Etch-Aware OPC Model Calibration via Etch Bias Filter
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing OPC models fail to accurately compute combined photolithography-etch effects, leading to inaccuracies due to non-uniform and non-linear etch bias, which degrades the overall model quality, especially at advanced technology nodes.
Innovation Solution
A system that calibrates the photolithography process model by constructing an etch bias model and fitting it based on measured CD data, incorporating etch effects by computing residues and using inverse etch bias models to normalize and adjust the photolithography process model, thereby generating an accurate OPC model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing OPC models are used to model photolithography-etch effects, then the modeling process is simple, but the accuracy of computing combined photolithography-etch effects is poor
Solution Approach 1:
The patent segments the combined photolithography-etch modeling into two separate models: a photolithography process model and an etch bias model. Each model is calibrated independently using measured CD data from respective process stages, allowing accurate representation of each process while maintaining manageable model complexity. The segmented models are then combined to achieve accurate prediction of final etch contours.
Solution Approach 2:
The patent introduces an intermediary etch bias model that bridges the photolithography process model and the final etch contour prediction. This intermediary model captures the non-uniform and non-linear etch effects by modeling the difference between photolithography CD and etch CD as a function of local pattern geometry, enabling accurate combined effect prediction without requiring a single complex monolithic model.
2Reliability
If etch effects are not incorporated into the photolithography process model, then the model is easier to calibrate, but the overall OPC model quality degrades
Solution Approach 1:
The patent performs preliminary calibration of the photolithography process model using measured CD data from the photolithography stage alone, before incorporating etch effects. This preliminary action establishes an accurate baseline model of the photolithography process, which is then combined with the etch bias model. This sequential approach ensures that each model is properly calibrated before integration, improving overall OPC model quality while managing calibration complexity.
Solution Approach 2:
The patent merges the calibrated photolithography process model with the etch bias model to create an integrated OPC model that accounts for both photolithography and etch effects. The merging combines the strengths of both models: the photolithography model accurately predicts resist contours, while the etch bias model accurately predicts etch-induced CD variations. This combination achieves high overall model quality by capturing the complete photolithography-etch process chain.
3Manufacturing precision
If non-uniform and non-linear etch bias is not modeled, then the modeling process is simpler, but the accuracy at advanced technology nodes deteriorates
Solution Approach 1:
The patent models non-uniform and non-linear etch bias by making the etch bias parameters dependent on local pattern geometry parameters such as line width, line spacing, and pattern density. This parameter change allows the etch bias model to adapt to different local conditions in the layout, capturing the non-uniform and non-linear nature of etch effects. The model uses polynomial functions of pattern geometry parameters to represent etch bias, achieving high manufacturing precision at advanced technology nodes while maintaining reasonable model complexity.
Data Source
AI summary
One embodiment of the present invention relates to a system that constructs and calibrates an etch-aware photolithography model. During operation, the system constructs an etch bias model which models a critical dimension (CD) difference between a measured CD value of a feature after the photolithography process and a measured CD value of the feature after the etch process. The system then fits the photolithography process model based at least on the post-lithography measured CD data and the etch bias model, thereby causing the photolithography process model to be aware of etch effects. The present techniques facilitate bridging the gap between the photolithography and the etch process in the OPC modeling flow. In particular, these techniques can be used to modify conventional staged OPC model or to construct a model based rule table for correcting a retarget model.