inquiry about bonds in a molecule

LAMMPS LAMMPS Mailing List Mirror
#1

Dear all,
I made MoDTC molecule with Avogadro software and when I minimize it some of the atoms bonds break and they create another bond with other type of the atoms, which they shouldnt make bonds with other type of atoms in that molecule. I want the exact initial structure bond to be hold.
I was wondering where does this problem can come from?
can it come from the reaxFF potential?
here is part of my minimization code.

echo both
units real
atom_style full
dimension 3
boundary p p p

read_data x.data

pair_style reax/c NULL checkqeq yes
pair_coeff * * CHOSMoNi.reax C H O S Mo

neighbor 2.0 bin #2 because the default parameter for skin is 2 in real units
neigh_modify every 1 delay 0 check yes
fix charge all qeq/reax 1 0.0 10.0 1e-6 reax/c
timestep 0.1

#---------------------ReaxFF Minimization---------------------------------
thermo 10
thermo_style custom step temp pe ke density press pxx pyy pzz lx ly lz

fix minconn all reax/c/bonds 100 emin_bonds.reaxc
dump 2 all custom 10 dumpsysminimize2.lammpstrj id type x y z
min_style cg
#fix 4 all box/relax iso 0.0 vmax 0.001
minimize 1.0e-4 1.0e-6 100 1000

unfix minconn

I highly appreciate any help.
Best,
Mina

#2

The answer is: it depends

There are two common problems that happened with people using ReaxFF on Molecular systems.

  1. the ReaxFF potential file was not parameterized for the kind of compound and conditions that it is being used for and thus does not provide sufficiently accurate forces
  2. the data file incorrectly includes topology information leading to pair wise interactions being excluded. In ReaxFF all topology (bonds, angles, dihedrals) is implicitly and computed as part of the pair style.

Axel