PREPARATION OF ANHYDROUS F-18 FLUORIDE, T. Tewson. Journal of Labelled Compounds and Radiopharmaceuticals S165; 52, Supplement 1 2009

The original specific aims of the grant where cut back considerably as the study section reduced both the time and the budget for the project. The objective of the grant was to show that fluorine-18 fluoride could be prepared completely anhydrous and thus substantially more reactive than conventionally dried fluoride using the method of Sun and DiMagno. This method involved using conventionally dried fluoride to prepare an aromatic fluoride in which the aromatic ring is substituted with electron withdrawing groups. The aryl fluoride is then dried and purified and the fluoride is displaced with an anhydrous nucleophile. Using fluorine-19 and macroscopic scale reactions the reactions work well and give anhydrous fluoride salts that are both more reactive and more selective in their reactions than conventionally dried fluoride. The original substrate chosen for the reaction was bromopentacyanobenzene (1). This compound proved to be easy to make but very hard to purify. As an alternative hexabromobenzene, which is commercially available in high purity, was tried. This reacted cleanly with conventionally dried F-18 fluoride in acetonitrile to give [{sup 18}F]-fluoropentabromobenzene (2), which could be dried by passage of the solution over alumina, which also removed any unreacted fluoride. The fluorine-18 fluoride could be liberated from (2) by displacement with an anhydrous nucleophilic tetra-alkylammonium salt but the anion had to be chosen with considerable care. The reaction is potentially reversible especially as, on the no carrier added scale, there is inevitably an excess of hexabromobenzene and so the displacing nucleophile is chosen to deactivate the aromatic compound to further nucleophilic displacement reactions. To this end tetrabutylammonium azide and tetrabutylammonium phenolate have been tried. Both work but the phenolate is probably the better choice. The F-18 fluoride produced by this process is substantially more reactive than conventionally dried fluoride. A solution of the 3'-anhydrothymidine-5-benzoate (3) was added to the fluoride solution and 30% of the fluoride was incorporated in less than 3 minutes at room temperature were as conventionally dried fluoride requires 10 minutes at 160 C and gives {approx}10% incorporation. These results are encouraging in that they show that the objective of truly anhydrous fluoride is worth pursuing but the problem is that you end up with too much 'stuff' in the solution. Four to five milligrams of hexabromobenzene are used for the initial fluorination reaction and enough of the tetra-alkylammonium salt has to be added to react with a substantial number of those bromides. No attempt has been made to optimize these amounts but there is clearly a lot of material in the solution before the final substrate is added. To avoid these difficulties experiments involving a different, low boiling carrier of the fluoride which can be distilled from the initial fluorination mixture have been tried. Phenyltrifluoromethane sulfonate reacts with fluoride to give trifluoromethane sulfonyl fluoride which boils at -20 C as shown. This reaction works with conventionally dried fluorine-18 fluoride and the no carrier added trifluoromethane sulphonyl fluoride distills out of the reaction as it forms. The choice of nucleophile to react it with to liberate the fluoride is limited and the obvious choice is tetrabutylammonium azide as the resulting trifluoromethane sulfonyl azide is unreactive. We have shown that this works in principle but the experimental details have not been explored