Semiconductor Yield Root Cause Detection via Spatial Metrology Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor manufacturing processes face challenges in identifying root causes for low yield in IC production, as issues are only determined after all process steps are completed, leading to delayed detection of process problems and potential recurrence of failures.
Innovation Solution
A method involving the comparison of yield distribution data with metrology data from different layers of the substrate, using a trained model to determine similarity metrics and flag potential issues, allowing for early detection of yield-related problems through scanner metrology data analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If electrical testing is performed only after all process steps are completed, then comprehensive yield assessment is achieved, but detection time is delayed and root cause identification is postponed
Solution Approach 1:
The patent applies preliminary action by performing metrology measurements on intermediate layers during the manufacturing process, before all process steps are completed. This allows early detection of potential yield issues through similarity comparison with final yield maps, enabling proactive intervention rather than waiting for final electrical testing.
Solution Approach 2:
The patent uses metrology data from intermediate layers as an intermediary indicator to predict final yield. Instead of directly measuring final electrical properties only at the end, the system uses intermediate structural measurements as proxies that correlate with eventual yield outcomes, enabling earlier assessment.
2Loss of time
If metrology measurements are performed on all layers during manufacturing, then early detection capability is improved, but process complexity and measurement time increase
Solution Approach 1:
The patent extracts only the necessary metrology data from intermediate layers that shows similarity to final yield patterns, rather than analyzing all possible measurements from all layers. This selective extraction approach reduces complexity while maintaining early detection capability.
Solution Approach 2:
The patent performs partial action by conducting metrology measurements on selected intermediate layers rather than all layers, and using similarity comparison on key features rather than complete data sets. This partial approach provides sufficient early detection benefit without the full complexity burden.
3Measurement precision
If comprehensive electrical testing is performed at the end, then accurate yield determination is achieved, but corrective actions cannot be implemented for already completed process steps
Solution Approach 1:
The patent enables preliminary detection of yield issues through intermediate metrology measurements and similarity analysis. When potential problems are detected early, corrective actions can be implemented on subsequent process steps before the substrate is fully manufactured, improving ease of manufacture by preventing defect propagation.
Solution Approach 2:
The patent implements feedback by comparing intermediate metrology data with expected yield patterns and using this information to guide subsequent manufacturing decisions. This feedback loop enables real-time process adjustment rather than waiting for final testing results.
Data Source
AI summary
A method for determining a root cause affecting yield in a process for manufacturing devices on a substrate, the method including: obtaining yield distribution data including a distribution of a yield parameter across the substrate or part thereof; obtaining sets of metrology data, each set including a spatial variation of a process parameter over the substrate or part thereof corresponding to a different layer of the substrate; comparing the yield distribution data and metrology data based on a similarity metric describing a spatial similarity between the yield distribution data and an individual set out of the sets of the metrology data; and determining a first similar set of metrology data out of the sets of metrology data, being the first set of metrology data in terms of processing order for the corresponding layers, which is determined to be similar to the yield distribution data.