[Wien] O2 molecule continue (rmt issue)

[Peter Blaha]

> 1- Computing the atomic energies of atoms like N and P in an FCC cell is 
> ok, however for O atom the high symmetry of the FCC cell results in 1/3 
> occupancies (for the 4th p electron of O) in the spin down case. Only 
> using a lower symmetry cell (orthorhombic) for O atom eliminates this issue.

This is true, however, for O this is a very small (negligible) energy 
difference. It is a bit more important (also to help convergence) for 
transition metals or even lanthanides.

> 2- For very small RMTs (e.g. 1.1 for O) one needs a large value of Gmax 
> (~20).

Again true. GMAX for small spheres has some effect on E-tot. I have not 
cross checked if this truncation cancels out in the energy differences.

> Using RMT = 1.1 for O in a compound will require one to use too small 
> RMTs for large atoms. For example if we use RMT = 1.1 for O in compounds 
> like SnO2, this will require one to use an RMT ~ 1.6 for Sn (RMTs 
> difference should not be > 30 %). This means that the ecut should be set 
> < -9.8 Ry which includes the 4s electrons to valence states. What are 
> the consequences?

Without test: I would "break" my own rules and set the Sn sphere still 
to 2.0 or 2.2 (such that no core leakage occurs at ecut=-6). I would not 
expect ghostbands coming up. This rule usually applies to cases where 
one atom could have very large RMT (2.5 or even larger) and probably 
converges rapidly (eg. alkali(earth) atoms in TM perovskites, ...)

But in principle there is nothing to say against Sn 4s as valence, 
besides the statement: be careful with the energy parameter (very fine 
E-search in case.in1) and make sure that RKmax is sufficient for such 
localized states.


> 
> Best Regards,
> 
> Alay
> 
> 
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
> 

-- 
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW: 
http://www.imc.tuwien.ac.at/tc_blaha-------------------------------------------------------------------------