Calibration Sample Grid Pattern for Coordinate Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current surface inspection apparatuses for semiconductor manufacturing face challenges in achieving high coordinate accuracy for defect detection due to non-uniform distribution of calibration points, leading to variations in correction results and insufficient accuracy, especially in areas with sparse error information.
Innovation Solution
A sample for coordinates calibration with intentional defects laid out at uniform spaces in radial and circumferential directions is used, featuring a circular substrate with a grid pattern of squares, where the center point coincides with the substrate center, and multiple defects are formed along each grid line, allowing for improved coordinate correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If calibration points are laid out only at center points of unit grids in radial directions, then the calibration process is simple, but the coordinate accuracy is insufficient due to non-uniform distribution of error information
Solution Approach 1:
The calibration sample is divided into multiple unit grids arranged in a two-dimensional matrix pattern, with each unit grid containing calibration points at its center point and four corner points. This segmentation approach ensures uniform distribution of calibration points across the entire sample surface, providing comprehensive error information for accurate coordinate calibration of the surface inspection apparatus.
Solution Approach 2:
The calibration pattern transitions from a one-dimensional radial arrangement to a two-dimensional matrix pattern. Unit grids are arranged both in radial directions and circumferential directions, creating a comprehensive two-dimensional calibration network. This dimensional expansion ensures that calibration points are uniformly distributed across the entire sample surface, eliminating sparse error information regions.
2Productivity
If calibration points are sparsely distributed in certain areas, then the calibration process is faster, but the correction results show significant variations in those areas
Solution Approach 1:
The calibration pattern achieves equipotential distribution of calibration points by arranging unit grids in a regular matrix pattern with consistent spacing in both radial and circumferential directions. Each unit grid contributes calibration points at equivalent positions (center and corners), ensuring uniform density of calibration information across the entire sample surface. This eliminates potential differences in correction accuracy between different regions.
Solution Approach 2:
The calibration sample employs homogeneous distribution of calibration points through the two-dimensional matrix pattern of unit grids. All unit grids follow the same structure with calibration points at identical relative positions, creating uniform calibration point density throughout the sample. This homogeneity ensures consistent correction accuracy across all areas of the sample surface.
3Reliability
If a virtual grid pattern with unit grids at equal spaces is used, then the calibration structure is regular, but the error information coverage is insufficient for high-power inspection requirements
Solution Approach 1:
The calibration pattern merges radial and circumferential grid arrangements into a unified two-dimensional matrix structure. Unit grids are positioned at intersections of radial and circumferential lines, combining the advantages of both radial symmetry and circumferential uniformity. This merging creates a dense network of calibration points that provides comprehensive error information coverage for reliable high-power defect detection.
Solution Approach 2:
The calibration pattern uses repeated copying of the unit grid structure across the entire sample surface. Each unit grid is an identical copy containing calibration points at the same relative positions (center and four corners). This systematic copying ensures uniform distribution of calibration points and provides sufficient error information coverage throughout the sample area for reliable defect detection.
Data Source
AI summary
A sample for coordinates calibration including (1) a substrate having a circular plate-shape, and (2) multiple intentional defects that form a grid pattern with squares as unit grids on a surface of the substrate, the intentional defect providing a center point of the grid pattern coinciding with a center point of the substrate and, letting the maximum value of a number of the unit grids arranged from the center point of the substrate in radial directions be N (a natural number equal to or larger than two), a number of the intentional defects formed at equal spaces along one side of the unit grid being N+1 including the two intentional defects providing a vertex of the unit grid is proposed.