Tunable High Spatio-Spectral Purity Undulator Radiation from a Transported Laser Plasma Accelerated Electron Beam

Abstract

Undulator based synchrotron light sources and Free Electron Lasers (FELs) are valuable modern probes of matter with high temporal and spatial resolution. Laser Plasma Accelerators (LPAs), delivering GeV electron beams in few centimeters, are good candidates for future compact light sources. However the barriers set by the large energy spread, divergence and shot-to-shot fluctuations require a specific transport line, to shape the electron beam phase space for achieving ultrashort undulator synchrotron radiation suitable for users and even for achieving FEL amplification. Proof-of-principle LPA based undulator emission, with strong electron focusing or transport, does not yet exhibit the full specific radiation properties. We report on the generation of undulator radiation with an LPA beam based manipulation in a dedicated transport line with versatile properties. After evidencing the specific spatio-spectral signature, we tune the resonant wavelength within 200-300 nm by modification of the electron beam energy and the undulator field. We achieve a wavelength stability of 2.6%. We demonstrate that we can control the spatio-spectral purity and spectral brightness by reducing the energy range inside the chicane. We have also observed the second harmonic emission of the undulator.

Figure 1

COXINEL Experimental set-up. Laser source (grey), gas jet (cyan), permanent magnet based quadrupoles (QUAPEVAs) (light grey), LANEX screen (black), electro-magnet dipoles (red) with an adjustable slit placed at the center (pink), electro-magnet quadrupoles (blue) with a 75 μm-thick Aluminum foil inserted at the center to remove plasma radiation and laser beam contamination (yellow), undulator (purple), dipole magnet (red) for electron beam dump (light purple), lens (grey) focusing the undulator radiation into a UV spectrometer (light grey).

Figure 2

Measured and simulated undulator spatio-spectral distribution. (a) Single shot measurement for an electron beam energy of 176 MeV, a 5 mm undulator gap, 3 mm electron slit, 2.2 mm spectrometer slit width, and an applied calibration of the grating and CCD camera (See Methods). (d) Simulation using SRW with parameters of Table 1. (b) and (e) Undulator spectra for different vertical positions at z = 0 (blue), 0.2 mm (green), 0.4 mm (yellow), 0.6 mm (orange), 0.8 mm (red). (c) and (f) Vertical radiation profiles with cuts at different wavelengths λ = 208 nm (blue), 228 nm (green), 248 nm (yellow), 268 nm (orange), 288 nm (red). Black curve: fit of the undulator resonance wavelength taking into account the chromatic aberrations of the lens (See Methods).

Figure 3

Wavelength tunability by undulator gap and energy change. Single shot spatio-spectra distribution measured for 161 MeV beam energy at different gaps: (a) 4.7 mm, (b) 5 mm, (c) 5.5 mm, (d) 6 mm, with an electron slit of 1 mm and a spectrometer slit of 2.2 mm. (e) Measured and theoretical (dashed) resonant wavelength versus undulator gap: 161 MeV (red) and 176 MeV (blue).

Figure 4

Resonant wavelength stability. Undulator resonant wavelength measured during 60 successive shots over 3 hours for an undulator gap of 4.7 mm and different electron slit widths. Average value (dashed), standard deviation (purple).

Figure 5

Undulator spatio-spectral distribution dependance on the electron beam energy selection. Single shot measured spatio-spectral distributions for a 4.7 mm undulator gap (with caibration) while varying the electron slit width: 4 (a), 3 (b), 2 (c) and 1 mm (d) with a 2.2 mm spectrometer entrance slit. Simulated spectra using SRW for a magnetic field of 1.17 T, with beam parameters taken from the simulations of the corresponding electron beam distribution transported along the line (see Table 1 for 161 MeV) for slit widths of 4 (e), 3 (f), 2 (g) and 1 mm (h) with their corresponding on-axis spectra (white curves). (i) Measured (red), analytically estimate of energy spread contribution (blue), analytically estimate of all the contributions (dashed) and simulated (line) FWHM relative bandwidth of the on-axis spectra.

Figure 6

Observation of the second undulator harmonic. Undulator radiation spatio-spectral distribution at a gap of 4.7 mm, electron slit opened at 4 mm and spectrometer slit at 2.2 mm. (a) Simulation showing the first and second harmonic, (b) zoom of (a), (c) simulation including chromatic effects of the lens on (b), (d) calibrated measurement.

Figure 7

Measured electron beam charge distribution (blue), vertical divergence (black) for 161 (a) and 176 (b) MeV, with σ′_x/σ′_z = 1.56. Twiss parameters (betatron and dispersion) evolution simulation for a monoenergetic electron beam of 161 (c) and 176 (d) MeV along the transport line. Twiss parameters (beta and alpha) at the undulator center versus energy slice for the 161 (e) and 176 (f) MeV cases. Energy distribution at the undulator center for different slit widths: no slit (red), 3.2 (orange), 2.2 (green), 1 (blue) mm width for the 161 (g) and 176 (h) MeV cases. Electron beam transverse distribution at undulator center for the 161 MeV: (i) no slit and (j) 1 mm slit. Parameters for the transport calculations: 1 mm.mrad initial emittance, 1 μm longitudinal size, 10⁶ macroparticles, 4.3 mm chicane strength. In the 176 (resp. 161) MeV case, QUAPEVA 1 of 40.7 mm magnetic length: +104.1 T m⁻¹ (resp. 113.5), QUAPEVA 2 of 44.7 mm magnetic length: −103.1 T m⁻¹ (resp. −111.3), QUAPEVA 3 of 26 mm magnetic length: +96.4 T m⁻¹ (resp. 103.4). QUAPEVA skew contribution (ratio of skew gradient over normal gradient) of +1.5 × 10⁻³ (QUAPEVA 1), −0.3 × 10⁻³(QUAPEVA 2), −0.7 × 10⁻³ (QUAPEVA 3) with a field variation of 2% at 4-mm radius due to a dodecapole component for the three QUAPEVA. Electromagnetic quadrupole gradients at 176 (reps. 161) MeV: −0.01, 4.7, −4.4, +0.29 T m⁻¹ (4.15, −3.45, −0.13,+ 1.7 T m⁻¹) for QEM 1, 2, 3, 4.