procedure to obtain infrared spectrum (IR) with abinit/anadd  [SOLVED]

Phonons, DFPT, electron-phonon, electric-field response, mechanical response…

Moderators: mverstra, joaocarloscabreu

Locked
bmatt
Posts: 24
Joined: Tue Aug 27, 2013 9:46 pm

procedure to obtain infrared spectrum (IR) with abinit/anadd

Post by bmatt » Sat Dec 26, 2015 6:51 pm

Hello abinit users,
I am trying to obtain an infrared spectrum using abinit with anaddb. I noticed on another forum that you must do the following:
-abinit run: additional variables needed
kptopt = 2
rfstrs = 3
rfelfd = 3 #using previously calculated d/dk response with rfelfd = 2
nqpt = 1 qpt = 0.0 0.0 0.0
rfatpol = 1 2 #complete set
rfdir = 1 1 1

-anaddb run
dieflag = 1
nfreq = 1000
frmin = 0.0
frmax = 2.0e-3 #(hartree)

When I do this I get 0.0 values for the dielectric and no values for the frequency dependent dieletric so I am pretty sure I am not giving specific parameters needed for this calculation. Is there a generic test case for obtaining an IR or does anyone know what I should put (above the normal ground state variables) for the response function calculation in abinit? and what values need to be there for anaddb?
Do I need rfphon = 1 in abinit? do I need the ifcflag activated in anaddb?
Any suggestions appreciated and sample working input files very much appreciated

bmatt
Posts: 24
Joined: Tue Aug 27, 2013 9:46 pm

Re: procedure to obtain infrared spectrum (IR) with abinit/a  [SOLVED]

Post by bmatt » Wed Dec 30, 2015 12:57 am

Feel dumb but here is my solution, not sure if it is perfectly right but I got some output...

abinit input file:
irdwfk = 1 irdddk = 1
rfatpol = 1 natom #all atoms
rfdir = 1 1 1
nqpt = 1 qpt = 0.0 0.0 0.0
rfelfd = 3 # where I have computed the d/dk perturbation in a previous dataset
rfphon = 1
rfstrs = 3

I guess since IR is a response to a lattice vibrations and e-field the rfelfd and rfphon need to be /=0

anaddb input file:
ifcflag = 1
dieflag = 1
nfreq = 1000
frmin = 0.0
frmax = number #(Hartree)

there are other variables but usually standard for calculating phonons in anaddb

- anaddb.out file contains ocilator strengths which can then be spread to resemble experiment
- nothing like answering your own question

Locked