[gmx-users] Alcohol molecules moving from water/vacuum interface

[William Foskett]

Hello,

I am completing simulations of 1-butanol and isobutanol molecules at a
vacuum/water interface for a bachelor final year research project. I have
set up a box of molecules with two vacuum/water interfaces (at either end
of the box on the xy plane) the following commands (example of 5 isobutanol
molecules per interface in a 3 x 3 x 10 box)

*genbox -ci isobutanol.gro -nmol 5  -box 3 3 1 -o layer.gro -p topol.top*

*editconf -f layer.gro -o layer2.g96 -box 3 3 2 -center 1.5 1.5 1*


*genconf -f layer2.g96 -o 2layers.gro -nbox 1 1 2 -dist 0 0 6*


genbox -cp 2layers.gro -cs spc216.gro -o box1.g96 -p topol.top -box 3 3 10


And then proceeded to run energy minimisation,a 20ps NVT run, a 40ps NPT
and finally a 20ns NVT run following compacting of the box based on a *gmxcheck
-c NPT_40ps.g96 *


The problem I am having is that alcohol molecules continually jump between
the interfaces. For examples, if I started with 5 molecules at one
interface and 5 at the other, I may end up with 6 and 4 at the end of the
series of simulations. This generally occurs in the NVT or NPT runs. The
isobutanol molecules seem to behave appropriately as surfactants, and
therefore I cannot isolate why they are moving between interfaces.


Is there any way that I can force the molecules to maintain their original
equal split without molecules jumping between interfaces?


If you require more information please let me know.


Thank you!