| CMOS image sensors have completely occupied the camera field with their high integration,low cost,and low power consumption.However,the two application areas with the largest market share are mobile consumer electronics and automotive image sensors.Their industry-leading technology is currently firmly in the hands of foreign large manufacturers.Among them,the imaging quality of HDR image sensors and the image overflow problem of small-sized pixels have become two major obstacles restricting the development of sensor technology.This thesis focuses on the two key factors that affect imaging quality in the above-mentioned technical difficulties: charge transfer efficiency,image overflow,and optimization of pixel structure in combination with existing processes.The main contents are as follows:1.Based on the active pixel structure of CMOS image sensors,the causes of nonlinear charge transfer phenomena from different sources are analyzed,and engineering solutions are proposed by combining simulation tools and actual data.2.A novel dual gate structure is proposed to address the charge backflow problem that affects the imaging quality of high dynamic range image sensors operating in highCG mode.After simulation,it has been verified that it can close the electron return channel from the transmission gate to the photodiode through advanced negative bias of the isolation gate,inhibit the return flow of electrons enriched below the gate to the photodiode,thereby allowing more electrons to be collected by the floating diffusion node,suppress the charge return flow,and improve the charge transfer efficiency.3.The image overflow of pixels is studied,and through the current channel analysis of pixels with overflowing electrons in simulation,it is found that the pixel state of image overflow is determined by the overflow current and the leakage current at the floating diffusion node,and the ratio of overflow current and leakage current basically determines the overflow ratio of pixels.Combined with the existing process to optimize and simulate the pixel structure for electron spillage and leakage,finally the method of shallow trench isolation on the frontside and deep trench isolation on the backside and N-type implant to control the electron leakage channel is selected to optimize the pixel overflow,and the tape-out is carried out on this structure,and the test results show that the image overflow between pixels can be better suppressed. |