Is it possible to plot transition densities between TDDFT excited states

With state_analysis = true and cc_trans_prop = 2, it is possible to calculate and visualize the transition densities between EOM-CCSD excited states. Is it possible to do that transition densities between TDDFT excited states?

I also need to calculate transition dipole moments between TDDFT excited states. The only way that I’m aware of is use the GMH diabatization code (sts_gmh = true). I wonder if any other approaches are available.

The state-to-state transition moments definitely exist; use STS_MOM = TRUE. I don’t know that state-to-state transition densities exist unless it’s possible through libwfa.

Hi, it definitely works for spin-flip TDDFT, see example 7.30 here

https://manual.q-chem.com/latest/ex_SFDFT-wfa.html

And I am pretty sure it would also work for normal TDDFT.

The crucial keywords would be something like

$rem
   STATE_ANALYSIS  = true
   WFA_REF_STATE   = 3
   MAKE_CUBE_FILES  = true
   PLOTS =           true
$end

$plots
grid_points                    50 50 50
$end

And I guess, you can do the same for CCMAN if you follow Example 7.165 here
https://manual.q-chem.com/latest/ex_eomip-wfa.html

Note: To get transition densities with libwfa, you have to set: MAKE_CUBE_FILES = true. Everything else should be as shown in that example

Hi Felix,

WFA_REF_STATE = N does work for normal TDDFT. Thank you very much.

For EOM-CCSD jobs, I noticed that cc_trans_prop = 2 will calculate transition density for all excited state pairs, which generates an excessive amount of data. Using cc_trans_prop = 1 and setting up the reference state using cc_state_to_opt might be a preferable option in some cases. It works in a similar way as WFA_REF_STATE. This is just for future reference.

Yuezhi

Thanks Felix. Can we put this information into the manual, both with regard to TDDFT state-to-state transition densities and Yuezhi’s trick for EOM-CCSD?

I think the information is already in the manual (the keyword wfa_ref_state is there). It probably just needs to be discussed more in the text of 10.2.12.

I can add a note in the EOM-CCSD property section regarding how to set cc_trans_prop for state analysis.