System and method for model based multi-patterning optimization
Some embodiments provide a method for optimally decomposing patterns within particular spatial regions of interest on a particular layer of a design layout for a multi-exposure photolithographic process. Specifically, some embodiments model the spatial region using a mathematical equation in terms of two or more intensities. Some embodiments then optimize the model across a set of feasible intensities. The optimization yields a set of intensities such that the union of the patterns created/printed from each exposure intensity most closely approximates the patterns within the particular regions. Based on the set of intensities, some embodiments then determine a decomposition solution for the patterns that satisfies design constraints of a multi-exposure photolithographic printing process. In this manner, some embodiments achieve an optimal photolithographic printing of the particular regions of interest without performing geometric rule based decomposition.
1. A method for decomposing a particular pattern on a layer of a design layout into a set of images for generating multiple sub-patterns for multiple photolithographic operations, the method comprising:
a) by a computer, imposing a grid on the particular pattern to define a set of nodes, each node associated with a cost;
b) based on said cost, searching through a space of lithographically feasible images to identify a set of at least two images for decomposing the particular pattern;
c) determining whether a union of the identified set of images adequately approximates the particular pattern; and
d) repeating operations (b) and (c) until a particular set of images is identified that has a union that adequately approximates the particular pattern.
2. The method of claim 1 further comprising using the identified particular set of images to produce a set of photomasks to generate the multiple sub-patterns during the multiple photolithographic operations.
3. The method of claim 1 , wherein determining whether the union of the identified set of images adequately approximates the particular pattern comprises optimizing a function.
4. The method of claim 3 , wherein the function comprises values of image intensities at positions selected according to the particular pattern.
5. A method for decomposing a particular pattern on a layer of a design layout into a set of images for generating multiple sub-patterns for multiple photolithographic operations, the method comprising:
a) by a computer, defining a mathematical model to bound a search through a solution space to only lithographically feasible images;
b) searching through a space of lithographically feasible images to identify a set of at least two images for decomposing the particular pattern;
c) determining whether a union of the identified set of images adequately approximates the particular pattern; and
d) repeating operations (b) and (c) until a particular set of images is identified that has a union that adequately approximates the particular pattern.
6. The method of claim 5 , wherein the mathematical model is an image-intensity model that bounds the search to only lithographically feasible images.
7. The method of claim 6 , wherein the mathematical model includes at least two Fourier decomposition equations for two intensity images.
8. A method for decomposing a particular pattern on a layer of a design layout into multiple sub-patterns for multiple photolithographic exposure operations, the method comprising:
a) by a computer, defining an objective function in terms of at least two different image intensities, each image intensity comprising values corresponding to a different photolithographic exposure operation, the image intensity values located at different spatial locations along a region of interest in the layer that includes the particular pattern;
b) solving the objective function by identifying two sets values for the two image intensities, said two sets of values producing an optimal value for the objective function; and
c) using the identified sets of image intensity values to produce said sub-patterns.
9. The method of claim 8 , wherein defining the objective function comprises:
a) identifying a plurality of locations in the pattern; and
b) for each location, identifying a cost expression that is dependent on an image intensity value for at least one lithographic exposure operation.
10. The method of claim 8 , wherein the particular pattern is in a region of the design-layout layer, wherein defining the objective function further comprises:
a) identifying a plurality of locations that are in the region but outside of the pattern; and
b) for each location in the region but outside of the pattern, identifying a cost expression that is dependent on an image intensity value for at least one lithographic exposure operation.
11. The method of claim 10 , wherein identifying the plurality of locations comprises imposing a grid on the pattern, said grid having a plurality of nodes, said locations corresponding to the locations of the nodes imposed on the pattern.
12. The method of claim 8 further comprising generating a photomask for each exposure operation using said sub-patterns.
13. The method of claim 12 , wherein using the identified sets of image intensity values to produce said sub-patterns comprises converting the image intensity values to a set of binary values used to produce the sub-patterns for the photomasks.
14. The method of claim 8 , wherein an intensity value for the region that exceeds a specified threshold defines a spatial location to be printed during a particular exposure corresponding to the intensity value, and wherein an intensity value for the region that does not exceed a specified threshold defines a spatial location that is not to be printed during a particular exposure corresponding to the intensity value.
15. The method of claim 8 , wherein a set of image intensity values represents a grayscale model for results that are feasibly achievable in a particular exposure operation according to printability constraints of the photolithographic exposure operations.
16. The method of claim 8 , wherein using the identified sets of image intensity values to produce said sub-patterns comprises identifying sets of image intensity values for each exposure such that the union of the sets of image intensity values generates a pattern that is sufficient to approximately reproduce the particular pattern on the layout.
17. A system comprising:
a) a first module for modeling a region of a layer of a design layout comprising at least one particular pattern in terms of at least two different image intensities, each image intensity comprising values corresponding to a different photolithographic exposure operation;
b) a second module for identifying a plurality of locations on said particular layer, each location associated with a particular cost defined in terms of the at least two image intensity values; and
c) a third module for iteratively searching through a space of lithographically feasible images based on said particular cost to identify a set of at least two images for decomposing the particular pattern when a union of the identified images adequately approximates the particular pattern.
18. The system of claim 17 further comprising a fourth module for using the identified particular images to produce a set of photomasks to generate multiple sub-patterns of the decomposed particular pattern during multiple photolithographic operations.
19. The system of claim 17 , wherein a set of image intensity values represents a grayscale model for results that are feasibly achievable in a particular exposure operation according to printability constraints of the photolithographic exposure operation.
20. A non-transitory computer readable storage medium that stores a computer program for decomposing a particular pattern on a layer of a design layout into multiple sub-patterns for multiple photolithographic operations, the computer program executable by a processor, the computer program comprising sets of instructions for:
a) imposing a grid on the particular pattern to define a set of nodes, each node associated with a particular cost;
b) searching, based on said cost, through a space of lithographically feasible images to identify a set of at least two images for decomposing the particular pattern;
c) determining whether a union of the identified set of images adequately approximates the particular pattern; and
d) repeating the sets of instructions for operations (b) and (c) until a particular set of images is identified that has a union that adequately approximates the particular pattern.
21. The non-transitory computer readable storage medium of claim 20 , wherein the computer program further comprises a set of instructions for using the identified particular set of images to produce a set of photomasks to generate the multiple sub-patterns during the multiple photolithographic operations.
22. The non-transitory computer readable storage medium of claim 20 , wherein the set of instructions for determining whether the union of the identified set of images adequately approximates the particular pattern comprises a set of instructions for optimizing a function.
23. The non-transitory computer readable storage medium of claim 22 , wherein the function comprises values of image intensities at positions selected according to the particular pattern.
24. A non-transitory computer readable storage medium that stores a computer program for decomposing a particular pattern on a layer of a design layout into multiple sub-patterns for multiple photolithographic operations, the computer program executable by a processor, the computer program comprising a set of instructions for:
a) defining a mathematical model to bound a search through a solution space to only lithographically feasible images;
b) searching through the solution space of lithographically feasible images to identify a set of at least two images for decomposing the particular pattern;
c) determining whether a union of the identified set of images adequately approximates the particular pattern; and
d) repeating operations (b) and (c) until a particular set of images is identified that has a union that adequately approximates the particular pattern.
25. The non-transitory computer readable storage medium of claim 24 , wherein the particular mathematical model is an image-intensity model that bounds the search to only lithographically feasible images.