Overlay Measurement Device Synthesizing Reference Images from Circuit Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing overlay measurement methods in semiconductor manufacturing rely on ideal reference images, which may not always be available, especially during the initial prototyping stages where quality variations occur, and fail to accurately calculate overlay without a suitable reference image.
Innovation Solution
A device and method that include an imaging unit, pattern recognition, reference image generation, quantification, and calculation units to synthesize reference images from multiple pattern images, allowing for overlay calculation even without an ideal reference image by quantifying differences and calculating offset amounts within the circuit patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If overlay measurement is performed using a dedicated mark around a chip with interpolation, then measurement can be performed, but measurement precision deteriorates when miniaturization of semiconductor process occurs and allowable overlay range reduces
Solution Approach 1:
The invention extracts the measurement function from dedicated marks around chips and moves it directly to the product circuit patterns themselves. By using the actual circuit patterns as measurement targets, the system eliminates the need for interpolation and directly measures overlay at arbitrary positions within the chip, thereby improving measurement precision to match the reduced allowable overlay range in miniaturized processes.
Solution Approach 2:
The invention transitions from measuring overlay at peripheral marks and interpolating to the center, to directly measuring overlay at arbitrary positions within the circuit pattern area. This dimensional shift enables measurement at the exact location where precision is critical, rather than relying on extrapolation from surrounding areas.
2Adaptability or versatility
If overlay measurement is performed in process initial step with quality variation, then measurement should be possible, but ideal reference image is not always obtainable
Solution Approach 1:
The system performs self-calibration by automatically generating reference images from the captured images themselves during the measurement process. Instead of requiring pre-prepared ideal reference images, the measurement device extracts pattern information from multiple captured images, synthesizes reference images, and uses these for overlay calculation, enabling reliable measurement even in process initial steps with quality variations.
Solution Approach 2:
The invention performs preliminary synthesis of reference images from multiple captured images before conducting the final overlay measurement. By pre-processing the images to create synthesized reference images that represent the circuit patterns, the system prepares reliable measurement references in advance, even when individual captured images may have quality variations.
3Adaptability or versatility
If reference image is synthesized from multiple images, then overlay calculation becomes possible without ideal reference image, but device complexity increases
Solution Approach 1:
The invention merges multiple captured images to synthesize reference images, combining the information from several images to create a more reliable reference. By superimposing and averaging the pattern information from multiple images, the system generates robust reference images that compensate for individual image quality variations, enabling overlay calculation without requiring perfectly ideal reference images.
Solution Approach 2:
The system creates synthesized reference images as copies or representations of the actual circuit patterns by processing multiple captured images. These synthesized references serve as virtual models of the circuit patterns, allowing overlay measurement without needing physical ideal reference images, thereby increasing adaptability while managing complexity through software-based image processing.
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
Enables accurate overlay measurement in the initial prototyping stages by synthesizing reference images from multiple pattern images, improving the stability and reproducibility of overlay calculations, even when ideal reference images are not obtainable.
Implementation Method 1
an imaging unit for capturing an image of a circuit pattern of a surface of an inspection object by an optical microscope or an electronic microscope
Implementation Method 2
an imaging unit for capturing an image of a circuit pattern of a surface of an inspection object by an optical microscope or an electronic microscope
Data Source
AI summary
An overlay measurement method using a reference image is an effective method for an overlay measurement in a product circuit. However, there is a problem that it is not possible to obtain an ideal reference image in a process of prototyping.A measurement device described in a specific embodiment of the present invention includes an imaging unit that captures an image of a circuit pattern of a semiconductor wafer surface by an optical microscope or an electronic microscope, a pattern recognition unit that extracts a first pattern and a second pattern from the image captured by the imaging unit, a reference image generation unit that synthesizes a first reference image using the first pattern extracted from a plurality of the images and synthesizes a second reference image using the second pattern extracted from the plurality of images, a quantification unit that quantifies a first difference that is a difference between the first reference image and the first pattern and a second difference that is a difference between the second reference image and the second pattern, and a calculation unit that calculates an overlay amount included in the circuit pattern using the first difference and the second difference.


