Finite electric field calculations without fixing ions

Total energy, geometry optimization, DFT+U, spin....

Moderator: bguster

Locked
Anwar
Posts: 6
Joined: Fri Jul 15, 2011 4:03 pm

Finite electric field calculations without fixing ions

Post by Anwar » Mon Sep 19, 2011 11:29 am

Hi everybody. I'm a newcomer and an inexperienced user of ABINIT.

I've been trying to calculate the polarization of perovskite BaTiO3 as a function of finite electric field. So, following the tutorial about polarization and electric filed, I calculated the polarization for various electric field (using a structure that I have optimized previously).

The problem I face is that the Polarization vs. Electric Filed curve comes out as a straight-line even for electric fields as high as 0.0001 a.u. Also there is no sign of hysteresis.

So I thought may be the problem is that the nuclear positions are fixed in the tutorial whereas the hysteresis / polarization switching in perovskites is due to the movement of central B atom and the O atoms under electric field.

Is there anyway to to calculate the polarization of a crystal without fixing the ions (i.e. with ionmov = 2 or 6)?

This is my input file:

Code: Select all




#Definition of the elementary cell
#*********************************
   acell 7.5120216890E+00  7.5120216890E+00  7.6915729471E+00 Bohr
   rprim 1 0 0
         0 1 0
         0 0 1

#Definition of the atoms
#***********************
   natom 5
   ntypat 3
   znucl   56 22 08
   typat 1 2 3 3 3
   ixc 11
   xred   0.0000000000E+00  0.0000000000E+00  0.0000000000E+00
                     5.0000000000E-01  5.0000000000E-01  5.1430000000E-01
                     5.0000000000E-01  5.0000000000E-01 -3.0760000000E-02
                     5.0000000000E-01  0.0000000000E+00  4.8422377187E-01
                     0.0000000000E+00  5.0000000000E-01  4.8422377187E-01

#Definition of the SCF procedure
#*******************************
   iscf 5
   nstep 100
   nband 12
   nbdbuf 0


#Definition of the plane wave basis set
#**************************************
   ecut  35
   ecutsm 0.5
   dilatmx 1.15
   ngkpt 6 6 6
   nshiftk 1
   shiftk 0.5 0.5 0.5
         
   toldfe 1.0d-12
   tphysel 300 K
   getwfk  -1

  ndtset  21
! ndtset   3
jdtset  11
        21     22  23  24  25     # The additional 8 values of the field have been suppressed to spare CPU time
        31     32  33  34  35
        41     42  43  44  45
        51     51  53  54  55
                                                                               
berryopt11 -1       rfdir11    1 1 1
                                                                               
berryopt21  4       efield21   0.000002  0.000002  0.000002   
berryopt22  4       efield22   0.000004  0.000004  0.000004   
berryopt23  4       efield23   0.000006  0.000006  0.000006   
berryopt24  4       efield24   0.000008  0.000008  0.000008   
berryopt25  4       efield25   0.000010  0.000010  0.000010   
                                                                               
berryopt31  4       efield31  0.000008 0.000008 0.000008   
berryopt32  4       efield32  0.000006 0.000006 0.000006   
berryopt33  4       efield33  0.000004 0.000004 0.000004   
berryopt34  4       efield34  0.000002 0.000002 0.000002   
berryopt35  4       efield35  0.000 0.000 0.000   





P.S. Another question has been bothering me. When I do the structural optimization run, the energy converges to a very high value (about 50 Ha). My input file for optimization:

Code: Select all


  ndtset   2       # There are 2 datasets in this calculation

# Set 1 : Internal coordinate optimization

  ionmov1   2       # Use BFGS algorithm for structural optimization
   ntime1   7       # Maximum number of optimization steps
  tolmxf1   1.0e-5  # Optimization is converged when maximum force
                    # (Hartree/Bohr) is less than this maximum
  natfix1   3       # Fix the position of two symmetry-equivalent atoms
                    #  in doing the structural optimization
  iatfix1   1 2 3     # Choose atoms 1 and 2 as the fixed atoms (see discussion)
  optcell1  0

# Set 2 : Lattice parameter relaxation (including re-optimization of
#         internal coordinates)

 dilatmx2   1.15    # Maximum scaling allowed for lattice parameters
 getxred2   -1      # Start with relaxed coordinates from dataset 1
  getwfk2   -1      # Start with wave functions from dataset 1
  ionmov2   2       # Use BFGS algorithm
   ntime2   10      # Maximum number of optimization steps
 optcell2   3       # Fully optimize unit cell geometry, keeping symmetry
  tolmxf2   1.0e-5  # Convergence limit for forces as above
 strfact2   100     # Test convergence of stresses (Hartree/bohr^3) by
                    # multiplying by this factor and applying force
                    # convergence test
  natfix2   3       
  iatfix2   1 2 3         

#Common input data

acell   3.992  3.992  4.036  angstrom           #this is a guess, with the c/a
                                        #ratio based on ideal tetrahedral
                                        #bond angles

  rprim  1 0 0   #  primitive vectors must be
         0 1 0   #specified with high accuracy to be
         0 0 1   #sure that the symmetry is recognized
                  #and preserved in the optimization
                  #process

#Definition of the atom types and atoms
 ntypat   3
  znucl   56 22 08
  natom   5
  typat   1 2 3 3 3

#Starting approximation for atomic positions in REDUCED coordinates
#based on ideal tetrahedral bond angles
   xred   0.0 0.0  0.0
          0.5 0.5  0.5143
          0.5 0.5 -0.03076
          0.5 0.0  0.4814
          0.0 0.5  0.4814
#Gives the number of bands, explicitely (do not take the default)
  nband   15           # For an insulator (if described correctly as an
                         # insulator by DFT), conduction bands should not
                         # be included in response-function calculations

#Definition of the plane wave basis set
   ecut   50          # Maximum kinetic energy cutoff (Hartree)
 ecutsm   0.5            # Smoothing energy needed for lattice paramete
                         # optimization.  This will be retained for
                         # consistency throughout.
ixc = 11                 !GGA, Perdew-Burke-Ernzerhof GGA functional
tphysel 300 K
#Definition of the k-point grid
  ngkpt   6 6 6          # 6 6 6 grid
nshiftk   1              # Use one copy of grid only (default)
 shiftk   0.5 0.5 0.5    # This choice of origin for the k point grid
                         # preserves the hexagonal symmetry of the grid,
                         # which would be broken by the default choice.

#Definition of the self-consistency procedure
 diemac   4000           # Model dielectric preconditioner
   iscf   7              # Pulay mixing of the potential
  nstep   40             # Maxiumum number of SCF iterations
 tolvrs   1.0d-18        # Strict tolerance on (squared) residual of the
                         # SCF potential needed for accurate forces and
                         # stresses in the structural optimization, and
                         # accurate wave functions in the RF calculations



The etotal and fcart are:

Code: Select all


           etotal1  -5.3094090418E+01
           etotal2  -5.3094307091E+01
            fcart1  -0.0000000000E+00 -0.0000000000E+00 -6.9446291206E-03
                    -0.0000000000E+00 -0.0000000000E+00 -2.4101663245E-02
                    -0.0000000000E+00 -0.0000000000E+00  3.1042348651E-02
                    -0.0000000000E+00 -0.0000000000E+00  1.9718573189E-06
                    -0.0000000000E+00 -0.0000000000E+00  1.9718573189E-06
            fcart2  -0.0000000000E+00 -0.0000000000E+00 -6.7464332558E-03
                    -0.0000000000E+00 -0.0000000000E+00 -1.9553866581E-02
                    -0.0000000000E+00 -0.0000000000E+00  2.6297646773E-02
                    -0.0000000000E+00 -0.0000000000E+00  1.3265321034E-06
                    -0.0000000000E+00 -0.0000000000E+00  1.3265321034E-06


User avatar
jzwanzig
Posts: 504
Joined: Mon Aug 17, 2009 9:25 am

Re: Finite electric field calculations without fixing ions

Post by jzwanzig » Mon Sep 19, 2011 12:27 pm

It is possible to compute structural relaxation in the presence of a finite electric field. You can use ionmov 2 for example. I don't think you can use optcell different from zero though (that is, you can't allow the cell size to vary, only the ion positions).
Josef W. Zwanziger
Professor, Department of Chemistry
Canada Research Chair in NMR Studies of Materials
Dalhousie University
Halifax, NS B3H 4J3 Canada
jzwanzig@gmail.com

Anwar
Posts: 6
Joined: Fri Jul 15, 2011 4:03 pm

Re: Finite electric field calculations without fixing ions

Post by Anwar » Mon Sep 19, 2011 2:34 pm

Thank you Professor. I shall append ionmov 2 to my input file. I hope allowing the structure to relax (by moving ions) will be reflected at the calculated wavefunction and subsequently, in the berry polarization.

User avatar
jzwanzig
Posts: 504
Joined: Mon Aug 17, 2009 9:25 am

Re: Finite electric field calculations without fixing ions

Post by jzwanzig » Mon Sep 19, 2011 2:40 pm

I have successfully computed \epsilon_\infty (clamped ions) and \epsilon_0 (relaxed ions) using this method. Don't forget that with ionmov 2 you are doing a structural relaxation so you also have to set ntime, tolmxf, and ecutsm.
Josef W. Zwanziger
Professor, Department of Chemistry
Canada Research Chair in NMR Studies of Materials
Dalhousie University
Halifax, NS B3H 4J3 Canada
jzwanzig@gmail.com

Anwar
Posts: 6
Joined: Fri Jul 15, 2011 4:03 pm

Re: Finite electric field calculations without fixing ions

Post by Anwar » Mon Sep 19, 2011 2:45 pm

jzwanzig wrote:I have successfully computed \epsilon_\infty (clamped ions) and \epsilon_0 (relaxed ions) using this method. Don't forget that with ionmov 2 you are doing a structural relaxation so you also have to set ntime, tolmxf, and ecutsm.


Once again I'm at your debt. I'm actually trying to calculate the optical dielectric tensor (as a function of changing electric field) for ferroelectrics. I can breath easier knowing it is not impossible :) . I'll add the appropriate commands to my input and let you know as soon as I get results. Since I'm not doing calculations in parallel, that may take a while.

sevket simsek
Posts: 18
Joined: Tue May 11, 2010 1:05 pm

Re: Finite electric field calculations without fixing ions

Post by sevket simsek » Mon Feb 26, 2018 2:09 pm

Dear Abinit user,

Is there anyone who has calculated the hysteresis curve for a ferroelectric material?
I calculated, but I got a straight line, I could not get the hysteresis curve.

ebousquet
Posts: 469
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: Finite electric field calculations without fixing ions

Post by ebousquet » Mon Feb 26, 2018 3:28 pm

Dear Sevket,
What about Fig.5.c of:
https://journals.aps.org/prb/abstract/10.1103/PhysRevB.84.115107
Done for PbTiO3?
Best wishes,
Eric

sevket simsek
Posts: 18
Joined: Tue May 11, 2010 1:05 pm

Re: Finite electric field calculations without fixing ions

Post by sevket simsek » Mon Feb 26, 2018 10:11 pm

Dear Eric,
Thank you very much for your reply before anything else.
I think you're talking about figure 5b.
I have calculated for tetragonal BaTiO3 and obtained a result as you indicated in Figure 5b.
but I want to get a curve like the one below.
if you share your advice and suggestion in this matter, I am delighted.
Best wishes
Attachments
Hysteresis loops of BaTiO3.png
Hysteresis loops of BaTiO3.png (75.61 KiB) Viewed 5851 times

ebousquet
Posts: 469
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: Finite electric field calculations without fixing ions

Post by ebousquet » Wed Feb 28, 2018 10:51 am

Dear Sevket,
Humm, what you want is beyond static 0K DFT, the experimental hysteresis involves a much more complex dynamical process with domains, defects, leakage, etc, which give "imperfect" hysteresis (see for example the discussion in http://iopscience.iop.org/article/10.1088/0953-8984/20/02/021001/meta at the experimental side).
As it is now, you get somehow the "hysteresis" of a perfect monodomain crystal at 0K. Going a bit further requires to make an effective Hamiltonian or atomic potential model.
Best wishes,
Eric

sevket simsek
Posts: 18
Joined: Tue May 11, 2010 1:05 pm

Re: Finite electric field calculations without fixing ions

Post by sevket simsek » Wed Feb 28, 2018 12:53 pm

Dear Eric,

I thank you for useful discussions.

Best wishes.

Locked