The Radioactivity Estimation of The Irradiated 13 MeV Cyclotron’s Concrete Shield

Received: 25 February 2020 Received in revised form: 17 March 2020 Accepted: 18 March 2020 The Center for Accelerator Science and Technology (PSTA) planned to install K500 concrete shield in its 13 MeV cyclotron facility (DECY13). However, fast neutrons that are generated by this cyclotron could activate materials of the concrete. It may harm the radiation workers. In this work, we conducted simulations using ORIGEN2 and PHITS computer code to estimate the formed radioactivity and the neutron flux distribution in the DECY-13 cyclotron's concrete shield. Based on the simulation, the induced radioactivity is 2.3478 × 10 Bq, while its gamma dose rate is 22.09 μSv/mh. The most contributed isotopes are Th-233, Ho-166, Al-28, Mn-56 and Si-31. This dose is quite high. Neutron fluxes in the rear of the simulated concrete shield are also still prominent. Accordingly, it is necessary to attach neutron shielding materials which do not generate high-intensity gamma-ray. The formed radioactivity is high; but it appears from the short half-life isotopes such as Th-233, Ho-166, Al-28, Mn-56 and Si-31. Its activity will diminish quickly after the cyclotron is off. Hence, it will be safe for radiation workers.


INTRODUCTION *
In recent years, the Centre for Accelerator Science and Technology (PSTA) has developed 13 MeV cyclotron (DECY-13) for the fluorine-18 production [1][2][3][4][5]. This cyclotron generates secondary gamma and neutron, referring to the previous analysis [6]. Therefore, the radiation shield with K500 concrete quality will be installed to reduce gamma exposure. It is the same material as previously used on its target module, for the experiment of proton beam energy measurements. In other hands, neutron might activate materials, including concrete. Radiation caused by the activated concrete material can endanger the radiation workers.
The other researchers have examined the activation of concrete in the nuclear reactors [7] and accelerators [8]. However, materials of concrete may differ depending on their sand, cement, and aggregate. In this case, the sand applied for concrete shields on the DECY-13 is Progo sand. The study elsewhere mentions that Progo sand contains thorium and the rare earth metal [9]. These elements are uncommon in the concrete. Furthermore, the previous analysis informs that DECY-13 produces mostly fast neutron [6]. It is distinct from preceding examinations using thermal neutrons [10]. Hence, the objective of this paper is to find out the formed radioactivity and the neutron flux distribution in the DECY-13's concrete shield.
The investigations elsewhere employed the ORIGEN2 [11,12] and PHITS [13,14] computer code in the various nuclear field. Both software is known to be reliable [15][16][17]. Accordingly, we used ORIGEN2 and PHITS to estimate the radioactivity

METHODOLOGY
DECY-13 implements the concrete shield with a density of 2.3 g/cm3. We conducted the Energy Dispersive X-Ray Fluorescence (EDXRF) analysis to acquired its elemental data.
ORIGEN2 performs calculations based on 1 gram of the concrete sample, for the simplicity in obtaining the specific activity. It is a commonly used unit in the radiation protection field. It describes the radioactivity concentration (Bq/g) of radionuclide in the material.
In ORIGEN2 calculation, we assumed the concrete sample as a slab with 1 cm length, since it has a density of 2.3 g/cm 3 , then its cross-section is 0.434783 cm 2 . These calculations utilized a neutron source of 1.91 × 10 11 neutrons/s, referring to the previous analysis [6]. Duration of irradiation was 1 hour. The radioactivity value procured from this algorithm is conservative because it ignores its distribution. PHITS also take an identical concrete sample for the simulation with monodirectional 1.5 MeV neutron source. Additionally, we varied the neutron energy and the duration of irradiation to find out its effect on the formed radioactivity.
PHITS applies the isotropic neutron source to examine radioactivity distribution. In these simulations, PHITS models the concrete as a large block with dimensions of 120 cm x 120 cm x 120 cm. There are small cubes with a size of 15 cm x 15 cm x 15 cm that arranged inside the block, regarding their axial (x) and radial (y) position as in Fig. 1. It is intended to estimate the neutron flux and radioactivity on the specific region. PHITS execute the [T-Track] and [T-DCHAIN] tally to obtain such value [18].

RESULTS AND DISCUSSION
The Energy Dispersive X-Ray Fluorescence (EDXRF) results confirm that the DECY-13's concrete shield contains thorium in the form of ThO 2 and the rare earth metal oxides such as CeO 2 , Er 2 O 3 , Table 1 shows its elemental content. ORIGEN2 uses all of these data, while PHITS dismiss the erbium (Er) and promethium (Pm). PHITS also replaces the natural lead with Pb-208 isotope.   In the fast neutron spectrum, the specific radioactivity of the irradiated concrete with an hour of irradiation based on PHITS simulation is lower than ORIGEN2 calculation. ORIGEN2 uses a conservative approach, in contrast to PHITS which is more realistic. Based on PHITS, specific radioactivity significantly increases, along with the decrease in neutron source energy, as in Fig. 3. Its trend is related to the probability of neutron interaction that is larger in the low energy spectrum, compared to the higher energy. At the source energy of 10 eV, it becomes equal to the ORIGEN2 result. It is in consequence of the thermal neutrons used by default in the ORIGEN2 calculation. Based on the PHITS simulation, the irradiation duration raises the specific radioactivity of the concrete shield, as displayed in Fig. 4. Its trend is logarithmic. This trend is related to the neutron fluence buildup and the decay of formed radioactive materials. The irradiation duration increases the neutron fluence and the probability of neutron interaction. Hence, the radioactivity rises. But, the formed radioisotopes are decaying along with time. Therefore, it compensates the increasing caused by neutron fluence and makes the trend logarithmic. Nonetheless, it does not remarkably rise, compared to the effect of neutron energy. The ORIGEN2 and PHITS calculation on the sample, cannot describe the condition of the entire concrete shield, because the flux and energy of the neutron are not constant along its journey. The concrete might moderate or even absorb the neutron beam. It is necessary to find out the distribution of the neutron flux and the formed radioactivity on the entire concrete shield. The solution is PHITS employ the full concrete block at a distance of 2 cm from the isotropic neutron source. Inside it, there are small cubes arranged on the x (axial) and y (radial) axis, as in Fig. 1. The coloured line represents neutron flux per source current. There are gradations of colour that show the flux differences in each region.
Total radioactivity that formed due to an hour of irradiation, in the whole concrete shield, is 2.3478 × 10 9 Bq. The prominent radioisotopes are Th-233, Ho-166, Al-28, Mn-56 and Si-31. Since Al-28 and Mn-56 emit high energetic gamma, thus both dominate the gamma dose rate of the irradiated concrete. The values are 11.25 and 9.26 µSv/m 2 h respectively from a total of 22.09 µSv/m 2 h. This gamma dose is quite high for the radiation workers. Figure 5 provides a detailed radioisotopes chart.
Ho-166 and Mn-56 have a relatively long halflive compared to Th-233. After 2 hours of irradiation, both dominate the formed radioactivity, replacing Th-233. Mn-56 also supersede Al-28 in the gamma dose rate with 16.34 µSv/m 2 h. It is because the formation rate of the short half-life radioisotopes cannot compensate for its decay rate compared to the relatively long half-life ones. However, neutron flux in the axial position is generally higher than radial ones. It was caused by neutron source distance to the concrete, which is further in the radial position. Neutron fluxes in the rear of the simulated concrete shield are substantial, ranging in order of 10 5 to 10 6 neutrons/cm 2 s. Therefore, it is needed to apply neutron shielding materials. The induced radioactivity increases from the axial position of 15 to 45 cm. Afterwards, it starts to attenuate, as shown in Fig. 7. But in the radial ones, it decreases from the beginning. The peak of radioactivity in the axial may be related to the nuclear cross-section, which is influenced by the incoming neutron energy. The cross-section represents the probability of interactions that might occur. The probability of neutron interaction is greater in the low energy neutron, compared to the fast one.
Neutrons are decelerating and lost its energy while passing through the concrete. If the neutron energy is decreasing, then its cross-section and the probability of absorption interaction are increasing. As a result, concrete materials absorb more neutrons. Hence, its radioactivity increases. This trend continues until neutron flux is not sufficient, thus, the radioactivity decrease. Based on the calculation using ORIGEN2 and PHITS, the formed radioactivity in the DECY-13's concrete shield caused by the fast neutron irradiation is quite high. These radionuclides also generate high gamma dose. The neutron absorber materials with low gamma emission are required. Thus, the gamma dose received by the radiation workers does not exceed the threshold.

CONCLUSION
The specific activity of DECY-13's concrete shield, based on a conservative approach with ORIGEN2 and PHITS is 9.66 × 10 7 Bq/g and 4.35 × 10 6 Bq/s respectively. Based on PHITS simulation, total radioactivity in the entire concrete shield is 2.3478 × 10 9 Bq, while its gamma dose rate is 22.09 µSv/m 2 h. It is quite high for radiation workers. Neutron fluxes in the end side of the simulated concrete shield are also noticeable. Accordingly, neutron absorber materials which do not produce high-intensity gamma-ray is needed. Although the formed activity and gamma dose rate is quite high, it comes from the relatively short half-life isotopes such as Th-233, Ho-166, Al-28, Mn-56 and Si-31. Hence, its activity will decrease along with non-operational time.