Conventional CMOS image sensors widely used in products currently on the market are mainly equipped with a rolling exposure function. This rolling exposure causes so-called "Jell-o effect" distortion when capturing a moving target. CMOS image sensors with a global-shutter function are one of the solutions to avoid this distortion. An in-pixel storage node is required to create a global-shutter CMOS image sensor. A floating diffusion and an additional capacitor can be used as an in-pixel storage node [1,2]. The light sensitivity of the in-pixel storage node is specified by the parasitic light sensitivity (PLS), which is the ratio of the light sensitivity of an in-pixel storage node and the light sensitivity of a photodiode. The PLS should be small enough so that the in-pixel storage is not lightsensitive. Artifacts are captured in an image from bright moving objects during read-out if the PLS is not small enough. The PLS of reported global-shutter CMOS image sensors is around -100dB. That would be small enough to use those image sensors in fields where the light source can be controlled. However, for DSC usage, users can easily encounter scenes with bright objects (e.g. sunlight or car headlights). Even if the in-pixel storage node is light-shielded, it is difficult to perfectly protect the in-pixel storage node from photo-generated carriers, as long as the in-pixel storage node and a photodiode are on the same silicon substrate. Meanwhile, 3D stacking technologies have been introduced for image sensors to give them more functionality and improved performance [3,4]. The reported minimum interconnection pitches for image sensors are over 20μm. These technologies do not fit the smaller pixel pitches of the image sensors in recent DSCs. In this paper, we report a rolling-shutter distortion-free 3D stacked image sensor with an in-pixel storage node of -160dB parasitic light sensitivity. The image sensor virtually achieves a global-shutter function using a 4times frame-shutter operation. The image sensor has 2 semiconductor substrates, where 1 substrate has a backside-illuminated photodiode array and the other a storage-node array. The image sensor achieves a PLS level of -160dB. The image sensor has 8.6μm pitched interconnections, and an interconnection yield of over 99.9% is achieved.

A microbial consortium capable of degrading turbine oil (TuO) consisting mainly of recalcitrant branched and polycyclic alkanes, has been established using a sludge sample collected from an oil-water separation tank of a hydraulic plant, by repeated enrichment cultivation. When this consortium, named the tank-2 consortium, was cultivated in a minimal salts medium containing 0.5% TuO, it degraded approximately 90% of TuO at 30 °C under shaking at 160 rpm in five days. Dur ing cultivation, the consor tium for med cell aggregates. Four teen bacter ial species were isolated from the consortium. Although the mixtures of all of the isolated str ains (IM) and predominant str ains selected from the IM (PM) degr aded approximately 60% of the TuO maximally in seven days after repeated enr ichment, the degradation r ate was not maintained. The decreased r ate of TuO degradation in the PM was recovered by inoculation with a mixture (AM) consisting of isolates involved in the cell aggregate for mation in the 3 tank-2 consor tium, suggesting that the for mation of bacter ial aggregates is an important factor for the degradation of TuO by the bacter ial consortia. Interestingly, the for mation of bacter ial aggregates was only obser ved when cyclic alkanes were added to the consortium as a source of car bon. The cell sur face hydrophobicity of the bacter ial aggregates was significantly high (77.6% ). These results suggest that the growth of bacter ia with high cell hydrophobicity can be induced by the cyclic alkanes fr action of TuO, and that the for mation of aggregates facilitates cells to for m contact with the recalcitr ant components of TuO and uptakes these components.

No abstract