Conceptual Shield Design for Boron Neutron Capture Therapy Facility Using Monte Carlo N-Particle Extended Simulator with Kartini Research Reactor as Neutron Source

1Department of Physics, Faculty of Mathematics and Natural Sciences, Universitas Negeri Yogyakarta, Yogyakarta 55281, Indonesia 2Department of Nuclear Engineering and Engineering Physics, Faculty of Engineering, Universitas Gadjah Mada, Yogyakarta 55281, Indonesia 3Center of Accelerator Science and Technology, National Nuclear Energy Agency, Yogyakarta 55281, Indonesia *Corresponding author: afifahhnts@gmail.com


INTRODUCTION
The World Health Organization (WHO) reported that 8.8 million people died due to cancer in 2015.Cancer treatments being used nowadays are chemotherapy, radiotherapy, and surgery (Benjamin 2014).The ineffectiveness of current cancer treatment methods has motivated some researchers to upgrade to advanced technology in cancer treatment (WHO 2017).
Boron Neutron Capture Therapy (BNCT) is a form of therapy that destroys specific cancer cells.BNCT uses nonradioactive nuclide 10 B enriched compounds concentrated in cancer tissues that must be put into a patient's body.The cancer tissues are then irradiated by a low energy source from neutron beam which induces a nuclear reaction of 10 B(n,α) 7 Li.The length path of the particles is induced at a range of 4.5 µm to 10 µm, approximately the diameter of a human cell.Thus, the probability of it reaching healthy cells is small.Japan, America, Finland, Argentina, and Taiwan are the countries which have used modified reactors for BNCT (Sauerwein and Moss 2009).Shaaban and Albarhoum (2015) previously conducted research about improving the neutron flux for BNCT in the Syrian Miniature Neutron Source Reactor with the power of 30 kW.It produces 2.53 × 10 15 n/s, along with an epithermal flux of 2.83 × 10 8 n/cm 2 s.Mokhtari et al. (2017) have conducted research on LPMR (low power medical reactor) using fuel U 3 Si 2 -Al.Their research encompassed in-creasing the epithermal flux by optimization, resulting in an epithermal flux of 1.01 × 10 9 n/cm 2 s.Some Indonesian researchers have been exploring the possibility of BNCT for cancer treatment.BNCT research by in vitro and in vivo method has been done in Kartini Reactor, Yogyakarta.Kartini Reactor is TRIGA MARK-II reactor type with a thermal power of 100 kW, to which it is planned that BNCT research facilities such as collimator and shield will be added.The collimator research was conducted by Warfi et al. (2016) in which optimization was by the design of Fauziah et al. (2015).The collimator is designed for the thermal column of the Kartini Reactor; it yields an epithermal neutron flux of 1.13 × 10 9 n/cm 2 s.
A BNCT research facility requires the shield to protect people from radiation.Radiation should be minimized in order to meet the requirements set up by BAPETEN (the Nuclear Energy Regulatory Agency of Indonesia) for safety reasons.The dose limit is based on BAPETEN regulations which explain that the dose limit for a radiation worker is 20 mSv/y.

Experimental method
The research is simulated using the Monte Carlo N-Particle Extended (MCNPX).MCNP is a computer code used to simulate the probability of neutron, photon, and electron as well as their interaction, including fission reaction, scattering, and absorption (Xoubi 2016).
The research is started by collecting data of BNCT facilities in Kartini Reactor.BATAN has designed the shield which is formed by paraffin blocks covered by aluminium with a thickness of 3 mm.The dimension of each paraffin block with its aluminium cover is ( 40

Source modelling
Kartini Reactor with the power of 100 kW is used for basic counting of multiplication factor.The number of fission reactions per second in Kartini Reactor is approximately given by Equation 3: The result of Equation 3 is used to calculate the multiplication factor which will be input to the program.The counting multiplication factor is given in Equation 4. For average 2.42 per fission rate, the normalization factor of neutron is as follows (Equation 5).Φ epi (n/cm 2 s) 1.13 × 10 9 > 1.00 × 10 9 Ḋy /Φ epi (Gy.cm 2 /n) 1.45 × 10 -13 < 2.0 × 10 -13 Ḋ f /Φ epi (Gy.cm 2 /n) 1.76 × 10 -13 < 2.0 × 10 -13 Φ th /Φ epi 0.0108 < 0.05 The modelling of Kartini Reactor and its collimator has been done by Warfi et al. (2016) (Figure 1).The collimator design is chosen because its output has met IAEA criteria.
The output of collimator by Warfi et al. ( 2016) is shown in Table 1.

Shield modelling
The components of the shielding are paraffin and aluminium with the dimension of (40 × 40 × 24) cm.The blocks formed were simulated by MCNPX and are shown in Figure 2.

RESULTS AND DISCUSSION
The dose rate is calculated in the soft tissue with the thickness of 5 cm.The components of the soft tissue are published by ICRP (1999).The tally is F4 (flux average over a cell).A kerma coefficient is used to convert flux to become the dose with the unit of Gy/s.Then Gy/s is converted to Sv/s by multiplying it with the weight factor for each radiation type.Soft tissue modeled in front of the collimator was used to calculate an initial dose of 45118.80 µSv/h.When calculating the dose rate over the shield, we modelled soft tissue in the surface.We tested the materials to compare each one's ability to reduce radiation, showing the dose rate when varying the thickness.The components were paraffin, aluminium, and lead.The results of the material testing is shown in Figure 3.
Paraffin is effective at absorbing the neutron because it contains hydrogen, but paraffin is not effective at absorbing gamma.Aygün and Budak (2012) found that the increase of  2017) concluded on their research that the ability of concrete shield will increase when hydrogen is added.The attention should be focused on the shield for real conditions.However, since paraffin melts due to the heat, it is covered by aluminum to keep the structure of the paraffin.Aluminium has a high corrosive resistance because it has an oxide layer on its surface (Pokhmurskii et al. 2011).
The output of MCNPX yields that the gamma dose rate increases with the increase of aluminium thickness.Neutrons that strike aluminium produce gamma because of inelastic scattering that makes the nuclei of the material exited (Padalino et al. 1999).Lead has a high atomic number and high density making it effective in reducing gamma radiation (Zeb et al. 2010).
It is found that dose rate in some soft tissue model locations still have exceeded the dose limit of 10.42 µSv/h.This indicated that the high dose rate is coming from the gamma contamination.The interaction of neutrons with aluminium during the inelastic scattering process also contributes to producing gamma radiation.Then radiation shielding has been optimized by adding several lead layers in the outer surface of the shield (Figures 4 and 5).The  measured dose rates before and after being optimized are shown in Tables 2 to 5. The largest dose rate is found in the front of shield because it is exactly in the front of the radiation source.Besides, the front side of shield is thinner than the others.From the tables are shown that dose rate has been reduced after optimization with lead with certain thickness.Some areas which are still above 10.42 µSv/h can be solved by distance and time principle, since the flux is inversely proportional to the square of the distance, or 1/r 2 (Ahmed 2007).In the design, aluminium keeps the structure of the paraffin, and lead is useful to reduce gamma radiation which is produced by the inelastic scattering of the neutron and aluminium reaction.With consideration of economic aspect, the maximum thickness of lead in the front side is 12 cm, the right side is 3.5 cm, the left side is 3.5 cm, and the top side is 2 cm.

CONCLUSIONS
The shield has been optimized with lead and the results are obtained with the largest dose rate of 57.60 µSv/h.Alu- minium is applied in the design to keep the structure of the paraffin, while lead reduces gamma radiation which is produced by the inelastic scattering of the neutron and aluminium reaction.
× 40 × 24) cm.The minimum effective dose used in this research is based on the recommendation of BAPETEN, and calculated using Equation 1: Working time: 8 hours/day × 5 days/week × 4 weeks/month × 12 months/year

FIGURE 3 .
FIGURE 3. Dose rate for different thicknesses of materials suggested as shielding radiation.

TABLE 2 .
Shield dose rate in the left side.

TABLE 3 .
Shield dose rate in the right side.

TABLE 4 .
Shield dose rate in the front side.

TABLE 5 .
Shield dose rate in the top side.