20170106_105232.jpg

Multiple Monoenergetic Gamma Radiography (MMGR)

 

The overarching goal of this project is to develop a novel approach to the active detection of shielded nuclear materials that combines a high sensitivity with low radiation dose. The use of MeV, intrinsically monoenergetic γ-rays for transmission imaging provides a means for both high contrast imaging of shielded objects and the means to confirm the fissile nature of the material. This technique relies on the large difference between the pair production cross section of mid- and low-Z materials and SNM (U and Pu) and, subsequently, the large difference in absorption at high energies. The required monoenergetic γ’s are easily produced using small state-of-the-art proton or deuteron accelerators. The number of objects such as containers which contain no SNM is many orders of magnitude larger than the rare container which might have SNM and, as a result, systems in use are plagued by time-consuming false alarms. To mitigate this problem, after determining that a particular object cannot be cleared, the system can also be used to generate both neutrons and γ’s to provide positive identification. Unlike other approaches attempted to date, the source is in fact a multi-particle source in which both neutrons and γ’s are produced simultaneously. The result is that SNM may be detected in three ways: imaging through shielded material, detection of photofission neutrons and γ’s, and detection of neutron-induced fission products. The multi-particle approach is particularly flexible with respect to the types of shielding that can be penetrated.

A common method of inspecting commercial cargoes for the presence of fissile materials involves the use of 1-10MeV bremsstrahlung photon beams. While simple and reliable, this technique has many downsides, such as the large doses involved, the non-specificity of the detected photons' energy, and the relative inefficiency at triggering NRF and photofission. Much progress in the field of radiography active interrogation can be achieved by developing monochromatic, and possibly tunable gamma sources. The specific, known energies of the photons from a multiple monoenergetic source can allow one to isolate and suppress contributions from photons which have undergone Compton scattering in the cargo and are thus dilute the information about a specific pixel - both in terms of the material information, as well as imaging contrast.

The 3MeV RFQ accelerator at MIT-Bates. CLICK TO ENLARGE.

Past Results with deuteron beams

MIT is part of a four university collaboration (along with GaTech, Penn State, and UMich) using proton and deuteron beams in 11B(d,nγ)12C and 12C(p,p’γ)12C reactions, which produce highly monochromatic photons. In its first for the program made use of a 3MeV deuteron source at MIT-Bates linear accelerator to experiment with (d,nγ) reactions. The 4.4 MeV and 15.1 MeV photons from this reaction are perfectly suited for performing dual energy radiography and thus determine the areal density (image) and effective Z (material) of a cargo pixel

A picture of the accelerator, with the overlaid schematics showing the beam and the target, can be seen above. An (early version) of the detector array can be seen to the right.

The NaI(Tl) detector array.  CLICK TO ENLARGE.

 

Data analysis and simulations

Mass attenuation coefficient's dependence on energy and Z. CLICK TO ENLARGE.

The observed photon spectra will contain information about the cargo material that has been traversed.  The 15.1MeV gammas' attenuation is highly dependent on Z, due to the Z^2 dependence of the pair production cross section. The plot below shows the total scattering cross section for three nuclei of very different Z (carbon, iron and uranium), showing a strong differentiation at 15MeV.  The transmission measurement at 4.4 MeV allows to determine the density, while at 15.1MeV the transmission allows to determine the Z.

 

Energy deposition spectra measurements from the transmitted beam for various materials of equal areal density. CLICK TO ENLARGE.

The spectra to the right shows the results of measurements for a variety of materials of equivalent areal density, showing a dramatic difference in the 15.1 MeV and 4.4 MeV transmission.  At 15.1MeV the interaction is entirely dominated by pair production, while at lower energies Compton scattering is dominant.  Since Compton scattering at ~4 MeV probes the electron density of the material, and since the electron density is approximately proportional to the mass density, transmission measurements at ~4 MeV are best for determining the areal (2D) density of the cargo.

 

Results

Cargo reconstructions for elements from low Z to high Z (such as lead and U).  The material reconstruction infers the the density (right) and the effective Z (center) to within percents.  CLICK TO EXPAND

The results of the reconstruction can be seen to the right. A set of objects were scanned through the 4.4/15.1 MeV beam. The absolute transmission (as compared to "open air" pixels) allows to determine the density, while the ratio of the transmission can be used in a Geant4 based model to reconstruct the Z. As a result the reconstruction fits the nominal values to just a few percent. Currently this is the only and first result where radiographic reconstructions, using monochromatic photons, are used to reconstruct not just images but actual quantitative information about the cargo. For a more detailed description of the experiment see the paper on arXiv.

The future?

The ARI collaboration -- faculty, research staff, students and program managers -- next to ION-12(sc).

While the RFQ system allowed a great proof of concept demonstration (see references 1,2, and 3 below), it is not necessarily the best system for a field application. Smaller, compact, lighter accelerators are needed which can produce continuous wave (CW) beams and enable higher yield and neutron-less reactions - such as 12C(p,p')12C and 16O(p,p')16O. Enter ION-12(sc)! This proton cyclotron is essentially a box with a proton beam fitted into a small, superconducting MRI magnet. It can produce 12.5 MeV proton CW beam, with currencies up to 50 uA. Stay tuned for more results!

Support

This work is supported in part by the U.S. Department of Homeland Security Domestic Nuclear Detection Office under a competitively awarded collaborative research ARI-LA Award, ECCS-1348328 and is part of a collaboration between the Massachusetts Institute of Technology, Georgia Institute of Technology, University of Michigan, and Pennsylvania State University. This support does not constitute an express or implied endorsement on the part of the United States Government.
We gratefully acknowledges the support of the Stanton Foundation Nuclear Security Fellowship. The authors wish to thank the MIT-Bates Research and Engineering Center staff for their invaluable contributions to the construction and operation of the experiment; in particular Peter Binns, Hamid Moazeni, and Ernest Ihloff.

Selected publications

  1. "Experimental Demonstration of Multiple Monoenergetic Gamma Radiography for Effective Atomic Number Identification in Cargo Inspection," Brian S. Henderson, Hin Y. Lee, Thomas D. MacDonald, Roberts G. Nelson, Areg Danagoulian, Journal of Applied Physics, vol. 123 (2018) 17 (arXiv:1802.04225v2)
  2. "Spectroscopic neutron radiography for a cargo scanning system," Jill Rahon et al., NIM A, 820 (2016)
  3. "Initial results from a multiple monoenergetic gamma radiography system for nuclear security," Buckley E. O'Day III et al., NIM A, 832 (2016)

Project team