CMOS Image Sensor Isolation Structure for Cross-Talk and Full Well Capacity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CMOS image sensors face challenges in maintaining reliability due to issues such as cross-talk between neighboring pixel regions and reduced full well capacity, which can be exacerbated by non-vertical deep device isolation patterns leading to bowing and void formation.
Innovation Solution
The implementation of a deep device isolation pattern with specific width ratios and vertical orientation, including a horizontal and vertical portion, to prevent cross-talk and enhance full well capacity by reducing bowing and voids.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a conventional deep device isolation pattern is used, then manufacturing is simpler, but cross-talk between neighboring pixel regions occurs and full well capacity is reduced
Solution Approach 1:
The deep device isolation pattern is divided into multiple vertical portions (first, second, third vertical portions) with different widths at different locations. This segmentation allows each portion to serve specific functions: wider portions provide stronger isolation to prevent cross-talk, while narrower portions maintain full well capacity, thereby resolving the contradiction between reliability and device complexity.
Solution Approach 2:
Different portions of the deep device isolation pattern have different width characteristics tailored to local requirements. The first vertical portion has a greater width to prevent cross-talk at critical interfaces, while the second and third portions have reduced widths to preserve full well capacity in regions where isolation pressure is lower. This local differentiation resolves the contradiction by optimizing isolation effectiveness without uniformly increasing complexity.
2Reliability
If the deep device isolation pattern is made wider to prevent cross-talk, then reliability improves, but full well capacity is reduced
Solution Approach 1:
The isolation pattern is segmented into portions with different widths placed at different locations. The first vertical portion has a greater width to provide strong isolation and prevent cross-talk, while the second and third portions have smaller widths to minimize impact on full well capacity. This segmentation resolves the contradiction by concentrating isolation strength where most needed while preserving charge storage capacity in other regions.
Solution Approach 2:
The width of the deep device isolation pattern varies locally according to isolation requirements. At interfaces between pixel regions where cross-talk risk is highest, the isolation pattern is wider (first vertical portion). In regions where cross-talk risk is lower, the isolation pattern is narrower (second and third vertical portions) to preserve full well capacity. This local quality differentiation resolves the contradiction between isolation effectiveness and full well capacity.
3Reliability
If the deep device isolation pattern has non-vertical orientation, then manufacturing is easier, but bowing and void formation occur reducing reliability
Solution Approach 1:
The method forms the deep device isolation pattern by sequentially creating trenches and filling them with insulating material in a controlled multi-step process. By preliminarily defining the trench geometry and then filling it, the process ensures vertical sidewalls and prevents bowing or void formation, thereby achieving high reliability without sacrificing manufacturability. The preliminary trench formation step sets up the precise geometry needed for subsequent filling.
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 improves the reliability and performance of CMOS image sensors by minimizing cross-talk and increasing full well capacity, thereby enhancing image quality and operational efficiency.
Implementation Method 1
a photoelectric conversion region in a pixel region among the plurality of pixel regions
Data Source
AI summary
An image sensor includes a gate electrode on a first surface of a substrate; a photoelectric conversion region in a pixel region among the plurality of pixel regions; and a deep device isolation pattern extending around the plurality of pixel region, The deep device isolation pattern includes a vertical portion with a lowermost portion; an uppermost portion; a central portion; a lower middle portion; and an upper middle portion. A ratio of a width of the uppermost portion to a width of the central portion is between 1:0.9 and 1:1.1. The deep device isolation pattern is spaced apart from the first surface of the substrate.