Recently I'm trying to get vibrationally-resolved electronic spectra for a series of molecules using 3-step methods described by "Vibrationally-resolved electronic spectra in GAUSSIAN 09 by V. Barone, J. Bloino and M. Biczysko",But it seems that the calculation fails in 90% cases showing the following error messege:
ERROR: The Franck-Condon factor corresponding to the overlap integral
between both vibrational ground states is too small:
|
The method is best used for medium-sized organic molecules with high symmetry. For other cases, the error that you mentioned can happen quite often.
In short, it appears that you use very small molecules which tend to differ qualitatively in terms of their minimum geometries and the associated Hessians. Consequently, the Duschinsky matrix might contain a lot of off-diagonal elements, which will spoil you calculation and/or the two minima will just be so far away from each other in orthonormal space that the harmonic approximation breaks down.
You can try a non-Duschinsky time-dependent implementation to enforce your spectrum, by using, e.g., the ORCA package (orca_asa application). The underlying physics might however just be unsuited to be well described in harmonic approximation, so the results might be meaningless.
The method is best used for medium-sized organic molecules with high symmetry. For other cases, the error that you mentioned can happen quite often.
In short, it appears that you use very small molecules which tend to differ qualitatively in terms of their minimum geometries and the associated Hessians. Consequently, the Duschinsky matrix might contain a lot of off-diagonal elements, which will spoil you calculation and/or the two minima will just be so far away from each other in orthonormal space that the harmonic approximation breaks down.
You can try a non-Duschinsky time-dependent implementation to enforce your spectrum, by using, e.g., the ORCA package (orca_asa application). The underlying physics might however just be unsuited to be well described in harmonic approximation, so the results might be meaningless.
- Badges
- Science topic