Machine learning based model builder and its applications for pattern transferring in semiconductor manufacturing

Patent metadata

Field	Value
Publication number	US-11610043-B2
Application number	US-202117193625-A
Country	US
Kind code	B2
Filing date	Mar 5, 2021
Priority date	Mar 5, 2021
Publication date	Mar 21, 2023
Grant date	Mar 21, 2023

Abstract

A system and a method of optimizing an optical proximity correction (OPC) model for a mask pattern of a photo mask is disclosed. A machine learning (ML) based model builder includes an OPC model, measurement data and a random term generator. Random terms are generated in a M-dimensional space by the random term generator. The ML based model builder classifies the random terms to clusters by applying a classifying rule. A representative subset of the random terms is determined among the classified clusters, and the representative subset is added to the OPC model.

First claim

What is claimed is:

1. A method of optimizing an optical proximity correction (OPC) model for a mask pattern of a photo mask, the method comprising:
obtaining an OPC model, measurement data and a random term generator;
generating random terms in a M-dimensional space by the random term generator;
classifying the random terms to clusters by applying feature selection, subset selection, and dimensionality reduction;
determining a representative subset of the random terms, wherein the representative subset comprises representative terms from the classified clusters, and wherein the representative terms are selected based on an impact of the representative terms on the OPC model and comprises the representative term having most significant effect on the OPC model;
adding the representative subset to the OPC model;
determining a performance of the OPC model with the added representative subset;
in response to the determination that the performance does not meet a design requirement, automatically adjusting the added representative subset until the performance of the OPC model with the added representative subset meets the design requirement and the OPC model with the added representative subset is obtained as an optimized OPC model; and
producing the mask pattern, using the optimized OPC model, on a mask blank.

2. The method of claim 1 , wherein classifying the random terms to clusters further includes generating a M-dimensional space based on a number N of random terms.

3. The method of claim 1 , wherein classifying the random terms to clusters further includes generating a n-dimensional space including a first subset of the measurement data based on a plurality of sample points.

4. The method of claim 3 , wherein classifying the random terms to clusters further includes generating classified clusters of the measurement data by applying a classifying rule to the n-dimensional space.

5. The method of claim 1 , wherein classifying the random terms to clusters includes selecting representative terms from each of the clusters.

6. The method of claim 1 , wherein determining the representative subset of the random terms includes selecting one or more representative terms from the clusters based on the impact of the representative terms on the OPC model.

7. The method of claim 1 , wherein the random terms include convoluted terms.

8. The method of claim 1 , wherein classifying the random terms to clusters includes generating classified clusters of the measurement data selected from the group consisting of a geographic center method, a moving average method, a data sampling rate method, a data magnification factor method, a data smoothing filter method, and types of sensor method.

9. The method of claim 1 , wherein the OPC model comprises an optical model and a resist model, wherein the optical model defines performance of optical components associated with a lithography scanner tool and the resist model defines distortion of an image in a photoresist layer.

10. The method of claim 1 , further comprising:
applying the OPC model on a mask pattern of a photo mask to produce simulated measurement data; and
comparing the simulated measurement data with the measurement data to verify the OPC model.

11. The method of claim 1 , wherein automatically adjusting the representative subset comprises performing steps of:
sorting the representative terms based on impact on the OPC model; and
adding a representative term, having most significant impact, to the representative subset until the performance of the representative subset meets the design requirement.

12. A method of optimizing an optical proximity correction (OPC) model for a mask pattern of a photo mask, the method comprising:
in response to a request of refining the OPC model, obtaining the OPC model and a random term generator;
obtaining selected measurement data from a characteristic data selector that sorts a plurality of measurement data in terms of impact to the OPC model and selects one or more measurement data having most significant impact on the OPC model as the selected measurement data;
adding the selected measurement data into the OPC model;
determining a performance of the OPC model with the selected measurement data;
in response to the determination that the performance does not meet a design requirement, automatically adjusting the characteristic data selector until the performance of the OPC model with the selected measurement data obtained by the adjusted characteristic data selector meets the design requirement and the OPC model with the added measurement data is obtained as an optimized OPC model; and
producing the mask pattern, using the optimized OPC model, on a mask blank.

13. The method of claim 12 , wherein obtaining selected measurement data includes obtaining selected measurement data based on a machine learning module.

14. The method of claim 12 , wherein the OPC model comprises an optical model and a resist model, wherein the optical model defines performance of optical components associated with a lithography scanner tool and the resist model defines distortion of an image in a photoresist layer.

15. The method of claim 12 , further comprising:
applying the OPC model on a mask pattern of a photo mask to produce simulated measurement data.

16. A system of optimizing an optical proximity correction (OPC) model for a mask pattern of a photo mask, comprising:
a random term generator;
a memory that stores computer executable components;
a processor that executes computer executable components stored in the memory, wherein the computer executable components comprise:
a statistical learning component that:
generates random terms in a M-dimensional space by the random term generator;
classifies the random terms to clusters by applying feature selection, subset selection, and dimensionality reduction;
determines a representative subset of the random terms, wherein the representative subset comprises representative terms from the classified clusters, and wherein the representative terms are selected based on an impact of the representative terms on the OPC model and comprises the representative term having most significant effect on the OPC model;
adds the representative subset to the OPC model;
determines a performance of the OPC model with the added representative subset;
in response to the determination that the performance does not meet a design requirement, automatically adjusts the added representative subset until the performance of the OPC model with the added representative subset meets the design requirement; and
obtains the OPC model with the added representative subset as an optimized OPC model.

17. The system of claim 16 , wherein the OPC model includes an optical model including optical parameters associated with a lithography scanner tool.

18. The system of claim 16 , wherein the OPC model includes a resist model based on an optical image and a mask pattern dependent functions associated with a lithography scanner tool.

19. The system of claim 16 , wherein the statistical learning component further includes a n-dimensional vector based on convoluted terms.

20. The system of claim 16 , wherein the statistical learning component further includes policy data representing a weighting policy associated with a machine learning module.

Assignees

• Taiwan Semiconductor Mfg Co Ltd

Inventors

• Yu Zhiru
• Zhang Lin
• Peng Danping
• Lei Junjiang

Classifications

• G03F7/70441

  Optical proximity correction [OPC] · CPC title

• G06F30/398 Primary

  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

• G03F7/705

  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