[lammps-users] LAMMPS/cif files/ chemical disorder

LAMMPS LAMMPS Mailing List Mirror
1

Hello,

I am building a feldspar slab on LAMMPS. When I run my simulations I have NAN in pressure , total energy and potential energy. To get my initial configuration I am using a .cif file for feldspar-albite from the American Mineralogist Crystal Structure Database and I am using “Mercury Software” to convert this file to a xyz.file.

I figure out that my problem with NAN is coming from overlapping atoms. This XYZ file that I am getting has Silicon and Aluminum with the same coordinates. Looks like that the authors of the cif files do this to represent a chemical disorder (isomorphic substitution). The problem is that LAMMPS recognizes it as 2 different atoms in the same position. I would like to know how to address this kind of problem on LAMMPS .

Thank you
Isa __________________________________________________________________

cif file
data_global
chemical_name_mineral ‘Albite’
loop
_publ_author_name
‘Harlow G E’
_journal_name_full ‘American Mineralogist’
_journal_volume 67
_journal_year 1982
_journal_page_first 975
_journal_page_last 996
_publ_section_title
;
The anorthoclase structures: The effects of temperature and composition
Sample: Or = 32.5, T = 23 C
;
_database_code_amcsd 0019652
_chemical_compound_source ‘Grande Calderira, Azores’
_chemical_formula_sum ‘Na.686 K.334 Ca.008 (Al Si3) O8’
_cell_length_a 8.290
_cell_length_b 12.966
_cell_length_c 7.151
_cell_angle_alpha 91.18
_cell_angle_beta 116.31
_cell_angle_gamma 90.14
_cell_volume 688.819
_exptl_crystal_density_diffrn 2.588
symmetry_space_group_name_H-M ‘C -1’
loop
space_group_symop_operation_xyz
‘x,y,z’
‘1/2+x,1/2+y,z’
‘-x,-y,-z’
‘1/2-x,1/2-y,-z’
loop
_atom_site_label
_atom_site_fract_x
_atom_site_fract_y
_atom_site_fract_z
atom_site_occupancy
Na 0.27500 0.00130 0.13660 0.68600
K 0.27500 0.00130 0.13660 0.33400
Ca 0.27500 0.00130 0.13660 0.00800
Al1 0.00850 0.17530 0.22070 0.25000
Si1 0.00850 0.17530 0.22070 0.75000
Al2 0.00690 0.81800 0.22520 0.25000
Si2 0.00690 0.81800 0.22520 0.75000
Al3 0.69540 0.11400 0.33540 0.25000
Si3 0.69540 0.11400 0.33540 0.75000
Al4 0.69410 0.88150 0.34650 0.25000
Si4 0.69410 0.88150 0.34650 0.75000
O1 0.00160 0.13880 0.99470 1.00000
O2 0.60560 0.99720 0.28280 1.00000
O3 0.82330 0.12790 0.21800 1.00000
O4 0.82270 0.85740 0.23290 1.00000
O5 0.02330 0.30120 0.26180 1.00000
O6 0.02410 0.69160 0.24210 1.00000
O7 0.18970 0.12040 0.40040 1.00000
O8 0.18830 0.87250 0.41250 1.00000
loop
_atom_site_aniso_label
_atom_site_aniso_U_11
_atom_site_aniso_U_22
_atom_site_aniso_U_33
_atom_site_aniso_U_12
_atom_site_aniso_U_13
_atom_site_aniso_U_23
Na -0.00392 0.05533 0.02892 0.00098 -0.00169 -0.02188
K -0.00392 0.05533 0.02892 0.00098 -0.00169 -0.02188
Ca -0.00392 0.05533 0.02892 0.00098 -0.00169 -0.02188
Al1 0.01566 0.01447 0.01061 -0.00195 0.00675 0.00000
Si1 0.01566 0.01447 0.01061 -0.00195 0.00675 0.00000
Al2 0.01483 0.01277 0.01061 0.00342 0.00627 0.00084
Si2 0.01483 0.01277 0.01061 0.00342 0.00627 0.00084
Al3 0.01427 0.00936 0.01352 0.00000 0.00627 0.00084
Si3 0.01427 0.00936 0.01352 0.00000 0.00627 0.00084
Al4 0.01510 0.00936 0.01311 0.00146 0.00651 0.00042
Si4 0.01510 0.00936 0.01311 0.00146 0.00651 0.00042
O1 0.03468 0.02554 0.02018 0.00293 0.01472 0.00168
O2 0.02406 0.01532 0.02580 0.00049 0.00772 0.00084
O3 0.02657 0.03490 0.03287 -0.00488 0.01809 0.00210
O4 0.02629 0.03320 0.03142 0.00537 0.01713 -0.00084
O5 0.02462 0.01873 0.02538 -0.00146 0.01061 -0.00126
O6 0.02350 0.01702 0.02455 0.00146 0.00724 0.00168
O7 0.02517 0.02128 0.01810 0.00293 0.00555 0.00337
O8 0.02406 0.02298 0.01976 -0.00146 0.00531 -0.00295

2

Hello,

I am building a feldspar slab on LAMMPS. When I run my simulations I have NAN in pressure , total energy and potential energy. To get my initial configuration I am using a .cif file for feldspar-albite from the American Mineralogist Crystal Structure Database and I am using “Mercury Software” to convert this file to a xyz.file.

I figure out that my problem with NAN is coming from overlapping atoms. This XYZ file that I am getting has Silicon and Aluminum with the same coordinates. Looks like that the authors of the cif files do this to represent a chemical disorder (isomorphic substitution). The problem is that LAMMPS recognizes it as 2 different atoms in the same position. I would like to know how to address this kind of problem on LAMMPS .

you could assign each of these sites a special atom type (so they are more easily recognized individually) in addition to the actual atom types needed for the model. please note that it is quite OK to have multiple atom types pointing to the same type of atom. then you read in the data and replicate it or create a geometry with replicated unit cells in a data file and read it in. now you can use the set command with the type/fraction option to set a desired fraction to the first of the two alternatives for each location (and special atom type). then you can use the set command with the type option to switch the remainder. rinse and repeat for each “special” atom type. if your system is large enough, it should have the same total stoichiometry as the compound and just randomly distributed type assignments on the lattice positions that have multiple possible atom types and have those in total with the correct probability.

axel.

3

Hello Axel,
Thank you very much for your email.I still do not understand a couple of things that you said. What do you mean by multiple atom types pointing to the same type of atom? How exactly this type/fraction command is addressing the location? My problem is that in the data file that I am reading all Aluminum and Silicon have the same xyz position. I should have 1 Al to 3 Si with different coordinates (NaAlSi3O8)
I am following what you said and I am using the type/fraction command but I am still getting NAN.

This is how my input is:

pair_style lj/cut/coul/cut 12.0 12.0
pair_modify mix arithmetic
read_data feldspar.txt

define feldspar parameters

mass 3 16.000000
mass 4 28.085500
mass 1 26.981539
mass 2 22.989769

pair_coeff 3 3 0.1554000000 3.5532
pair_coeff 4 4 0.0000018405 3.7064
pair_coeff 1 1 0.0000018405 3.7064
pair_coeff 2 2 0.1301000000 2.6378

dump 1 all xyz 1000 feldspar_piston.xyz

thermo_style custom step temp pe etotal press vol density
thermo 1000

reset_timestep 0

fix 1 all wall/lj126 zhi EDGE 1.0 1.0 2.5 units box

fix 2 all wall/piston zlo pos -2 vel 0.001 units box

#Feldspar :NaAlSi3O8
set group all type/fraction 1 0.25 12393

set group all type/fraction 4 0.75 12393

#minimize 0.0001 0.000001 100 1000

run 100000

unfix 1
unfix 2

write_restart equil.restart
write_data data_water_feldspar.txt

4

You are having some more fundamental problems of understanding what you are doing than what I have time and energy to explain and teach you.

I can give you a pointer of what to do, but the rest is up to you and your adviser. After all it is the job of your adviser to tutor and advise you in what you are trying to do.

Sorry,
Axel