Harmonic Detectivity Metric for Overlay Metrology Recipe Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for recipe selection in overlay metrology are subjective and prone to score saturation, leading to the selection of low-quality or unstable recipes.
Innovation Solution
A metrology system that uses a Harmonic Detectivity Metric to quantify optical signals and calibrate metrology tools, allowing for objective recipe selection and improved measurement accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If current methods for recipe selection are used (combining normalized metrics into a single fused score), then the selection process is simplified, but the method suffers from score saturation and measurement condition ambiguity leading to low-quality or unstable recipe selection
Solution Approach 1:
The patent changes the parameter used for recipe selection from normalized fused scores to unnormalized raw metric values. This parameter change eliminates score saturation while maintaining ease of operation, as the system still provides automated selection but based on metrics that retain their full dynamic range and discriminatory power across different measurement conditions
Solution Approach 2:
The patent introduces an intermediary step that identifies and selects recipes based on the distribution characteristics of raw metric values across the measurement landscape. This intermediary selection process based on metric distributions rather than fused scores resolves the contradiction by providing reliable recipe selection without sacrificing operational simplicity
2Device complexity
If multiple metrics are normalized and combined into a single fused score for recipe selection, then the selection criteria are unified, but the dynamic range for recipe differentiation is limited causing score saturation
Solution Approach 1:
The patent segments the recipe selection process into two independent parts: (1) evaluation of multiple raw metrics without normalization to preserve their individual dynamic ranges and differentiation capabilities, and (2) selection based on the distribution patterns of these unnormalized metrics. This segmentation maintains measurement precision while managing complexity through a structured two-stage approach
3Adaptability or versatility
If subjective user decisions are used to weight and normalize metrics, then flexibility in metric selection is provided, but the process becomes subjective and leads to inconsistent recipe selection
Solution Approach 1:
The patent implements self-service by allowing the measurement landscape and metric distributions to speak for themselves without subjective user intervention. The system automatically identifies optimal recipes based on the inherent distribution patterns of raw metric values across the measurement space, eliminating subjective weighting decisions while maintaining the ability to select from multiple metrics and providing consistent, objective recipe selection
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
An image-based overlay metrology system is disclosed. The system includes a controller couplable to a metrology sub-system. The controller is configured to receive a set of image signals of a first metrology target disposed on the sample from the metrology sub-system and determine a plurality of harmonic detectivity metric values by calculating a harmonic detectivity metric value for each of the plurality of image signals. The controller is also configured to identify a set of optical measurement conditions of the metrology sub-system based on the plurality of harmonic detectivity metric values, wherein the set of optical measurement conditions define a recipe for optical metrology measurements of the metrology sub-system. The controller then provides the recipe to the metrology sub-system for execution of one or more optical metrology measurements of one or more additional metrology targets.