Image analysis – Applications – Manufacturing or product inspection
Reexamination Certificate
1999-12-23
2004-08-17
Bali, Vikkram (Department: 2623)
Image analysis
Applications
Manufacturing or product inspection
C382S145000
Reexamination Certificate
active
06778695
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to inspection of reticles for use in manufacturing integrated circuits. More particularly, the invention relates to use of circuit designs (e.g., netlists, layouts) in inspecting reticles for use in manufacturing integrated circuits.
BACKGROUND OF THE INVENTION
FIG. 1
 is one embodiment of an integrated circuit processing arrangement. The general uses of the components of 
FIG. 1
 are known in the art. Light source 
100
 provides light towards wafer 
130
. Mask/reticle 
110
 blocks light for certain predetermined portions of wafer 
130
. Stepper/scanner 
120
 directs the patterns of mask/reticle 
110
 to one of multiple integrated circuits being developed on wafer 
130
.
High-level integrated circuit (IC) description languages such as VHDL and Verilog® are commonly used to design circuits. One embodiment of VHDL is described in greater detail in “IEEE Standard VHDL Language Reference Manual,” ANSI Std. 1076-1993, Published Jun. 6, 1994. One embodiment of Verilog® is described in greater detail in IEEE Standard 1364-1995. These high-level IC description languages allow a circuit designer to design and simulate circuits by using high-level code to describe the structure and/or behavior of the circuit being designed.
The integrated circuit design is used to generate an IC layout. Verification and other procedures can also be performed on the IC design or the IC layout. A set of reticles are created based on the IC layout and the reticles are used to manufacture the IC. Prior to mass production of the IC, the reticles are inspected to determine whether the IC manufactured from the reticles will perform the desired function. At least one reticle is generated for each layer of the IC to be manufactured.
FIG. 2
 illustrates a prior art approach to reticle inspection. The high-level IC description language code is used to produce netlist 
200
 that describes an interconnection of circuit components that provide the desired functionality. Netlist 
200
 can include coordinate information to describe the position of various circuit components within the overall circuit. Netlist 
200
 can then be used to develop the layout and ultimately fabricate an integrated circuit IC having the functionality of the designed circuit. Netlist 
200
 can also be used for emulation and/or verification purposes.
Netlist 
200
 is used to generate layout 
210
, which describes the physical layout of an IC. Verification tool 
220
 can compare layout 
210
 to netlist 
200
 to verify that layout 
210
 is an accurate representation of netlist 
200
. Layout 
210
 can be modified as necessary based on results generated by verification tool 
220
.
Layout 
210
 is used to generate reticle database 
230
, which is used to generate reticle 
240
. Reticle production equipment (not shown in 
FIG. 2
) generates reticle 
240
 from reticle database 
230
 in any manner known in the art. This can include exposure to electron beams, laser beams, or ion beams using mask writing tools such as, for example, the MEBES 4500 or ALTA 3500 available from ETEC Corporation of Hayward, Calif.
Once fabricated, reticle 
240
 is then inspected for defects using inspection device 
250
. Inspection device 
250
 can be, for example, Model 351 Inspection System, manufactured by KLA-Tencor of San Jose, Calif. Several techniques for identifying defects are known in the art that typically compare the image of a reticle under inspection with a reference image. The reference image can be, for example, an adjacent identical layout (die:die, or die-to-die inspection), or a reference image derived from the layout (die:database, or die-to-database inspection). In either case, mismatched geometries are flagged as a defect. In either case, defects identified by inspection device 
250
 are output as defect list 
260
. Defect list 
260
 can also be referred to as an inspection report.
In a typical prior art reticle inspection system, inspection device 
250
 flags defects based solely on a mismatch of feature geometries between the image and the database pattern. Defect list 
260
 is then reviewed by a human operator, who examines the image of each defective location on the reticle and classifies the defect. From the classification, the operator determines which defects are insignificant and which require repair. Defect list 
260
 may be prioritized by the human operator, for example, by size of defect. Defects are repaired at 
280
 in any manner known in the art.
The reticle inspection and repair technique of 
FIG. 2
 is also generally human labor intensive and therefore expensive. As integrated circuits become smaller and smaller, the potential for defects increases and the percentage of defects that must be repaired increases, which results in more work for the human operator and a greater expense to manufacture an operable integrated circuit. In other words as the size of an IC decreases and/or the number of layers of the IC increases, the cost of reticle inspection increases.
A prior art solution to this is to prioritize the defects and eliminate the insignificant defects from the defect list using simulation. In this case, the effect the defect will have on the IC is simulated using conventional simulation techniques, and the defects are prioritized by the effective size on the wafer. However, this solution is not ideal. The amount of time for an accurate simulation can be larger than for human review, and all defects must be simulated to allow the prioritization to occur. As the number of defects increases with smaller feature size, this can potentially increase the total time required for defect evaluation.
What is needed is an improved reticle inspection and repair technique that results in decreased simulation requirements and/or faster defect evaluations, while still accurately identifying all defects that will impact the device being fabricated.
SUMMARY OF THE INVENTION
Methods and apparatuses for netlist based reticle defect prioritization are described. A defect list is generated in response to an inspection of a reticle for use to manufacture an integrated circuit. The defect list is compared to a processed netlist corresponding to the integrated circuit and/or a processed layout corresponding to the integrated circuit. The defect is prioritized based, at least in part, on the comparison.
In one embodiment, the processed netlist provides a device location that can be used to identify regions of interest to be used for prioritization of the defect list. For example, a predetermined region around a transistor gate can be defined as a region of interest. In one embodiment, the processed layout provides device geometry and/or critical dimensions that can be used to identify regions of interest to be used for prioritization of the defect list.
REFERENCES:
patent: 4717644 (1988-01-01), Jones et al.
patent: 4758094 (1988-07-01), Wihl et al.
patent: 5598341 (1997-01-01), Ling et al.
patent: 5686206 (1997-11-01), Baum et al.
patent: 5994030 (1999-11-01), Sugihara et al.
patent: 6483937 (2002-11-01), Samuels
patent: 04 340740 (1992-11-01), None
patent: 07 239306 (1995-09-01), None
patent: WO00/36525 (2000-06-01), None
International Seach Report, European Patent Office, Date of Mailing: May 7, 2002, International application No.: PCT/US 00/42830, International filing date: Dec. 22, 2000, pp. 1-4.
Moore Andrew J.
Schellenberg Franklin M.
Bali Vikkram
Christensen O'Connor Johnson & Kindness PLLC
LandOfFree
Design-based reticle defect prioritization does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Design-based reticle defect prioritization, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Design-based reticle defect prioritization will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3341659