Re: [ESPResSo-devel] User Guide on ELC and noneutralization

[Axel Arnold]

Hi!

> I was wondering why there is no information in the user guide about how to
> use ELC with dielectric contrast. The testsuite seems to work and the basic
> command is the same as for MMM2D as far as I see, so maybe that would be a
> good idea to add that.

MMM2D and ELC do the same for contrasts.

> > Also, there is this thing with the "-noneutralization" command when
> > dealing with non-neutral systems. There is no information about that
> > command in the user guide as well. I now figured out for myself, that
> > one has to add that command, but I find its name somehow misleading.
> > Actually when activating that feature there IS a neutralization done by
> > the artificially added homogeneous background, so I would say its name
> > should rather be neutralization, or even better let it get that
> > information from the total charge of the system itself.
> 
> The name "-noneutralization" for this option is indeed somewhat misleading.
> It originates from the fact that this option kills an unwanted effect
> (creation of a parabolic potential) associated to the homogeneous
> neutralization inherited from the underlying P3M method in P3M/ELC.

The standard approach of P3M leads to a homogeneous, neutralizing background, 
which gives the parabolic potential that Vincent talks about. With "-
noneutralization", ELC does not add any neutralizing background, hence the 
name. Formally, that can be understood as two equally charged, neutralizing 
walls at infinite distance, which does not exert forces, but effectively, 
simply the neutralizing background is switched off. Therefore I find the name 
quite reasonable.

> Personally, I think that this option should be activated automatically in
> Espresso when simulating a charged slab with P3M/ELC. The name
> "noneutralization" could be dropped entirely. When simulating a slab system
> with a net charge, Espresso could output a message "Warning: system with a
> non-zero net charge. Using background charge correction, see Ref. ...".

Another case where you need this flag are charge plates, which also add 
charge. To automatically detect whether one has to neutralize or not, ELC 
would also need to check whether charge plates are present, which is not 
implemented. Therefore, one has to manually tell ELC whether to apply the 
neutralizing, homogeneous background or not. And since P3M by default 
neutralizes, ELC also does.

I'll put a bit more documentation to the UG.

Axel

-- 
JP Dr. Axel Arnold
ICP, Universität Stuttgart
Pfaffenwaldring 27
70569 Stuttgart, Germany
Email: address@hidden
Tel: +49 711 685 67609