IEDM 2024 Program is Live

70th Annual IEEE International Electron Devices Meeting (IEDM) will be held December 7-11, 2024 in San Francisco California. Session #41 is on the topic of "Advanced Image Sensors":

https://iedm24.mapyourshow.com/8_0/sessions/session-details.cfm?scheduleid=58

Title: 41 | ODI | Advanced Image Sensors
Description:
This session includes 6 papers on latest image sensor technologies developments. To be noticed this year the multiple ways of stacking layer with new features. The first stack involves a dedicated AI image processing layer based on neural networks for a 50 Mpix sensor. The second one shows progress on small pixel noise with 2-layer pixel and additional intermediate interconnection. Third stack, very innovative with organic pixel on top of conventional Si based ITOF pixel for true single device RGB-Z sensor. All three papers are authored by Sony Semiconductors. InAs QD image sensors are also reported for the first time as a lead-free option for SWIR imaging by both IMEC and Sony Semiconductors Also progress in conventional IR global shutter with newly nitrated MIM capacitor and optimized DTI filling for crosstalk and QE improvement is presented by Samsung semiconductor.

Wednesday, December 11, 2024 - 01:35 PM
41-1 | A Novel 1/1.3-inch 50 Megapixel Three-wafer-stacked CMOS Image Sensor with DNN Circuit for Edge Processing
This study reports the first ever 3-wafer-stacked CMOS image sensor with DNN circuit. The sensor was fabricated using wafer-on-wafer-on-wafer process and DNN circuit was placed on the bottom wafer to ensure heat dissipation. This device can incorporate the HDR function and enlarge the pixel array area to remarkably improve image-recognition.

Wednesday, December 11, 2024 - 02:00 PM
41-2 | Low Dark Noise and 8.5k e− Full Well Capacity in a 2-Layer Transistor Stacked 0.8μm Dual Pixel CIS with Intermediate Poly-Si Wiring
This paper demonstrates a 2-layer transistor pixel stacked CMOS image sensor with the world’s smallest 0.8μm dual pixel. We improved the layout flexibility with intermediate poly-Si wiring technique. Our advanced 2-layer pixel device achieved low input-referred random noise of 1.3 e−rms and high full well capacity of 8.5k e−.

Wednesday, December 11, 2024 - 02:25 PM
41-3 | A High-Performance 2.2μm 1-Layer Pixel Global Shutter CMOS Image Sensor for Near-Infrared Applications
A high performance and low cost 2.2μm 1-layer pixel near infrared (NIR) global shutter (G/S) CMOS image sensor (CIS) was demonstrated. In order to improve quantum efficiency (QE), thick silicon with high aspect ratio full-depth deep trench isolation (FDTI) and backside scattering technology are implemented. Furthermore, thicker sidewall oxide for deep trench isolation and oxide filled FDTI were applied to enhance a modulation transfer function (MTF). In addition, 3-dimensional metal-insulator-metal capacitors were introduced to suppress temporal noise (TN). As a result, we have demonstrated industry-leading NIR G/S CIS with 2.71e- TN, dark current of 8.8e-/s, 42% QE and 58% MTF.

Wednesday, December 11, 2024 - 03:15 PM
41-4 | First Demonstration of 2.5D Out-of-Plane-Based Hybrid Stacked Super-Bionic Compound Eye CMOS Chip with Broadband (300-1600 nm) and Wide-Angle (170°) Photodetection
We propose a hybrid stacked CMOS bionic chip. The surface employs a fabrication process involving binary-pore anodic aluminum oxide (AAO) templates and integrates monolayer graphene (Gr) to mimic the compound eyes, thereby enhancing detection capabilities in the ultraviolet and visible ranges. Utilizing a 2.5D out-of-plane architecture, it achieves a wide-angle detection effect (170°) equivalent to curved surfaces while enhancing absorption in the 1550 nm communication band to nearly 100%. Additionally, through-silicon via (TSV) technology is integrated for wafer-level fabrication, and a CMOS 0.18-µm integrated readout circuit is developed, achieving the super-bionic compound eye chip based on hybrid stacked integration.

Wednesday, December 11, 2024 - 03:40 PM
41-5 | Pseudo-direct LiDAR by deep-learning-assisted high-speed multi-tap charge modulators
A virtually direct LiDAR system based on an indirect ToF image sensor and charge-domain temporal compressive sensing combined with deep learning is demonstrated. This scheme has high spatio-temporal sampling efficiency and offers advantages such as high pixel count, high photon-rate tolerance, immunity to multipath interference, constant power consumption regardless of incident photon rates, and motion artifact-free. The importance of increasing the number of taps of the charge modulator is suggested by simulation.

Wednesday, December 11, 2024 - 04:05 PM
41-6 | A Color Image Sensor Using 1.0-μm Organic Photoconductive Film Pixels Stacked on 4.0-μm Si Pixels for Near-Infrared Time-of-Flight Depth Sensing
We have developed an image sensor capable to simultaneously acquire high-resolution RGB images with good color reproduction and parallax-free ranging information by 1.0-μm organic photoconductive film RGB pixels stacked on 4.0-μm NIR silicon pixels for iToF depth sensing.

Wednesday, December 11, 2024 - 04:30 PM
41-7 | Pb-free Colloidal InAs Quantum Dot Image Sensor for Infrared
We developed an image sensor using colloidal InAs quantum dot (QD) for photoconversion. After spincoating the QDs on a wafer and standard semiconductor processing, the sensor exhibited infrared sensitivity and imaging capability. This approach facilitates easier production of lead-free infrared sensors for consumer use.

Wednesday, December 11, 2024 - 04:55 PM
41-8 | Lead-Free Quantum Dot Photodiodes for Next Generation Short Wave Infrared Optical Sensors
Colloidal quantum dot sensors are disruptingimaging beyond the spectral limits of silicon. In this paper,we present imagers based on InAs QDs as alternative for 1stgeneration Pb-based stacks. New synthesis method yields 9nm QDs optimized for 1400 nm and solution-phase ligandexchange results in uniform 1-step coating. Initial EQE is17.4% at 1390 nm on glass and 5.8% EQE on silicon(detectivity of 7.4 × 109 Jones). Using metal-oxide transportlayers and >300 hour air-stability enable compatibility withfab manufacturing. These results are a starting point towardsthe 2nd generation quantum dot SWIR imagers.

Also of interest, the following talk in Tuesday's session "Major Consumer Image Sensor Innovations Presented at IEDM"

Description: Authors: Albert Theuwissen, Harvest Imaging
Image Sensors past, and progress made over the years