[Wien] calculation with different RMT for the same atom

[Peter Blaha]

When you do calculations with different RMT for hydrogens, chances are very high that the results
are bad. The H-wavefunctions from the large RMT will be much better converged than the other ones
and this may lead to some artificial charge transfer,.....

So you have to repeat your "other" calculations and reduce the spheres there.

PS: What "orbital dependend" functional did you use for hydrogen ???? (I hope NOT LFA+U ?)

Yurko Natanzon schrieb:
> 2009/11/10 Laurence Marks <L-marks at northwestern.edu>:> If you have different RMT's for the same type of atom, in effect they> are different atoms. Run with the same RMT for all the H, this should> work without problem in principle, although other things can go wrong,> and your QTL-B/Select errors suggest that you might have other> problems.
> The problem is that when I introduce an impurity, I cannot use thesame RMT as for other H atoms, because this causes RMT overlap. I alsocannot reduce RMT of all the H atoms, because I want to compare mycalculations with other ones calculated for large H RMT (especiallywith orbital dependent functionals, where RMT matters a lot). so theonly solution I have is to reduce the RMT of the impurity.
> 2009/11/10 Peter Blaha <pblaha at theochem.tuwien.ac.at>:> You MUST obey the rules and use the same RMT for all hydrogens (eg. 0.8> bohr).> Also remember, that small H spheres do NOT require the usual RKMax, but even> RKMax=3.5> might be sufficient (test it !).
> Yes, thank you, I reduced the RKmax to 4, and it now works. Do youmean, that if I use different RMTs for the same type of atom, I getsome results which are in some way unphysical?
> with kind regards,Yurko
> --Yurko (aka Yuriy, Iurii, Jurij etc) NatanzonPhD studentDepartment for Structural Research (NZ31)Henryk Niewodniczański Institute of Nuclear PhysicsPolish Academy of Sciencesul. Radzikowskiego 152,31-342 Krakow, PolandE-mail: Yurii.Natanzon at ifj.edu.pl, yurko.natanzon at gmail.com_______________________________________________Wien mailing listWien at zeus.theochem.tuwien.ac.athttp://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 

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