Optimization of Back-Side Illuminated CMOS Image Sensors

Abstract

We optimized back-side illuminated (BSI) CMOS image sensors (CIS) to improve optical efficiency. We used three-dimensional (3D) finite-difference time-domain (FDTD) simulation to optimize the radius of the curvature of the micro-lens and the thickness of the over-coat layer. We also used the Poynting theorem to calculate the power absorption of the photo-diode, and the transfer matrix to design the optimum structure of anti-reflection coating. Using the theory of frequency selective surface (FSS), we designed IR pass filter and IR cutoff filter possible to integrate into 3D BSI pixels. In this paper, we represented optimization methods of the BSI pixel structure and design methods of metal IR pass filter and IR cutoff filter. These results can be used to enhance optical properties of CISs and reduce the cost of the CIS module.