[QE-users] Thermal trap ionizations energy evaluation

[Тимофей Перевалов]

Dear QE users and developers,
I need your help. I am evaluating the thermal ionizations energy of O vacancy in HfO2 as it described the article [PRB 75, 205336, 2007]. One of the calculation formulas (in adapted form) used: (E(perf, q=-1) - E(perf, q=0))  –  (E(def, q= -1) -  E(def, q= 0)).  I know about the compensation background charge presence and I hope that its contribution is cancelled since in formulas, there is a double subtraction. How true is this? I am exploiting B3LYP functional. What should be considered in calculations of this kind?
The resulting values depend on the supercell size. How do I get rid of this? Obviously due to the background charge contribution. The calculations with the Makov-Payne correction (for c-HfO2) give completely different values and also depend on the on the supercell size. 

Sincerely,
Timofey Perevalov
ISP SBRAS, Russia