Multi-OPC Method for Semiconductor Mask Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing optical proximity correction (OPC) methods require excessive runtime due to the high number of simulations needed for complex OPC models, leading to increased execution time and potential errors in pattern correction, especially as design complexities increase in semiconductor manufacturing.
Innovation Solution
A multi-OPC method that uses a combination of a simple and complex OPC model, where the simple model is calibrated to have an error tendency similar to the complex model, reducing the number of iterations and execution time by generating a re-target pattern and performing simulations with both models to achieve accurate OPC results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a complex OPC model is used to achieve accurate pattern correction, then manufacturing precision is improved, but execution time increases due to excessive runtime
Solution Approach 1:
The patent divides the OPC process into two distinct simulation phases: a first simulation using a simplified OPC model for initial corrections, and a second simulation using a complex OPC model for final precision adjustments. This segmentation allows the system to benefit from both speed (first simulation) and accuracy (second simulation) without paying the full computational cost of the complex model throughout the entire process.
Solution Approach 2:
The first simulation using the simplified OPC model performs preliminary corrections on the layout pattern before the second simulation. By pre-processing the pattern with the faster simplified model, the system reduces the workload and iteration requirements for the subsequent complex model simulation, thereby reducing total execution time while maintaining final accuracy.
2Manufacturing precision
If a complex OPC model is used to ensure accurate edge placement error correction, then manufacturing precision is improved, but the number of simulations increases leading to longer runtime
Solution Approach 1:
The patent segments the simulation workload by assigning different levels of model complexity to different simulation stages. The first simulation uses a simplified model with fewer calculations per iteration, while the second simulation uses the complex model only when needed for final precision. This segmentation improves overall productivity by reducing the total number of expensive complex simulations required.
Solution Approach 2:
The patent applies partial action by using the simplified OPC model for the majority of the correction process where high precision is not yet required, and only employs the complex model partially for final refinement. This avoids the excessive computational action of running the complex model throughout the entire correction process, thereby improving simulation efficiency.
3Adaptability or versatility
If design complexity increases in semiconductor manufacturing, then product functionality is improved, but the number of simulations required increases leading to excessive runtime
Solution Approach 1:
The patent implements segmentation by creating a two-tier simulation architecture where design complexity is handled differently at each tier. The simplified model in the first simulation handles basic geometric corrections efficiently, while the complex model in the second simulation handles sophisticated optical effects. This segmentation allows the system to adapt to complex designs without proportionally increasing runtime.
Solution Approach 2:
The first simulation performs preliminary processing that simplifies complex design patterns before they are passed to the second simulation. By pre-processing complex geometries with the faster simplified model, the system reduces the effective complexity that the second simulation must handle, thereby mitigating the runtime increase that would otherwise result from high design complexity.
Data Source
AI summary
The inventive concept provides an optical proximity correction (OPC) method using a multi-OPC model that can reduce a runtime of an entire OPC method by reducing an iteration number of simulations using a complex OPC model, and a method of manufacturing a mask by using the OPC method. The OPC method using the multi-OPC model can generate a re-target pattern to be applied to a simple OPC model, and can perform a simulation by using a complex OPC model on a target pattern after performing a simulation using the simple OPC model on the re-target pattern. Therefore, an iteration number of simulations using the complex OPC model can be reduced and, accordingly, an entire execution time of the OPC method can be reduced.