Intelligent photomask disposition

• US 6,526,164 B1
• Filed: 05/27/1999
• Issued: 02/25/2003
• Est. Priority Date: 05/27/1999
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Wong, Alfred Kwok-Kit, Mansfield, Scott Marshall
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): CHAWAN, SHEELA C
• Application Number:
  US09/320,901
• Time in Patent Office: 1,370 Days
• Field of Search: 382/144, 382/147, 382/148, 382/149, 356/237.47, 356/390, 356/394, 356/239.8, 356/237.4, 356/237.1, 356/394.5, 356/237.2, 356/237.3, 356/237.5, 356/237.7, 356/238.8, 324/501, 324/537, 348/125, 348/126, 430/5, 430/22, 430/30
• US Class Current: 382/144
• CPC Class Codes: G03F 7/7065 Defects, e.g. optical inspection of patterned layer for defects
• CPC Class Codes: G06T 7/0004 Industrial image inspection

Abstract

A method for determining whether a defect that is detected by photomask inspection will adversely affect a semiconductor device, such as a wafer. The method has the ability of relating defect specifications directly to device performance and wafer yields, and assessing the impact of combining the defect with the critical dimension error using standard inspection tools. More specifically, the method includes the steps of: inspecting the photomask for defects; measuring the size and location of the defects relative to features on the photomask; classifying the defects by type of defect; assigning an equivalent mask critical dimension error (EME) value to each of the features based on size, location and type of defect; assigning a total mask error to each of the features by adding EME values to each defect impacting the features; and comparing the equivalent critical dimension error to a mask critical dimension error tolerance to determine whether the defects adversely affect the performance of the semiconductor device.

Claims

1. A method for determining if an undesirable feature on a photomask will adversely affect the performance of a semiconductor device On a wafer, said device being created by said photomask through a lithographic process, said method comprising the steps of:

(a) inspecting said photomask for defects;

(b) measuring the size and location of said defects relative to features on said photomask;

(c) classifying said defects by type of defect;

(d) assigning an equivalent mask critical dimension error (EME) value to each of said features based on said size, said location and said type of defect, wherein said EME value for each defect is assigned by the equation;

2. The method as recited in claim 1, wherein said EME value is based on a functional relationship for said defect type, said relationship consisting of measurable parameters selected from the group consisting of size, location relative to said features, phase and transmission.

3. The method as recited in claim 2, wherein said relationships are precomputed and are first generated for a test structure and then generalized to other ones of said features on said photomask.

4. The method as recited in claim 3, wherein said test structure is an isolated line.

5. The method as recited in claim 2, wherein said precomputed relationships are determined for a first photomask type and then used for a second one of said photomask type.

6. The method as recited in claim 1, wherein said photomask is a binary photomask made of substantially opaque regions on substantially clear regions.

7. The method as recited in claim 1, wherein said defect types comprise:

clear in opaque, opaque in clear, critical dimension error and registration error.

8. The method as recited in claim 1, wherein said photomask is an attenuated phase shifting photomask.

9. The method as recited in claim 8, wherein defects in said attenuated phase shifting photomask are classified by defect types, said defect types comprising:

clear in attenuator, attenuator in clear, opaque in attenuator, opaque in clear, phase error, transmission error, critical dimension error and registration error.

10. The method as recited in claim 1, wherein said photomask is an alternating phase shifting photomask.

11. The method as recited in claim 10, wherein defects in said alternating phase shifting photomask are classified by defect types, said defect types comprising:

clear in opaque, opaque in clear, phase bump, phase hole, phase error, transmission error, critical dimension error and registration error.

12. The method as recited in claim 1, wherein said defect types include:

clear in opaque, opaque in clear, clear in attenuator and attenuator in clear.

13. The method as recited in claim 1, wherein said B, C, and K are precomputed constants.

14. The method as recited in claim 13, wherein the photomask is a binary photomask and B=0.

15. The method as recited in claim 13, wherein the photomask is an attenuated phase shifting photomask and B is allowed to be different from zero.

16. The method as recited in claim 15, wherein B<

0.

17. The method as recited in claim 13, wherein said defect type is classified as opaque in clear, and 1.3 um⁻

1<

C<

1.7 um<sup>−

1</sup> and 2.0 um<sup>−

2</sup><

K<

6.0 um<sup>−

2</sup>.

18. The method as recited in claim 13, wherein said defect type is classified as clear in opaque, and −

1.3 um<sup>−

1</sup><

C<

−

1.7 um<sup>−

1</sup> and 2.0 um<sup>−

2</sup><

K<

to 6.0 um<sup>−

2</sup>.

19. The method as recited in claim 13, wherein said defect type is classified as clear in attenuator, and −

1.3 um<sup>−

1</sup><

C<

−

1.7 um<sup>−

1</sup> and 1.0 um<sup>−

2</sup><

K<

5.0 um<sup>−

2</sup>.

20. The method as recited in claim 13, wherein said defect type is classified as opaque in attenuator and C is in the range −

1.0 to 1.0 um<sup>−

1</sup> and K is in the range 20.0 to 40.0 um<sup>−

2</sup>.

21. The method as recited in claim 12, wherein the EME for each defect is assigned using the equation:

22. The method as recited in claim 1, wherein said TME value at location (x_o,y_o) is computed according to $\mathrm{TME}$

(xo,yo)=∑

EMEnN

(x-xo,y-yo)

wherein EME_n(x,y) is the EME of the n^th defect and N is the total number of defects that were found.

23. The method as recited in claim 22, wherein the summation is only carried out over a small subset of defects that are close to location (x_o,y_o).

24. The method as recited in claim 22, wherein the worst case location (x_o′

, y_o′

) is found by taking partial derivatives with respect to x_o and y_o and setting the result equal to 0 at x_o=x_o′ and

y_o=y_o′

. This system of equations then being solved for x_o′ and

y_o′

.

25. The method as recited in claim 1, wherein said photomask contains optical proximity correction or subresolution assist features.

26. The method as recited in claim 25, wherein said features consist solely of features that exist in the design of said photomask and that are supposed to be printed on said wafer.

27. The method as recited in claim 25, wherein said features consist of both features that are supposed to be printed on said wafer and features that exist in the design data for said photomask but are not supposed to be printed on said wafer.