Multiwfn forum

may01dz

Member

Registered: 2018-07-17

Posts: 58

Interpretation of hole-electron analysis results using ChatGPT

I wrote the prompt below in ChatGPT:

So I want to determine whether transition number 1 is a π–π*, σ–π*, or n–π* transition using the hole-electron analysis framework described in the attached paper ( "Liu et al. - 2020 - An sp-hybridized all-carboatomic ring, cyclo[18]carbon Electronic structure, electronic spectrum, a.pdf").
I used the Multiwfn program to obtain the parameters needed to identify the nature of this transition, which are shown in the table below.
Integral of hole:        0.960000
Integral of electron:    0.959942
Integral of transition density:    0.000039
Transition dipole moment in X/Y/Z:   0.428149   2.521187  -2.158261 a.u.
Sm index (integral of Sm function):   0.45245 a.u.
Sr index (integral of Sr function):   0.69978 a.u.
Centroid of hole in X/Y/Z:        7.818786    0.960151    2.729048 Angstrom
Centroid of electron in X/Y/Z:    7.534434    0.737712    2.573506 Angstrom
D_x:   0.284  D_y:   0.222  D_z:   0.156    D index:   0.393 Angstrom
Variation of dipole moment with respect to ground state:
X:    0.515837  Y:    0.403521  Z:    0.282165    Norm:    0.713116 a.u.
RMSD of hole in X/Y/Z:       1.365   1.709   2.496   Norm:   3.319 Angstrom
RMSD of electron in X/Y/Z:   1.173   1.932   2.114   Norm:   3.095 Angstrom
Difference between RMSD of hole and electron (delta sigma):
X: -0.191  Y:  0.223  Z: -0.382    Overall: -0.224 Angstrom
H_x:  1.269  H_y:  1.821  H_z:  2.305  H_CT:  1.654  H index:  3.207 Angstrom
t index: -1.261 Angstrom
Hole delocalization index (HDI):       7.27
Electron delocalization index (EDI):   7.41
Ghost-hunter index:    -33.291 eV, 1st term:  3.340 eV, 2nd term:    36.631 eV
Excitation energy of this state:     3.091 eV
Based on the attached paper and the data in the results table, are there clear indicators of the nature of transition 1?
Note that transition 1 is characterized by the following:
STATE  1:  E=   0.113590 au      3.091 eV    24930.0 cm**-1 <S**2> =   0.000000 Mult 1
    74a ->  75a  :     0.959961 (c=  0.97977605)
And the response was as follows.

Based on the data you provided and the criteria described in Liu et al. (2020) for hole-electron analysis, we can now interpret the nature of transition 1 (0–1A → 1–1A) using quantitative descriptors:

Key Descriptors from Your Multiwfn Output:

Integral of hole / electron
≈ 0.96
Very good quality, near-conservation of charge.

Integral of transition density
0.000039
Very small ⇒ transition is not charge-transfer (CT).

D index (distance between hole and electron centroids)
0.393 Å
Short ⇒ local excitation.

Sm index
0.452
Indicates strong spatial overlap.

Sr index
0.699
Also consistent with non-CT excitation.

Δσ (delta RMSD)
-0.224 Å
Small, no major shape difference.

Hole & Electron Delocalization Indices (HDI/EDI)
~7.3 / 7.4
Strongly delocalized ⇒ characteristic of π orbitals.

t index
–1.261 Å
Negative ⇒ not typical charge-transfer behavior.

Transition dipole vector norm
3.31 a.u.
Large, consistent with intense π–π* transition.

Excitation energy
3.091 eV (401.1 nm)
In the visible–near UV range.

Dominant configuration
74a → 75a with coefficient ≈ 0.98
Clean single excitation between adjacent orbitals.

Interpretation Using Hole-Electron Analysis Framework:
Liu et al. emphasize that π–π* transitions:
Occur with strong overlap between hole and electron regions (high Sm, moderate-to-high Sr).
Exhibit short D index and delocalized densities.
Show large oscillator strengths and transition dipole magnitudes due to strong spatial overlap.
Involve delocalized MOs such as πin or πout in conjugated systems.
By contrast:
n–π* transitions generally show more localized hole, small HDI, larger D and smaller oscillator strength.
σ–π* transitions usually originate from localized σ orbitals with small delocalization and different energy levels.

Conclusion:
All numerical evidence (short D, large Sm, strong delocalization indices, clean HOMO→LUMO excitation, high transition dipole moment) points toward:
Transition 1 is a strongly allowed π–π* excitation.
There is no indication of n or σ orbital involvement (e.g., low HDI, large D, weak overlap, or low fosc), which further excludes n–π* or σ–π* character.

How accurate were its answers and interpretations of the results, and can I rely on them?

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 2,030

Website

Re: Interpretation of hole-electron analysis results using ChatGPT

Hello,

I'm not willing to make any comment on the output generated by any AI (this is a complete waste of my time). Sorry...

may01dz

Member

Registered: 2018-07-17

Posts: 58

Re: Interpretation of hole-electron analysis results using ChatGPT

Thank you for your response. I understand your position regarding AI-generated content. My intention was to validate and discuss the scientific interpretation, not the tool itself. I appreciate your time either way.