I am preparing a QMMM system of ionic liquids using CHARMM42sp (for squantum) and CHARMM45m (for mndo97). There's no linker atom. The QM part is composed of two isolated molecules, ammonium (cation) and an anion with a sugar ring.
I did the minimization in several ways with a constrain of 20 -- turning off qmmm, turning on qmmm with mndo97 pm3 method, turning on qmmm with quantum pm3 method, using SHAKE for all bonds including qm part.
However, the hydrogens of either ammonium or the anion kept coming off and flowing away. Also, the sugar ring couldn't stay in the correct chair conformation and is twisting in a strange way (and sometimes it even became flat).
The charges are set correctly, and there's no missing atom. The input file is attached.
May I know if there's anything else I can check to figure out what is wrong here?