GROMACS version: 2020
GROMACS modification: Yes/No
Dear all
I want to anneal a system containing water(tip4p) and polymer(PPO). I want first increase temperature to 2000 K then decrease it step by step to 295 K. I want to know if does this process lead to water be vaporized and bring instability and wrong analysis?
Thank you
Overall, I think there is nothing wrong with just trying. A question is what you would like to achieve with the simulation setup. If your system is still okay after you cooled it down you are theoretically in a sane state. One issue that might arise during such a procedure is that you might find a low-energy phase, that is not well parametrized by the force-field, because you overcome some very high energy-barriers. Thus, I would recommend thinking about what phases you would expect, checking order parameters and where you take the polymer parameter from. Here, I would expect the largest effect come from the force-field rather than the way you simulate.
What exactly will happen is also a question of how you control pressure during the process, the simulation time, and what thermostat you use during the process.
If you use an NVT simulation, I would expect the water not to be vaporized, but your simulation might become unstable and you might have to decrease the simulation step size, because at very high temperatures, atoms can move a lot in a short time.
Thank you very much for replying
Actually, I want to study water diffusion and its properties in polymer structure and with your deep description I convinced not to use such method. There is no compulsion for me to use annealing process for equilibration.
I want my system to be well equilibrated to show a system very close to real system. I just saw an article used annealing from 2000 K to 300K for the same system and it was a bit confusing for me. All in all, what technique do you suggest for more equilibration of system before main run(100 ns of NPT)?
Some body tell me use output of NVT as input for minimization and then use output of minimization as an input for NVT and do this process for example four times. Does it beneficial?
Thank you again