Fixing charge of a particle type in ReaxFF

[MISH]

Hi everyone,

Is it possible to apply a constraint in the charge equilibration process in ReaxFF simulation so that the charge on a particular particle type stays constant? This is very easily done in the serial ReaxFF code (using the MOLCHARGE keyword) but I don’t think it is available as part of pair_style reaxff command, or is there a way around that?

A similar questions was asked 11 years ago and the answer at that time was no.

how to fix a charge in ReaxFF simulation

Dear all lammps user’s: i would ask if it is possible to fix the charge for a given atom when we use ReaxFF potential. i have tried to do this by playing on the fix qeq/reax but this dont work. in the manual i read that this is possible with comb potential. so i would ask if there is some manner to do this with reax/c best regards

Thanks,
MH

[akohlmey]

The functionality of fix qeq/reaxff honoring the group name of the fix command and applying the charge equilibration only to atoms within the fix group has since been implemented and thus you can have charges unchanged by defining a group that doe not include those atom types and then use that group instead of “all” when defining you qeq/reaxff fix.

[MISH]

Thanks very much Axel,

this does work but I noticed that when a type is excluded from qeq/reaxff it is not integrated in time at all. The atoms stay where they are.

I attached my input files to this message. I ran it once with qeq/reaxff operating on all atoms and once where it was only acting on molecular species. The atoms in the second case did not move at all. I can’t think of a physical reason for that.

I put all my input files in this link:

ReaxFF input files

I am wondering if you observe the same thing.

Thanks,
MH

[akohlmey]

@MISH

but I noticed that when a type is excluded from qeq/reaxff it is not integrated in time at all. The atoms stay where they are.

This must be due to some unrelated cause. There is absolutely no connection in the code between fix qeq/reaxff and time integration.

[akohlmey]

@MISH

I am wondering if you observe the same thing.

I don’t.

[MISH]

Thanks,

I will update my version of LAMMPS.

MH