Hy Slimane You can perform the calculation by changing one parameter in the fdf file ... BUT the results you will get from that are not take for grant. There are several problems related to calculating charged states by using Periodic Boundary Conditions. The most serious in my opinion is that the infinite replication of the charged cell generates an spurious electrostatic background - which is partially compensated by introducing a jellium background on the calculation ... but it does not completely vanish - which means that your Formation energies will be based on comparing total energies calculated in different potential references - which is senseless.
In most of papers I´ve refereed they use a potential aligment procedure to `correct` it ... but these aligment procedures are also very questionable. The main problem is that the correction ,as it is usually applied, corresponds just to an arbitrary shift on the Fermi Energy - i.e. the energy of the free electron you have to take into account to calculate the defect ionization energy. Thus, it corrects the energy of the electron reservoir which is interacting with your system ... but the problem of the mismatching reference potentials in you total energy calculations still remains. Another issue is the electrostatic energy generated by the charged periodic replicas ... there is simply no agreement if it must be corrected or not ... and even among those who say yes, there few agreement on how. If you take a look on the literature you will find formation energies of Charged Oxygen vacancies in ZnO ranging from -1 to +4 eV ... which clearly shows that we still don't have a definitive methodology to handle with charged supercells. In conclusion : Thats a very tricky subject! NH On Mon, Feb 7, 2011 at 12:59 PM, Slimane Haffad <[email protected]>wrote: > hi neyhmor, > > thank you very much for your quick response and for the links you gave me > , I already read some papers, but what i want to know exactly as follows: > for the calculation of the energy of an transition level (0 / q), we must > find the energy of the system in the neutral charge state (0) and the > enrgy at charge state (q), it is the latter that intrigues me, is this > energy that of the defect level or it must change few things in the input > file of siesta and recalculate the energy? > > slimane; > > > --- En date de : *Dim 6.2.11, N H <[email protected]>* a écrit : > > > De: N H <[email protected]> > Objet: Re: [SIESTA-L] semi metalic caracter > À: [email protected] > Date: Dimanche 6 février 2011, 21h36 > > The answer for the first question is : No > > For the second you can take a look on: > > Van der Walle CG, Neubauer J *J. App. Phys* *2004 *95, 3851. > > Janotti, A.; Van de Walle, C.G. *Phys. Rev. B.* *2007*, *76*, 165202. > > Lany, S.; Zunger, A. *Phys. Rev B* *2008*,* 78*, 235104. > > Makov, G.; Payne, M.C. *Phys. Rev. B* *1995*, *51*, > 4014-4022.<http://link.aps.org/doi/10.1103/PhysRevB.51.4014> > > > Lany, S.; Zunger, A. *Phys. Rev. B* *2005,* *72*, 035215. > > Lany, S.; Zunger, *Phys. Rev. B*, *2010*, *81*, 113201. > > Gerstmann, U.; Deak, P.; Aradi, B.; Frauenheim, Th.; Overhof, H. *Physica > B* *2003*, *340*, 190-194. > > Moreira, N. H.; Aradi, B.; Rosa A. L.; Frauenheim, Th. *J.* *Phys. **Chem. > C* *2010, **114*, 18860-18865. > > From the beginning be aware that calculating charged defects is a very > tricky subject! > > > > On Sun, Feb 6, 2011 at 5:02 PM, Slimane Haffad > <[email protected]<http://fr.mc283.mail.yahoo.com/mc/[email protected]> > > wrote: > > > Dear siesta users: > > I calculated the band structure of nitrogen doped ZnO nanowire and showing > a acceptor defect level just above the highest occupied state, knowing > that the Fermi level is below the highest occupied level. I want to know > if this looks like a semi-metallic character and how to calculate the > energy of the system at charge state q (-1) in siesta. > I'll be very grateful if someone can help > > slimane, > > > >
