3.1 Neutron source simulation and moderation

There are many factors that affect the neutron flux via ⁷Li(p,n)⁷Be reaction such as target thickness, outgoing angle, and proton energy. In addition, with the increase of the incident proton energy, the proton has a large range, which can produce more accompanied γ-rays via ⁷Li(p,p′)⁷Li and ⁷Li(p,γ)⁸Be reactions [24] that are the harmful biological dose. An optimal experimental scheme should be based on the balance of these three facts for neutron production and moderation.

The neutron energy and yield via near threshold ⁷Li(p,n)⁷Be reaction were investigated using TARGET [20, 21] code. In this work, two proton separation threshold energies (1.885 MeV and 1.930 MeV) and two target separation thicknesses (6.2 μm as the thin target and 152 μm as the thick target) were chosen to calculate neutron production.

In addition, different kinds of moderator materials were used to enhance the intensity of resonance neutrons. These materials should have a high scattering cross-section and a low capture cross-section in the keV range. They also do not exhibit decomposition in the radiation field, high long-term radioactivity or moisture during long time operation [22]. The following materials as water (H₂O), graphite, heavy water (D₂O), and polythene (CH₂) have been suggested. The moderation effects of the aforementioned materials were calculated using the Geant4 (Version: 10.2) Monte Carlo code [28]. In this study, the geometric model was built according to the actual experimental conditions presented in Sec. 3.2, and the cylindrical moderator close to the neutron source was placed in the collimator aperture. The GGST_BERT_HP [29, 30] has been used as a standard physics model to describe neutron transportation, which employs a high precision neutron package used for neutrons with energies below 20 MeV, down to thermal energies.

3.2 Experimental neutron source

The time of flight (TOF) method was used to measure the neutron beam via near threshold ⁷Li(p, n)⁷Be reactions based on the HI-13 Tandem Accelerator at the China Institute of Atomic Energy [25]. The minimum energy of the proton beam generated by the accelerator was approximately 7 MeV, and the pulse length was less than 3 ns (full-duration at half-maximum). An aluminum slice with a thickness of 280 μm was used as the absorber to reduce the proton energy to a mean energy of 2.5 MeV.

Although ⁷Li(p,n)⁷Be reaction is favorable for enhancing neutron intensity in the keV range, some difficulties are still encountered when lithium is used as an ideal target. As shown in Table 2, the low melting point and poor thermal conductivity of pure lithium make it a poor candidate. A good water-cooling system is essential for the timely removal of heat from the target, otherwise, the integrity of the lithium target will be compromised. The chemical properties of lithium cause it to rapidly react with air to form compounds, which directly affects the neutron energy and yield. During the fabrication and installation of the lithium target, it is not allowed to come into contact with air [16].

A schematic of the experimental setup is shown in Fig. 2. A lithium glass detector (ϕ 1.6 cm×0.95 cm) used to monitor the neutron beam was positioned at 0° with respect to the proton beam axis. The detector signals were used to trigger a Time-to-Digital Converter (TDC) to reduce dead-time, while the delayed signals of proton beams were recorded for data acquisition of neutron flight time. Two parameters were registered simultaneously for each event: the pulse height (PH) and the TOF. The detection efficiency of the lithium glass detector was previously calibrated by partners [23]. The experimental parameters are summarized in Table 2.

Fig. 2.Full-screen View

(Color online) The experimental setup. Left figure shows the profile of the neutron source shield and collimator system, and right figure shows the structure of the lithium target.

4.1 Simulated neutron spectra and discussion

The calculated neutron energy spectra obtained via the near threshold ⁷Li(p,n)⁷Be reaction under different conditions are shown in Fig. 3, and the neutron yields under the different energy ranges are shown in Table 3. The neutron spectra at 0° angle have an obvious peak at 30 keV, while the resonance neutron yields are very small. Compared with the target thickness, the incident proton has a greater influence on the neutron spectra and yields. According to the TARGET results, the neutron energy spectrum corresponding to the 0° output angle in Fig. 3(d) was selected to study neutron moderation at a high flux.

Fig. 3.Full-screen View

(Color online) The neutron energy spectra via ⁷Li(p,n)⁷Be reaction under different conditions. (a) represents the conditions of 1.88 MeV proton beams and a thick target (6.2 μm), (b) represents the conditions of 1.88 MeV proton beams and a thick target (152 μm), (c) represents the conditions of 1.93 MeV proton beams and a thick target (6.2 μm), (d) represents the conditions of 1.93 MeV proton beams and a thick target (152 μm).

The moderation effects of different kinds of materials was calculated using the Geant4 code; the results are shown in Fig. 4. The resonance neutron yields increased gradually with the thickness of the moderators until there was a dynamic balance between the neutron absorption and moderation. Graphite could sufficiently moderate fast neutrons (>30 keV) in order to enhance the intensity of the 13.5 keV neutrons, while the hydrogen-containing materials slowed down more neutrons to epithermal energies. The resonance neutron yields in the 12.6-15.8 keV range could reach 6.45×10⁵ sr^-1 μA^-1, 6.93×10⁵ sr^-1 μA^-1, 1.29×10⁶ sr^-1 μA^-1 and 5.97×10⁵ sr^-1 μA^-1 respectively. According to the calculated results, graphite is an optimal moderator to generate more resonance neutrons.

Fig. 4.Full-screen View

(Color online) The neutron energy spectra after (a) H₂O, (b) D₂O, (c) Graphite, and (d) CH₂ moderation of different thicknesses with calculated neutron flux.

4.2 Experimental neutron spectrum and discussion

A two-dimensional spectrum of "PH-TOF" based on the measured raw data is provided to allow for the discrimination of the n-γ signals, as shown in Fig. 5. The region labeled peak-1 corresponds to 240 keV neutrons because of the high detection efficiency of the lithium glass detector. The region labeled peak-2 corresponds to the 560 keV neutrons produced by the 2.5 MeV protons at the surface of the lithium target. The neutron flight distance was so short that these neutrons (< 20 keV) could not be discriminated using the TOF method. They were therefore treated as fast neutrons (> 700 keV) produced in the next pulse because their flight times were longer than that of the proton pulse period. According to the calculated results, the neutron peak at 30 keV should be clearly visible. In fact, the recorded neutron flux was moderated by the water used to cool the lithium target.

Fig. 5.Full-screen View

(Color online) The two-dimensional spectrum of "PH-TOF" measured with lithium glass.

We can obtain the neutron energy spectrum by combining the two-dimensional spectrum (Fig. 5) and the detection efficiency of the lithium glass after ticking out the γ-ray background, as shown in Fig. 6. The right peak corresponds to the region labeled peak-2 in Fig. 5, while the left peak corresponds to the approximately 180 keV neutrons produced by the 1.9-2.0 MeV proton beams. There are three issues that affect these low energy proton beams: (1) the energy of protons after the absorber was broadened and the spectrum followed a Gaussian distribution (Full Width at Half Maximum > 0.5 MeV); (2) the energy loss of protons in the Li target leads to low energy neutrons (approximately 180 keV) production; (3) these neutrons could be produced via ⁷Li(p, n)⁷Be^★ reaction (E_th = 2.373 MeV).

Fig. 6.Full-screen View

The final neutron energy spectrum.

According to the experimental results, the neutron flux reached 1.4110³ sr^-1 μA^-1 at a distance of 58 cm from the lithium target. Moreover, the neutron source strength reached 4.7410⁶ sr^-1 μA^-1, which is lower than the calculated results by an order of magnitude. The difference is mainly attributed to the moderation and absorption of the cooling water, scattering, and absorption of the collimator.

With the experimental neutron flux, the same moderators were chosen to study the moderation effects. The Geant4 simulation results are shown in Fig. 7. The resonance neutron yield values in the 12.6-15.8 keV range, and could reach 4.76×10⁴ sr^-1 μA^-1, 1.23×10⁵ sr^-1 μA^-1, 5.23×10⁴ sr^-1 μA^-1 and 4.76×10⁴ sr^-1 μA^-1 respectively. It is possible to obtain more 13.5 keV neutrons with deuteron-containing materials compared to those that contain carbon and hydrogen. However, none of these materials were capable of fully moderating the maximum number of fast neutrons (hundreds of keV) down to the desired resonance energy range, while increasing the resonance neutron yield. A high incident proton energy is adverse to neutron production and moderation in terms of increasing the resonance neutron yield near the threshold of ⁷Li(p, n)⁷Be reaction. This is an essential requirement for further SNCT neutron source studies.

Fig. 7.Full-screen View

(Color online) The neutron energy spectra after (a) H₂O, (b) D₂O, (c) graphite, and (d) CH₂ moderation of different thicknesses with experimental neutron flux.