Two devices were installed for the quasimonoenergy neutron activation cross-section measurement experiment using the cyclotron facility at ARTI. One device measures proton beam energy and the other generates a neutron. The energy distribution and neutron flux need to be determined to measure the activation cross-section. The neutron energy distribution is obtained by Monte Carlo simulation using Monte Carlo N-Particle (MCNP), which requires the proton beam energy data as input. The neutron flux is estimated by measuring the accumulated current through the 50 μm thick molybdenum foil placed in front of the graphite stopper of the neutron generator. The neutron generator consists of a stainless steel cross pipe (STS304), a lithium or beryllium target, and a graphite stopper. The diameter, length, and thickness of the cross pipe are 10 cm, 10 cm, and 2 mm, respectively. The lithium target (6 cm×6 cm×2 mm) and graphite stopper (6 cm×6 cm×6 mm) are placed 2.2 cm apart in the center of the cross pipe. Figs. 1 and 2 show the experimental devices for measuring the neutron activation cross-sections.

The neutrons produced by this neutron generator will be irradiated onto samples, including gold, cobalt, and niobium, to measure the neutron activation cross-sections. The cyclotron has a beam energy of 15–30 MeV. This experiment involves the irradiation of a 30 MeV, 1 μA proton beam into the neutron generator.

Monte Carlo simulations are performed to investigate the effects of neutrons and secondary particles generated during the experiment on the interior and surroundings of the facility. The MCNP6.1 code [1] was used with the KNE80 neutron library [2] generated by Korea Atomic Energy Research Institute based on ENDF/B-VIII.0 for radiation shielding and activation analyses in this study. MCPLIB84 and EL03 in the MCNP code package were utilized as photon and electron data libraries, respectively. Fig. 3 shows the facility modeled for the simulation, with the target room on the right where the neutron activation is measured. The neutron generator produces neutrons by bombarding Li-7 or Be-9 targets with a proton beam, with the latter yielding over four times the number of neutrons compared to the former. Therefore, the simulation results using the Be-9 target were considered for the shielding analysis.

We assumed irradiation of a Be-9 target with a 30 MeV, 1 μA proton beam for 1 hour for activation analysis, thereby generating neutrons that activate the facility. The International Commission on Radiological Protection 74 dose conversion factors [3] for neutrons and photons incident on anterior–posterior geometry of an adult anthropomorphic computational model were used in the dose rate calculations using the MCNP code. This work involves geometry splitting and Russian roulette as a variance reduction technique, which is performed by adjusting the cell importance. The importance of a cell, IMP, within the beam interrogation room was set at 1, adjacent concrete walls to the tally at 3, and the tallies’ IMP at 9, anticipating uniform neutron irradiation in all directions.

The rigorous-two-step method (R2S) [4] was used, coupling the Monte Carlo N-Particle Extended (MCNPX) 2.7.0 [5] with the SP-FISPACT-2010 code [6, 7], patched with a function to simulate spallation reactions [8], to assess the activation of devices, air, and concrete walls within the facility. Following the R2S analysis method [4], we performed neutron transport calculations using MCNP to obtain neutron spectra and total flux. Subsequently, activation inventory estimations were performed using FISPACT based on these results. Afterward, the MCNP code was used to compute the decay gamma transport of the activated facility. The 175-group neutron energy group structure was used in FISPACT calculations. For activation analysis at energies above 20 MeV, where neutron activation data is unavailable, HISTP files generated from physics model calculations by the MCNP code were used.

We first calculated the spectrum of neutrons generated by the collision of a 30 MeV proton beam with a Be-9 target and graphite stopper to improve the efficiency of the simulations. We used the MCNP code to obtain a neutron spectrum divided into 80 logarithmically spaced energy groups ranging from 1 eV to 40 MeV. This spectrum, depicted in Fig. 4, was used to conduct subsequent radiation shielding and activation analyses. In this work, the isotropic neutron source was assumed instead of considering angular distribution because the low beam current and the sufficient thickness of the shielding wall enable the conclusion that considering angles would not significantly affect the shielding analyses. Furthermore, the target, moderator, reflector, and shielding system, which is depicted as a green square in Fig. 3B, is expected to shield higher neutron fluxes in the forward direction, thereby further reducing the effect of the forwardness of the neutron source.

Public dose limits have to be applied for radiation shielding analysis because the upper floors of the cyclotron facility and the exterior of the building are public areas. The dose rate should be less than 0.5 μSv·hr⁻¹ in public areas, assuming 2,000 hours of operation per year and considering the public dose limit of 1 mSv·yr⁻¹. Conversely, the outside of the duct and the vicinity of the entrance, where there may be significant radiation leakage, are radiation-controlled areas. Similarly, the dose rate in radiation radiation-controlled areas should be less than 10 μSv·hr⁻¹, considering the occupational dose limit. Figs. 5 and 6 show the neutron and photon dose rate distributions intraoperatively in the target room. Additionally, the dose rate is high inside the target room during operation, but decreases as the radiation passes through the concrete walls and maze-like passage, thereby reducing the dose rate outside the target room.

We calculated the dose rates at the key locations around the target room shown in Fig. 7. In particular, the dose rate calculation locations around the entrance, where radiation leakage is expected, are subdivided (Fig. 8). Table 1 summarizes the calculated dose rates for the locations shown in Figs. 7 and 8. The dose rates at all locations are less than 0.5 μSv·hr⁻¹, which is below the occupational dose limit as well as the public dose limit. This revealed that the Monte Carlo simulation uncertainties in areas where the dose rates are relatively high (>0.1 μSv·hr⁻¹) are less than 5%. The uncertainty ranges between 10% and 30% for regions with dose rates of <0.1 μSv·hr⁻¹, including the outside duct. Areas expected to have relatively low dose rates (between 19 and 21) demonstrate calculated dose rates of <10⁻³ μSv·hr⁻¹, with uncertainties exceeding 30% (Fig. 8, areas).

Neutrons produced by the neutron generator activate inside the facility, including devices, air, and concrete walls. Therefore, activation analysis was conducted by investigating the types of radioisotopes produced and their radio activities caused by the 1 hour neutron beam irradiation (BI). Fig. 9 illustrates the activities of the major nuclides caused by the irradiation. All isotopes shown in Fig. 9, being closer to the source, originated from the neutron generator, including the Be-9 target. Mn-56 and Co-58 decay rapidly over time due to their relatively short lifetimes, whereas other isotopes show little variation.

The radionuclides expected to be generated due to the activation in the ordinary concrete composition include Na-22, Na-24, and Mn-54, among others. However, the gamma radiation caused by concrete structure activation is mainly due to radioactive nuclides, such as Eu-152, Eu-154, and Co-60, originating from impurities present within concrete. Therefore, several impurities were selected from the reference [9–11] and incorporated into the concrete composition. Table 2 summarizes their types and quantities. Fig. 10 shows the activities of the isotopes listed in Table 2 overtime on the concrete wall closest to the neutron generator.

Moreover, we conducted a study to evaluate the radiation dose received by workers from the activated facility. The dose by activated materials was evaluated at the location indicated by the red dot in Fig. 11, assuming the worker is positioned 10 cm away from the device. Radiation doses received from the facility were calculated using the R2S method with uncertainties of approximately 0.1%. Whereas, the gamma dose rates obtained from FISPACT were used for assessment instead of the R2S method due to the widespread presence of air and concrete making it difficult to specify the location. Therefore, the actual dose rates received by workers are lower than the results. The BI time varies depending on the sample. This study calculated dose rates of the facility, air, and concrete activated by neutrons for 30 minuites and 1 hour of BI time (Fig. 12, Tables 3 and 4).

The dose rate will not decrease sufficiently until 5 hours after turning off the beam if the BI time is 1 hour, and it still requires 3 hours even if the BI time is 30 minutes. However, as mentioned earlier, the actual dose rate is expected to be lower than the calculated results, and access will be allowed if the values measured by the dosimeter are lower than the permissible dose limit. We are considering the option of quickly collecting only the sample within a short timeframe while considering the annual dose limit in case of measuring the activation cross-section of short-lived radioisotopes.