Generation of a femtosecond electron microbunch train from a photocathode using twofold Michelson interferometer

The interest in producing ultrashort electron bunches has risen sharply among scientists working on the design of high-gradient wakefield accelerators. One attractive approach generating electron bunches is to illuminate a photocathode with a train of femtosecond laser pulses. In this paper we describe the design and testing of a laser system for an rf gun based on a commercial titanium-sapphire laser technology. The technology allows the production of four femtosecond laser pulses with a continuously variable pulse delay. We also use the designed system to demonstrate the experimental generation of an electron microbunch train obtained by illuminating a cesium-telluride semiconductor photocathode. We use conventional diagnostics to characterize the electron microbunches produced and confirm that it may be possible to control the main parameter of an electron microbunch train.

Figure 1

Schematic diagram of a laser system based on chirp-pulse amplification.

Figure 2

Schematic of a method to produce four laser pulses.

Figure 3

Top row: Schematic of a femtosecond single-shot autocorrelator for the cases where (a) one and (b) two laser pulses interact in a potassium dihydrogen phosphate (KDP) crystal. Bottom: The examples of measured autocorrelation dependencies.

Figure 4

Schematic layout of the LUCX facility showing the locations of the quadrupole magnets, bending magnet and diagnostic screens.

Figure 5

Typical transverse spatial distribution of the UV laser pulse on the virtual photocathode with the associated projection (red solid traces) and the Gaussian fit functions (green dashed curves).

Figure 6

Top row: The typical electron density distribution for the different magnet currents corresponds to: (a) 7.63 MeV, (b) $7.50\mathrm{MeV}$, and (c) $7.30\mathrm{MeV}$, respectively. Bottom: Beam image centroid position against electron beam energy (red dots) with a linear fit function (red solid curve): amplitude $7.82\pm 0.03\mathrm{MeV}$ and slope $3.18\times {10}^{-2}\pm 2.65\times {10}^{-3}\mathrm{MeV}/\mathrm{mm}$; beam image centroid against the rf phase of acceleration structure (blue triangles) with a linear fit function (green dashed curve): slope $0.11\pm 5.44\times {10}^{-3}\mathrm{rf}\mathrm{deg}/\mathrm{mm}$.

Figure 7

Examples of the microbunch electron density distributions in a train measured on the Ce:YAG screen while the linac rf phase was set to zero-crossing phase. The acceleration phase of the rf gun is 20 rf deg and time delay determined by movable mirror No. 2 is equal to 0.7 ps. Images (a), (b), (c) and (d) respectively correspond to the time delay between pairs of laser pulses equal to 1.20, 1.10, 0.88 and 0.66 ps.

Figure 8

Measured microbunch horizontal spacing between the electron pairs of bunches (red dots) as a function of the time delay between pairs of laser pulses with the linear fit function (green solid curve): slope $123.86\pm 16.26\mathrm{fs}/\mathrm{mm}$. The blue triangular dots correspond to simulations carried out with astra.

Figure 9

Examples of the microbunch electron density distribution in trains measured on the Ce:YAG screen while varying the acceleration phases of the rf gun: (a) 25, (b) 15 and (c) 10 rf deg. Note that time delays determined by movable mirrors No. 1 and No. 2 are 0.66 and 0.7 ps, respectively.

Figure 10

Examples of the microbunch electron density distribution in trains measured on the Ce:YAG screen while varying the rotation position of the half-wave plates in MI. The acceleration phase of the rf gun is 20 rf deg and time delays determined by movable mirrors No. 1 and No. 2 are 0.66 and 0.7 ps, respectively.