[Wien] Charge density questions

[Peter Blaha]

There's no special normalization. Of course the charge has units of 
electrons/bohr^3, thus when you integrate the charge density over the unit 
cell you need to find the number of electrons in this cell. (maybe 11 in 
your case, beware of semicore and that a numerical integration due to the 
strongly varying density near the nuclei is probably highly inaccurate.)
 
> 1) First I would like to ask about the normalization of the Wien charge 
> denisty over the unit cell. Let me take a simple example: If the case, 
> for instance, is ordinary FCC Cu, then the number of valence electrons 
> per unit cell is 11, (3d^10 4s^1). If I then integrate the total valence 
> band charge density over a 3D volume equal to the unit cell, would the 
> integrated charge density then be equal to 11 elementary charges divided 
> by the unit cell volume in Å^3? Or could the density be normalized in 
> some other way? If so, how?
> 
> 2) And then a general question on how to calculate a good charge 
> density. If one is primarily interested in the charge density, are there 
> any special things one needs to consider, besides converging the 
> calculation with respect to k-mesh, RKmax and choice of appropriate 
> linearisation energies, as usual? Is it for instance better to use the 
> charge density as convergence criteria?

For a good charge density you may in addition to the usual parameters you 
mentioned, also inclease the LM list in case.in2 to L=8 or 10. (For the 
latter you may need to redimension some parameters).



                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------