Mechanisms to control laser-plasma coupling in laser wakefield electron acceleration

Experimental results, supported by precise modeling, demonstrate optimization of a plasma-based injector with intermediate laser pulse energy ($<1\mathrm{J}$), corresponding to a normalized vector potential $a_0=2.15$, using ionization injection in a tailored plasma density profile. An increase in electron bunch quality and energy is achieved experimentally with the extension of the density downramp at the plasma exit. Optimization of the focal position of the laser pulse in the tailored plasma density profile is shown to efficiently reduce electron bunch angular deviation, leading to a better alignment of the electron bunch with the laser axis. Single peak electron spectra are produced in a previously unexplored regime by combining an early focal position and adaptive optic control of the laser wavefront by optimizing the symmetry of the prefocal laser energy distribution. Experimental results have been validated through particle-in-cell simulations using realistic laser energy, phase distribution, and temporal envelope, allowing for accurate predictions of difficult to model parameters, such as total charge and spatial properties of the electron bunches, opening the way for more accurate modeling for the design of plasma-based accelerators.

Figure 1

Experimental setup at the Lund Laser Centre. The laser is focused into the gas cell by the off-axis parabola. The interaction between the laser and the plasma produces an accelerating cavity and electron bunch which is illustrated in the simulated inset. Accelerated electrons exiting the gas cell are then dispersed with the permanent dipole magnet and produce scintillating radiation on a LANEX screen which is then imaged onto a 16-bit CCD. An adaptive optic, set after the compressor, is used to tune the beam wavefront. Laser diagnostics are performed in vacuum using attenuators before the compressor: using the flip mirror, the beam (in pink) can be extracted to measure wavefront curvature using a Phasics wavefront sensor; the energy distribution in the focal volume is recorded in vacuum using a camera movable on axis in place of the gas cell. The adaptive optic settings are altered to produce the three laser setting cases displayed in Fig. 2 as measured by the focal spot camera under vacuum. Energy measurements are taken using the leak beam (shown in light red) through a dielectric mirror and a calorimeter-calibrated camera.

Figure 2

Laser energy distribution in the transverse plane around the focal position—relative position marked above—for three different settings of the AO and their corresponding energy profiles displayed from their modal description used in the simulation (denoted by Sim.). Flat phase setting (FPS) and two manually altered AO setting fluence profiles, 1bFPS and 2bFPS, are displayed in pink, blue, and green, respectively. Each image is normalized to its maximum value for visibility.

Figure 3

(a) Rotational asymmetry parameter through focus for the three experimental laser energy distributions shown in Fig. 2; (b) Cumulative fraction of energy contained in successive angular modes in the simulated laser distributions. For both (a) and (b), FPS, 1bFPS, and 2bFPS are displayed in red squares, blue triangles, and green circles, respectively.

Figure 4

Normalized electron density profile along laser axis in the gas cell from fluid simulations. Short and long exit configurations are indicated by purple and green plasma densities and inset dimensions, respectively. The laser travels from negative to positive $z$ values as marked by the red arrow.

Figure 5

Experimental electron charge density in divergence-energy space and their corresponding spatially integrated $dQ/dE$ ($\mathrm{p}\mathrm{C}\mathrm{M}{\mathrm{eV}}^{-1}$) within a $\pm 7.5\mathrm{m}\mathrm{rad}$ window around the laser axis indicated by the dashed horizontal gray line with the laser focus at $z=-0.35\mathrm{m}\mathrm{m}$ for the two exit plate configurations illustrated in Fig. 4: (a) short exit configuration (SE) case, and (b) long exit configuration (LE). Standard deviation of four and three consecutive shots for (a) and (b), respectively, are illustrated by the shaded green region. All $dQ/dE$ plots are plotted from zero ($\mathrm{p}\mathrm{C}\mathrm{M}{\mathrm{eV}}^{-1}$) in linear scaling. The purple and red dashed lines indicate the maximum value of $dQ/dE$ for the displayed spectra.

Figure 6

Same as Fig. 5 for laser focus at $z=0\mathrm{m}\mathrm{m}$ with three consecutive shots included in the standard deviation.

Figure 7

Representative experimental electron spectra within $\pm 7.5\mathrm{m}\mathrm{rad}$ window around the laser axis between 11 and 175 MeV, corresponding to laser focal positions along the longitudinal spatial axis, $z$, (a) preplateau: $z=-0.8\mathrm{m}\mathrm{m}$, (b) periplateau: $z=0\mathrm{m}\mathrm{m}$, and (c) postplateau: $z=0.8\mathrm{m}\mathrm{m}$ for the LE case and 1bFPS AO settings; (d) total charge within a $\pm 4\mathrm{m}\mathrm{rad}$ of the electron bunch peak spatial location, and (e) electron bunch peak displacement from laser axis, defined as zero angular displacement, are shown as blue circles as functions of position along the laser axis. Simulated results for $n_e=6.7\times {10}^{18}{\mathrm{cm}}^{-3}$ and $a_0=2.15$ are plotted as black hexagons for a Gaussian laser bunch and realistic 1bFPS sim laser as blue squares; for comparison $n_e=7.5\times {10}^{18}{\mathrm{cm}}^{-3}$ and $a_0=2.15$ are indicated by cyan down triangles ($a_0=2.0$ by magenta up triangles). Plasma density profile is illustrated by the gray line. Errors are given by the standard deviation of the values for both parameters and $dQ/dE$ from consecutive shots.

Figure 8

Simulation results for (a) the evolution of the center of mass of the laser fluence inside a disk of $20\mu \mathrm{m}$ radius, and (b) of the average angle of the electrons accelerated to the energy peak, as a function of position on the laser axis, for three focal positions: $-0.8\mathrm{m}\mathrm{m}$ (purple square symbols), 0.4 mm (gray triangles), and 0.7 mm (black circles).

Figure 9

Experimental spectra (a)–(c) illustrating the effect of phase front optimization on accelerated electron bunches at laser focus $z_{\mathrm{foc}}=-0.8\mathrm{m}\mathrm{m}$ for wavefront configurations (a) FPS, (b) 1bFPS, and (c) 2bFPS. Electron charge density in divergence-energy space ($\mathrm{p}\mathrm{C}\mathrm{M}{\mathrm{eV}}^{-1}\mathrm{m}{\mathrm{rad}}^{-1}$) and their corresponding spatially integrated $dQ/dE$ ($\mathrm{p}\mathrm{C}\mathrm{M}{\mathrm{eV}}^{-1}$) within a $\pm 7.5\mathrm{m}\mathrm{rad}$ window around the laser axis indicated by the dashed horizontal line and in an energy window of 11.3 and 175 MeV. Standard deviation calculated over five shots and plotted here in cyan. (d) simulation results for the evolution along the propagation distance $z$ of the normalized vector potential $a_0$ of the laser pulse: the black curve represents the normalized plasma density profile, while the focus position, $z_{\mathrm{foc}}$, is marked by the red arrow; (e) evolution with $z$ of the charge of the electrons having final energy above 10 MeV (solid lines) and the average energy of the electrons contributing to the peak in energy normalized by its maximum values (dashed lines). For (d) and (e), the red curves correspond to FPS, blue curves to 1bFPS, and green to 2bFPS.