[Wien] When to add a U.

[Peter Blaha]

Ca is know to have "almost" active 3d electrons (the next element in the
periodic table, Sc,  of course has). 
E.g. in an LCAO description, an accurate calculation would need d-basis 
fuctions for Ca. Thus it
is reasonable and clear that LDA+U for Ca-d has an effect (I'll not duscuss
whether it is physically meaningful or not, but it might be that the Ca-d
bands are not fully correctly positioned within LDA alone !).

Mg does NOT have d-electrons! An LCAO calculation does not need d-basis 
fuctions. So putting LDA+U on Mg-d is meaningless!
The  d-partial waves of Mg within our APW formalism are tails of O-p 
wavefunctions!
If you want to play around, you may want to add LDA+U to O-p ! (or
Mg s,p - although I'd call the latter also "unphysical").


> Just to expand the scope of the discussion (?) further, for fun I did
> LDA+U for MgO. With U=6eV, J=0 there is a change of about 0.01 Ryd/Mg
> atom. Although MgO has no "d-electrons", it does have d-harmonics in the
> density and if I understand DFT/U correctly, there is no distinction
> between d-electrons and d-harmonics. This is probably similar to the
> effects in Ca....??
> > systems". However I recently read a nice paper where they calculated NMR
> > shielding parameters, which depend on the unoccupied bands:
> > J. Am. Chem. Soc.; 2004; 126(39); 12628-12635
> > The conclusion here is that one need a U for the unoccupied d-states of Ca.
> > This
> > would off course not influence any of the ground state properties, but it
> > surprised me because I would guess the shape off the d-state to be highly
> > diffuse.


                                      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/
--------------------------------------------------------------------------