In Situ Measurement of Electron Energy Evolution in a Laser-Plasma Accelerator

We report on a novel, noninvasive method applying Thomson scattering to measure the evolution of the electron beam energy inside a laser-plasma accelerator with high spatial resolution. The determination of the local electron energy enabled the in-situ detection of the acting acceleration fields without altering the final beam state. In this Letter we demonstrate that the accelerating fields evolve from $(265\pm 119)\mathrm{GV}/\mathrm{m}$ to $(9\pm 4)\mathrm{GV}/\mathrm{m}$ in a plasma density ramp. The presented data show excellent agreement with particle-in-cell simulations. This method provides new possibilities for detecting the dynamics of plasma-based accelerators and their optimization.

Figure 1

(a) Schematic of the experimental setup. (b) Mean measured longitudinal plasma density profile with its standard deviation and the region where the Thomson measurement was performed.

Figure 2

Simulation of x-ray transport and detection. (a) An x-ray source spectrum (green), the detector signal of this x-ray beam (solid blue), and the double Gaussian fit used to determine the peak of the deposited energy distribution (dotted blue). (b) Peak of the detector signal and 68% confidence intervals from fitting routine as a function of the peak energy of the source spectrum (blue). The calibration function (red) is used for the adjustment of measured data.

Figure 3

Normalized electron spectrum of 340 shots with Thomson laser turned on at an overlap position of $1330\mathrm{\mu }\mathrm{m}$ (gray line), the standard error from individual normalized shots (gray area) and the normalized electron spectrum at the same overlap position without the Thomson laser (orange dashed line). The error band without the Thomson laser is similar to the displayed band with the Thomson laser activated. The same level of agreement between signal and background shots was also observed at overlap positions of $1190\mathrm{\mu }\mathrm{m}$ and $1560\mathrm{\mu }\mathrm{m}$. The region around the peak is shown in more detail in the inset. The spectrum obtained from PIC simulations is depicted as a red dotted line. All lines are normalized individually.

Figure 4

(a) Electron beam energy measured via Thomson scattering as function of the overlap position (blue). The final electron beam energy at each interaction point is shown as the average peak of Gaussian fits to the single electron spectra with its standard error (gray). The energy evolution obtained from the PIC simulation is shown as a red line, for better comparison again using Gaussian fits. (b). Acceleration gradient as function of the overlap position. The acceleration gradient was calculated using a linear fit of $\pm 3$ data points. At the edges only available data points were included. The error bars display the 68% confidence interval.