Dual-FOV Chip Mounting Alignment for Sub-μm Bonding Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semiconductor chip bonding technologies face challenges in achieving high precision mounting due to the need for sub-μm accuracy and the impact of mechanical and thermal changes, particularly in flip chip bonding, which requires high temperature and pressure, leading to deteriorated mounting precision.
Innovation Solution
A mounting device and method utilizing a dual-field-of-view (FOV) optical system with an image sensor to simultaneously capture alignment marks on both objects, allowing for precise adjustment of their relative position through a bonding head and stage, and incorporating fiducial marks to correct for image sensor misalignments and optical changes over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Strength
If high temperature and pressure are applied during flip chip bonding, then bonding strength is improved, but mounting precision deteriorates due to mechanical and thermal changes
Solution Approach 1:
The patent performs alignment mark recognition and position adjustment before the bonding process. The bonding head or stage is adjusted based on recognized alignment marks to achieve precise positioning prior to applying high temperature and pressure, preventing precision deterioration during bonding
Solution Approach 2:
The patent replaces mechanical alignment methods with an optical system that uses image sensors to recognize alignment marks. This optical measurement system enables high-precision positioning without the mechanical errors that would be exacerbated by thermal and mechanical changes during bonding
2Manufacturing precision
If sub-μm precision is required for TSV chip bonding, then bonding quality is improved, but the complexity of the mounting device increases
Solution Approach 1:
The patent divides the optical system into two fields of view: a first FOV for recognizing alignment marks and a second FOV for observing the bonding process. This segmentation allows each optical system to be optimized for its specific function, achieving high precision without requiring an overly complex unified system
Solution Approach 2:
The patent introduces alignment marks as intermediary elements that facilitate precise positioning. These marks serve as reference points that the optical system can detect and use to calculate the exact position of the chip, simplifying the overall positioning task while achieving sub-μm precision
3Device complexity
If a single optical system is used for alignment, then device complexity is reduced, but alignment precision deteriorates due to inability to simultaneously capture both objects
Solution Approach 1:
The patent segments the optical observation into two distinct fields of view: one for capturing the first object and its alignment mark, and another for capturing the second object and its alignment mark. This allows simultaneous recognition of both objects with high precision without requiring complex mechanical coordination
Solution Approach 2:
The patent merges the alignment recognition function and bonding observation function into a single integrated optical system with dual FOV. This combining approach achieves the precision benefits of separate systems while maintaining the simplicity of a unified device structure
Data Source
AI summary
A mounting device includes: a bonding head configured to hold a first object, a bonding stage configured to hold a second object, and a dual-field-of-view (FOV) optical system including an image sensor configured to simultaneously capture an image of a first alignment mark on the first object and an image of a second alignment mark on the second object to obtain a first image. At least one of the bonding head and the bonding stage is configured to adjust a relative position between the first object and the second object based on the first image, and bond the first object to the second object.