Temporal characterization of ultrashort linearly chirped electron bunches generated from a laser wakefield accelerator

A new method for diagnosing the temporal characteristics of ultrashort electron bunches with linear energy chirp generated from a laser wakefield accelerator is described. When the ionization-injected bunch interacts with the back of the drive laser, it is deflected and stretched along the direction of the electric field of the laser. Upon exiting the plasma, if the bunch goes through a narrow slit in front of the dipole magnet that disperses the electrons in the plane of the laser polarization, it can form a series of bunchlets that have different energies but are separated by half a laser wavelength. Since only the electrons that are undeflected by the laser go through the slit, the energy spectrum of the bunch is modulated. By analyzing the modulated energy spectrum, the shots where the bunch has a linear energy chirp can be recognized. Consequently, the energy chirp and beam current profile of those bunches can be reconstructed. This method is demonstrated through particle-in-cell simulations and experiment.

Figure 1

Output of a synthetic numerical experiment from OSIRIS simulation. (a) Conceptual illustration of the electron bunch interacting with the back of the laser pulse (the plasma is not shown for clarity). (b) Schematic of the virtual energy spectrometer with slit which is similar to the experimental set up. The energy spectrums recorded by a virtual screen as seen without (c) and with the slit (d), respectively. The color scales are different in (c) and (d).

Figure 2

The blue dots in (a) are the modulated energy spectrum corresponding to Fig. 1. It is fitted by four Gaussian peaks (the green solid lines) and the red dashed line represents the superposition of these four peaks. (b) The longitudinal phase space of the bunch. The blue and green dots represent the electrons received by the virtual energy spectrometer without and with the slit, respectively. The red dots are calculated from (a) and the red line is a linear fit. (c) The beam current profile. The blue line is the current profile directly obtained from the simulation, while the red dots and line is the recovered current profile using the fitting data in (a). The error bars in (b) and (c) arise from the fluctuations induced by the variation of the slit width (from 1.4 mm to 2.2 mm) and the displacement between the bunch center and the slit center (from $-0.8\mathrm{mm}$ to 0.8 mm).

Figure 3

The left column (a) shows the longitudinal phase space of electron bunches with different energy chirps. The middle column (b) shows the corresponding bunch modulated by the laser, where the bunch length is 6 fs and the laser wavelength is $0.8\mu \mathrm{m}$. The corresponding energy of the electron increases as the color changes from blue to yellow. The amplitude of the modulation is 3 mm, and the width of the slit is 2 mm, as shown by the dashed lines in the last subfigure in column (b). The corresponding energy spectrums measured with the presence of the slit are shown in the right column (c).

Figure 4

Experiment setup.

Figure 5

(a)–(d) are the raw experimental spectra as seen on the Lanex screen. The color bars represent for the charge density. The corresponding linearized energy spectrums are shown in (e). The plasma density for these shots is $1.26\pm 0.11\times {10}^{19}{\mathrm{cm}}^{-3}$.

Figure 6

(a) shows a typical deconvoluted energy spectrum (the blue dots) fitted by four Gaussian peaks (the solid green lines), where each peak represents a single beamlet. The red line is the superposition of the fitted peaks. (b) The recovered longitudinal phase space. (c) The recovered current profile. The deconvoluted energy spectrum, the recovered longitudinal phase space and the recovered current profile for another shot with a higher density [$1.31\times {10}^{19}{\mathrm{cm}}^{-3}$ versus $1.10\times {10}^{19}{\mathrm{cm}}^{-3}$ in (a)–(c)] are shown in (d)–(f), respectively. The error bars in (b), (d), (e), and (f) represent for the uncertainty of bunchlet separation (see main text).