CELL_OPT to obtain lattice parameters close to experimental values.

1,048 views

Skip to first unread message

ganta.pra...@gmail.com

unread,

Sep 21, 2017, 5:02:59 AM9/21/17

to cp...@googlegroups.com

Hi all,

I am trying to do cell optimization of a goethite FeOOH unit cell. The lattice parameters of the considered unit cell are a=9.91340 b=3.0128, c=4.58 ( the same values I used in cp2k).

I have tested all three cases (direct cell opt, geo opt, MD) of the CELL_OPT and

The cell optimized values are approximately a=9.33, b=3.063 c=4.179.

The experiment values of lattice constants are a=9.96 b=3.02 c=4.62.

I want to obtain the optimized values close to expt. values.

In trail and error, I have used STESS_TENSOR NUMERICAL, I have disabled D3 correction and I have also used zero pressure.

Below is my input file.

&Global

PROJECT goethite_cellopt

RUN_TYPE CELL_OPT

PRINT_LEVEL LOW

&END GLOBAL

&FORCE_EVAL

METHOD Quickstep

STRESS_TENSOR ANALYTICAL

&DFT

BASIS_SET_FILE_NAME BASIS_MOLOPT

POTENTIAL_FILE_NAME GTH_POTENTIALS

CHARGE 0

MULTIPLICITY 1

&MGRID

NGRIDS 5

CUTOFF 600

REL_CUTOFF 60

&END

&QS

METHOD GPW

EPS_DEFAULT 1.0E-12

EXTRAPOLATION ASPC

EXTRAPOLATION_ORDER 4

&END

&SCF

SCF_GUESS ATOMIC

MAX_SCF 1500

EPS_SCF 1.0E-8

&OUTER_SCF

MAX_SCF 30

EPS_SCF 1.0E-8

&END

ADDED_MOS 30

CHOLESKY INVERSE

&SMEAR ON

METHOD FERMI_DIRAC

ELECTRONIC_TEMPERATURE [K] 300

&END SMEAR

&DIAGONALIZATION

ALGORITHM STANDARD

&END DIAGONALIZATION

&MIXING

METHOD BROYDEN_MIXING

ALPHA 0.1

BETA 1.5

NBROYDEN 8

&END MIXING

&END SCF

&XC

&XC_FUNCTIONAL PBE

&END XC_FUNCTIONAL

&VDW_POTENTIAL

POTENTIAL_TYPE PAIR_POTENTIAL

&PAIR_POTENTIAL

PARAMETER_FILE_NAME dftd3.dat

TYPE DFTD3

REFERENCE_FUNCTIONAL PBE

R_CUTOFF [angstrom] 15

&END

&END VDW_POTENTIAL

&END XC

&END DFT

&SUBSYS

&CELL

ABC [angstrom] 9.91340 3.01280 4.58000

PERIODIC XYZ

&END CELL

&TOPOLOGY

COORD_FILE_NAME goethite_unitcell.xyz

COORD_FILE_FORMAT XYZ

&CENTER_COORDINATES T

&END CENTER_COORDINATES

&END

&KIND H

BASIS_SET DZVP-MOLOPT-SR-GTH

POTENTIAL GTH-PBE-q1

&END KIND &KIND Fe

BASIS_SET DZVP-MOLOPT-SR-GTH

POTENTIAL GTH-PBE-q16

&END KIND

&KIND O

BASIS_SET DZVP-MOLOPT-SR-GTH

POTENTIAL GTH-PBE-q6

&END KIND

&END SUBSYS

&END FORCE_EVAL

&MOTION

&CELL_OPT

OPTIMIZER BFGS

MAX_ITER 5000

KEEP_ANGLES T

EXTERNAL_PRESSURE 1.01325 0 0 0 1.01325 0 0 0 1.01325

&END

Goethite unit cell

Fe H O2

Fe 1.446365 0.753200 4.357412

Fe 8.467034 2.259600 0.222588

Fe 3.510335 2.259600 2.067412

Fe 6.403065 0.753200 2.512588

H 8.912147 0.753200 2.752580

H 1.001253 2.259600 1.827420

H 5.957953 2.259600 0.462580

H 3.955446 0.753200 4.117420

O 7.940633 0.753200 1.305300

O 1.972767 2.259600 3.274700

O 6.929466 2.259600 3.595300

O 2.983933 0.753200 0.984700

O 9.400877 0.753200 3.682320

O 0.512523 2.259600 0.897680

O 5.469223 2.259600 1.392320

O 4.444177 0.753200 3.187680

Please let me know if you have any suggestions.

Goethite_unitcell.xyz

Vladimir Rybkin

unread,

Sep 22, 2017, 5:27:42 AM9/22/17

to cp2k

Dear ... ,

here is a couple of comments:
1) the results are not that bad bearing in mind all approximations made in the PBE functional;
2) MOST IMPORTANTLY: one needs to converge the the calculations with respect to k-points. In practice, as you are performing a Gamma-point calculation, you should use a supercell, i.e. replicate your unit cell at least twice in b and c directions.
3) Why are you using smearing: is it a metal?
4) MOLOPT basis sets are better suited to molecular calculations. Try GTH basis sets.
5) GEO_OPT does not optimize the cell at all, that's not an option.
6) I don't think you need external pressure as soon as you get a reasonable large cell.

Yours,

Vladimir

четверг, 21 сентября 2017 г., 11:02:59 UTC+2 пользователь ganta.pra...@gmail.com написал:

Reply all

Reply to author

Forward

0 new messages