I'm studying a mechanism of fluorine substitution in my molecule (the project isn't that much about actual research as learning computational chemistry). There are two $\ce{C6F5}$ rings and I want to know which one will be substituted first (I have more or less consistent experimental data for N-nicleophiles).

  • Two sides of each ring aren't equivalent, should I study nucleophile's approach each ring from each side (4 calculations in total)?

I assume a classical mechanism via a Meisenheimer complex, but when I try to optimize its structure it just expels the nucleophile (returns to reagents).

So I run relaxed PES scan, bringing closer the nucleophile ($\ce{Me2NH}$) to $\ce{C6F5}$ ring, optimize the maximum as TS and compute its vibration. I get only one imaginary vibration (as intended) but it corresponds to $\ce{C-F}$ bond breaking, not $\ce{C-N}$ bond forming (which I'm not certain about).

  • Can I make a conclusion from it, that reaction proceeds without an intermediate (I've found such examples in the literature) and has late TS (although the reagents have the energy higher than the products, which should produce early TS)?
  • Is Hammond's postulate even applicable for an arbitrary non-model reaction?
  • $\begingroup$ +1 But I had to make some edits to get your chemical notation presented properly, can you take a look and do the same for the rest? Also your second question about what level of theory is accurate enough ought to be a second question, as it is completely different from the question of whether or not you can draw that conclusion about proceeding without an intermediate. $\endgroup$ Jan 5, 2021 at 8:26
  • 1
    $\begingroup$ @NikeDattani thank you for the edits, I didn't know \ce command from latex works here. The question about theory level really was not related to general methodology. $\endgroup$
    – Roman
    Jan 5, 2021 at 8:37
  • $\begingroup$ I've finished with changing functional from B3LYP to M06-2X as it gives me the intermediate that I'm searching for. Also there is a paper about B3LYP failing to predict some mechanisms doi.org/10.1039/C2CP40438A . $\endgroup$
    – Roman
    Jan 25, 2021 at 14:03
  • $\begingroup$ I see three questions there. Is it possible for you to reduce it down to just one question? Multi-question posts are extremely discouraged because they can become hard to deal with when someone has an answer to 1 of the questions but not the other two, and in other circumstances too. You can ask 2 or 3 separate questions if you want to, but it would be nice if you could reduce this post down to just one question. $\endgroup$ Feb 28, 2021 at 4:58
  • 1
    $\begingroup$ @NikeDattani I've almost completely figured it out by myself during this two months, will it be better to render all my experience and findings in an answer? $\endgroup$
    – Roman
    Feb 28, 2021 at 6:50


Browse other questions tagged .