Skip to content

Multi-filter diamond array time-of-flight particle detector in laser–plasma experiments

FieldValue
JournalReview of Scientific Instruments
Published2026-02-01
AuthorsArmando Raso, Edoardo Domenicone, Giada Petringa, F. Consoli, Farmesk Abubaker
DOI10.1063/5.0302424

Six 35 µm single-crystal CVD diamond channels produced about 1 ns pulses and about 100% charge collection efficiency. The array was tested at the Prague Asterix Laser System, where different aluminum filters supported partial discrimination of overlapping ion species. It measured a mean proton energy of about 2.25 ± 0.05 MeV, close to 2.3 ± 0.1 MeV from a Thomson parabola spectrometer.

Time-of-flight diagnostics infer particle energy from velocity, so ions with the same velocity overlap in time. The paper notes that a 1 MeV proton, a 4 MeV alpha particle, and a 12 MeV C6+ ion share one TOF value. The MFDA pairs six nominally identical detectors with aluminum filters of different thicknesses. Differential stopping power changes both signal amplitude and shape according to particle energy and species. These responses form a spectral fingerprint and permit partial discrimination when arrival times overlap. Discrimination remains model-based because each contribution depends on the assumed particle count and energy distribution.

Detector construction and characterization

Section titled “Detector construction and characterization”

The detectors used synthetic single-crystal radiation detectors with metal contacts on both sides of the sensitive layer. A high-purity intrinsic layer was grown by MPCVD on 4 × 4 × 0.5 mm³ HPHT diamond. Laser cutting separated it from the substrate, leaving 35 µm total diamond thickness. Thermally evaporated chromium contacts were 80 nm thick with a 2.5 × 2.5 mm² cross section. Each unit used a PCB bias tee with a 1 MΩ resistor and a 1 µF capacitor. Six 2.8 mm pinholes aligned the filters with the detectors, which sat about 2 cm behind the filters. At 1 m, each unit subtended about 6 × 10⁻⁴ sr. A minimum target distance of 35 cm avoided housing shadowing. The aluminum KF-100 housing also held two CR-39 detectors.

Vacuum tests used 5.48 MeV alpha particles from an Am-241 source. A CAEN A1423B preamplifier fed a 2 GHz, 20 GS/s Lecroy 620ZI scope. The single-alpha pulse had a FWHM of about 1 ns for basically all six units. An ORTEC charge chain gave about 100% CCE in pulse-height analysis. IBIC at AN2000 with 1 MeV protons showed uniform contact-area response apart from two wire-bond shadows. A 1 MeV peak in the parallel pulse-height spectrum corresponded to the map.

The nominal aluminum filters were 3, 6, 15, 20, 25, and 35 µm thick. Measurements with 5.48 MeV alphas and Geant4 gave effective thicknesses of 2.7, 6.5, 14.8, and 20.3 µm. Their uncertainty was about 3%. The 25 and 35 µm filters were not calibrated because the alpha range was insufficient.

The Python SToF code used SRIM stopping powers to calculate residual energy after each filter. It combined relativistic time of flight with a diamond charge relation using 13.2 eV per electron-hole pair. The voltage model used the 50 Ω total readout resistance. It assumed each filter sat close to the detector relative to the target path. An illustrative beam combined Maxwell-Boltzmann proton and C6+ distributions with a narrow alpha distribution near 4 MeV. The proton and carbon maxima were 3 and 18 MeV. Those maxima were arbitrary but linked through E = z eφ for a shared sheath potential. The alpha model omitted low-energy fusion products and proton momentum transfer, so it served only as a reference case.

PALS used an iodine laser with 600 J energy at 1315 nm and a 350 ps pulse duration. Solid targets were irradiated to study proton-boron interactions and front-side alpha emission. The MFDA sat 1.27 m from the target, at 53° from target normal and 21° above the equatorial plane. It operated at -90 V and fed six 1 GHz, 10 GS/s scope channels through RG223 cables. A Faraday cage and UPS isolated the acquisition system from electromagnetic pulse noise.

Shot No. 60 776 produced a narrow photopeak near 0 ns and a broad signal from protons and heavier ions. Each temporal cutoff agreed with the proton threshold of its aluminum filter. The thinnest channel saturated under the high ion flux. The TPS measured a maximum C6+ energy of about 14.5 MeV in that shot. This energy lay below the transmission thresholds of most filters. Only the first two channels registered partial carbon contributions, while the others mainly detected protons.

Cross-validation with particle diagnostics

Section titled “Cross-validation with particle diagnostics”

The MFDA reconstructed proton spectra by differentiating the time-domain signal after converting arrival time to energy. The comparison covered the upper half of the measured proton energy range, where the authors expect protons alone. The TPS sat at 45° from target normal in the equatorial plane. For shot No. 60 776, MFDA maximum energies ranged from 2.20 to 2.27 MeV, while the TPS measured 2.33 MeV. Mean energies agreed at about 2.25 ± 0.05 MeV for MFDA and 2.3 ± 0.1 MeV for TPS. The plotted vertical uncertainties were about 5% for MFDA and 3% for TPS.

For shot No. 60 843, TPS-derived proton spectra reproduced the shape, amplitude, and modulation in channels 3 through 6. Channels 1 and 2 saturated at high flux, and their measured amplitudes fell below the proton-only predictions. The measured excess between 130 and 160 ns suggests a possible heavier-ion contribution. Synthetic alpha signals were up to two orders of magnitude lower than proton signals, so the alpha contribution was negligible in that shot. C6+ ions were stopped by the thicker filters or produced negligible signals in the first two channels.

CR-39 plates inside the MFDA flange used filters matching the diamond channels. Each pit-energy bin was mapped to TOF and an expected diamond charge. The corresponding MFDA signal was integrated over that bin’s time interval. For shot No. 60 866, the test used 2.7, 6.5, and 14.8 µm filters. Proton-based reconstruction agreed with MFDA measurements better than alpha-based reconstruction. Overlapping CR-39 pit size-energy relations left ion identity ambiguous.

The prompt x-ray photopeak amplitude was measured as the maximum signal near the start of each TOF track. Its normalized intensity decreased exponentially with aluminum thickness. Shot No. 60 843 at about 480 J gave an attenuation coefficient of 0.04 µm⁻¹ and an estimated mean energy of about 6 keV. The 35 µm filter was replaced with an 18 µm filter for that shot. This estimates mean energy but does not reconstruct the full x-ray spectrum. Detector response can bias the estimate toward efficiently detected photon energies. The authors present it as a rough comparison benchmark.

The MFDA also supports an internal consistency check by using one channel to generate predicted responses for the others. For shot No. 60 777, a proton-generated prediction matched well, while a C6+-generated prediction did not. The method can test particle identity assumptions in selected intervals. A forward-fitting procedure was proposed but not implemented because data availability was limited.

The present aluminum filters were aggressive for heavier ions, leaving the array largely proton dominated. The authors plan modular filter assemblies with finer thickness steps for later campaigns. They also propose smaller pinholes to reduce the sensitive area and mitigate saturation. The analysis assumes that all six diamonds receive identical signals before filtering. An angular spread of about 0.9° at PALS supports this approximation, but anisotropic emission can still change each channel.

For detector engineering, the measured design point is a 35 µm single-crystal diamond layer operated at -90 V with about 1 ns response. Filter selection should match target ion energies and intended species. Pinhole area should match expected flux to reduce saturation. The internal self-check tests consistency within the array, while TPS and CR-39 cross-checks address species ambiguities.

The paper’s laser-plasma time-of-flight detector calls for an MPCVD-grown single-crystal diamond layer with metal contacts, and contact compatibility should be reviewed through custom fabrication support. 6CCVD’s diamond materials catalog lists single-crystal material from 0.1 to 500 µm, covering the paper’s 35 µm layer. It also lists Au, Pt, Pd, Ti, W, and Cu metallization and global shipping, although the paper used 80 nm chromium contacts.

View Original Abstract

Recent advances in ion diagnostics for laser-induced plasma experiments have improved system design and data analysis. Measuring charged particle emissions from laser-irradiated targets provides valuable insights into laser-matter interactions. Among real-time diagnostics, Time-of-Flight (TOF) detectors are reliable systems for analyzing particle beam properties such as kinetic energy and shot-to-shot reproducibility. Diamond-based detectors are ideal for TOF measurements due to their fast response time, radiation hardness, and low leakage current. However, TOF detectors lack particle discrimination. To overcome this, a Multi Filter Diamond Array (MFDA) was developed using six nominally identical single crystal diamond detectors paired with aluminum foils of different thicknesses to exploit particle stopping power differences. The MFDA was tested at the Prague Asterix Laser System during an experimental campaign in the framework of the FUSION project, and data analysis was performed. A cross-validation with other diagnostics, including a Thomson Parabola Spectrometer and CR-39 detectors, is also presented.

  1. 2013 - A Superintense Laser-Plasma Interaction Theory Primer
  2. 2004 - The Physics of Inertial Fusion: Beam Plasma Interaction, Hydrodynamics, Hot Dense Matter