CNT moving out of the simulation box (containing water) in its axial direction

Dear All,

I am trying to simulate the Kapitza resistance between SWCNT and water molecules at room condition. For that I am using 7200 molecules of water (density-1000 kg/m3) and (5,5) chirality SWCNT. I have used LJ and Coulomb potential for water (SPC-E model) and AIREBO with torsion ON for CNT.

I performed NVT followed with NPT and NVE with rescaling of CNT temperature to a higher value using fix rescale temp command. Apart form pressure, the system was equlibrated at constant temp, density and energy. I also performed minimization at the beginning.

In VMD, it was observed that the CNT moved out of the simulation box completely in the axial direction during NVT, NPT and NVE creating a cylindrical void in the simulation box. But even if the CNT was outside the box, radial breathing mode was still present showing that it was still interacting with the water molecules. Also, the water molecules did not occupy the void space.

What could be the reason for such behavior of CNT?

Also, when I switched off the torsion flag in the AIREBO potential, CNT did not move out of the box.

So, is there any problem with the potential field I am using or its only a matter of visualization?

Hoping for a positive and helpful suggestions. Thankyou.

Dear All,

I am trying to simulate the Kapitza resistance between SWCNT and water
molecules at room condition. For that I am using 7200 molecules of water
(density-1000 kg/m3) and (5,5) chirality SWCNT. I have used LJ and Coulomb
potential for water (SPC-E model) and AIREBO with torsion ON for CNT.

I performed NVT followed with NPT and NVE with rescaling of CNT temperature
to a higher value using fix rescale temp command. Apart form pressure, the
system was equlibrated at constant temp, density and energy. I also
performed minimization at the beginning.

In VMD, it was observed that the CNT moved out of the simulation box
completely in the axial direction during NVT, NPT and NVE creating a
cylindrical void in the simulation box. But even if the CNT was outside the
box, radial breathing mode was still present showing that it was still
interacting with the water molecules. Also, the water molecules did not
occupy the void space.

What could be the reason for such behavior of CNT?

because of periodic boundary conditions, the CNT does not really move
out of the box, but it simply moves into a periodic copy of the box
and a periodic copy of the CNT takes its place.
it is just normal diffusive behavior. it only is a concern, if there
is a significant bias observed over multiple simulations starting from
similar, but decorrelated initial configurations.

if you absolutely want to suppress it, you can use a weak tethering
potential via fix spring.

Also, when I switched off the torsion flag in the AIREBO potential, CNT did
not move out of the box.

again, only if there is a strong correlation between multiple
simulations started from decorrelated initial positions, you can draw
a conclusion from that (and start to worry).

So, is there any problem with the potential field I am using or its only a
matter of visualization?

neither.

axel.