[Wien] Supercell and point simmetry

[francesco d'acapito]

Dear wien users

I'm using wien in the calculation of relaxed structures around dopants in 
diamond/zincblende structures. In some papers [ex. S. Lany et al. PRB 62 
(2000) R2259] i've found that, in order to reduce computational time a Td 
point group simmetry around the dopant is mantained in the supercell (SC). 
How is this realized in practice ?

Normally I define a 16 atoms SC with all independent atoms (by applying 
small perturbations to the coordinates in order to remove equivalency) but 
doing this to a 32 or 64 SC should lead to unacceptable minimization times 
(in my case 4-5 days for a 16 atom SC).

Is there a smart way to modify the original output of the program supercell 
to obtain a new one that preserves the Td point symmetry ?
****************************************************************************
dr. Francesco d'Acapito
GILDA CRG c/o ESRF
BP220 F-38043 Grenoble (F)

Tel +33 4 76882426   Fax +33 4 76882743

e-mail: dacapito at esrf.fr
URL: http://www.esrf.fr/exp_facilities/BM8/handbook/control.html
****************************************************************************