2024

Luntinen, Miha

Electron cyclotron resonance ion sources (ECRISs) have been pivotal for modern experimental nuclear physics in extending the range of known nuclides on the nuclear chart. In recent years, the use of radioactive ion beams (RIBs) of highly charged heavy ions has become vital for nuclear physics research. Many of the isotopes have half-lives shorter than 1s, and as the half-lives decrease nuclear decay begins to degrade the RIB intensity. ECRISs are used as charge breeding devices for RIBs: The ECRIS is modified to accept an input beam of low charge state ions, produced e.g. by the ISOL-method. The injected ions are captured by the magnetically confined plasma sustained in the ion source, and subsequently ionized to high charge states via multiple collisions with the plasma electrons. The ions exit the plasma stochastically, and a portion of them become available for beam formation and are extracted from the ion source by means of electric fields for post-acceleration. To meet the challenge of producing RIBs with ever shorter half-lives, ECRIS researchers require advanced plasma diagnostic methods to study the charge breeding times and the underlying physical processes. In this thesis, we present the Consecutive Transients (CT) method developed for this purpose. It extends the continuum of transient material injection methods dating back to the1990s. Like its predecessors, the CT-method is based on pulsed injection of material in to the plasma, and study of the timeseries of the extracted beam currents arising as a consequence. It improves the previous methods by implementing a model based data analysis, which minimizes the number of assumptions - importantly, it requires no presupposed model for ion confinement. We describe the derivation of the model, the computational procedures, and the experimental campaigns undertaken to establish and test the method. The plasma electrondensity ne and average energy (Ee), and the characteristic times of ionization Τinz, charge exchange Τcx and ion confinement Τconf are obtained as results. We show that the CT-method is sensitive enough to study variations in these plasma parameters as the ECRIS operating condition is changed, and to identify bottlenecks in the charge breeding process. The hierarchy of Τconf, Τinz, Τcx for highly charged ion production is refined. The uncertainty sources are investigated, and avenues for the improvement of result precision are identified. It was found that ne 5 ~ 1011cm -3, ja (Ee) 100 eV–1 keV. The characteristic times ranged from 1ms to ~ 10 ms. All plasma parameters were found to be chargestate dependent, reflecting the spatial distribution of ion populations in the non-equilibrium plasma.