All the ideas were there to suggest the idea of an insulating tube to focus an ion beam. Pr. Y.Yamazaki (RIKEN, Japan) and its team, members of the collaboration, have first proposed the use of borosilicate capillaries  [3]. Hot stretched over a length of several centimeters, the capillary can be worked to obtain a cone with a 2mm diameter input and less than 1 µm at the exit. However it remains to check that highly charged ions could leave such a “micro-funnel” without modification of its charge number.

The transmission of an Ar⁸⁺ beam and more recently of a Xe²³⁺ one, with energies of a few keV, obtained jointly at RIKEN (Japan) and CIRIL (Caen), proves the feasibility of the process. This appears to be possible by the auto-organized electric loading of the walls of the micrometric insulating tube. In addition, the first experiments show that the transmission remains stable over several hours and is not a transitory phenomenon to be related to the progressive loading of the walls. The increase by one order of magnitude of the beam current density shows finally that the capillary plays an active part, beyond the simple focusing of the beam. It was thus possible to obtain a beam (Ar⁸⁺ at 80 keV) with an initial diameter of about 1 µm (size of the capillary exit). Its divergence is lower than 0.4 milliradian (150 µm diameter at 20 cm of the capillary exit, see figure).

The system already works and is thus applicable to the production of micrometric beams. Remains to be better experimentally and theoretically understood, the finest details of the ion-surface interaction for the selected geometry. Finally, the optical quality of the beam at the exit of capillary would allow additional focusing by more conventional means (electromagnetic lenses) to further decrease the beam size in the nanometric range.