I designed and synthesized a fluorescent sensor with an ICT mechanism. I want to analyze the excited states of this sensor. That is, I need to observe the electron transitions in the sensor’s excited state. In short, to determine whether ICT is present or not. I read that the way to do this is through NTO visualization, and I did some work on this. First, I performed S0 optimization, followed by Td-DFT analysis. Using this final output, I created three files named hole.cub, electron.cub, and transitionDensity.cub for NTO visualization using software called MultiWfn. My question is: Which of these three files should I use to demonstrate the ICT transition? Or should I use a different file? Thank you in advance for your answers.
The hole (occupied) and electron (virtual) NTOs are telling you about the occupied → virtual excitation, so if there’s intramolecular CT to be found, that’s where I would look for it. The transition density involves the overlap of ground and excited state wave functions and typically I find those hard to interpret. You can find more information about visualizing TDDFT calculations in this recent mini-review,
Thank you for your answer. In what cases and when can I use the Transition Density visualization? Or in what cases and when can I use the hole-electron visualization? Is the hole-electron visualization sufficient to prove ICT? If Transition Density visualizations show the excited state electron motions, wouldn’t it be easier to prove ICT from there? By the way, I can’t access your article, but thank you for the valuable information you provided.
There isn’t any “electron motion” per se, since this is time-independent QM (eigenstate picture). The NTOs provide the best possible particle/hole description of the excitation, as close as possible to describing the excitation as a transition between one occupied and one virtual orbital (as described in the review cited above). In my opinion, transition densities are hard to interpret due to their nodal structure, which (roughly speaking) looks like the product of nodal structure of the occupied and virtual NTOs.
Visualization is necessary but may not be sufficient depending on the use case. For quantitative CT analysis, I usually use the LIBWFA implementation using the STATE_ANALYSIS and NTO_PAIRS keywords in Q-Chem, which would generate NTO files along with the ctnum_mulliken.om file, followed by the computation of the CT numbers using these files with the TheoDORE package (Transition density matrix analysis — theodore 3.1.1 documentation). This analysis quantifies the charge transfer between different fragments of the molecule. Please refer to both the Q-Chem and TheoDORE user manuals for theoretical details.
CIS_AMPL_ANAL = TRUE will also print changes in both Mulliken and Loewdin charges upon vertical excitation.
Since I don’t know much about QM or MM, I need to make an effort to understand them. I’m still new to computational chemistry, but your answers are invaluable. Thank you very much.
Thank you very much, I will look into the details.