Laser-Accelerated, Low-Divergence 15-MeV Quasimonoenergetic Electron Bunches at 1 kHz

Open Access

Laser-Accelerated, Low-Divergence 15-MeV Quasimonoenergetic Electron Bunches at 1 kHz

F. Salehi, M. Le, L. Railing, M. Kolesik, and H. M. Milchberg

Phys. Rev. X 11, 021055 – Published 11 June 2021

Abstract

We demonstrate laser wakefield acceleration of quasimonoenergetic electron bunches up to 15 MeV at 1-kHz repetition rate with 2.5-pC charge per bunch and a core with $< 7 − mrad$ beam divergence. Acceleration is driven by 5-fs, $< 2.7 − mJ$ laser incident on a thin, near-critical-density hydrogen gas jet. Low beam divergence is attributed to reduced sensitivity to laser carrier-envelope phase slip, achieved in two ways using gas jet positon control and laser polarization: (i) electron injection into the wake on the gas jet’s plasma density downramp and (ii) use of circularly polarized drive pulses. These results demonstrate the generation of high-quality electron beams from a few-cycle-pulse-driven laser plasma accelerator without the need for carrier-envelope phase stabilization.

Received 29 October 2020
Revised 1 February 2021
Accepted 20 April 2021

DOI:https://doi.org/10.1103/PhysRevX.11.021055

Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI.

Published by the American Physical Society

Physics Subject Headings (PhySH)

Laser wakefield acceleration

Accelerators & BeamsPlasma PhysicsAtomic, Molecular & Optical

Authors & Affiliations

F. Salehi ¹, M. Le ¹, L. Railing ¹, M. Kolesik², and H. M. Milchberg ^1,*

¹Institute for Research in Electronics and Applied Physics, University of Maryland, College Park, Maryland 20742, USA
²College of Optical Sciences, University of Arizona, Tucson, Arizona 85712, USA

^*milch@umd.edu

Popular Summary

High-energy electron beams are in large demand for many applications in science, industry, and medicine. However, traditional accelerators are costly and require a lot of physical space, thus limiting their utility. A promising alternative is to accelerate electrons using the enormous electric fields generated in the wake of a laser propagating through a plasma. This laser wakefield acceleration could replace conventional accelerators for some tasks because of its compact footprint and bright, ultrashort electron pulses, especially if the repetition rate of the pulses could be increased. Here, by packing a modest amount of energy into laser pulses just two optical cycles in duration and by controlling the laser polarization, we demonstrate generation of record high-energy and low-divergence relativistic electron beams at a high repetition rate of 1000 pulses per second.

In the laser-dense plasma interaction leading to electron acceleration, our laser pulses naturally self-collapse and increase their intensity to relativistic levels, where plasma electrons exposed to the pulse can quiver at nearly the speed of light. The pulse intensity becomes so high that electrons are expelled by light pressure, leaving an electron void or bubble immediately following the pulse. This bubble acts like a relativistic electron accelerator, pulling electrons from the bubble wall and forming a 15-MeV electron beam in the direction of the laser pulse.

In past accelerator experiments using few-cycle pulses, the bubble was violently wiggled by the laser pulse, leading to low-quality electron beams. In this paper, we demonstrate how to tame the bubble, leading to much lower divergence and higher-energy electron beams.

Key Image

Article Text

Click to Expand

Supplemental Material

Click to Expand

References

Click to Expand

Issue

Vol. 11, Iss. 2 — April - June 2021

Subject Areas

Reuse & Permissions

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
Experimental setup. (a) Laser pulses with few-cycle duration, generated through a hollow core fiber (HCF) and a chirped mirror compressor, are used to drive electron acceleration from a near-critical-density hydrogen jet target. A probe pulse split from the main drive pulse is used for (b) probing of the few-cycle pulse interaction with the jet and (c) interferometric measurement of the jet density. The white arrows show the drive beam propagation direction, the white dashed lines [(b)] indicate the laser beam $4 e^{- 2}$ Gaussian intensity envelope, and the red dashed lines [(c)] show the density FWHM contour. (d),(e) Sample electron beam profile and energy spectrum as imaged on the LANEX screen. (f) HCF output spectrum as a function of input polarization (input energy 6 mJ). Elliptical input polarization ( $ϵ = 0.5$ ) generates the broadest output spectrum, which is near-circularly polarized. LP beams are polarized left-right in electron beam profile and spectrum images such as (d) and (e).
Reuse & Permissions
Figure 2
Experimental conditions: 2.2-mJ, 5-fs LP drive pulse, peak jet plasma density $N_{e, peak} / N_{cr} = 0.08$ , ${(P / P_{cr})}_{\max} \sim 2$ . (a) Electron beam profiles, $θ_{x}$ and $θ_{y}$ lineouts, and Lorentzian fits for varying position of the jet center ( $z_{jet}$ ) relative to the laser beam waist at $z = 0$ (ten-shot burst). The Lorentzian fit quality is reduced as the jet center moves closer to the beam waist, with rms fit error $G$ [32] shown in each panel. (b) PIC simulations [32], showing normalized laser vector potential $a$ (color map and dashed white line) overlaid on electron density. To illustrate plasma wave injection at locations progressively farther into the density downramp, laser propagation is launched at beam waists located (i) $10 μ m$ past the jet center, (ii) $30 μ m$ past the jet center, and (iii) $50 μ m$ past the jet center. The horizontal scale is distance from the jet center. Simulation parameters: $a_{0} = 1$ , 5-fs LP drive pulse, peak plasma density $N_{e, peak} / N_{cr} = 0.08$ , and jet FWHM $140 μ m$ .
Reuse & Permissions
Figure 3
Experimental conditions: 2.6-mJ, 5-fs LP drive pulse, peak jet plasma density $N_{e, peak} / N_{cr} = 0.12$ , ${(P / P_{cr})}_{\max} \sim 3.5$ . (a) Electron beam profiles for the vacuum beam waist near the jet center ( $z_{jet} = 20 μ m$ ) and in the density downramp ( $z_{jet} = - 110 μ m$ ). (b) Corresponding electron bunch spectra and (c) shadowgraphic images, which show the beam focus near the jet center (bottom) and in the downramp (top). The dashed white lines show the $4 e^{- 2}$ envelope of the vacuum beam. (d) Lineouts of angle-integrated electron spectra of (b) on a linear energy scale. The resolution of the QME peaks is reduced by some penetration of high-energy electrons through the spectrometer slit edges. The electron beam profiles and spectra are averages from ten-shot bursts.
Reuse & Permissions
Figure 4
Experimental conditions: 2.7-mJ, 5-fs $C P$ drive pulse, peak jet plasma density $N_{e, peak} / N_{cr} = 0.10$ , ${(P / P_{cr})}_{\max} \sim 3.1$ . (a) Single-shot electron beam profiles for varying $z_{jet}$ . The middle panel inset for $z_{jet} = - 60 μ m$ (near half-peak density in the downramp) shows the electron ring image enhanced by $5 \times$ . (b) Corresponding energy spectra. (c) Energy spectra lineouts plotted on a linear scale. (d) PIC simulation result for a $C P$ pulse with injection near half-maximum density in the downramp. Electrons (magenta dots) injected in the second bucket form a ring in the transverse plane (right) as they outrun the first bucket. The electrons in the ring have a lower energy than those in the central spot (bottom left). Simulation parameters: pulse energy 2.7 mJ, pulse width 5 fs, vacuum spot FWHM $4.5 μ m$ , jet FWHM $140 μ m$ , $z_{jet} = - 60 μ m$ , and $N_{e} / N_{cr} = 0.1$ .
Reuse & Permissions
Figure 5
Simulation parameters: LP or $C P$ , 2.7 mJ, 5 fs, $λ_{0} = 650 nm$ , vacuum spot FWHM $4.5 μ m$ , jet $140 μ m$ FWHM, $N_{e} / N_{cr} = 0.11$ , and center of neutral hydrogen jet placed at the beam waist ( $z_{jet} = 0$ ). (a) Normalized bubble centroid position $(\bar{x} / R_{b}, \bar{y} / R_{b})$ in transverse plane v. time. $CP = circular$ polarization, $LP = linear$ polarization. Numbers near points are values of $ω_{0} t$ . (b) $\bar{x} / R_{b}$ and $\bar{y} / R_{b}$ vs time, showing the CEP slip-induced bubble centroid oscillations (see the text). (c) Average axial field $⟨ E_{z} (0, 0, ξ) ⟩$ in the accelerating phase of the bubble vs time for $C P$ and LP pulses. The open circles are for times after electron injection and before bubble breakup. (d) Top: electron beam profiles ( $> 2 MeV$ ) at $ω_{0} t = 2000$ from $C P$ and LP pulses, where the bunch is near $z = 80 μ m$ . The rms beam divergence $Δ θ_{div} = {[(Δ θ_{x}^{2} + Δ θ_{y}^{2}) / 2]}^{1 / 2}$ is shown in the panels. Bottom: beam lineouts along $x$ and $y$ . The LP direction is $y$ (up-down).
Reuse & Permissions
Figure 6
Experimental conditions: 2.7-mJ, 5-fs LP or $C P$ drive pulse, peak jet plasma density $N_{e, peak} / N_{cr} = 0.11$ , ${(P / P_{cr})}_{\max} \sim 3.4$ , $z_{jet} = 0$ . (a) LP driver: electron beam profile (left) and corresponding energy spectrum (right). (b) CP driver: electron beam profile (left) and corresponding energy spectrum (right). The two values of $Δ θ_{div}$ shown are for the central peak and the ringlike pedestal. (c) Electron spectrum lineouts on a linear scale for LP and $C P$ drivers. Electron beam profiles and spectra are averaged over ten-shot bursts. (d) Electron spectra for ten consecutive $C P$ shots at 0.5 Hz.
Reuse & Permissions

Physical Review X