My Community

Dear Members and developers,

I am not able to reproduce results for total energy for ferromagnetic gama-Fe given in Marsman and Hafner's paper, PRB B 66, 224409 (2002). For a=3.55 A I am getting E= -8.13069 eV while in the paper it is about -8.16 eV. The parameters in INCAR file are as follows:

ENCUT = 500
PREC= High
GGA = PE
VOSKOWN = 1
ISMEAR = -5
ISPIN =2
MAGMOM =3

I am using 13X13X13 Monkhorst mesh and PAW PBE potential.

I would very much appreciate if somebody could help.

R. Prasad
IIT Kanpur

Which Fe potential do they use (do you use the same)? I.e. is the semi-core electrons included as valence electrons? It may give this order of difference in total energy for Fe.

Best regards,
/Dan Fors




<span class='smallblacktext'>[ Edited Thu Jul 10 2008, 07:16PM ]</span>

Thank you for your response. I have used the PAW PBE potential from the VASP site. I believe that Marsman and Hafner have used the same potential.

Regards,
R. Prasad

First to clarify my earlier response. There are two Fe potentials (both of them PAW-GGA-PBE). One of them have 8 valence electrons and the second also include the "3p" electrons so it has 14 valence electrons.

Second even if you set all settings (ENCUT, ENAUG, PREC, LREAL, NG(X,Y,Z), smearing, ISYM etc. ) the same way as in the paper, I guess that there is always the possibility that the changes in the code during the six last years might have slightly shifted the numerical evaluation of the total energy (to how large extend I don't know).

But you shouldn't be that concerned with that you obtain a slightly different value in the total energy compared with the published paper. The important thing is the energy differences, so as long as you get the same energy differences between the different phases and volumes, you should be satisfied.

Best regards,
/Dan Fors



<span class='smallblacktext'>[ Edited Sun Jul 13 2008, 12:28PM ]</span>

Thank you for your help. I tried the other potential which has 14 electrons. It gives total energy about -8.26 eV.

Regards,
R. Prasad