What to do when Transition State Theory fails ?

aitouna

Hello Dear all,
an organic reaction that yields experimentally a four products (one pair of regioisomers and one pair of stereoisomers). Four Transition States were located at three different level of theory, B3LYP, M06-2X and wB97XD along with the basis set 6 -311G(d,p). The calculation of the rate constant k and find the products ratio gave a disagrrement with the experimental results for the three DFT methods.
I'm disappointed, and i don't know what to do in this situation.
Please, don't hesitate to offer your advices, and accept my sincere thanks.

wzkchem5

aitouna 发表于 2022-4-13 16:12
Dear wzkchem5,
First, i thank you for your prompt and informative reply, and I added diffuse func ...

The reactants and transition states of a reaction can adopt many different conformations. If you haven't done a conformational search, but rather picked an arbitrary conformation, then there is a considerable chance that the conformation that you chose is not the lowest energy (or more accurately speaking, lowest Gibbs free energy) conformer, which means that you will overestimate the Gibbs free energy of this species. The chances of getting the correct conformer may improve somewhat if you chose conformations from experimental crystal structures, for example, but there is still no guarantee that the conformers you chose are the correct ones. The only way to be reasonably sure is to do a conformational search, using programs like Molclus or CREST.

aitouna

wzkchem5 发表于 2022-4-12 22:47
What you said does not necessarily mean that transition state theory fails for your system. There ar ...

Dear wzkchem5,
First, i thank you for your prompt and informative reply, and I added diffuse function only to heavy atoms, and i used solvation model : PCM, the result is a change in energies but the interpretation still the same.
I didn't quite understand the point (1) "you chose the wrong conformation", could you clarify more, please.

wzkchem5

What you said does not necessarily mean that transition state theory fails for your system. There are many other possible causes, like: (1) you chose the wrong conformation; (2) your calculation requires diffuse functions for the result to be qualitatively correct; (3) your system has multireference character, so that most density functionals yield qualitatively wrong results; (4) you did not explicitly solvate some functional groups that require explicit solvation for a correct description; and (5) your proposed mechanisms are wrong.
If you have solid evidence that you have taken due care in all of the above aspects and that the transition state theory itself must be the sole source of error, then you should do variational transition state (VTST) calculations, probably with tunneling corrections, with programs like PolyRate. If some of the reaction barriers are extremely low or if reaction path bifurcation is present, then an AIMD simulation may be necessary. But to stress again, it is much more probable that the problem is not about transition state theory, but one or more of the five points I have mentioned.