[gmx-users] Mix two solvents

[Mark Abraham]

----- Original Message -----
From: teklebrh at ualberta.ca
Date: Tuesday, August 10, 2010 13:50
Subject: [gmx-users] Mix  two solvents
To: gmx-users at gromacs.org

> Dear Gromacs,
> 
> I want to combine two separately equilibrated solvents into one 
> mixed solvent file. How can I do that? I want to check the 
> mixing effect of those solvents and try to check the density and 
> volume change before and after mixing those solvents. I have 
> equilibrated and run MD for 2ns for each solvent and the density 
> I got is pretty the same as of the experimental value. But now I 
> want to combine both these files into one .gro file so that to 
> see the volume/density change after MD and calculate the mixing 
> effect. Can anybody help me on this issue?

First, decide the solvent ratio.

One approach is to use the "replicate this small box" feature in genconf, where the problem now reduces to generating a small box with about the right ratio of solvent molecules. For a 1:1 ratio, that might be as simple as getting a single example of each molecule, translating one of them slightly with editconf, concatenating the coordinate files in a text editor, and adjusting the box size suitably. You'll want to leave some gap between the molecules so that they don't have steric clashes, but not so much gap that the replicated box will have an excessive volume.

Then, use genbox to replicate this. If the solvent ratio of this box isn't quite satisfactory, randomly select enough molecules to delete to fix this (remember to adjust the total atom count for .gro files!).

Construct the .top file by #including suitable solvent .itp files from your earlier work.

Then do very gentle EM and equilibrate. You might need to use NVT to let the temperature adjust first, then NPT to let the volume adjust to a sensible density. You might need to use a very small timestep, or very weak pressure coupling to stop things exploding.

Once T, P and density look about right, switch to whatever ensemble you want to sample and equilibrate for a while. The random generated velocities will break up the order that was created by the replication, but not immediately!

Mark

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20100810/d3bec7ba/attachment.html>