Question on losing atoms during minimization

Dear all,

Thank you so much for your advice and I have already tried my best to
simplify my input files once more. New code is listed as follows to describe
my question which emerges when I try to calculate the grain boundary energy
of a fcc Ni bicrystal. Before an error occurred, periodic boundary was set
for each of the 3 dimensions and a bicrystal with specific orientations was
constructed. Then this error occurs during the energy minimization after 34
time-steps in accompany with a message which shows that an atom was lost
during the minimization. Besides, this error occurs only when the code was
performed in a parallel run via a mpirun command on Linux, but not happen
when it runs on Windows via the lmp_win_no-mpi.exe for Windows. So my
question is why an atom can be lost during energy minimization in a box with
period boundaries in all dimensions, and if there is a solution to this
problem.
I will be so lucky and appreciate if you could give me some suggestions
about my question at your convenient.

Yours sincerely

My input file:
################################## Initiallize Simulation

Dear all,

Thank you so much for your advice and I have already tried my best to
simplify my input files once more. New code is listed as follows to describe
my question which emerges when I try to calculate the grain boundary energy
of a fcc Ni bicrystal. Before an error occurred, periodic boundary was set
for each of the 3 dimensions and a bicrystal with specific orientations was
constructed. Then this error occurs during the energy minimization after 34
time-steps in accompany with a message which shows that an atom was lost
during the minimization. Besides, this error occurs only when the code was
performed in a parallel run via a mpirun command on Linux, but not happen
when it runs on Windows via the lmp_win_no-mpi.exe for Windows. So my
question is why an atom can be lost during energy minimization in a box with
period boundaries in all dimensions, and if there is a solution to this
problem.
I will be so lucky and appreciate if you could give me some suggestions
about my question at your convenient.

this input does not lose any atoms for me, even if run in parallel.

axel.