Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly  
We are NATO-free zone

Re^3: DFT Functionals: what is the one most similar to the M062x?

Alex Granovsky

Dear Enrico,

A source of some information on Rydberg states is:

I personally would suggest you to use pbe0 or a version of pbe0
with the fraction of HF exchange of 1/3 . For the latter, you may
look at:

This is consistent with my own experiences. Several years ago
I was trying to calibrate pbe0 and bhhlyp functionals to better
reproduce charge transfer states of push-pull derivatives of
stilbene and some di-aryl-benzenes and found that hybrids with
HFX=0.33333 work best.  

Kind regards,
Alex Granovsky

On Tue Aug 6 '13 10:32pm, Enrico wrote
>Dear Thom,
>Thanks for your suggestion...I edited my previous be on the safe side!
>Best regards
>On Tue Aug 6 '13 10:04pm, Thomas wrote
>>Dear Enrico,

>>Your question doesn't have a straightforward answer as comparing XC functionals is very difficult. Even if we limit the performance scope to TDDFT on D-A complexes, the behavior of functionals can vary widely with the system being studied. However, in general I think that hybrid functionals using a large portion of exact (i.e. HF) exchange will give you the best result. In FF 8.0.0 RC, the following functionals can be used with TDDFT:


Exchange Correlation Exchange-Correlation Hybrid
SLATER   VWN1RPA     SVWN1RPA             B1LYP
B88      VWN5        BVWN1RPA             BHHLYP
XPW91    PW91LDA     SVWN5                B1PW91
GILL96   LYP         BVWN5                B3PW91
XPBE96   CPW91       SLYP                 B3LYP1
OPTX     CPBE96      BLYP                 B3LYP5
                     GLYP                 O3LYP1
                     OLYP                 O3LYP5
                     XLYP                 X3LYP
                     BPW91                PBE1PW91
                     PW91                 PBE0

>>One popular choice for TDDFT is BHHLYP, which uses 50 % exact exchange. Note that the amount of exact exchange in a hybrid functional can be modified with the HFX keyword of the $DFT group. For example,


 $DFT HFX(1)=0.75 $END

>>sets the amount of exact exchange to 75 %.

>>Finally, I would like to advise you to edit your original message and remove the last sentence as it might cause problems with your G09 license.
>>Kind regards,
>>On Tue Aug 6 '13 5:51pm, Enrico wrote
>>>Hi, I'm performing TDDFT calculations on molecular D-A complexes, from what i read on literature one of better DFT functionals for such purposes is M06 (in particular the one with the doubled HF term). I read here in previous threads that these Mx functional are not implemented and they won't be implemented in FF for a while.
>>>In such circumstances i was wondering if any one of you could suggest me a good substitute of the M062x functional to be used in FF (or in general a good functional for UVvis prediction in donor-acceptor complexes where CT transition are involved).
>>>Thank you

>>>Enrico Chiavazza, PhD
>>>Dept. Chemistry
>>>Univ. of Torino, Italy

[ Previous ] [ Next ] [ Index ]           Tue Aug 6 '13 11:28pm
[ Reply ] [ Edit ] [ Delete ]           This message read 918 times