Backside Illuminated Image Sensor with P+ Implantation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As image sensors and their pixels continue to shrink, they face challenges in efficiently capturing photonically generated charges, leading to reduced image quality due to increased dark current and decreased sensitivity, especially for longer wavelengths.
Innovation Solution
The implementation of a front side illuminated imaging sensor with a highly doped P+ implantation layer and optimized remaining substrate thickness, using boron implantation and laser annealing, to enhance quantum efficiency and reduce dark current, while selectively reflecting longer wavelengths back towards the front surface for improved sensitivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If image sensors and pixels are shrunk to reduce size, then device miniaturization is achieved, but quantum efficiency decreases and dark current increases
Solution Approach 1:
The patent inverts the conventional light incident direction by using backside illumination instead of frontside illumination. This allows light to enter through the back surface of the substrate, bypassing metal interconnect layers and other front-side structures that block photons in miniaturized sensors, thereby maintaining quantum efficiency despite reduced sensor size
Solution Approach 2:
The patent changes the doping concentration parameter by forming a highly doped P+ implantation layer with concentration of 1×10^19 to 1×10^21 atoms/cm³ at the back surface. This parameter change creates strong electric field effects that suppress dark current generation while maintaining efficient charge collection, addressing the dark current issue that arises from sensor miniaturization
2Productivity
If pixel size is reduced to increase array density, then array integration is improved, but sensitivity to longer wavelengths decreases
Solution Approach 1:
By inverting the illumination direction to backside illumination, longer wavelength photons (red and near-IR) can penetrate deeper into the substrate without being blocked by front-side metal layers. This enables miniaturized pixels to maintain sensitivity to longer wavelengths while achieving high array integration
Solution Approach 2:
The patent utilizes the substrate thickness dimension by optimizing it to 3-10 micrometers, creating a sufficient light path length for longer wavelengths to be absorbed and generate charges. This dimensional optimization enables wavelength-sensitive detection in compact pixel structures
3Length of stationary object
If substrate thickness is reduced to minimize sensor depth, then device integration is improved, but charge collection efficiency decreases
Solution Approach 1:
The patent optimizes the substrate thickness parameter to a specific range of 3-10 micrometers and forms a highly doped P+ layer at the back surface. This combination of parameters creates an efficient charge collection mechanism where the thin substrate enables device integration while the high doping concentration ensures sufficient electric field for charge collection
Solution Approach 2:
The P+ implantation layer is formed in advance at the back surface before final device assembly. This preliminary action creates a built-in electric field structure that actively collects photogenerated charges, compensating for the reduced substrate thickness and maintaining collection efficiency
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 increases quantum efficiency and spectral performance, reduces dark current, and enhances sensitivity to red and near-IR wavelengths, resulting in improved image quality by effectively capturing and processing photonically generated charges.
Implementation Method 1
selectively reflecting longer wavelengths back towards the front surface for improved sensitivity
Implementation Method 2
using boron implantation and laser annealing
Implementation Method 3
using boron implantation and laser annealing
Data Source
AI summary
An image sensor includes a semiconductor layer that filters light of different wavelengths. For example, the semiconductor layer absorbs photons of shorter wavelengths and passes more photons of longer wavelengths such that the longer wavelength photons often pass through without being absorbed. An imaging pixel having a photodiode is formed near a front side of the semiconductor layer. A dopant layer is formed below the photodiode near a back side of the semiconductor layer. A mirror that primarily reflects photons of longer visible wavelengths is disposed on the back side of the semiconductor layer.


