My Community

Dear All,

I want to use hybrid functionals in the VASP. 5* version. For the HSE06 functional, the INCAR file likes this:
GGA=PB
LHFCALC =. TRUE. HFSCREEN =0.2
ALGO =Damped
AEXX = 0.25
ALDAC =0.0
AGGAc =1.0

Are these keyword right?

How to set up the keyword for b3lyp functional?

Any suggestions are very appreciated!

Thanks in advance!

Dear Lyzhu,

The INCAR file for the B3LYP functional should contation the following tags:

LHFCALC = . TRUE.
ALGO = D; TIME = 0.5

GGA = B3
AEXX = 0.2
AGGAX = 0.72
AGGAC = 0.81
ALDAC = 0.19

Hope it helps.

Dear Mahlaga,

Thanks for your kindly reply!

For b3lyp calculation, in the POTCAR file, which kind of potential should we use? PAW_GGA is fine?

How about for HSE06 calculation? Do you konw my keyword is right? For the POTCAR, do we need PAW_PBE potential?

Thanks!

Dear Lyzhu,

You can use the PAW_PBE POTCAR files for the B3LYP calculations. Please mind, set the LEXCH tags in the POTCAR file to LEXCH = B3, for consistence with GGA = B3 tag in the INCAR.
I'm not so sure about HSE06, but I think it is used in combination with PAW_PBE potentials. Typical INCAR files for this functional are given on Page 100 of the manual.

Thanks

Could I use B3LYP based on the LDA ,compute the character of the systems based on the LDA ?

Dear Mahlaga,
I am doing the B3LYP calculations in VASP 5.2 using the keywords you specified. But I meet the following error messages:

internal error in FOCK_ACC: GRIDHF is not real
the gamma only version in combination with Hartree-Fock requires
to compile the code using -DNGXhalf or -DNGZhalf (mpiversion)

What is the problem of my calculations?
I am very appreciated for your responses!
Thanks!



<span class='smallblacktext'>[ Edited Thu Nov 12 2009, 05:42AM ]</span>

Solved!