Towards highest peak intensities for ultra-short MeV-range ion bunches

Laser-based ion acceleration as a source for intense, MeV-range ion bunches is discussed for many possible applications: in the context of fusion science1, the creation of warm dense matter2,3,4 or as diagnostic tool5,6,7 as well as for medical applications8,9. A well understood and widely-used mechanism for laser-based ion acceleration is the TNSA (target normal sheath acceleration10,11). Typically accelerated protons show excellent beam properties with respect to bunch intensity and emittance12 and also the feasibility of efficiently accelerating heavier ions could be demonstrated experimentally13. However, the beam suffers from a large divergence and continuous broad energy spectrum, while for most applications a collimated bunch with defined energy spread is necessary. First promising results in beam shaping could be achieved via the application of pulsed solenoids14,15, permanent magnetic quadrupoles16,17 or laser-triggered microlenses18. For the also necessary manipulation of the longitudinal bunch dynamics, injection into a synchronous radiofrequency (rf) field yields high potential and a first conceptual demonstration was performed in Japan19,17. As interest in such novel beamline concepts arises20,21, the German national collaboration LIGHT (Laser Ion Generation, Handling and Transport22), has built a test beamline at the GSI Helmholtz center for heavy ion research as the central part of the collaboration’s agenda. This beamline exploits the TNSA mechanism to provide a very compact proton source with energies currently reaching up to 28.4 MeV. The acceleration is driven by GSI’s PHELIX laser (Petawatt High Energy Laser for Ion eXperiments23), which is focused at a laser intensity of 5 × 1019 W/cm2 onto a thin metal foil target (typically 5 or 10 μm thin gold or titanium foils). From the continuous and highly divergent source spectrum a specific energy can be selected and collimated via a pulsed high-field solenoid, which can be operated at up to 9T field strength. The obtained source parameters are in the typical range for TNSA experiments within the given laser and target parameters and it is possible to accelerate more than 1012 protons above 4 MeV energy in total with about 1010 protons in a 1 MeV energy bin around 10 or 8 MeV. Via chromatic focusing with the pulsed solenoid, up to one third of the protons in such an energy bin can be captured and transported through the beamline within a collimated bunch with still relatively large energy spread (about 20% FWHM). Although thus the overall capture efficiency is on the sub-percent level, still large single-bunch particle numbers above 109 can be created. However, the beamline is routinely not operated at its limit and also typical shot-to-shot fluctuations of up to a factor of 2 in particle numbers are observed. This first step of the experiment is described in detail in24 and an illustration given in Fig. 1. Figure 1 Shown are the accesssible source parameters via TNSA (energy-dependent proton distribution function and half opneing divergence angle) at the used experimental area Z6 at GSI Darmstadt. The next step has been the longitudinal phase rotation of the bunch via applied electrical fields within a rf cavity, running at 108.4 MHz and providing a total electrical potential of more than ±1 MV. Injection of the bunch at a synchronous phase of ΦS = −90 deg leads to a rotation around the central energy in longitudinal phase space and at a certain rf input power to an efficient energy compression of the bunch; less than 3% energy spread could be achieved in a previous experimental run25 for protons at 9.4 MeV energy and particle numbers larger than 109. With increasing rf power the bunch can also be ‘over-rotated’ in phase space, leading to a situation of a well-ordered energy distribution within the bunch with the slower particles at the front and the faster particles at the back. Along a further propagation length, the faster will catch up with the slower particles and at one specific distance a minimum in the bunch length will be reached. The mechanism is called phase focusing and is illustrated in Fig. 2 together with the alternative operation mode for energy compression. While the latter was already demonstrated in a previous run in 2013, finally the phase focusing could be experimentally accomplished recently. This completes the initial commissioning phase of this novel laser-driven ion beamline, available now at GSI and representing the focus of this paper. Figure 2 Illustration of phase rotation in longitudinal phase space via applied rf. The comparative simulations are performed with the TraceWin code from cea26 and use beam parameters, that are adapted to the experimental findings to most precisely model the experiment. The specific parameters used here will be discussed later in context with the experimental results.