Automated Risk Inventory and Common Process Window for IC Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for manufacturing integrated circuits (ICs) face challenges in identifying process-sensitive geometries and defects, leading to increased uncertainty and risk of defects or impaired operability, particularly as IC components shrink below 22 nanometers, due to the reliance on manual or limited automated techniques for visual inspection and correction.
Innovation Solution
A method that converts images of manufactured circuits into representative contours, generates a risk inventory to identify process-sensitive geometries, and creates a common process window to indicate manufacturing reliability, allowing for automated adjustment of manufacturing tools to improve reliability and reduce defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual or limited automated visual inspection techniques are used to identify process-sensitive geometries and defects, then the complexity of the inspection system is reduced, but the measurement precision and reliability of defect identification deteriorate
Solution Approach 1:
The patent replaces manual visual inspection with an automated computational system that converts manufactured circuit images into representative contours and applies rule-based analysis. This substitution of mechanical/manual inspection with automated image processing and computational analysis directly improves measurement precision while managing system complexity through algorithmic approaches.
Solution Approach 2:
The patent creates representative contours as simplified copies or abstractions of the actual manufactured circuit features. By working with these representative contour models rather than raw images, the system achieves precise defect identification while reducing the computational complexity of analyzing full-resolution circuit imagery.
2Productivity
If automated techniques are implemented to model and identify defects, then the productivity and manufacturing efficiency improve, but the device complexity increases due to the need for automated analysis systems
Solution Approach 1:
The patent implements a self-service automated system that independently performs defect identification and process-sensitive geometry detection without requiring manual intervention. The system automatically converts images to contours, applies optical rule checks, and generates risk inventories, thereby improving manufacturing productivity while containing complexity through autonomous operation.
Solution Approach 2:
The patent transforms the inspection problem by changing parameters from analyzing raw pixel data to working with extracted contour features and geometric representations. This parameter transformation enables automated analysis to achieve high productivity while managing complexity by focusing computational resources on critical geometric features rather than entire images.
3Reliability
If visual inspection is used to identify problem areas, then the manufacturing tool adjustment can be made, but the loss of time and increased cost during testing occur
Solution Approach 1:
The patent performs preliminary automated analysis by converting images to representative contours and identifying process-sensitive geometries before final manufacturing decisions are made. This preliminary automated screening reduces the time required for manual inspection while maintaining or improving reliability through consistent, objective defect identification.
Solution Approach 2:
The patent implements a feedback mechanism where the automated analysis results (risk inventory, common process window) directly inform manufacturing tool adjustments. This closed-loop feedback system improves manufacturing reliability by ensuring that identified defects and sensitive geometries are addressed through systematic tool parameter modifications, reducing the need for repeated manual inspection cycles.
Data Source
AI summary
A method disclosed herein includes: converting an image of a manufactured circuit to a plurality of representative contours, the plurality of representative contours corresponding to printed features in the manufactured circuit; generating a risk inventory for the manufactured circuit based on the plurality of representative contours, the risk inventory being configured to identify at least one process sensitive geometry (PSG) in the manufactured circuit; generating a common process window (CPW) for the manufactured circuit based on the plurality of representative contours and the risk inventory, the CPW being indicative of manufacturing reliability of each feature in the manufactured circuit; and generating instructions to adjust a manufacturing tool for creating the manufactured circuit, based on the generated CPW.