[Pw_forum] Limit of e-field magnitude for 2-D sheet like systems

Rajdeep Banerjee rajdeep.jzs at gmail.com
Wed Apr 23 13:49:17 CEST 2014


Dear all,
           I am doing relax calculations on monolayer of graphene-like
material (2-D) under the presence of static homogeneous electric field in
z-direction (i.e. perpendicular to the layer). I am using a electric field
range of 0.1-0.5 V/angstrom. The problem is that the bfgs steps are not
converging (i.e. CASE: energy  _new > energy  _old) after 4-5 steps, if I
apply a field value > 0.2 V/angstrom (0.0055 Ry.a.u). Following is the
relevant part of the code:
...
/
 &system
    ibrav = 4,
    celldm(1) = 7.5284, celldm(3) = 5.020259255 ,
    nat = 10, ntyp = 3,
    ecutwfc = 40, ecutrho = 400,
    london = .true.,
/
 &electrons
    diagonalization = 'david', mixing_mode = 'local-TF',
    mixing_beta = 0.4, conv_thr = 1.0d-8, electron_maxstep = 125,
    efield_cart(1) = 0.0,
    efield_cart(2) = 0.0,
    efield_cart(3) = .008
/
 &ions
...
K_POINTS automatic
 12 12 1 0 0 0

            Is it because the applied field is too high? I have checked the
paper Souza et al., PRL, 89, 117602 (2002) where they have put a condition :

            e|E.ai| < e|Ec.ai| ~ Egap/Ni
(E = magnitude of the electric field, ai = lattice vector, Ni = number of
k-points in i direction, Egap = Band gap)
Is this condition true for aperiodic systems like mine where there's
artificial periodicity in z-direction ? If it's true then:
for, Egap = 1.15 eV, N3 = 1 and a3 = 20 angstrom gives an Emax = 0.05
V/angstrom
whereas if I use a3 as the layer thickness which is around 5.55 angstrom
then Emax =  0.0207 V/angstrom which is the value after which the relax
calculations are not converging. Is this interpretation correct?

           Any kind of help is highly appreciated.

Thanks,
Rajdeep Banerjee
(Ph. D. Student)
Theoretical Sciences Unit
Jawaharlal Nehru Centre for Advanced Scientific Research
Bangalore, India
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20140423/6a33d486/attachment.html>


More information about the users mailing list