ABINIT Discussion Forums

Dear all,

I set usepotzero=1 with very new JHT-v0.2 PAW-PBE atomic datasets.
However, the calculated eigenvalues are too high to be printed in output files, and shown as follows.

e.g. (BaO: JHT ver0.2)
Fermi (or HOMO) energy (eV) =********** Average Vxc (eV)= -10.30742
Eigenvalues ( eV ) for nkpt= 256 k points:
kpt# 1, nband= 20, wtk= 0.00013, kpt= 0.0000 0.0000 0.0000 (reduced coord)
********************************************************************************
********************************************************************************
****************************************

I found the same erroneous results in the CaO calculation, too.

On the other hand, when using previous Jollet and Torrent PAW (PBE) potentials, usepotzero=1 looks fine.

e.g. (BaO: Jollet and Torrent, paw_setup version="0.5")
Fermi (or HOMO) energy (eV) = 11.15187 Average Vxc (eV)= -9.73628
Eigenvalues ( eV ) for nkpt= 256 k points:
kpt# 1, nband= 20, wtk= 0.00013, kpt= 0.0000 0.0000 0.0000 (reduced coord)
-14.68665 -3.83095 0.78599 0.78599 0.78599 10.72484 10.72484 10.72484
14.70197 15.90527 15.90527 15.90527 16.34543 18.13936 18.13936 19.55865
26.06028 26.06028 26.06028 28.60952

Input file for BaO is as follows:
------------------------------
#Cell and atoms definition
acell 3*1 angstrom
rprim
0.00000 2.77898 2.77898
2.77898 0.00000 2.77898
2.77898 2.77898 0.00000
ntypat 2 natom 2 typat 1 2
xred 0.0 0.0 0.0
1/2 1/2 1/2
znucl 56 8 nband 20

#Grid definitions
ecut 20.
pawecutdg 40.

#SCF cycle parameters
tolvrs 1.0d-10
nstep 20

#K-points and sym
# nsym 0
occopt 1
ngkpt 20 20 20
nshiftk 1
shiftk 0.0 0.0 0.0

#I/O parameters
optforces 2 optstress 1
prtwf 0 prtden 0 prteig 0
prtvol 2
enunit 1

usepotzero 1
------------------------------

These potentials are obtained from
http://www.abinit.org/downloads/PAW2/OLD-VERSIONS/JTH-TABLE-0.1/index.html
http://www.abinit.org/downloads/PAW2

Is this a bug of the usepotzero tag?
Does someone have a comment/suggestion?

Thank you very much for your help.
Best regards,
Yu Kumagai

Dear Yu Kumagai,

Thanks for the bug report.
The usepotzero was not concerned.
This was a small bug occuring for some specific psps.
It has now been corrected and the correction will be available in the next 7.10 version of abinit (to be released in some weeks).

Marc

Dear Dr. Marc Torrent,

Thank you very much for your info.
You mean the problem occurs in the abinit or PAW potential set?
I tried abinit-ver7.10.1, but it doesn't resolve the problem...

Best,
Yu

This was too late for the 7.10.1.
Will be in 7.10.2