Band gap calculation: L(S)DA+U INCAR FILE

Message

hyon-jee · #1 Post by **hyon-jee** » Mon Mar 17, 2008 11:01 pm

Hello

I'm getting into the band gap business and had a test run with Si. With a regular DFT, I get only 0.54 eV. So I tried with LDA+U.

I made a INCAR file based on what I read from the manual (I wish it had more easy to follow examples), but my calculation just doesn't start. Can someone provide a simple INCAR file for LDA+U method?

Thanks!

#2 Post by **admin** » Thu Mar 27, 2008 8:48 am

Please set LDAU=.TRUE. and LDAUTYPE, LDAUL, LDAUU and LDAUJ in INCAR A simple example is given in excercise 4.3: NiO +LSDAU of the vasp-workshop tutorials, you can download this file from our homepage http://cms.mpi.univie.ac.at/vasp-workshop, it is untarred from Handson4.tgz . However, please note that the underestimation of the bandgap is a known feature of all standard DFT methods, and that Si is certainly not a strongly correlated system (as would be suitable for LDAU). The implementation of exact exchange contributions (hybrid functionals) as will be incuded in vasp.5.1, gives a significant improve of the calculated band gap widths.

pablop · #3 Post by **pablop** » Tue Apr 01, 2008 4:30 pm

Hello, when this significant improve can be make true for normal users?. When will vasp.5.1 be finally distributed?

#4 Post by **admin** » Thu Apr 03, 2008 1:35 pm

we will inform all registered VASP users as soon as the release date and conditions to vasp.5.1 are fixed. It will be this year.