> Unfortunately the process I am simulating says otherwise from experimental
> point of view. I determined the values and now I need to apply them to my
> system.
> >
> > "I need to change the dielectric constant of water in
> > this droplet from 78.5 to 1 and at the same time preserve the vacuum."
The concept of a dielectric constant is important in the context of
*absent* solvent. If you want to "mimic" the dielectric effect of water
(where the dielectric is 1 at close range--no solvent screening-- and 78.5
at long range (where there is a lot of water screening the Coulombic
effect between two charges) in simulations *without* water you could
implement screening as a function of distance. In the old days, we did
this with a distance dependent dielectric. It would be a really bad idea
to turn on distance dependent dielectric and use a solvent blob (although
this has been done previously) since you doubly screen and drastically
inhibit motion.
However, in simulations in explicit solvent, as Roitberg says, we always
set the dielectric to be one since solvent is present to screen the
charges. So, you want to run MD with dielectric equal to one, but somehow
effectively measure what the dielectric is as a function of distance from
your solute to see how you compare to experiment. This is non-trivial
since dielectric screening is a bulk property, and calculating it in
different shells around the solvent will likely be difficult to converge
and/or require a fair amount of sampling...
--tec3
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Apr 21 2017 - 14:00:02 PDT
