Study of limitations on pixel size of very high resolution image sensors

Abstract

The employment of small pixel size is very crucial to the physical implementation of very high resolution CMOS APS image sensors. This is because of the restriction imposed by the required lithography method utilized in modern sub-micron CMOS fabrication processes on the size of the image sensor chip. Furthermore, the cost of a CMOS active pixel sensor (APS) image sensor chip substantially goes down as the pixel size goes down, whether the image sensor chip is very high resolution or not. However, there exist several optical limitations on the pixel size that make it imprudent to further reduce the pixel size beyond those limitations. A major limitation on the reduction of pixel size is that pixel sensitivity is significantly reduced as pixel size goes down. One may take refuge in using low f-number optics to allow more photons into the pixel to compensate for the reduced sensitivity of the small pixel. However, the employment of low f-number optics is associated with high cost, which may offset the low cost advantage of image sensor chips that have small pixels. Additionally, low f-number optics may introduce undesirable aberrations. Another major limitation on the reduction of pixel size is that cross-talk is considerably increased as the pixel size is decreased. Keeping cross-talk low is very critical to the proper operation of an image sensor, particularly color image sensors.