3D structure inspection or metrology using deep learning

Patent metadata

Field	Value
Publication number	US-11644756-B2
Application number	US-202117393979-A
Country	US
Kind code	B2
Filing date	Aug 4, 2021
Priority date	Aug 7, 2020
Publication date	May 9, 2023
Grant date	May 9, 2023

Abstract

Methods and systems for determining information for a specimen are provided. Certain embodiments relate to bump height 3D inspection and metrology using deep learning artificial intelligence. For example, one embodiment includes a deep learning (DL) model configured for predicting height of one or more 3D structures formed on a specimen based on one or more images of the specimen generated by an imaging subsystem. One or more computer systems are configured for determining information for the specimen based on the predicted height. Determining the information may include, for example, determining if any of the 3D structures are defective based on the predicted height. In another example, the information determined for the specimen may include an average height metric for the one or more 3D structures.

First claim

What is claimed is:

1. A system configured to determine information for a specimen, comprising:
an imaging subsystem configured to generate images of a specimen, wherein one or more three-dimensional structures are formed on the specimen;
one or more computer systems; and
one or more components executed by the one or more computer systems, wherein the one or more components comprise a deep learning model configured for predicting a height of the one or more three-dimensional structures based on one or more of the images; and
wherein the one or more computer systems are configured for:
locating and isolating portions of the images corresponding to the one or more three-dimensional structures, wherein the images input to the deep learning model comprise only the isolated portions of the images corresponding to the one or more three-dimensional structures; and
determining information for the specimen based on the predicted height of the one or more three-dimensional structures.

2. The system of claim 1 , wherein the one or more three-dimensional structures are one or more bumps formed on a wafer.

3. The system of claim 1 , wherein determining the information comprises determining if any of the one or more three-dimensional structures are defective.

4. The system of claim 1 , wherein the information comprises an average height metric for the one or more three-dimensional structures.

5. The system of claim 1 , wherein the images input to the deep learning model are collected by the imaging subsystem in a single pass of the specimen.

6. The system of claim 1 , wherein the images input to the deep learning model are collected by the imaging subsystem at a single focus value.

7. The system of claim 1 , wherein the images comprise bright field images of the specimen or dark field images of the specimen.

8. The system of claim 1 , wherein the images comprise bright field images of the specimen and dark field images of the specimen.

9. The system of claim 1 , wherein the one or more computer systems are further configured for training the deep learning model with the images generated by the imaging subsystem of the specimen or a different specimen with two or more focus offsets.

10. The system of claim 1 , wherein the one or more computer systems are further configured for training the deep learning model with images generated by the imaging subsystem of a different specimen having the three-dimensional structures formed thereon with multiple, known values of a characteristic of the three-dimensional structures.

11. The system of claim 1 , wherein the one or more computer systems are further configured for locating and isolating one or more of the portions of the images corresponding to the one or more three-dimensional structures, respectively, and generating individual one or more cropped patch images for individual one or more three-dimensional structures, respectively, based on the isolated one or more portions.

12. The system of claim 1 , wherein the one or more computer systems are further configured for said locating and isolating the portions of the images corresponding to the one or more three-dimensional structures by template matching.

13. The system of claim 1 , wherein the one or more computer systems are further configured for said locating and isolating the portions of the images corresponding to the one or more three-dimensional structures based on design information for the specimen.

14. The system of claim 1 , wherein the one or more computer systems are further configured for said locating and isolating the portions of the images corresponding to the one or more three-dimensional structures by inputting the images into a YOLO network configured for the locating and included in the one or more components executed by the one or more computer systems.

15. The system of claim 1 , wherein the one or more computer systems are further configured for said locating and isolating the portions of the images corresponding to the one or more three-dimensional structures by inputting the images into an additional deep learning model configured for the locating and included in the one or more components executed by the one or more computer systems.

16. The system of claim 1 , wherein the deep learning model is further configured as a convolutional neural network.

17. The system of claim 1 , wherein the deep learning model comprises a combination of convolution layers and fully connected layers.

18. The system of claim 1 , wherein the deep learning model is further configured as an AlexNet.

19. The system of claim 1 , wherein the deep leaning model is further configured as a YOLO network, and wherein the YOLO network is further configured for said locating and isolating the portions of the images corresponding to the one or more three-dimensional structures.

20. The system of claim 1 , wherein the imaging subsystem is further configured as an inspection subsystem.

21. The system of claim 1 , wherein the imaging subsystem is further configured as a metrology subsystem.

22. The system of claim 1 , wherein the imaging subsystem is further configured as a light based subsystem.

23. The system of claim 1 , wherein the imaging subsystem is further configured as an electron beam subsystem.

24. The system of claim 1 , wherein the specimen is a wafer.

25. A non-transitory computer-readable medium, storing program instructions executable on one or more computer systems for performing a computer-implemented method for determining information for a specimen, wherein the computer-implemented method comprises:
generating images of a specimen with an imaging subsystem, wherein one or more three-dimensional structures are formed on the specimen;
locating and isolating portions of the images corresponding to the one or more three-dimensional structures;
predicting a height of the one or more three-dimensional structures based on one or more of the images by inputting the one or more of the images into a deep learning model included in one or more components executed by the one or more computer systems, wherein the images input to the deep learning model comprise only the isolated portions of the images corresponding to the one or more three-dimensional structures; and
determining information for the specimen based on the predicted height of the one or more three-dimensional structures.

26. A computer-implemented method for determining information for a specimen, comprising:
generating images of a specimen with an imaging subsystem, wherein one or more three-dimensional structures are formed on the specimen;
locating and isolating portions of the images corresponding to the one or more three-dimensional structures;
predicting a height of the one or more three-dimensional structures based on one or more of the images by inputting the one or more of the images into a deep learning model included in one or more components executed by one or more computer systems, wherein the images input to the deep learning model comprise only the isolated portions of the images corresponding to the one or more three-dimensional structures; and
determining information for the specimen based on the predicted height of the one or more three-dimensional structures, wherein the locating and isolating and the determining are performed by the one or more computer systems.

Assignees

• Kla Corp

Inventors

• Young Scott A
• Bhaskar Kris
• Nicolaides Lena

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06T2207/30148

  Semiconductor; IC; Wafer · CPC title

• G06T2207/30141

  Printed circuit board [PCB] · CPC title

• G06T2207/20084

  Artificial neural networks [ANN] · CPC title