Firefly and PC GAMESS-related discussion club

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

Re^3: wrong oscillator strengths for high symmetry molecules in TDDFT calculation


On Tue Dec 6 '11 5:59pm, Andrii Kulinich wrote
>>1. I did not see anything wrong with the example you attached.
>The first seven transitions in the TDDFT spectrum of ferrocene have zero oscillator strengths, and this should be wrong (on my opinion).

Does your calculated spectrum on the one hand and your opinion on the other hand agree with this spectrum: ? And the conformer is wrong: you use staggered conformation, while the correct one is eclipsed (see discussion on Chemport).

>>2. As to second part of your question, could you please be a bit
>>more specific. It is hard to answer without looking into actual
>>output files. Note, in C2 group, there are no forbidden transitions
>>at all. Does your question mean that you got large transition
>>moments for forbidden transitions in Ci group?
>I did not get large transition moments for forbidden transitions in Ci group. Nevertheless, I believed that for a molecule with C2 symmetry the B-symmetry transitions should be "more allowed" than A-symmetry transitions, but in some cases of C2-symmetry bis-dyes it was vise versa.

I know systems like this (I study very similar dyes at the Photochemistry Center). TDDFT cannot correctly reproduce spectra of such dimers. Instead of so called Davydov splitting of one pi->pi* absorption band into two (allowed and forbidden ones), you obtain spurious charge-transfer transitions (the forbidden ones) due to wrong asymptotic behavior of any LDA or GGA functional. First four transitions (from MOs #190, 191 to MOs #192, 193) are of interest. Others are from lower lying orbitals, they do not contribute to absoption much. Transitions #1 and #3 are allowed. They are normal delocalized transitions 191->192 and 190->193, respectively. Transitions #2 and #4 are spurious charge-transfer transitions (190->192 and 191->193). This is not seen directly from MOs, but you may construct NOs for each transition and analyze them visually. Allowed transition #1 has lower oscillator strength, while transition #2 is more intense.

>The other point is that (according to the classical molecular exciton theory) in bis-dyes with Ci symmetry of their molecules (angle of 180 degrees between their chromophores) the long-wavelength allowed transition should be more intensive then the short-wavelength one(1-st and 3-rd transitions correspondingly in the attached file), but TDDFT calculation gives opposite result.

This is not quite true. The key point is not the angle between chromophores, but the angle between transition dipoles in each chromophore. "Classical" Davydov's model is built on point dipole approximation, which is not valid for such short distances. Moreover, extended dipole model (by Kuhn) is not valid, too. Both theories are rather speculative. Neither TDDFT, nor CIS, nor CASSCF (which are free from spurious charge-transfer transitions) can reproduce exciton splitting, because it is governed by the detailed structure of the dimer (and monomers are not necessarily planar!).

Anyway, we can discuss this problem privately (in Russian). By the way, this is the experimental spectrum of the monomer dye

[ This message was edited on Wed Dec 7 '11 at 2:09am by the author ]

[ Previous ] [ Next ] [ Index ]           Wed Dec 7 '11 2:09am
[ Reply ] [ Edit ] [ Delete ]           This message read 1849 times