General Model for Defect Dynamics in Ionizing-Irradiated SiO2-Si Structures

Irradiation damage is a key issue for the reliability of semiconductor devices under extreme environments. For decades, the ionizing-irradiation-induced damage in transistors with silica-silicon (SiO2-Si) structures at room temperature has been modeled by a uniform generation of E-gamma ' centers in the bulk silica region through the capture of irradiation-induced holes, and an irreversible conversion from E-gamma ' to P-b centers at the SiO2/Si interface through reactions with hydrogen molecules (H-2). However, the traditional model fails to explain experimentally-observed dose dependence of the defect concentrations, especially at low dose rate. Here, it is proposed that the generation of E-gamma ' centers is decelerated because the holes migrate dispersively in disordered silica and the diffusion coefficient decays as the irradiation goes on. It is also proposed that the conversion between E-gamma ' and P-b centers is reversible because the huge activation energy of the reverse reaction can be reduced by a "phonon-kick" effect of the vibrational energy of H-2 and P-b centers transferred from nearby nonradiative recombination centers. Experimental studies are carried out to demonstrate that the derived analytic model based on these two new concepts can consistently explain the fundamental but puzzling dose dependence of the defect concentrations for an extremely wide dose rate range.