I am trying to simulate a metal-organic framework in LAMMPS using the UFF potential. It's working quite well for some structures where all molecules have an occupancy of 1.
However, when I have a solvent molecule that has an occupancy of 0.5, what I do is randomly select one of the two positions for each unit cell. The structure collapses upon simulation. I'm not sure how to fix this. My MOF is about 6x3x3 unit cells big; would there be any beneficial reason to try a huge system?
I've also tried simulating at low temperatures, NPT, NVT; it falls apart everytime. You're not really supposed to use partial charges with UFF, but I tried it with and without.
It's just amazing to me that one structure is so incredibly stable and then a slightly different structure that has 0.5 occupancies falls apart. Does anyone have any experience with this?