The ABINIT Users Mailing List ( CLOSED )

[Myrta Gruning <myrta.gruning@uclouvain.be>]

Hallo Shinekin,


1- You cannot get the band gap from Kohn-Sham DFT, independently from
which xc you are using (J. P. Perdew, R. G. Parr, M. Levy, and J. L.
Balduz, Phys. Rev. Lett. 49, 1691 (1982); J. P. Perdew and M. Levy,
ibid. 51, 1884 (1983); L. J. Sham and M. Schlüter, Phys. Rev. Lett. 51,
1888 (1983); Phys. Rev. B 32, 3883 (1985).). 
So you cannot really compare with experimental data. 

If nevertheless you need/want to do that, note that
 
2- The band gap depends on the geometry (at equilibrium, or in presence
of strain...). So to have a fair comparison between LDA and GGA (and be
sure you are not just observing strain effects...)  you should compute
the Kohn Sham gap at the corresponding equilibrium geometry.  

cheers,
m

On Sat, 2008-11-01 at 09:23 +0100, shinekin@hotmail.com wrote:
> Hi Dear all,
> 
> In the band structure calculation of SnO2, I want to evaluate the 
> dependence of
> band gap on the exchange-correlation potential.
> 
> I found that the GGA underestimated the band gap substantially; the 
> calculated
> value was only 0.82 eV, but the LDA give a value of 2.9eV while
> 
> the experimental value was 3.6eV. Can anyone tell me why this two potentials
> given a totally different results? Why the LDA gets so
> 
> close to the experimental result while The GGA not? 
> 
> Any suggestion or advice would be greatly appreciated! Thanks in advance.
> 
> Below is the detail of calculation: (Except the exchange-correlation 
> potential,
> every parameter keeps unchanged)
> 

-- 

___________________________________
dr. Gruning Myrta
Université catholique de Louvain
SE/FSA/MAPR/PCPM
Croix du Sud,1
1348 Louvain-la-Neuve

e-mail: myrta.gruning@uclouvain.be
phone1:  +32 (0)10 478721
phone2:  +32 (0)10 478681 
fax:    +32 (0)10 473452
URL:    http://www.myrta.eu/
___________________________________