[gmx-users] how to calculate the conc in the genion

Tsjerk Wassenaar tsjerkw at gmail.com
Thu Aug 11 17:35:58 CEST 2011


>> I'd be amazed if the error was in the code and not in your calculation. The
>> number of water molecules doesn't matter for the calculation of the ion
>> concentration, of course. Pay attention to your box shape. And do consider
>> the number of ions has to be an integer, so for a given volume you cannot
>> get arbitrarily close to a given concentration.

The error is, to my humble opinion, in the reasoning. Concentration is
a macroscopic property, and when dealing with a minute volume, the
concentration of something in it is ill defined. Especially when
there's something else in that volume, taking up a significant amount
of space, like a membrane, protein or void, it becomes troublesome. I
would argue that the worst you can do in that case is take the volume
of the box and calculate the number of things to add from there to
reach a given concentration.
Whether the number of water molecules matters for the calculation of
the ion concentration depends on the unit you use for concentration.
Probably  molality is a better option than molarity. For that you do
take the number of water molecules. Frankly, that's what I usually do.
Doing so will give a desired concentration of ions in the solvent,
regardless of volume occupied by other (big) solutes or by nothing.
There is just one problem that stays nonetheless; in how far does the
bulk concentration you use as target correspond to the local
concentration you might need to use? Solutes, membranes and voids may
alter the local concentration significantly.

By the way, Lina, it would have helped if you had given the equations,
numbers and outcomes that lead you to believe there is something

Hope it helps,


Tsjerk A. Wassenaar, Ph.D.

post-doctoral researcher
Molecular Dynamics Group
* Groningen Institute for Biomolecular Research and Biotechnology
* Zernike Institute for Advanced Materials
University of Groningen
The Netherlands

More information about the gromacs.org_gmx-users mailing list