Browse Prior Art Database

Shifting pattern data to avoid mask blank defects

IP.com Disclosure Number: IPCOM000005967D
Original Publication Date: 2001-Nov-20
Included in the Prior Art Database: 2001-Nov-20
Document File: 5 page(s) / 74K

Publishing Venue

Motorola

Related People

Scott Hector: AUTHOR [+3]

Abstract

Obtaining defect free mask blanks will be a significant challenge as lithographic feature size continues to be reduced. To enable the use of mask blanks that are not defect free but have a reasonably small number of defects, the pattern on the mask can be moved so that the mask blank defects do not affect the printed pattern. With the known defect locations from a mask-blank defect inspection report, the pattern data can be positioned to make the defects non-printable or suitable for easy repair after mask patterning. The pattern data can be translated, rotated, or perhaps even magnified slightly to avoid known defects. Critical areas of the mask pattern can be identified to determine how to change the pattern to more efficiently avoid the effects of mask blank defects. With this approach, the effective yield of mask blank fabrication will be increased and the cost of mask production will be reduced.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 37% of the total text.

Shifting pattern data to avoid mask blank defects

Scott Hector and Lloyd C. Litt

John Maltabes

Abstract

                    Obtaining defect free mask blanks will be a significant challenge as lithographic feature size continues to be reduced.  To enable the use of mask blanks that are not defect free but have a reasonably small number of defects, the pattern on the mask can be moved so that the mask blank defects do not affect the printed pattern.  With the known defect locations from a mask-blank defect inspection report, the pattern data can be positioned to make the defects non-printable or suitable for easy repair after mask patterning.   The pattern data can be translated, rotated, or perhaps even magnified slightly to avoid known defects. Critical areas of the mask pattern can be identified to determine how to change the pattern to more efficiently avoid the effects of mask blank defects.  With this approach, the effective yield of mask blank fabrication will be increased and the cost of mask production will be reduced.

I.                   Present mask fabrication method

                    Presently, mask blanks are not routinely inspected upon arrival at the mask patterning facility.  The relatively small number of defects on the mask blank might manifest themselves as defects during patterning. Some of these printing defects are located in regions that are not critical.  If these blank defects become defects in the pattern, the pattern defects are usually repairable.  Figure 1 shows a flow chart of the present mask fabrication process, emphasizing decisions based on defect inspection results.  Some mask blank defects result in mask patterning yield loss, 1-Yp, or final mask yield loss, 1-YF, where YP is the yield after pattern defect inspection and YF is the final mask yield. 

                    Mask blank defects in the new materials needed for the AEPSM (Attenuated Embedded Phase Shift Mask)  have increased.  This increase in blank defects has in turn reduced YP and YF. Next generation lithography techniques, such as electron projection lithography (EPL) and extreme ultraviolet lithography (EUVL),  require mask blanks made from different materials than binary optical masks or AEPSM.  These new mask blanks will initially have more defects than present mask blanks for optical projection lithography.  Hector et al. have shown that printable defect density on EPL1 and EUVL2 mask blanks has a significant impact on mask price.  The preferred solution to yield loss from blank defects is to lower printable defect count on the blank with cleaner processing and/or repair.  However, lowering blank defect count will be difficult, especially as the minimum printable defect size decreases.  When EPL or EUVL are introduced the initial mask blank defect count might be high. Therefore, a method to pattern masks with high yield on mask blanks that have defects is desirable. 

Figure 1.  Flow chart of the present method of fabricating lithographic masks.

II.                 Modified mask fabrication method to enable using mask blanks with known defects

If mask...