Method and system for obtaining bounds on process parameters for OPC-verification

1. A method for performing printability verification of a mask layout, the method comprising:

creating one or more tight clusters;

computing a set of process parameters associated with a point on said mask;

comparing said set of process parameters to said one or more tight clusters; and

reporting an error when at least one of said process parameters is away from said one or more tight clusters,

wherein said one or more tight clusters is a cluster selected from a group consisting of convex hull and ellipsoid.

2. The method of claim 1 , wherein said creating one or more tight clusters comprises:

measuring a group of calibration data from a plurality of test structures;

creating a process model in a space having one or more dimensions;

applying said process model to said group of calibration data thus obtaining a group of process parameters; and

dividing said group of process parameters into said one or more tight clusters in said space of one or more dimensions.

3. The method of claim 2 , wherein said dividing said group of process parameters comprises:

applying a grid based unsupervised clustering approach thus approximating a number of clusters from said group of process parameters; and

applying a supervised clustering approach thus smoothing said number of clusters.

4. The method of claim 3 , wherein smoothing said number of clusters comprises applying convex hulls or ellipsoids to corresponding process parameters in said number of clusters.

5. The method of claim 2 , wherein said computing a set of process parameters comprises:

identifying said point on said mask;

applying said process model to said point thus obtaining an output point corresponding to said point on said mask; and

deriving said set of process parameters from said output point.

6. The method of claim 1 , wherein said comparing said set of process parameters comprises

providing one or more pre-determined values relating to a level of predictability

establishing one or more inequalities for said one or more tight clusters using said one or more pre-determined values;

applying said one or more inequalities in testing said set of process parameters; and

determining whether one or more of said process parameters are away from said one or more tight clusters from said testing.

7. The method of claim 2 , wherein data in said group of calibration data are about 7 to about 20 times more than the dimensions of said space.

8. The method of claim 1 , wherein said reporting an error comprises reporting said error when said at least one of said process parameters is away from said one or more tight clusters by a pre-determined value.

9. A system for performing printability verification of a mask layout, the system comprising:

at least one storage device to store machine-readable codes;

a central processing unit (CPU) operationally connected to the storage device; and

a controller controlling the storage device and the CPU,

wherein the CPU is adapted to execute the machine-readable codes to:

creating one or more tight clusters;

computing a set of process parameters associated with a point on said mask;

comparing said set of process parameters to said one or more tight clusters; and

reporting an error when at least one of said process parameters is away from said one or more tight clusters,

wherein said one or more tight clusters is a cluster selected from a group consisting of convex hull and ellipsoid.

10. The system of claim 9 , further comprising at least one input/output device, said input/output device is adapted to accepting instructions and input from an operator and for outputting results from CPU executing the machine-readable codes during simulation.

11. The system of claim 9 , wherein to create said one or more tight clusters, the CPU is adapted to:

measuring a group of calibration data from a plurality of test structures;

creating a process model in a space having one or more dimensions;

applying said process model to said group of calibration data thus obtaining a group of process parameters; and

dividing said group of process parameters into said one or more tight clusters in said space of one or more dimensions.

12. The system of claim 11 , wherein to divide said group of process parameters, the CPU is adapted to:

applying a grid based unsupervised clustering approach thus approximating a number of clusters from said group of process parameters; and

applying a supervised clustering approach thus smoothing said number of clusters.

13. The system of claim 12 , wherein to smooth the number of clusters, the CPU is adapted to applying convex hulls or ellipsoids to corresponding process parameters in the number of clusters.

14. A machine-readable storage medium having stored thereupon a set of instruction that, when executed by a machine, result in:

creating one or more tight clusters;

computing a set of process parameters associated with a point on said mask;

comparing said set of process parameters to said one or more tight clusters; and

reporting an error when at least one of said process parameters is away from said one or more tight clusters,

wherein said one or more tight clusters is a cluster selected from a group consisting of convex hull and ellipsoid.

15. The machine-readable storage medium of claim 14 , wherein said creating one or more tight clusters comprises:

measuring a group of calibration data from a plurality of test structures;

creating a process model in a space having one or more dimensions;

applying said process model to said group of calibration data thus obtaining a group of process parameters; and

dividing said group of process parameters into said one or more tight clusters in said space of one or more dimensions.

16. The machine-readable storage medium of claim 15 , wherein said dividing said group of process parameters comprises:

applying a grid based unsupervised clustering approach thus approximating a number of clusters from said group of process parameters; and

applying a supervised clustering approach thus smoothing said number of clusters.

17. The machine-readable storage medium of claim 16 , wherein smoothing said number of clusters comprises applying convex hulls or ellipsoids to corresponding process parameters in said number of clusters.

18. The machine-readable storage medium of claim 15 , wherein said computing a set of process parameters comprises:

identifying said point on said mask;

applying said process model to said point thus obtaining an output point corresponding to said point on said mask; and

deriving said set of process parameters from said output point.

19. The machine-readable storage medium of claim 14 , wherein said comparing said set of process parameters comprises

providing one or more pre-determined values relating to a level of predictability

establishing one or more inequalities for said one or more tight clusters using said one or more pre-determined values;

applying said one or more inequalities in testing said set of process parameters; and

determining whether one or more of said process parameters are away from said one or more tight clusters from said testing.

20. The machine-readable storage medium of claim 15 , wherein data in said group of calibration data are about 7 to about 20 times more than the dimensions of said space.

21. The machine-readable storage medium of claim 14 , wherein said reporting an error comprises reporting said error when said at least one of said process parameters is away from said one or more tight clusters by a pre-determined value.