confused ypp interpolated bandstructures

[arlonne]

band_s47.jpg

Hi all,

I just finished G0W0 calculation for a organic crystal. the ypp ploted bandstructure is very strange.
firstly, the old version ypp (3.3.0) plot very correct VB,VB-1,CB and CB+1, but band cross occurs at other bands, see the Figure.
secondly, I also use the newest ypp (3.4.0), band cross are disappeared, however, the energy difference between VB and VB-1 and some other VBs is larger than that from the output of G0W0 (o-gwppa.qp). See in the Figure.

Finally, I checked the o-gwppa.qp. the result explain the band cross in old ypp. the G0W0 eigenvalues is not linearly decrease or increase. Therefore, why it happens?

P.S. I have tested the convergence of G0W0(PPA) calculations.
-----------------------
ypp.in:

bnds # [R] Bands
electrons # [R] Electrons (and holes)
cooIn= "rlu" # Points coordinates (in) cc/rlu/iku/alat
% BKpts
0.50 | 0.00 | 0.00 | # Bands circuit
0.00 | 0.00 | 0.00 | # Bands circuit
0.00 | 0.50 | 0.00 | # Bands circuit
0.00 | 0.50 | 0.50 | # Bands circuit
0.00 | 0.00 | 0.50 | # Bands circuit
0.00 | 0.00 | 0.00 | # Bands circuit
%
BANDS_steps=11 # Number of divisions
% BANDS_range
247 | 307 | # Bands Range
%
GfnQPdb= "E < ./ndb_LF100.QP" # [EXTQP G] Database
---------------

[Daniele Varsano]

Dear Arlonne,
thanks for reporting the problem,
We are aware that could a problem in the input QP database in the devel version.
We will correct it in the next days.

Best,

Daniele

[andrea marini]

I just finished G0W0 calculation for a organic crystal. the ypp ploted bandstructure is very strange.
firstly, the old version ypp (3.3.0) plot very correct VB,VB-1,CB and CB+1, but band cross occurs at other bands, see the Figure.
secondly, I also use the newest ypp (3.4.0), band cross are disappeared, however, the energy difference between VB and VB-1 and some other VBs is larger than that from the output of G0W0 (o-gwppa.qp). See in the Figure.

Finally, I checked the o-gwppa.qp. the result explain the band cross in old ypp. the G0W0 eigenvalues is not linearly decrease or increase. Therefore, why it happens?

Dear Arlonne, in the latest revision to the devel GPL (revision 48) there was indeed a bug. So the loading of the QP database in version 3.4.0 was wrong.

This is now fixed. However, as far as I understand, there is a problem with some bands crossing that seems to be related to the interpolation procedure. We will work further on this issue.

Can you, in the meantime, confirm the problem in the newest ypp 3.4.0 (revision 48) ?

Thanks

Andrea

[ljzhou86]

Yes, this problem has been resolved after updating my source according this post viewtopic.php?f=9&t=497.