I’m trying to implement NEB method to an alkali-ion conducting solid. The system is closely packed in the sense that all the crystallographic sites for alkali-ions are occupied. I’ve taken a single alkali ion and it’s position as the initial configuration and the nearest neighbor position as the final state.
The problem I am facing is that the final state position is already occupied with another alkali ion. And maybe that’s why I’m getting the following error message: “Pair distance < table inner cutoff” because my potential table won’t let two alkali-ion come closer than a certain distance (for my case the final replica and the original alkali-ion).
I tried by removing the existing alkali-ion at the final position by some measures (maintaining the charge neutrality), but that appeared to be problematic.
As a matter of fact, the neb is converging in a crooked way.
It will be really helpful if someone could shed some light on the problem.
P.S - This query is perhaps very much ‘system-specific’ than about lammps itself. So pardon me.
The final state conformation you give to NEB is supposed
to be a relaxed state with one or more atoms in new
positions relative to the initial state. Have you fully
minimized both your initial and final state before
starting the NEB calc?
I have taken a single mobile-ion as the neb atom for this case and for the final state I’m giving the coordinate of it’s nearest neighbor where it should jump.
In order to do so, I had to remove the nearest neighbor and add it’s charge to another alkali ion to maintain the charge neutrality of the whole system.
But most likely this very act is disturbing the system’s relaxation as a whole. I have used the ‘minimization’ before entering into the neb calculations though.
I’m attaching the input files for your reference. Please let me know if I am missing something.
read_NZP4.dat (81 KB)
final.hop1 (32 Bytes)
in.nasx2 (878 Bytes)
I have taken a single mobile-ion as the neb atom for this case and for the final state I'm giving the coordinate of it's nearest neighbor where it should jump.
In order to do so, I had to remove the nearest neighbor and add it's charge to another alkali ion to maintain the charge neutrality of the whole system.
what you describe is not a NEB calculation. it is 100% bogus. you
cannot change the number of atoms and/or arbitrarily change the charge
of atoms (as that changes their chemical nature).
But most likely this very act is disturbing the system's relaxation as a whole. I have used the 'minimization' before entering into the neb calculations though.
I'm attaching the input files for your reference. Please let me know if I am missing something.
you are missing the basic assumption of a NEB calculation: you are
trying to reconstruct the path of lowest energy between two different
geometries of the same system. if this will imply a change in the
chemistry, e.g. breaking or transferring of a bond, then this has to
happen due to the method of how the individual energies are computed.
in the case of a chemical change, that requires a model, that includes
such changes in the energy function (e.g. using a reactive force field
or a quantum chemical calculation). under no circumstances, can you
remove or add an atom.