[Wien] Pt atom in a box

[xn22 at drexel.edu]

The problem is that you used GGA ex-potential. You will get converging very fast
if you use LDA. The reason is that it has second-derivative in GGA which leads to
sharp osscillation for small density area in vacuum region but no problem with LDA.
You can avoid this kind problem by following two methods:
1. Reducing the mixing parameter as small as possible (1x10^(-4)). You need to test
how small is ok for different system;
2. Occuping the vacuum region by filling as many empty spheres if you really need
GGA-ex. Otherwise just use LDA.

Good luck.

Xiliang Nie

----- Original Message -----
From: "M. Todorova" <todorova at fhi-berlin.mpg.de>
Date: Monday, October 20, 2003 11:44 am
Subject: [Wien] Pt atom  in a box

> Dear all,
> 
> I am trying to calculate a Platinum atom in a box. The problem is, 
> thatno convergence can be reached even after 100 iterations. The 
> chargedistance seems to oscilate between values of 0.04-0.01.
> 
> Here are some details about the calculation:
> The dimensions of the box are (13x14x15) bohr, the muffin-tin 
> radius is
> 2.45 bohr.
> The cutoff is 16 Ry and ROOT sampling is used for the k space
> intergration.
> I have tried using both the Gamma point, as well as shifting it.
> For the mixing Broyden's scheme with a mixing factor of 0.01 is used.
> Appart from a spin-polarized calculation, I have also tried a
> fixed-spin-moment calculation, but to no avail.
> I attach my case.struct,case.in* and case.klist files.
> 
> Any suggestion are welcome. Thank you in advance.
> Regards
> Mira 
> 
> 
> ------------------------------------------------------------------
> Mira Todorova                           todorova at fhi-berlin.mpg.de
> Fritz-Haber-Institut, Abt. Theorie      Phone:  +49 30 8413 4808
> Faradayweg 4-6                          FAX:    +49 30 8413 4701
> D-14195 Berlin (Germany)                www.fhi-berlin.mpg.de/th
> ------------------------------------------------------------------
>