Sub-picosecond, high-brightness electron bunch trains are routinely produced at SPARC-LAB via the velocity bunching technique. Such bunch trains can be used to drive multi-color Free Electron Lasers (FELs) and plasma wake field accelerators. In this paper we present recent results at SPARC-LAB on the generation of such beams, highlighting the key points of our scheme. We will discuss also the on-going machine upgrades to allow driving FELs with plasma accelerated beams or with short electron pulses at an increased energy.
We present the experimental evidence of the generation of coherent and statistically stable Free-Electron Laser (FEL) two color radiation obtained by seeding an electron double peaked beam in time and energy with a single peaked laser pulse. The FEL radiation presents two neat spectral lines, with time delay, frequency separation and relative intensity that can be accurately controlled. The analysis of the emission shows a temporal coherence and regularity in frequency significantly enhanced with respect to the Self Amplified Spontaneous Emission (SASE).
The SPARC LAB complex hosts a 150 MeV electron photo-injector equipped with an undulator for FEL production (SPARC) together with a high power TW laser (FLAME). Recently the synchronization system reached the performance of < 100 fs<sub>RMS</sub> relative jitter between lasers, electron beam and RF accelerating fields. This matches the requirements for next future experiments: (i) the production of X-rays by means of Thomson scattering (first collisions achieved in 2014) and (ii) the particle driven PWFA experiment by means of multiple electron bunches. We report about the measurements taken during the machine operation using BAMs (Bunch Arrival Monitors) and EOS (Electro-Optical Sampling) system. A new R and D activity concerning the LWFA using the external injection of electron bunches in a plasma generated by the FLAME laser pulse is under design. The upgrade of the synchronization system is under way to guarantee the < 30 fs RMS jitter required specification. It foresees the transition from electrical to optical architecture that mainly affects the reference signal distribution and the time of arrival detection performances. The new system architecture is presented together with the related experimental data.
In this paper we discuss the spectra of the electrons produced in the laser-plasma acceleration experiment at
FLAME. Here a <30 fs laser pulse is focused via an f/10 parabola in a focal spot of 10 μm diameter into a 1.2
mm by 10 mm rectangular Helium gas-jets at a backing pressure ranging from 5 to 15 bar. The intensity achieved
exceeds 10<sup>19</sup> Wcm<sup> −2</sup>. In our experiment the laser is set to propagate in the gas-jet along the longitudinal axis to use the 10 mm gas-jet length and to evaluate the role of density gradients. The propagation of the laser pulse in the gas is monitored by means of a Thomson scattering optical imaging. Accelerated electrons are set to
propagate for 47,5 cm before being detected by a scintillating screen to evaluate bunch divergence and pointing.
Alternatively, electrons are set to propagate in the field of a magnetic dipole before reaching the scintillating screen in order to evaluate their energy spectrum. Our experimental data show highly collimated bunches (<1 mrad) with a relatively stable pointing direction (<10 mrad). Typical bunch electron energy ranges between 50 and 200 MeV with occasional exceptional events of higher energy up to 1GeV.
In the SPARC photoinjector, the amplified Ti:Sa laser system is conceived to produce an UV flat top pulse profile
required to reduce the beam emittance by minimizing the non-linear space charge effects in the photoelectrons
pulse. Beam dynamic simulations indicate that the optimal pulse distribution must be flat top in space and time
with 10 ps FWHM duration, 1 ps of rise and fall time and a limited ripple on the plateau. In a previous work
it was demonstrated the possibility to use a programmable dispersive acousto-optics (AO) filter to achieve pulse
profile close to the optimal one. In this paper we report the characterization of the effects of harmonics conversion
on the pulse temporal profile. A technique to overcome the harmonics conversion distortions on the laser pulses
at the fundamental wavelength in order to obtain the target pulse profile is explained too. Measurements and
simulations in the temporal and spectral domain at the fundamental laser wavelength and at the second and
third harmonics are presented in order to validate our work. It is also described a time diagnostic device for the