Hi Everyone,
I have some questions regarding geometry optimizations for excited states, and I hope someone can clarify them for me.
Question:
When performing calculations involving excited states in Gaussian, I should use TD-SCF methods, correct? In that case, I need to include the "td" keyword in my route line; otherwise, the calculations will be performed in the ground state.
But, specifically, for molecules with a net spin of 1, if I am interested in optimizing the geometry for T1 and T2, I should proceed as follows:
- For T1, I should change the multiplicity to 3 and not include the "td" keyword (e.g., `# opt freq L.O.T`).
- For T2, I should include "td=root=1" and set the multiplicity to 3 (e.g., `# opt freq td=(root=1) L.O.T`).
Is this specifically for molecules with a net spin of 1 because T1 is essentially their ground state on triplet?
Additionally, should all my calculations follow this route line format if I want to calculate the geometry for excited states S1, S2, S3, etc.?
- (e.g., `# opt freq td=(root=1,2,3…) L.O.T`).
Thank you!
You must keep the multiplicity as 1 even if you want to calculate a triplet because thats the multiplicity of the ground state.
For T1 would be: TD(triplets, root=1) opt
For T2 :TD(triplets,root=2) opt
If you don´t specify "triplets" inside TD the default is "singlets" so for singlets you only need to specify the root.
If you search for "TD gaussian" in google you can check the online manual in options you have every keyword explained
I recommend you to add density=current as well in case you want to check mulliken charges or related for the ex. state
I also don´t recommend to run frequencies in the same calculation as an optimization. I would do it in independent calculations
It is not the same TD(triplets,root=1) and changing the multiplicity to 3 because in the last case you are treating it as a triplet ground state "T0" and not as the first excited state
Fernando Muñoz-Alba, after talking with some colleagues, the procedure that I followed seems correct.
The route line format to calculate the geometry for excited states S1, S2, S3, etc, is:
- # opt freq td(root=1,2,3…) L.O.T
However, when we calculate the optimization for the triplet state, T1 is considered the ground state, but the next triplet states (T2, T3, and so on) should have the keyword td(root=1,2,3…).
For T1, we should change the multiplicity to 3 and not include the "td" keyword, e.g.:
- # opt freq L.O.T
Yes, exactly as a normal optimization for the ground state, but with a multiplicity of the triplets.
For T2, T3, T4... we should include td(root=1, 2, 3...) and set the multiplicity to 3, e.g:
- # opt freq td(root=1,2,3…) L.O.T
obs: LOT = Level of theory, for example: M062X/6-31+G(d,p)
Calculating the T1 using DFT with 3 multiplicity, lacks rigor compared to TD-DFT because it treats T1 as a triplet ground state rather than an excited state.
This approach doesn´t account for the S0→T1 electronic transition. As a result, it provides the minimum energy of the triplet potential energy surface but does not yield the true excitation energy. In contrast, TD-DFT explicitly models T1 as an excited state derived from S0 being more accurate and consistent description.
- Badges
- Science topic