System and methods for OPC model accuracy management and disposition
View Patent ↗System and methods for OPC model accuracy and disposition using quad matrix are presented. A method includes obtaining wafer data from a calibration test pattern. The method also classifies the wafer data into four quadrants of a quad matrix. The method further utilizes at least one of the four quadrants to quantify OPC model accuracy.
1. A method for generating a mask used in lithographic processes to form a semiconductor device on a wafer, the method comprising:
employing a computer for performing quad matrix management to quantify optical proximity correction (OPC) model accuracy comprising
obtaining wafer data from a calibration test pattern,
classifying the same wafer data obtained from the calibration test pattern into four quadrants of a quad matrix, and
utilizing one of the four quadrants to quantify OPC model accuracy; and
creating a photomask after performing the quad matrix management.
2. The method of claim 1 , wherein classifying the wafer data into four quadrants comprises:
classifying the wafer data into two sets of data based at least on data integrity and confidence level.
3. The method of claim 2 , wherein the two sets of data comprise one set of data with a confidence level less than a user-defined confidence level and another set of data with a confidence level within the user-defined confidence level.
4. The method of claim 2 , wherein classifying the wafer data into four quadrants further comprises:
classifying the wafer data into two sets of data based at least on a user-defined model specification.
5. The method of claim 4 , wherein the two sets of data comprise one set of data meeting the user-defined model specification and another set of data not meeting the user-defined model specification.
6. The method of claim 1 , wherein the four quadrants comprise:
a first quadrant for a first portion of the wafer data that is within a user-defined confidence level and meets a user-defined model specification;
a second quadrant for a second portion of the wafer data that is less than the user-defined confidence level and meets the user-defined model specification;
a third quadrant for a third portion of the wafer data that is less than the user-defined confidence level and does not meet the user-defined model specification; and
a fourth quadrant for a fourth portion of the wafer data that is within the user-defined confidence level and does not meet the user-defined model specification.
7. The method of claim 1 , wherein utilizing one of the four quadrants to quantify OPC model accuracy comprises at least one of the following:
identifying a best model;
identifying one or more weak points in the wafer data; or
identifying a best number of measurement.
8. The method of claim 7 , wherein identifying a best model comprises:
comparing at least two models; and
selecting one of the at least two models that has fewer data points in the fourth quadrant than other models of the at least two models as the best model.
9. The method of claim 7 , wherein identifying one or more weak points in the wafer data comprises:
identifying one or more data points that fall in the fourth quadrant as the one or more weak points.
10. The method of claim 7 , wherein identifying a best number of measurement comprises:
determining a number of measurement that results in a level of uncertainty below a user-defined level of uncertainty.
11. A method comprising:
quantifying optical proximity correction (OPC) model accuracy by performing operations comprising:
selecting an OPC model at least by performing quad matrix management which comprises
obtaining wafer data from a calibration test pattern,
classifying the same wafer data obtained from the calibration test pattern into four quadrants of a quad matrix, and
utilizing one of the four quadrants to quantify OPC model accuracy;
creating a photomask based at least in part on the selected OPC model after performing the quad matrix management;
performing wafer verification; and
verifying lithography performance.
12. The method of claim 11 , wherein utilizing one of the four quadrants to quantify OPC model accuracy comprises identifying a best model.
13. The method of claim 11 , wherein classifying the wafer data into four quadrants comprises:
classifying the wafer data into first two sets of data based at least on data integrity and confidence level; and
classifying the wafer data into second two sets of data based at least on a user-defined model specification.
14. The method of claim 13 , wherein the first two sets of data comprise one set of data meeting the user-defined model specification and another set of data not meeting the user-defined model specification, and wherein the second two sets of data comprise one set of data meeting the user-defined model specification and another set of data not meeting the user-defined model specification.
15. The method of claim 11 , wherein the four quadrants comprise:
a first quadrant for a first portion of the wafer data that is within a user-defined confidence level and meets a user-defined model specification;
a second quadrant for a second portion of the wafer data that is less than the user-defined confidence level and meets the user-defined model specification;
a third quadrant for a third portion of the wafer data that is less than the user-defined confidence level and does not meet the user-defined model specification; and
a fourth quadrant for a fourth portion of the wafer data that is within the user-defined confidence level and does not meet the user-defined model specification.
16. The method of claim 11 , wherein utilizing one of the four quadrants to quantify OPC model accuracy comprises at least one of the following:
identifying a best model;
identifying one or more weak points in the wafer data; or
identifying a best number of measurement.
17. The method of claim 16 , wherein identifying one or more weak points in the wafer data comprises:
identifying one or more data points that fall in the fourth quadrant as the one or more weak points.
18. The method of claim 16 , wherein identifying a best number of measurement comprises:
determining a number of measurement that results in a level of uncertainty below a user-defined level of uncertainty.
19. A method comprising:
quantifying optical proximity correction (OPC) model accuracy by performing operations comprising:
selecting an OPC model at least by simulation with quad matrix management which comprises
obtaining wafer data from a calibration test pattern,
classifying the wafer data into four quadrants of a quad matrix
utilizing at least one of the four quadrants to quantify the OPC model accuracy which comprises identifying a best model, wherein identifying the best model comprises
comparing at least two models, and
selecting one of the at least two models that has less data points in the fourth quadrant than other models of the at least two models as the best model;
creating a photomask based at least in part on the selected OPC model;
performing wafer verification; and
verifying lithography performance.