Multi-Stage Series Photoresist Characterization Network
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current photoresist models in IC manufacturing face challenges with high complexity, low computational efficiency, and large simulation errors, particularly in advanced nodes, due to their inability to accurately describe nonlinear physical and chemical processes, leading to errors in mask optimization and calibration processes.
Innovation Solution
A multi-stage series photoresist characterization network is developed using Wiener-Padé form sub-cascading modules, which construct a series model by convolving and combining Wiener nonlinear orders and kernel functions to accurately characterize nonlinear responses with reduced computational resources, and a calibration method based on constrained quadratic convex optimization to adjust parameters efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If strict theoretical methods are used to simulate photolysis exposure, reaction diffusion, and photopolymerization, then accuracy is improved, but computational efficiency deteriorates due to high complexity
Solution Approach 1:
The photoresist model is segmented into multiple stages: exposure stage (photolysis), post-exposure stage (reaction diffusion), and development stage (photopolymerization). Each stage is modeled separately using appropriate mathematical methods, allowing accurate simulation of complex processes while maintaining computational efficiency through modular structure
Solution Approach 2:
The model uses parameter changes to represent the progression of photoresist transformations through different stages. By changing key parameters (exposure dose, diffusion coefficients, polymerization rates) at each stage transition, the model accurately captures nonlinear physical and chemical effects without requiring exhaustive computational simulation of all molecular interactions
2Measurement precision
If deep learning neural networks are used to characterize photoresist internal reactions, then characterization accuracy is improved, but simulation progress deteriorates due to heavy dependence on training samples and large calculation errors
Solution Approach 1:
The patent introduces an intermediary calibration process that bridges theoretical models and experimental data. Instead of directly training deep learning networks on raw data, the model uses intermediate representations (exposure patterns, diffusion profiles, development curves) that are physically meaningful and require fewer training samples, reducing computational burden while maintaining accuracy
Solution Approach 2:
The model performs preliminary analytical calculations for exposure, diffusion, and polymerization stages before final pattern formation. This preliminary action provides initial conditions and constraints that guide subsequent simulations, reducing the computational search space and improving simulation speed while maintaining characterization accuracy
3Adaptability or versatility
If a universal photoresist model is developed through large number of samples in different scenarios, then model universality is improved, but calibration process complexity and time consumption increase
Solution Approach 1:
The patent develops a universal photoresist model framework that can handle multiple photoresist types and processing conditions through a single unified mathematical structure. The model uses universal parameters (exposure characteristics, diffusion coefficients, polymerization kinetics) that can be adjusted for different scenarios without requiring separate models, achieving versatility while reducing calibration time through parameter scaling rather than complete re-calibration
Data Source
AI summary
Disclosed in the invention are a method and system for modeling, calibration, and simulation of a multi-stage series photoresist characterization network, pertaining to the field of semiconductor lithography. The invention comprises: firstly dividing a photoresist reaction process into several key stages, using a new idea of modeling a multi-stage series system network, constructing multiple stages of series Wiener-Padé form sub-cascading modules according to characteristics of lithography processes, and utilizing a joint calibration strategy based on a constrained quadratic convex optimization algorithm to provide a simulation means based on library matching and low-order multivariate polynomial equivalence of model parameters. The invention emphasizes and leverages universal advantages of the Wiener-Padé system theory in the characterization of non-linear system response characteristics, thereby achieving accurate and efficient modeling and calibration of complex physical, optical, and chemical highly-nonlinear response characteristics of photoresists in different process flows, while avoiding over-fitting and reducing model complexity and redundancy.