Andrew Hoffman and Dr. Lawrence Rees, Department of Physics and Astronomy
The BYU nuclear research group has over the past few years been attempting to design neutron detectors for the use of both scientists and homeland security applications. When radioactive materials are being transported, they can emit a broad spectrum of particles including gammas, betas, alphas, and neutrons. Specifically plutonium is fissile enough (it can sustain a chain fission reaction) such that it emits enough neutrons to detect while being transported. Neutrons unlike gamma rays can penetrate material used for shielding gammas like lead and bismuth. They are also unique to heavy metals used for building nuclear weapons so they can be used to distinguish between isotopes used for most medical purposes which are primarily gamma emitters. It is for these purposes that the use of neutron detectors has become so crucial for portal monitoring at ports and border checkpoints.
One of the problems with the detectors being designed by our research group is that they tend to have higher gamma sensitivities than some alternative designs. Gamma sensitivity means that the plastic scintillator (material used to detect events) will detect both gamma rays and neutron events. This can be the cause of some problems with a detector, because multiple gamma rays could potentially be mistaken for an actual neutron. It also increases the amount of noise in the data that needs to be filtered to distinguish a neutron event.
It was my idea to use some kind of gamma shielding to reduce the number of gamma rays that reached the detector while still minimizing the number of neutrons that may be lost or scattered by the shielding. The two materials studies were lead and bismuth. Both lead and bismuth are dense heavy metals so they are able to absorb gamma rays extremely well. Their nuclei, however, are very stable and therefore do not absorb or scatter neutrons very well. The advantage of lead is its density. Since it has a higher density than bismuth it can shield more gamma rays. Bismuth has several advantages. It is more stable than lead, and therefore scatters and absorbs fewer neutrons than lead. It is non-toxic and safe to handle. It also consists of only non radioactive isotopes so it would not have its own background radiation (typically lead contains a small fraction of radioactive isotopes as well as radium).
Initially computation computations were run using MCNP (Monte Carlo Neutral Particle) and PoliMI. The results of these computations showed that there is no significant advantage to adding additional shielding beyond four to six inches of lead or bismuth. With three to six inches of either material the results showed over 99% of low energy gammas being shielding and about 90% of higher energy gammas being shielded. These same calculations were run to calculate the number of neutrons being absorbed by the shielding material. It was predicted that bismuth would allow significantly more neutrons through the shielding. Computational results showed that bismuth shielding allowed more low energy neutrons through by a factor of ten. Because of the computational results, timing, and equipment restraints only bismuth shielding was tested with the detector.
The bismuth shielding was produced by taking raw bismuth and melting it down into inch or half inch discs made to fit the diameter of the detector. These discs were then stacked and taped together for a total of three inches of shielding.
Data was first taken with a bare source in order to compare the shielding experiment results. To reduce room return (neutrons bouncing off the floor/walls) reaching the detector, we set the source and detector on a scissor lift 22′ high. A 252Cf source was placed 12.75 in away from the detector. A sixty minute data run was taken, as well as a thirty minute background run. Three more runs were taken to compare the results with the gamma shielding. The first run was with 3 inches of bismuth shielding with a 252Cf source placed 12.75 in away from the detector. The second run was with a 60Co source placed 12.75 in above the bare detector. The third run was taken with the same 60Co source above the detector with 3 inches of bismuth shielding. All runs were taken for one hour. Another one hour background run was taken to ensure accuracy of the background subtraction. The same hardware and software settings were used as with the bare 252Cf run.
The way the data is recorded, the software distinguishes between a neutron and a gamma event. A neutron will bounce around in a plastic scintillator material dissipating energy, and then is captured by a cadmium nucleus which will then emit a gamma. A gamma will just dissipate energy into the plastic. This means that a neutron will have a “double pulse” in the detector, or will have two events. There were issues in the actual analysis of the data recorded with the formatting of the raw data. Because of time restraints I was not able to rewrite analysis code to further filter out gamma events that may have triggered a “neutron event”. The actual efficiencies calculated may be very inaccurate due to the hardware and software calibration issues discussed above, but the patterns give us an idea of the usefulness of the gamma shielding.
Without shielding the calculated intrinsic efficiency of the detector was around 13.4% (meaning 13.4% of all neutrons emitted in the direction of the detector were actually detected). When shielding was added this efficiency dropped to about 3.4%. When using the bare gamma source 1.31% of the gammas hitting the detector were mistaken as neutrons, and when the bismuth shielding was added this dropped to .124%. The correlation between the neutron efficiency dropping by a factor of ten and the gammas being mistaken for neutrons dropping by a factor of ten could be evidence that the shielding is does shield the gammas without harming the neutron efficiency significantly as predicted. The research conducted for this ORCA grant was published as my senior thesis and can be found on the BYU physics webpage with much more detail: http://www.physics.byu.edu/Thesis. My research will be continued by other undergraduate students in the group, and will hopefully (if successful) be implemented in portal monitors are border checkpoints and ports.