ABINIT Discussion Forums

Dear colleagues,

is there any more profound way to modify the .freq file, which one obtains with anaddb when trying to plot phonon dispersions (PD) in the case when one has different non-analyticities from different direction approaching the Gamma point?

We are trying to reproduce the PD of rutile TiO2. We have calculated non-analyticities using two directions:

nph2l 2 ! number of directions in list 2
qph2l 1.0 0.0 0.0 0.0
0.0 1.0 1.0 0.0


and obtained the corrections:

Phonon at Gamma, with non-analyticity in the
direction (cartesian coordinates) 1.00000 0.00000 0.00000
Phonon energies in Hartree :
-2.950590E-07 -2.718936E-07 -1.570178E-07 5.329891E-04 5.739427E-04
8.163704E-04 9.414841E-04 1.617812E-03 1.837808E-03 1.914023E-03
1.928310E-03 2.042815E-03 2.194253E-03 2.194396E-03 2.290724E-03
2.897259E-03 3.755997E-03 3.844283E-03

Phonon at Gamma, with non-analyticity in the
direction (cartesian coordinates) 0.00000 0.70711 0.70711
Phonon energies in Hartree :
-3.225039E-07 -2.565986E-07 -1.505021E-07 5.329890E-04 5.739427E-04
8.165679E-04 8.616599E-04 1.773011E-03 1.837807E-03 1.914023E-03
1.928310E-03 2.142726E-03 2.194253E-03 2.194396E-03 2.290654E-03
2.897259E-03 3.655566E-03 3.844283E-03


But when we try to plot them with band2eps, we always obtain unacceptable figure. After several attempts, we figured that if we put another point just before (or after) and very very close to the Gamma point in the anaddbs' 1st q point list, and replace this point (in the .freq file) with the correction for one direction and replace the Gamma point (in the .freq file) with the correction for the other direction, the plot would be ok. But it still isn't.


Could someone help us with this small issue?

Thank You in advance!

Igor Lukacevic

Dear all,

after many tries, I found no solution via band2eps.

But one can do it by hand using xmgrace and PHFRQ file instead of .freq file. One just has to put in additional Gamma point for every occurence of the Gamma point in the PHFRQ file and apply the the appropriate non-analyticity correction.

Cheers!

Igor Lukacevic

ilukacevic wrote:But one can do it by hand using xmgrace and PHFRQ file instead of .freq file. One just has to put in additional Gamma point for every occurence of the Gamma point in the PHFRQ file and apply the the appropriate non-analyticity correction.

Well, actually, it's not good. I gave it a better view now, and compared with experiments. Modes are not reproduced well, if one follows this way.

I still haven't found the way how to modify the .freq file in order to have the band2eps to plot the PD correctly. I think that it is not possible to do it by adding a point, since band2eps takes q-points one by one (adds one point to the x-axis for each q-point) and PD curves will, thus, always be displaced.

If someone has come up with an idea, I'm open for any suggestion.

Igor Lukacevic

ilukacevic wrote:Dear all,

after many tries, I found no solution via band2eps.

But one can do it by hand using xmgrace and PHFRQ file instead of .freq file. One just has to put in additional Gamma point for every occurence of the Gamma point in the PHFRQ file and apply the the appropriate non-analyticity correction.

Cheers!

Igor Lukacevic

When I checked more carefully, this method is good. The problem was that I "accidentally" missed the last mode, so it didn't look good to me. One just has to be careful and apply the right correction corresponding to the direction to and from the Gamma point.

Igor L.

At Igol L,
please what command did u issue to xmgrace to plot phonon band structure from PHFRQ. Thanks

Hi ilukacevic,

Please could you tell me what command you used to plot .PHFRQ in xmgrace. How did you set gamma, X , Y etc. points along x-axis. band2eps is not working for me.

Thanks,

Suvas

BAMGBOSE, M. K. wrote:At Igol L,
please what command did u issue to xmgrace to plot phonon band structure from PHFRQ. Thanks

None. I opened it from within xmgrace.

suvas wrote:Hi ilukacevic,

Please could you tell me what command you used to plot .PHFRQ in xmgrace. How did you set gamma, X , Y etc. points along x-axis. band2eps is not working for me.

Thanks,

Suvas

Answer to the first question is given above. For the second...just edit the x-axis in xmgrace and enter your own axes tick labels.