[Pw_forum] tranmission calculation

Sclauzero Gabriele gabriele.sclauzero at mat.ethz.ch
Tue Jan 28 12:18:04 CET 2014 



Thank you very much for your reply, it is really appreciated. I will changed all parameters of my input file. I have a planer molecule in X-Y plane and  I want to use a planar lead in Y-Z. I realized I have to change the transmission direction to Z.

Good!

Could I have a lead in X-Y plane and calculate transmission for all structure (as an extended molecule or scattering region)?
Reference article for leads: Physica B: Condensed Matter, Volume 406, Issue 20, 15 October 2011, Pages 3969-3974

You can also calculate transport for a system geometry with leads made of "thick" nanowires or extend 2D surfaces as in the paper you cite. The latter case requires to sample the k_perpendicular-dependent transmission in the 2D Brillouin zone. Moreover, the calculation for each k_perp. is much more demanding than in the monatomic chain case.
I would suggest you to start with the simple case of the monatomic chain and then move to the massive leads.
The two PRB papers I have mentioned in my first reply can give you a good idea for the two cases. Those calculations have been performed with PWscf and PWCOND.

GS


Wishes


On Tue, Jan 28, 2014 at 1:41 PM, Gabriele Sclauzero <gabriele.sclauzero at epfl.ch<mailto:gabriele.sclauzero at epfl.ch>> wrote:
Dear Khadije,

    On top of the possible memory problem pointed out by Paolo, I see several issues in your input.

1. Unless things have changed recently, you must use fully-relativistic pseudopotentials in combination with lspinorb=.true. (I wonder why the code does not stop earlier with an error message). Anyway, I suggest you to start with scalar-relativistic collinear calculations. They are much faster and the relaxed structure should not change much. Once you got familiar with collinear calculations in PWscf and PWCOND you might consider investigating spin-orbit effects.

2. You are using PAW datasets, which is a relatively recent feature. I don't think that it has been extended to the transmission calculations with PWCOND yet (please check).

3. The new atomic geometry is still far from what you need for the scattering region. You are still simulating an isolated (0D) object, not a 1D object. I would suggest you to use XCrysDen (http://www.xcrysden.org/) to visualize your input geometry. If you repeat your unit cell along z (using shift-n key), you will probably realize it immediately. If your left/right leads are monatomic Au chains, then you have to put a short linear chain on each side of the molecule (not a square as you do now) and then appropriately choose the length of the unit cell along z.

4. Beware: in the transmission calculations the transport direction is always along z, not along y as you instead seem to assume now. Also, the scatterer has to be at the center of the unit cell, and the continuations of the leads (atomic chains) at the edges. Please have a VERY CAREFUL look at the examples in PWCOND/examples/, also with the aid of xcrysden.

5. You will need to select the k-points in a more suitable way: one k point in the xy plane should be sufficient, while you need to check convergence along z. But before doing this, you need to solve issue #3.


HTH

GS
ETHZ, Materials Theory

Dear Gabriele Sclauzero

Thanks alot for your answer. I changed my molecule and added 4 Au atoms instead of 2. I am getting an error after running it. could you help me please?

input file:
 &control
    calculation = 'relax'
    restart_mode='from_scratch',
    prefix='adn',
    tprnfor = .true.
    pseudo_dir = '/home/khalili/espresso-5.0.2/pseudo',
    outdir='./'
 /
 &system
    ibrav= 0, celldm(1)=6.0,
    nat= 23, ntyp= 5,
    noncolin=.true.,
    lspinorb=.true.,
    starting_magnetization(1)=0.0,
    ecutwfc = 35.0,
    ecutrho = 350.0,
    occupations='smearing',
    smearing='methfessel-paxton',
    degauss=0.01
 /
 &electrons
    diagonalization='david'
    electron_maxstep = 500,
    mixing_mode = 'plain'
    mixing_beta = 0.3
    conv_thr =  1.0d-8
 /
&ions
            ion_dynamics = 'bfgs'
/
ATOMIC_SPECIES
 C   12.0107   C.pbe-n-kjpaw.UPF
 H   1.0       H.pbe-kjpaw.UPF
 N   14.0067   N.pbe-kjpaw.UPF
 Au  196.96655 Au.pbe-dn-kjpaw.UPF
 S   32.065    S.pbe-n-kjpaw.UPF
ATOMIC_POSITIONS
 N   0.170976000  0.000000000  0.000000000
 C  -0.170976000  0.000000000  0.000000000
 N  -0.382005000  0.268863000  0.000000000
 C  -0.210257000  0.563407000  0.000016000
 C   0.142155000  0.600825000 -0.000027000
 C   0.334612000  0.302284000  0.000800000
 N   0.676126000  0.302006000  0.001939000
 N   0.233362000  0.942208000 -0.004019000
 C  -0.064351000  1.105626000 -0.011570000
 N  -0.330002000  0.881899000  0.000315000
 S  -0.270976000 -0.450000000  0.000000000
 H   0.798025000  0.039749000  0.001986000
 H   0.805314000  0.502022000 -0.000734000
 S  -0.064351000  1.568242000 -0.011570000
 H  -0.553279000  0.914575000  0.131427000
 Au -0.464351000  1.914575000 -0.011570000
 Au  0.335649000  1.914575000 -0.011570000
 Au -0.064351000  1.914575000 -0.411570000
 Au -0.064351000  1.914575000  0.388430000
 Au -0.670976000 -0.950000000  0.000000000
 Au  0.129024000 -0.950000000  0.000000000
 Au -0.270976000 -0.950000000  0.400000000
 Au -0.270976000 -0.950000000 -0.400000000
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 1 1 1 1

mpirun -np 20 /home/khalili/espresso-5.0.2/bin/pw.x -in adn.rx.in<http://adn.rx.in/> | tee adn.rx.out


output:
     .
     .
     .
     per-process dynamical memory:   627.4 Mb

     Self-consistent Calculation

     iteration #  1     ecut=    35.00 Ry     beta=0.70
     Davidson diagonalization with overlap

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   EXIT CODE: 9
=   CLEANING UP REMAINING PROCESSES
=   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================
YOUR APPLICATION TERMINATED WITH THE EXIT STRING: Killed (signal 9)
This typically refers to a problem with your application.
Please see the FAQ page for debugging suggestions.

_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org<mailto:Pw_forum at pwscf.org>
http://pwscf.org/mailman/listinfo/pw_forum



--
Khadije Khalili
Ph.D Student of Solid-State Physics
Department of Physics
University of Mazandaran
Babolsar, Iran
kh.khalili at stu.umz.ac.ir<mailto:kh.khalili at stu.umz.ac.ir>

_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org<mailto:Pw_forum at pwscf.org>
http://pwscf.org/mailman/listinfo/pw_forum


Dr. Gabriele Sclauzero
Materials Theory - ETHZ
ETH Zurich, HIT G 43.2
Wolfgang-Pauli-Str. 27
8093 Zürich, Switzerland

Phone +41 44 633 94 10
Fax +41 44 633 14 59
gabriele.sclauzero at mat.ethz.ch<mailto:gabriele.sclauzero at mat.ethz.ch>
www.theory.mat.ethz.ch

More information about the users mailing list