Lammps NEB with ReaxFF


is there a way to restrain atoms in NEB and use the final data ? I have a system with water molecules and i am getting debonded water molecules in my NEB calculation. The second image is already a debonded hydrogen from a molecule and i do not know how to fix this

Attached you can find 16 replicas at step 0. The water molecules are atoms 6 and 7 (Oxygen and hydrogen). As you can see in the second replica i have already a debonded Hydrogen.
dump.neb.0.1 (37.6 KB)
dump.neb.0.2 (37.6 KB)
dump.neb.0.3 (37.6 KB)
dump.neb.0.4 (37.6 KB)
dump.neb.0.5 (37.6 KB)
dump.neb.0.6 (37.6 KB)
dump.neb.0.7 (37.6 KB)
dump.neb.0.8 (37.6 KB)
dump.neb.0.9 (37.6 KB)
dump.neb.0.10 (37.6 KB)
dump.neb.0.11 (37.6 KB)
dump.neb.0.12 (37.6 KB)
Uploading: dump.neb.0.13…
Uploading: dump.neb.0.14…
dump.neb.0.15 (37.5 KB)

Two comments:

  • The initial geometry is (mostly) irrelevant for this. What matters is the result.
  • LAMMPS will generate a linear interpolation between beginning and end structure. If that is not what you want, you have to provide suitable intermediate geometries and then LAMMPS will interpolate between beginning, end and those.

That means if i do not have intermediate geometries these two states are physically not relevant or the transition from A to B is most likely impossible ?

Sorry, but I don’t have the time to fill in for your adviser / tutor. I have explained the LAMMPS side of things. The rest is up to you and the people who advise and tutor you.

you already helped me. thanks a lot