ABINIT Discussion Forums

Dear ABINIT Users / Developers,

I wish to compute the optical absorption and optical conductivity of graphene.
I know ABINIT has a program called 'conducti' and 'conducti_paw' but I cannot find a manual or tutorial for them.

I also searched the ~/tests/v6/ folder and the README file contains no mention to them, neither
the ~/tests/v6/Inputs folder contains a .in file to show the usage.

How can I learn to use them?

Best,

Hi Temok,

I have tentatively used conducti (I may be wrong and appreciate any suggestions) but this is the procedure I used ( I got these from the ABINIT test files but don't remember where):

Step 1: generate the wavefunctions for Step 2. A possible input file is as follows:

ndtset 5
enunit 2
prtvol 1
# - Parameters governing the convergence
ecut 25
nband 30
occopt 7
tsmear 0.01
# - Options for the choice of calculations
ixc 11
nstep 200
# SC run
iscf1 3
toldfe1 1.d-10
nqpt1 0
prtden1 1
# NSC run
iscf2 -2
tolwfr2 1.0d-22
nqpt2 0
getden2 1
getwfk2 1
# ddk response function
iscf3 -3
tolwfr3 1.0d-22
nqpt3 1
qpt3 0.0d0 0.0d0 0.0d0
rfdir3 1 0 0
rfelfd3 2
getwfk3 2
# ddk response function
iscf4 -3
tolwfr4 1.0d-22
nqpt4 1
qpt4 0.0d0 0.0d0 0.0d0
rfdir4 0 1 0
rfelfd4 2
getwfk4 2
# ddk response function
iscf5 -3
tolwfr5 1.0d-22
nqpt5 1
qpt5 0.0d0 0.0d0 0.0d0
rfdir5 0 0 1
rfelfd5 2
getwfk5 2
# - Definition of special k-points
kptopt 1
ngkpt 8 8 8 # This may be increased to test convergence
nshiftk 1
shiftk 0 0 0
wtk 1.0
kptnrm 1
# - Symmetries of the lattice
nsym 0
# This line added when defaults were changed (v5.3) to keep the previous, old behavior
iscf 5
# Data common to the three different datasets
# Definition of the unit cell
acell #specify
angdeg or rprim #specify
# Definition of the atom types
ntypat # specify
znucl #specify
# Definition of the atoms
natom #specify
typat #specify
xred #specify

After running ABINIT, you get the following files: H.o_DS3_1WF145, H.o_DS4_1WF146, H.o_DS5_1WF147, H.o_DS2_WFK.

Step 2: Create an input file such as the following (con.in):
1
H.o_DS3_1WF145
H.o_DS4_1WF146
H.o_DS5_1WF147
H.o_DS2_WFK
1E-02
# paste all the wtk values from the output file of Step 1 here. This may be a long list.
0.005 2.0
!Here is a description of the above lines
!first order wave function file for direction 1 0 0
!first order wave function file for direction 0 1 0
!first order wave function file for direction 0 0 1
!tsmear in Hartree
!wtk as it is written in the outfile
!space between two points, conductivity is compute up to this value in Ha

Then issue the command conducti < con.in
The files with extensions out and sig contain the conductivity data.

Hope this helps.

Vahid Askarpour
Physics Department
Dalhousie University
Halifax, Canada

Thanks for your example Vahid,

Similar examples are found in the ABINIT tree at
../abinit-6.0.3/tests/v3/Input/t78.in AND t79.in (PSPNC)
../abinit-6.0.3/tests/v5/Input/t09.in AND t10.in (PAW)

Basic manuals are found at
../abinit-6.0.3/doc/users/conducti_manual.pdf
../abinit-6.0.3/doc/users/conductivity_paw_manu.pdf

(Thanks T. Rangel)

-------

Now I wish to know if it is possible to introduce a scattering [time] rate
in computing the conductivity for a metal.

Thanks all,

Dear Vahid,
I am getting started with 'conducti', I was wondering how to pick the second last value, the '!space between two points', should i find it in the output file?
Thank you for your attention,
Valentina

Hi,
I have a guess about how to pick the second last value, the '!space between two points', of the input file for conducti. I think that it should be less or of the order of the energy difference between the last occupied and the first empty energy level, information that i can get from the _EIG file. Am i right?
Valentina

Hello Temok and Vahid,
i am using abinit-6.0.4,and i wana to calculate conductivity for metal but i can't find tutorial.
would you please help me to get information about it?
Best regards.
Mina.

Hi Valentina and Mina,

The last two values define your energy axis. The first value defines the energy spacing between the calculated conductivity values. You can choose it to be small (for finer curves). The second value defines the maximum energy up to which you would like to calculate the conductivity.

For the manual please see the posting by Temok on this thread.

Vahid

Dear Vahid Askarpour,

Did you try to calculate this conductivity using parallel case? That is to say, in principle, for the first step, calculate the density and wavefunction, we can use parallel calculations to get them with KGB parallelation. Thus, we can add some parameters in the input file:
paral_kgb 1
npkpt
npband
npfft
bandpp

This parallerism is OK for the SCF calculation, but it will stop after SCF calculation, .ie., for the send dataset "iscf2 -2". The hints show that there may be some problem in the reading previous DEN and WFK files. Did you meet some problems like this?

Thanks a lot.

Jiayu Dai
NUDT

Dear Jiayu,

I just saw your posting. First I did the SFC and nSFC runs, then I set up three separate runs for the three directions in order to generate the wavefunctions always using the following tags.

wfoptalg 1
nbdblock 100

Cheers!
Vahid

Dear Vahid,

Thanks for your reply. In fact, i found that the using of nbdblock can not acceleret the speed much. Therefore, i would like to use the KGB parallelation. Recently, i changed the set of my disk system, and use the accesswff = 1, then the calculations can be completed and the speed is accelerated much.

Thanks a lot.

Jiayu
NUDT