ABINIT Discussion Forums

Dear all ,please
My name is Abdulmutta . I am Ph.D programme department of Physics . Prince of Songkla University
I have a problems in optimizes structure tetragonal phase.
This my work I want to calculated optimizes structure in tetragonal phase
of PbTiO3 (Lead Titanate) perovskite structure I want to find lattice constant (a) and
(c/a) in tetragonal phase I used ionmove=3 and this is my input file
Please suggestion and comment my input file and How I have going

# PbTiO3 : computation of the total energy
# pure of Ti.The convergence in acell

ndtset 10

#Definition of the unit cell and ecut,
#for which one will have to make a convergence study
acell: 7.4 7.4 7.4 acell+ 0.02 0.02 0.02
optcell 6
ecutsm 2.5
dilatmx 2.0
#natfix 1
#iatfix 1

# find the optimal total energy
natom 5
ionmov 3
ntime 20
tolmxf 5.0d-8
xred 0.0 0.0 0.0
0.5 0.5 0.5
0.5 0.5 0.0
0.0 0.5 0.5
0.5 0.0 0.5

toldff 1.0d-6

nband 14

rprim 1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0

#Definition of the atom types
ntypat 3
znucl 82 22 8

#Definition of the atoms
#natom 5
typat 1 2 3 3 3

#Definition of the planewave basis set
ecut 40.0

#Definition of the k-point grid
kptopt 1
ngkpt 4 4 4
nshiftk 1
shiftk 0.5 0.5 0.5

#Definition of the SCF procedure
nstep 100
#toldfe 1.0d-6
diemac 12.0

Dear Abdulmutta,

A rather robust technique to optimize a phase without constraint is to use two datasets :
(1) optimization of the atomic positions without relaxing the lattice parameters (so a and c are fixed,
and one uses optcell 0 and ionmov 2), then
(2) on the basis of the geometry determined by dataset 1 (so using getxcart 1), to follow with optcell 3 and ionmov 2 .

Now, coming to your input file, the first obvious problem is that you feed into the calculation
a fully cubic geometry. ABINIT will not break this cubic symmetry by itself ! The first think to do is to
use a different values for a and b lattice parameters, than for c lattice parameter. Starting with the experimental
geometry will be the best.
Then,
(1) I remark that you feed different values for a , for different datasets. For your information,
this technique is not coherent with the technique outlined above.
(2) The dilatmx value seems very high.
(3) tolmxf is very small 1.0d-5 should be enough (especially, tolmxf should never be as small or even smaller than toldff)

Best wishes,
X.

Dear GONZE
I am following with you introduce in my input file
the final I found that a, b, c lattice parameter is 7.5332 7.5332 7.5354
it become to nearly cubic phase . The data in experimental is 7.3775 7.3775 7.8570
it so far. What I have to going please

# PbTiO3 : computation of the total energy
# pure of Ti.The convergence in acell
# Dataset 1: ground state calculation,optimization of the atomic positions without relaxing the lattice parameters
# Dataset 2: on the basis of the geometry determined by dataset 1 (so using getxcart 1), to follow with optcell 3 and ionmov 2 .

ndtset 2

#Definition of the atomic positions without relaxing the lattice parameters and fixed Pb
optcell1 0
ionmov1 2

#Definition of the basis of the geometry determined by dataset 1 (so using getxcart 1)
getxcart2 1
optcell2 3
ionmov2 2
dilatmx2 1.5
ecutsm2 0.5

# COMMON INPUT DATA

#Definition of the unit cell
acell 7.3775 7.3775 7.8570
rprim 1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0

#Definition of the atom types
ntypat 3
znucl 82 22 8

#Definition of the atoms
natom 5
typat 1 2 3 3 3
xred
0.0 0.0 0.0
0.5 0.5 0.5
0.5 0.5 0.0
0.0 0.5 0.5
0.5 0.0 0.5

#Definition of the planewave basis set
ecut 35.0

# find the optimal total energy
ntime 25
tolmxf 1.0d-5
toldff 1.0d-6
nband 14

#Definition of the k-point grid
kptopt 1
ngkpt 4 4 7
nshiftk 1
shiftk
0.5 0.5 0.5
getwfk -1

#Definition of the SCF procedure
nstep 100
#toldfe 1.0d-6
diemac 12.0

Dear Mutta,

One more point : you should use
ngkpt 4 4 4

Apart of this ... well ... try !
You have to make yourself your own experience ...
X

Dear GONZE
I try it but the lattice parameter not consist with experimental data
and I try many method for this my input one more equation I want to
fixed Pb(Lead) atom and displacement Ti(Titanium) atom along z axis
and O(Oxygen) atom to be move freedom.
PbTiO3 is a perovskite oxide which is known to behave as a ferroelectric.
The ferroelectric response of BaTiO3 is commonly thought to be the result
of a displacive transition. In this problem, we will study the energetics of
cubic BaTiO3 and use first principles calculations to gather information
pertaining to the displacive transition to the tetragonal phase.
For displacive transition atom in the input file how I have going please

Abdulmutta

Dear all
PbTiO3 ,Tetragonal structure was optimized by letting the fractional coordinates of the ions change in the way symmetry allows (distortion in z direction) at fixed
c/a ratio by keeping the volume fixed. Then, by using the optimized fractional coordinates of the ions, c/a
ratio was optimized at constant volume. Thus, c/a ratio, ∆T −T i , ∆T −O1 and ∆T −O2 were the quantities
that were optimized during the minimization.
My equation is how I can set only acell(3) are optimized at constant volume ? optcell= 3 not consist with acell(3)

Abdulmutta

Dear Mutta
I was following your problem in optimization of PbTiO3 in tetragonal phase. I have the same problem. The optimized structure is very close to cubic phase.
Could you please let me know what was your structural optimization procedure? How did you solve the problem?did you get different values for a and c lattice parameters?
Best regards,Payam Norouzzadeh

Dear Payam
I have an idea for solve as your problem by displacement only Ti atom about 0.01 as z direction, this could be solve.
And another thinks you could used EXP. data for lattice parameters and fraction coordinates.

Please let me know if it work
Abdulmutta

Dear Mutta
I start from experimental data but do you mean I should fix all atoms but Ti in z direction and then let the crystal to be relaxed?
Best regards,Payam Norouzzadeh

Dear Payum
Yes, start with optcell 2 , ionmov 2 and then
for examples
Pb 0.0 0.0 0.0
Ti 0.5 0.5 0.51
O 0.5 0.5 0.0
O 0.5 0.0 0.5
O 0.0 0.5 0.5

Hope this help !
Mutta
Department of Physics, Faculty of Science
Prince of Songkla Universuty, Thailand

Dear Payum,
To get the optimized structure of PTO you can start with optcell 2, ionmov 2 and distort the z- direction of the atomic displacements of Pb, Ti, O1,O2 and O3 i.e you need to break the symmetry. You can also use ecut 45 Hartrees as it works well for PTO.

I hope this could help you.
H. Sharma.

Dear Payum,
To break the symmetry just set nsym to 1.

Best wishes,
Maxim

Hello
I know this post is not new, but have you guys solved the problem? because I have the same problem with tetragonal perovskite structure. it is turning to cubic
any help plz?

Nawzad