[lammps-users] Is there the way to create particle randomly consecutively?

Dear All,

I want to simulate interaction of oxygen molecule and water molecule with Silicon surface. To do so, I create a water molecule/oxygen molecule just above silicon surface. For long time kinetic I should create many water/O2 molecule randomly and consecutively.

First,I tried to use “displace_atoms” command, following command “create”, with argument “random”, but it turns out to move each atom of molecule not move molecule it self. While with argument “move”, it move the molecule but not randomly. Both command also produce what I did not expect , the particles which move are not the newest particle which is just created rather other older particles.

–part of my input file--------------------------

There is no way to create a molecule, with atoms
bonded to each other, as opposed to individual atoms,
which is what you are doing with create_atoms. The
only way to read in a molecule is in a data file, where
you can also specify the bonds between atoms. So I
suggest you create your molecules one at a time,
from a data file, by reading a config (with a new molecule),
running for a while, writing out a restart file, convert to
a data file, add a new molecule, restart, etc.

Then the randomizing you want can be done outside
of LAMMPS.

Steve

Right now I am using reaxFF potential, for simulate molecule water should I use TIP3P/TIP4P/SPC model with Lennard-Jones potential?. Could I keep using reaxFF?

I don't know anything about modeling water with ReaxFF. If
you can, then you do not need to specify the bond topology of
the water molecule as in my earlier email. So you should
be able to do it by creating 3 atoms (one by one) in the correct
configuration.

Steve