PUMA Experiment at CERN

• PUMA is a CERN-based experiment that probes nucleon distributions in stable and radioactive nuclei via low-energy antiproton annihilation.
• It integrates sophisticated beamline components, high-resolution mass purification, and ultrahigh vacuum engineering to ensure precise measurements.
• The system enables advanced studies of neutron skin in exotic nuclei while providing benchmark calibration with stable isotopes.

The antiProton Unstable Matter Annihilation (PUMA) experiment at CERN is designed to probe the nucleonic composition in the matter density tail of stable and radioactive nuclei via low-energy antiproton-induced annihilation. PUMA uniquely combines low-energy antiprotons provided by the Extra Low ENergy Antiproton (ELENA) facility with isotopically pure ion bunches from either stable sources or ISOLDE-produced radioactive nuclei, enabling direct measurements of neutron and proton distribution in rare isotopes. To achieve these scientific objectives, PUMA integrates a sophisticated antiproton deceleration beamline, dedicated offline ion source infrastructure, and high-resolution mass purification systems, all operating under ultrahigh vacuum conditions and sub-μs bunch timing constraints (Fischer et al., 2024, Schlaich et al., 11 Apr 2025, Schlaich et al., 2023).

1. Experimental Architecture and Beamline Components

PUMA's infrastructure consists of several functionally distinct but interconnected beamlines tailored to the transport, deceleration, purification, and focusing of antiprotons and ions:

• Antiproton Deceleration Beamline (from ELENA):

The LNE51 transfer line (≈15 m) channels antiprotons from ELENA through a fast deflector switch (ZDFA–ZDSA), electrostatic quadrupole/HV-corrector modules (ZQNA), and profile-monitoring SEM grids. A 37.7° electrostatic bender branches the beam to PUMA, optimizing transverse matching to acceptance of the Penning trap (spot ∼2 mm rms; emittance 6 mm·mrad H, 4 mm·mrad V).

• High-Voltage Deceleration Section:

Pre-focusing is achieved via a high-voltage (up to –85 kV) injection einzel lens, delivering antiprotons to a pulsed drift tube (PDT; 700 mm length, 100 mm bore), which switches potential from –96 kV to 0 V in ≤500 ns, reducing kinetic energy from 100 keV to ≈4 keV. Ejection is performed with a +5 kV einzel lens for refocusing.

• Low-Energy Transport to Trap:

Comprising two LV einzel lenses (±5 kV, X–Y steerers), an electrostatic quadrupole bender allows off-axis ion injection, with a beam-spot diagnostic station (BTV, phosphor screen and camera) for destructive analysis.

• Offline Ion Source Beamline (POLIS):

Structured as four vacuum-isolated sections: electron-impact ion source (SPECS IQE 12/38), multi-reflection ToF mass spectrometer (MR-ToF MS), pulsed drift tube decelerator plus RFQ cooler–buncher (RFQcb), and a, multi-stage differentially-pumped transfer line to the handover location. Interspersed assemblies allow for focusing and steering using five einzel-lens/steer stations, three PDTs for energy modulation, and multiple MagneTOF detectors alongside a Faraday cup (Schlaich et al., 11 Apr 2025).

2. High-Voltage Deceleration Principle and Operational Timing

The antiproton kinetic energy is reduced using a fast-switched high-voltage PDT. When the bunch is entirely inside the field-free region:

$E_\mathrm{out} = E_\mathrm{in} - q \Delta V$

For $q=e$ and $\Delta V=96$ kV, this results in $E_\mathrm{out} \approx 4$ keV, compatible with downstream Penning trap trapping requirements (Fischer et al., 2024). The switching is performed by a Behlke HTS1501-20-LC2 unit (250 ns trigger-to-action delay, 500 Ω/170 pF RC network yields τ ≈ 85 ns, measured transient ≈80 ns). The timing sequence is tailored such that the PDT is ramped to full voltage ≈9.5 s before antiproton arrival (minimizing outgassing), and switched within ≈3.85 μs of detection.

3. Mass Purification and Ion Handling

The multi-reflection time-of-flight mass spectrometer (MR-ToF MS) is integral for beam purification:

• Design Features:

Adoption of a 62 mm inner diameter for mirror stacks, uniform electrode lengths, and a 480 mm lift mechanism markedly enhance transverse acceptance (≈8× higher than the reference Greifswald device) (Schlaich et al., 2023). SIMION Monte-Carlo simulations demonstrate a solution volume increase from 1.2% to 6.5%, correlating to an acceptance improvement from ≈9.75π mm mrad to ≈75.4π mm mrad.

• Operation:

Ions are injected (3 keV), chopped into ≈250 ns FWHM bunches, and subjected to mass separation (resolving power $R_m ≥ 5 \times 10^4$ typically, up to $R_m ≈ 10^5$; for current ion source, $R ≈ 50,000$ after 150 turns). Deflection gates synchronize to revolution periods for isobaric suppression—critical for isotope-specific studies.

• Cooling and Bunching:

Purified ions are decelerated to ≈300 eV for injection into the RFQcb, which cools and accumulates multiple packets (buffer gas pressure $p ≈ 2 \times 10^{-5}$ mbar). Subsequent pulsed acceleration yields bunches at ≤5 keV, compressed to <300 ns lengths.

4. Vacuum Engineering and UHV Performance

Extreme vacuum integrity is mandatory for both antiproton and ion beamlines, preventing annihilation on residual gas and contamination between species:

• Antiproton Line:

Aluminum tube walls and MACOR insulators exhibit minimal outgassing; non-evaporable getter (NEG) coatings absorb and chemically bind atmospheric gases (H$_2$, O$_2$, N$_2$, CO) after activation at 200–400 °C. All components are rated for bake-out at 250 °C. With fully activated NEG coatings, base pressure at the trap entrance is $\leq 2 \times 10^{-11}$ mbar; typical cycle maintains < $2 \times 10^{-10}$ mbar for ≈75% uptime (Fischer et al., 2024).

• Ion Source/Transfer Line:

Three differential-pumping stages reduce pressure from $10^{-5}$ mbar at the RFQcb down to < $5 \times 10^{-10}$ mbar at the interface with the antiproton beamline—even during active operation. Buffer-gas purity moderates charge exchange losses and influences stacking efficiency, with bake-purified systems supporting >400 ion injections without measurable saturation (Schlaich et al., 11 Apr 2025).

5. Beam Quality Characterization and Transmission Metrics

Experimentally, the system achieves:

• Antiproton Beam:

Final mean energy $\mu = (3898 \pm 3)$ eV, energy spread $\sigma_E = (127 \pm 4)$ eV (88% within ±200 eV), transmission fraction $T(4\,\mathrm{keV}) = (55 \pm 3)\%$ (Fischer et al., 2024). Bunch lengths after deceleration are $(93 \pm 3)$ ns (simulation: 89 ns), compatible with the downstream Penning trap’s temporal acceptance. Spot-size characterization yields 2D Gaussian spreads:

$$\sigma_\mathrm{horiz} = (3.0 \pm 0.1)\, \mathrm{mm}, \quad \sigma_\mathrm{vert} = (3.8 \pm 0.2)\, \mathrm{mm}$$

This is within the Penning trap aperture (radius $r_a = 5.6$ mm; 64% geometric acceptance). Envelope formalism and focusing condition for einzel lenses govern spot evolutions:

$$\sigma^2 = \varepsilon_\mathrm{rms} \beta, \quad 1/f = V_L/(2E) \cdot kL$$

• Ion Beam (Kr Isotopes):

$1 \times 10^4$ ions per <300 ns bunch for all six stable Kr isotopes, including $^{78}$Kr (0.36% abundance). Accumulated FWHM: 250–300 ns; mass resolution $R_m > 5 \times 10^4$ suppresses neighboring isobars below 0.01%. Energy spread $\sigma_E = 13$ eV at 3 keV, transverse emittance $\ll$ acceptance for Penning trap injection (Schlaich et al., 11 Apr 2025).

6. Safety, Conditioning, and Operational Solutions

High-voltage operations required meticulous conditioning and design modifications:

• Leakage and Conditioning:

Initial in-vacuum leakage ($\sim$100 μA at –96 kV) was suppressed to $\sim$11 μA by mechanical rounding of electrode edges, raising isolation distances, and surface polishing.

• Field Gradient Management:

Guard rings at metal-insulator-vacuum triple junctions reduced field spikes.

• System Interlocks:

HV cage meets IP3X standards, with magnetic and trapped-key interlocks; gate valves are tied to CERN access protocols to safeguard ELENA ring integrity.

• Timing Management:

Voltage ramp scheduling restricts pressure increases to <25% of the cycle, maintaining UHV and beam quality for ≥75% of each operational period (Fischer et al., 2024).

7. Implications for Physics Goals and Future Developments

Successful integration and commissioning demonstrate that PUMA can deliver simultaneous, species-pure, high-intensity, sub-μs ion and antiproton bunches to a nested Penning trap under ultrahigh vacuum:

• Neutron Skin Probing in Exotic Nuclei:

By ensuring isotope-pure $N_{\mathrm{ions}} > 10^4$ and $p < 10^{-11}$ mbar, PUMA enables studies of nucleon distribution in neutron-rich nuclei via annihilation localization (Schlaich et al., 11 Apr 2025).

• Reference and Calibration Measurements:

Offline infrastructure (POLIS, MR-ToF MS) supports benchmarking against stable isotopes.

• Ongoing Hardware Evolution:

Upgrades in bunch compression (potential reduction in chopping width), HV supply stability, and RF focusing remain priorities. POLIS hardware is scheduled for integration on ELENA’s line, facilitating the first antiproton–radioisotope capture runs.

• MR-ToF Collaboration:

Distributed fabrication and further optimization via a multi-institutional effort (TU Darmstadt, Univ. Greifswald, Groningen, Manchester, MIT, Mainz, Innsbruck).

In sum, the PUMA experiment’s architecture and subsystem performance confirm readiness for precision studies of antiproton-nucleus interactions at the limits of stability, with demonstrated capabilities exceeding original design specifications (Fischer et al., 2024, Schlaich et al., 11 Apr 2025, Schlaich et al., 2023).