Backside Image Sensor Pixel With Vertical Color Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional back-side illumination image sensors suffer from reduced sensitivity and poor color fidelity due to light absorption by color filters and thermal noise, leading to issues like aliasing and moiré effects.
Innovation Solution
The image sensor employs a back-side illumination structure with vertically superposed doped photosensitive regions framed by vertical annular gates, generating electrostatic potential wells and barriers to capture all colors of the light signal without filters, enhancing sensitivity and reducing artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If color filters are placed on the pixel surface to extract color information, then color reconstruction is enabled, but sensitivity is reduced due to light absorption by filters
Solution Approach 1:
The patent transitions from lateral color filtering in the pixel plane to vertical color separation along the depth dimension. Multiple photosensitive regions are stacked at different depths, with each region capturing a specific color band based on its depth position. This vertical arrangement eliminates the need for lateral color filters that absorb light, as each depth-layered region directly captures its designated color wavelength range without filtering losses.
Solution Approach 2:
Instead of using color filters to block unwanted wavelengths and extract color information, the patent inverts the approach by using depth-positioned photosensitive regions to selectively absorb specific color bands. Rather than filtering out colors, the system lets different colors penetrate to different depths where corresponding photosensitive regions capture them, effectively inverting the color separation mechanism from filtering to selective absorption by depth.
2Measurement precision
If conventional color-selection techniques based on absorption depth are used, then color information can be separated, but thermal noise increases detrimentally
Solution Approach 1:
The pixel is segmented into multiple vertically stacked photosensitive regions at different depths, with each region dedicated to capturing a specific color band. This segmentation is achieved through controlled doping profiles that create distinct potential wells at different depths. By separating the color capture function into discrete depth-layered segments rather than using continuous absorption-depth techniques, the patent reduces thermal noise while maintaining color separation capability.
Solution Approach 2:
The patent modifies the electrical parameters of the photosensitive regions through controlled doping to create distinct potential wells at different depths. By adjusting dopant concentrations and distribution, the system creates electric field configurations that guide charge carriers to specific readout nodes based on their generation depth. This parameter control allows precise color separation without relying on thermal diffusion processes that generate noise.
3Loss of energy
If vertically superposed photosensitive regions are used to capture all colors, then sensitivity improves, but device complexity increases
Solution Approach 1:
The vertically stacked photosensitive regions serve multiple functions simultaneously: they act as both color separation elements and charge generation regions. Each depth-layered region captures a specific color band while also functioning as a photodetector that generates charge carriers. This multi-functionality eliminates the need for separate color filter layers and simplifies the overall pixel structure despite the vertical stacking, as each region performs both color selection and light detection.
Solution Approach 2:
The patent implements a nested structure where multiple photosensitive regions are vertically contained within a single pixel volume, with each region nested at a different depth. The doping profiles are nested such that each deeper region is surrounded by the electrical influence of upper regions, creating a hierarchical charge collection structure. This nesting allows compact integration of multiple color-capture functions within a single pixel footprint without requiring lateral expansion.
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 approach improves sensitivity and color fidelity while minimizing light loss and thermal noise, allowing for effective capture and reconstruction of original colors without filter-related losses and noise artifacts.
Implementation Method 1
bias the vertical annular gates during an integration phase, so as to generate an electrostatic potential comprising potential wells in the central portion of the volume of each photosensitive region and a potential barrier at each interface between two neighboring photosensitive regions
Implementation Method 2
As the semiconductor substrate absorbs various colors of the light at various depths, each photosensitive region captures a different color
Implementation Method 3
each pixel comprising a plurality of, for example three, doped photosensitive regions that are superposed vertically in the substrate between the back side and the front side
Data Source
AI summary
A back side illuminated image sensor includes a pixel formed by three doped photosensitive regions that are superposed vertically in a semiconductor substrate. Each photosensitive region is laterally framed by a respective vertical annular gate. The vertical annular gates are biased by a control circuit during an integration phase so as to generate an electrostatic potential comprising potential wells in the central portion of the volume of each doped photosensitive region and a potential barrier at each interface between two neighboring doped photosensitive regions.