[Wien] w2web / was: insulator in virtual crystal approximation

[buero at luitz.at]

On Sat, 8 Oct 2005, Moritz Hoesch wrote:

> A word of caution: the w2web struct-generator doesn't like virtual atoms and 
> will change them to real integer charge atoms. I edited the case.struct file 
> by hand.

w2web was designed to help the novice user to perform a SCF calculation 
easily, and to help an experienced user to perform tasks that he/she 
rarely uses.

It was not designed to replace the expert!

w2web deduces from the nuclar charge the atom selected and inserts 
automatically the apropriate atomic symbol (which can only be done for 
integer numbers, indeed). Therefor it "does not like" virtual atoms.

But w2web is readily willing to assist an expert, i.e. if you set up 
(manually) a VCA calculation, you can then perform a SCF calculation and 
further calculate properties like DOS, spectra, etc. with w2web.


> To recap: I have to change the values (1) in the case.struct file [per 
> atom] (2) in the case.inst file [per atom] and (3) in the case.in2 file 
> [per unit cell].

I would recommend not change case.inst (except for anti-ferromagnetic 
calculations). Instead set up your case with "real" atoms (hint: use a 
unique label to ensure that a particular site remains an inequivalent 
position), and change Z in case.struct und number of electrons in case.in2 
AFTER initialisation.

Best
   Joachim

PS: Kampai  :-)


--
luitz.at | interfacing art, science and technology

Dipl.-Ing. Joachim Luitz KEG 
Wohlmuthgasse 18 . A-3003 Gablitz . T +43 2231 61254 . Fax +43 2231 612544
buero at luitz.at . http://www.luitz.at . skype://jluitz