As it's widely known - reaction of 2nd order nucleophilic substitution at the Si atom (SN2@Si) is believed to proceed in the gas phase via addition-elimination process involving stable, hypervalent Si(V) intermediate (for small substituents attached to Si: H, Me, etc.) resulting in single-well PES for the addition part while approaching nucleophile bonds with the silicon atom.
I was asked lately if there's nothing wrong at the very basis with this picture, because every reaction (and creation of a Si-Nu bond certainly IS a chemical reaction in most chemists' eyes) should (ought to at least) proceed via a transition state. It is true for more sterically demanded substituent at Si atom, e.g. t-Bu ones. Steric Pauli repulsion is then sufficient to increase repulsion energy and to cause additional stable complex and TS to occur.
But should not it be true for small substituents in the gas phase as well? I know the appropriate literature and I am practiced in computational chemistry to a sufficient degree to be aware of the fact that (practically) all computational results indicate that there is NOT such a barrier in that case.
But what's the reason actually? In principle the Pauli repulsion (i.e. overlapping of molecular orbitals of both substituted tetravalent Si(IV) compound and approaching nucleophile) should be quite neglibile but exsisting nevertheless, should't it? The orbitals eventually extend to infinity. On the other side there is no TS without reactants and products complexes' basins. I realize that existing algorithms for finding an energetic minimum on PES operate within a range set by certain copmutational tresholds. Could be that the case?
Is there a possibility that single-well PES is an artefact of our imperfect algorithms and other mathematical tools we use (e.g. incomplete fuctional basis set) and other approximations we made? (I don't think in particular about Born-Oppenheimer approximation here because I realize that B-O. approx. is necessary for being able to talk about PES whatsoever).
If there's yet a serious quantum physics standing behind that phenomenon I would be grateful for everyone who could explain it to me. Thank you all in advance.
Mr. Gostyński,
Which theoretical approach you have used studying TS?
It was DFT - B3LYP/6-31+G(d). But I don't think this is the case - I get the same results using MP2. I'm aware of inefficiencies of these methods and relatively small basis set.
The question more relevant to me is somewhat more abstract: during a coordinate scan (Si-F- distance) for backside attack H3SiF + F- the PES takes form of 'classical' hook curve as in distance squeezing/expanding for Lennard-Jones potential. An analogous minimum can be obserwed when one substitutes Me for H (although I used a much longer scan step). These are typical examples of reactions occuring without energy barrier.
When one allows fluoride anion F- to side attack (i.e. approach on the edge between two methyl groups) only then Pauli repulsion causes appearance of energy barrier. For t-Bu TS exists even for a backside attack.
And there comes my inquiry regarding two questions:
1) is there an "in-principle" possibility that a negligible TS (more precisely: an energy increase due to Pauli repulsion or other factors) could occur somewhere "in-between" while scanning the F-Si distance for H and Me as substituents and we aren't just able to "catch" and visualize it or:
2a) there's no TS due to the fact that even the weakest attractive dispersion/van der Waals forces are able to overcome the negligible Pauli repulsion and the whole curve takes the shape of L-J potential, or:
2b) the interplay between the electronic factors unfold just the way it does and there is virtually NO Pauli repulsion or any other factors that could cause the energy barrier to appear? (No Pauli repulsion is in my opinion impossible because of what I said before about spatial expansion of orbitals and of the fact that the mathematical form of L-J potential encompasses Pauli repulsion.)
Wouldn't the 2a-b) statements be tantamount to farewell to claiming that every reaction must (at least in principle) go through a transition state?
If the statement 1) should be yet valid, where is in my case located the reactants' basin to connect it through TS with the basin of products? There is none, since H3SiF + F- posess a stable minimum only for products - at equilibrium distance for my calculated "L-J" potential. (A possible explanation could be above-mentioned imperfection of our computational algorithms and methods.)
One of these statements is in my opinion untenable - whether be it the existence of TS for each chemical reaction (I acquiesce to that view point, having mentioned earlier reactions without TS), whether the correctness of results obtained (due to lack of precision or some other factors - question is open) - in such a case a minimal, even negligible TS and reactants complex sould exist.
The flashpoint is - which one view is correct and what's the explanation for it?
Mr. Gostyński,
Looking at your posting I have understood that you have used keyword" 'Scan'.
The topic of your quetion, however, requires to examine CT effect, furthermore there can be mechanism with more than one CT effect.
In this context you should use the following theoretical approaches:
I.
1.1. IRC
1.2. AIM population per IRC step
II.
2.1. ADMP
2.2. AIM analysis per ADMP step
So when you obtain the relations energy per step (or time in trajectory) you should compute AIM populations per each of the steps (It is shown as attachment).
There is needed to account for the electron excess on both F-atoms in your case, and the electronic populations on Si and 3Hs.
Ok, I'll try out the approach you suggested. Thank you for your cooperation and advice.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Quantum Physics