Tunable All-Optical Quasimonochromatic Thomson X-Ray Source in the Nonlinear Regime

We present an all-laser-driven, energy-tunable, and quasimonochromatic x-ray source based on Thomson scattering from laser-wakefield-accelerated electrons. One part of the laser beam was used to drive a few-fs bunch of quasimonoenergetic electrons, while the remainder was backscattered off the bunch at weakly relativistic intensity. When the electron energy was tuned from 17–50 MeV, narrow x-ray spectra peaking at 5–42 keV were recorded with high resolution, revealing nonlinear features. We present a large set of measurements showing the stability and practicality of our source.

Figure 1

Experimental setup. $0.8\mu \mathrm{m}$-wavelength, 28 fs-duration pulses from the ATLAS-60 TW Ti:sapphire laser system at MPQ are split into a driver (1.2 J) and colliding beam (0.3 J). They are focused to $4.2\times {10}^{19}\mathrm{W}/{\mathrm{cm}}^2$ ($a_0=4.4$) and $1.8\times {10}^{18}\mathrm{W}/{\mathrm{cm}}^2$ ($a_0=0.9$), respectively. The driver accelerates electrons from a plasma ($n_e=5\times {10}^{19}{\mathrm{cm}}^{-3}$) in a laser-ionized supersonic He gas flow from a $300\mu \mathrm{m}$ conical de Laval nozzle. A razor blade creates a shock-front electron injector. The electron beam is analyzed on a scintillator screen behind a calibrated 1 T-dipole magnet spectrometer [30]. The colliding beam is focused 1.4 mm behind the electron injection point at a collision angle of 3.7°. X rays are detected on axis after $30\text{−}\mu \mathrm{m}$ aluminum and $250\mu \mathrm{m}$ Kapton windows by either a scintillator-based or a single-photon counting x-ray CCD camera.

Figure 2

Images of the x-ray beam from the scintillator-intensified camera. Series (a), (b), and (c) show raw data at an electron energy of 30, 50, and 70 MeV, corresponding to x-ray photon energies of 15, 42, and 83 keV, respectively. The four top images show the CCD signal with the colliding beam on, the two lower images in each section were taken with a blocked colliding beam. The gain of the MCP was doubled for image (c) to compensate for the lower scintillator efficiency at this energy. Because of enhanced beaming, the brightness of series (a) and (b) seems to be equal in spite of an eightfold reduction in the scintillator efficiency at the higher energy in (b).

Figure 3

Electron (left) and corresponding x-ray photon spectra (right). Each horizontal trace is a single laser shot. Shown are the best 50% of shots by photon number in each run. Different horizontal sections correspond to different razor-blade positions and electron beam energy. X-ray spectra are corrected for filter and vacuum window transmission (see Fig. 1) and CCD sensitivity. Run-averaged spectra are shown in red, white lines show the expected x-ray spectrum for each averaged electron spectrum. The simulation was performed with spectra 9.0 [42], assuming an equivalent undulator model with a Gaussian field envelope and a peak undulator parameter of $K=a_0=0.9$.

Figure 4

Single representative x-ray spectrum (multiplied by 3), corresponding electron spectrum (inset), and x-ray spectrum calculated by spectra 9.0 [42] for an interaction at $a_0=0.75$.

Figure 5

Spectral peaks positions, determined by a Gaussian fit to the measured spectra, whose rms width defines the error bars. Each color stands for a different shock position, indicating the reproducibility of each setting. The blue line labeled $a_0=0.83$ is a quadratic best fit [according to Eq. (1)] to the measured x-ray scaling, while the one for $a_0=0$ serves as a comparison. The inset shows the measured electron and x-ray photon number/msr for each shot.