Workshop on the Role of the Nuclear Physics Research
Community in Combating Terrorism
Washington, DC, July 11-12, 2002
Conventional Explosive and Weapon Detection Working Group
Subgroup on Gamma-Ray Beams
B.L. Berman (Chair, GWU), A. Afanasev (JLab), W. Bertozzi (MIT),
T. Brethauer (TSWG), P. Farrell (BTG), G. Merkel (ARL), P. Thieberger (BNL), and
L. Wielopolski (BNL)
The Subgroup on Gamma-Ray Beams met on July 12 to frame its report, which
was delivered orally in preliminary form that day to the Conventional Explosive and
Weapon Detection Working Group. This document summarizes the substance of that
report. Recognizing that the technologies presented in the poster session were a limited
subset of the potentials approaches utilizing gamma-ray beams, the Subgroup focused on
the broader technical and operational characteristics of such beams which make them
attractive as a tool in this application.
The principal problems addressed by the Subgroup had to do with detecting and
imaging explosives in luggage and/or cargo. The challenge is to find small (sub-
kilogram) quantities of high explosives in air-freight containers (of the order of
8810), in the presence of normal cargo and, perhaps, deceptive shielding, and to find
large quantities (hundreds of kilograms) in maritime containers (of the order of
8840). In either case, the scanning would have to be done in a few minutes. Several
schemes based on gamma-ray beams have been proposed to do this.
A related problem is to equip first responders to the threat of a ―mysterious
possible terrorist infernal device‖ with a portable or transportable interrogation system
with which to examine the device without moving it. Here, however, since the device
would be singular and therefore the first responders need not be in a hurry, a variety of
imaging devices could be used and good statistics could be exploited. If the device were
chemical or biological in nature, ionizing radiation (at doses significantly higher than
those used for diagnostic imaging and analysis) might be able to neutralize it; if it were a
bomb, conventional, nuclear, or ―dirty,‖ its components could be identified. Gamma-ray
beams could play several roles in the above scenarios.
Gamma-ray beams have certain characteristics that are markedly advantageous to
explosives and weapons detection:
Penetrability. Gamma rays in the energy region of 1-10 MeV pass through matter,
especially low-Z materials, with very little attenuation, relative to other kinds of
radiation. In this energy region, photoelectric absorption has fallen sharply from its high
level at lower energies, pair production has not yet reached its high level at higher
energies, and Compton scattering does not have a large cross section. This characteristic
is particularly advantageous for rapidly probing large cargo containers, which frequently
contain large amounts of low-Z material, such as food and textiles. Of course, cargo
containers sometimes contain higher-Z materials, such as appliances and auto parts, but
even these materials are usually characterized by low packing densities, and hence
amenable to interrogation by gamma-ray beams. By contrast, the ability of neutron
beams to penetrate to the center of large cargo containers is severely limited.
Insensitivity to hydrogen. Gamma-ray absorption and scattering are minimal for
hydrogen, unlike the case for neutrons. This characteristic is also particularly
advantageous for large cargo containers, which often contain large amounts of
hydrogenous material that tends to thermalize neutrons and then absorb them.
Low activation. Gamma-ray beams of less than 10-12 MeV induce little activation in
common low-Z materials (C, N, O, Al, Si), which have either high photonucleon
thresholds (12C, 14N, 16O, 27Al, 28Si) or else lead to stable nuclei with (, n) reactions (2H,
C, 15N, 17,18O, 29Si, 30Si), and gamma-ray beams of less than 7-8 MeV induce no
activation in nearly all stable nuclides. Also, secondary neutrons from (, n) reactions do
not have the high flux of a neutron beam. This is made apparent by the fact that U.S. and
overseas regulatory agencies have approved MeV gamma-ray beams for food irradiation
and they are used widely.
High intensity. Gamma-ray beams, especially from linacs and microtrons, can be very
intense, and the electron accelerators that produce them constitute a mature technology,
are long-lived (>104 hours between maintenance periods), and, especially important for
counter-terrorism cargo-screening applications at air and sea ports of entry, are very
Ease of image processing. Image processing, a vital requirement for explosives and
weapons detection, is made far easier and more reliable with high-flux beams and the
high counting rates they make possible. Spatial resolution of gamma-ray detectors is
significantly higher than is the case for neutron detectors. Millimeter resolution without
distortion is feasible, and while still not quite as fine as for x-ray film, the difference is
Multi-functionality. The same gamma-ray beams that can probe unknown materials
in order to identify their isotopic composition can also serve to perform an imaging
function. Even chemical discrimination might be feasible, but this has not yet been
demonstrated in the field.
Universality. Gamma-ray beams can be used to identify any material, since
photonuclear cross sections, including those for photoneutron production and
photofission (of the actinides) are substantial (for sharp resonances and in the giant-
resonance energy region) throughout the periodic table.
Specificity. Virtually any isotopic species is uniquely identifiable, using resonant
gamma absorption (RGA) or scattering (RGS, or resonance fluorescence).
Reliability. We should repeat this twice. No technology that is not robust can be
operationally deployed and entrusted with our safety.
For our purposes here, there are two main classes of gamma-ray beams—
monochromatic and continuous. Monochromatic gamma-ray beams are suited to
transmission measurements like RGA. They are made by nuclear reactions of specific
interest, and can be as narrow-band as the (thermally Doppler-broadened) nuclear energy
levels themselves. For example, the 13C(p, )14N reaction is used to produce gammas that
are absorbed preferentially by 14N, a major component of most modern explosives. Use
of compound (layered) targets can produce more than one energy beam, so that isotopic
ratios (e.g., 16O:14N) can be measured as well. This method thus has a high degree of
specificity (for those elements for which it can be used), and is compatible with existing
elemental-ratio databases for the most common types of explosives and drugs of abuse.
Continuous gamma-ray beams, well suited to scattering measurements like RGS
and to pulsed-gamma analysis (PGA) can be produced as the familiar bremsstrahlung
from an electron beam striking a radiator, usually a high-Z target. Because
bremsstrahlung beams can be very intense, and because all elements emit characteristic
―signature‖ gamma rays, either from scattering or following short-lived activation, these
techniques can be used to interrogate a container to identify all of the materials contained
therein by identifying the signature energies of the gammas measured with high-energy-
resolution detectors. These methods, in addition to their specificity, thus have a high
degree of universality.
Other gamma-ray beams, partly monochromatic, produced from positron
annihilation in flight, tagged bremsstrahlung, or coherent radiation from crystals (e.g.,
channeling radiation or coherent bremsstrahlung), can be used for special purposes. Over
the years, all of these production techniques have been proven as well.
Finally, we wish to point out that both the RGA and RGS techniques (as well as
the techniques based on photofission for detection of 235U and 239Pu, as well as other
fissionable materials, such as 233U, 237Np, and 241Am) depend on having precise and
reliable nuclear data, including the properties of nuclear energy levels and transitions
(energy, width, multipolarity [to determine the angular distribution of scattered photons]),
photonuclear cross sections [(, n), (, 2n), (, p), (, f)] and neutron multiplicities, and
gamma-ray attenuation coefficients. There was a series of posters from the new HIGS
facility at TUNL which outlined the capability (present and future) of this facility to
produce these kinds of needed input data, using the intense, polarized gamma-ray beam
produced by the Compton backscattering of photons produced in their FEL. We were
impressed by the possibilities for making substantial improvements to the existing body
of such data, for many crucial isotopes throughout the periodic table, especially when the
HIGS facility will be able to realize its full potential in terms of intensity.
We refer the reader to a previous AFTAC report [S.D. Gardner, C.M. Frankel,
and B.L. Berman, NIS6-94:341SDG (1994)] that outlines the underlying nuclear theory
and compares and contrasts several of these techniques with regard to explosives and
This report endorses no particular technique or practitioner, and does not reflect
an official position of any academic institution or government agency. We refer the
reader to several of the poster papers that were presented at the Workshop, and we
include herewith short summaries of their contents.
We thank A. Fainberg (TSA), who served as a consultant to the Subgroup, for his
Nuclear Resonance Fluorescence in Material Detection and Object Imaging
All nuclei have characteristic excited states at energies that are unique. Many of
these states have strong probabilities for electromagnetic excitation and can usually be
found in the lowest 7 MeV. By using a beam of photons that is continuous in energy
(bremsstrahlung), we ensure that there are photons at all energies and all elements in the
beam can be excited to their characteristic states (resonance absorption). These states
then decay mostly back to the ground states with the emission in all directions of photons
of the characteristic energies (resonance fluorescence, or RGS). Detecting these
characteristic photons provides a signal that identifies the elements in the sample
uniquely. Collimating both the photon beam and the viewing direction of the detectors
provides spatial imaging. An object can be identified by its spatial position and its
The resonant absorption makes a ―hole‖ in the transmitted spectrum of photons
and this is also a signal of great importance. Typically, for a 5-MeV photon the resonant
cross section has a peak value of about 500 barns, depending on the angular momentum
of the states involved. Because of thermal Doppler broadening, these states, which are a
fraction of an electron volt wide, are spread out over about 20 electron volts for a typical
nucleus with mass of about 16. The effective peak cross section is thus reduced to about
3 barns, depending on the photon energy. These resonant nuclear cross sections are
significantly larger than the photoelectric, Compton and pair production cross sections
that normally absorb photons out of the incident beam. The resonant absorption that
makes these ―holes‖ can be measured by detecting the resonant fluorescence (scattering)
of the transmitted beam by a sample that contains all the elements of interest. A separate
shadow image results for each resonant energy. Each image is like a standard X-ray
shadow image, but now each of these images is element specific. For a single object,
these element specific images all coincide in the shadow.
Ideal detectors for such a system would be Ge photon detectors. Typical photon
energies (among many others) are 2.313 MeV for 14N, 4.438 MeV for 12C, 6.917 MeV for
O, 2.982 MeV for 23Na, and 5.248 MeV for 29Si. Placing detectors at angles greater
than 90 degrees with respect to the incident beam reduces the background due to
Compton scattering because the scattered photons are all reduced below 0.5 MeV and
easily filtered. A system using an electron accelerator of 10 ma at about 8 MeV would
allow one to examine a large sea-going container for practically every element we know
with good counting rates and scans might take as little as a few minutes. All the elements
for such a system are commercially available. Fissionable materials could also be
identified by the technique. Examples of spectra were presented [Bertozzi poster] for
materials containing C, N, O, Si and Na. Even the 4% isotope of silicon (29Si) in glass
was easily detected. The method is compatible with the detection of almost any
Identification of Explosives and Fissile Material Using Pulsed Gamma Analysis
The Pulsed Gamma Analysis (PGA) technique [Afanasev] is an alternative
implementation of the Alvarez method [Alvarez]. It is based on photoneutron reactions
with a threshold near 30 MeV. An electron accelerator with an output energy of 50 MeV
is used to produce bremsstrahlung gamma-rays. The gamma-rays are directed at the
object under inspection in order to knock out neutrons and protons from nitrogen, carbon
and oxygen nuclei, thereby producing very short-lived radioactive isotopes. Beta decay
of these isotopes typically occurs in a fraction of a second and results in the emission of
element-specific monoenergetic gamma-rays in the range 2-5 MeV that can be detected
to determine the relative concentration of the above elements. The counting rates and
penetration depth of this approach should exceed competitive methods and provide a
needed solution for the inspection of cargo containers and other large volumes, even in
the presence of large quantities of hydrogenous materials.
If the interrogated object contains concealed fissile material , its presence will be
indicated by delayed fast neutrons from photofission. Therefore the same accelerator-
based detection system equipped with neutron and gamma detectors will be able to
identify conventional explosives, fissile material or their combination that may occur, for
instance, in Radiation Dispersion Devices (‗dirty bombs‘).
The pulsed time structure of the scanning photon beam is needed to avoid
background neutrons and gamma-rays and to observe clean signals from the target
[Afanasev] A. Afanasev, V. Afanasev, V. Rudychev, Pulsed Gamma Analysis for
Detection of Explosives, US Patent pending.
[Alvarez] Luis W. Alvarez, Nitrogen Detection, US Patent 4,756,866 (1988).
Vacuum Insulated Tandem Accelerator for Counter-Terrorism Applications
A new high current vacuum insulated tandem accelerator can be used for
production of monochromatic photons for detection of explosives and other contraband.
Prior efforts to use the method of nuclear resonance absorption (NRA) to detect nitrogen
using the inverse reaction 13C(p,)14N failed due to limitations of the accelerator to
transmit the necessary proton current. The vacuum insulated tandem (VITAN)
accelerator is designed to transmit up to 40 mA of protons at energies up to 2.5 MeV.
The accelerator features:
No glass or ceramic accelerating columns for beam transport. The charge-
exchange canal is situated in a vacuum tank and high voltage is applied to it through an
insulating column that is remote from the transport region of the high current ion beam.
The gas stripper canal is enclosed in a system of coaxial cylindrical shells that
provide an optimum potential gradient for beam focusing and for grading the potential to
prevent high voltage discharge. The shells are vacuum transparent because they possess
a large number of openings.
Apertures in the walls of the vacuum tank and in the coaxial shields focus the
accelerated ion beam.
Efficient pumping of the inner cavity, which contains the gas stripper, is
accomplished by a combination of cryogenic recirculating pump in the vacuum vessel
and conventional turbo-molecular pumping through the removable covers of the
Energy resolution is ~ 0.1%.
The accelerator can be used to drive several reactions that are useful for detection
of materials and mapping of hidden spaces. Example processes include:
Reaction Energy (MeV) Application
C(p, 1.7476 MeV photons for Nuclear Resonance Absorption (NRA)
Positrons for producing tunable source of MeV photons
F(p,ee)16O 1.8 – 2.3
by in-flight annihilation for NRA
MeV photons for Photon-Induced Positron Annihilation
F(p,)16O 1.8 – 2.5
Li(p,n)7Be 1.8 – 2.5 Epithermal to MeV neutrons
The accelerator will be built through a collaboration of Brookhaven Technology
Group, Inc., Setauket, NY and Budker Institute of Nuclear Physics, Novosibirsk, Russia.
Gamma Nuclear Resonance Absorptiometry
Broad-beam gamma radiation can be used for inducing elemental emission lines
through (,n) or nuclear-resonance reactions that can be measured using nuclear
spectroscopy detection systems. However, these measurements can be augmented when
using a finely tuned monoenergetic gamma beam that overlaps with one of the levels in
element of interest. Gamma Nuclear Resonance Absorptiometry (GNRA) is made
possible by nuclear reactions of specific interest, and can be as narrow-band as the
(Doppler-broadened) nuclear energy levels themselves. For example, the 13C(p,)14N
reaction is used to produce 9.17-MeV gammas that are absorbed preferentially (at the
appropriate laboratory angle, about 80.7) by the 9.17-MeV energy level in 14N, a major
component of most modern explosives. Use of compound (layered) targets can produce
more than one energy beam, so that isotopic ratios (e.g., 16O:14N) can be measured as
well. The narrowness of the nuclear-resonance levels (122 eV width in the case of the
9.17-MeV level in nitrogen), dictates the use of proton accelerators for production of
resonance recoil-compensated gamma radiation. Bremsstrahlung radiation with a total
width of about 10 MeV would need to be much more intense to produce the same yield of
interactions. Thus the monochromatic radiation has the crucial advantage of reducing
any activation of or radiation damage to the cargo content by a large factor. Imaging
with a resonance-detector array for GNRA allows measuring simultaneously the total and
the nitrogen attenuations employing pulse-shape discrimination. The system can be used
for small (luggage) or large (shipping) containers, interrogation. The narrow fan-beam
configuration of the resonance beam makes it particularly amenable for high throughput,
using multiple inspection stations on the same system.
This method thus has a high degree of specificity. The feasibility of the GNRA
method using resonance detectors has been demonstrated in several experiments where
small amounts of explosive simulants were used. At present the effort is concentrated on
constructing a full-scale facility.