[Pw_forum] Where to find old format pseudopotential files?

[Paolo Giannozzi]

On Tuesday 25 July 2006 18:59, José Rui Faustino de Sousa wrote:

> >  For the core correction:
> >   rho_core(r) = (a_nlcc + b_nlcc*r^2) exp ( -alpha_nlcc*r^2)
>
> Do you know any citation to that formula? 

there is nothing to cite: it is a simple formula that does a lousy
fit of the tail of the core charge. As the chemists say, "you can fit
any function with three gaussians", but one is not enough. I don't 
recommand using it.

> And, by the way, for the scheme to construct the radial logarithmic grid?

as above: there is nothing to cite. The radial logarithmic grid that doesn't
start from r=0 makes no sense, but it has always been there and the 
atomic code is based on it, so it is not even simple to get rid of it. Its
usage for pseudopotentials makes even less sense but it is simpler 
to use the same grid for the atom and for the pseudoatom.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy