order overlay error according to claim 1, wherein: Step 1 includes the following sub-steps: exposing the wafer using the photomask before correction; Measuring the wafer after exposure to obtain overlay error data of each layer; The first overlay error introduced by each process step of the wafer is calculated by performing software fitting on the overlay error data of each layer.

3. The method for compensating high-order overlay error according to claim 1, wherein: Step 2 further includes: calculating a characteristic distribution value based on the second overlay error.

4. The method for compensating high-order overlay error according to claim 1, 2 or 3, wherein: Step 3 includes the following sub-steps: Importing the characteristic distribution value into direct writing software; The mask writing is performed under the control of the direct writing software.

5. The method for compensating high-order overlay error according to claim 4, wherein: After step 4 is completed, the iterative operation of steps 1 to 4 is repeated.

6. The method for compensating high-order overlay error according to claim 5, wherein: Also includes: Establishing a first database through the iterative operation; A relationship model between each of the process steps and the overlay error variation is established through the first database.

7. The method for compensating high-order overlay error according to claim 6, wherein: Also includes: Introducing the relational model into direct write software; The direct writing software obtains the correction amount in the mask writing according to the relationship model and the process and performs the mask writing.

8. The method for compensating high-order overlay error according to claim 6, wherein: In processes with the same process platform and stable process steps, the same relationship model is used to perform mask correction.

9. The method for compensating high-order overlay error according to claim 7, wherein: The data in the first database include the first overlay error, the second overlay error or the characteristic distribution value of each section of the process obtained in each iterative calculation.

Images