stereoselectivity in additions to carbonyl groups

a presentation by Rudy Kohn
and Sheila Wang

Stereoselectivity in carbonyl reactions is often handled in two ways: with explanations based on the formation of a chelate (a claw-like chemical formation, figure 1), or with an appeal to a Felkin-Anh transition state (figure 2).

A chelate is basically a not-really there 5-membered ring in which a metal on a Grignard reagent (that's a Magnesium alkylate) interacts with the double-bonded oxygen and a fairly strong nucleophile three bonds away. Felkin-Anh, so far as we can tell, is just a transition state in which the nucleophile and metal interact with the two ends of the carbonyl at the same time.

These ideas are important because they can help determine the stereochemistry of additions. Chelation is best for telling us what side will be less hindered for attack because the atom keeps its stereochemistry in the rigid chelate form.

In figure 3, this is shown. The nucleophile can only attack from the 'bottom' of our view. The R1 group hinders a topside attack.

In fact, one of the compounds in this article:
Carda, M.; Castillo, E.; Rodriguez, S.; Gonzalez, F.;
Marco, J. A. Tetrahedron Asym., 2001, 12, 
1417-1429.

And when the reaction shown below was done, the Si addition isomer was greatly favored. Why? The group that hangs off behind the chelating group in the reactant hinders many of the Grignard reagents' attempts to attack from the Si side. The ratio of this reaction was 33:67.

Felkin-Anh is based on a reaction in which the metal and nucleophilic group both interact with the ends of the carbonyl bond simultaneously. They come together so that the orbitals of the newly formed sp3 carbon do not eclipse with the already existing bonds, or so that the largest substituent on the neighboring carbon is anti to the addition.

This is another way to control, or at least explain, stereochemistry. Addition on a carbonyl tends to follow this trend in any case; it is the counter-effects which make the other isomers appear in general.  Chelation works in opposition to this.

In this reaction, the Re attack is favored for the reason that it pushes the oxygen forward and there is no eclipsing. The Re attack ends up being favored 4:1, due to Felkin-Anh interactions.

Here is a look at both of our reactions.  The top reaction is our example of chelation, and the bottom process is an example of Felkin-Ahn stereoselectivity.