I observed bond breaking during the optimization of negatively charged molecules using various computational methods. This issue was persistent across different levels of theory. I am seeking assistance from the community to understand the potential causes of this phenomenon and possible solutions to obtain stable optimized structures for negatively charged molecules.
For example, optimization of the 2,3,5-Trichlorothiophene molecule with a charge of -1 in DFT B3LYP 6311G++(d, P) resulted in broken bonds, I have attached the image of optimized structure.
There is a multitude of options here:
a) If that is an actually optimized structure and not a saddle point (verifiable by calculation of vibrational frequencies), I would not rate this a "bond breaking" but more like an elongation of a bond. The bonds which most editors draw are exclusively assigned from interatomic distances and do not come from bond order analyses (unless you tell me explicitely that you've used special tools to do so).
b) Assuming it's an actually broken bond verified by AIM or something like that (i) what was your multiplicity? Uncharged it would be a singlet for sure, did you model this as a doublet? (ii) is the negative charge localized on the dissociated Cl? It would at least make sort of sense that one Cl breaks off heterolytically as Cl- leaving behind a neutral (radical) molecule.
"I observed bond breaking during the optimization of negatively charged molecules using various computational methods."
No, you didn't, as there are no bonds in QM methods. You did observe that one of the C-Cl bonds got a bit longer. This is an artefact stemming likely from the fact, that 2,3,5-Trichlorothiophene is a neutral molecule and not an anion. If you define the system so that it has some strange charge/spin-state, you will likely get strange results.
I think Jürgen Weippert brought to you very good options.
The first thing I should do is to measure the distance between the chlorine and carbon. Perhaps it is just an elongation of the bond as Jürgen Weippert said. If not, check the charge and also the multiplicity.
After this, if all is right until this point, run an Opt+Freq calculation, and check the frequencies you have. If there is any of them resulting in a negative value, that means your molecule is a transition state.
If you are working on that, you can also perform a IRC method just to know the reaction coordinate.
Bests,
- Badges
- Science topic
- Similar topics
- Chemistry, Organic
- Organic