WIEN2k is an "ab initio" method. You cannot assign a certain number of VALENCE electrons to a specific atom. Only core electrons are specific to an atom and its occupation can be constraint.

The term "charge transfer insulator" describes a situation if the CBM and VBM has states on different atoms, thus an excitation changes the "valence" of the atoms. In contrast, a Mott insulator would have CBM and VBM on the same atoms (eg. a d-d transition and an excitation does not change the charge state. A charge transfer insulator does not require that you move electrons around yourself.

However, often such systems are not well described by standard GGA calculations and you may need LDA+U, EECE or hybrid-DFT. This is also not "ab initio", because you have to make the decision and use for metallic fcc Co standard GGA, but for CuO GGA+U, but thats at the moment the state of the art.

Am 18.11.2019 um 16:32 schrieb ??????:
Respected Prof. Blaha,

Thanks for your detailed explation. The method I followed in the core-EELS calculation is consistent with what you said, setting core hole, adding electrons to background charge, etc. However, if I simply remove one oxygen 2p electron and simply add one electron in Cu 3d orbit to simulate Cu-based HTSC oxides charge transfer insulators, is this method feasible in Wien2k ?

Best wishes !

Yifan Ding (Ph.D candidate)

Institute of Physics, Chinese Academy of Sciences (CAS)
Address: No.8 Zhongguancun South 3rd Street, Haidian District, Beijing
100190 China

Wien mailing list

Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email:    WIEN2k:
Wien mailing list

Reply via email to