Rh-catalyzed intramolecular C-H insertion, exemplified by the cyclization of 62 to 63 and 64 to 65 , inherently has many of the features one would look for in a catalytic asymmetric process. In the course of this reaction, an inexpensive, achiral starting material is converted into a cyclic product having versatile and useful functionality. The only by-product is the environmentally-innocuous nitrogen.
Based on our work and the work of many others, we have developed ( Rh insertion) a computationally-based model 66 of the transition state for the C-H insertion . In this model, the dirhodium tetracarboxylate is minimized with ZINDO, and then locked. The point of commitment to bond formation is mimicked by establishing weak bonds between the bonding carbons and between the target H and the proximal Rh, and then minimizing with Molecular Mechanics.
Our conclusion is that at the point of commitment to bond formation, the forming ring is chair-like, with the ester axial, and with the target C-H bond aligned with the C-Rh bond. We believe that bond formation then proceeds by the electron density in the target C-H bond swinging down to form the new C-C bond, and the electron density in the C-Rh bond moving over to form the new C-H bond, releasing the neutral Rh catalyst. We know that the reaction is concerted - insertion proceeds with retention of absolute configuration - but formation of the C-C bond may well precede formation of the C-H bond.
Design of the Chiral Rhodium Catalyst: There are two competing transition states for C-H insertion, 73 and 74. In transition state 73 , insertion is taking place into HA. In transition state 74 , insertion is taking place into the enantiotopic HB. The challenge is to design a chiral rhodium catalyst such that transition state 73 is favored over transition state 74 by at least the 2.5 kcal / mol we have observed is necessary to expect substantial diastereoselectivity in the C-H insertion reaction.
With the computationally-based model in hand, we have been able to design and assess a wide range of chiral Rh(II) carboxylates. Although we had initially thought that chirality at C-2 and C-7 of the C-8 diacid would be important, this turned out not to be the case. Indeed, the diacid that, computationally, is the most effective so far is 75 which only has substituents at C-4 and C-5. What makes the derived complex 76 so effective is that the 1-naphthyl substituents are more stable away from each other, quasi-axial, rather than on top of each other, quasi-equatorial. This places the naphthalene ring in just the right position to interact with the folding substrate on the apical position of the Rh catalyst.
The simple a-diazo ester illustrated can cyclize to an enantiomer which is S at the cyclopentane carbon bearing the ester, or R. Given the syn transition states illustrated, there is still the choice of cyclizing toward the front, away from the naphthalene ring, or to the rear, into the naphthalene ring. The former transition state would give the S product, and so for this discussion is termed the S transition state. The latter transition state would give the R product. Our calculations, using the method outlined above, indicate that the S transition state, with the ligand illustrated, is favored over the R transition state by 2.9 kcal / mol.
For computationally-guided applications of Rh-mediated C-H insertion in natural product synthesis, see: Rh insertion
Home Page: Taber home page.