EELS for Benzene with Abinit+Yambo

[gpk]

Dear Yambo List,
I am trying to simulate experimental EELS spectra of benzene Molecule with abinit + Yambo.
The EELS spectrum is plotted by, Imaginary[1/[calculated Epsilon (or dielectric function)]],
as it is mentioned in: http://www.yambo-code.org/doc/docs/doc_LR.php

But also from the site, it states that: "We also note that Eq. (2) is only valid for angular resolved EELS on bulk materials and not for spatially resolved EELS on isolated nanoobjects.(probably molecules ?)"

So can I use abinit+Yambo to simulate EELS spectrum? Or is it still have some tricks? [our plan is to simulate, the plasmon oscillations wrt. the EELS peak especially in the core region].

Many thanks in advance,
Krishna Mohan G P
------------------ ------------------ ------------------ ------------------
Please note that:
am using Yambo in this order: with kss_o_KSS,
[1] /scratch3/krishnan/Scientific/abinitBin_16_etsf/bin/a2y -F kss_o_KSS
[2] /scratch3/krishnan/Scientific/abinitBin_16_etsf/bin/yambo
[3] /scratch3/krishnan/Scientific/abinitBin_16_etsf/bin/yambo -o c # to create yambo.in
[4] /scratch3/krishnan/Scientific/abinitBin_16_etsf/bin/yambo
and finally, I got a EELS at ~ 5eV (which is something similar to the Benzene pi-pi* transition!).
[2] Files used for EELS
--- --- --- --- --- --- --- --- ---
band.files [just for band calc]
::::::::::::::
band.in
band.out
band_i
band_o
band_t
06c.pbe_hgh
01h.pbe_hgh


***************************************************

::::::::::::::
relax.in
::::::::::::::
acell 10 10 10 Angstrom
rprim 1 0 0
0 1 0
0 0 1

xangst 0.000 1.390 0.000
1.204 0.695 0.000
0.000 -1.390 0.000
-1.204 -0.695 0.000
1.204 -0.695 0.000
-1.204 0.695 0.000
0.000 2.470 0.000
2.139 1.235 0.000
0.000 -2.470 0.000
-2.139 -1.235 0.000
2.139 -1.235 0.000
-2.139 1.235 0.000

ntypat 2
znucl 6 1
natom 12
typat 1 1 1 1 1 1 2 2 2 2 2 2

kptopt 0
nkpt 1
kptnrm 1


ecut 20
ixc 11
iscf 7
npulayit 7
wtk 1.0
nstep 150
diemac 12.0
toldfe 1.0d-7


symmorphi 0

ionmov 3
ntime 10


::::::::::::::
band.in
::::::::::::::
acell 10 10 10 Angstrom
rprim 1 0 0
0 1 0
0 0 1
ntypat 2
znucl 6 1
natom 12
typat 1 1 1 1 1 1 2 2 2 2 2 2

xangst 2.8538984130E-21 1.4178793921E+00 0.0000000000E+00
1.2270791794E+00 7.0935608787E-01 0.0000000000E+00
2.8538984130E-21 -1.4178793921E+00 0.0000000000E+00
-1.2270791794E+00 -7.0935608787E-01 0.0000000000E+00
1.2270791794E+00 -7.0935608787E-01 0.0000000000E+00
-1.2270791794E+00 7.0935608787E-01 0.0000000000E+00
2.8538984130E-21 2.5166752386E+00 0.0000000000E+00
2.1795461820E+00 1.2568551003E+00 0.0000000000E+00
2.8538984114E-21 -2.5166752386E+00 0.0000000000E+00
-2.1795461820E+00 -1.2568551003E+00 0.0000000000E+00
2.1795461820E+00 -1.2568551003E+00 0.0000000000E+00
-2.1795461820E+00 1.2568551003E+00 0.0000000000E+00

ndtset 2

# Definition of the k-point grids
kptopt1 0
nkpt1 1
kptnrm1 1

# output the density of states
prtdos1 1
# output the charge density
prtden1 1

# Definition of the planewave basis set
ecut 20 # Maximal kinetic energy cut-off, in Hartree
ixc 11 # XC potential Perdew-Zunger

# Definition of the SCF procedure
nstep 50 # Maximal number of SCF cycles
toldfe1 1.0d-7 # Will stop when this tolerance is achieved on total energy
diemac 12.0 # Although this is not mandatory, it is worth to
# precondition the SCF cycle. The model dielectric
# function used as the standard preconditioner
# is described in the "dielng" input variable section.
# Here, we follow the prescription for bulk silicon.

# Calculation of the band structure
iscf2 -2 # non self consistent calculation
getden2 -1 # read the charge density
kptopt2 -7 # 7 segments
ndivk2 4 21 16 8 18 11 21 # number of divisions of the segments
kptbounds2
0.25 0.625 0.625 # U
0.25 0.5 0.75 # W
0 0 0 # Gamma
0 0.5 0.5 # X
0.25 0.5 0.75 # W
0 0.5 0 # L
0 0 0 # Gamma
0.375 0.375 0.75 # K

tolwfr2 1.0d-12 # tolerance on wavefunction squared residual
nband2 34 # number of bands *** *** since 17band found in the relax.out
enunit2 1 # Will output the eigenenergies in eV
prtvol2 2 # output printing option

symmorphi 0


occopt1 3 # otherwise ERROR ...

::::::::::::::
kss.in
::::::::::::::
acell 10 10 10 Angstrom
rprim 1 0 0
0 1 0
0 0 1
ntypat 2
znucl 6 1
natom 12
typat 1 1 1 1 1 1 2 2 2 2 2 2

xangst 2.8538984130E-21 1.4178793921E+00 0.0000000000E+00
1.2270791794E+00 7.0935608787E-01 0.0000000000E+00
2.8538984130E-21 -1.4178793921E+00 0.0000000000E+00
-1.2270791794E+00 -7.0935608787E-01 0.0000000000E+00
1.2270791794E+00 -7.0935608787E-01 0.0000000000E+00
-1.2270791794E+00 7.0935608787E-01 0.0000000000E+00
2.8538984130E-21 2.5166752386E+00 0.0000000000E+00
2.1795461820E+00 1.2568551003E+00 0.0000000000E+00
2.8538984114E-21 -2.5166752386E+00 0.0000000000E+00
-2.1795461820E+00 -1.2568551003E+00 0.0000000000E+00
2.1795461820E+00 -1.2568551003E+00 0.0000000000E+00
-2.1795461820E+00 1.2568551003E+00 0.0000000000E+00

# Definition of the planewave basis set
ecut 20 # Maximal kinetic energy cut-off, in Hartree
ixc 11 # XC potential Perdew-Zunger

# Definition of the SCF procedure
nstep 50 # Maximal number of SCF cycles
#toldfe 1.0d-7 # Will stop when this tolerance is achieved on total energy
diemac 12.0 # Although this is not mandatory, it is worth to


ngfft 80 80 80

# Definition of parameters for the calculation of the kss file
iscf -2 # non self-consistency, read previous density file
#getden 1 # read the density file
prtvol 2 # output printing option
tolwfr 1.0d-12 # tolerance on wavefunction squared residual
nband 21 # number of bands for calculations
nbandkss 21 # number of bands for eigenstate output (_KSS file)
nbdbuf 10 # number of bands for the buffer
npwkss 3000 # number of planewave components for eigenstate output
kssform 3 # kss in double precision
symmorphi 0

istwfk 1
paral_kgb=0

# Definition of the k-point grids
kptopt 0
nkpt 1
kptnrm 1

[Daniele Varsano]

Dear krishna,

Plasmon oscillation of a finite system I think is not well defined.
Anyway, what you can do is: first of all instruct Yambo you are dealing with a finite system:
In order to do that you have to set the non_periodic_direction (NonPDirs="XYZ"), have
a look here.

In this way, Yambo will give you as output of a linear response calculation, the dynamical
polarizability:
See here the definition and
pay attention to the dimensions. This is what should be meaningful for a finite molecule.
If I remember well, for finite system, the absorption and EELS definition should coincide
in the limit of infinite supercell value (Sottile et al. INT. J. OF QUANTUM CHEMISTRY, 102, 684, 2005.).
Have also a look to this post that
can be useful for you.

Moreover when dealing with finite system, you can also truncating the Coulomb interaction in order
to avoid the repeated images of the system (dictated by the plane wave representation) to
interact each other. You can have a look at this paper ( Phys. Rev. B 73, 205119 (2006).)
To do that, include the -c option when building up your input file and fill the corresponding field
to set the shape and size of the cutoff potential. Anyway this effects is much more bigger in GW/BSE calculations
than in RPA.

Anyway, if you are interested in core electrons, pay attention to the fact that Yambo use pseudopotentials!!

Cheers,
Daniele

[gpk]

Dear Dan

Many thanks for the help and the explanations...
I will mail to this forum If I can simulate EELS of molecules

Thanks and regards
Krishna Mohan