Measurement of thermal neutron fluence distribution ... - Springer Link

1 downloads 0 Views 307KB Size Report
May 16, 2009 - distribution of induced radioactivity, but also optimizing the shield design for radiation safety in preparation for the decommissioning process.
Radiol Phys Technol (2009) 2:159–165 DOI 10.1007/s12194-009-0060-7

Measurement of thermal neutron fluence distribution with use of 23Na radioactivation around a medical compact cyclotron Toshioh Fujibuchi Æ Ichiro Yamaguchi Æ Tetsuharu Kasahara Æ Takashi Iimori Æ Yoshitada Masuda Æ Ken-ichi Kimura Æ Hiroshi Watanabe Æ Tomonori Isobe Æ Takeji Sakae

Received: 29 September 2008 / Revised: 12 April 2009 / Accepted: 15 April 2009 / Published online: 16 May 2009 Ó Japanese Society of Radiological Technology and Japan Society of Medical Physics 2009

Abstract A medical compact cyclotron produces about 1015 neutrons per day along with 100 GBq of 18F. Therefore, it is important to establish radiation safety guidelines on residual radioactivity for routine operation, maintenance work, and decommissioning. Thus, we developed a simple method for measuring the thermal neutrons in a cyclotron room. In order to verify the feasibility of our proposed method, we measured the thermal neutron distribution around a cyclotron by using the activation of 23Na in salt. We installed 78 salt dosimeters in the cyclotron room with a 50 cm mesh. The photopeak of 24Na was measured, and the neutron flux distribution was estimated. Monitoring the

T. Fujibuchi (&) Department of Radiological Sciences, School of Health Sciences, Ibaraki Prefectural University, Ami, Ami-machi, Inashiki-gun, Ibaraki 300-0394, Japan e-mail: [email protected] T. Fujibuchi  T. Isobe  T. Sakae Graduate School of Comprehensive Human Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8575, Japan I. Yamaguchi National Institute of Public Health, 2-3-6 Minami, Wako, Saitama 351-0197, Japan T. Kasahara  T. Iimori  Y. Masuda Department of Radiology, Chiba University Hospital, 1-8-1 Inohana, Chuo-ku, Chiba 260-8687, Japan K. Kimura Fujita Corporation, 2025-1, Ono, Atsugi, Kanagawa 243-0125, Japan H. Watanabe Yokohama Rousai Hospital, 3211, Kodzukue-cho, Kohoku-ku, Yokohama, Kanagawa 222-0036, Japan

neutron flux distribution in a cyclotron room appears to be useful for not only obtaining an accurate estimate of the distribution of induced radioactivity, but also optimizing the shield design for radiation safety in preparation for the decommissioning process. Keywords Medical compact cyclotron  Positron emission tomography  Radioactivation  Neutron  Decommissioning  Sodium chloride

1 Introduction Recently, the number of medical institutions with an installed medical compact cyclotron has been increasing owing to the prevalence of the positron emission tomography (PET) examination [1]. More than 100 cyclotrons (124 units by 31 March 2007) [2] have been installed in Japan. A cyclotron irradiates the target box with accelerated protons. Through a (p, n) reaction, positron emission radionuclides (11C, 13N, 15O, and 18F) with a short half-life (from 2 to 120 min) are produced. Relatively large numbers of neutrons are produced by these nuclear reactions in the target box. Radioactivity is generated in the cyclotron itself. The concrete also becomes radioactive, mainly due to the capture of thermal neutrons. The handling of this residual radioactivity has become an issue in radiation safety [3]. Therefore, evaluating the residual radioactivity in a cyclotron, in indoor air, and in concrete has become an essential task in relation to protecting indoor workers from radiation and decommissioning countermeasures [4–7]. However, it is difficult for a hospital to measure the distribution of residual radioactivity in a medical cyclotron as a routine radiation safety measure. Thus, we have developed a simple, cost-saving method that is suitable for

160

T. Fujibuchi et al.

use in a medical institution with a cyclotron. We measured the distribution of neutron flux in the cyclotron room using this method in order to assess its feasibility.

2 Materials and methods All measurements were carried out at the cyclotron (CYPRIS HM-18; Sumitomo Heavy Industries, Ltd., Tokyo, Japan) installed at the Chiba University Hospital. This facility produces only 18F by an [18O] H2O (p, n) 18F reaction. Since April 2003, the cyclotron has been operated for 60 min per day about four times a week at a mean beam current of 20 lA and a proton beam acceleration energy of 18 MeV. 2.1 Thermal neutron measurement using the radioactivation of sodium chloride Neutron flux is often determined by a radioactivation method that measures the radioactivity induced in a sample foil by an (n, c) reaction. Typically, gold foils and a germanium semiconductor detector are used for the measurement [8]. However, these materials and this equipment are rarely available at hospitals. Thus, we developed a method of thermal neutron measurement using the radioactivation of 23Na contained in sodium chloride (NaCl) [9] and a NaI detector that is routinely employed at hospitals for quality assurance testing of radiopharmaceuticals. We used an aluminum-vial casing (diameter 15 mm, height 7 mm, and thickness 0.1 mm) filled with NaCl (2.5 g, 99.5% purity, Nakarai Tesque, Inc., Kyoto, Japan) and covered by a sealing film (Parafilm M: American National Can Company, USA) as our sample (Fig. 1a). To verify the measurement precision associated with the use of NaCl, we prepared gold foils (diameter 9.8 mm, thickness 0.05 mm, and weight 68 mg) (Fig. 1b). The basic characteristics of the radioactivation of 197Au and 23Na are shown in Table 1 [10, 11]. The cross-sections of neutron capture for each nuclide are shown in Fig. 2. We performed the following measurements to verify the basic features of this method. 2.1.1 Confirmation of the effect of self-shielding of the salt for neutron activation To define whether correction for neutron self-shielding of the sample is necessary, we arranged five NaCl samples of different thicknesses, 1.4, 2.8, 4.2, 5.6, and 7.0 mm, the weights of which were 0.5, 1.0, 1.5, 2.0, and 2.5 g, respectively, closely around point E in Fig. 3 for simultaneous neutron measurement. We examined the relationship

Fig. 1 Samples used for the activation method: a sodium chloride and b gold foil. Sodium chloride was packed inside an aluminum cap. The inside diameter of the cap was 15 mm, its height was 9 mm, and the weight of sodium chloride was 2.0 g. The gold foil was 9.8 mm in diameter, 0.05 mm in thickness, and 68 mg in weight

between the thickness of each sample and the total radiation counts measured by the NaI spectrometer. 2.1.2 Comparison with a gold foil sample The samples were placed at nine points around the target on the floor during cyclotron operation. These points are labeled A to I in Fig. 2. We waited for 6 h before collecting the irradiated samples to allow the residual radioactivity of the cyclotron room to drop to a safe level. These samples were measured with a NaI scintillation-type detector (802-292: Canberra Industries, Inc., Meriden, CT, USA) as well as a multichannel pulse-height analyzer (GENIE 2000: Canberra Industries, Inc.) and a NaI spectrometer, which were employed to quality test the radiopharmaceuticals used for PET. To measure the samples, we attached the uncovered top to the surface of a NaI detector. The influence of the background was reduced by enclosing the sample in a 5cm-thick lead block. During the measurements, the rated voltage of the NaI spectrometer was set to 1,000 V, and the number of measurement channels was 2,048. The energy window of each channel was 0.908 keV, as determined by calibration. The measurement time for each sample was three minutes. Collection windows were set up at all of the energy absorption peaks (24Na: 1369 keV and 198Au: 412 keV). To convert the radiation counts per minute for the sample measured with the NaI spectrometer into the thermal neutron flux, we measured a standard radioactive source. In view of the decay during the measurements, we

Measurement of thermal neutron fluence distribution Table 1 Activation elements (Au, Na) and basic nuclear data [12, 13]

161

Activation foils

Nuclear reaction

197

197

23

23

Gamma-ray energy (keV)

Au (n, c)

Au

Na (n, c)

Na

198

Au

24

Na

412 1,369

Half-life

Thermal neutron cross-section (barn)

2.695 days

98.5 ± 0.4

14.96 h

0.53

2,754

Fig. 2 Neutron capture cross-sections for

23

Na and

197

Au [12, 13]

where S is the total radiation count during the measurement, C is the count rate at the beginning of this measurement, k is the disintegration constant of the activated radionuclide, and t is the measurement time. t = 0 corresponds to the time at which measurements were started. The Chiba University Hospital does not have any standard sources. Thus, the four samples were measured with an HP-Ge detector (GEM20: ORTEC, IL, USA) at the National Institute of Public Health (NIPH). This system was calibrated with standard radioactive gamma volume sources (MX033U8: Japan Radioisotope Association). These standard sources are composed of 109Cd, 57Co, 139 Ce, 51Cr, 85Sr, 137Cs, 54Mn, 88Y, and 60Co. These standard sources take the form of a type U8 container (MX033U8). Because this is roughly the same shape as that of the present samples, we corrected only for the sample height. The correction method obtained the count efficiency function at each energy using a standard sample of five different heights, and interpolated via the height of the sample. As for the self-shielding of Au, we calculated it with EGS [10], and corrected for it. We estimated the detection efficiency with a NaI spectrometer by comparing the count rate to that of the Ge detector for 412 keV (198Au) and 1,369 keV (24Na) gamma rays. The count rate obtained by Eq. 1 was converted into the radioactivity using the following equation: C ¼ AeIc expðkTC Þ:

Fig. 3 Layout of the cyclotron room and the sampling points. The samples of sodium chloride and gold foil were installed at points A–I. The neutron and photon distributions at the surface of the floor were measured using salt dosimeters and glass dosimeters, installed using 50 cm mesh

used the following formula to convert the total radiation counts into the counting rate: S¼

Zt 0

C exp ðktÞdt;

ð1Þ

ð2Þ

Here, A is the induced radioactivity, e is the detection efficiency of the NaI detector, Ic is the branching ratio of the gamma rays, and Tc denotes the time from the end of irradiation to gamma-ray measurement. To obtain the saturated activity, we used the equation w AS ¼ r/ NA hf1  expðkTi Þg; ð3Þ M where As is the saturated activity, r is the nuclear reaction cross-section, / is the neutron flux, w is the mass of the target nuclide, M is the atomic weight of the target nuclide, NA is Avogadro’s number, h is the isotopic abundance of the target nuclide, and Ti denotes the exposure time. To convert from radioactivity to thermal neutron flux, we used the equation [11] /th ¼ Fth

As  FCd As ðCdÞ ; th Þ NT rð1  FCd TCd

ð4Þ

162

where /th is the thermal neutron flux, Fth is a factor that corrects for self-shielding of thermal neutrons in the foil and neutron perturbation, As is the radioactivity of the nuclide activated by thermal and the epithermal neutrons, As (Cd) is the radioactivity of the activated nuclide resulting from epithermal neutrons alone, due to the effect of cadmium, FCd is a factor that corrects for the epithermal neutrons that are absorbed by the cadmium filter, NT is the number of atoms of the target nuclide, Tth Cd is the probability that the thermal neutrons penetrate the Cd cover, and r is the nuclear reaction cross-section. In this study, we used a Cd cover 1.0 mm thick. We also used 98.5 barns as the cross-section, a neutron energy of 0.025 eV [12], a value of 10-6 was adopted for Tth Cd, a value of 0.91 was adopted for Fth, and 1.21 was adopted for Fcd [8]. As for the uncertainty in the measurement, we treated the radioactive decay as a stochastic phenomenon. The total radiation count per 3 min, as measured by the NaI spectrometer, follows the Poisson distribution. Therefore, we assumed that the variance of the distribution of total radiation counts was equal to the mean of the distribution. Using the following formula, we obtained the uncertainty along with the randomness of the counting rate: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi N NBG r¼ ð5Þ þ 2 : t2 tBG N is the total radiation count for the sample, t is the preset time for measuring the sample, NBG is the total radiation count from the background, and tBG is the preset time for measuring the background. To obtain the factor that converts the count rate measured by the NaI spectrometer into the induced radioactivity, we measured each sample using a Ge detector and a NaI spectrometer, and then calculated the mean. We used the standard deviation of the conversion factor as the standard uncertainty of the measurement.

T. Fujibuchi et al.

3 Results 3.1 Basic characteristics of the radioactivation method using NaCl 3.1.1 Confirmation of the effect of self-shielding of the salt for neutron activation The relationship between the thickness and the count rate of 24Na is shown in Fig. 4. The count rates for each thickness are normalized to a sample of 1.4 mm thickness to offset the amount of atoms in each sample. The value decreased as the thickness increased. The value was 91% in the case of 7.0 mm thickness. A statistically significant positive correlation between the thickness of NaCl and the 24 Na count rate was observed; the coefficient of determination (R2) was 0.98 with Cd covered and 0.94 uncovered. 3.1.2 Comparison with the gold foil The radioactivities, Cd ratios, and thermal neutron fluxes of Au and NaCl are shown in Table 2. The relationship between the activities of Au and Na is shown in Fig. 5. Figure 6 shows the results obtained when we converted the activity into a thermal neutron flux. The results for the thermal neutron fluxes obtained from the measurements of the gold foil and the NaCl sample are compared in Fig. 6. The CVs (coefficients of variation, relative standard deviation), corresponding to the standard uncertainty of measuring total radiation counts using the NaI spectrometer, were 4.0% for NaCl and 0.2% for the gold foil. The CVs for the factors used to convert the total radiation count measured by the NaI spectrometer into the induced radioactivity were 13.5% for NaCl and 1.2% for

2.2 Measurement of the thermal neutron flux distribution in the cyclotron room To evaluate the thermal neutron flux distribution in the cyclotron room, we placed 78 NaCl samples on the floor with a 50 cm mesh and irradiated them with neutrons during cyclotron operation. After the cyclotron had been operated, the total energy absorption peaks of 24Na were measured with a NaI spectrometer, and the radioactivated distribution was determined. In addition to this neutron measurement, the X-ray and gamma-ray distributions during operating and nonoperating cyclotron conditions were measured with glass dosimeters (GD-352: Chiyoda Technol Corporation, Tokyo, Japan).

Fig. 4 Relationship between thickness of NaCl and the count rate at the same measurement point. The count rates for each thickness are normalized to a sample of 1.4 mm thickness to offset the amount of atoms in each sample. Error bars indicate standard uncertainties based on counting errors

Measurement of thermal neutron fluence distribution Table 2 Radioactivities, Cd ratios, and thermal neutron fluxes of Au and NaCl

Pointa

Au

See Fig. 3

NaCl

Activity (9106 Bq/g) Cd (-)

a

163

Cd ratio

Cd (?)

A

3.4

1.3

2.7

B

5.5

2.4

C D

7.3 11.6

2.9 5.1

E

11.2

F

6.9

G H

Thermal neutron flux (9105 cm-2 s-1)

Activity (9104 Bq/g) Cd (-)

Cd ratio

Cd (?)

Thermal neutron flux (9105 cm-2 s-1)

5.7

7.3

1.4

5.4

4.0

2.3

8.0

10.6

2.1

4.9

8.2

2.6 2.3

11.7 16.6

13.0 23.1

3.1 4.2

4.1 5.5

8.2 18.2

5.2

2.2

14.9

25.1

5.1

4.9

10.2

2.6

2.7

11.4

14.4

2.8

5.1

10.8

3.9

1.6

2.5

6.1

8.5

2.3

3.7

5.0

1.0

0.6

1.7

0.9

2.0

0.5

4.2

1.2

I

0.9

0.5

1.9

1.0

1.8

0.5

3.4

0.6

Mean

5.7

2.4

2.3

8.5

11.8

2.4

4.6

7.4

Fig. 5 Relationship between the radioactivities of each sample, with Cd covered or uncovered

Fig. 6 Relationship between the thermal neutron fluxes of 198Au and 24 Na at nine measurement positions

the gold foil. The combined standard uncertainty was 12 times higher for the gold foil than for NaCl. R2 was 0.88.

4 Discussion 4.1 Basic characteristics of the measurement

3.2 Neutron flux distribution in the cyclotron room Figure 7a and b show, respectively, the absorbed dose distribution for the photons and the radioactivity of 24Na during cyclotron operation, and Fig. 7c shows the distribution of the photons under nonoperating conditions. The maximum 1-cm-equivalent dose rates, as measured by an electronic pocket dosimeter (Dose3 and Dose3-N, Chiyoda Technol Co., Tokyo, Japan), were 176 mSv/h (see Fig. 7a), 146 mSv/h (see Fig. 7b), and 7.7 lSv/h (see Fig. 7c).

Compared to 197Au, the relative residual radioactivity of Na was only 6.5% for the duration of the irradiation and the time between the end of irradiation and the initial measurement in the present study. This low induced radioactivity was caused by the cross-sections and different half-lives of 197Au and 23Na. There was no tendency for bias in the neutron flux estimation between these methods; one sample’s uncertainty was assumed to be small compared to the other sample’s uncertainty. The uncertainty in 23

164

T. Fujibuchi et al.

Fig. 7 Thermal neutron flux and photon distribution in the cyclotron room. a Photon distribution during cyclotron operation, b activated 24Na distribution during cyclotron operation, and c photon distribution during cyclotron nonoperation. Maximum points were normalized to 1, and doses at the other points are indicated by their relative values

the factor that was used to convert total radiation counts measured by the NaI spectrometer into induced radioactivity was larger for NaCl than for the gold foil because of the lower count rate of the former and the shorter half-life of 24Na compared to that of 198Au. Carrying out the measurement process earlier would result in a smaller uncertainty due to decay. The mean Cd ratio was 4.6 for NaCl and 2.3 for Au in these measurements. This is due to the influence of fast neutrons in the cyclotron room and the shapes of the crosssection curves. 4.2 Advantage of using salt to measure neutrons in a hospital 197

Au is commonly used in radioactivation methods due to its relatively large thermal neutron cross-section, the relatively long half-life of radioactivated 198Au, and an adequate gamma-ray energy for measurement [8, 12–14]. However, gold foils are rather expensive, especially for multiple measurement points. However, the cross-section of thermal neutron capture for 23Na is 183 times smaller than that for 197Au (see Fig. 2). The resonance absorption in the energy range beyond that of thermal neutrons is smaller, and the thermal neutrons are captured comparatively specifically. Furthermore, because the decay constant of 24Na is 4.32 times larger than that of 198Au, when the saturation factor is taken into consideration, the induced radioactivity is 3.19 times larger for 24Na than for 198 Au for a 1 h and an 8 h cool-down. As a result, the radioactivation method that uses salt for thermal neutron measurements in the cyclotron room is suitable for practical use. In this study, we employed high-purity NaCl (Nakarai Tesque, Inc., Kyoto, Japan) as the sample because impurities influence the amount of undesirable induced radioactivity. Even the impurity levels in table salt sold in markets are well controlled if it is manufactured by chemical treatment. Because NaCl samples are difficult to prepare,

aluminum casings were used to hold the NaCl samples. The cross-section of aluminum for thermal neutron capture is comparable to that of Na. However, these casings are very thin, about 0.1 mm. Thus, the induced radioactivity in the casing is extremely low because of the smaller target nuclide in the irradiated aluminum casing. Consequently, the induced radioactivity in the aluminum casing seems to be negligible for thermal neutron measurements, given that we could not detect any significant number of counts using the NaI spectrometer. Moreover, there was no difference between the NaI spectrometer counts observed when Parafilm was or was not used. Therefore, the activation of Parafilm is negligible in this measurement. However, the present method has a limitation: the amount of 24Na decreases by decay, which limits the number of samples that can be tested by an individual detector. However, the autoradiograph seems to be able to make collective measurements of multiple samples possible. A method that uses a high-sensitivity imaging plate to measure the radioactivity induced in gold foils has already been reported [9, 15]. Facilities that have this system could measure sample radioactivity efficiently. The amount of 198Au radioactivity generated in the present study was about 800 Bq per gold foil. Even if 100 samples were to be used, the total amount becomes 8 kBq, which is only 8% of the exemption level of the Basic Safety Standard of International Atomic Energy Agency (1 MBq). 4.3 Photon and neutron distributions in the cyclotron room The distribution patterns for thermal neutrons and photons during cyclotron operation may differ because the shield structure of the target box was different for neutrons and photons. The photon distribution under nonoperating conditions was similar to the neutron distribution during operation. During nonoperation, most of the photons appeared to come from the radionuclides in the cyclotron

Measurement of thermal neutron fluence distribution

itself and in the floor (induced by thermal neutrons). To reduce the induced radioactivity in the facility, it is important to take the shielding of neutrons into account when designing an accelerator. To reduce the radioactivity induced in concrete, lowradioactivation concrete has been developed, and some facilities have installed this type of concrete [16]. It is difficult to design a general method for reducing the induced radioactivity for different cyclotron energy levels and different types of self-shielding, taking the cost into account. However, a provisional estimate of this cost, including the cost of decommissioning the cyclotron facilities, is important for each institute. A high-accuracy estimation would require neutron data. Because the size and layout of the cyclotron room are different for each facility, each institution should estimate its own neutron dose distribution. This study suggests that each facility should be able to evaluate the neutron flux conveniently using salt.

5 Conclusion A radioactivation measurement method using salt was examined as a practical method of measuring the thermal neutron fluxes in medical facilities. It was possible to measure the thermal neutron flux using salt as well as a NaI spectrometer for measuring the gamma rays emitted by 24 Na. By measuring the thermal neutron flux, it is possible to promote radiation safety and prepare for decommissioning work by reducing excess induced radioactivity. Acknowledgments This survey was performed as a part of research conducted by the Japanese Society of Radiological Technology (JSRT) research group on the ‘‘Control and disposal of radiological waste.’’ Measurements were undertaken under a grant-in-aid for Scientific Research from the Ministry of Health, Labour and Welfare for research on medical safety and a health technology assessment entitled ‘‘Research on securing safety of medical radiation’’ (H19medical treatment-common-003) (chief researcher: Makoto Hosono). This study was discussed at the debriefing sessions of the research project on radioactivity held at the small-scale radiation generator facilities and convened by MEXT. Sessions were held at the High Energy Accelerator Research Organization (KEK). Part of this study was presented at the 6th Annual Meeting of the Japanese Society of Radiation Safety Management, where significant advice was received from the chairperson, Dr. Ohishi.

165

References 1. Kumada Y. Development of compact cyclotron for PET (in Japanese). J Part Accelerator Soc Jpn. 2005;2(2):191–6. 2. Japan Radioisotope Association. Statistics on the use of radiation in Japan 2007. Tokyo: Japan Radioisotope Association; 2007, 24:9. 3. Sectional Chief, Radiation Protection Division, Nuclear Safety Bureau, Science and Technology Agency. About the handling of the radioactivate nuclide at facilities of radiation generator. Saitama: Science and Technology Agency; 1998. 4. Ishimoto T, Kubota H, Mori T, et al. Evaluation of activation and decommissioning of the medical compact cyclotron (in Japanese). Radioisotopes. 2005;54:185–92. 5. Masumoto K, Ohtsuki T, Kasagi J, et al. Evaluation of radioactivities induced in the shielding of the neutron scattering facility and its decommissioning. Hoken Buturi. 1990;34(2):151–60. 6. Shibata T. Radioactive material management for accelerator dismantlement. Radioisotopes. 1999;48(3):208–15. 7. Yoshimura A, Uehara S, Ohsaki S, et al. Activation of 40Ar in air induced by neutrons from small size cyclotron. Radioisotopes. 1993;42(6):325–9. 8. Beckurts KH, Wirtz K. Neutron physics. New York: Springer; 1964. 9. Kasahara T, Fujibuchi T, Iimori T, et al. Measurement of thermal neutron distribution from a medical cyclotron using auto radiography with 23Na activation detector and medical imaging plate (in Japanese). Jpn J Radiat Saf Manag. 2008;7:49–54. 10. Sakama M, Saze T, Tanii T, et al. Determination of neutron fluxes at the self-shielded PET cyclotron and the electron linac apparatus of Tokushima University Hospital using activation foil method (in Japanese). Jpn J Radiat Saf Manag. 2008;7:138–47. 11. Hirayama H, Namito Y, Bielajew AF, et al. The EGS5 code system (SLAC-R-730; KEK Report 2005–8). Stanford: SLAC; 2005 12. Shibata K, Kawano T, Nakagawa T, et al. Japanese evaluated nuclear data library, version 3, revision 3. J Nucl Sci Technol. 2002;39:1125–36. 13. Japan Radioisotope Association. Radioisotope pocket data book. 10th ed. Tokyo: Maruzen; 2001. 14. Yamamoto K, Kumada H, Yamamoto T, et al. Reproducibility of thermal neutron flux distribution on patient’s brain surface with a realistic phantom. Trans At Energy Soc Jpn. 2004;3(2):193–9. 15. Masumoto K, Toyoda A, Eda K, et al. Measurement of the spatial distribution of neutrons in an accelerator room by the combination of activation detectors and an imaging plate. Radiat Saf Manag. 2002;1:12–6. 16. Kinno M, Kimura K, Nakamura T. Raw materials for low-activation concrete neutron shields. J Nucl Sci Technol. 2002; 39(12):1275–80.