Method and apparatus for macro defect detection using scattered light

• US 5,859,698 A
• Filed: 05/07/1997
• Issued: 01/12/1999
• Est. Priority Date: 05/07/1997
• Status: Expired due to Fees

• Current Assignee: Nikon Research Corporation Of America (Nikon Corporation)
• Original Assignee: Nikon Corporation
• Inventors: Chau, Henry K., Aiyer, Arun A.
• Primary Examiner(s): Pham, Hoa Q.
• Application Number:
  US08/852,849
• Time in Patent Office: 615 Days
• Field of Search: 356/237, 356/239, 356/394, 356/371, 382/149, 382/218, 382/259, 382/108, 382/281, 382/141, 348/125, 348/87
• US Class Current: 356/237.2
• CPC Class Codes: G01N 21/95607 using a comparative method

Abstract

Macro defects in a processed or partly processed semiconductor wafer, liquid crystal display element, disk drive element or the like, are detected using scattered light. By use of automated image processing techniques, a reference image and a sample image are formed from the scattered light and edge enhanced. A difference image is formed by comparing the edge enhanced reference and sample images. The difference image is evaluated using one or more automated image processing techniques such as thresholding, morphological transformations and blob analysis to identify macro defects.

Claims

1. A method for detecting macro defects on a sample surface comprising the steps of:

forming a first edge enhanced image of a reference;

directing light from a light source onto said sample surface at an oblique angle wherein said sample surface corresponds to said reference;

detecting light from said light source scattered by said sample surface;

forming a test image of said sample surface from the scattered light;

performing an edge enhancement of said image to form a second edge enhanced image;

forming a difference image by subtracting said first edge enhanced image from said second enhanced image;

forming a thresholded difference image by applying a pre-selected threshold method to said difference image, wherein said thresholded difference image further comprises an array of pixels; and

determining a presence or absence of macro defects wherein each pixel or contiguous group of pixels, of said thresholded difference image having a gray scale value in excess of a value determined by said threshold method is defined as a macro defect.

2. The method of claim 1, wherein said reference comprises a regular pattern of light scattering features on a reference surface that represent a desired pattern.

3. The method of claim 2, wherein said first edge enhanced image is formed from an image of said reference surface.

4. The method of claim 2, wherein forming said first edge enhanced image comprises the steps of:

directing light from said light source onto said reference surface at said oblique angle;

detecting light from said light source scattered by said reference surface;

forming an image of said reference surface from the scattered light; and

performing an edge enhancement of said image.

5. The method of claim 4, wherein forming said image of said reference surface comprises forming a plurality of temporary images, each temporary image formed of a different reference surface, and averaging said plurality of temporary images.

6. The method of claim 4, wherein forming said image of said reference surface comprises the steps of:

forming one or more first temporary images, each first temporary image formed of a different reference surface;

forming a second temporary image, wherein said second temporary image is formed of a database representation of said reference surface; and

averaging said one or more first temporary images and said second temporary image.

7. The method of claim 1, wherein said reference is a database representation of a regular pattern of light scattering features that represent a desired pattern.

8. The method of claim 1 wherein forming a difference image comprises normalizing said second edge enhanced image to said first edge enhanced image.

9. The method of claim 1 wherein determining a presence or absence of macro defects further comprises performing a morphological operation of said thresholded difference image.

10. The method of claim 1 wherein determining a presence or absence of macro defects further comprises performing a direct or differential blob analysis.

11. A method of scattered light macro defect detection comprising the steps of:

training a macro defect detection apparatus, wherein said training comprises forming a reference template image;

acquiring a sample image of a sample surface from scattered light with a camera;

forming an edge enhanced image of said sample image;

normalizing said edge enhanced image to said reference template image;

forming a difference image of said normalized edge enhanced image and said reference template;

forming a thresholded difference image of said difference image;

performing a direct blob analysis of said thresholded difference image or a differential blob analysis of said difference image and said thresholded difference image; and

determining a presence or absence of macro defects using said direct or differential Blob analysis and thresholded difference image;

wherein forming said edge enhanced image, normalizing said edge enhanced image, forming said difference image, forming said thresholded difference image, performing said direct or differential blob analysis and said determining a presence or absence of macro defects is performed by said macro defect detection apparatus.

12. The method of claim 11, wherein forming said reference template image comprises forming an edge enhanced image of a reference surface.

13. The method of claim 12, wherein forming said edge enhanced image comprises substituting a database representation of a regular pattern of light scattering features that are not macro defects for said reference surface.

14. The method of claim 12, wherein forming said edge enhanced image comprises the steps of:

forming a plurality of temporary images wherein each temporary image is formed of a different reference surface; and

forming an average image of the plurality of temporary images, wherein said edge enhanced image is formed of said average image.

15. The method of claim 11, wherein acquiring a sample image of a sample surface comprises illuminating said sample surface with a light source positioned at an oblique angle to said sample surface.

16. An apparatus to inspect for macro defects on a substrate comprising:

a support for holding said substrate;

a light source for directing light onto a surface of said substrate when said substrate is on said support;

a detector located to receive light scattered from said surface of said substrate; and

an automated processor coupled to said detector for comparing an edge enhanced image of said surface to an edge enhanced reference template, thereby to determine a presence or absence of macro defects on the surface of the substrate.

17. The apparatus of claim 16, further comprising at least one additional detector located to receive scattered light from the surface of the substrate, the additional detector being located to detect scattered light at a different angle than is the first detector.

18. The apparatus of claim 16, further comprising a mechanism coupled to the substrate for moving the substrate relative to the detector.

19. The apparatus of claim 18, wherein the support includes means for rotating the substrate in a plane defined by its surface.

20. The apparatus of claim 16, wherein the detector comprises a charged coupled device camera.