Dear Prem Sen,
Yes, you are right, both codes (turbo_lanczos.x and turbo_davidson.x) currently
cannot be used for metallic systems (using smearing). If a smearing is needed,
then it must be implemented (playing with the parameters mixing_mode,
mixing_beta, etc. will not help without the smearing, I think).
There is also another code turbo_eels.x which can be used with the smearing.
But this code was created to compute electron energy loss spectra (EELS) for
solids. You may try to use it for your system in a supercell with k=0 (do not
use K_POINTS{gamma}, because it is not supported). The code will produce a file
with the dielectric function and inverse dielectric function for a specified
transferred momentum q. You may try to set q very small (but not exactly zero),
perform the calculation, and then plot the imaginary part of the dielectric
function, and see if this it what you need. Be aware that I did not test the
turbo_eels.x code for finite systems, so there is no warranty that the results
will be correct. Please report what you get, if you decide to try it out. And
use the latest version of QE (i.e. 6.0). Thanks!
HTH
Regards,
Iurii
--
Dr. Iurii Timrov
Postdoctoral Researcher
Swiss Federal Institute of Technology Lausanne (EPFL)
Laboratory of Theory and Simulation of Materials (THEOS)
CH-1015 Lausanne, Switzerland
________________________________
From: [email protected] <[email protected]> on behalf of prem
sen <[email protected]>
Sent: Sunday, January 29, 2017 11:58 AM
To: [email protected]
Subject: Re: [Pw_forum] Learning TDDFT calculation in Atomic label (CH4
molecule) using turbo_lanczos.x and turbo_davidson.x
Dear Dr. Iurii Timrov,
Thanks again for your valuable reply.
I knew, how to check the convergence using both the codes but to cross verify
myself, I asked you the question. Now I get the same absorption spectra for CH4
molecule using both the code (turbo_lanczos.x & turbo_davidson.x).
I was also trying for Carbon-dimer. But I am facing a basic problem. Without
using, occupations = "smearing", this system is not getting converged. If I use
"smearing", turbo_davidson.x or turbo_lanczs.x is not possible. I have tried
with different kinds of PPs, "mixing_mode" and "mixing_beta". Is there any
solution for this problem ??
For Example I have used ::
&CONTROL
calculation = "scf",
restart_mode = "from_scratch"
prefix = "C2",
pseudo_dir = "../pseudo/",
outdir = "./tmp/",
/
&SYSTEM
ibrav = 0,
! celldm(1) = 20,
nat = 2,
ntyp = 1,
ecutwfc = 25,
ecutrho = 250,
nosym = .true.,
! occupations = "smearing",
! smearing = "mv",
! degauss = 0.001,
/
&ELECTRONS
diagonalization='david'
mixing_mode = 'plain'
mixing_beta = 0.6,
conv_thr = 1.0d-8
/
CELL_PARAMETERS angstrom
15.0 0.0 0.0
0.0 15.0 0.0
0.0 0.0 15.0
ATOMIC_SPECIES
C 12.01 C.pbe-rrkjus.UPF
ATOMIC_POSITIONS (angstrom)
C 1.255833278 0.000000000 0.000000000
C 0.000000000 0.000000000 0.000000000
K_POINTS {Gamma}
Thanks & regards,
Prem Sen
Ph.D. Student,
IIT Bombay,
Mumbai,India.
_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum
