Cancelling Molecule Interactions

Hello,

So in my project, I was able to create MOFs and insert DMMP molecules into the file successfully. Now, to make sure neighboring DMMP molecules do not interfere with one another, I wish to turn off DMMP-DMMP interactions. Is there a right way to go about this? I have read the neigh_modify command may help, but that is for atoms not molecules. I feel like the numbers in the pair_coeff command may be of use too, but I wanted to ask you just in case.

Best,
Dorian Thompson

LAMMPS normally decides how a pair of particles interact entirely due
to their type. There are two ways that come to mind:

1) You could create a list of "pair_coeff" commands which turn off
interactions between any atom type which would normally appear in a
DMMP molecule, with any other atom type (which would also appear in a
DMMP molecule). If their are N atom types in DMMP, this will require
N*(N-1)/2 different pair_coeff commands. (In each "pair_coeff i j"
command, make sure that i<=j.)
https://lammps.sandia.gov/doc/pair_coeff.html

2) However, probably the easiest way to do this would be to assign
these N atom types to their own group (eg "gDMMP") and use the
"neigh_modify exclude group gDMMP gDMMP" command.
https://lammps.sandia.gov/doc/neigh_modify.html
https://lammps.sandia.gov/doc/group.html
(That way you don't have to generate a long list of pair_coeff commands.)

Either way, if some of these atom types also appear elsewhere in your
simulation (in other molecules, for example), then you will have to
define two different versions of these atoms: one version for the atom
as it appears in DMMP, and another version as it appears everywhere
else.

Cheers

Andrew

There are also neigh_modify exclude options for atoms
in the same or different molecule.

Steve

To canel using pair_coeff command, which numbers do I set to zero? I sell all the numbers that were not atom types to zero, and the run failed, most likely due to the dynamics.

Best,
Dorian Thompson

That depends on your pair style. Just look at the functional forms in the docs. For lj/cut it has to be epsilon that needs to become zero. For any coulomb based pair style this method is not suitable, since there are no parameters to be set with pair_coeff.
Axel

To canel using pair_coeff command, which numbers do I set to zero?

That depends on which pair_style you are using. If you are using one of the many Lennard-Jones based pair styles (beginning with “pair_style lj/…”), then it is enough to set the “epsilon” parameter to zero for that pair of atoms. The syntax for doing this also depends on the pair_style. For example, see
https://lammps.sandia.gov/doc/pair_lj.html

for other pair_styles, see:
https://lammps.sandia.gov/doc/pairs.html

I sell all the numbers that were not atom types to zero, and the run failed, most likely due to the dynamics.

That could be for many reasons. It’s difficult to diagnose the reason. Turning off interactions between pairs of atoms will always have a sudden and dramatic effect on the local pressure of the system. Perhaps you should try using the minimize command before continuing?