Multiwfn forum / Triplet state optimizations beyond T1 using TD-DFT

Re: Triplet state optimizations beyond T1 using TD-DFT

dummy@example.com (rikaaardoss) — Sat, 23 Jan 2021 02:44:21 +0000

Dear Tian Lu, I did the calculations with the suggested commands, however the "state tracking" problem persists, I have tried to reduce the maxstep to 1, change the functional to b3lyp and camb3lyp, also the option verytight and also I could not solve the problem. I found very interesting the article that uses NTOs to solve the problem (SDNTO), however I don't have the knowledge to implement something like this (I'm still waiting for the response from the authors of the SDNTO code to have access).

I wanted to update the situation regarding my problem and I want to thank you for your kind help. Apparently if traditional Gaussian algorithms are used, the study of triplet states from T2 onwards for transition metal complexes becomes complicated.

Best regards

Re: Triplet state optimizations beyond T1 using TD-DFT

dummy@example.com (sobereva) — Fri, 22 Jan 2021 02:57:14 +0000

If you mean your TDDFT optimization doesn't follow the potential energy surface specified by "root", try to decrease trust radius of optimization, using e.g. "maxstep=3,notrust" in opt keyword. As can be seen in Section 4.3 of my blog article http://www.shanxitv.org/468, the S2 state was successfully tracked with these options.

Triplet state optimizations beyond T1 using TD-DFT

dummy@example.com (rikaaardoss) — Thu, 21 Jan 2021 16:29:11 +0000

I have been trying to optimize triplet states of transition metal complexes beyond T1 (more precisely T2 and T3) using TDDFT and Gaussian16. The level of theory used has been the B3LYP functional and the 6-31g* and LanL2DZ pseudopotential bases for the metal however I have had the problem of the "crossover" of triplet states (or state tracking). Basically the T2 converges to T1, the same thing happened with T3, at some point the T3 state also converged to a T2 state. I read Professor Tian Lu's post: "Talking about the influence of potential energy surface crossing on the optimization of excited state" (sorry, I don't speak Chinese so I used a translator, I hope the title is correct) http://www.shanxitv.org/468, however I have not been able to correct this error. My last attempt was to add difuse functions to the calculation but apparently I will have the same problem (the calculations are still running).

From what I understand, this type of problem occurs when the states are relatively close to each other, since this problem is not observed when I optimize the singlet states (which are quite far apart). I want to emphasize that I am trying to evaluate the possibility that the compound I am studying does not satisfy Kasha's rule (I have experimental data from an experimental collaborator).

I would appreciate any help or advice to avoid and/or solve this problem.

Best regards.