Hello everyone,
I just wonder can we calculate potential energy of the system without consider one pair style or one group of atoms? Just like the method mentioned in: Jiang, J.-W.; Park, H. S., Mechanical Properties of Mos2/Graphene Heterostructures. Applied Physics Letters 2014, 105, 033108.
The author calculate the PE of graphene/MoS2 system without considering the LJ pair style between the two layers.
Any answer or suggestion will be appreciated.
Regards,
Qingchang
Hello everyone,
I just wonder can we calculate potential energy of the system without
consider one pair style or one group of atoms? Just like the method
mentioned in: Jiang, J.-W.; Park, H. S., Mechanical Properties of
Mos2/Graphene Heterostructures. Applied Physics Letters 2014, 105, 033108.
The author calculate the PE of graphene/MoS2 system without considering
the LJ pair style between the two layers.
​there are multiple ways to do this. LAMMPS allows you to remove entries
between groups of atoms or types of atoms from the neighbor list. entries
that are excluded from the neighbor list cannot contribute to the potential
energy. please see the neigh_modify exclude command for details. or in a
more general way, you can just define those atoms with additional atom
types and set the interaction​ energy (epsilon) to zero for pairs of types
that you don't want to interact. this will work with almost any MD code.
axel.
Thanks to all.
I understand that cutting off the interaction between them will work.
But I just wonder if I can avoid the pair style when I calculate PE but keep
their interaction. Actually I want to dig out the PE contributed by other parts
although it seems to make less physical sence.
Best.
Here is what the paper says:
"All MD simulations in this work were performed using the publicly
available simulation code LAMMPS ...
The MoS2 interatomic interactions are described by a recently
developed Stillinger-Weber potential, while the carbon-carbon
interactions are described by the second generation Brenner (REBO-II)
potential. The MoS2 and graphene layers in the GMG heterostructure are
coupled by van der Waals interactions, which are described by the
Lennard-Jones potential. The energy and distance parameters in the
Lennard-Jones potential are 3.95 meV and 3.625 A, while the cutoff is
10.0 A.
"
This description is a little incomplete, but it sounds like a good use
case for pair style hybrid, combining their Stillinger-Weber
potential (Mo,S) with REBO (C) and LJ. It is not clear which atom
pairs the LJ interaction is applied to. It might be C-S and C-Mo, or
it might be just C-S. The LAMMPS commands would look something like:
pair_style hybrid lj/cut 10.0 sw rebo
pair_coeff * * sw MoS.sw Mo S NULL
pair_coeff * * rebo CH.airebo NULL NULL C
pair_coeff 1 3 lj/cut 3.95 3.625
pair_coeff 2 3 lj/cut 3.95 3.625
Aidan
See the compute pair command. One of its uses is with
pair hybrid (or overlay), to extract the pair contributions
from specific sub-styles.
Steve