TERSOFF PARTICLE IN WATER

LAMMPS LAMMPS Mailing List Mirror
1

Hello all users,

I’m trying to solve for BN particle in water so this is the numbering of my
atoms:

1-O
2-H
3-B
4-N

So I’m using this pair_coeff commands to use SPC/E for water, Tersoff for B
and N, and Lennard-Jones for cross relations. I carefully read the manual
about how to use pair_style hybrid command but the results are not as
expected, therefore I’m trying to understand the problem if it is in
potential definition or in post-processing. Do you think is there a problem
with these force field definitions? I understand that it should be
overwritten if we define the same type of atoms more than once in pair_coeff
so I guess once the tersoff is defined for B and N, water parameters and
cross relations should get the new values in the following pair_coeff
commands right? I will appreciate if you give me a couple of suggestions if
I am doing something wrong or how can I get better force field definition
for this problem? Thank you for your attention.

pair_style hybrid lj/cut/coul/long tersoff coul/long 9.0
pair_coeff * * tersoff slotmant.tersoff NULL NULL B N
pair_coeff 1 1 lj/cut/coul/long 0.1533 3.166
pair_coeff 2 2 lj/cut/coul/long 0 0
pair_coeff 1 2 none

this one is *wrong*. you may not have lj interactions between H and O,
but you have to have coulomb.

you could have either:

pair_coeff 1 1 lj/cut/coul/long 0.1533 3.166
pair_coeff 1 2 coul/long
pair_coeff 2 2 coul/long

or:

pair_coeff 1 1 lj/cut/coul/long 0.1533 3.166
pair_coeff 1*2 2 lj/cut/coul/long 0 0

pair_coeff 1 3 lj/cut/coul/long 0.12 3.31
pair_coeff 1 4 lj/cut/coul/long 0.149 3.414
pair_coeff 2 3 none
pair_coeff 2 4 none

i don't understand this segment either. do you have coulomb
interaction between O and B or N? or only lj? and how about the
charges on H atoms?

axel.

2

Since B and N are not charged atoms as I specified in my data file, I should
not have coulombic interactions involving B and N atoms and only LJ should
be employed between O-B and 0-N. I expect from lj/cut/coul/long to do this
as I assigned the charges of B and N as 0 already. So I did not need to
define anything between H-B and H-N since there is no LJ between them in
addition the absence of Coulomb.

I corrected the mistake between H and O but there is no significant change
son I think the actual problem is the interactions between O-B and O-N
because I get more consistent results without those cross-relations(O-B,
O-N). But this is still not acceptable as we expect, because potential
between 0-B and O-N must be important.

May be LJ parameters are not appropriate for simulating B-N what do you
think? I had to try this because there is no way to combine Tersoff and LJ
for cross relations, right? Thank you for your response Dr. Kohlmeyer.