I am trying to calculate reorganization energy for rISC between T1 and S1. In order to perform T1 optimization in gaussian 16w do I use ground state method or TD-SCF?
You need first to determine the ground state (GS), singlet or triplet.
1. Do GS (opt/freq jobs) for both singlet and triplet and you shall be able to compare energies and properties.
2. From GS, either S0 or T0 (but not both), the first singlet S1 or the first triplet T1 excited states can be calculated. Even though unrestricted GS can do this, formally one should use TD methods.
Observe that if the GS is S0, T0 does not exist, and vice versa.
Thank you for your response. I already optimized the molecule at ground state by keeping spin= singlet. I will now calculate S1 and T1 states from this GS. But I am unsure about the next steps, especially T1 optimization. Some people are saying for T1 optimization my method should be ground state in gaussian while others are saying I can go for td-scf method and specify "state=triplets only". Could you clear out this confusion?
If you optimize a triplet state using unrestricted GS, this state may be probably T1 because a GS method will try to find the lowest electronic state.
That's OK, and it works fine when T1, T2, ..., Tn are well separated.
However, formally, one should use TD to calculate Tn, n = 1, 2, ..., n.
T1 calculated with unrestricted GS could be equivalent to T1 calculated with TD, but this is not always true and needs verification, particularly if the levels of triplet states are not well separated like in large molecules or organimetallic complexes.