Systems and methods for defect detection using image reconstruction

• US 10,416,087 B2
• Filed: 11/21/2016
• Issued: 09/17/2019
• Est. Priority Date: 01/01/2016
• Status: Active Grant

• Current Assignee: KLA-Tencor Corporation
• Original Assignee: KLA-Tencor Corporation
• Inventors: Zhang, Jing, Nesbitt, Jeremy, Chen, Grace Hsiu-Ling, Wallingford, Richard
• Primary Examiner(s): Werner, Brian
• Application Number:
  US15/357,888
• Publication Number:
  US 20170191945A1
• Time in Patent Office: 1,030 Days
• CPC Class Codes: G01N 2021/8883 involving the calculation of gauges, generating models
• CPC Class Codes: G01N 2021/8887 based on image processing techniques
• CPC Class Codes: G01N 21/8851 Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges G01N21/8806 and G01N21/93 - G01N21/95692 take precedence; optical measurement of dimensions G01B11/00; optical scanning G02B26/10; image transformation G06T3/00; computerised image enhancement G06T5/00; image processing per se for flaw detection G06T7/0002
• CPC Class Codes: G01N 21/9501 Semiconductor wafers manufacturing processes per se of semiconductor devices implementing a measuring step H01L22/10
• CPC Class Codes: G01N 21/95607 using a comparative method
• CPC Class Codes: G06T 2207/10004 Still image; Photographic image
• CPC Class Codes: G06T 2207/10056 Microscopic image
• CPC Class Codes: G06T 2207/20056 Discrete and fast Fourier transform, [DFT, FFT]
• CPC Class Codes: G06T 2207/20076 Probabilistic image processing
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/30148 Semiconductor; IC; Wafer
• CPC Class Codes: G06T 3/4053 based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• CPC Class Codes: G06T 7/0006 using a design-rule based approach
• CPC Class Codes: G06T 7/001 using an image reference approach

Abstract

An inspection system includes an illumination sub-system, a collection sub-system, and a controller. The illumination sub-system includes an illumination source configured to generate a beam of illumination and a set of illumination optics to direct the beam of illumination to a sample. The collection sub-system includes a set of collection optics to collect illumination emanating from the sample and a detector configured to receive the collected illumination from the sample. The controller is configured to acquire a test image of the sample, reconstruct the test image to enhance the resolution of the test image, and detect one or more defects on the sample based on the reconstructed test image.

Claims

1. An inspection system, comprising:

an illumination sub-system, comprising;

an illumination source configured to generate a beam of illumination; and

a set of illumination optics to direct the beam of illumination to a sample;

a collection sub-system, comprising;

a set of collection optics to collect illumination emanating from the sample; and

a detector configured to receive the collected illumination from the sample; and

a controller communicatively coupled to the detector, the controller including a memory device and one or more processors configured to execute program instructions configured to cause the one or more processors to;

acquire a test image of the sample;

estimating a point spread function of the inspection system as a transform of a combination of an illumination aperture and a collection aperture of the inspection system;

reconstruct the test image, based on the estimated point spread function, to enhance the resolution of the test image; and

detect one or more defects on the sample based on the reconstructed test image.

2. The inspection system of claim 1, wherein the estimating a point spread function of the inspection system as a transform of a combination of an illumination aperture and a collection aperture of the inspection system comprises:

estimating the point spread function of the inspection system as a Fourier transform of a linear combination of the illumination aperture and the collection aperture of the inspection system.

3. The inspection system of claim 1, wherein reconstructing the test image comprises:

deconvolving the test image based on the estimated point spread function.

4. The inspection system of claim 3, wherein deconvolving the test image comprises:

deconvolving the test image using at least one of direct inversion, a Wiener filter, maximum likelihood deconvolution, or maximum a posterior deconvolution.

5. The inspection system of claim 4, wherein deconvolving the test image using maximum likelihood deconvolution comprises:

deconvolving the test image using noise having at least one of a Gaussian or a Poisson distribution.

6. The inspection system of claim 5, wherein deconvolving the test image using maximum likelihood deconvolution comprises:

deconvolving the test image using Richardson-Lucy deconvolution based on the estimated point spread function.

7. The inspection system of claim 4, wherein deconvolving the test image using maximum a posterior deconvolution comprises:

deconvolving the test image using noise having at least one of a Gaussian or a Poisson distribution.

8. The inspection system of claim 3, wherein a cost function associated with deconvolving the test image includes a sparse distribution regularization parameter.

9. The inspection system of claim 8, wherein a cost function associated with deconvolving the test image further includes an image gradient regularization parameter.

10. The inspection system of claim 9, wherein the image gradient regularization parameter comprises:

at least one of a Total Variation regularization parameter, a Tikhonov-Miller regularization parameter, or a Bilateral regularization parameter.

11. The inspection system of claim 9, wherein detecting one or more defects on the sample based on the reconstructed test image comprises:

acquiring a reference image of the sample;

deconvolving the reference image based on the point spread function;

generating a difference image between the deconvolved test image and the deconvolved reference image; and

detecting one or more defects on the sample based on the difference image.

12. The inspection system of claim 11, wherein a cost function associated with deconvolving the reference image includes the image gradient distribution regularization parameter.

13. The inspection system of claim 1, wherein acquiring the test image of the sample includes acquiring the test image of the sample using an anti-aliasing technique, wherein the anti-aliasing technique suppresses aliasing noise in the test image, wherein reconstructing the test image to enhance the resolution of the test image recovers one or more spatial frequencies associated with the sample impacted by the anti-aliasing technique.

14. The inspection system of claim 13, wherein reconstructing the test image comprises:

interpolating the test image.

15. The inspection system of claim 14, wherein interpolating the test image comprises:

interpolating the test image using at least one of bilinear interpolation, cubic interpolation, Lanczos interpolation, nearest neighbor interpolation, B-spline interpolation, or sinc interpolation.

16. The inspection system of claim 13, wherein reconstructing the test image comprises:

estimating the point spread function of the inspection system; and

deconvolving the test image based on the estimated point spread function.

17. The inspection system of claim 1, wherein detecting one or more defects on the sample based on the reconstructed test image includes reporting one or more locations of the one or more defects on the sample.

18. An inspection system, comprising:

an illumination sub-system, comprising;

an illumination source configured to generate a beam of illumination; and

a set of illumination optics to direct the beam of illumination to a sample;

a collection sub-system, comprising;

a set of collection optics to collect illumination emanating from the sample; and

a detector configured to receive the collected illumination from the sample; and

a controller communicatively coupled to the detector, the controller including a memory device and one or more processors configured to execute program instructions configured to cause the one or more processors to;

acquire a test image of the sample;

detect one or more defects on the sample based on the test image;

generate one or more patch images associated with the one or more detected defects on the sample;

estimating a point spread function of the inspection system as a transform of a combination of an illumination aperture and a collection aperture of the inspection system;

reconstruct the one or more patch images, based on the estimated point spread function, to enhance the resolution of the one or more patch images; and

classify the one or more defects based on the one or more reconstructed patch images.

19. The inspection system of claim 18, wherein reconstructing the test image comprises:

deconvolving the one or more patch images based on the estimated point spread function.

20. A multi-mode inspection system, comprising:

an illumination sub-system, comprising;

an illumination source configured to generate two or more modes of illumination; and

a set of illumination optics to sequentially direct the two or more modes of illumination to a sample;

a collection sub-system, comprising;

a set of collection optics to collect illumination emanating from the sample, wherein the set of collection optics includes an adjustable collection aperture, wherein the adjustable collection aperture is configured to generate two or more collection modes, wherein the illumination sub-system includes two or more system modes formed from the two or more modes of illumination and the two or more collection modes; and

a detector configured to receive the collected illumination from the sample; and

a controller communicatively coupled to the detector, the controller including a memory device and one or more processors configured to execute program instructions configured to cause the one or more processors to;

acquire two or more test images of the sample based on the two or more system modes;

estimate a point spread function of the inspection system as a transform of a combination of an illumination aperture and a collection aperture of the inspection system;

reconstruct the two or more test images, based on the estimated point spread function, to enhance the resolution of the two or more test images based on the estimated point spread function; and

identify one or more defects on the sample based on the two or more reconstructed test images.

21. The multi-mode inspection system of claim 20, wherein estimating a point spread function of the inspection system comprises:

estimating the point spread function of the inspection system as a Fourier transform of a linear combination of an illumination aperture and a collection aperture of the inspection system.

22. The multi-mode inspection system of claim 20, wherein reconstructing the two or more test image comprises:

deconvolving the two or more test image based on the estimated point spread function.

23. The multi-mode inspection system of claim 22, wherein deconvolving the test image comprises:

deconvolving the two or more test images using Richardson-Lucy deconvolution based on the estimated point spread function.

24. The inspection system of claim 23, wherein a cost function associated with deconvolving the two or more test images includes a sparse distribution regularization parameter.

25. The multi-mode inspection system of claim 23, wherein a cost function associated with deconvolving the two or more test images further includes an image gradient regularization parameter.

26. The multi-mode inspection system of claim 25, wherein the image gradient regularization parameter comprises:

at least one of a Total Variation regularization parameter, a Tikhonov-Miller regularization parameter, or a Bilateral regularization parameter.

27. The multi-mode inspection system of claim 26, wherein detecting the one or more defects on the sample based on the two or more reconstructed test images comprises:

combining the two or more reconstructed test images into a combined reconstructed test image; and

detecting the one or more defects based on the combined reconstructed test image.

28. The multi-mode inspection system of claim 24, wherein identifying one or more defects on the sample based on the two or more reconstructed test images comprises:

detecting one or more defects on the sample based on the two or more reconstructed test images.

29. The multi-mode inspection system of claim 28, wherein identifying one or more defects on the sample based on the two or more reconstructed test images further comprises:

generating one or more patch images associated with the one or more detected defects on the sample; and

classifying the one or more defects based on the one or more patch images.

30. An inspection system, comprising:

an illumination sub-system, comprising;

an illumination source configured to generate a beam of illumination; and

a set of illumination optics to direct the beam of illumination to a sample;

a collection sub-system, comprising;

a set of collection optics to collect illumination emanating from the sample; and

a detector configured to receive the collected illumination from the sample; and

a controller communicatively coupled to the detector, the controller including a memory device and one or more processors configured to execute program instructions configured to cause the one or more processors to;

configure a defect detection recipe, wherein configuring the defect detection recipe comprises;

acquiring a configuration image of the sample;

estimating a point spread function of the inspection system as a transform of a combination of an illumination aperture and a collection aperture of the inspection system; and

reconstructing the configuration image of the sample, based on the estimated point spread function, to enhance the resolution of the configuration image; and

detect one or more defects on the sample based on the defect detection recipe.

31. The inspection system of claim 30, wherein configuring the defect detection recipe further comprises:

defining one or more care areas on the sample to be inspected based on the reconstructed configuration image.

32. The inspection system of claim 30, wherein configuring the defect detection recipe further comprises:

measuring a dynamic range of the reconstructed configuration image; and

adjusting a run-time parameter of the illumination source based on the measured dynamic range of the reconstructed configuration image.

33. The inspection system of claim 32, wherein the run-time parameter comprises:

a light level of the illumination source.

34. The inspection system of claim 30, wherein configuring the defect detection recipe further comprises:

training a design-based defect detection technique.

35. The inspection system of claim 34, wherein the design-based defect detection technique comprises:

at least one of a pixel-to-design alignment technique, a context-based imaging technique, or a template based imaging technique.