Abstract
Abstract
Radioluminescent silica-based fiber dosimeters offer great advantages for designing miniaturized realtime sensors for high dose-rate dosimetry. Rise and fall kinetics of their response must be properly understood to better assess their performances in terms of measurement speed and repeatability. A standard model of radioluminescence (RL) has already been quantitatively validated for doped silica glasses, but beyond conclusive comparisons with specific experiments, a comprehensive understanding of the processes and parameters determining transient and equilibrium kinetics of RL is still lacking. We analyze in detail the kinetics inherent in the standard RL model. Several asymptotical regimes in the RL growth are demonstrated in the case of a pristine sample (succesive quadratic, linear and power-law time dependencies before the plateau is reached). We show how this situation is modified when a pre-irradiation partly fills traps beforehand. RL growth is then greatly accelerated because of the pre-formation of recombination centers (RCs) from dopant ions, but not due to pre-filling of trapping levels. In all cases, the RL intensity eventually tends to a constant level equal to the pair generation rate, long before all carrier densities themselves reach equilibrium. This occurs late under irradiation, when deep traps get to saturation. The fraction of dopants converted into RCs is then ‘frozen’ at a lower level the smaller the density of deep traps. Controlling RL kinetics through the engineering of material traps is not an option. Pre-irradiation appears to be the simplest way to obtain accelerated and repeatable kinetics.
Funder
Agence Nationale de la Recherche