Backside Illumination Image Sensor With Tapered Pixel Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CMOS image sensors face challenges in achieving improved light-receiving efficiency and auto-focus functionality, particularly in backside illumination configurations, where the existing designs are complex and not optimized for concise processes.
Innovation Solution
The image sensor design incorporates a substrate with pixel isolation and separation patterns that decrease in width from the backside to the light-receiving surface, allowing for efficient light detection and auto-focus capabilities by arranging unit pixels in different directions, utilizing a backside illumination structure and FDTI process for concise implementation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a backside illumination structure is used to improve light-receiving efficiency, then light sensitivity is improved, but the structure becomes more complex
Solution Approach 1:
The patent inverts the conventional front-illumination structure by implementing backside illumination, where light enters through the back surface of the substrate rather than the front. This inversion allows light to directly reach the photoelectric conversion parts without passing through metal interconnect layers, thereby improving light sensitivity while maintaining a relatively simple overall structure through the FDTI process
Solution Approach 2:
The patent extracts and removes the pixel isolation pattern from the traditional deep trench structure, replacing it with a simplified isolation implementation. This extraction of the complex isolation structure reduces manufacturing complexity while maintaining the backside illumination functionality and light-receiving efficiency
2Manufacturing precision
If pixel isolation patterns are used to define unit pixels, then pixel boundaries are clearly defined, but the manufacturing process becomes more complex
Solution Approach 1:
The patent extracts and simplifies the pixel isolation pattern by removing the deep trench structure traditionally used for pixel isolation. Instead, it uses a simplified isolation pattern that still clearly defines pixel boundaries but can be manufactured with less complex processes, thereby reducing manufacturing complexity while maintaining precise pixel boundary definition
Solution Approach 2:
The patent changes the parameters of the isolation pattern by using width variations (narrower at the light-receiving surface, wider at the back surface) and strategic positioning (between photoelectric conversion parts in different directions) to achieve effective pixel isolation with simpler manufacturing compared to traditional uniform deep trench structures
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
This design enhances light-receiving efficiency and auto-focus functionality while maintaining a concise structure and process, improving the performance of CMOS image sensors without adding complexity.
Implementation Method 1
The first unit pixel includes a first photoelectric conversion part and a second photoelectric conversion part arranged along a first direction. The second unit pixel includes a third photoelectric conversion part and a fourth photoelectric conversion part arranged along a second direction intersecting the first direction.
Data Source
AI summary
An image sensor includes a substrate with a first surface opposite a second surface, a pixel isolation pattern defining first and second unit pixels adjacent to each other in the substrate, and first and second separation patterns in the substrate. The first unit pixel includes first and second photoelectric conversion parts along a first direction. The second unit pixel includes third and fourth photoelectric conversion parts along a second direction intersecting the first direction. The first separation pattern extends in the second direction between the first and second photoelectric conversion parts. The second separation pattern extends in the first direction between the third and fourth photoelectric conversion parts. A width of the pixel isolation pattern, a width of the first separation pattern, and a width of the second separation pattern each decrease from the second surface of the substrate toward the first surface of the substrate.