Self-Amplified Spontaneous Emission Free-Electron Laser with an Energy-Chirped Electron Beam and Undulator Tapering

We report the first experimental implementation of a method based on simultaneous use of an energy chirp in the electron beam and a tapered undulator, for the generation of ultrashort pulses in a self-amplified spontaneous emission mode free-electron laser (SASE FEL). The experiment, performed at the SPARC FEL test facility, demonstrates the possibility of compensating the nominally detrimental effect of the chirp by a proper taper of the undulator gaps. An increase of more than 1 order of magnitude in the pulse energy is observed in comparison to the untapered case, accompanied by FEL spectra where the typical SASE spiking is suppressed.

Figure 1

Propagation of a field spike (blue) developing at the rear of a chirped $e$-beam (green) for (a) untapered case, and (b) tapered case as a function of $z$. Vertical axes indicate the resonant condition as a function of the relative position between the beam and radiation, $\delta$ is the final detuning in the untapered case. In (a) the slippage leads the spike out of resonance, when the chirp is combined with an appropriate undulator taper (b), the resonance can be preserved.

Figure 2

Single shot spectrum acquired with all undulators set at nominal resonance of 540 nm, with spectrometer entrance slit opened at $100\mu \mathrm{m}$, corresponding to a resolution of $\sim 0.17\mathrm{nm}$. The vertical axis represents the nondispersive dimension at the slit.

Figure 3

Resonant frequency for a 115.2 MeV electron beam in the six undulators, untapered and tapered case (Taper A). Taper B after compensation of the observed wavelength blueshift.

Figure 4

Energy vs relative spectral width for untapered and tapered cases.

Figure 5

Histogram of the pulse energy distribution given by 100 measurements (exp), and by 100 GENESIS 1.3 simulations, where statistical fluctuations of peak current and bunch length (A) and also emittances and energy spread (B) were introduced. The listed beam or undulator parameters were used, except for the slice emittances ($1.4\mathrm{mm}\mathrm{mrad}$) and relative slice energy spread ($1.7\times {10}^{-3}$) which were chosen to match measured mean pulse energy.

Figure 6

Typical FEL spectrum showing a single spike from the experiment (a) and from the GENESIS 1.3 simulation (b). The total pulse energy is $\sim 300\mu \mathrm{J}$.