Problem MD simulation LAMMPS with ReaxFF

Dear lammps User’s,
i have a problem using ReaxFF.
I want to simulate a box (35AngstromX35AngstromX110Angstrom) with a supercell of Tricalcium Aluminate + water molecules + ions Ca2+ + ions Cl- + small polymer.
LAMMPS script is this:

clear

units real

dimension 3
boundary p p p
atom_style charge
neighbor 2.0 bin
neigh_modify delay 2 every 1 check yes
read_data ./common/box_PCE_wat_Ca_C3A.data
pair_style reax
pair_coeff * * ffield.reax 6 1 4 9 2 10 3
timestep 0.5
velocity all create 300.0 9999 #dist gaussian
thermo_style custom step etotal pe ke temp press vol
thermo 1
dump 1 all custom 10 ./dcd/box_PCE_C3A_wat_Ca.xyz.* id element q xu yu zu
dump_modify 1 sort id element Al C Ca Cl H Na O
fix 1 all nvt temp 300.0 300.0 100.0
run 200000
print “All done!”
print “MD Equilibration of the system is completed”

All calculations are done on computer clusters.
LAMMPS shows these messages:

LAMMPS (11 Jan 2012)
Reading data file …
orthogonal box = (0 0 0) to (35 35 110)
2 by 2 by 4 MPI processor grid
12441 atoms
Setting up run …

Hi Nicola,

Without more information, such as full input script, ffield.reax and
data files, nobody can tell what went wrong. You likely ran out of
physical memory as pair_style reax is quite memory intensive.

Ray

All calculations are done on computer clusters.

It is unusual for LAMMPS not to print an error message. Are you
running in batch mode? If so, can you run interactively, e.g.
on your desktop or on the cluster? Is there any screen output or logfile
output that has an error message?

Steve