Femtosecond Compression Dynamics and Timing Jitter Suppression in a THz-driven Electron Bunch Compressor

We present the first demonstration of THz driven bunch compression and timing stabilization of a relativistic electron beam. Quasi-single-cycle strong field THz radiation is used in a shorted parallel-plate structure to compress a few-fC beam with 2.5 MeV kinetic energy by a factor of 2.7, producing a 39 fs rms bunch length and a reduction in timing jitter by more than a factor of 2 to 31 fs rms. This THz driven technique offers a significant improvement to beam performance for applications like ultrafast electron diffraction, providing a critical step towards unprecedented timing resolution in ultrafast sciences, and other accelerator applications using femtosecond-scale electron beams.

Figure 1

(a) Schematic of the in-vacuum components for the THz driven compression and streaking setup. Within the PPWG compressor, coupling between the electron beam and THz pulse reverses the beam’s energy chirp, resulting in velocity bunching during the subsequent drift. (b) Snapshots of the simulated longitudinal phase space (electron energy $\gamma$ vs $z$ position) are shown for the beam (i) entering the compressor, (ii) exiting the compressor, and (iii) at the streaking slit. In the slit, the second THz pulse imparts a transverse momentum kick to the beam which streaks the longitudinal profile onto the $y$ axis on a downstream imaging detector. THz source 1 is polarized along the $z$ axis; THz source 2 is polarized along the $y$ axis.

Figure 2

(a) Diagram of the compressor cross section, with incoming THz pulse polarized along the $z$ axis. (b) Close-up view of the interaction region, showing the $250\mu \mathrm{m}$ diameter beam tunnel and the $210\mu \mathrm{m}$ PPWG gap. A short reflects the incoming THz pulse. (c) Photograph of the assembled structure showing the adiabatic horn opening. (d) Photograph of the disassembled structure before the final etch. The structure is translated along the $y$ axis to switch between compressor “on” with beam passing through the beam tunnel and compressor “off” with beam passing through the 2.54 mm diameter clearance hole. A GaP crystal was used for in situ electro-optic (EO) sampling of the THz field.

Figure 3

Comparison of the uncompressed beam to the maximally compressed beam. Single shot beam images in (a) and (c) show examples with THz driven compression off and on, respectively, with the corresponding fit to their projected distribution shown in (b) and (d). (e)–(f) show the rms bunch length of the uncompressed beam over 1000 shots, with a mean of 105 fs and standard deviation of 19 fs. (g)–(h) The rms bunch length of the compressed beam over 1000 shots, with a mean of 39 fs and standard deviation of 7 fs. (i)–(j) The time of arrival of the uncompressed beam over 1000 shots, with a timing jitter of 76 fs rms. (k)–(l) The time of arrival of the compressed beam over 1000 shots, giving a reduced timing jitter of 31 fs rms.

Figure 4

(a) Scan of the bunch length and timing jitter of the electron beam, along with the deflection angle in the $x\text{−}z$ plane (scale shown on right axis), as a function of the THz pulse arrival time within the compressor structure. The linewidth of the deflection data reflects the standard deviation of deflection angle. (b) EO sampling within the PPWG structure, characterizing the THz field in the unshorted region. The ringing apparent after $\sim 6\mathrm{ps}$ is an artifact from the EO crystal, not the true waveform.