Charged particle induced reactions on beryllium as a fast neutron source for irradiation experiments

.


Introduction
The accelerator-driven fast neutron sources are devices where neutrons are produced in bombardment of target by a charged particle beam extracted from accelerator. They represent cheaper, smaller, and cleaner alternative to research nuclear reactors. These neutron sources can be used (as a primary or supplementary tool) for many research activities realized at the nuclear reactors, such as the material research, isotope production, neutron imaging, neutron activation analysis, boron neutron capture therapy, nuclear data measurement, sub-surface exploration, etc. Just as the research nuclear reactors, the compact accelerator-based neutron sources are also capable to be involved in interdisciplinary research. Based on used neutron source reaction, the neutron fields produced by the accelerator-driven neutron sources represent the useful tools, e.g., for application of fast neutron activation analysis, nuclear data validation, radiation harness tests of electronics, and material research. For such applications, the neutron sources that utilize the deuteron or proton beams with energies up to 10 to 30 MeV and thick targets are more convenient. Many cyclotron-based sources in the world utilize thick beryllium layers as targets, and their neutron fields are based on the p+Be and d+Be source reactions.

Materials and methods
The p+Be and d+Be reactions are distinguished by the high value of neutron spectral yield. This unique characteristic is further enhanced by the favourable high value of Be melting point (1300 °C) that enables to use intensive charged particle beams on targets.

The p + Be neutron source reaction
Main production reactions of neutrons during proton bombardment of Be target are 9 Be(p,n) 9 B, 9 Be(p,pn) 8 Be, 9 Be(p,pa) 5 He * , and 9 Be(p,na) 5 Li. The most important reaction on beryllium is the (p,n) reaction, it has a Qvalue about -1.85 MeV [1], and it forms the high energy region of neutron spectrum. Other three reactions produce the general continuum due to the multi-body break-up processes. Typically from a thick beryllium target, a broad or white neutron spectrum is obtained.

The d + Be neutron source reaction
Main neutron producing reactions for Be-target bombardment by a deuteron beam are 9 Be(d,n) 10 B, 9 Be(d,pn) 9 Be, 9 Be(d,p2n) 8 Be, 9 Be(d,na) 6 Li, and 9 Be(d,2n) 9 B. The (d,n) reaction is responsible for high energy part of neutron spectrum, it has a Q-value about +4.4 MeV [1]. Other four reactions form the broad maximum around 40% of the deuteron energy.

Multi-foil activation technique
The multi-foil activation technique is experimental technique that allows determination of the neutron spectrum from the set of experimentally measured reaction rates utilizing the unfolding code (e.g., SAND-II [16]). The reaction rates are obtained from the gamma-ray spectrometry measurements of activation foils that were irradiated nearby the source target. Multifoil activation method is advantageously used for a nonpoint-like geometrical arrangement of irradiation system where other experimental methods are not utilizable.

Beryllium target station NG-2 at NPI
Nuclear Physics Institute (NPI) of the Czech Academy of Sciences (CAS) in Rez operates the NG-2 fast neutron source. This neutron source uses proton beam up to 35 MeV and deuteron beam up to 20 MeV and Be-target station for broad spectrum production or Li-target station for quasi-monoenergetic neutron field production. Neutron generators NG-2 are operated in negative ion mode of acceleration at U-120M cyclotron (which has pulse capabilities). They are primarily focused on nuclear data provisioning and validation for the fusion related research programs DEMO [17], IFMIF-DONES [18], and ITER [19].
The Be-target station (see Fig.3) is standardly operated together with the p+Be source reaction and 35 MeV proton beam, and it provides the white spectrum up to 33 MeV with fast neutron flux up to 10 11 cm -2 s -1 nearby the target. The Be-target used in the NG-2 neutron source has a thickness of 8 mm.

Neutron field determination
To extend the experimental utilization of neutron sources of the NPI mostly towards more traditional research reactor applications, the neutron fields generated by the p+Be and d+Be source reactions for various energies of charged particle beams provided by the U-120M cyclotron have been recently studied. Several new fast neutron fields were measured and investigated at close source-to-sample distances.
In particular, the neutron spectra of the p+Be interaction were measured for 35 MeV, 30 MeV, 24 MeV, and 20 MeV proton beams. Energy spectra of the d+Be neutron field were investigated for three values of deuteron beam at the NPI, namely for 20 MeV, 15 MeV, and 10 MeV. For the neutron spectra determination of the p+Be and d+Be source reactions, the multi-foil activation method was selected. In each experiment, set of several activation detectors for irradiation experiments usually consisted of 10 materials that were sensitive to various parts of the neutron spectra according to activation cross-sections. Irradiation experiments usually lasted for 10 to 20 hours, and the value of deuteron beams current and proton beams currents were 6 to 12 μA. The stacks of foils were located on the Al-holder nearby the Betarget, usually at position P0 and position P14 on the holder (see Fig.4), that means samples were located at distances of 14 mm and 154 mm from the Be-target. Irradiated foils were investigated by means of the nuclear gamma-ray spectrometry technique (HPGe detector). Measured reaction rates were used for the neutron spectrum reconstruction. Usually, about 30 activation reactions were observed in analysed foils in each experiment.

Results
Broad neutron energy spectra of neutron fields based on proton induced reactions on Be were measured for four proton energies, results for experiments with 35 MeV, 30 MeV, 24 MeV, and 20 MeV protons on beryllium together with the MCNPX simulations are depicted in Fig.5-8. The MCNPX [20] simulations are well consistent with neutron spectra reconstructed from measured reaction rates. There is only small deviation in the region below 15 MeV probably due to space integration effect in real activation foils. Table 1 summarizes the total fast neutron fluxes reached in individual experiments with various proton beam energies. The mean energy of each neutron field is also included in Table 1, and the experimental values agree well with expected values based on equation (1).    Results of neutron field measurements for deuteron induced reactions on thick beryllium for three various values of deuteron energies (20 MeV, 15 MeV and 10 MeV) are provided in Fig.9-12. On account the fact that the MCNPX code does not take into consideration the forward orientation of neutron spectrum emitted from the d+Be reaction, the MCNPX simulations are underestimate by factor 2-3 with respect to SAND-II spectra reconstructed from measured reaction rates. Table 2 sums up the total fast neutron fluxes reached in corresponding experiments with d+Be reaction.

Conclusions
At the NPI, several new neutron fields based on the p+Be and d+Be interactions were developed, i.e. four types of the p+Be and three types of the d+Be neutron fields of white spectra are available now. Mean energies of obtained spectra are well consistent with empirical equations proposed by M.A. Lone and C.J. Parnell. The shapes of neutron spectra are in very good agreement with results reported by other authors.
These neutron spectra represent an important tool for integral benchmark experiments, integral validation of cross-sections for fusion related program IFMIF-DONES and testing the radiation hardness of electronics against the fast neutron fields. They also represent a very useful tool for application of fast neutron activation analysis with great potential for interdisciplinary research, such as investigation of historical, environmental, and geological samples, and study of cultural heritage.
The irradiation experiments at the NG-2 neutron source carried out at the CANAM infrastructure of the NPI CAS Rez were