w-tuning for n-th excited state

wham09 · 2026-01-28 02:42:36

Dear Prof. Lu,

I am currently trying to conduct electron excitation analysis for the 5th excited state of my molecule.

The 5th excitation is characterized by charge transfer from alpha-HOMO to alpha-LUMO+3.

If I wanted to do w-tuning of long-range-corrected functional for this state, should I then look at N and N-1 (for HOMO), and separately, N+7 and N+6 (for LUMO+3)?

sobereva · 2026-01-29 18:50:41

No. The electronic states involving in w-tuning are independent of the excited state(s) of interest. You should follow the standard way of tuning the w parameter, and then the w parameter is used for calculating all excited states.

wham09 · 2026-01-30 00:33:07

Thank you for the answer.

I’d like to ask just one follow-up question.

The ground state of my molecule is doublet D0, and the first excited state (D1) is characterized by alpha-HOMO (30A) to alpha-LUMO (31A), while beta-HOMO stays 29B.

In this case, when I calculate N, N+1, and N-1 states for w-tuning, should I set the N+1 state as triplet? In this way the N+1 state will have the 31A populated.

sobereva · 2026-02-01 23:26:41

It is seemingly reasonable, but I am not sure if the result will be better than tuning w with only considering N and N-1 states, especially when other excited states are also to be studied.

wham09 · 2026-02-09 07:31:37

Dear Prof. Lu,
I have two questions about TD calculation, and since one of them is closely related to this thread, I’ll leave these here.

1. When the excited states of interest are local excitation states, is w-tuned LRC-functional still more ideal than PBE0?

2. If I wanted to look at electronic structure of a geometry-optimized excited state in solution (by extracting natural orbitals from density=current TD calculation), should using SMD model yield more realistic electronic structure than PCM, at least in principle?

sobereva · 2026-02-09 16:20:26

1 I cannot say it must be better, but w-tuned LRC has high probability better than PBE0 in this case.

2 No. The polar part of SMD is just PCM, and only polar part affects excitation energies and electronic structure of excited state.

wham09 · 2026-02-10 04:24:40

About the second question, could you please clarify a little more? I understand that the “change” of energy and electronic structure from the ground state is only affected by the polar part of solvation model. But, isn’t the reference electronic structure of the ground state still affected by the nonpolar part of solvation? For example, the electron density curve of fluoride anion is different when PCM or SMD model is used, although not much. If the ground state electronic structure is better represented by SMD model, shouldn’t this be consistently applied for the corresponding excited state?

sobereva · 2026-02-10 10:38:32

Nonpolar part of SMD only contributes to electronic energy, you will understand this point if you read original paper of SMD. Moreover, it is independent of actual electron distribution but only dependent of coordinate, so nonpolar contributions to ground state and excited state energies are exactly the same, namely it doesn't contribute to vertical excitation energy, and doesn't affect electronic structure.

Note that the polar part of SMD is slightly different to PCM, because SMD defined a set of atomic radii focusing on reproducing experimental molecular solvation energy, these radii are commonly different to the default atomic radii used by the PCM in quantum chemistry codes under the default setting.

Multiwfn forum

#1 2026-01-28 02:42:36

w-tuning for n-th excited state

#2 2026-01-29 18:50:41

Re: w-tuning for n-th excited state

#3 2026-01-30 00:33:07

Re: w-tuning for n-th excited state

#4 2026-02-01 23:26:41

Re: w-tuning for n-th excited state

#5 2026-02-09 07:31:37

Re: w-tuning for n-th excited state

#6 2026-02-09 16:20:26

Re: w-tuning for n-th excited state

#7 2026-02-10 04:24:40

Re: w-tuning for n-th excited state

#8 2026-02-10 10:38:32

Re: w-tuning for n-th excited state

Board footer