Dear Marcos,
Thank you very much for your answer. I was
working with (n,n) tubes. I have performed a single cell calculation with
the 1x1x75 kgrid_monkhorst_Pack and i got the correct results. Indeed, I was
seeing the effect of the folding of the 1-d BZ.
regards,
pablo
----- Original Message -----
From: "Marcos Verissimo Alves" <[EMAIL PROTECTED]>
To: <SIESTA-L@listserv.uam.es>
Sent: Tuesday, January 16, 2007 2:42 PM
Subject: Re: [SIESTA-L] DOS SWCNT
Hola Pablo,
1- for metallic tubes at least 3 unit cells are needed to obtain a
metallic
DOS
Does that mean that with a single unit cell you cannot obtain a metallic
DOS? For what kind of nanotubes (zigzag (3*m,0), m integer, or armchair)?
This result is weird. Have you relaxed the structure completely, including
the nanotube axis? (A good value for the force tolerance would be 0.04
eV/Ang or smaller.) What is the minimum inter-image distance in the x-y
plane?
From this statement I might guess (quite wildly) that you haven't gone to
sufficiently high numbers of k-points for the calculation of the DOS. Do
an "experiment": try using
%block kgrid_Monkhorst_Pack
1 0 0 0.0
0 1 0 0.0
0 0 75 0.0
%endblock kgrid_Monkhorst_Pack
in a calculation with a single (totally relaxed) unit cell and check it
against the result of the three unit cells, which I suppose you did using
the MP grid from item 2. If they match, what you are seeing is the effect
of the folding of the 1-d BZ, I would guess.
2- %block kgrid_Monkhorst_Pack
1 0 0 0.0
0 1 0 0.0
0 0 25 0.0
%endblock kgrid_Monkhorst_Pack
This has to be checked by the total energy convergence with the number of
points, although it is not far from the number of points I had to give for
good VC relaxation of a graphene unit cell in one of the directions in the
sheet plane. Good for structural relaxation, but not DOS calculation (see
a previous e-mail by Andrei Postnikov on the subject), which requires a
substantially higher amount of k-points.
Also, from the block above I suppose that your nanotube axis is aligned
with the z-axis. Correct?
3- PAO.BasisSize DZP
Ok, good basis, but be carfeul with the Energy shift.
4- MeshCutoff 400. Ry
Might be a bit too much, unless you are going to a very strict energy
convergence. My experience is that 300 Ry does the job fine - even 250
wouldn't be too bad.
5- PBE perfoms better than LDA but, the results with the later are
reasonable.
Makes sense. Again, depends on what you want. Check against the infinite
number of calculations in the literature both with siesta and PW basis
sets.
6- the r for the TM pseudo was 1.5
If it gives you good agreement of pseudo-eigenvalues and logarithmic
derivatives against the all-electron results, then don't worry, be
happy... :)
Marcos
--
Dr. Marcos Verissimo Alves
Post-Doctoral Fellow
Condensed Matter and Statistical Physics Sector
International Centre for Theoretical Physics
Trieste, Italy
--------
I have become so addicted to vi that I try to exit OpenOffice by typing
:wq!
--
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.1.410 / Virus Database: 268.16.12/631 - Release Date:
16/01/2007
