Semiconductor devices make up all of modern electronics, so, in addition to modeling performance, it is necessary to also understand and model the degradation processes that cause reliability issues. Defects, such as Ga vacancies in GaN-based HEMTs, exist in these devices, but they are hydrogenated in the fabrication process, making them neutral and without a level in the gap to capture carriers. Highly-energetic (hot) carriers and multiple hits from colder carriers can transfer energy to the defect and cause hydrogen to be released. Dehydrogenated defects are considered active as they are often charged and can scatter or capture carriers. In prior work, the activation probability has been examined through a defect-activation cross section that was assumed to be a step function with a minimum activation energy. No work had been done from first-principles to find this cross section, but there had been extensive work on carrier-capture cross sections. However, prior applications of the capture theory were limited by considering energy dissipation into a single phonon mode because there is an exploding number of possible phonon configurations as more modes are allowed to participate. First, we revamped the approach to the capture problem by simplifying and implementing a time-domain integration method that allows for quick calculations of all possible phonon configurations. We then developed an algorithm to map the multiple-carrier-scattering problem onto the capture problem to calculate the average activation lifetime of a hydrogenated defect. We present the results obtained for capture and for the hydrogen-release problem in a singly-hydrogenated Ga vacancy in GaN.
