NUMERICAL SIMULATION OF ONE DIMENSIONAL SILICON PN-JUNCTION BETAVOLTAIC CELL WITH MONTE CARLO METHOD

• Main Author: Jaya, Christopher
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/65113
• Institution: Institut Teknologi Bandung

The betavoltaic cell is a device that converts radioactive decay energy to electrical energy directly. This device has a small dimension with a long lifetime, henceforth it’s suitable to power pacemakers or other microelectromechanical systems/MEMS. A betavoltaic cell is consist of a beta emitter source and semiconductor junction. The beta particle that had been emitted by its source interacts with the semiconductor junction and creates several electronhole pairs or EHPs that diffuse to the depletion region and increase junction potential. In this final assignment, simulation of a one-dimensional planar betavoltaic cell has been done using several analytical models from various literature. The type of semiconductor of the cell model in this assignment is Si pn-junction. The simulation was done numerically using Python with the Monte Carlo method, which is a simulation principle that generates several amounts of random numbers. The output of the cell was done by generating 0.25mCi and 1mCi equivalent of beta particles‘ energy. Variables variations besides source activity are stopping power interaction model, surface recombination velocity, dopant concentration, cell temperature, and battery dimension. Output parameters from the simulation are compared with references. The betavoltaic cell model in this study has consisted of a Ni-63 beta emitter, 0.7?m type-P Si, and 1?m type-N Si. For the main calculation, the type-P dopant concentration is 1026/m3, and type-N dopant concentration is 1022/m3. The main assumption is that the surfaces of both ends of the cell are completely passivized, therefore the surface recombination velocity on both sides are 0m/s. From this study, beta emitter activity has a positive correlation with the cell’s power output and its efficiency, with the power output of the cell with 1mCi source is between 0.568nW-0.915nW and its efficiency is between 0.557%-0.898%, and for the cell with 0.25mCi source its power output is between 0.097nW-0.161nW and its efficiency is 0.382%-0.634. Those values are varied depending on the interaction model in the calculation. From the variation of the surface recombination velocity, it was shown that the conversion efficiency of the cell has a negative correlation with its surface recombination velocity, hence the highest efficiency is achieved when the cell surfaces are completely passivized. Furthermore, from the variation of the dopant concentration, it was shown that there’s an optimal dopant concentration for the P-type and N-type that give the highest cell efficiency. In this study, the optimal values for the dopant concentration are ???????? = ???????? = 1.468 × 1025/m3. However, these values vary highly depending on the diffusion coefficient and length in the semiconductor junction, cell’s crystal uniformity, and cell’s design. With the mathematical models that this study used, it was found that efficiency has a negative correlation with cell temperature; the higher the temperature, the lower its efficiency and its power output becomes. Also, from the variation of P-type and N-type thickness, while maintaining the overall thickness to be constant, there’s an optimal thickness for P-type and N-type of the cell that gives the highest efficiency and power output which, in this study, are 0.294?m for the P-type and 1.506?m for the N-type.