Laser Wakefield Source Enables Single-Shot Muon Radiography

Muon radiography is attractive because GeV-scale muons traverse dense material that stops photons, electrons, and many hadrons. The usual trade-off is flux: cosmic-ray muons are free but slow, while accelerator muon sources are large facilities. This paper reports a proof-of-principle experiment that moves the source toward the laboratory scale by using a laser wakefield accelerator to drive secondary particle production from a tungsten converter.

The central result is not a clean monochromatic muon beam. It is a single-shot radiographic signal from a laser-driven secondary beam containing muons, pions, neutrons, and other particles, with simulations and decay timing used to show that the penetrating high-energy component is dominated by muons after transport through shielding and the object.

Core Contribution

The paper demonstrates single-shot high-energy particle radiography from a multi-GeV LWFA electron beam. The authors image a dense absorber placed in front of two detectors inside a truck, after the particle beam has passed through a beam dump, shielding, two laboratory walls, the exterior wall of the building, and the truck wall. The source-to-object scale is about 15 m, which matters because it tests the particles after realistic attenuation rather than at the converter exit.

The result extends prior laser-driven muon-source work from production and detection toward imaging. The authors also directly check that muons are present by measuring delayed scintillator signals consistent with stopped-muon decay, rather than relying only on particle-transport simulation.

Technical Approach

The experimental chain starts with the CSU ALEPH laser system. About 1 J, sub-100 fs pulses form a 30 cm nitrogen/hydrogen plasma channel using an axicon-assisted self-guiding scheme. The main drive pulse, up to 20 J and roughly 40 fs FWHM, is focused into that channel to generate multi-GeV electron beams. A magnetic spectrometer measures the electron spectrum and charge, while tungsten material in the collimator/converter region drives an electromagnetic and hadronic shower.

Figure 1 summarizes the source, spectrometer, shielding, and muon-decay detector geometry.

The electron data show why the source is useful but also why it is noisy. In a representative set of more than 1150 consecutive shots, electron energies span roughly 1–8 GeV; about 70% of shots produced multi-GeV beams, and under better operating conditions the fraction exceeded 80%. Many shots include electrons above 5 GeV, with mean total charge estimated above 100 pC per shot. That electron distribution is then used as the input for secondary-particle estimates.

Muon identification relies on timing. The PMT-based scintillator detector sees a prompt saturated pulse from the electromagnetic shower, followed by delayed candidate peaks. The analysis applies a 100 mV pulse-height threshold and uses a 0.7–10 µs time window. Candidate delays are fit with

where $b$ accounts for backgrounds. The main dataset gives 392 muon-decay candidates over 4955 laser shots and a fitted lifetime of $\tau=2.3\pm0.3$ µs, close to the PDG value of 2.1969811 µs. A cosmic-ray calibration run gives 2.12 ± 0.49 µs with the same 100 mV threshold and 2.13 ± 0.12 µs without the threshold, supporting the timing interpretation.

Results and Analysis

For radiography, the detector array was placed in a truck outside the laboratory, roughly 15 m downstream of the tungsten slit assembly. The object was 60 cm of lead plus 15 cm of HDPE, inserted to obscure two detectors in a vertical ten-detector array. The experiment recorded 82 shots without the object and 75 shots with it. Both the averaged image and a representative single shot show a clear shadow at the obscured detector positions.

The authors estimate that the highest-charge electron beams generate up to 3.6×10⁴–4.7×10⁴ muons per shot, with 5000–6500 muons per shot in the representative 1150-shot series. FLUKA simulations place the muon fluence incident on the object at about 50 muons/cm² for the highest-energy, highest-charge shots. The paper compares that single-shot fluence to more than 8 hours of naturally occurring cosmic-ray muons near the horizon.

The species composition is the most important caveat in interpreting the image. Simulations show that neutrons dominate particle counts below about 600 MeV, accounting for roughly 90% of transmitted particles in that low-energy range. But above roughly 1 GeV, after the lead/HDPE object and propagation, the transmitted particles are nearly all muons. The detector response is therefore a mixed signal, not a particle-tagged muon image, but the penetrating component responsible for the high-energy transmission is well supported by the transport model.

Caveats in Practice

This is a proof-of-principle imaging experiment, not yet a deployable radiography system. The detector array does not discriminate particle species, the object geometry is simple, and the source has large shot-to-shot variation. The authors also state that the supporting data are not publicly available, only available on reasonable request. Even with those limits, the paper makes a credible case that LWFA-driven secondary beams can compress a muography exposure from hours to one laser shot for some dense-object scenarios.