Sodium Chloride solution Simulation

Dear LAMMPS users,

I prepared a data file for sodium chloride solution with mole template. I defined potential of water and ions with pair_style lj/charmm/coul/long 8.0 10.0. Although I minimized system, some bond of water molecules are missing at the beginning. Any suggestions?

echo both
units metal
atom_style full

bond_style harmonic
angle_style charmm
read_restart min
neigh_modify delay 0 every 1 check yes

------------------------------- group -------------------------------

group oxygen type 1

group hydrogen type 2
group water type 1 2
group Na type 3
group Cl type 4
group ions type 3 4

------------------------------- potential -------------------------------

pair_style lj/charmm/coul/long 8.0 10.0 7 9

pair_coeff 1 1 0.00673885 3.169
pair_coeff 2 2* 0.00000 0.00000
pair_coeff 1 2 0.00000 0.00000
pair_coeff 3 3 0.004336412 2.583
pair_coeff 4 4 0.004336412 4.4
pair_coeff 1 3 0.0054 2.8760

pair_coeff 1 4 0.0054 3.7845
pair_coeff 3 4 0.004336412 3.3302
velocity all create 300.0 12345678 dist gaussian

kspace_style pppm 4.3e-6

special_bonds charmm
bond_coeff 1 19.51 0.9572
angle_coeff 1 2.39 104.52 0 0

fix fxnvt all nvt temp 300.0 300.0 100.0

------------------------------- minimization -------------------------------

min_style fire

minimize 0 0 100000 1000000
write_restart min

------------------------------- Deletion -------------------------------

delete_atoms overlap 1.5 water ions mol yes

------------------------------- run -------------------------------

timestep 1.0

dump 1 all custom 100 traj_nvt.lammpstrj id mol type x y z ix iy iz
dump 2 ions custom 100 traj_nvt_ion.lammpstrj id mol type x y z ix iy iz
thermo 100

run 50000

why do you use delete_atoms? AFAIU from the manual, it requires bond yes option and proper molecule IDs in data file , otherwise some bonds can go missing

perhaps you just need to do some small timestep (0.005-0.1fs) equilibration after minimization to resolve overlapping issues, and there will be no need to delete atoms