Lithography Pattern Shape Optimization for Critical Dimension Variation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current lithography methods face challenges in accurately reproducing small critical dimensions on substrates due to manufacturing variations and computational inefficiencies, particularly in optical lithography and inverse lithography technology, leading to poor yield and increased costs.
Innovation Solution
A method involving neural networks is employed to calculate and optimize patterns on substrates by generating multiple possible neighborhoods and adjusting parameters to reduce manufacturing variation, using charged particle beam systems with individually controllable beamlets to enhance precision and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If optical lithography or inverse lithography technology is used to fabricate small critical dimensions, then manufacturing capability is achieved, but manufacturing precision deteriorates due to variations leading to poor yield
Solution Approach 1:
The patent applies preliminary action by pre-calculating multiple possible neighborhoods and their corresponding patterns before actual manufacturing. The system generates a plurality of possible patterns for each physical design element, evaluates their manufacturing variations in advance, and selects the optimal pattern that minimizes variation. This proactive approach allows the system to anticipate and compensate for manufacturing variations before they occur, thereby improving both precision and yield.
Solution Approach 2:
The patent employs parameter changes by adjusting the pattern design parameters based on calculated manufacturing variations. The system modifies pattern dimensions, shapes, and configurations to optimize manufacturing outcomes. By changing pattern parameters proactively based on pre-calculated variations, the system compensates for expected manufacturing deviations, thereby maintaining high precision and yield.
2Productivity
If conventional lithography methods are used, then manufacturing process is simple, but computational time increases and efficiency decreases
Solution Approach 1:
The patent applies segmentation by dividing the computational task into processing individual neighborhoods independently. The system divides the reticle design into multiple neighborhoods, calculates possible patterns for each neighborhood separately, and then combines results. This segmentation enables parallel processing and reduces overall computational time while maintaining comprehensive evaluation of manufacturing variations.
3Manufacturing precision
If pattern optimization is performed to reduce manufacturing variation, then manufacturing accuracy improves, but device complexity increases
Solution Approach 1:
The patent applies self-service by implementing an automated system that independently calculates manufacturing variations, evaluates multiple possible patterns, and selects optimal designs without requiring manual intervention. The system automatically performs all necessary computations and selections, reducing the need for complex manual processes and expert involvement, thereby managing complexity while improving precision.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach improves manufacturing accuracy and reduces computational time, enabling more precise pattern formation with reduced variation, thereby enhancing the yield and reducing costs in semiconductor and display fabrication.
Implementation Method 1
In all types of charged particle beam lithography, charged particle beams shoot energy to a resist-coated surface to expose the resist
Data Source
AI summary
Methods for calculating a pattern to be manufactured on a substrate include inputting a physical design pattern, determining a plurality of possible neighborhoods for the physical design pattern, generating a plurality of possible mask designs for the physical design pattern, calculating a plurality of possible patterns on the substrate, calculating a variation band from the plurality of possible patterns, and modifying the physical design pattern to reduce the variation band. Embodiments also include inputting a set of parameters for a neural network to calculate a pattern to be manufactured on a substrate, calculating a plurality of patterns to be manufactured on the substrate for the physical design in each possible neighborhood of the plurality of possible neighborhoods, training the neural network with the calculated plurality of patterns, and adjusting the set of parameters to reduce the manufacturing variation for the calculated plurality of patterns to be manufactured on a substrate.