[Pw_forum] Fermi energy

[Giuseppe Mattioli]

Dear Valentina

> Is it already the Fermi Energy defined in the article by Zunger?

Definitely not! You should individuate the valence band maximum and  
use the corresponding eigenvalue in the formula as a constant. The  
Fermi level is the variable (you calculate the formation energy at a  
given Fermi level, and, possibly, the transition level between two  
different charge states of the same defect). The rationale is that the  
position of transition levels should be independent on the Fermi  
level. I may suggest (but do not tell prof. Zunger, please...) that  
Van de Walle's paper is somewhat clearer:

Van de Walle, C. G.; Neugebauer, J.; J. Appl. Phys. 2004, 95, 3851.

A very recent review by Pacchioni and Di Valentin, which improves in  
some details the calculation of transition levels, can be found here:

http://pubs.acs.org/doi/pdf/10.1021/ar4002944

There is also another detail, often hidden into the papers. When you  
perform the calculations in a plane-wave/pseudopotential PBC  
framework, if you change the total charge, you also insert a  
background charge that shift the potential (defined up to a constant  
value) and, in turn, the eigenvalues. You should insert some DeltaV  
value in the formation energy formula to recover the shift. If the  
supercell is large, and the dielectric constant of the host matrix is  
high (let us say a >100-atom cell of TiO2, for instance), then DeltaV  
is negligibly small. Otherwise you can find something around some  
tenths of eV. However, Pacchioni and Di Valentin have tested a  
workaround for the estimate of DeltaV (see the review).

> occupations='smearing' with degauss =1d-20

I hope that it is degauss =1d-2

HTH

Giuseppe

Giuseppe Mattioli
ISM-CNR
Italy

Quoting "DELLACA' Valentina (CRF)" <valentina.dellaca at crf.it>:

> Hi,
> We are computing the formation energies of charged defects in  
> supercells. We are referring to Lany and Zunger, Phys. Rev. B 78,  
> 235104 (2008) Eq. (1).
> The Fermi energy used here is defined as  Ev (valence band maximun)  
> + delta E Fermi ; i.e. is defined with respect the valence band  
> maximum.
> The system we are working with is an insulator, and we are using  
> occupations='smearing' with degauss =1d-20 in order to have the  
> Fermi Energy written in the log file.( Please, tell me if we are  
> doing something wrong here!)
> The question is: the Fermi Energy that I read in the log file, is  
> defined respect to what? Is it already the Fermi Energy defined in  
> the article by Zunger?
> Thanks,
> Regards
> Valentina
>
>
> Valentina Dellacà
> Group Materials Labs
> PA&CT - Materials CAE Development & Virtual Analysis
>
>
> Centro Ricerche Fiat S.C.p.A.
> Sede legale e amministrativa: Strada Torino, 50
> 10043 Orbassano (TO), Italia
> Tel  +39 011 9083138
> Fax +39 011 9083666
>
> www.crf.it<http://www.crf.it>


-- 
********************************************************
- Article premier - Les hommes naissent et demeurent
libres et ègaux en droits. Les distinctions sociales
ne peuvent être fondèes que sur l'utilitè commune
- Article 2 - Le but de toute association politique
est la conservation des droits naturels et
imprescriptibles de l'homme. Ces droits sont la libertè,
la propriètè, la sùretè et la rèsistance à l'oppression.
********************************************************

    Giuseppe Mattioli
    CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
    v. Salaria Km 29,300 - C.P. 10
    I 00015 - Monterotondo Stazione (RM)
    Tel + 39 06 90672836 - Fax +39 06 90672316
    E-mail: <giuseppe.mattioli at ism.cnr.it>