Defect monitoring in semiconductor device fabrication

1. A method of forming a device comprising:

providing a substrate containing at least a partially formed device thereon, wherein the device comprises at least one defect site;

acquiring a pixilated image of the defect site selected, wherein each pixel comprises a grey level value (GLV);

eliminating surrounding noises of the defect site;

identifying a point of the image as the center of the defect;

performing a plurality of iterations to exclude outer edge pixels surrounding the center of the defect image; and

categorizing the defect as a killer defect or a non-killer defect.

2. The method of claim 1 wherein eliminating surrounding noises comprises:

defining a bright noise threshold value (BNTV) in a bright voltage contrast (BVC) defect site, eliminating pixels having GLV less than the BNTV; or

defining a dark noise threshold value (DNTV) in a dark voltage contrast (DVC) defect site, eliminating pixels having GLV greater than the DNTV.

3. The method of claim 2 wherein the center of the defect comprises:

one pixel or adjacent pixels having the highest GLV in a BVC defect site; or

one pixel or adjacent pixels having the lowest GLV in a DVC defect site.

4. The method of claim 3 wherein the outer edge pixels comprises pixels with a drastic change in GLV with respect to the center of the defect.

5. The method of claim 4 wherein excluding outer edge pixels comprises:

defining an edge exclusion number (EEN) to indicate the number of iterations of edge exclusion processes to be performed,

wherein the EEN comprises a whole number except zero which is less than the radius from the center of the defect image.

6. The method of claim 4 further comprises:

defining a minimum number of pixels (MNP) remain after edge exclusion processing;

comparing the number of pixels remaining in the image with MNP after each iteration of edge exclusion; and

terminating the edge exclusion process if the number of pixels remaining in the image is equal to MNP.

7. The method of claim 4 further comprises:

defining a minimum number of pixels (MNP) remain after edge exclusion processing;

comparing the number of pixels remaining in the image with MNP after each iteration of edge exclusion;

terminating the edge exclusion process if the number of pixels remaining in the image is less than MNP; and

reinstating the pixels removed in the current iteration of edge exclusion processing.

8. The method of claim 7 wherein reinstating the pixels removed comprises:

reinstating the pixels with higher GLVs before the pixels with lower GLVs in a BVC defect site until the image has the number of pixels equal to MNP; or

reinstating the pixels with lower GLVs before the pixels with higher GLVs in a DVC defect site until the image has the number of pixels equal to MNP.

9. The method of claim 4 wherein the edge exclusion process is terminated when there is at least one pixel remaining.

10. The method of claim 1 wherein categorizing the defect comprises:

determining a defect grey level value (DGLV) for the defect site by averaging the GLVs of the remaining pixels;

determining a threshold DGLV defining an acceptable range for a killer defect; and

comparing the DGLV for the defect site with the threshold DGLV.

11. The method of claim 10 further comprises:

categorizing the defect as a killer defect if the DGLV is greater than the threshold DGLV and categorizing the defect as a non-killer defect if the DGLV is less than or equal to the threshold DGLV in a BVC defect site; or

categorizing the defect as a killer defect if the DGLV is less than the threshold DGLV and categorizing the defect as a non-killer defect if the DGLV is greater than or equal to the threshold DGLV in a DVC defect site.

12. The method of claim 1 wherein the GLV of the pixel is derived from subtracting a measured GLV by a reference GLV.

13. The method of claim 12 wherein the GLV of the pixel comprises:

a positive GLV relating to a bright voltage contrast (BVC); or

a negative GLV relating to a dark voltage contrast (DVC).

14. The method of claim 13 further comprises:

processing the image to determine outer edge pixels of the defect site,

wherein the outer edge pixels comprises pixels with a change in polarity in GLV.

15. The method of claim 14 wherein excluding outer edge pixels comprises:

defining an edge exclusion number (EEN) to indicate the number of iterations of edge exclusion processes to be performed,

wherein the EEN comprises a whole number except zero which is less than the radius from the center of the defect image.

16. The method of claim 15 wherein categorizing the defect comprises:

determining a defect grey level value (DGLV) for the defect site by averaging the GLVs of the remaining pixels;

determining a threshold DGLV defining an acceptable range for a killer defect; and

comparing the DGLV for the defect site with the threshold DGLV.

17. A method of forming a device comprising:

providing a substrate containing at least a partially formed device thereon, wherein the device comprises at least one defect site;

acquiring a pixilated image of the defect site selected, wherein each pixel comprises a grey level value (GLV);

eliminating surrounding noises of the defect site;

identifying a point of the image as the center of the defect wherein the center of the defect comprises

one pixel or adjacent pixels having the highest GLV in a bright voltage contrast (BVC) defect site, or

one pixel or adjacent pixels having the lowest GLV in a dark voltage contrast (DVC) defect site;

performing a plurality of iterations to exclude outer edge pixels surrounding the center of the defect image; and

categorizing the defect as a killer defect or a non-killer defect.

18. The method of claim 17 wherein the outer edge pixels comprises pixels with a drastic change in GLV with respect to the center of the defect.

19. The method of claim 18 wherein categorizing the defect comprises:

determining a defect grey level value (DGLV) for the defect site by averaging the GLVs of the remaining pixels;

determining a threshold DGLV defining an acceptable range for a killer defect;

comparing the DGLV for the defect site with the threshold DGLV; and

categorizing the defect as a killer or non-killer defect which further comprises

categorizing the defect as a killer defect if the DGLV is greater than the threshold DGLV and categorizing the defect as a non-killer defect if the DGLV is less than or equal to the threshold DGLV in a BVC defect site, or

categorizing the defect as a killer defect if the DGLV is less than the threshold DGLV and categorizing the defect as a non-killer defect if the DGLV is greater than or equal to the threshold DGLV in a DVC defect site.

20. A system for monitoring defects comprises:

means for providing a substrate containing at least a partially formed device thereon, wherein the device comprises at least one defect site;

means for acquiring a pixilated image of the defect site selected, wherein each pixel comprises a grey level value (GLV);

means for eliminating surrounding noises of the defect site;

means for identifying a point of the image as the center of the defect;

means for performing a plurality of iterations to exclude outer edge pixels surrounding the center of the defect image; and

means for categorizing the defect as a killer defect or a non-killer defect.