# US 2023/0408932 A1 · Deep learning based adaptive alignment precision metrology for digital overlay

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0408932 A1
- Document ID: US20230408932A1
- Status: Published application
- Kind: A1
- Application number: 18/034,903
- Filed: 2021-11-30
- Published: 2023-12-21
- Earliest claimed: 2020-12-03
- Assignees: Applied Materials, Inc.
- Applicants: Applied Materials, Inc.
- Inventors: Tamer COSKUN; Yen-Shuo LIN; Aidyn KEMELDINOV
- Web page: https://patentgrep.com/patent/US20230408932A1

## CPC

- G03F 7/706841: {Machine learning}
- G03F 9/7046: {Strategy, e.g. mark, sensor or wavelength selection}
- G03F 9/7003: {Alignment type or strategy, e.g. leveling, global alignment}
- G03F 9/7088: {Alignment mark detection, e.g. TTR, TTL, off-axis detection, array detector, video detection}
- G03F 9/7092: {Signal processing}
- G06N 20/00: Machine learning
- G06T 7/001: {using an image reference approach}
- G06T 7/337: {involving reference images or patches}
- G06T 7/73: using feature-based methods
- G06V 10/44: Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
- G06V 10/7515: {Shifting the patterns to accommodate for positional errors}
- G06V 10/764: using classification, e.g. of video objects
- G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
- G06T 2207/20081: Training; Learning
- G06T 2207/30148: Semiconductor; IC; Wafer
- G06T 2207/30204: Marker
- H04N 7/181: {for receiving images from a plurality of remote sources}
- G03F: PHOTOMECHANICAL PRODUCTION OF TEXTURED OR PATTERNED SURFACES, e.g. FOR PRINTING, FOR PROCESSING OF SEMICONDUCTOR DEVICES; MATERIALS THEREFOR; ORIGINALS THEREFOR; APPARATUS SPECIALLY ADAPTED THEREFOR; (phototypographic composing devices B41B; photosensitive materials or processes for photographic purposes G03C; electrophotography, sensitive layers or processes therefor G03G)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

Embodiments described herein relate to a system, methods, and non-transitory computer-readable mediums that accurately align subsequent patterned layers in a photoresist utilizing a deep learning model and utilizing device patterns to replace alignment marks in lithography processes. The deep learning model is trained to recognize unique device patterns called alignment patterns in the FOV of the camera. Cameras in the lithography system capture images of the alignment patterns. The deep learning model finds the alignment patterns in the field of view of the cameras. An ideal image generated from a design file is matched with the camera with respect to the center of the field of view of the camera. A shift model and a rotation model are output from the deep learning model to create an alignment model. The alignment model is applied to the currently printing layer.

## Claims

### Claim 1

A method, comprising:
capturing an image in a field of view (FOV) of each camera of a plurality of cameras of a processing unit of a lithography system, the processing unit having a plurality of image projection systems, each image projection system of the plurality of image projection systems including at least one of each camera of the plurality of cameras, the image including one or more real objects;
transferring the image in the FOV to a server in communication with the lithography system, the server including a deep learning (DL) module, the DL module including a deep learning (DL) model, a template searcher, and an affine modeler;
identifying a real class and a real location of the real objects in the FOV;
locating a template box within the FOV, the template box located based on object matching with a template, the template including a plurality of ideal objects; and
fitting a shift model and a rotation model to a location difference between the plurality of ideal objects and the one or more real objects in the FOV.

### Claim 2 (depends on 1)

The method of claim 1, wherein the lithography system further includes a lithography server, the lithography server operable to receive the image from each camera of the plurality of cameras.

### Claim 3 (depends on 2)

The method of claim 2, wherein the plurality of image projection systems are in communication with an image projection computer, the image projection computer operable to send the image from each camera of the plurality of cameras to the lithography server.

### Claim 4 (depends on 3)

The method of claim 3, wherein the image projection computer is operable to instruct the camera to capture the image in the FOV of each camera of the plurality of cameras.

### Claim 5 (depends on 1)

The method of claim 1, wherein the one or more real objects correspond to a resulting pattern based on a mask pattern data, the resulting pattern formed in a photoresist by the processing unit.

### Claim 6 (depends on 5)

The method of claim 5, further comprising altering a second mask pattern data corresponding to a second pattern according to the shift model and the rotation model, the second pattern patterned above the resulting pattern.

### Claim 7 (depends on 1)

The method of claim 1, wherein the FOV of the camera is positioned to capture an alignment mark of the image.

### Claim 8 (depends on 1)

The method of claim 1, wherein the FOV of the camera is positioned to capture a target feature of the image.

### Claim 9 (depends on 1)

The method of claim 1, wherein the DL model is operable to identify the real class and the real location of the real objects in the FOV.

### Claim 10 (depends on 1)

The method of claim 1, wherein the template searcher is operable to locate the template box within the FOV.

### Claim 11 (depends on 1)

The method of claim 1, wherein the affine modeler is operable to fit the shift model and the rotation model.

### Claim 12

A method, comprising:
capturing a sample image in a field of view (FOV) of each camera of a plurality of cameras of a processing unit of a lithography system, the processing unit having a plurality of image projection systems, each image projection system of the plurality of image projection systems including at least one of each camera of the plurality of cameras, the sample image including one or more sample objects and one or more sample alignment marks;
transferring the sample image in the FOV to a server in communication with the lithography system, the server including a deep learning (DL) module, the DL module including a geometry processing software and a deep learning training engine;
identifying a positon of the one or more sample alignment marks with the geometry processing software;
creating a ground truth table with the geometry processing software, the ground truth table including an ideal class and an ideal location of one or more ideal objects and of one or more alignment marks; and
producing an X shift value, a Y shift value, and a rotation value in the deep learning training model, the X shift value, the Y shift value, and the rotation value corresponding to a position difference between the sample alignment marks compared to the ground truth table.

### Claim 13 (depends on 12)

The method of claim 12, wherein the lithography system further includes a lithography server, the lithography server operable to receive the sample image from each camera of the plurality of cameras.

### Claim 14 (depends on 13)

The method of claim 13, wherein each image projection system of the plurality of image projection systems includes an image projection computer, the image projection computer operable to send the sample image from each camera of the plurality of cameras to the lithography server.

### Claim 15 (depends on 14)

The method of claim 14, wherein the image projection computer is operable to instruct the camera to capture the sample image in the FOV of each camera of the plurality of cameras.

### Claim 16

A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause a computer system to perform the steps of:
capturing an image in a field of view (FOV) of each camera of a plurality of cameras of a processing unit of a lithography system, the processing unit having a plurality of image projection systems, each image projection system of the plurality of image projection systems including at least one of each camera of the plurality of cameras, the image including one or more real objects;
transferring the image in the FOV to a server in communication with the lithography system, the server including a deep learning (DL) module, the DL module including a deep learning (DL) model, a template searcher, and an affine modeler;
identifying a real class and a real location of the real objects in the FOV;
locating a template box within the FOV, the template box located based on object matching with a template, the template including a plurality of ideal objects; and
fitting a shift model and a rotation model to a location difference between the plurality of ideal objects and the real objects within the FOV.

### Claim 17 (depends on 16)

The non-transitory computer-readable medium of claim 16, wherein the lithography system further includes a lithography server, the lithography server operable to receive the image from each camera of the plurality of cameras.

### Claim 18 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein each image projection system of the plurality of image projection systems includes an image projection computer, the image projection computer operable to send the image from each camera of the plurality of cameras to the lithography server.

### Claim 19 (depends on 18)

The non-transitory computer-readable medium of claim 18, wherein the image projection computer is operable to instruct the camera to capture the image in the FOV of each camera of the plurality of cameras.

### Claim 20 (depends on 13)

The non-transitory computer-readable medium of claim 13, wherein the one or more real objects correspond to a resulting pattern based on a mask pattern data, the resulting pattern formed in a photoresist by the processing unit.