Layer Class Relative Density for IC Lithography Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current photolithography modeling techniques for multi-patterning in semiconductor manufacturing are inaccurate in representing capacitive or resistive effects, particularly in modeling the thickness variations of metal lines within a layer, due to the lack of consideration for layer classes, leading to inefficiencies in IC chip production.
Innovation Solution
The method introduces a relative layer-class density parameter to model thickness variations by calculating the weighted average of values based on various window sizes, allowing direct representation of thickness tables and avoiding numerical errors associated with fitting polynomial expressions to data dependent on three variables, thereby accounting for layer-class effects in multi-patterning processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If multi-patterning is used to resolve smaller features, then feature resolution capability is improved, but modeling accuracy of capacitive or resistive effects deteriorates due to lack of layer class consideration
Solution Approach 1:
The patent segments the layer into multiple layer classes based on the number of patterning exposures required. By dividing the layer into distinct classes (e.g., first layer class for features formed by first exposure, second layer class for features formed by second exposure), the model can accurately capture the different capacitive and resistive effects of each class while maintaining overall feature resolution capability.
Solution Approach 2:
The patent applies local quality by assigning different modeling parameters and characteristics to different layer classes within the same layer. Each layer class has its own specific capacitive and resistive effect models that reflect the local patterning history and process conditions, thereby improving modeling accuracy without compromising the overall feature resolution.
2Adaptability or versatility
If polynomial expressions are fitted to model thickness variations, then modeling flexibility is improved, but numerical errors increase due to data dependency on three variables
Solution Approach 1:
The patent introduces a new dimension by incorporating layer class as a discrete categorical variable in addition to the continuous spatial coordinates. Instead of fitting polynomial expressions to three-variable data, the model uses layer class-specific parameters and lookup tables that directly represent thickness variations for each layer class, thereby eliminating numerical fitting errors while maintaining modeling flexibility through the additional layer class dimension.
Solution Approach 2:
The patent creates separate thickness models for each layer class by copying and adapting the base thickness model to account for layer-specific characteristics. This approach avoids the need to fit a single complex polynomial to all data points, instead using multiple simpler, layer-class-specific models that are more numerically stable and accurate.
3Adaptability or versatility
If separate tools are used to model capacitive and resistive effects, then functional specialization is improved, but system complexity increases
Solution Approach 1:
The patent merges the capacitive and resistive effect modeling into a unified photolithography process model that handles both effects simultaneously. The unified model uses a common layer class framework and integrated parameter sets, allowing both capacitive and resistive effects to be calculated and applied together in a single modeling pass, thereby reducing system complexity while maintaining the functional specialization needed for accurate effect prediction.
Data Source
AI summary
A method and apparatus of a novel modeling scheme for performing optical lithography simulation for a multi-color layer fabrication process is described. The method interpolates for simulation use between test or experimental data or descriptions to more accurately apply color differentiated parameters to the model creation and lithography simulation.