The development of compact neutron sources

Many small research reactors used as neutron sources are being shut down. To replace them, new facilities are being developed. In particular, compact accelerator-based neutron sources can take up many of the activities previously supported by reactor-based facilities.


Small research reactors
During the 1950s and 1960s, numerous laboratories and universities built and operated dozens of research and test reactors 1 , smaller and less powerful than power reactors and configured to provide neutron beams and pool access. That was the optimistic era of 'atoms for peace' programmes around the world, aimed at promot ing peaceful uses of nuclear reactors, balancing the fear of nuclear weapons.
The reactors were of many different types, opera ting at total thermal powers between 1 and 10 MW, with steady thermal neutron fluxes between 10 12 and 10 13 n cm −2 s 1 (for comparison, large facilities the fluxes are ~10 15 -10 17 n cm −2 s −1 ) and supporting mostly local activities. Most offered nearcore materials irradiation (neutron damage) and activation analysis facilities, used to study neutroninduced radiation damage in power reactor materials and to determine the concentration of elements in materials in a nondestructive way, as well as for medical and industrial isotope production. Also common were external neutron beams using arrays of instruments for slowneutron scattering research examining atomic structures and motions (materials science) and for neutron cross section measurements, which enable design and production of medical iso tope production systems. Moreover, among many other applications, they enabled reactor physics tests and laboratorycourse demonstrations and supported nuclear data and nuclear astrophysics measurements by supplying cross sections for neutronmodulated nuclear reactions relevant to astrophysical isotope production.
Around most of these reactors grew programmes of nuclear science education, along with the development and improvement of smallsource technology (espe cially moderators -media adjacent to the primary fastneutron source that reduce the speed of neutrons, turning them into thermal and cold neutrons -relevant to both small and large sources) and of neutron physics and materials science instrumentation, neutron detec tors, reactor physics tests, radiation applications and training. Many of these became fullfledged university nuclear engineering departments, with undergraduate and graduate education in broad programmes on radi ation physics and power reactor development. About 15 neutron summer schools have been established -a number that is destined to grow -to train students in the use of local facilities and to help them when entering research at the major international user facilities, neutron sources based on fission and spallation reactions.
What's being lost A large fraction of these original small neutron sources have now been or will soon be shut down. Some have been dismantled because of fear of reactorrelated inci dents, none of which ever involved laboratorybased or universitybased reactors. Capabilities are being lost as the small reactors are shut down, especially device proto typing, technique development and education and prac tice in the field, as well as opportunities for handson student projects and laboratory experiments.
Also being lost, as some of the larger of the small facil ities close, are sources for production of medical isotopes (especially 99 Mo), for materials science experiments based on neutron scattering and for neutroninduced transmutation doping of highpurity silicon to pro duce precisely controlled volumetric ntype 31 Pdoped silicon for use, for example, in photovoltaic cells (the reaction is 30 Si + n→ 31 Si→ 31 P + e + ). The concentration of expertize in terms of engineers and scientists residing at the smaller reactors is also diminishing, with a con sequent loss of resources for the education and training of young scientists.

New sources
These changes motivated the development of new neu tron facilities, in particular compact acceleratordriven neutron sources (CANS) based on lowenergy (typically less than 100 MeV) chargedparticle reactions produc ing neutrons. A schematic representation of a CANS is shown in Fig. 1. Simply put, the point is not that CANS are much better than reactors, but that they are different, supporting a broad range of applications. Most CANS operate with deuteron-deuteron (~2 MeV), deuterontritium (~14 MeV), proton-beryllium (~2 MeV) or The development of compact neutron sources

John M. Carpenter
Many small research reactors used as neutron sources are being shut down. To replace them, new facilities are being developed. In particular, compact accelerator-based neutron sources can take up many of the activities previously supported by reactor-based facilities.
Argonne National Laboratory, Lemont, IL, USA. e-mail: carpenterjohnm@ yahoo.com proton-lithium (~2.4 MeV) reactions. Operation can be either steady or pulsed. Neutrons slow down in room temperature moderators, resulting in Maxwellian neu tron spectra with a temperature of ~300 K and a mean energy of 0.025 eV. As sources of neutron beams, small acceleratordriven sources can compensate for the loss of some of the reactorbased installations. Furthermore, these acceleratorbased facilities are also sources of chargedparticle beams, which are useful for other purposes such as ion implantation.
By comparison, reactors are steady sources operat ing continuously, with neutron energies and moderated neutron spectra similar to those of CANS. Both CANS and reactorbased sources can be paired to cold mod erators, producing neutron spectra with temperatures as low as 20 K, depending on the moderator temperature and configuration.

UCANS
The Union for Compact Acceleratordriven Neutron Sources (UCANS) was formed in 2009 to support the ongoing development of small acceleratorbased neutron sources around the world and to promote the exchange of information on emerging science and novel applications relevant to longpulsed and/or mediumflux neutron sources.
UCANS is a loosely organized collaboration of inter national laboratory and university personnel meeting about once a year. Laboratories volunteer their spon sorship and lead the meetings in turns. Scientists from In the 2018 UCANS meeting 2 the discussion focused on accelerator and beam optics, neutron detection, nuclear astrophysics and nuclear data, target radiation damage and heat removal, CANS projects and inno vative instrumentation, medical applications, material characterization, computer simulations and data analysis software, optical devices and education.

Applications of CANS
CANS allow for easier tailoring of the neutron spectra compared with large facilities. Some materials science investigations carried out with high resolution and throughput rate at major sources rely on preliminary tests at a small source. However, existing CANS also support a variety of applications that go beyond meas urements at bigger facilities 3 . For example, they are used in the developing field of characterizing cultural arte facts 4 , and several groups are developing and deploy ing portable neutron radiography technologies for the nondestructive inspection of infrastructure, such as roads and bridges, and cargo to detect contraband nuclear materials. Fissile materials give themselves away by emitting fissiongenerated neutrons with a time and energy signature detectably distinct from that of neu trons scattered from common nonfissionable materials. CANS cargo inspection facilities are probably inexpen sive enough to be deployed permanently at ports that need nuclear detection capabilities on a regular basis, whereas mobile inspection CANS could be stationed and moved around when called for.
CANS are also well suited for the simulation of singleevent upsets, which are bit flips in semiconductor devices caused by fast neutrons generated by scattering events in the high atmosphere. Another application of CANS that was long foreseen and hoped for is boron neutron capture therapy, a noninvasive therapy in which a boron isotope with a high cross section for capturing slow neutrons is concentrated in cancerous tumour cells, so that when the tumours are irradiated with neu trons the isotope absorbs them and emits highenergy alphaparticles that kill the tumour cells. However, pro gress in this area is hampered by the need to develop pharmaceuticals containing boron that target cancer cells. Finally, there is a pressing need for alternatives to reactorproduced medical isotopes, particularly (but not only) 99  useful for different types of applications and a thorough discussion of the existing facilities can be found in reF. 3 . All these applications benefit from the intrinsic advantages of CANS over reactors and large, multiuser neutron facilities, including low acquisition and mainte nance costs, flexible scheduling, lighter safety concerns, easier access and local management.
To conclude, let's look at the case of the 2 MW Ford Nuclear Reactor at the University of Michigan's Phoenix Memorial Laboratory, which opened in 1957 and closed in 2003 and is a good example of the response to the shutting down of an atoms for peace reactor. The local Nuclear Engineering and Radiological Sciences Department has installed a deuteron-deuteron (2.2 MeV) and a deuteron-tritium (14 MeV) neutron generator in its new Neutron Sciences Laboratory and an array of lowenergy chargedparticle accelerators in its Ion Beam Laboratory 5 (using, for example, protons, deuterons and alpha particles). These two small facilities serve both teaching and research purposes involving students and faculty.