Challenges in CMOS‐based images

Abstract

High resolution CMOS image sensor demand is driven by consumer applications like mobile phone or digital still camera. To maintain the image quality, we have to save the signal to noise ratio (SNR) despite the number of photons reduction gathered by the pixel. In this paper, we present how back‐side illumination technology could help to recover signal and how to optimize the process and device in order to reduce the noise. A quantum efficiency larger than 70% is targeted to justify efforts done on BSI process development. We also have to pay attention to pixel to pixel crosstalk coming from free carrier diffusion. To overcome this parasitic phenomenon, deep trench isolation (DTI) should be considered in order to significantly reduce crosstalk between two neighboring pixels. The second main noise contributor related to BSI process is the dark current signal. Dark current less than 1 aA per pixel at room temperature is mandatory to save the image quality at low light level. Therefore, all the Si‐SiO2 interfaces at the pixel side walls have to be processed without any process damage to reduce as much as possible the surface defects. On top of that, side wall implantation and side wall dielectric fabrication have been optimized to reduce this noise source. The pixel layout has to be fully optimized also to collect and store charges properly. With respect to the 3D pixel shape, a new fully depleted pinned photodiode with vertical charge storage and no lag has been developed. (© 2014 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Curing Process on Passivation Layer for Backside-Illuminated CMOS Image Sensor Application

“…Thus, a backside-illuminated (BSI) structure was introduced to address the critical issues encountered in scaled pixels [7]. However, thinning the substrate in the BSI structure can cause many defects on its surface, which increases its fatal dark-current [7], [8], [9]. In these systems, the passivation layer is in direct contact with the rough surfaces of the substrate [7].…”

Low-Temperature Solution-Processed All Organic Integration for Large-Area and Flexible High-Resolution Imaging

“…However, many application scenarios are not able to be addressed by the CIS technology, especially when large-area coverage and free form factor (e.g. being flexible or conformable) is required [11][12][13]. The typical examples include x-ray imagers [14][15][16][17], scanners [18], in-display full or half screen fingerprint recognition [19], interactive flat-panel displays and emerging bio-inspired visual systems [20][21][22][23].…”

Bias-selected full Red/Green/Blue color sensing and imaging based on inversely stacked radial PINIP junctions

“…Blue/Green/Red (RGB) color sensing is prevalent in a wide range of applications, including imaging, charge-coupled device (CCD) camera, environmental surveillance and biological sensing [1][2][3][4][5][6], which are always in pursuit of more compact and more efficient color sensing configuration to achieve higher imaging resolution, faster detection and more flexibility, catering to the needs of wearable and soft electronics. Conventional RGB color sensors in CCD camera use Bayer mosaic matrix array with 4 photodetectors and 4 filters (2 for green, 1 for red and 1 for blue) for a single color unit, as depicted schematically in figure 1(a), which is a fundamental limit to shrink the unit size further, reduce the cost and adapt to flexible or curved surfaces [7][8][9][10][11]. To break this limit, also to avoid color aliasing in mosaic imaging [12][13][14], filter-less stacked PIN/PIN junction sensors were proposed, where the different incident wavelengths get absorbed at different penetration depths in stacked PIN junctions (see figure 1(b)) [15][16][17][18].…”