الفهرس | Only 14 pages are availabe for public view |
Abstract A solar cell is a photovoltaic device designed to convert the sun light into electrical power-. When the sunlight falls on the solar ceH, the incident energy is converted directly into electricity without any mechanical movement and delivers this power to a suitable load in an efficient manner. One possibility of making cheap solar cells arrays lies in the fabrication method. Since the diffusion process, which is used to create a p-n junction, is costly, there has been. a steady search for alternate and potentially lower cost methods of forming a photovoltaic barrier or junction. One of these techniques consists of inducing a conductivity-type change at the surface of the semiconductor by the application an ultra-thin metal or a relatively thick transparent conducting semiconductor. Such cells rely on a thin interfacial layer ({ 10-30 } A) between the top contact (metal) and the base semiconductor. This interfacial layer is generally an oxide or some other compound, which is normally an insulator in its bulk form. Hence these cells are referred to as metal-insuhitor-semiconductor (MIS) solar cells. This thesis is devoted to study the theoretical analysis ofMISnL solar cells, Also the thesis presents a developed model to study the performance of a Metal-lnsulator¬ Semiconductor with induced inversion layer solar cell (MIS/IL) as the Al/tunnel-oxide/p-Si structure. The model included the effect of change in cell parameters namely: doping concentration, oxide thickness, mobile charge density and metal work function. It also included the dependence on the mobile charge density and fixed oxide charge density. A back bias applied between substrate and metal inversion grid was added to the model. The genetic algorithm used to optimize the MIS/IL efficiency by changing all these parameters to extract its values to give the optimum efficiency. The optimization results compared with respect to theoretical optimization and experimental optimization, the comparison shows good agreement for the genetic algorithm optimization. |