[Wien] Energy resolved structure factors

[Peter Blaha]

> I would like to (as far as possible) get energy-resolved
> structure factors from the valence/semicore states. The
> particular structure I am looking at is MgO, and I would
> like the structure factors for the states associated with
> Mg 2p. I've tried setting the seperation between the core
> and valence states to -3, but as one might expect I run
> into all sorts of problems. Any suggestions?

This is quite trivial. Do the regular scf calculation!!!!!
Do NOT change core-separation,....

Once the scf has finished, check the eigenvalues listed in the scf file
and edit case.in2

For instance put EMIN to some value like -2 and run x lapw2
This generates a clmval density without states below -2 (without Mg 2s,2p)
run lapw3 (VAL) to get the structurfactor contribution for all states
above -2 Ry.

You can repeat that as desired with different EMIN.

When interested in F comming from the lowest eigenvalues, you simply
change NE in case.in2. E.g. to get the Mg 2s contribution, put NE=2.0
for Mg 2s+2p put it to 8,....

(You can also combine the recipies and truncate than with EMIN the 2s,...)


Never do   run_lapw   with such in2 files!!!


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