[Pw_forum] tranmission calculation

[Gabriele Sclauzero]

Dear Khadije Khalili

> Dear QE users
> I am doing transmission calculation for a molecular wire, for example, a Benzene molecule sandwiched between two gold electrode. I did relax and scf calculation for a benzene molecule that 2 H atoms were substituted with 2 Au atoms.

I could not find the file for the scattering region in your attachments. In any case, I think it is not sufficient to include one gold atom on each side of the molecule, because the scattering cell has to describe a 1D geometry, not just an isolated molecule (scatterer). In other words, the scattering region must contain part of the left and the right leads at the sides of the scatterer.
You need to include enough gold atoms on each side such that the Kohn-Sham potential at the apex of the scattering region does not "feel" the perturbation of the molecule in the middle. See, for instance, Fig.1 of http://prb.aps.org/abstract/PRB/v85/i16/e165411.
You can check this by comparing the complex band structure of the one atom cell obtained by cutting out an edge of the scattering cell (using prefixt="benzene" and bdl=..., see the PWcond user guide) with that of the lead region. See Alexander Smogunov's paper http://www.sciencedirect.com/science/article/pii/S0039602803004539, and following ones citing this one.

By the way, I think that in your case you could make the scattering cell symmetric and use ikind=1 for the transmission calculation. There is no need for distinct left and right leads.
Moreover, have you actually understood the meaning of ikind? ikind=0 is for the complex band structure calculation, while for the transmission you need ikind=1 (equal leads) or ikind=2 (distinct left and right leads). Please see here http://iramis.cea.fr/Pisp/alexander.smogunov/ and have a careful look to the examples in PWcond/examples. 


> Then, I did scf calculation for a single Au atom with the coordination just like as later calculation. Could you say me if my calculation is true? Or I have to do relax and scf calculation for a single benzene?

Relaxing or not the atomic structure in the scattering region is a delicate matter with this geometry, because you do not have massive leads and you cannot just fully relax all atoms in the scattering region. You might try to relax the molecule and a few gold atoms on each side, while keeping fixed those toward the edges (after you include them). You might also repeat this operation for different values of the cell size along the transport direction and take the one with the lowest energy.
In any case, to get realistic conductance value you need to include massive leads (i.e., gold surfaces or tips, see our other 2012 PRB: http://prb.aps.org/abstract/PRB/v85/i16/e165412).

HTH

Gabriele Sclauzero
Materials Theory, ETHZ

> 
> aur.scf.in:
> &control
>    calculation='scf'
>    restart_mode='from_scratch',
>    pseudo_dir = '/home/client3/espresso-5.0.2/pseudo',
>    outdir='./'
>    prefix='aur'
> /
> &system
>    ibrav = 0,
>    celldm(1) =6.0,
>    nat= 1,ntyp= 1,nspin = 1,
>    ecutwfc = 35.0,
>    ecutrho = 350.0
>    occupations='smearing',
>    smearing='methfessel-paxton',
>    degauss=0.01
> /
> &electrons
>    conv_thr = 1.0e-8
>    mixing_beta = 0.7
> /
> ATOMIC_SPECIES
>  Au  196.96655 Au.pbe-dn-kjpaw.UPF
> ATOMIC_POSITIONS
> Au      -0.444314932  -0.706149130   0.008586376
> CELL_PARAMETERS cubic
>  6.0   0.0   0.0
>  0.0   6.0   0.0
>  0.0   0.0   6.0
> K_POINTS (automatic)
> 2 2 2 1 1 1
> 
>  aul.scf.in
> &control
>    calculation='scf'
>    restart_mode='from_scratch',
>    pseudo_dir = '/home/client3/espresso-5.0.2/pseudo',
>    outdir='./'
>    prefix='au'
> /
> &system
>    ibrav = 0,
>    celldm(1) =6.0,
>    nat= 1,ntyp= 1,nspin = 1,
>    ecutwfc = 35.0,
>    ecutrho = 350.0
>    occupations='smearing',
>    smearing='methfessel-paxton',
>    degauss=0.01
> /
> &electrons
>    conv_thr = 1.0e-8
>    mixing_beta = 0.7
> /
> ATOMIC_SPECIES
>  Au  196.96655 Au.pbe-dn-kjpaw.UPF
> ATOMIC_POSITIONS
> Au      -0.174740372   1.847908104   0.017126043
> CELL_PARAMETERS cubic
>  6.0   0.0   0.0
>  0.0   6.0   0.0
>  0.0   0.0   6.0
> K_POINTS (automatic)
> 2 2 2 1 1 1
> 
> benzene.cond.in
> &inputcond
>     outdir='./'
>     prefixl='aul'
>     prefixr='aur'
>     prefixs='benzene'
>     band_file ='bands.dat'
>     ikind=0
>     energy0=1.d0
>     denergy=-0.1d0
>     ewind=3.d0
>     epsproj=1.d-4
>     nz1=1
> /
>     1
>     0.0 0.0 1.0
>     40
> 
> 
> Waiting for your reply.
> Thanks and Regards
> -- 
> Khadije Khalili
> Ph.D Student of Solid-State Physics
> Department of Physics
> University of Mazandaran
> Babolsar, Iran
> kh.khalili at stu.umz.ac.ir


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20140122/de4a412f/attachment.html>