Bond atom missing in box size check

LAMMPS LAMMPS Beginners
, , ,
1

Dear everyone,

I am trying to simulate cu oxidation with reaxff, when i try to relax my model i get the error "bond atom missing in box size check"after all the steps.I notice that all the o2 molecule just get apart into atoms ,i have changed some reaxff potential files but just the same error.Here is my files,can anyone help me?i’ve been troubled for 2 days :sob:

units real
dimension 3
boundary p p p
atom_style full
neighbor 0.5 bin
neigh_modify every 5 delay 0 check yes

pair_style reaxff lmp_control safezone 2.5 mincap 3000

region box block 0 200 0 200 0 200 units box
create_box 2 box bond/types 1 extra/bond/per/atom 1

mass 1 16.00
mass 2 63.55

region oxygen sphere 100 100 100 15 side out
molecule O2 O2.data
create_atoms 0 random 550 20010 oxygen mol O2 25367 overlap 1.5

region cunp sphere 100 100 100 13 side in
lattice fcc 3.61
create_atoms 2 region cunp

pair_coeff * * CuOH.ff O Cu

#cu relax

fix 1 all qeq/reax 10 0.0 10.0 1e-6 reax/c

minimize 1.0e-6 1.0e-8 1000 10000

timestep 0.3
fix relax all npt temp 300.0 300.0 100.0 iso 1.0 1.0 1000 drag 0.1

thermo 100
thermo_style custom step temp pe
run 150000

2

Are you defining bonds in the O2.data file?

If the energy interactions in your system are described solely by ReaxFF then you should not use the bond/types 1 extra/bond/per/atom 1 options in the create_box command.

3

yes，here is my O2.data：
2 atoms
1 bonds

Coords

1 -0.0576 0.0000 0.0000
2 1.1584 0.0000 0.0000

Types

1 1
2 1

Bonds

1 1 1 2

β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”
but if i delete "bond/types 1 extra/bond/per/atom 1"options i will get this error :Invalid bond type in Bonds section of molecule file: 1 1 1 2 (…/molecule.cpp:1044) .How should i modify my commands?Thanks so much for you reply!

4

From a technical perspective, delete the lines

1 bonds
Bonds
1 1 1 2

in the O2.data file.

From a physical perspective, check that atoms are not very close to its other after creation, which can lead to high forces thus high accelerations and atoms flying too fast in the simulation box.

HTH

5

Thank you so much for your advice.
I tried your method and the O2 molecules are not getting apart during relaxing now.But when simulating oxidation, oxygen molecules still maintain their molecular structure and react on the copper surface, which should have get apart into atoms and reacted with copper. That is to say, two oxygen atoms are always together, and they should break down into atoms during the reaction. Can you tell me any modification methods? Sorry to trouble you so much :sob: :sob:

6

This sounds more like you need to spend some time picking up a text book on physical chemistry and learn some more about reaction kinetics. Most chemical reactions require some activation, happen in multiple individual steps, rarely achieve 100% completion, and are β€œslow” compared to the time scales available to classical MD simulations.

Besides, has the ReaxFF parameterization you are using been created for those reactions?

7

Sorry I cannot assist you with this issue!