Defect Location Refinement Using Coordinate Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current two-phase wafer inspection and review process, involving an optical wafer inspection system and a charged particle review system, faces inefficiencies due to inaccuracies in location measurements and misalignment, leading to a time-consuming search process for suspected defects.
Innovation Solution
A method that includes receiving location information from the optical wafer inspection system, filtering and grouping suspected defects, calculating permutation-related mappings to account for location errors, and selecting the most accurate mapping to refine the locations of suspected defects for precise review by the charged particle review system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the charged particle review system uses large search windows to account for location inaccuracies, then the reliability of defect identification is improved, but the productivity decreases due to time-consuming search process
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing the transformation matrix that maps coordinates between the optical inspection system and the charged particle review system. This matrix, derived from calibration data and error characteristics, is used to predict defect locations before the actual review process, allowing the system to focus search efforts in predetermined accurate regions rather than searching large areas blindly.
Solution Approach 2:
The patent replaces the mechanical brute-force search approach with a computational coordinate transformation system. Instead of physically scanning large areas with the charged particle beam, the system uses mathematical models and transformation matrices to calculate precise defect locations, substituting mechanical search with intelligent computational prediction.
2Measurement precision
If the charged particle review system uses large search windows centered around optical inspection locations, then the measurement precision is improved by accounting for errors, but the loss of time increases due to extended search area
Solution Approach 1:
The system performs preliminary coordinate transformation and error compensation calculations before the actual defect review. By pre-computing the most likely defect locations using the transformation matrix and error characteristics, the system prepares accurate search targets in advance, eliminating the need for time-consuming broad searches during the actual review process.
Solution Approach 2:
The patent creates a transformed coordinate model that copies and adapts the optical inspection coordinate system to the charged particle review system coordinate system. This virtual copy includes embedded error compensation information, allowing the system to predict defect locations in the review system's coordinate space without physical trial-and-error searching.
3Manufacturing precision
If the system accounts for multiple sources of location errors (X-Y offset, scaling, rotation), then the manufacturing precision is improved, but the device complexity increases due to multiple error sources to manage
Solution Approach 1:
The patent merges multiple independent error sources (X-Y offset, scaling errors, rotation errors) into a single unified transformation matrix. Instead of handling each error source separately through multiple adjustment mechanisms, the system combines their effects into one comprehensive mathematical model that can be applied through a single coordinate transformation operation.
Solution Approach 2:
The patent transforms the physical error parameters (offset distances, scaling factors, rotation angles) into mathematical transformation parameters within the coordinate system. By changing the representation of errors from physical adjustments to mathematical parameters, the system can compensate for all error types through computational parameter application rather than physical system modifications.
Data Source
AI summary
A method, a non-transitory computer readable medium and a detection system for determining locations of suspected defects of a substrate.