The electron cyclotron resonance (ECR) charge state breeding method based on the PHOENIX source is first dedicated to the continuous high charge state radioactive ion production for cyclotrons, due to their q/m requirements, and to the high charge state production efficiency. It can be used as an injector for RFQ, linacs, and synchrotrons either in cw or pulsed operation. The efficiency of the 1+ to the n+ ion production and the charge breeding time are the fundamental parameters of the method, their dependence with regard to the 1+ ion beam emittance injected has been measured and will be discussed with regard to the radioactive ion beam acceleration. The efficiency of the process slightly decreases when injecting ion beam emittances higher than 30 π mm mrad, and remains constant in the 1+ radioactive intensity range available today and in the future. The ECR ion trap tuning associated with the afterglow mode permits the ionization, the accumulation, and the extraction of the injected primary beam keeping interesting efficiencies suitable to pulsed accelerators.

The latest 28 and 37.5 GHz developments show that very intense pulsed beams of medium charge states can be produced with an ECRIS. We will present latest results obtained with PHOENIX 28 GHz in afterglow mode operation and specially the capacity of the source to produce high current of medium charge states despite a relatively low radial magnetic confinement system. Initially, these developments have been made for Pbn+ ion production for the LHC. A program is now underway by using a very strong ionic pumping effect in order to produce highly bunched beams of radioactive ions suitable for synchrotron acceleration for the CERN Beta-beam project. The purpose of this program consists of producing about 1012 of He6 or Ne18 ions/bunch within 100 μs pulse durations and with a 10 Hz repetition rate. This development must use two ECRIS advantages: the gas efficiency and the bunching capability. We will present a concept showing how we can bunch and ionize the radioactive gas injected in a cw operation source. Experimental data will be presented to illustrate this approach. To match the ionizing time to the time structure of the synchrotron used for the postacceleration we have to imagine a pulsed device using 60 GHz rf power and above; this is the purpose of this presentation.