Defect classification by fitting optical signals to a point-spread function

Patent metadata

Field	Value
Publication number	US-10957035-B2
Application number	US-201916355584-A
Country	US
Kind code	B2
Filing date	Mar 15, 2019
Priority date	Nov 30, 2018
Publication date	Mar 23, 2021
Grant date	Mar 23, 2021

Abstract

A semiconductor die is inspected using an optical microscope to generate a test image of the semiconductor die. A difference image between the test image of the semiconductor die and a reference image is derived. For each defect of a plurality of defects for the semiconductor die, a point-spread function is fit to the defect as indicated in the difference image and one or more dimensions of the fitted point-spread function are determined. Potential defects of interest in the plurality of defects are distinguished from nuisance defects, based at least in part on the one or more dimensions of the fitted point-spread function for respective defects of the plurality of defects.

First claim

What is claimed is:

1. A non-transitory computer-readable storage medium storing one or more programs for execution by one or more processors of a semiconductor-inspection system that includes an optical microscope, the one or more programs comprising instructions for:
deriving a difference image between a test image of a semiconductor die and a reference image, wherein the test image is generated by inspecting the semiconductor die using the optical microscope;
identifying a defect in the difference image;
fitting a point-spread function to the identified defect as indicated in the difference image and determining one or more dimensions of the fitted point-spread function; and
based at least in part on the one or more dimensions of the fitted point-spread function for the identified defect, determining whether the identified defect is a defect of interest or a nuisance defect.

2. The computer-readable storage medium of claim 1 , wherein:
the point-spread function is a two-dimensional Gaussian function; and
the one or more dimensions comprise a first dimension indicative of a width of the fitted Gaussian function in a first direction and a second dimension indicative of a width of the fitted Gaussian function in a second direction.

3. The computer-readable storage medium of claim 2 , wherein:
the first dimension is selected from the group consisting of a standard deviation of the fitted Gaussian function in the first direction and a full width at half maximum of the fitted Gaussian function in the first direction; and
the second dimension is selected from the group consisting of a standard deviation of the fitted Gaussian function in the second direction and a full width at half maximum of the fitted Gaussian function in the second direction.

4. The computer-readable storage medium of claim 1 , wherein the one or more dimensions comprise:
a first distance between maximal gradients of the fitted point-spread function in a first direction; and
a second distance between maximal gradients of the fitted point-spread function in a second direction.

5. The computer-readable storage medium of claim 1 , wherein the one or more dimensions comprise:
a first area under a cross-section of the fitted point-spread function in a first direction through a maximum of the fitted point-spread function, normalized by the height of the maximum; and
a second area under a cross-section of the fitted point-spread function in a second direction through the maximum, normalized by the height of the maximum.

6. The computer-readable storage medium of claim 1 , wherein the point-spread function is a sinc function.

7. The computer-readable storage medium of claim 1 , wherein the point-spread function is a polynomial function.

8. The computer-readable storage medium of claim 1 , wherein the instructions for fitting the point-spread function to the defect comprise instructions for fitting a numeric simulation of the point-spread function to the defect.

9. The computer-readable storage medium of claim 1 , wherein the instructions for determining whether the identified defect is a defect of interest or a nuisance defect comprise instructions for:
determining whether a first dimension of the one or more dimensions of the fitted point-spread function exceeds, or equals or exceeds, a first threshold; and
in response to determining that the first dimension exceeds, or equals or exceeds, the first threshold, classifying the identified defect as a nuisance defect.

10. The computer-readable storage medium of claim 9 , wherein the instructions for determining whether the identified defect is a defect of interest or a nuisance defect further comprise instructions for:
determining whether the first dimension is less than, or less than or equal to, a second threshold, wherein the second threshold is less than the first threshold; and
in response to determining that the first dimension is less than, or less than or equal to, the second threshold, classifying the identified defect as a nuisance defect.

11. The computer-readable storage medium of claim 1 , wherein the one or more dimensions of the fitted point-spread function comprise two dimensions, and the instructions for determining whether the identified defect is a defect of interest or a nuisance defect comprise instructions for:
determining whether a metric that is a function of the two dimensions satisfies a threshold; and
in response to a determination that the metric does not satisfy the threshold, classifying the identified defect as a nuisance defect.

12. The computer-readable storage medium of claim 11 , wherein the instructions for determining whether the identified defect is a defect of interest or a nuisance defect comprise instructions for providing the one or more dimensions of the fitted point-spread function for the identified defect as input to a machine-learning algorithm trained to distinguish potential defects of interest from nuisance defects.

13. The computer-readable storage medium of claim 11 , wherein:
the one or more programs further comprise instructions for determining a goodness of fit of the fitted point-spread function;
the goodness of fit is a sum of squares of the residuals for the fitted point-spread function or a largest single residual for the fitted point-spread function; and
determining whether the identified defect is a defect of interest or a nuisance defect is further based at least in part on the goodness of fit.

14. The computer-readable storage medium of claim 11 , wherein the identified defect is not resolved in the difference image.

15. The computer-readable storage medium of claim 11 , wherein the one or more programs further comprise instructions for generating a report specifying defects of interest.

16. A semiconductor-inspection system, comprising:
an optical microscope;
one or more processors; and
memory storing one or more programs for execution by the one or more processors, the one or more programs including instructions for:
deriving a difference image between a test image of a semiconductor die and a reference image, wherein the test image is generated by inspecting the semiconductor die using the optical microscope;
identifying a defect in the difference image;
fitting a point-spread function to the identified defect as indicated in the difference image and determining one or more dimensions of the fitted point-spread function; and
based at least in part on the one or more dimensions of the fitted point-spread function for the identified defect respective defects, determining whether the identified defect is a defect of interest or a nuisance defect.

17. A method, comprising:
inspecting a semiconductor die using an optical microscope to generate a test image of the semiconductor die; and
in a computer system comprising one or more processors and memory storing instructions for execution by the one or more processors:
deriving a difference image between the test image of the semiconductor die and a reference image;
identifying a defect in the difference image;
fitting a point-spread function to the identified defect as indicated in the difference image and determining one or more dimensions of the fitted point-spread function; and
based at least in part on the one or more dimensions of the fitted point-spread function for the identified defect, determining whether the identified defect is a defect of interest or a nuisance defect.

18. A non-transitory computer-readable storage medium storing one or more programs for execution by one or more proc

Assignees

• Kla Tencor Corp
• Kla Corp

Inventors

• Konecky Soren
• Brauer Bjorn

Classifications

• H10P74/203 Primary

  Structural properties, e.g. testing or measuring thicknesses, line widths, warpage, bond strengths or physical defects · CPC title

• H10P74/27

  Structural arrangements therefor · CPC title

• H10P72/06

  Apparatus for monitoring, sorting, marking, testing or measuring · CPC title

• G06T2207/30148

  Semiconductor; IC; Wafer · CPC title

• G06T2207/10056

  Microscopic image · CPC title