System to identify a wafer manufacturing problem and method therefor

• US 6,885,977 B2
• Filed: 12/20/2002
• Issued: 04/26/2005
• Est. Priority Date: 12/20/2002
• Status: Expired due to Term

• Current Assignee: Applied Materials Incorporated
• Original Assignee: Applied Materials Incorporated
• Inventors: Gavra, Yifah, Dor, Amos
• Primary Examiner(s): Tsai, Carol S. W.
• Application Number:
  US10/327,174
• Publication Number:
  US 20040122859A1
• Time in Patent Office: 858 Days
• Field of Search: 702182-185, 702/35, 700109-110, 382/145, 382/149, 382/141, 714/732, 714/737, 438/14, 356/237.1, 356/237.2, 356/237.3, 356/237.4
• US Class Current: 702/185
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: H01L 22/20 Sequence of activities consisting of a plurality of measurements, corrections, marking or sorting steps

Abstract

A method (400) for determining a response to a wafer manufacturing process problem includes the steps of generating a wafer map signature (406) containing defect attributes to be identified and retrieving (410) one or more wafer map signatures from a database. The wafer map signature to be identified is matched (414) with the one or more wafer map signatures from the database in order to help find a root cause of a defect in a wafer manufacturing process.

This provides at least the advantage that a signature matching capability allows an Operator to more accurately identify a root manufacturing cause of a wafer defect. Furthermore, instant feedback to the wafer manufacturing process can be provided to facilitate rapid correction of manufacturing errors.

Claims

1. A method for determining a response to a wafer manufacturing process problem, the method comprising the steps of:

generating a wafer map signature corresponding to a wafer that contains defect attributes associated with a wafer manufacturing process;

retrieving one or more stored wafer map signatures from a database, wherein each of the one or more stored wafer man signatures include a description of process tools and/or equipment that the wafer was exposed to prior to signature capture;

matching said wafer map signature with said one or more stored wafer map signatures from said database;

calculating a conditional probability distribution of a process tool type or a determined wafer defect type based on the wafer map signature to be identified; and

combining the calculation with a wafer map signature similarity metric to arrive at a probabilistic recommendation of the wafer manufacturing process to be corrected.

2. The method for determining a response to a wafer manufacturing process problem according to claim 1, further comprising the step of:

determining, for a number of said retrieved wafer map signatures, a similarity metric relating to a signature similarity between said retrieved wafer map and said wafer map signature to be identified.

3. The method for determining a response to a wafer manufacturing process problem according to claim 1, the method further comprising the step of adding a new case to a database of cases if said step of matching fails to yield an acceptable match, wherein said wafer map signature is one defect attribute of a case.

4. The method for determining a response to a wafer manufacturing process problem according to claim 3, the method further comprising the step of applying one or more thresholds to said step of matching in order to select a subset of said stored wafer map signatures that exhibit similar defect attributes to said wafer map signature to be identified.

5. The method for determining a response to a wafer manufacturing process problem according to claim 1, wherein said step of generating a wafer map signature includes the step of automatically generating a wafer map signature based on defect attributes of said wafer map signature to be identified.

6. The method for determining a response to a wafer manufacturing process problem according to claim 1, the method further comprising the preceding steps of storing and selecting a plurality of cases in a database, wherein the plurality of cases include wafer map signatures, a number of which are to be retrieved in said step of matching.

7. The method for determining a response to a wafer manufacturing process problem according to claim 1, the method further comprising the step of storing said wafer map signature to be identified in a database.

8. The method for determining a response to a wafer manufacturing process according to claim 1, the method further comprising the step of grading a result of said step of matching to enable a most similar case that exhibits a most similar wafer map signature to be identified.

9. The method for determining a response to a wafer manufacturing process problem according to claim 1, the method further comprising the step of estimating a manufacturing process problem associated with said wafer map signature based on a previously identified wafer manufacturing process problem associated with said closest matching wafer map signature.

10. The method for determining a response to a wafer manufacturing process problem according to claim 9, the method further comprising the step of adapting a manufacturing process based on said step of estimating.

11. The method for determining a response to a wafer manufacturing process problem according to claim 1, the method further comprising the step of indexing said wafer map signatures to organize the signature features in the database to facilitate rapid retrieval.

12. The method for determining a response to a wafer manufacturing process problem according to claim 1, wherein said step of matching a wafer map signature further includes matching at least one of product defects, substrate differences, and imaging modality in order to determine a set of closest matches.

13. The method for determining a response to a wafer manufacturing process problem according to claim 1, wherein said step of generating a wafer map signature includes generating a signature of a bare wafer, a patterned wafer, a sawn wafer, a whole wafer, and a multi-chip module.

14. The method for determining a response to a wafer manufacturing process problem according to claim 1, wherein the method further includes the step of:

defining at least one threshold value based on an analysis of said wafer map signature of said wafer to be identified.

15. The method for determining a response to a wafer manufacturing process problem according to claim 14, wherein said at least one threshold value is pre-determined for a particular wafer type.

16. The method for determining a response to a wafer manufacturing process problem according to claim 15, wherein said at least one threshold value is programmed into a computer to he used for automatically determining a response to a wafer manufacturing process problem.

17. The method for determining a response to a wafer manufacturing process problem according to claim 14, the method further comprising the step of:

applying at least one threshold value to said similarity metric to determine a set of closest matches between said wafer map signature to be identified and said retrieved wafer map signature.

18. The method for determining a response to a wafer manufacturing process problem according to claim 1, wherein the step of generating a wafer map signature is based on at least one of the number of defects, defect positions, cluster shape, surrounding density, and orientation, bins.

19. The method for determining a response to a wafer manufacturing process problem according to claim 1, wherein the step of matching employs a nearest neighbor algorithm.

20. The method for determining a response to a wafer manufacturing process problem according to claim 1, wherein the step of retrieving includes retrieving a set of wafer map signatures that are similar in structural content and/or visual appearance to said wafer map signature to be identified.

21. A system to identify a manufacturing problem adapted to perform the method steps of claim 1.

22. A storage medium storing processor-implementable instructions and/or data for controlling a processor to carry out the method of claim 1.

23. The storage medium storing processor-implementable instructions and/or data according to claim 22, wherein the processor-implementable instructions and/or data includes at least one of the following:

(i) A signature creator to generate a wafer map signature indicative of a wafer manufacturing process problem;

(ii) A signature storage algorithm;

(iii) A signature retrieval function operably coupled to a database for retrieving a wafer map signature from the database corresponding to the wafer manufacturing process problem;

(iv) A signature-matching engine to select the closest set of wafer map signatures to the signature being searched;

(v) An algorithm to set or adjust confidence or similarity metrics, and/or related threshold levels;

or

(vi) Information relating to the product, wafer type or layer of a wafer that has been used to generate the wafer map.

24. A system to identify a manufacturing problem comprising:

a signature creation function configured to create a wafer map signature corresponding to a wafer that contains defect attributes to be identified and associated with a wafer manufacturing process;

a database, operably coupled to said signature creation function and configured to store a plurality of wafer map signatures;

a signature retrieval function, operably coupled to said signature creation function and configured to retrieve one or more wafer map signatures from said database, wherein each of the one or more stored wafer map signatures include a description of at least one of process tools and equipment that the wafer was exposed to prior to signature capture; and

a signature matching function, operably coupled to said signature retrieval function and configured to match said wafer map signature to be identified with said one or more wafer map signatures retrieved from said database, to calculate a confidence metric based on a similarity between said wafer map signature to be identified with said one or more wafer map signatures retrieved from said database, to calculate a conditional probability distribution of a process tool type or a determined wafer defect type based on the wafer man signature to be identified, and to combine the calculation with the confidence metric to arrive at a probabilistic recommendation of the wafer manufacturing process to be corrected.

25. The system to identify a manufacturing problem according to claim 24, wherein one or more signal processors is operably coupled to said database to perform one or more of:

the signature creation function, signature retrieval function and/or signature matching function.

26. The system to identify a manufacturing problem according to claim 25, wherein one or more signal processors calculates a wafer map signature similarity between said retrieved wafer map signature and said wafer map signature to be identified.

27. The system to identify a manufacturing problem according to claim 25, wherein said one or more signal processors applies a threshold in said matching function to select a subset of said stored wafer map signatures that exhibit similar defect attributes to said wafer map signature to be identified.

28. The system to identify a manufacturing problem according to claim 25, wherein one of said one or more signal processors stores said wafer map signature to be identified in said database.

29. The system to identify a manufacturing problem according to claim 25, wherein one of said one or more signal processors estimates a wafer manufacturing fault associated with said wafer to be identified, said wafer manufacturing fault including at least one of product defects, substrate differences, imaging modality, and is based on a wafer manufacturing fault associated with said retrieved wafer map signature that closely matches said wafer map signature to be identified.

30. The system to identify a manufacturing problem according to claim 24, wherein said database is indexed to facilitate rapid retrieval of said wafer map signatures.

31. The system to identify a manufacturing problem according to claim 24, wherein said signature creation function creates a wafer map signature of defect attributes to be identified, where said wafer is one of a bare wafer, a patterned wafer, a sawn wafer, a whole wafer, and a multi-chip module.

32. The system to identify a manufacturing problem according to claim 24, wherein said signature matching function applies at least one threshold value to said confidence metric calculation to determine which wafer map signatures are to be retrieved from said database by said signature retrieval function.

33. The system to identify a manufacturing problem according to claim 24, wherein said signature matching function employs a nearest neighbor algorithm.

34. The system to identify a manufacturing problem according to claim 24, wherein said signature creation function creates a wafer map signature based on at least one of a number of defects, defect positions, cluster shape, surrounding density, orientation, and bins.

35. The system to identify a manufacturing problem according to claim 24, wherein said signature creation function creates a wafer map signature containing a description of process tools and/or equipment that said wafer was exposed to prior to signature creation.

36. The system to identify a manufacturing problem according to claim 25, wherein said one or more signal processors adds a new case to said database of cases if said matching function fails to yield an acceptable match, wherein said wafer map signature is one defect attribute of a case.