ABINIT Discussion Forums

Dear all,
I've been trying the tree following tests in abinit-7.10.2 version : As results of AlAs tests we have:

1- the input "trf1-5.in" (Response-function of an homogeneous electric field) >>>>>>>>>>>>>>>>>>>>>> the result of dielectric constant is 9.7606048428


2- the input "tnlo-2.in"(BERRY PHASE (field=0 )Compute various quantities from Delta polarization) : >>>Linear optical dielectric constant : >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>Electronic dielectric tensor = 9.20199931 (9.94846084) >>>Static dielectric constant : >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>Relaxed ion dielectric tensor = 10.95641891 (11.84823634)


3- the input ''~abinit/tests/tutorespfn/Input/tffield_6.in'' (Finite electric field calculation of alas at clamped atomic positions) >>>>>>>>>>>>>>>>>>>>>>>>>><>>>>>>>>>>>>\epsilon_11 = 1+ 4 \pi \chi(1)_11= 8.13

please can i know the difference between the dielectric constants generated , and which is the appropriate methode to my calculations given that i aim to find the static optical dielectric constant of BaTiO3 in spontaneous state

best regards

Most of the dielectric constants are comparable, but will only be similar if you converge everything very scrupulously wrt k-points ecut and maybe bands.

The static dielectric constant adds the phonon contribution to the "infinity" electron-only dielectric constant. This is explained in Kittel or Ashcroft and Mermin, or in Xavier Gonze's double PRB on the method.