DESIGN CONSIDERATIONS FOR TABLE-TOP FELs∗

Refinements in laser technology (few-cycle pulse generation, chirped pulse amplification) combined with supercomputer-based plasma simulations have brought the discipline of relativistic laser-matter interaction to a new level of predictability. This was recently demonstrated by the generation of brilliant electron bunches with energies on the 0.1-1-GeV-scale. Our plan is to utilize such laseraccelerated electron beams to realize table-top FELs. The essential feature of those electrons is their ultra-high beam current of up to few 100 kA in 10 fs. Such high currents make small-period undulators realistic, which require less electron energy for the same FEL wavelength. Together with low emittance and relatively large Pierce parameters the undulator length for reaching SASE saturation should be as small as only meter-scales. In this paper we present our first basic design considerations based upon start-toend simulations including 3d PIC codes and GENESIS 1.3. In contrast to large-scale XFELs, which will be dedicated user facilities, our aim is to deliver the proof-of-principle of table-top FELs, starting from the VUV to the X-ray range. The present paper will give a short overview of the basic design-considerations for table-top FELs, while a more detailed manuscript will be submitted to Nucl. Instrum. Methods A. LASER-PLASMA ACCELERATORS The year 2004 marked a breakthrough in the field of laser-plasma accelerators [1]. Three independent groups demonstrated the acceleration of electrons up to relativistic energies with quasi-monoenergetic distributions. There are different mechanisms for laser-plasma accelerators. Here we focus on the so-called “bubble regime”, which was theoretically predicted by one of us (MtV). A laser-pulse with a pulse duration smaller than the plasma wavelength is focused upon a gas jet, where due to its ponderomotive force plasma electrons are kicked away (in transverse direction), leaving an electron-free cavity the so-called “bubble” behind the laser pulse. These electrons return to the axis some micrometer behind the laser, where due to the enhanced space charge electrons are scattered into the bubble. Due to the absence of negative charges inside the bubble, the captured electrons experience a strong electrical field gradient of up to TV/m generated by the inertial positive ion background. Due to the strong acceleration field the necessary acceleration distances can be as small as mil∗Work supported by DFG TR18 † [email protected] limeters. Besides this downscaling in size, bubble acceleration delivers high-current beams. Typically about 10 9...10 electrons are captured into the bubble, as found both experimentally and from scaling laws [2]. As can be seen in Fig. 1 the length of the bubble stem is in the order of a few microns only. Therefore, beam currents of the order of 100kA can be reached. The diameter of the bubble stem allows to realize very small source sizes. Utilizing (discharge) capillaries instead of gas-jets allows longer acceleration distances (cm-scale due to laser guiding beyond the Rayleigh length) and even smaller energy spreads (due to the so-called de-phasing, which causes faster electrons to be slowed down and slower ones to be accelerated) [3]. It is thus expected that the energy spread for 100 MeV is about 2 percent, but 0.2 percent for 1 GeV. By capillaries maximum electron energies reached today are 1.2 GeV [4]. Normalized emittances are as good as from classical accelerators. Figure 1: Snap-shot from PIC simulation of bubble acceleration: electron density map, propagation direction z. The typical length scale is the plasma wavelength, thus micrometers. The “bubble” behind the laser can trap nC charge, thus yielding electron beam currents on the scale of 100 kA. CONDITIONS FOR TABLE-TOP SASE FELS The construction of laser-plasma accelerators as described above clearly allow a table-top electron accelerator. For realizing a table-top FEL one also requires a tableProceedings of FLS 2006, Hamburg, Germany PLT04