Newbie question: calculated bandgap of Si is not correct

Message

qql_2011 · #1 Post by **qql_2011** » Sat Jun 04, 2011 2:16 pm

Dear VASPers,

I did a simple band-structure calculation following the 2_4_diamondSi example in the VASP online workshop, and found that the calculated bandgap of Si is only ~0.6eV. I checked the plot in the handsonII.pdf and found the results (from the plot and my simulation) are consistent. Can anyone guide me to find out what is wrong? Thanks!

PS: the example and pdf files are in the following URLs
http://cms.mpi.univie.ac.at/vasp-worksh ... ndson2.tgz

http://cms.mpi.univie.ac.at/vasp-worksh ... dsonII.pdf

<span class='smallblacktext'>[ Edited ]</span>

forsdan · #2 Post by **forsdan** » Sat Jun 04, 2011 5:44 pm

Only one of the many many articles on the subject:

http://prl.aps.org/abstract/PRL/v96/i24/e246401

Cheers,
/Dan

qql_2011 · #3 Post by **qql_2011** » Sun Jun 05, 2011 4:53 am

Hi Dan,

Thanks a lot for kindly reply! Seems that it is a well-known issue... obviously I should do more homework on this.

Then a lame question follows: is there a clean solution from the simulation: e.g. any input tags can avoid the DFT issue? I tried using B3LYP method (by turning on GGA=B3 in INCAR and LEXCH = B3 in POTCAR), what I got is about 0.77eV. closer but not enough...

<span class='smallblacktext'>[ Edited Sun Jun 05 2011, 05:26AM ]</span>