Semiconductor Wafer Defect Root Cause Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for identifying the cause of manufacturing defects in self-aligned double patterning (SADP) techniques, such as Cross SADP, cannot distinguish the root cause of unexpected shifts in hole patterns formed during the semiconductor manufacturing process.
Innovation Solution
A system and method that captures images of a semiconductor wafer, calculates geometric centers and average geometric centers, and performs coordinate transformations to determine the precise shift amount and identify the manufacturing process responsible for the shift, using a processing unit and image capture unit to analyze geometric features and their serial numbers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional optical inspection equipment is used to check hole patterns, then the inspection can be performed, but the root cause of unexpected shifts cannot be distinguished
Solution Approach 1:
The inspection system segments the wafer surface into multiple regions and captures images from different locations. Each image is independently processed to calculate geometric centers, allowing the system to track shifts across different areas and identify whether shifts are localized or widespread, thereby providing more precise shift measurement while preserving information about shift patterns and potential causes.
Solution Approach 2:
The system adds the dimension of spatial distribution by capturing multiple images from different locations on the wafer. Instead of a single global inspection, the method creates a multi-point measurement system that tracks geometric centers across different spatial positions, enabling differentiation between random variations and systematic shifts caused by specific manufacturing process issues.
2Loss of information
If multiple images are captured and processed to identify shift causes, then root cause analysis is enabled, but the inspection complexity increases
Solution Approach 1:
The inspection system performs multiple functions using the same hardware: capturing images, calculating geometric centers, determining shifts, and identifying root causes. By making the system multi-functional, the patent avoids adding separate dedicated equipment for each function, thereby enabling comprehensive root cause analysis while minimizing the increase in overall system complexity.
Solution Approach 2:
The system creates digital copies of the wafer surface through multiple images and processes these copies computationally to identify patterns and causes of shifts. This approach replaces complex physical inspection mechanisms with computational analysis of image data, reducing hardware complexity while enhancing diagnostic capability.
Data Source
AI summary
A method for identifying a cause of manufacturing defects is provided. The method includes capturing, by an image capture unit, a number N of images from a semiconductor wafer, wherein each of the s umber N of images comprises a number M of geometric features, calculating, by a processing unit, a geometric center for each of the geometric features of the number N of images, calculating, based on the number N of images, a number M of average geometric centers associated with the number M of geometric features, and calculating a shift amount for each geometric feature of the number N of images.