towhee-users Mailing List for MCCCS Towhee

On Wed, Nov 25, 2009 at 7:17 AM, Lawrence R. Pratt <lp...@tu...> wrote:

> Xinli and I are collaborating on this work.   I compliment you on the
> utility of Towhee which is helping us a lot.   *But* an opinion that
> Ewald does not make sense for the study of physically non-neutral
> systems is false.   There have been >10^3 well-posed and informative
> calculations on physically non-neutral systems that use Ewald.

I'm afraid we are going to have to agree to disagree on this point.
Charge neutrality may not be a law, but I continue to believe it is an
extremely good idea.

> The explanation of this point is that Ewald methods implicitly
> incorporate a uniform neutralizing background charge.   The treatment
> is thus "mathematically neutral" even if the system is physically
> non-neutral.   If the system treated is physically neutral, the

And perhaps this makes sense in the case of plasma, but it seems to me
quite a step to consider details of water structure about one ion as
being important enough to simulate atomistically while then treating
its very similar counterion as a uniform background charge.

That mixed model makes me uncomfortable especially when I try to
understand how it would be generalized to a multiple box simulation
such as the Gibbs ensemble, or even an ensemble that is at a fixed
pressure instead of volume.  Would an isobaric-isothermal simulation
of a box with 500 waters and 4 Cl ions with a uniform background of Na
predict the same liquid density as a simulation with 500 waters, 4 Na
and uniform background Cl; or as a simulation of 500 waters, 4 Cl and
4 Na?  If that sort of proof of the method exists in the literature I
would find it much more compelling than 1000 studies calculating
chemical potentials.  I have been betrayed too many times by chemical
potential to consider it a compelling measure, while density is so
reassuringly comparable to experiment.

> So .. if you need motivation (or cohorts) for this slight extension of
> Towhee, count us.   We'd be happy to use Towhee with that slight
> extension, to do serious calculations for physically non-netural
> systems of high current interest, and to prove to you (and that
> fraction of the world yet again) that Ewald and "infinitely
> replicated" non-neutral systems go together very nicely.

If you wish to implement a version of Ewald into Towhee that uses the
methods you describe then I would be willing to incorporate it into
future releases, provided it is an option that can be turned on or off
as the user desires.  It would not be the first feature implemented
into Towhee that I was not originally a big believer in, and some of
those have turned out quite well.

Marcus

-- 
Marcus G. Martin
Director, Useful Bias Incorporated
88 Martinez Road
Edgewood NM 87015-8222
ph. (505) 286-4457
www.usefulbias.com
www.photobirder.com