Experience using microfluidic reactors for aromatic fluorination with F-18 fluoride

Objectives To determine if a microfluidics reaction system is practical for the routine production of F-18 radiopharmaceuticals using aromatic fluorination.

Methods Fluorine-18 was prepared by the ¹⁸O(p,n)¹⁸F with 17Mev protons on >95% enriched ¹⁸O-water. It was trapped in a conventional fashion using an ion exchange column, eluted with potassium carbonate/Kryptofix (300 ul of a solution containing 6.8 mg of potassium carbonate, 38 mg of Kryptofix/ml) and dried by azeotropic distillation of acetonitrile. The fluoride was dissolved in DMSO and the precursor, 6-nitroveratraldehyde 6 mgs, in DMSO was run through the microfluidics reactor. Runs were performed using different length reactors, from 8 to 1 meter. All were 100 microns in diameter. Temperatures were kept at185°C. System pressures of ~100psi were maintained. Shorter reactors allowed for a 3x increase in flow. For comparison reactions were also performed directly in the vial used to dry the fluoride

Results Of the fluorine-18 that went into the microfluidics reactor 65-70% was incorporated into the aromatic ring regardless of the conditions. The length of the reactor or the contact time had no effect on the yield. The total reaction time was dictated by the time required to pump the solution through the reactor. However only about 50% of the fluorine-18 fluoride could be loaded into the microfluidics reactor, the rest was left behind in the drying vessel, tubing and valves, so the overall yield was closer to 35%. Direct heating of the precursor and fluoride in DMSO gave yields of ~50% in the same reaction times.

Conclusions The fluorination yield in the microfluidic reactor is a function of the state of the F-18 fluoride not the contact time in the reactor. Fluorine utilization is limited by losses in the system.