High-precision beam profile measurement with a microchannel-plate detector in the high magnetic field of the WISArD experiment

We present the development and characterization of a compact low-energy ion beam diagnostic for the WISArD (Weak Interaction Studies with $\mathrm{^{32}Ar}$ Decay) experiment at ISOLDE/CERN. The microchannel plate (MCP) detector, which is configured in a Z-stack and has a resistive position sensitive anode, was tested with both stable and radioactive beams. This work focuses on the image reconstruction method, which corrects the pincushion distortion inherent to the square-shaped resistive anode, and investigates the influence of the magnetic field on the detector performance. Our results demonstrate that the detector achieves beam profile measurements with sub-millimeter accuracy, while coping with the spatial and high-magnetic field (4 T) constraints of the experiment. These capabilities meet the precision requirements of the WISArD experiment for extracting the modified beta-neutrino angular correlation coefficient, ã$_{βν}$, with an uncertainty of 0.1%.

💡 Research Summary

The paper reports the design, construction, and performance evaluation of a compact microchannel‑plate (MCP) detector developed for the WISArD (Weak Interaction Studies with ³²Ar Decay) experiment at ISOLDE/CERN. The WISArD experiment aims to determine the modified beta‑neutrino angular correlation coefficient (ã₍βν₎) in the β‑decay of ³²Ar with a relative precision of 0.1 %, a goal that requires sub‑millimetre knowledge of the radioactive ion beam spot on the catcher foil. Existing uncertainties of about 3 mm in beam position and radius contributed a systematic error of Δã₍βν₎ ≈ 4 ‰, motivating the development of a high‑precision beam diagnostic that can operate inside the 4 T superconducting magnet of the WISArD detection tower.

Detector design and constraints
The detector must fit within a maximum thickness of 15 mm and width of 25 mm, and it must survive the strong magnetic field. A Z‑stack of three Hamamatsu F1551‑01 MCPs (12 µm channel diameter, 8° bias angle, 17.9 mm active diameter) was chosen because the smaller channel size mitigates the gain loss that typically occurs in high fields (a 1 T field can reduce the gain of a 25 µm MCP by an order of magnitude). The three‑plate Z‑stack provides higher overall gain than a standard chevron configuration, which is essential when the gain is partially suppressed by the 4 T field.

For position read‑out, a custom square‑shaped resistive anode was fabricated. The resistive layer consists of a graphite‑glyceride paint (mass ratio 1 : 1.33) applied to a 200 µm‑deep, 16 mm × 16 mm milled pocket in a PEEK substrate, yielding a sheet resistance of 1–2 kΩ/□. Four thin stainless‑steel strips make contact with the four corners of the anode, providing the charge signals needed for position reconstruction.

Mechanical integration
The detector is mounted on a rotatable rod inside the WISArD tower, allowing it to be moved out of the beam axis when not in use. The entire assembly, including two PCBs for high‑voltage bias and signal extraction, fits within the spatial constraints of the magnet bore.

Electronics and data acquisition
Front‑MCP bias voltages range from –2.0 kV to –3.2 kV, the back MCP is grounded, and the anode is held at +100 V. Fast signals from the back MCP are decoupled via a 2.2 nF capacitor and a 1 MΩ resistor, then amplified by a ×10 fast current pre‑amplifier. The four corner charge signals are processed by charge pre‑amplifiers (≈1 V/pC gain) and digitized by the FASTER DAQ system (CARAS for the fast MCP signal, MOSAHR for the charge signals). All channels are time‑stamped with 8 ns resolution, enabling precise coincidence selection.

Image reconstruction and distortion correction
The raw position is first obtained from the classic four‑corner charge ratios: X₍d₎ = (–C₁ + C₂ + C₃ – C₄)/ΣC, Y₍d₎ = (–C₁ – C₂ + C₃ + C₄)/ΣC.
Because the square resistive anode introduces a strong pincushion distortion, a logarithmic correction is applied: cᵢ = ln(Cᵢ/ΣC) and the same linear combination of the cᵢ yields corrected coordinates (X, Y). This step removes most of the non‑linearity but leaves residual asymmetries.

A permanent aluminium mask (0.5 mm thick, 2 mm pitch, 1.4 mm × 1.4 mm holes) is mounted in front of the first MCP. The mask provides 45 well‑defined reference points (the four corners of each hole). The authors model each hole with four linear edge functions and define a binary transmission function Iⱼ(X,Y) that is 1 inside the hole and 0 elsewhere. The full mask transmission I_mask is the sum over all holes, which is then convolved with a 2‑D Gaussian R(δX,δY) to account for detector resolution. The analytical expression (involving error functions) is fitted to the measured hole pattern, extracting the corner coordinates, the Gaussian widths (δX, δY), and an overall amplitude. A second iteration allows each hole to have its own amplitude and resolution, improving the fit stability.

Bilinear interpolation maps the corrected (X,Y) coordinates onto the physical mask coordinates (x,y). After this full calibration, the detector achieves a spatial resolution of σₓ ≈ 0.22 mm and σᵧ ≈ 0.24 mm, and the gain imbalance between corners is reduced to < 3 %.

Performance in magnetic fields
The detector was characterized with a stable 30 keV ³⁹K⁺ beam under magnetic fields ranging from 0 T to 4 T. In the 4 T field the MCP gain drops by about 8 dB, but by increasing the pre‑amplifier gain the signal‑to‑noise ratio remains above 20 dB. The coincidence time distribution between the MCP trigger and the four corner signals shows a main peak at –170 ns; a selection window of –190 ns to –100 ns retains ~92 % of events while rejecting random coincidences and cable reflections.

Radioactive beam test and impact on WISArD
During the 2025 ISOLDE run, a 30 keV ³²Ar⁺ beam was directed onto the detector. Using the calibrated mask, the beam spot was reconstructed with a centroid uncertainty of ±0.08 mm and a full‑width at half‑maximum of 2.8 mm. This sub‑millimetre knowledge of the implantation region reduces the systematic uncertainty on the beam position from 3 mm to 0.1 mm, which translates into a reduction of the systematic error on ã₍βν₎ from 4 ‰ to below 0.1 ‰—well within the experiment’s precision goal.

Conclusions and outlook
The work demonstrates that a compact Z‑stack MCP coupled to a custom square resistive anode can operate reliably in a 4 T magnetic field, delivering sub‑millimetre spatial resolution and stable timing performance. The combination of logarithmic charge correction and a mask‑based non‑linear calibration effectively removes the intrinsic pincushion distortion. The detector meets the stringent requirements of the WISArD experiment for precise beam profiling, enabling the targeted 0.1 % measurement of the beta‑neutrino angular correlation. The authors suggest that the same design principles could be extended to even higher magnetic fields (> 5 T) or to other low‑energy radioactive ion beam lines where space is limited and precise beam diagnostics are essential.