LMONO and IDIPOL
Posted: Mon Jan 06, 2014 6:28 pm
Hello VASP
For calculations of charged defects in bulk semiconductors, which are the appropriate settings for LMONO and IDIPOL?
Currently, I adjust the number of electrons (NELECT) to reach the desired charge state (q) of the defect from the neutral configuration. I find the line with NELECT in the OUTCAR file and then add or subtract q for the charged defect total energy calculation. I set EPSILON to be equal to the experimental value.
I notice that when I include LMONO=.TRUE. the energy changes by about ~1 eV. Is this the first order Makov-Payne correction term? Is there any other effect (in the manual it seems to imply the LMONO flag concerns only the static potential shift due to the excess charge). I also notice that when LMONO=.TRUE. then IDIPOL has little (< 0.001 eV) or no effect on the total energy. Should IDIPOL be used? What is the correct procedure for computing charged defect total energies with VASP?
Thanks,
Michael Shaughnessy
For calculations of charged defects in bulk semiconductors, which are the appropriate settings for LMONO and IDIPOL?
Currently, I adjust the number of electrons (NELECT) to reach the desired charge state (q) of the defect from the neutral configuration. I find the line with NELECT in the OUTCAR file and then add or subtract q for the charged defect total energy calculation. I set EPSILON to be equal to the experimental value.
I notice that when I include LMONO=.TRUE. the energy changes by about ~1 eV. Is this the first order Makov-Payne correction term? Is there any other effect (in the manual it seems to imply the LMONO flag concerns only the static potential shift due to the excess charge). I also notice that when LMONO=.TRUE. then IDIPOL has little (< 0.001 eV) or no effect on the total energy. Should IDIPOL be used? What is the correct procedure for computing charged defect total energies with VASP?
Thanks,
Michael Shaughnessy