Convolutional neural network-based mode selection and defect classification for image fusion
US-10115040-B2 · Oct 30, 2018 · US
US10867108B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-10867108-B2 |
| Application number | US-201916250128-A |
| Country | US |
| Kind code | B2 |
| Filing date | Jan 17, 2019 |
| Priority date | Sep 18, 2018 |
| Publication date | Dec 15, 2020 |
| Grant date | Dec 15, 2020 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
According to some embodiments, the present disclosure provides a method for determining wafer inspection parameters. The method includes identifying an area of interest in an IC design layout, performing an inspection simulation on the area of interest by generating a plurality of simulated optical images from the area of interest using a plurality of optical modes, and selecting, based on the simulated optical images, at least one of the optical modes to use for inspecting an area of a wafer that is fabricated based on the area of interest in the IC design layout.
Opening claim text (preview).
What is claimed is: 1. A method for determining wafer inspection parameters, the method comprising: identifying an area of interest in an integrated circuit (IC) design layout, wherein a specific defect of interest (DOI) is in the area of interest; performing a plurality of inspection simulations on the area of interest using a plurality of optical modes to thereby generate a plurality of simulated optical images, wherein each optical mode has at least one optical parameter value that differs from other optical modes within the plurality of optical modes; generating a plurality of image attributes from each simulated optical image from the plurality of optical images; selecting a preferred optical mode for the specific DOI from the plurality of optical modes based on a comparison of the plurality of image attributes from each simulated optical image from the plurality of simulated optical images, wherein the plurality of image attributes of the preferred optical mode identifies the specific DOI better than the plurality of image attributes associated with any other optical mode; and after performing the plurality of inspection simulations, inspecting the area of interest on a wafer using the selected optical mode. 2. The method of claim 1 , wherein, during the plurality of inspection simulations on the area of interest, each simulated optical image is generated using a respective optical mode by performing the steps of: generating a first image from the area of interest without the DOI using the optical mode; generating a second image from the area of interest with the DOI defect using the optical mode; and generating the simulated optical image as a difference image by subtracting one of the first and second images from the other. 3. The method of claim 2 , wherein a type of the DOI and a location of the DOI in the area of interest are specified. 4. The method of claim 3 , wherein the area of interest surrounding the DOI has a size of about 200×200 um ∧2 or less. 5. The method of claim 3 , further comprising adding data relevant to the DOI to a database, wherein the data relevant to the DOI includes the selected optical mode, the type of the DOI, the location of the DOI, and the location of the area of interest. 6. The method of claim 1 , wherein each of the plurality of optical modes include optical parameters selected from the group consisting of: pixel size, optical wavelength, aperture shape, optical polarization, and focus setting. 7. The method of claim 1 , wherein the one or more attributes include a signal to noise (S/N) ratio. 8. The method of claim 1 , wherein the selecting the preferred optical mode includes selecting optical mode from the plurality of optical modes corresponding to a first image attribute value that is higher than the first image attribute value of each of the other of the plurality of simulated optical images. 9. A non-transitory machine-readable medium having stored thereon instructions for performing a method comprising machine executable code which when executed by at least one machine, causes the machine to: select a portion of an integrated circuit (IC) design layout, wherein a specific defect of interest (DOI) is in the portion of the IC design layout; perform a plurality of inspection simulations on the portion of the IC design layout, wherein the plurality of inspection simulations comprises using a plurality of optical modes to generate a plurality of simulated optical images from the portion of the IC design layout, wherein each optical mode has at least one optical parameter value that differs from other optical modes within the plurality of optical modes; determine one or more attributes from each simulated optical image of the simulated optical images; compare the one or more attributes of each optical image; and select a preferred optical mode for the specific DOI from the plurality of optical modes based on the comparison of the one or more attributes of each simulated optical image from the plurality of simulated optical images, wherein the preferred optical mode identifies the specific DOI better than any other optical mode based on the one or more attributes of the simulated optical image produced by the preferred optical mode. 10. The non-transitory machine-readable medium of claim 9 , further comprising an optical system configured to optically inspect an area of a wafer using the selected optical mode, wherein the area of the wafer is fabricated based at least in part on the portion of the IC design layout. 11. The non-transitory machine-readable medium of claim 9 , wherein the machine is further configured to designate the selected optical mode in a database as a preferred mode for inspecting the portion of the IC design layout. 12. The non-transitory machine-readable medium of claim 9 , wherein each optical mode of the plurality of optical modes includes optical parameters selected from the group consisting of: pixel size, optical wavelength, aperture shape, optical polarization, and focus setting. 13. The non-transitory machine-readable medium of claim 9 , wherein using an optical mode of the plurality of optical modes to generate a simulated optical image from the portion of the IC design layout comprises: generating a first image from the portion of the IC design layout without the DOI using the optical mode; generating a second image from the portion of the IC design layout with the DOI using the optical mode; and generating the simulated optical image as a difference image by subtracting one of the first and second images from the other. 14. The non-transitory machine-readable medium of claim 13 , wherein a type of the DOI and a location of the DOI are specified. 15. The inspection system of claim 9 , wherein the portion of the IC design layout is based on a 3D model, and wherein the inspection simulation further comprises using finite difference time-domain (FDTD) equations to solve Maxwell boundary conditions. 16. The non-transitory machine-readable medium of claim 9 , wherein evaluating each of the plurality of optical modes by comparing the plurality of attributes of each of the plurality of simulated optical images comprises using at least one of machine learning, random forest, and mixed model algorithms. 17. The non-transitory machine-readable medium of claim 9 , wherein the select the preferred optical mode includes selecting the optical mode corresponding to a first image attribute value that is higher than the first image attribute value of each of the other of the plurality of simulated optical images. 18. A method for retraining wafer inspection parameters, the method comprising: providing a database listing defects and their corresponding defect detection signals; performing a first optical inspection on an area of a wafer that has been fabricated based on an integrated circuit (IC) design layout, wherein the performing of the first optical inspection comprises using a first set of optical parameters to generate an optical image; generating a first defect detection signal from the optical image, wherein the first defect detection signal is not associated with any defects listed in the database; identifying a second defect detection signal in the database that is closest to the first defect detection signal; performing a second optical inspection of the area to identify a defect not listed in the database by using a second set of optical parameters associated with the second defect detection signal; and determining whether to modify values of the first set of optical par
Structural properties, e.g. testing or measuring thicknesses, line widths, warpage, bond strengths or physical defects · CPC title
comprising optical enhancement of defects or not-directly-visible states · CPC title
Yield analysis or yield optimisation · CPC title
Information management, e.g. software; Active and passive control, e.g. details of controlling exposure processes or exposure tool monitoring processes · CPC title
Defects, e.g. optical inspection of patterned layer for defects · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.