minimize negative coulomb energy

As Axel said, you probably have some close contacts. Only one pair of atoms needs to be a little too close together for the energy of the entire system to be enormous. (And when you have 50000 atoms, it’s actually hard to avoid this.)

Often the starting geometry is so bad that even the minimizers fail to converge and I have to slightly modify my starting geometry a few times. I often change the minimizer algorithm, and reduce the “dmax” parameter to something small, like 0.02.

have fun

andrew

(Incidentally, the 1/r^12 repulsion used in the Lennard Jones interactions is more likely than the coulomb repulsion to be responsible for your initial large energies.)