[Wien] band gap with EV-GGA and PBE-GGA

Hong Jiang jiang at fhi-berlin.mpg.de
Thu May 24 16:11:04 CEST 2007


Hi Mary,
You can try using LDA+U, (orb in Wien2k) for the systems you are 
interested in. (see wien2k userguide 4.5.6) 

Good luck,
   Hong

Mary White wrote:
> Dear all,
>      Thank you for all your help. I once considered the GW 
> approximation.But what I calculate is heavy atoms,I can't deal with 
> heavy atoms with the GW method.Maybe in the pseudo potential,can add 
> the spin-orbit coupling.Is there indeed no other ways to calculate the 
> small gap in wien2k?
>     Thank you for your attention.
>                                                                    Mary 
>
> */Hong Jiang <jiang at fhi-berlin.mpg.de>/* wrote:
>
>     Hi,
>     Your problem is actually not related to which structure you use.
>     Usually the band gap calculated from LDA or GGA can not be taken too
>     seriously because of the well-known band gap problem of LDA/GGA, i.e.
>     the LDA/GGA band gaps are always under-estimated, and therefore for
>     small gap systems, LDA/GGA usually predict metallic behavior; for
>     transition metal oxides like MnO, FeO, CoO and NiO, even experimental
>     gaps are not small (~4 eV), LDA/GGA can also give zero or very small
>     gaps. In terms of the band gap, EV93 is known to give larger gaps,
>     but
>     usually not large enough compared to experiment. For accurate and
>     reliable calculations of band gaps, one has to use the GW
>     approximation, which is unfortunately very expensive and only
>     feasible
>     for small systems. See the review paper in Rep. Prog. Phys. 61, 237
>     (1998) for more details.
>     Hope this is helpful.
>
>     Hong
>
>
>
>
>
>     Mary White wrote:
>     > Dear R.Ahmed and all users,
>     > Thank you for your attention. After I first optimized the
>     > experimental structure (both lattice parameters and atom
>     position), I
>     > calculated the band structure and got the gap of 0.2ev with EV-GGA
>     > (0.1ev with the experimental structure EV-GGA). The optimized
>     lattice
>     > parameters is a little overestimated compared the experimental
>     lattice
>     > parameters. But I found the force on the two atoms is a little
>     large.
>     > So I optimized the atom position in min_lapw (not for lattice
>     > parameters).The force is small. Then I calculate the band gap
>     with the
>     > new optimized structure, but very strange,the gap is disappear.
>     That
>     > appears metallic. I don't know what I can do. Only optimized the
>     atom
>     > position,the result is very different (compared with the first
>     > optimized structure 0.2ev).I think the optimized structure is ok.
>     > There is no experimental band structure. I don't know how to
>     > deal with it. I tried to adjust the parameters,but that is no
>     effect.
>     > Any comments do help me.Thanks in advance.
>     >
>     > */Rashid Ahmed /* wrote:
>     >
>     > Dear Mary
>     > hi
>     > This is known fact that GGA usually overestimates lattice
>     > parameters which susequently decrease band gap value. As your
>     > result with experimental lattice parameters is already very small.
>     > so with further decrease in band gap, it shows wrong mettalic band
>     > structure.
>     > bye
>     > R. Ahmed
>     >
>     >
>     ------------------------------------------------------------------------
>     > From: /Mary White /
>     > Reply-To: /A Mailing list for WIEN2k users
>     > /
>     > To: /wien2k /
>     > Subject: /[Wien] band gap with EV-GGA and PBE-GGA/
>     > Date: /Tue, 22 May 2007 05:10:18 -0700 (PDT)/
>     >
>     > Dear all users,
>     >
>     > I calculated the energy band with EV-GGA and PBE-GGA.
>     > The system is semiconductor.And the gap I calculated with the
>     > experimental structure is about 0.1ev with EV-GGA.But when I
>     > calculate with the optimized structure,the gap is
>     > disappear.From the band structure,it appears
>     > metallic.The band from EV-GGA and PBE-GGA seems very
>     > similar.The parameters in the calculation with experimental
>     > structure is the same as that with optimized structure.I think
>     > the optimized structure is OK. I am puzzled with that.
>     >
>     > Does anyone know it? Look forward to your advice. Any
>     > help will be appreciated.Thank you in advance.
>     >
>     >
>     >
>     > Mary White
>     >
>     >
>     ------------------------------------------------------------------------
>     > Moody friends. Drama queens. Your life? Nope! - their life,
>     > your story.
>     > Play Sims Stories at Yahoo! Games.
>     >
>     > >_______________________________________________
>     > >Wien mailing list
>     > >Wien at zeus.theochem.tuwien.ac.at
>     > >http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>     >
>     >
>     >
>     ------------------------------------------------------------------------
>     > Catch suspicious messages before you open them�with Windows Live
>     > Hotmail.
>     > _______________________________________________
>     > Wien mailing list
>     > Wien at zeus.theochem.tuwien.ac.at
>     > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>     >
>     >
>     >
>     ------------------------------------------------------------------------
>     > Get your own web address.
>     >
>     > Have a HUGE year through Yahoo! Small Business.
>     >
>     >
>     >
>     ------------------------------------------------------------------------
>     >
>     > _______________________________________________
>     > Wien mailing list
>     > Wien at zeus.theochem.tuwien.ac.at
>     > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>     >
>
>     _______________________________________________
>     Wien mailing list
>     Wien at zeus.theochem.tuwien.ac.at
>     http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>
>
> ------------------------------------------------------------------------
> Boardwalk for $500? In 2007? Ha!
> Play Monopoly Here and Now 
> <http://us.rd.yahoo.com/evt=48223/*http://get.games.yahoo.com/proddesc?gamekey=monopolyherenow> 
> (it's updated for today's economy) at Yahoo! Games.
> ------------------------------------------------------------------------
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>   



More information about the Wien mailing list