Optically active 2,5-disubstituted bicyclo[2.2.2]octa-2,5-dienes (bod) have found use within synthetic organic chemistry as chiral ligands for rhodium asymmetric catalysis reactions. These chiral ligands often provide greater enantioselectivity than their phosphorus-based chiral ligand cousins when employed in the formation of carbon-carbon bonds under rhodium catalysis. The drawback to using these types of 2,5-disubstituted bicyclo[2.2.2]octa-2,5-diene ligands is that they are very expensive to buy if they are commercially available or they must be synthesized if they are not commercially available. Present literature syntheses of these bod ligands rely on the physical separation of diastereomeric derivatives via recrystallization with low recovery or the separation of enantiomers via chiral HPLC, in which only small amounts of material may be separated.
We have devised a synthesis of phenyl, benzyl, and methyl substituted bod ligands based on a bridged Robinson annulation reaction of 1,5-diketones and 1,5-ketoaldehydes, which gives the bicyclic core necessary for the bod ligand. Furthermore, our synthesis is designed to create optically active 1,5-diketones and 1,5-ketoaldehydes which transfer their chirality to the bicyclic core of the molecule. Our synthesis does not rely upon separation of racemic material at any step; instead we provide a method to synthesize any optically active bod ligand that is desired.
We successfully synthesized the chiral 3-allylcyclohexanone (>95% ee), a key intermediate in our synthesis, using a chiral conjugate allylation of α,β-unsaturated β-ketoesters using Cu(OTf)2 and the chiral tBu-box ligand. We then successfully completed the racemic synthesis of 2,5-diphenylbicyclo[2.2.2]octa-2,5-diene (Ph-bod) over 11 steps in 4.4% yield. The chiral Ph-bod ligand is projected to take 14 steps and proceed in a 1.1% overall yield. We also synthesized the racemic 2,5-bis(phenylmethyl)bicyclo[2.2.2]octa-2,5-diene (Bn-bod) over 7 steps in 1.4% overall yield. The synthesis of 2,5-dimethylbicyclo[2.2.2]ocat-2,5-diene (Me-bod) was attempted but was unsuccessful at this time. Ultimately we proved the utility of the bridged Robinson annulation reaction for the synthesis of various bicyclo[2.2.2]octa-2,5-dienes.