Predictive SEM Imaging for Wafer Inspection Without Photoresist Damage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wafer inspection methods using scanning electron microscopes (SEM) cause damage to the photoresist on the wafer, altering the pattern and resulting in inaccurate key performance indicators, and existing physics-based models fail to generate accurate predictive images of the wafer after etching due to unknown etching parameters.
Innovation Solution
A machine learning model is trained using SEM images of the wafer before and after etching to generate predictive images without damaging the wafer, utilizing a convolutional neural network (CNN) to minimize SEM-induced pattern alterations and account for etching variability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If SEM imaging is used to inspect the wafer after photoresist development, then inspection capability is achieved, but the photoresist is damaged and pattern accuracy deteriorates
Solution Approach 1:
The system performs SEM imaging on a first portion of the wafer before the etching process to capture the original pattern. This preliminary imaging preserves the undamaged pattern information that can be used for accurate key performance indicator measurement, avoiding the need to image the damaged photoresist after development.
Solution Approach 2:
The system creates a predictive image of the etched wafer pattern by training a machine learning model on SEM images taken before etching. This predictive image serves as a copy of what the etched wafer will look like without actually imaging the damaged photoresist, thereby maintaining pattern accuracy while achieving inspection capability.
2Difficulty of detecting and measuring
If SEM imaging is performed after photoresist development, then defect detection is enabled, but photoresist shrinkage occurs and key performance indicators become inaccurate
Solution Approach 1:
The system performs SEM imaging on a first portion of the wafer before the etching process to capture the original pattern. This preliminary imaging preserves the undamaged pattern information that can be used for accurate key performance indicator measurement, avoiding the need to image the damaged photoresist after development.
Solution Approach 2:
The system introduces a machine learning model as an intermediary that predicts the etched wafer pattern based on pre-etch SEM images. This intermediary allows defect detection capability while avoiding direct imaging of the damaged photoresist, thereby maintaining measurement precision for key performance indicators.
3Ease of manufacture
If physics-based models are used to generate predictive images, then computational approach is provided, but accuracy deteriorates due to unknown etching parameters
Solution Approach 1:
The system transitions from using fixed physics-based model parameters to using data-driven parameters learned from actual SEM images through machine learning. This parameter change allows the model to adapt to specific process conditions and achieve higher predictive accuracy without requiring precise knowledge of etching parameters.
Solution Approach 2:
The system replaces the physics-based computational model with a machine learning model that learns patterns directly from image data. This substitution eliminates the need for accurate physics-based etching parameter knowledge while maintaining computational efficiency and improving predictive accuracy.
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
The machine learning model allows for accurate, non-damaging inspection and higher throughput by generating predictive SEM images of the wafer after development, improving metrology and reducing wafer damage.
Implementation Method 1
The primary electrons interact with the wafer and may be backscattered or may cause the wafer to emit secondary electrons
Implementation Method 2
The primary electrons interact with the wafer and may be backscattered or may cause the wafer to emit secondary electrons
Data Source
AI summary
A system and method for generating predictive images for wafer inspection using machine learning are provided. Some embodiments of the system and method include acquiring the wafer after a photoresist applied to the wafer has been developed; imaging a portion of a segment of the developed wafer; acquiring the wafer after the wafer has been etched; imaging the segment of the etched wafer; training a machine learning model using the imaged portion of the developed wafer and the imaged segment of the etched wafer; and applying the trained machine learning model using the imaged segment of the etched wafer to generate predictive images of a developed wafer. Some embodiments include imaging a segment of the developed wafer; imaging a portion of the segment of the etched wafer; training a machine learning model; and applying the trained machine learning model to generate predictive after-etch images of the developed wafer.