Hello,everyone, I want to simuate a system with reactive part and non-reactive part, which means I need to mix reaxff forcefield and OPLS-AA forcefield. In past discussions, I understand that using pair_style hybrid command may cause energy leakage, and I noticed that mdi package can use lammps as engine or dirver, so if it is possible to use this way to do what I was mentioned before without energy leakage.Any advice will be appreciate.
I mean lammps is used as a MM code for OPLS-AA part and also applied as QM code to do the reaxff part, Any advice will be appreciate.
The energy leakage is caused by mixing polarizable and non-polarizable components in a simulation. The non-polarizable part can polarize the polarizable part but not vice versa. That is independent from how you combine those two parts.
Thanks for your reply, another question is that if a non-reactive forcefield applied in this system is not charged, which means that atoms in non-reactive part have no charge, such energy leakage will still happend or not?
OPLS-AA does have mandatory partial charges.
Thanks for your kind reply，using another forcefiled without charge rather than opls-aa, will still cause energy leakage? opls-aa is best choice for my non-reactive part, but it can be replaced by other forcefields.
Then you will replace electrostatic coupling between the two subsystem with a plain mechanical coupling and you essentially replace one error source with a different one.
Thanks for your reply, I will try to use pure reaxff forcefield for a smaller system, such discussion is very impressive. Have a nice day!