Mass separation of 225Ac for medical applications: first proof of principle at CERN-MEDICIS

Due to its promise as a medical radioisotope, global demand for 225Ac is beyond the current possible production rate. Several new techniques have already been proposed to increase the production, but further research is required. For its use in treatment, other isotopes, that could potentially be harmful, have to be removed. This purification can be achieved with several methods, but the most promising are chemical separation and mass separation. While the former generally has a higher efficiency, the latter is known for the high isotopic purity that can be achieved with it. Within the scope of this general research lies MED-024, which is an experimental campaign that is investigating the production and mass-separation of 225Ac. In this thesis, the focus will be on characterising the ISOL process efficiency of two specific collections. Both of these used a nitrate solution of 225Ac, with known activity, obtained from the Joint Research center (JRC) in Karlsruhe (Germany). The first run dried the solution on a rhenium foil directly. On the other hand, the second evaporated the 225Ac on a thorium-dioxide felt, where it was assumed that the 225Ac would be distributed homogeneously throughout the porous target material, though without penetrating the target grains. These runs mainly concentrate on the ionization and separation efficiency of 225Ac offered by the laser ionization setup at CERN-MEDICIS. Zinc-coated gold foils were implanted with mass-separated 225Ac and sent to IKS for a detailed activity analysis. Three different techniques were used to measure the activity at different times, which enabled the end of collection activity to be determined by extrapolation. These methods were single g-ray spectroscopy, g-g coincidence measurements and a-spectroscopy. Each of these has its own advantages and disadvantages. For the a-spectroscopy, more accurate geometric efficiency calculations were required. For this, three models were created. The first of these is a semi-analytical method, in which the source distribution is assumed. Once the geometric efficiency is put in terms of an integral that can not be solved analytically, numerical Monte Carlo techniques were used. The other two models (NRAM and FNM) started from a more numerical basis. SRIM was used to model the stopping and redirection of a-particles and recoiling daughter nuclei within and out of the sample. The NRAM uses extrapolation of the particle trajectories in combination with assumptions on probabilities of physical processes once the particles leave the sample. The FNM, on the other hand, models the entire source-detector geometry directly. The resulting collection efficiencies were equal to 9.77(19)%for the first collection and 8.99(48) % for the second. These are only lower bounds to the best achievable efficiency, as the optimization of the setup was performed during the collection and the samples were temporarily removed for mass scans. Furthermore, the second collection had to be ended before saturation was reached. The results from the second run indicate that the effusion efficiency is negligible.