[lammps-users] Non-reactive MD with ReaxFF

Dear LAMMPS community,

I have a question to which your advice and help is greatly appreciated.

Does anyone know how I can run an MD simulation using ReaxFF as the forcefield at a high temperature of, for example, 3000 K, while there happens no bond breaking/formation in the system? Should I set some special parameters in the ReaxFF force field? As you know, bonds break and form very easily at this high temperature, if we use ReaxFF. I need to do the non-reactive run to equilibrate the system using an NVT ensemble.

Thank you in advance.

Farshad

Dear LAMMPS community,

I have a question to which your advice and help is greatly appreciated.

Does anyone know how I can run an MD simulation using ReaxFF as the forcefield at a high temperature of, for example, 3000 K, while there happens no bond breaking/formation in the system? Should I set some special parameters in the ReaxFF force field? As you know, bonds break and form very easily at this high temperature, if we use ReaxFF. I need to do the non-reactive run to equilibrate the system using an NVT ensemble.

actually, enforcing certain bonding conditions would keep the
system _away_ from equilibrium. the real high temperature
equilibrium would be exactly the state where bonds are often
broken.

that being said, nothing keeps you from adding explicit bonded
forces to your topology. doing this on the reaxff side would be
very complicated, since reax is recomputing bonds through a
charge equilibration and bond-order scheme.

so using additional explicit bonds as harmonic restraints to enforce
stability of molecular structures should work.

however, a big concern is the transferability of the parametrization
to you high temperature regime. do you have any confirmation for
that from some published literature?

cheers,
     axel.

Dear Axel,

Thanks for your help and good points.

Regarding the reference, as far as I know esp. for non-reacting systems, it is a common practice to equilibrate the structure, say for 200 ps, before starting the production stage of the MD simulation. But, I don't know whether that equilibration stage is also necessary for reacting systems being simulated by force fields such as ReaxFF. I haven't seen in the literature that equilibration process being done in conjunction with Reax at high temperatures.

So do you think whether it is right to heat up the system quickly (in 5 ps) from a nonreactive temperature such as 2000 K to a reactive temperature 3000 K and run the production stage without any equilibration at 3000 K?

Thanks.

Farshad

Dear Axel,

Thanks for your help and good points.

Regarding the reference, as far as I know esp. for non-reacting systems, it is a common practice to equilibrate the structure, say for 200 ps, before starting the production stage of the MD simulation. But, I don't know whether that equilibration stage is also necessary for reacting systems being simulated by force fields such as ReaxFF. I haven't seen in the literature that equilibration process being done in conjunction with Reax at high temperatures.

i have not much practical experience with reaxff, but the question is
beyond that.
equilibrium is when your system is in equilibrium with its environment
(i.e. the average
potential energy does not drift significantly). how long that takes is
very system dependent
and in some cases, it might even be impossible to get to equilibrium
with regular MD
at the target temperature. 200ps may be adequate, or may be far to short.

So do you think whether it is right to heat up the system quickly (in 5 ps) from a nonreactive temperature such as 2000 K to a reactive temperature 3000 K and run the production stage without any equilibration at 3000 K?

well, the question is whether you are interested in collecting
equilibrium data or not?

regardless of what you do to bring your system to the desired temperature of
3000K, it _will_ equilibrate and you can only collect equilibrium data
_after_ it
has (sufficiently) equilibrated. now if you want to disallow reactions
to happen
while already running at the target temperature, is a choice that you can make,
and it can be a good one or it can be a waste of time. that is impossible to
say without knowing anything about your system, your strategy, and your goals.

for certain, you will not be in equilibrium directly after you remove the bonds,
and how the result will be different from a system where you do not inhibit
bond breaking while heating and how fast you should heat the system and
whether that is a good approach is beyond my predictive capabilities.

axel.