Stacked Image Sensor Readout for Low-Noise Color Separation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional photo-sensitive devices, particularly those using silicon substrates, suffer from limitations such as reduced detectable wavelengths, high noise characteristics, low conversion gains, and significant loss of light and resolution due to their optical and dimensional constraints, as well as inferior kTC noise performance and high dark current.

Innovation Solution

A thin-film, stacked photo-sensitive device with a fully depleted pixel structure and a common floating electrical connection is developed, allowing for improved light absorption, enhanced resolution, and reduced noise by vertically stacking multiple photo-sensitive layers with specific wavelength responses and integrating them with a read-out integrated circuit (ROIC) for efficient signal processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If a color filter array is stacked on top of a pixel to selectively absorb each color signal, then color signal separation is achieved, but one-third loss of light signal and three-quarters loss of resolution occur

Engineering Contradiction:
Improvecolor signal separationVSAvoidlight signal loss
Core Design Contradiction:
Ease of manufactureVSLoss of energy

Solution Approach 1:

The patent transitions from a planar color filter array to a vertically stacked three-dimensional architecture where multiple photodiodes are stacked in the depth direction. This dimensional change allows simultaneous capture of multiple color signals without lateral filtering, eliminating light loss while maintaining color separation capability through vertical stacking of blue, green, and red photodiodes.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The pixel is segmented into multiple independent photodiode units stacked vertically, each responsible for detecting a specific wavelength range. This segmentation allows direct conversion of incident light to electrical signals in each layer without requiring a color filter array, thereby eliminating the one-third light loss associated with traditional CFA approaches.

Inventive Principle:
Principle #1Segmentation

2Adaptability or versatility

If a three-transistor pixel circuit is used with quantum dot detector material, then CMOS read-out integration is achieved, but unacceptably high reset noise occurs due to un-correlated read-out

Engineering Contradiction:
ImproveCMOS read-out integrationVSAvoidreset noise
Core Design Contradiction:
Adaptability or versatilityVSObject-generated harmful factors

Solution Approach 1:

The patent merges the read-out operations of multiple stacked photodiodes into a single correlated read-out process. By sharing common read-out circuitry and synchronizing the read-out timing of all stacked photodiodes, the system achieves correlated double sampling that eliminates reset noise, while maintaining CMOS integration through shared transistor structures.

Inventive Principle:
Principle #5Merging (Combining)

3Device complexity

If the photodiode area is directly connected to a pixel circuit below the photodiode area, then integration is achieved, but a defective contact region generates high dark current

Engineering Contradiction:
ImproveintegrationVSAvoiddark current
Core Design Contradiction:
Device complexityVSObject-generated harmful factors

Solution Approach 1:

The patent introduces an intermediate transfer region and transfer gate structure between the stacked photodiode and the pixel circuit. This intermediary layer provides a controlled interface that isolates the photodiode from direct contact with the pixel circuit, preventing dark current generation from defective contact regions while maintaining electrical connectivity through the transfer gate mechanism.

Inventive Principle:
Principle #24Intermediary (Mediator)

4Device complexity

If a photodiode capacitance is added to a floating diffusion node where charge-to-voltage conversion takes place, then integration is achieved, but conversion gain is low and noise characteristics are poor

Engineering Contradiction:
Improvecharge-to-voltage conversion integrationVSAvoidconversion gain
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent extracts the photodiode capacitance from the floating diffusion node by implementing separate read-out circuits for each stacked photodiode. This extraction allows the photodiode to operate with its full capacitance benefit for charge storage while the floating diffusion node is used solely for low-noise charge-to-voltage conversion, thereby maximizing conversion gain and improving noise characteristics.

Inventive Principle:
Principle #2Taking out (Extraction)

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

The solution enables a compact, optically efficient image sensor with improved noise characteristics and high conversion gain, maximizing light absorption and minimizing signal loss, thereby enhancing image quality and resolution.

Implementation Method 1

an active layer (113), configured to generate electrical charges in response to a first predefined range of wavelengths of light incident on the device (100)

Methodology Applied
Scientific EffectPhotoelectric effect: Photoelectric Effect

Implementation Method 2

a bottom electrode (116) separated by a dielectric material (115) from the second charge transport layer (114) wherein the bottom electrode (116) is configured to provide a first electrical voltage for at least partially depleting the first portion (114a) of the corresponding second charge transport layer (114)

Methodology Applied
Scientific EffectElectric field: Electric Field

Implementation Method 3

a transfer gate electrode (117) separated by a dielectric material (115) from the second charge transport layer (114), wherein the transfer gate electrode (117) is configured to control transfer of electrical charges accumulated in the first portion (114a) via the transfer region (114b)

Methodology Applied
Scientific EffectElectrical conduction: Conduction (electrical)

Data Source

PatentEP4030480B1Image sensor comprising stacked photo-sensitive devices
Publication Date: 2024.02.28 INTERUNIVERSITAIR MICRO ELECTRONICS CENT (IMEC VZW)
  • EP4030480B1 patent drawingFigure 1
  • EP4030480B1 patent drawingFigure 2
  • EP4030480B1 patent drawingFigure 3

AI summary

An image sensor comprises at least two vertically stacked photo-sensitive devices wherein each respective photo-sensitive device comprises a stack of a top electrode, a first charge transport layer and an active layer. Each respective stack generates electrical charges in response to a corresponding predefined range of wavelengths of light incident on the image sensor. Each photo-sensitive device further comprises a second charge transport layer having a first portion, vertically aligned underneath the active layer, and a second portion, transfer region, protruding laterally to extend beyond the active layer. A dielectric layer separates the first portion from a bottom electrode providing a voltage for depleting the first portion, and the transfer region from a transfer gate providing a voltage for transferring the generated electrical charge to a floating electrical connection, shared by all stacked photo-sensitive devices. The floating electrical connection couples to a read-out-circuitry.