Nuclear Materials Identification

Measurements of proton-induced fission of uranium were performed as a surrogate for photon-induced fission that can be induced in cargo for homeland security applications using high energy bremsstrahlung x-ray sources. The measurements demonstrate that continuous or near-continuous beams can use the multiplicity of prompt neutrons from fission to identify nuclear material much more quickly than typical slow-pulsed (a few hundred Hz) LINAC drivers.

Earlier experiments with a pulsed electron accelerator at the Idaho Accelerator Center demonstrated the ability to distinguish fissile from non-fissile targets by prompt neutron multiplicity. The HRIBF experiments demonstrate dramatic gain in analyzing speed using CW beams to induce fission. The HRIBF experiments employed beams of 10, 14 and 18 MeV protons on ^natU and ^natTi and an array of eight liquid scintillator neutron detectors employing pulse shape discrimination. The observable used in the experiment is twofold neutron coincidences with a neutron energy threshold of 1 MeV. The proton the p+Ti reaction provides a source of events with no more than one neutron in the final state. At 10 MeV proton energies the total cross section for events with a single neutron in the final state is many orders of magnitude (>10⁴) larger that those producing multiple neutrons. Thus neutrons from the Ti target provides a surrogate for neutrons produced by x-ray interrogation of innocuous high-Z materials via (γ,n) reactions. Such single neutron events can give rise to random coincidences (Poisson process) and coincidences generated by scatter of a single neutron from detector to detector.

Fig. 1 - Schematic illustration of relevant neutron emission processes in photofission.

Results from the initial run are illustrated in Figure 2 and 3. To simulate low detection efficiencies which might result from the presence of shielding materials under realistic conditions, measurements similar to those illustrated in Figure 3 were made at a variety of target to detector distances ranging from 0.3 m to 2.0 m (efficiencies of ~0.04 to 0.002). With the lowest efficiency, an unambiguous signal was obtained in about 60 s. Further experiments at HRIBF are planned.

Fig. 2 - Coincident neutron time difference spectra. True coincidences from proton-induced fission of uranium show up as a peak (red). Uncorrelated neutrons from Ti(p,n) show no peak at Δt=0, but can have peaks away from zero due to multiple detection of a single neutron.

Fig. 3 - True twofold neutron coincidence rate is shown as a function of counting time. An ~5σ signal is obtained in ~30s. Using a 120 Hz pulsed beam at IAC and a similar detector setup, ~1 hour was required to obtain an ~3σ signal.

Collaborators