Lammps reactive force field not working

LAMMPS LAMMPS General Discussion
1

Respected Lammps Users, I am using lammps 2023 mpi on windows 10. I am trying to execute the reactive wetting with localized chemical reaction at the interface of iron and diamond with an intermediate oxygen in between the two. After I run the simulation using the following script; there is a sudden expansion and spreading of diamond crystal while the iron just outer booundary is expanded and moves towars atoms of oxygen. There is no clear attachement or bonding to for oxides of both elements and it looks no proper reaction and wetting occured. Also even after applying temperature control temperature surges to very hig values. I created data file using material studio for oxygen and then called it in lammps script along with adding iron and diamond the way the lammps script is written. The input script along with all files are as under. Kindly assist. May the diamond also need airebo and iron need eam or just reactive filed is enough. Is it possible that the problem originated due to creating data file and merge into lammps script using read_data? thanks

 log log.mreact

set potential function and parameters

units metal
dimension 3
boundary p p p
atom_style charge

#neighbor 0.5 bin
#neigh_modify delay 5

neighbor 3.0 bin
neigh_modify delay 0 every 1 check yes

region simbox block -35 128 -35 150 -39 145 units box
create_box 3 simbox

lattice custom 3.57 &
a1 1.0 0.0 0.0 &
a2 0.0 1.0 0.0 &
a3 0.0 0.0 1.0 &
basis 0 0 0 &
basis 0.5 0.5 0 &
basis 0 0.5 0.5 &
basis 0.5 0 0.5 &
basis 0.25 0.25 0.25 &
basis 0.75 0.25 0.75 &
basis 0.75 0.75 0.25 &
basis 0.25 0.75 0.75

region diamondslab block -25 118 -30 130 -20 11 side in units box
create_atoms 1 region diamondslab

lattice custom 3.5 &
a1 1.0 0.0 0.0 &
a2 0.0 1.0 0.0 &
a3 0.0 0.0 1.0 &
basis 0 0 0 &
basis 0.5 0.5 0 &
basis 0 0.5 0.5 &
basis 0.5 0 0.5

region ironcube block 15 74 20 75 20 70 side in units box

create_atoms 2 region ironcube

read_data FeOC.data add append shift 0 0 12

region oxygenslab block -5 98 -50 105 12 19 side in units box

group slab region diamondslab
group iron region ironcube
group oxygen region oxygenslab

set group slab type 1
set group iron type 2
set group oxygen type 3

mass 1 12
mass 2 55.7
mass 3 15.9994

write_data data.complreact

pair_style reax/c lmp_control

pair_coeff * * ffield.reax.Fe_O_C_H C Fe O

fix 7 all qeq/reaxff 1 0.0 10.0 1e-6 param.qeq

#fix 8 all qeq/reaxff 1 0.0 10.0 1.0e-6 reax/c

thermo 1
thermo_style custom step press pxx pyy pzz pe temp
#thermo_modify norm yes

#minimize 1.0e-21 1.0e-21 10000 10000
timestep 0.001
#fix 1 iron nvt temp 1100 1100 1
#velocity iron create 1100 12345 mom no rot no

#velocity iron create 1100 1234567654321 mom yes rot yes dist gaussian loop geom
#run 0
#velocity iron scale 1100

#velocity oxygen create 273 12345 mom no rot no
#velocity oxygen create 273 1234567654321 mom yes rot yes dist gaussian loop geom
#run 0
#velocity oxygen scale 273

#velocity slab create 1100 12345 mom no rot no
#velocity slab create 1100 1234567654321 mom yes rot yes dist gaussian loop geom
#run 0
#velocity slab scale 1100

#fix 1 all nvt temp 1100 1100 1
#velocity all create 1100 1234567654321 mom yes rot yes dist gaussian loop geom
#run 0
#velocity all scale 1100

fix 2 iron nve
#fix 3 oxygen nve
fix 4 iron temp/berendsen 1000.0 1000.0 100.0
#fix 5 oxygen temp/berendsen 1000.0 1000.0 100.0

fix 2 all nve
fix 5 all temp/berendsen 1000.0 1000.0 100.0

timestep 0.001
dump movie all atom 1 m.react

thermo_modify lost ignore flush yes
run 1100000
undump movie
write_restart restart.reaction ```

2

Shouldn’t you use units real with ReaxFF?

3

Sir, I just have seen and applied the units real with timestep 0.2 fs rather than 0.001 fs and rerun the script but the same higher temperature surges from 0 to several millions occurs which again randomly spread diamond crystal while the iron just outer layer is errupted and moved outwards and the central atoms remains static without any vibratory or random motion. I used the berendsen or nvt fix along with velocity seed but the same happened.

In addition as I had made the oxygen.data file using material studio only type id 1 can be assigned as just one element oxygen to make it useable and then called in the lammps script using read_data. The other two types including iron and diamond are written adding atom basis in lammps script. Now there are total three types of atoms where the oxygen data file is put into the creating the new region oxygen slab in lammps script around the defined xlo xhi as per data file otherwis just two atom types are displayed. type 1 oxygen and again type 1 silver and type 2 diamond. so I have created 3 simbox and one region around oxygen then group it as oxygen then type three. Is this way right or there may be a trouble here it to interfere the whole simulation?

Is there any other way rather than definig a region to assign oxygen as a new atom type?

thanks

4

I understand the words you are using, but still cannot make much sense of what you are trying to tell us. This mostly looks like it is a problem of setting up the simulation as a whole and not a LAMMPS issue. That is outside the scope of this forum and very specific to the research you are trying to do. Thus all I can suggest is to just do what is generally suggested when simulations of complex system misbehave: go back and test each component of your system for correctness and that the force field you are using can reproduce relevant data. E.g. that you are not using a ReaxFF parameter set generated for studying combustion reactions for regular condensed matter simulations and so on.

5

Sir, If I made these files to run for eam and airebo potential but replacing oxygen by copper data file as its gemoetry; oxygendatafile, is not suitable for style charge and real units and gives error. when lammps script is run with metal units and atomic style then it will work properly. This may or may not suggest troubling with how to apply reaxff or after applying need some additional command to generate menaingful results without temperature surge.

6

It mostly suggest that you need more training and did not understand the recommendation in my previous post. I have nothing more to say.

As already stated, this is not a LAMMPS problem but a PEBCAC situation.