Demonstration of a Narrow Energy Spread, $\sim 0.5\mathrm{GeV}$ Electron Beam from a Two-Stage Laser Wakefield Accelerator

Laser wakefield acceleration of electrons holds great promise for producing ultracompact stages of GeV scale, high-quality electron beams for applications such as x-ray free electron lasers and high-energy colliders. Ultrahigh intensity laser pulses can be self-guided by relativistic plasma waves (the wake) over tens of vacuum diffraction lengths, to give $>1\mathrm{GeV}$ energy in centimeter-scale low density plasmas using ionization-induced injection to inject charge into the wake even at low densities. By restricting electron injection to a distinct short region, the injector stage, energetic electron beams (of the order of 100 MeV) with a relatively large energy spread are generated. Some of these electrons are then further accelerated by a second, longer accelerator stage, which increases their energy to $\sim 0.5\mathrm{GeV}$ while reducing the relative energy spread to $<5\%$ FWHM.

Figure 1

Schematic of the experimental setup showing the 800 nm Callisto laser beam (in red), the two-stage gas cell, the 800 nm probe beam (in blue), a measured interferogram with its associated Abel inverted density profile (above the interferometer CCD camera), the plasma emission spectrometer and CCD camera, plasma emission images (shown inside the gas cell windows) with the spatially resolved plasma emission spectrum along the laser axis from each stage of the gas cell, the vacuum laser axis after the gas cell (red dashed line), the 0.42 T dipole magnet (20 cm long, centered 66 cm from the exit of the gas cell), the deflected electron trajectory (green dashed line) onto the image plates (located 132 and 192 cm from the exit of the gas cell), and the optical path of the transmitted laser light (in gray) to an imaging system and a prism spectrometer. The transverse size of the plasma observed in the interferogram is larger in the accelerator section because the noncoupled (and therefore unguided) diffracting laser light ionizes a volume larger than the sub-$50\mu \mathrm{m}$ wake, which is not resolved by this diagnostic. Optical access for the transverse diagnostics is provided by constructing the walls of the gas cell from microscope slides pressed against rubber gaskets to form a seal. The source of the Si line in the plasma emission image is suspected to be minute amounts of Si outgassing from the gaskets under vacuum. The gas cell entrance (exit) aperture is 0.5 (2.0) mm.

Figure 2

(a) Magnetically dispersed electron beam images from a 4 mm injector-only gas cell (top) and the 8 mm two-stage cell (bottom). (b) Electron spectra above 70 MeV for the 8 mm two-stage injector-accelerator cell (blue curve) filled to an electron density of $3\times {10}^{18}{\mathrm{cm}}^{-3}$ in each stage for a coupled laser power of 40 TW and the 4 mm injector-only cell (red curve) filled to an electron density of $3.4\times {10}^{18}{\mathrm{cm}}^{-3}$ for a coupled laser power of 50 TW. The injector gas fill in each case is 99.5% He and 0.5% ${\mathrm{N}}_2$, and the total charge is indicated for each spectrum. (c) The total observed charge above 70 MeV for injector gas fills of pure He (red squares) and 99.5% He with 0.5% ${\mathrm{N}}_2$ (blue circles) for coupled laser powers between 30 and 60 TW.

Figure 3

(a),(b) Images of the transmitted laser light at the exit of the 4 mm injector stage and the 8 mm two-stage gas cell, respectively. Vertically integrated lineouts indicate the guided laser spot size, while the vacuum laser spot size at the exit of the cell is denoted for each case by a yellow circle. A long-pass filter is placed in front of the camera to attenuate unshifted and blueshifted light. (c),(d) The transmitted laser spectrum from the injector-only and the two-stage cell, respectively. A mask is used to block the fundamental 800 nm light to take better advantage of the dynamic range of the 8-bit, frame-grabbed, infrared camera. The spectral fringes are due to the etalon effect of the uncoated pellicle beam splitter ($5\mu \mathrm{m}$ thick at 45°) seen in Fig. 1. For each spectrum, the recorded false-color image is shown.