3D Stacked CMOS and Bolometer Imager for Infrared Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional silicon-based CMOS imagers have limited light absorption/detection properties, particularly in the infrared range, due to their transparency to infrared light, requiring substantial path lengths and absorption depths to detect longer wavelengths effectively.
Innovation Solution
The development of semiconductor devices with 3D architectures, where a defect-free semiconductor layer is formed on a front side and a bolometer is coupled to a processed surface on the backside, separated by a cavity, allowing for improved light absorption through texturing and low-temperature processing to maintain defect-free conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional silicon-based CMOS imagers are used, then manufacturing cost is reduced and visible light detection is achieved, but infrared light absorption is poor and detection capability is limited
Solution Approach 1:
The patent combines a visible light CMOS imager and an infrared bolometer imager into a single integrated device. The CMOS imager detects visible light (400-1100 nm) while the bolometer detects infrared wavelengths (8-14 μm), allowing both detection capabilities to coexist in one device without requiring separate manufacturing processes.
Solution Approach 2:
The patent positions the bolometer on the backside of the CMOS imager, creating a three-dimensional stacked architecture. This vertical arrangement allows infrared detection to occur in a different spatial dimension from visible light detection, eliminating spectral interference and enabling both functions to operate simultaneously without compromising manufacturing simplicity.
2Reliability
If silicon photodetectors are used for visible light, then manufacturing is simple and cost-effective, but path length and absorption depth must be substantial to detect infrared photons
Solution Approach 1:
The patent uses a bolometer as an alternative detection mechanism for infrared wavelengths rather than relying on silicon photodetector absorption. The bolometer measures temperature changes caused by absorbed infrared radiation, providing a different physical detection approach that is more sensitive for long-wavelength infrared photons without requiring increased absorption depth.
3Reliability
If backside illumination is used to improve light absorption, then detection efficiency increases, but surface defects and noise increase
Solution Approach 1:
The patent segments the detection function into two separate components: a frontside-illuminated CMOS imager for visible light and a backside-mounted bolometer for infrared detection. This segmentation allows each component to operate in its optimal illumination mode without suffering from the defects associated with backside illumination of silicon photodetectors.
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 enhances light absorption and detection capabilities across a broader spectral range, including infrared, by reducing surface defects and increasing the effective absorption length, leading to improved signal-to-noise ratios and detection efficiency.
Implementation Method 1
coupling a bolometer to the processed surface
Implementation Method 2
silicon based detectors are mostly transparent to infrared light... can covert visible incident light into a photocurrent
Data Source
AI summary
Semiconductor devices having three dimensional (3D) architectures and methods form making such devices are provided. In one aspect, for example, a method for making a semiconductor device can include forming a device layer on a front side of a semiconductor layer that is substantially defect free, bonding a carrier substrate to the device layer, processing the semiconductor layer on a back side opposite the device layer to form a processed surface, and bonding a smart substrate to the processed surface. In some aspects, the method can also include removing the carrier substrate from the semiconductor layer to expose the device layer.