ABINIT Discussion Forums

Dear ABINIT users and developers,

I am using ABINIT 6.12.3 and considering that the release notes state that "Phonon calculations using DFPT can be done reliably with the PAW methodology, for the insulating case" I tried to do it for the L and Gamma points of InP in the zincblend phase (checking purposes). However, I found the following error message in the log file:

Subroutine Unknown:0:BUG
nkxc must be 1 or 3 !
Action : contact ABINIT group.

I would like to know if this is a limitation of ABINIT (phonon calculations with PAW at specific k points) or I am doing something wrong.
Please, find my input file below.
Thanks in advance,
Luis Ogando

PS: I found nothing related to this topic in the forum
PS2: the execution was parallelized only for the k points (mpirun --n option)

===========================================================================================================================

ndtset 2

#Ground state calculation

tolwfr1 1.0d-22

#Response-function calculation, with q=(1/2,1/2,1/2)

pawxcdev2 0
rfphon2 1
rfatpol2 1 2
rfdir2 1 1 1
nqpt2 1
qpt2 0.50000000 0.50000000 0.50000000
getwfk2 -1

tolvrs2 1.0d-22
#######################################################################
#Common input variables

#Definition of the unit cell
acell 11.235840 11.235840 11.235840

rprim 0 1/2 1/2
1/2 0 1/2
1/2 1/2 0

#Definition of the atom types
natom 2
ntypat 2
typat 1 2

znucl 49 15

xred
0.00000000 0.00000000 0.00000000 # Triplet giving the REDUCED coordinate of atom 1 (In).
0.25000000 0.25000000 0.25000000 # Triplet giving the REDUCED coordinate of atom 2 (P).

#Definition of the planewave basis set

ecut 25.0000

pawecutdg 62.0000

#Definition of the k-point grid

kptopt 3

kptrlatt -6 2 2
2 2 2
2 2 -6

shiftk 0.50000000 0.50000000 0.50000000

kptrlen 3.8922E+01

#Definition of the SCF procedure

nstep 600

diemac 12.0

Dear Luis,

This message is not really explicit... but this means that you are trying to use GGA which is not yet ready for DFPT+PAW.
Try LDA.
GGA to be released soon...

Marc

Dear Prof. Marc Torrent,

Thank you very much for your help !
I will try it and post the result.
All the best,
Luis

Dear Marc,
Is it possible to do phonon dispersion calculations with DFPT in the regime of DFT+U in the current version of ABINIT?

Thanks,
Yuanjun

torrent wrote:Dear Luis,

This message is not really explicit... but this means that you are trying to use GGA which is not yet ready for DFPT+PAW.
Try LDA.
GGA to be released soon...

Marc

Dear Prof. Marc Torrent,

I am sorry for the huge delay !!
I would like to report the result of my phonon calculation using PAW with LDA.
Well, the calculation goes until the end, but the calculated phonon frequencies at Gamma are completely wrong ! Just take a look:

Phonon frequencies in cm-1 :
- -4.340096E+03 -2.379470E+03 2.094202E+02 2.409050E+03 2.804881E+03
- 7.378877E+03

Please, I would thank you if you could take a glance at the input file I used to obtain these horrible results (below) and say me if there is something obviously wrong. In a negative case, I would like to ask you if this kind of problem can be attributed to the PAW functions I used.

Thanks in advance,
Luis

==========================================================================================================================

ndtset 2

#Ground state calculation
kptopt1 2

tolwfr1 1.0d-22

#Response-function calculation, with q=0

pawxcdev2 0

rfphon2 1

rfatpol2 1 2


rfdir2 1 1 1

nqpt2 1

qpt2 0.00000000 0.00000000 0.00000000

getwfk2 -1

kptopt2 2

tolvrs2 1.0d-22

#######################################################################
#Common input variables

#Definition of the unit cell
acell 11.090177 11.090177 11.090177

rprim 0 1/2 1/2
1/2 0 1/2
1/2 1/2 0

#Definition of the atom types
natom 2

ntypat 2

typat 1 2

znucl 49 15

xred
0.00000000 0.00000000 0.00000000
0.25000000 0.25000000 0.25000000

#Definition of the planewave basis set
ecut 25.0000

pawecutdg 62.0000

#Definition of the k-point grid.
ngkpt 8 8 8

#Definition of the SCF procedure
nstep 600

diemac 12.0

Hi, I just ran the Gamma and L phonons of InP with PAW, with two versions of the code: 6.12.3, where I get similarly problematic results as you, and 7.0.1 (I'm one of the developers so have access to the current development branch) and I get excellent results. Clearly Marc Torrent and his team have made improvements to the PAW DFPT since version 6.12.3 came out; I hope soon that a 7.X version will be publically available, that should help you a lot.

Dear Prof. Zwanziger,

Thank you very much for the good news !! I am waiting for ABINIT 7.0.1 with the best expectations !!
All the best,
Luis

jzwanzig wrote:Hi, I just ran the Gamma and L phonons of InP with PAW, with two versions of the code: 6.12.3, where I get similarly problematic results as you, and 7.0.1 (I'm one of the developers so have access to the current development branch) and I get excellent results. Clearly Marc Torrent and his team have made improvements to the PAW DFPT since version 6.12.3 came out; I hope soon that a 7.X version will be publically available, that should help you a lot.