Lost atoms in lammps

Dear Lammps users,
I’m using lammps version 23-May-2014. I’m trying to make the relaxation of the atomic positions of the Lizardite using CLAY-FF as a force-field.

Such relaxation ends without failure, but when I look at the geometry obtained I find that the oxygen and hydrogen atoms of the hydroxyl groups are overlapping.
I have made a NVE calculation (whitout minimization) and lammps throws an error.
ERROR: Lost Atoms: Original 18 current 12
I don’t know what is wrong.
what is the reason for this error?

Thanks in advance

Rubén Soria Martínez

Institut des Sciences de la Terre (ISTerre) / Géochimie
Université Joseph Fourier (UJF)
FRANCE

lammps.inp (4.86 KB)

When I run your example, I get crazy values for the potential energy, -3.2736804e+15 . I think there is probably something wrong with the force field you are using?