Method for determining a mask pattern comprising optical proximity corrections using a trained machine learning model

Patent metadata

Field	Value
Publication number	US-12566368-B2
Application number	US-202117796751-A
Country	US
Kind code	B2
Filing date	Feb 4, 2021
Priority date	Feb 12, 2020
Publication date	Mar 3, 2026
Grant date	Mar 3, 2026

Abstract

A method for determining a mask pattern and a method for training a machine learning model. The method for determining a mask pattern includes obtaining, via executing a model using a target pattern to be printed on a substrate as an input pattern, a post optical proximity correction (post-OPC) pattern; determining, based on the post-OPC pattern, a simulated pattern that will be printed on the substrate; and determining the mask pattern based on a difference between the simulated pattern and the target pattern. The determining of the mask pattern includes modifying, based on the difference, the input pattern inputted to the model such that the difference is reduced; and executing, using the modified input pattern, the model to generate a modified post-OPC pattern from which the mask pattern can be derived.

First claim

The invention claimed is:

1 . A non-transitory computer-readable medium comprising instructions therein, the instructions, when executed by one or more processors, configured to cause the one or more processors to at least:
obtain, via execution of a model using a target pattern to be printed on a substrate as an input to the model, a post optical proximity correction (post-OPC) pattern; determine, based on the post-OPC pattern, a simulated pattern that will be printed on the substrate; and determine a mask pattern based on a difference between the simulated pattern and the target pattern by:
modification, based on the difference, of the target pattern inputted to the model such that the difference is reduced; and
execution, using the modified target pattern, of the model to generate a modified post-OPC pattern from which the mask pattern can be derived.

2 . The non-transitory computer readable medium of claim 1 , wherein the model is a trained machine learning model configured to generate a post-OPC pattern for an input pattern.

3 . The non-transitory computer readable medium of claim 2 , wherein the instructions configured to cause the computer system to determine the mask pattern are further configured to cause the computer system to determine the mask pattern in an iterative manner, each iteration comprising:
modification, based on a gradient of the difference, of at least a portion of the target pattern inputted to the model to reduce the difference between the target pattern and the simulated pattern, the gradient being indicative of how the target pattern should be modified to reduce or minimize the difference; execution, using the modified target pattern, of the trained machine learning model to generate the modified post-OPC pattern; determination of the simulated pattern based on the modified post-OPC pattern; determination of whether the difference between the simulated pattern and the target pattern is reduced or minimized; and responsive to the difference being reduced or minimized, extraction of polygon shapes from the modified post-OPC pattern to generate the mask pattern.

4 . The non-transitory computer readable medium of claim 3 , wherein the at least a portion of the target pattern inputted to the model comprises a contour corresponding to a target feature within the target pattern, and/or
wherein the difference between the simulated pattern and the target pattern comprises a difference between a target contour of the target pattern and a simulated contour of the simulated pattern, and/or wherein the instructions configured to cause the computer system to modify the target pattern inputted to the model are further configured to cause the computer system to modify, based on the gradient of the difference, a contour of the target pattern to reduce the difference between a target contour of the target pattern and a simulated contour of the simulated pattern, the gradient being indicative of how a contour of the target pattern should be modified to reduce or minimize the difference.

5 . The non-transitory computer readable medium of claim 3 , wherein the instructions configured to cause the computer system to modify the target pattern inputted to the model are further configured to cause the computer system to:
assign control points on a target contour of the target pattern; and adjust, based on the gradient of the difference, a position of one or more control points such that the difference is reduced or minimized.

6 . The non-transitory computer readable medium of claim 1 , wherein the instructions configured to cause the computer system to determine the simulated pattern are further configured to cause the computer system to execute a process model of the patterning process using the post-OPC pattern or the modified post-OPC pattern to generate a simulated pattern.

7 . The non-transitory computer readable medium of claim 1 , wherein the instructions are further configured to cause the computer system to:
process, via thresholding, an image of the modified post-OPC pattern to detect edges associated with one or more features within the modified post-OPC pattern; and generate the mask pattern using the edges of the one or more features.

8 . The non-transitory computer readable medium of claim 1 , wherein the post-OPC pattern or the modified post-OPC pattern comprises: a main feature corresponding to the target feature, and at least one assist feature located around or beside the main feature.

9 . The non-transitory computer readable medium of claim 8 , wherein the model is a trained machine learning model and wherein the at least one assist feature is not modified in a second or a subsequent execution of the trained machine learning model when the modified target pattern is used.

10 . The non-transitory computer readable medium of claim 1 , wherein the target pattern, the post-OPC pattern, and/or the modified post-OPC pattern is a gray-scale pixelated image.

11 . The non-transitory computer readable medium of claim 6 , wherein the process model is an after development process model, and the simulated pattern is an after development image.

12 . The non-transitory computer readable medium of claim 11 , wherein the process model is a resist model and the simulated pattern is a resist pattern.

13 . A method for determining a mask pattern to be employed in a patterning process, the method comprising:
obtaining, via executing, by a hardware computer system, a computer model using a target pattern to be printed on a substrate as an input to the model, a post optical proximity correction (post-OPC) pattern; determining, based on the post-OPC pattern, a simulated pattern that will be printed on the substrate; and determining the mask pattern based on a difference between the simulated pattern and the target pattern, the determining of the mask pattern comprising:
modifying, based on the difference, the target pattern inputted to the model such that the difference is reduced; and
executing, using the modified target pattern, the model to generate a modified post-OPC pattern from which the mask pattern can be derived.

14 . The method of claim 13 , wherein the model is a trained machine learning model configured to generate a post-OPC pattern for an input pattern.

15 . The method of claim 14 , wherein the determining of the mask pattern is done in an iterative manner, each iteration comprising:
modifying, based on a gradient of the difference, at least a portion of the target pattern inputted to the model to reduce the difference between the target pattern and the simulated pattern, the gradient being indicative of how the target pattern should be modified to reduce or minimize the difference; executing, using the modified target pattern, the trained machine learning model to generate the modified post-OPC pattern; determining the simulated pattern based on the modified post-OPC pattern; determining whether the difference between the simulated pattern and the target pattern is reduced or minimized; and responsive to the difference being reduced or minimized, extracting polygon shapes from the modified post-OPC pattern to generate the mask pattern.

16 . The method of claim 15 , wherein the at least a portion of the target pattern inputted to the model comprises a contour corresponding to a target feature within the target pattern, and/or
wherein the difference between the simulated pattern and the target pattern comprises a difference between a target contour of the target pattern and a simulated contour of the simulated pattern, an

Assignees

• Asml Netherlands Bv
• Asml Netherlands

Inventors

• Cao Yu
• Tao Jun
• Zhang Quan
• Shu Yongsheng
• Fong Wei-Chun

Classifications

• G06F30/27

  using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model · CPC title

• G06F30/398

  Design verification or optimisation, e.g. using design rule check [DRC], layout versus schematics [LVS] or finite element methods [FEM] (optical proximity correction [OPC] design processes G03F1/36) · CPC title

• G03F7/70441

  Optical proximity correction [OPC] · CPC title

• G03F7/705

  Modelling or simulating from physical phenomena up to complete wafer processes or whole workflow in wafer productions · CPC title

• G03F1/36 Primary

  Masks having proximity correction features; Preparation thereof, e.g. optical proximity correction [OPC] design processes · CPC title