Graph correction method

Abstract

A method for optical proximity correction, comprising: providing a layout to be corrected; obtaining an original light intensity distribution corresponding to the layout to be corrected; obtaining a diffusion distribution corresponding to the layout to be corrected; processing the original light intensity distribution and the diffusion distribution by using a free convolution model to obtain an effective light intensity distribution; performing simulation exposure by using an effective light intensity greater than an effective light intensity threshold to obtain a simulation exposure layout; obtaining a lithography layout according to the simulation exposure layout; and obtaining an average offset according to sizes of the lithography layout and the simulation exposure layout at the same position. The method reduces the difference between the prediction ability of the optical proximity correction model and the true value.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor, in particular to a pattern correction method. BACKGROUND

[0002] Optical proximity correction (OPC) adds a calculated bias to a target to maximize fidelity of a pattern on a wafer. An effective model for OPC operation relies on an accurate OPC model that is constructed by optimizing the form of the fitting coefficients.

[0003] We need to continually optimize the OPC model to improve the accuracy of the pattern on the wafer. SUMMARY

[0004] The present application solves the technical problem of providing a pattern correction method to improve the accuracy of the pattern on the wafer.

[0005] To solve the above technical problems, the technical scheme of the present application provides a pattern correction method, comprising: providing a to-be-corrected layout; obtaining an original light intensity distribution corresponding to the to-be-corrected layout; obtaining a diffusion distribution corresponding to the to-be-corrected layout; processing the original light intensity distribution and the diffusion distribution by using a free convolution model to obtain an effective light intensity distribution; simulating exposure by using the effective light intensity greater than an effective light intensity threshold to obtain a simulated exposure layout; obtaining a lithography layout according to the simulated exposure layout; and obtaining an average offset according to the size of the lithography layout and the simulated exposure layout at the same position.

[0006] Optionally, the diffusion distribution includes a photoacid distribution and a photobase distribution.

[0007] Optionally, the original light intensity distribution, the photoacid distribution and the photobase distribution are processed according to the free convolution model to obtain the effective light intensity distribution. wherein I effective is the effective light intensity distribution, I is the original light intensity distribution, I -b is the photoacid distribution, I +b is the photobase distribution, C0, C1, …, C 15 are to-be-optimized coefficients, b is a to-be-optimized parameter, and FMK is a free convolution model. is a variable in the free convolution model, and k has a value range of 1 to 13 and i has a value range of a natural number of 1 to N.

[0008] Optionally, the free convolution model FMK is an m×m matrix, the number of N is m×m×13, and m is a natural number greater than 1.

[0009] Optionally, the relationship between the t+1th variable adjustment of the free convolution model and the tth variable adjustment is wherein is the t+1th variable adjustment, is the tth variable adjustment, DRMS is the average offset convergence step, is the finite difference, i ranges from 1 to N, and k ranges from 1 to 13.

[0010] Optionally, the method for obtaining the average offset comprises: providing the coefficient values of C0, C1, …, C 15 ; adjusting the parameters b and the variables to obtain the effective light intensity; providing an effective light intensity threshold; performing simulation exposure with the effective light intensity greater than the effective light intensity threshold to obtain a simulation exposure pattern; obtaining a lithography pattern according to the simulation exposure pattern; and obtaining the average offset according to the sizes of the lithography pattern and the simulation exposure pattern at the same position.

[0011] Optionally, the method for obtaining the average offset comprises: providing the parameter values of C0, C1, …, C ; providing the parameter b; adjusting the coefficients C0, C1, …, C 15 to obtain the effective light intensity; providing an effective light intensity threshold; performing simulation exposure with the effective light intensity greater than the effective light intensity threshold to obtain a simulation exposure pattern; obtaining a lithography pattern according to the simulation exposure pattern; and obtaining the average offset according to the sizes of the lithography pattern and the simulation exposure pattern at the same position.

[0012] Optionally, the method for obtaining the average offset comprises: providing the coefficient values of C0, C1, …, C 15 ; providing the variable values of C0, C1, …, C ; providing the parameter b; adjusting the effective light intensity threshold to obtain an effective light intensity distribution; performing simulation exposure with the effective light intensity greater than the effective light intensity threshold to obtain a simulation exposure pattern; obtaining a lithography pattern according to the simulation exposure pattern; and obtaining the average offset according to the sizes of the lithography pattern and the simulation exposure pattern at the same position.

[0013] Optionally, the average offset is obtained according to the sizes of the lithography pattern and the simulation exposure pattern at the same position wherein, RMS is the average offset, CD i,s is the critical dimension of the simulation exposure pattern obtained at the ith variable adjustment, CD i,w is the critical dimension of the lithography pattern at the same position obtained at the ith variable adjustment, and wi is the weight related to the ith variable adjustment, and i ranges from 1 to N.

[0014] Optionally, the method for obtaining a lithography pattern according to a simulation exposure pattern comprises: providing a target layout; obtaining an edge placement error between the simulation exposure layout and the target layout; correcting the to-be-corrected layout according to the edge placement error to obtain a corrected layout; taking the corrected layout as a mask layout; and obtaining a lithography layout according to the mask layout.

[0015] Compared with the prior art, the technical scheme of the present application has the following beneficial effects:

[0016] In the technical scheme of the present application, the original light intensity distribution, the photoacid distribution and the photobase distribution are processed by using a free convolution model to obtain an effective light intensity distribution, the adjustability and flexibility of the free convolution model are high, the optical proximity correction model is not limited by any functional form, the number of adjustable parameters of the free convolution model is large, so that more diffusion mechanisms can be explored in the photoresist stage, and the average deviation of the final model is reduced, and the gap between the prediction ability of the optical proximity correction model and the true value is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a flowchart of a pattern correction method in an embodiment of the present application.

[0018] Figure 2 is a schematic diagram of an effective light intensity distribution. DETAILED DESCRIPTION

[0019] As described in the background, we need to constantly optimize the model of optical proximity correction to improve the accuracy of the pattern on the wafer.

[0020] Specifically, the traditional optical proximity correction model uses a Gaussian function as a convolution kernel to simulate the diffusion process in the exposed photoresist. We can certainly adjust the deviation related to the Gaussian to describe the speed of diffusion. However, changing the deviation does not really change the form of the function, it is still a Gaussian function, which cannot capture physical parameters other than standard diffusion, and in some cases, the fitting ability of the traditional model may not be enough.

[0021] In the technical scheme of the present application, a free-style model form is proposed, which is not limited by any functional form, the number of adjustable parameters of the convolution kernel is large, which allows us to explore more diffusion mechanisms in the photoresist stage, and the average deviation of the final model is reduced by 10%, so that the gap between the prediction ability of the model and the true value is reduced.

[0022] In order to make the above-mentioned purposes, characteristics and beneficial effects of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings.

[0023] Figure 1 is a flowchart of a pattern correction method in an embodiment of the present application.

[0024] Please refer to Figure 1 , the pattern correction method comprises:

[0025] Step S10: providing a to-be-corrected layout;

[0026] Step S20: obtaining an original light intensity distribution corresponding to the to-be-corrected layout;

[0027] Step S30: obtaining a diffusion distribution corresponding to the to-be-corrected layout;

[0028] Step S40: processing the original light intensity distribution and the diffusion distribution by using a free convolution model to obtain an effective light intensity distribution;

[0029] Step S50: simulating exposure by using an effective light intensity greater than an effective light intensity threshold to obtain a simulated exposure layout;

[0030] Step S60: obtaining a lithography layout according to the simulated exposure layout;

[0031] Step S70: obtaining an average offset according to the size of the lithography layout and the simulated exposure layout at the same position.

[0032] The method processes the original light intensity distribution, the light acid distribution and the light base distribution by using the free convolution model to obtain the effective light intensity distribution, the adjustability and flexibility of the free convolution model are higher, so that the optical proximity correction model is not limited by any functional form, and the number of adjustable parameters of the free convolution model is larger, so that more diffusion mechanisms can be explored in the photoresist stage, and the average offset of the final model is reduced, so that the prediction ability of the optical proximity correction model is closer to the true value.

[0033] Next, each step is analyzed and described.

[0034] Please continue to refer to Figure 1 , step S10 is performed: providing a to-be-corrected layout.

[0035] The to-be-corrected layout is a layout of a mask plate without optical correction.

[0036] Please continue to refer to Figure 1 , step S20 is performed: obtaining an original light intensity distribution I corresponding to the to-be-corrected layout.

[0037] The method for obtaining the original light intensity distribution I corresponding to the image to be corrected includes: performing simulated exposure on the image to be corrected to obtain a simulated exposure image; obtaining the distribution of light intensity I within the simulated exposure image; establishing a coordinate system, wherein the coordinate system is used to define the relationship between each position of the simulated exposure image and the light intensity within the simulated exposure image at the corresponding position; and obtaining the functional relationship between the light intensity within the simulated exposure image and each position within the simulated exposure image.

[0038] Please continue to refer to this. Figure 1 Step S30: Obtain the diffusion distribution corresponding to the layout to be corrected.

[0039] In this embodiment, the diffusion distribution includes photoacid distribution I. -b Distribution of light-induced alkali I +b .

[0040] The photoacid distribution I -b The light intensity distribution is defined as light intensity greater than the threshold b.

[0041] Obtain the photoacid distribution I corresponding to the layout to be corrected. -b The method includes: obtaining the distribution of light intensity I within the simulated exposure pattern; taking a light intensity greater than a threshold b; performing simulated exposure with a light intensity greater than the threshold b to obtain the simulated exposure pattern; and obtaining the light intensity I within the simulated exposure pattern. -b The distribution of the simulated exposure pattern; establishing a coordinate system, which is used to define the coordinates of each position in the simulated exposure pattern and the corresponding position of the photoacid I in the simulated exposure pattern. -b The relationship; obtaining photoacid I within the simulated exposure pattern -b The functional relationship between the simulated exposure pattern and each position within the pattern.

[0042] The light alkali distribution I +b For light intensity distribution where the light intensity is less than the threshold b, the light intensity distribution I is... +b Photoacid distribution I -b The transpose of .

[0043] Obtain the photoalkali distribution I corresponding to the pattern to be corrected. +b The method includes: obtaining the distribution of light intensity I within the simulated exposure pattern; taking a light intensity less than a threshold b; performing simulated exposure with a light intensity less than the threshold b to obtain the simulated exposure pattern; and obtaining the light intensity I within the simulated exposure pattern. +b The distribution of light alkali I in the simulated exposure pattern; establishing a coordinate system, which is used to define the coordinates of each position in the simulated exposure pattern and the corresponding position within the simulated exposure pattern. +b The relationship; obtaining photoalkali I within the simulated exposure pattern +b The functional relationship between the simulated exposure pattern and each position within the pattern.

[0044] Please continue to refer to this. Figure 1Step S40: Use the free convolution model FMK to process the original light intensity distribution I and the light acid distribution I. +b and light base I -b The distribution is processed to obtain the effective light intensity distribution I. effective .

[0045] Based on the free convolution model FMK, the original light intensity distribution I and the photoacid distribution I are analyzed. +b and light base I -b The distribution is processed to obtain the effective light intensity distribution. , among which, I effective For effective light intensity distribution, I -b For photoacid distribution, I +b I represents the light intensity distribution, C0, C1, ..., C1 represents the light intensity distribution. 15 Let be the coefficients to be optimized, b be the parameters to be optimized, and FMK be a free convolution model. Let k be a variable in the free convolution model, where k ranges from 1 to 13 and i ranges from 1 to N natural numbers.

[0046] The effective light intensity distribution has 16 gradients. When optimizing the model, the RMS value will be searched along the 16 gradients to complete the model optimization process.

[0047] The effective light intensity distribution I effective Distribution diagram as shown Figure 2 As shown.

[0048] In this embodiment, the free convolution model FMK is a two-dimensional function, which can be represented as an m×m matrix. The variables in the free convolution model... The quantity is m×m×13, where m is a natural number greater than 1.

[0049] Obtain the effective light intensity I effective During the process, the free convolution model FMK has m×m×13 parameters that can be adjusted, which can be iterated N times, thereby reducing the average offset of the final model and narrowing the gap between the prediction ability of the optical proximity correction model and the true value.

[0050] The relationship between the (t+1)th variable adjustment and the tth variable adjustment is: in For the (t+1)th variable adjustment, For the t-th variable adjustment, DRMS ​​is the average offset convergence step size. For finite differences, the value of i ranges from 1 to N, and the value of k ranges from 1 to 13.

[0051] In this embodiment, the value of DRMS ​​is less than 0.2. This results in a progressively smaller average offset (RMS) being obtained subsequently, achieving the optimization objective.

[0052] Please continue to refer to this. Figure 1 Execute step S50: with an effective light intensity I greater than the effective light intensity threshold T effective Perform simulated exposure to obtain the simulated exposure map.

[0053] The simulated exposure pattern is the pattern obtained by simulating the exposure of the pattern to be corrected using an optical correction model.

[0054] Please continue to refer to this. Figure 1 Step S60: Obtain the photolithography pattern based on the simulated exposure pattern.

[0055] The method for obtaining a photolithography pattern based on a simulated exposure pattern includes: providing a target pattern; obtaining an edge placement error based on the simulated exposure pattern and the target pattern; correcting the pattern to be corrected based on the edge placement error; after multiple simulated exposure-correction-simulated exposures, obtaining a corrected pattern when the edge placement error is less than a preset value; using the corrected pattern as a mask pattern; transferring the corrected pattern onto photoresist through exposure and development; and obtaining the photolithography pattern based on the mask pattern.

[0056] The corrected pattern is the pattern of the mask after optical correction, and the photolithography pattern is the photoresist pattern that needs to be actually etched.

[0057] The method for obtaining edge placement error based on simulated exposure pattern and target pattern includes: dividing the outline of the simulated exposure pattern into several line segments; calculating the offset between any line segment and the corresponding line segment in the target pattern, which is the edge placement error.

[0058] Please continue to refer to this. Figure 1 Step S70: Obtain the average offset based on the dimensions of the photolithography pattern and the simulated exposure pattern at the same position.

[0059] The average offset is obtained based on the dimensions of the photolithographic pattern and the simulated exposure pattern at the same position. Where RMS is the average offset, and CD is the mean offset. i,s CD is the key dimension of the simulated exposure pattern obtained during the i-th variable adjustment. i,w Let wi be the key dimension at the same position of the lithographic pattern obtained during the i-th variable adjustment, and let wi be the weight related to the i-th variable, with i ranging from 1 to N.

[0060] The smaller the value of the average offset (RMS), the smaller the difference between the prediction ability of the optical proximity correction model and the true value, the better the correction effect, and the more accurate the prediction result.

[0061] The process of obtaining the average offset (RMS) involves selecting the coefficients to be optimized, C0, C1, ..., C. 15 Parameter b to be optimized, variables in the free convolution model The process of iteratively optimizing the effective light intensity threshold T results in a smaller average offset (RMS) after model optimization, which is close to the true value.

[0062] In this embodiment, the method for optimizing the model includes the following iterative process: providing C0, C1, ..., C 15 The coefficient values ​​provide the effective light intensity threshold T; based on the provided C0, C1, ..., C 15 The coefficient values ​​and effective light intensity threshold T are related to parameter b and variable Optimize; provide Provide parameter b based on the provided parameter value; And parameter b with respect to the coefficients C0, C1, ..., C 15 Optimize; provide C0, C1, ..., C 15 The coefficient value provides The variable value is provided, and parameter b is provided; based on the provided C0, C1, ..., C 15 , The variable values ​​and parameter b are used to optimize the effective light intensity threshold T.

[0063] In this embodiment, the parameter b and the variable Optimization methods for obtaining the average offset (RMS) include providing C0, C1, ..., C 15 The coefficient values; for parameter b and variable Multiple adjustments were made to obtain the effective light intensity I. effective Provide an effective light intensity threshold T; with an effective light intensity I greater than the effective light intensity threshold T. effective Perform simulated exposure to obtain a simulated exposure pattern; obtain a photolithography pattern based on the simulated exposure pattern; obtain the average offset (RMS) based on the dimensions of the photolithography pattern and the simulated exposure pattern at the same position.

[0064] This process involves parameter b and variable b. The optimization process employs a gradient algorithm to optimize the parameter b and the variable. Optimize.

[0065] For parameter b and variable Multiple adjustments were made to obtain the effective light intensity I. effective The methods include: for variables The t-th adjustment is performed, and the free convolution model FMK is used to adjust the original light intensity distribution I and the light acid distribution I. +b and light base I-b The distribution is convolved to obtain the effective light intensity I. effective With an effective light intensity I greater than the effective light intensity threshold T effective Perform simulated exposure to obtain a simulated exposure pattern; obtain a photolithography pattern based on the simulated exposure pattern; obtain the average offset (RMS) based on the dimensions of the photolithography pattern and the simulated exposure pattern at the same position; then adjust the variables... The (t+1)th adjustment is performed, and the free convolution model FMK is used to adjust the original light intensity distribution I and the light acid distribution I. +b and light base I -b The distribution is convolved to obtain the effective light intensity I. effective With an effective light intensity I greater than the effective light intensity threshold T effective Perform simulated exposure to obtain a simulated exposure pattern; obtain a photolithography pattern based on the simulated exposure pattern; obtain the average offset RMS based on the dimensions of the photolithography pattern and the simulated exposure pattern at the same position; after N adjustments, the obtained average offset RMS becomes smaller and smaller.

[0066] In this embodiment, the method for obtaining the average offset (RMS) includes: providing The parameter values; providing parameter b; for the coefficients C0, C1, ..., C 15 Adjustments were made to obtain the effective light intensity I. effective Provide an effective light intensity threshold T; with an effective light intensity I greater than the effective light intensity threshold T. effective Perform simulated exposure to obtain a simulated exposure pattern; obtain a photolithography pattern based on the simulated exposure pattern; obtain the average offset (RMS) based on the dimensions of the photolithography pattern and the simulated exposure pattern at the same position.

[0067] This process involves the coefficients C0, C1, ..., C 15 The optimization process employs a gradient algorithm to optimize the coefficients C0, C1, ..., C. 15 Optimize.

[0068] In this embodiment, the method for obtaining the average offset RMS includes: providing C0, C1, ..., C 15 The coefficient value is provided. The variable value; providing parameter b; adjusting the effective light intensity threshold T to obtain the effective light intensity distribution I. effective With an effective light intensity I greater than the effective light intensity threshold T effective Perform simulated exposure to obtain a simulated exposure pattern; obtain a photolithography pattern based on the simulated exposure pattern; obtain the average offset (RMS) based on the dimensions of the photolithography pattern and the simulated exposure pattern at the same position.

[0069] This process optimizes the effective light intensity threshold T using a gradient algorithm.

[0070] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A method for correcting graphics, characterized in that, include: Provide the layout to be revised; Obtain the original light intensity distribution corresponding to the layout to be corrected; Obtain the diffusion distribution corresponding to the pattern to be corrected, wherein the diffusion distribution includes photoacid distribution and photobase distribution; The original light intensity distribution and diffusion distribution are processed using a free convolution model to obtain the effective light intensity distribution. The original light intensity distribution, photoacid distribution, and photoalkali distribution are then processed using the free convolution model to obtain the effective light intensity distribution. Where Ieffective is the effective light intensity distribution, I is the original light intensity distribution, and Ieffective is the original light intensity distribution. -b For photoacid distribution, I +b The distribution of light and alkali is represented by C0, C1, ..., C15, which are coefficients to be optimized, b is a parameter to be optimized, and FMK is a free convolution model. In the free convolution model, k ranges from 1 to 13, and i ranges from 1 to N natural numbers. Simulated exposure is performed using an effective light intensity greater than the effective light intensity threshold to obtain a simulated exposure pattern. Obtain the lithography pattern based on the simulated exposure pattern; The average offset is obtained based on the dimensions of the photolithographic pattern and the simulated exposure pattern at the same position.

2. The graphic correction method as described in claim 1, characterized in that, The free convolution model FMK is an m×m matrix, and N has a size of m×m×13, where m is a natural number greater than 1.

3. The graphic correction method as described in claim 1, characterized in that, The relationship between the (t+1)th variable adjustment and the tth variable adjustment is: ,in For the (t+1)th variable adjustment, For the t-th variable adjustment, The average offset convergence step size. For finite differences, the value of i ranges from 1 to N, and the value of k ranges from 1 to 13.

4. The graphic correction method as described in claim 1, characterized in that, Methods for obtaining the average offset include: providing the coefficient values ​​of C0, C1, ..., C15; and adjusting the parameter b and variables. Multiple adjustments are made to obtain the effective light intensity; an effective light intensity threshold is provided; simulated exposure is performed with an effective light intensity greater than the effective light intensity threshold to obtain a simulated exposure pattern; a lithographic pattern is obtained based on the simulated exposure pattern; and the average offset is obtained based on the dimensions of the lithographic pattern and the simulated exposure pattern at the same position.

5. The graphic correction method as described in claim 4, characterized in that, Methods for obtaining the average offset include: providing The parameter values ​​are determined; parameter b is provided; the coefficients C0, C1, ..., C15 are adjusted to obtain the effective light intensity; an effective light intensity threshold is provided; simulated exposure is performed with an effective light intensity greater than the effective light intensity threshold to obtain a simulated exposure pattern; a lithographic pattern is obtained based on the simulated exposure pattern; and the average offset is obtained based on the dimensions of the lithographic pattern and the simulated exposure pattern at the same position.

6. The graphic correction method as described in claim 5, characterized in that, Methods for obtaining the average offset include: providing coefficient values ​​for C0, C1, ..., C15; providing... The variable value is provided; parameter b is provided; the effective light intensity threshold is adjusted to obtain the effective light intensity distribution; simulated exposure is performed with an effective light intensity greater than the effective light intensity threshold to obtain the simulated exposure pattern; the lithography pattern is obtained based on the simulated exposure pattern; the average offset is obtained based on the size of the lithography pattern and the simulated exposure pattern at the same position.

7. The graphic correction method as described in claim 1, characterized in that, The average offset is obtained based on the dimensions of the photolithographic pattern and the simulated exposure pattern at the same position. Where RMS is the average offset, CDi,s is the key dimension of the simulated exposure pattern obtained when the variable is adjusted for the i-th time, CDi,w is the key dimension of the same position of the lithographic pattern obtained when the variable is adjusted for the i-th time, wi is the weight related to the i-th variable, and the value of i ranges from 1 to N.

8. The graphic correction method as described in claim 1, characterized in that, The method for obtaining a lithographic pattern from a simulated exposure pattern includes: providing a target pattern; obtaining the edge placement error between the simulated exposure pattern and the target pattern; correcting the pattern to be corrected based on the edge placement error to obtain a corrected pattern; using the corrected pattern as a mask pattern; and obtaining a lithographic pattern based on the mask pattern.

Images