TD DFT - excited state natural orbitals

[qadir]

Hello,

Is there any way to generate natural orbitals corresponding to an excited state with TD DFT?
I tried to dump the excited state wavefunction in WFN format with egrad (I would like to analyze the electronic
density with AIM), however, the WFN interface does not find any occupied orbitals (all occupation numbers
are NaNs).

Thanks!

Regards,
Qadir

[dw1032]

"Push"
I'm having the same question and can't find a proper solution/explanation on how to do it. I want to calculate the first excited state orbital of my compound. Using escf with n=1 one can of course calculate the transition including oscillator strength but when i print out the orbitals after the calculation I'm only getting the Orbitals of the ground state right?
I see people using Gaussian for that purpose and they show S1' wave functions as well.
Best regards!

[uwe]

Hi,

may I ask you what you need the wavefunction of the excited state for? The transition density can be generated and plotted, this is the property which cares all the physical information.

Natural transition orbitals can be generated for TDDFT by using the tool proper. Just call proper after a TDDFT calculation, go into the 'mos' submenu and there call 'dftnto'. The tool is interactive and prints all the options and guides you through the process. In the end, you will have the NTOs on disk. To generate 3D files to plot (e.g. from within TmoleX), switch to the 'grid' menu and use the 'nto <number of excitation>' command to generate plot files.

Regards, Uwe

[dw1032]

Thank you for the answer!
Regarding your question: I want to show that by breaking the symmetry in the excited state of my molecules the fluorescence emission is no longer forbidden.

[dw1032]

Follow up question because maybe my understanding of NTOs isn't quite correct.
Are NTOs the overlapp between occupied and virtual orbitals that contribute to the excitation as in: I calculate the NTO for the first excitation which has contributions from several orbital transitions (HOMO to LUMO, HOMO-1 to LUMO and so on). So the NTO for that excitation would be the overlapp of these transitions. If so, what is the difference between a virtual and an occupied NTO?

[finrod]

If we switch to the electron/electron-hole picture then a (set of) virtual NTO(s) corresponds to the (excited) electron wavefunction and the occupied NTO(s) corresponds to the hole wavefunction.

cf. https://manual.q-chem.com/5.2/Ch7.S13.SS2.html

The contributions of the different molecular orbitals to the excitations, like H=>L, H-1=>L etc., are then listed together with their 'weight' (i.e. contribution to the excitation) in the file 'escf.out'. You'll find it towards the end under 'excitation'.