It can be seen that the bulk of the space that is asymmetrically discriminated by the BINOL ligand is on the side of the chiral axis that is opposite that where the reaction is taking place at the metal center. Thus, the simple idea was to move the annulated benzene rings from the back of BINOL to the front to give a molecule of the type 151, which we call VANOL. It was necessary to add phenyl groups at the 3 and 3’-positions to allow for resolution. Space filling models reveal that there is a deeper chiral pocket surrounding the metal center in VANOL than in BINOL. Addition of an extra set benzene rings onto the front of VANOL gives 152 (VAPOL). a molecule with an even deeper chiral pocket around the metal center. The names VANOL and VAPOL are derived from the fact that both compounds are members of the family of vaulted biaryls that have annulated benzene rings curving around the nascent active site of the ligand-metal complex.

The syntheses of the VANOL and VAPOL ligands are similar and are illustrated by the synthesis of the VAPOL ligand shown in Scheme XVI [1]. The synthesis begins with the protected phenanthrol 9 which was prepared from the benzannulation of the naphthalene carbene complex 7 with phenylacetylene as indicated in Scheme I. Borrowing from an esoteric reaction, we were able to reduce the acetate and cleave the methyl ether in a single step with aluminum chloride and propanethiol to give the phenanthrol 153 in 81 % yield on scales up to 300 grams. The oxidative coupling of the phenanthrol 153 failed with traditional methods involving metal oxidants, but was successful simply by heating in air.  This was done neat by melting 153 at 190 –210 ^oC in the presence of air in a test tube. On bigger scale, the optimal temperature was 185 ^oC or 365 ^oF.

As indicated in the photograph, Dr. Eugene Grant found that the best method for large scale synthesis was to load a lasagna pan with a thin layer of 153 (40 – 80 g) and bake it in an oven at 365 ^oC for 24 hours. The resolution is carried out in a classical manner involving selective crystallization of diastereomeric salts. The salts were prepared by initial conversion of racemic VAPOL 152 to the cyclic phosphonic acid 154. Crystalline salts of this acid were obtained by the addition of (-)-cinchonidine. The resolution was made even easier by the fact that both diastereomeric salts were crystalline. After the first salt was collected, the solvent was removed and then crystallization gave the second salt. Liberation of VAPOL from each salt gave access to both enantiomers of VAPOL in optically pure form. We have recently found that the resolution can be avoided by employing a deracemization procedure involving copper (II) salts and (-)-spartiene [2].