SNM detection by fission signatures induced by low-energy neutrons

The identification of fast neutron interactions in liquid scintillation detectors, by means of the pulse shape discrimination method (PSD), has been demonstrated by many to be a valid technology both as an alternative to standard neutron detectors and as a direct detection method of fast neutrons. We have previously reported on the use of this technology combined with a pulsed neutron source for the detection of coincident prompt fission neutrons in multiple liquid scintillation detectors. The intention is to develop this technique into a method for detection of special nuclear materials (SNM) in shielded containers. In this paper we present a more detailed investigation of the effectiveness of total, neutron and photon, counting as well as analysis of two and three fold coincidences in observation intervals of 20 nanosecond duration. The observation of neutrons and photons in such coincidence events is a strong indicator for the presence of fissile materials. As the observation interval is very short, the background rate of coincidence events is very small thus favouring a relatively low fissile mass detection limit. We use a pulsed 14-MeV neutron source in a graphite assembly as external neutron source. In the time period of 55-150 μsec after the 14-MeV neutron burst only low-energy source neutrons persist. These source neutrons are not identified as neutrons in the PSD analysis of the detector response whereas the induced fast fission neutrons are. Interrogation only by low-energy neutrons is achieved by time-gating the acquisition system and avoiding the period when fast source neutrons interact with the detectors. The interest in using a neutron source with an energy component above cadmium cut-off is to assure detection of fissile materials which are deliberately shielded against an external thermal neutron source. The paper shows the experimentally obtained sensitivities of the method to the detection of small low enriched uranium samples. The sensitivity to the fissile mass of prompt fission neutron counting versus counting of coincidence events is discussed