Modeling transient response of PN junction diode under radiation beams considering nonlinear effects.

摘要

Diodes are used as dosimeter to measure the delivered dose from the radiation beams of linear accelerators used in the radiation therapy field. Diode sensitivity S, defined as charge collected per unit absorbed dose, has been found to increase with an increase in instantaneous dose rate r of the radiation beam. In order to calculate the dependence of S on r, one has to solve for the excess minority-carrier concentration Deltap (= Deltan) from the carrier transportation equation. This equation is a nonlinear partial different equation (PDE), with two variables of space coordinate x and time t. The nonlinear term in the PDE is the net recombination-generation (R-G) rate U. The historical solutions were either the transient solution from the linear PDE or the steady state solution from the nonlinear differential equation.; This dissertation solves the nonlinear PDE with simplified nonlinear term U. The analytical solution is the first to include both transient and nonlinearity at the same time.; The analytical solution for Deltap (x, t) from the PDE is used to derive the current density J p of the minority carriers and the diode sensitivity S. A computer software program has also been developed to numerically solve the non-simplified PDE. The agreement of the calculations for Deltap (x) and Jp (t) is excellent between the numerical and the analytical solutions, proving the analytical solution is legitimate. A theoretical model is established with the analytical solution to explain the r dependence of diode sensitivity S under pulsed radiation beam.; The sensitivity of several different types of diodes was measured with the 18 MV photon beam of a medical linear accelerator. The diode sensitivity from the experiment is compared with the calculations from both the analytical and the numerical solutions. The theoretical calculations yield the instantaneous dose rate dependence of the diode sensitivity. The agreement is generally good between the theory and the experiments. The results are discussed and possible error sources are analyzed.