# [lammps-users] The Neighbor Index of REBO potential of Carbon Nanotube

Dear Authors and users,

I am new to lammps. I tried to compare the calculated force and
potentials from lammps (turn-off LJ terms and Torsion terms) with my
own code of rebo (based on Brenner 2002) potential. I don't quite
understand the way lammps index the first nearest neighbor lists. For
example, I studied a (8,4) pristine CnT. So there are total 112 atoms
in one unit cell and each Carbon is bonded to three nearest
neighbors.

I get the total energy and temperature from lammps, but feel the
temperature is lower than I expected. I also printed out the neighbor
lists (table below). The 1st column indicates the "own" atom, and the next
three columns are the nearest neighbors. I have three questions based on
my output and print out. Hope some one could help me and answer the
questions.

1. I found there are some wired numbers in the neighbor list such as (221,
231, 156). Since only 112 atoms in one unit cell, I didn't define these
atoms. Where they come from?

2. After I print out the x,y,z coordinates of each atom, I found the atoms
index are not the same as what I defined in input file. Can anyone explain
how the neighbor index define or modified in codes?

3. The total Energy is 850 Kcal/mol. Should I interpret this value by
850Kcal/Avagadro as the total energy per atom?

own bond1 bond2 bond3
0 1 2 12
1 0 3 14
2 221 0 6
3 1 4 5
4 3 16 56
5 3 6 7
6 2 5 19
7 5 24 63
8 231 9 30
.... (short cut here)
109 97 99 110
110 109 111 156
111 105 107 110

Thanks you for any comments or suggestions.

bests

Haibin Chen Ph.D.
Mechanical Engineering Dept
Carnegie Mellon University
Pittsburgh, Pa, 15213

Answers below. Note that AIREBO with terms turned
off is not REBO. There are some differences.

Steve

Dear Authors and users,

``````I am new to lammps\. I tried to compare the calculated force and
``````

potentials from lammps (turn-off LJ terms and Torsion terms) with my
own code of rebo (based on Brenner 2002) potential. I don't quite
understand the way lammps index the first nearest neighbor lists. For
example, I studied a (8,4) pristine CnT. So there are total 112 atoms
in one unit cell and each Carbon is bonded to three nearest
neighbors.

I get the total energy and temperature from lammps, but feel the
temperature is lower than I expected. I also printed out the neighbor
lists (table below). The 1st column indicates the "own" atom, and the next
three columns are the nearest neighbors. I have three questions based on
my output and print out. Hope some one could help me and answer the
questions.

1. I found there are some wired numbers in the neighbor list such as (221,
231, 156). Since only 112 atoms in one unit cell, I didn't define these
atoms. Where they come from?

Ghost atoms.

2. After I print out the x,y,z coordinates of each atom, I found the atoms
index are not the same as what I defined in input file. Can anyone explain
how the neighbor index define or modified in codes?

The neighbor index is to a local atom, not a global atom ID. XYZ coords
are also modified across PBC for ghost atoms.

3. The total Energy is 850 Kcal/mol. Should I interpret this value by
850Kcal/Avagadro as the total energy per atom?

Depends on the "norm" factor in your themo output. It's either total or
per-atom. See the thermo_modify norm command.