A lithography model optimization method

Abstract

The present application relates to the technical field of photoetching modeling, in particular to a photoetching model optimization method for reducing the influence of grid point dependency in a photoetching model, comprising the following steps: obtaining actual critical dimensions, calculating the grid point dependency of the corresponding preset critical dimensions at preset grid point offset positions according to the obtained actual critical dimensions; obtaining at least two key values of the maximum offset position and the maximum grid point dependency of each preset critical dimension according to the grid point dependency of the preset critical dimensions at the preset grid point offset positions; evaluating the grid point dependency of the preset critical dimensions according to the at least two key values corresponding to each preset critical dimension; and adjusting the corresponding preset critical dimensions according to the evaluation results; this design can solve the problem of too large grid point dependency in the existing photoetching model.

Description

[Technical Field]

[0001] This invention relates to the field of photolithography modeling technology, and in particular to a photolithography model optimization method. [Background Technology]

[0002] In modern digital image processing, images are stored in a raster structure. A raster structure divides an image into uniformly distributed grids, each called a pixel, and explicitly records the luminance value of each pixel. Images consist of a digital array of information, with all pixel positions arranged in a regular pattern.

[0003] In modern photolithography, a method of storing image content using a similar grid structure is called grid-based mapping. Combining mathematical theory, a mask can be described as a polygon with different transmittance and phase. According to the Nyquist sampling theorem, the sampling frequency should be twice the maximum frequency of the original signal. Different sampling positions contain different information, which is always a fragment of the original data, leading to grid dependence. Image quality and grid dependence are ensured through upsampling / downsampling processes. The main pixel size is primarily determined by the spatial frequency passband or empirical formulas. The main mask errors are corner rounding and bias, which are physically derived from electron beam proximity effects / etching processes, etc.

[0004] Normally, grid quantization error is reduced by selecting a high-resolution subpixel size. However, with the continuous advancement of technologies related to various process nodes in integrated circuit manufacturing, the feature size of integrated circuit manufacturing processes is constantly shrinking. Therefore, the requirements for reducing grid dependence are becoming increasingly stringent. Existing computational lithography models still have a high dependence on grids and can no longer meet process requirements. [Summary of the Invention]

[0005] To address the problem that existing computational lithography models are too dependent on grid points, this invention provides a lithography model optimization method.

[0006] The solution to the technical problem of this invention is to provide a photolithography model optimization method to reduce the influence of lattice dependency in the photolithography model, comprising the following steps:

[0007] Obtain the actual key dimensions, and calculate the grid dependency of the corresponding preset key dimensions at preset grid offset positions based on the obtained actual key dimensions; obtain at least two key values ​​for each preset key dimension, namely the maximum offset position and the maximum grid dependency, based on the grid dependency of the preset key dimensions at preset grid offset positions; evaluate the grid dependency of the preset key dimensions based on at least two key values ​​corresponding to each preset key dimension; adjust the corresponding preset key dimensions based on the evaluation results.

[0008] Preferably, obtaining the actual key dimensions and calculating the grid dependency of the corresponding preset key dimensions at the preset grid offset position based on the obtained actual key dimensions specifically includes the following steps: generating an initial optical model; obtaining the pixel size of the initial optical model and calculating the grid offset range based on the pixel size; obtaining the actual key dimensions from the initial optical model and calculating the grid dependency of the preset key dimensions at the preset grid offset position based on the actual key dimensions.

[0009] Preferably, the formula for defining the grid offset range is as follows: pixel represents the pixel size of the initial optical model, and shift represents the grid offset range.

[0010] Preferably, the step of obtaining at least two key values ​​for each preset key dimension—the maximum offset position and the maximum grid point dependency—based on the grid point dependency of the preset key dimension at the preset grid point offset position specifically includes the following steps:

[0011] The maximum offset position corresponding to each preset key dimension is obtained by calculation and / or comparison; the maximum grid point dependency corresponding to each preset key dimension is obtained by calculation and / or comparison.

[0012] Preferably, based on the grid point dependency of the preset key dimensions at preset grid point offset positions, at least two key values ​​are obtained: the maximum offset position and the maximum grid point dependency of each preset key dimension. The method further includes the following steps:

[0013] The average grid value of each preset key dimension is calculated based on the number of preset grid offset positions and the corresponding grid dependency of each preset key dimension.

[0014] Preferably, the step of obtaining at least two key values—the maximum offset position and the maximum grid point dependency—of each preset key dimension based on the grid point dependency at the preset grid point offset position further includes the following steps:

[0015] The discrete value of each preset key dimension is calculated based on the number of preset grid offset positions for each preset key dimension, the corresponding grid dependency, and the average grid value of each preset key dimension.

[0016] Preferably, the definition formula for the grid dependency is as follows: Wherein, grid represents the grid dependency, cd represents the actual critical size, subscript i represents the preset grid offset position, subscript j represents the preset number of critical sizes, and N represents the total number of grid points, which is obtained based on the grid offset range.

[0017] Preferably, the formula for defining the grid point discrete value is as follows: Where σ represents the discrete value, cd represents the actual critical dimension, subscript i represents the preset grid offset position, subscript j represents the preset number of critical dimensions, N represents the total number of grid points, which is obtained based on the grid offset range, and M represents the number of critical dimensions.

[0018] Preferably, the grid dependency of the preset key dimensions is evaluated based on at least two key values ​​corresponding to each preset key dimension, specifically including the following steps:

[0019] The maximum grid point dependence of each preset key dimension is compared; the grid point discrete values ​​of each preset key dimension are compared; and based on the comparison results, at least one preset key dimension that is more easily affected by the grid points is determined.

[0020] Preferably, the corresponding preset key dimensions are adjusted based on the evaluation results, specifically including the following steps:

[0021] Adjust the positions of preset key dimensions that are more easily affected by grid points.

[0022] Compared with existing technologies, the photolithography model optimization method of the present invention has the following advantages:

[0023] 1. The lithography model optimization method of the present invention first obtains the actual key dimensions and calculates the grid dependency of the corresponding preset key dimensions at preset grid offset positions. Then, based on the grid dependency, it calculates at least two key values: the maximum offset position and the grid dependency corresponding to each preset key dimension. The calculation results of at least two key values ​​are combined to evaluate the grid dependency of the model. This design enables the evaluation results to accurately reflect the grid dependency of each preset key dimension of the model on each grid point, so that the position of the key dimensions of the model can be precisely optimized in subsequent steps based on the judgment results to reduce the grid dependency of the model. This method is efficient and effective, and can effectively solve the problem that the existing computational lithography model has too much dependence on grid points.

[0024] 2. The grid offset range in this invention is determined based on the pixel size, which is determined by the initial optical model generated for each modeling object, resulting in high specificity and accuracy.

[0025] 3. In this invention, the absolute value of the grid offset range is less than or equal to half the pixel size of the initial optical model.

[0026] 4. In addition to obtaining the two key values ​​corresponding to each preset key dimension, the maximum offset position and grid point dependency value, this invention also calculates the average grid point value of each preset key dimension. The average grid point value can help evaluate the grid point dependency of the model, further improve the accuracy of the evaluation results, and thus ensure the optimization effect of the computational lithography model.

[0027] 5. In addition to obtaining the two key values ​​corresponding to each preset key dimension, the maximum offset position and the grid point dependency value, this invention also calculates the grid point discrete value of each preset key dimension. The grid point discrete value can help evaluate the grid point dependency of the model, further improve the accuracy of the evaluation results, and thus ensure the optimization effect of the computational lithography model.

[0028] 6. In this invention, grid point dependence refers to the degree to which the image formed depends on the grid points when imaging the modeled object. When the grid point dependence is large, even a slight shift in the point position will have a great impact on the imaging. From the data, this means that the actual key dimensions will have a large error. Understandably, when the grid point dependence is small, the shift in the point position will have a smaller impact on the imaging.

[0029] 7. In this invention, the grid point discrete value is used to evaluate the grid point dependency of each preset grid point offset position, and to help determine whether the calculation result of the grid point dependency deviates from the normal range.

[0030] 8. This invention compares the corresponding key values ​​of each preset key dimension to determine which preset key dimension is more significantly affected by grid points than other key dimensions. When further optimizing the model, the position of the key dimension that is more easily affected by grid points can be adjusted, which can greatly reduce the impact of grid point division on the lithography model. [Attached Image Description]

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a block diagram of the photolithography model optimization method provided in the first embodiment of the present invention.

[0033] Figure 2 This is a block diagram of step S1 provided in the first embodiment of the present invention.

[0034] Figure 3 This is a block diagram of step S2 provided in the second embodiment of the present invention.

[0035] Figure 4 This is a block diagram of step S3 provided in the third embodiment of the present invention.

[0036] Figure 5 This is an example table of grid-point dependency data for some key dimensions provided in the fourth embodiment of the present invention.

Detailed Implementation Methods

[0037] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0038] The terms “vertical,” “horizontal,” “left,” “right,” “up,” “down,” “upper left,” “upper right,” “lower left,” “lower right,” “lower left,” “lower right,” and similar expressions used in this article are for illustrative purposes only.

[0039] Please combine Figures 1 to 4 The first embodiment of the present invention provides a photolithography model optimization method, including the following steps:

[0040] Step S1: Obtain the actual key dimensions, and calculate the grid dependency of the corresponding preset key dimensions at the preset grid offset positions based on the obtained actual key dimensions;

[0041] Step S2: Based on the grid dependency of the preset key dimensions at the preset grid offset positions, obtain at least two key values ​​for each preset key dimension: the maximum offset position and the maximum grid dependency.

[0042] Step S3: Evaluate the grid dependency of the preset key dimensions based on at least two key values ​​corresponding to each preset key dimension;

[0043] Step S4: Adjust the corresponding preset key dimensions based on the evaluation results.

[0044] Understandably, the lithography model optimization method of the present invention first obtains the actual key dimensions and calculates the grid dependency of the corresponding preset key dimensions at preset grid offset positions. Then, based on the grid dependency, it calculates at least two key values: the maximum offset position and the grid dependency corresponding to each preset key dimension. The calculation results of at least two key values ​​are combined to evaluate the grid dependency of the model. This design enables the evaluation results to accurately reflect the grid dependency of each preset key dimension of the model on each grid point, so that the position of the key dimensions of the model can be precisely optimized in subsequent steps based on the judgment results to reduce the grid dependency of the model. This method is efficient and effective, and can effectively solve the problem that the existing computational lithography model has too much dependence on grid points.

[0045] Furthermore, in step S1, the actual key dimensions are obtained by using the preset grid offset position within the grid offset range as the starting grid point.

[0046] Furthermore, the number of preset key dimensions can be adjusted according to the actual situation, and can be one or more. In this embodiment, the number of preset key dimensions is more than three thousand; in other embodiments, the number of preset key dimensions can reach tens of thousands or even higher.

[0047] Specifically, in this invention, grid point dependence refers to the degree to which the image formed depends on the grid point position when imaging a modeled object. When the grid point dependence is large, even a slight shift in the position of the point will have a significant impact on the imaging. From a data perspective, this means that the actual key dimensions will have a large error. It is understandable that when the grid point dependence is small, the shift in the position of the point will have a smaller impact on the imaging. For example, one situation where the grid point dependence is relatively large is when the shape near the preset grid point offset position is relatively complex. Even a slight change in the preset grid point offset position will have a significant impact on the imaging.

[0048] Furthermore, step S1 specifically includes the following steps:

[0049] Step S11: Generate the initial optical model;

[0050] Step S12: Obtain the pixel size of the initial optical model and calculate the grid offset range based on the pixel size; it can be understood that the grid offset range in this invention is determined based on the pixel size, and the pixel size is determined by the initial optical model generated for each modeling object, which is highly targeted and accurate.

[0051] Step S13: Obtain the preset key dimensions, and calculate the grid dependency of the corresponding grid points within the corresponding grid offset range based on the preset key dimensions.

[0052] Specifically, in this embodiment, the formula for defining the grid offset range is as follows: `pixel` represents the pixel size of the initial optical model, and `shift` represents the grid offset range. For example, when `pixel` is 24, -12 ≤ `shift` ≤ 12.

[0053] Furthermore, the definition formula for lattice dependency is as follows: Where grid represents grid dependency, cd represents the actual critical size, subscript i represents the grid offset position, subscript j represents the preset number of critical sizes, and N represents the total number of grid points, which is obtained based on the grid offset range, that is, the total number of grid points within the grid offset range.

[0054] Furthermore, step S2 specifically includes the following steps:

[0055] Step S21: Combine the offset positions of each preset grid point and the corresponding grid point dependency of each preset key dimension, and obtain the maximum offset position of each preset key dimension by comparison;

[0056] Step S22: Combine the offset positions of each preset grid point and the corresponding grid point dependency of each preset key dimension, and obtain the maximum grid point dependency of each preset key dimension by comparison.

[0057] Furthermore, step S2 also includes the following steps:

[0058] Step S23: Calculate the average grid value for each preset key dimension based on the number of preset grid offset positions and the corresponding grid dependency. It is understandable that the average grid value helps evaluate the grid dependency of the model, further improving the accuracy of the evaluation results, thereby ensuring the optimization effect on the computational lithography model.

[0059] Specifically, in this embodiment, the formula for defining the average grid point value is as follows: mean represents the average grid point value, cd represents the actual critical dimension, subscript i represents the preset grid point offset position, subscript j represents the preset number of critical dimensions, N represents the total number of grid points, which is obtained based on the grid point offset range, and M represents the number of critical dimensions.

[0060] Furthermore, in step S2, the preset grid offset position must be within the grid offset range.

[0061] Furthermore, step S2 also includes the following steps:

[0062] Step S24: Calculate the grid discrete value for each preset key dimension based on the number of preset grid offset positions for each preset key dimension, the corresponding grid dependency, and the average grid value for each preset key dimension. Understandably, the grid discrete value helps evaluate the grid dependency of the model, further improving the accuracy of the evaluation results, thereby ensuring the optimization effect on the computational lithography model.

[0063] Understandably, in this invention, the grid point discrete values ​​are used to evaluate the grid point dependency of each preset grid point offset position, and to help determine whether the calculation result of the grid point dependency deviates from the normal range.

[0064] Furthermore, the grid point discrete values ​​are calculated according to the 3sigma principle.

[0065] Specifically, in this embodiment, the formula for defining the grid point discrete value is as follows: σ represents the grid point discrete value, mean represents the average grid point value, cd represents the actual critical dimension, i represents the preset grid point offset position, j represents the preset number of critical dimensions, N represents the total number of grid points, which is obtained based on the grid point offset range, and M represents the number of critical dimensions.

[0066] Furthermore, step S3 specifically includes the following steps:

[0067] Step S31: Compare the maximum grid point dependency for each preset key size;

[0068] Step S32: Compare the grid point discrete values ​​for each preset key dimension;

[0069] Step S33: Based on the comparison results, determine at least one preset critical dimension that is more easily affected by grid points.

[0070] Furthermore, step S4 specifically includes the following steps:

[0071] Step S41: Adjust the positions of preset key dimensions that are more easily affected by grid points.

[0072] Furthermore, when the maximum grid point discrete value of the preset critical dimension is greater than or equal to the first preset value, and the average grid point value is greater than or equal to the second preset value, it is determined that the preset critical dimension is significantly affected by the grid points.

[0073] Furthermore, when the maximum grid point discrete value of a preset key dimension is greater than or equal to the first preset value, and the average grid point value is less than the second preset value, the grid point discrete value is used to determine whether the grid point dependency value of the preset key dimension is normal. If it is not normal, the data set is excluded, and the cause is traced back to find the source.

[0074] Please see Figure 5 In the table, Grid in shift represents the grid dependency value of the preset key dimension at a certain grid offset position, max_grid represents the maximum grid dependency value of the preset key dimension within the grid offset range, meam_grid represents the average grid value of the preset key dimension within the grid offset range, and 3sigma_grid represents the grid discrete value of the preset key dimension within the grid offset range.

[0075] Furthermore, based on the information in the table, it can be seen that the first preset critical dimension has max_grid = 0.6 and 3sigma_grid = 0.74, indicating that this critical dimension is more significantly affected by the grid than other critical dimensions. When performing step S4, a calculation model can be designed for its optimal position, which can greatly reduce the impact of grid division on the lithography model.

[0076] Compared with existing technologies, the photolithography model optimization method of the present invention has the following advantages:

[0077] 1. The lithography model optimization method of the present invention first obtains the actual key dimensions and calculates the grid dependency of the corresponding preset key dimensions at preset grid offset positions. Then, based on the grid dependency, it calculates at least two key values: the maximum offset position and the grid dependency corresponding to each preset key dimension. The calculation results of at least two key values ​​are combined to evaluate the grid dependency of the model. This design enables the evaluation results to accurately reflect the grid dependency of each preset key dimension of the model on each grid point, so that the position of the key dimensions of the model can be precisely optimized in subsequent steps based on the judgment results to reduce the grid dependency of the model. This method is efficient and effective, and can effectively solve the problem that the existing computational lithography model has too much dependence on grid points.

[0078] 2. The grid offset range in this invention is determined based on the pixel size, which is determined by the initial optical model generated for each modeling object, resulting in high specificity and accuracy.

[0079] 3. In this invention, the absolute value of the grid offset range is less than or equal to half the pixel size of the initial optical model.

[0080] 4. In addition to obtaining the two key values ​​corresponding to each preset key dimension, the maximum offset position and grid point dependency value, this invention also calculates the average grid point value of each preset key dimension. The average grid point value can help evaluate the grid point dependency of the model, further improve the accuracy of the evaluation results, and thus ensure the optimization effect of the computational lithography model.

[0081] 5. In addition to obtaining the two key values ​​corresponding to each preset key dimension, the maximum offset position and the grid point dependency value, this invention also calculates the grid point discrete value of each preset key dimension. The grid point discrete value can help evaluate the grid point dependency of the model, further improve the accuracy of the evaluation results, and thus ensure the optimization effect of the computational lithography model.

[0082] 6. In this invention, grid point dependence refers to the degree to which the image formed depends on the grid points when imaging the modeled object. When the grid point dependence is large, even a slight shift in the point position will have a great impact on the imaging. From the data, this means that the actual key dimensions will have a large error. Understandably, when the grid point dependence is small, the shift in the point position will have a smaller impact on the imaging.

[0083] 7. In this invention, the grid point discrete value is used to evaluate the grid point dependency of each preset grid point offset position, and to help determine whether the calculation result of the grid point dependency deviates from the normal range.

[0084] 8. This invention compares the corresponding key values ​​of each preset key dimension to determine which preset key dimension is more significantly affected by grid points than other key dimensions. When further optimizing the model, the position of the key dimension that is more easily affected by grid points can be adjusted, which can greatly reduce the impact of grid point division on the lithography model.

[0085] The above provides a detailed description of a photolithography model optimization method disclosed in the embodiments of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention. Any modifications, equivalent substitutions, and improvements made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A photolithography model optimization method for reducing the influence of lattice dependency in a photolithography model, characterized in that, Includes the following steps: Generate an initial optical model, obtain the actual key dimensions from the initial optical model, and calculate the grid dependency of the corresponding preset key dimensions at preset grid offset positions based on the actual key dimensions; Based on the grid dependency of the preset key dimensions at the preset grid offset positions, at least two key values ​​are obtained for each preset key dimension: the maximum offset position and the maximum grid dependency. The grid dependency of the preset key dimensions is evaluated based on at least two key values ​​corresponding to each preset key dimension; Adjust the corresponding preset key dimensions based on the evaluation results.

2. The photolithography model optimization method as described in claim 1, characterized in that, After generating the initial optical model, the following steps are also included: Obtain the pixel size of the initial optical model, and calculate the grid offset range based on the pixel size.

3. The photolithography model optimization method as described in claim 2, characterized in that: The formula for defining the grid offset range is as follows: , where pixel represents the pixel size of the initial optical model, and shift represents the grid offset range.

4. The photolithography model optimization method as described in claim 1, characterized in that, The step of obtaining at least two key values—the maximum offset position and the maximum grid point dependency—from each preset key dimension based on the grid point dependency at the preset grid point offset position specifically includes the following steps: The maximum offset position corresponding to each preset key dimension is obtained by calculation and / or comparison. The maximum grid point dependency for each preset key size is obtained by calculation and / or comparison.

5. The photolithography model optimization method as described in claim 4, characterized in that, The step of obtaining at least two key values—the maximum offset position and the maximum grid point dependency—of each preset key dimension based on the grid point dependency at the preset grid point offset position also includes the following steps: The average grid value of each preset key dimension is calculated based on the number of preset grid offset positions and the corresponding grid dependency of each preset key dimension.

6. The photolithography model optimization method as described in claim 5, characterized in that, The step of obtaining at least two key values—the maximum offset position and the maximum grid point dependency—of each preset key dimension based on the grid point dependency at the preset grid point offset position also includes the following steps: The discrete value of each preset key dimension is calculated based on the number of preset grid offset positions for each preset key dimension, the corresponding grid dependency, and the average grid value of each preset key dimension.

7. The photolithography model optimization method as described in claim 4, characterized in that: The definition formula for the lattice dependency is as follows: In this context, grid represents the grid dependency, cd represents the actual critical size, subscript i represents the preset grid offset position, subscript j represents the preset number of critical sizes, and N represents the total number of grid points, which is obtained based on the grid offset range.

8. The photolithography model optimization method as described in claim 6, characterized in that: The definition formula for the grid point discrete value is as follows: Where σ represents the discrete value, cd represents the actual critical dimension, and the subscript i represents the preset grid offset position. The subscript j represents the number of preset critical dimensions, N represents the total number of grid points, which is obtained based on the grid point offset range, and M represents the number of critical dimensions.

9. The photolithography model optimization method as described in claim 6, characterized in that, The evaluation of the grid dependency of the preset key dimensions based on at least two key values ​​corresponding to each preset key dimension specifically includes the following steps: Compare the maximum grid point dependency for each preset key size; Compare the discrete values ​​of the grid points for each preset key dimension; Based on the comparison results, at least one preset critical dimension that is more easily affected by grid points is determined.

10. The photolithography model optimization method as described in claim 9, characterized in that, The adjustment of the corresponding preset key dimensions based on the evaluation results specifically includes the following steps: Adjust the positions of preset key dimensions that are more easily affected by grid points.

Images