Re: How can I find the proper number of solvent molecules in the simulation?

From: Brian Radak (
Date: Mon Jun 06 2016 - 09:49:42 CDT

Unfortunately I think you are asking a question with no proper answer.
Some practical answers (depending on what insights you are after) might be:

1.) as many molecules as you can afford to simulate - a vague and often
un-useful criteria

2.) enough molecules to give you the desired "concentration" of polymer,
perhaps based on an idealized density of the solvent - I like to think
of this as creating a virtual "beaker" of solvent and then adding a
single molecule of solute. The problem is that the desired concentration
is often much lower than one can realistically afford to model.

3.) if using PBCs, enough molecules to give an appropriate clearance
from the cell edge - Appropriate clearance might be plausibly taken as
the cutoff distance plus a few angstroms. You should also consider cell
volume fluctuations due to NpT.


On 06/06/2016 09:25 AM, faride badalkhani wrote:
> Dear all,
> I am going to investigate the effect of solvent on the structure of a
> polymer. Therefore. I performed the NPT simulations on the
> water+polymer and now I want to repeat the simulations with the other
> solvent which is a fluoroalkane. The problem is that I do not know
> what is the proper number of solvent molecules I should add to the
> solvent box. Could you help me on this, please?
> Regards,
> Farideh

Brian Radak
Postdoctoral Appointee
Leadership Computing Facility
Argonne National Laboratory
9700 South Cass Avenue, Bldg. 240
Argonne, IL 60439-4854
(630) 252-8643

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:22:14 CST