Detection of alpha particle emitters originating from nuclear fuel inside reactor building of Fukushima Daiichi Nuclear Power Plant

We measured alpha emitters obtained from a reactor building in the Fukushima Daiichi Nuclear Power Plant (FDNPP) by using an alpha particle imaging detector. For developing the detector, we used a very thin (0.05-mm-thick) a cerium-doped Gd3(Ga,Al)5O12 (Ce:GAGG) scintillator and silicon photomultiplier (SiPM) arrays as the photodetector. The floor of the reactor building in FDNPP was wiped off by using smear papers, and the radioactivity of these papers was measured by the alpha particle imaging detector. In addition, we measured a Plutonium (Pu) sample (mainly 5.5 MeV alpha particles from 238Pu) obtained from a nuclear fuel facility by using of the same detector for comparison with the smear papers. The alpha spectrum was in the energy range of 5–6 MeV, which corresponds to the alpha particle energy of 238Pu (5.5 MeV). The correlation coefficient of the alpha spectra of the smear papers and the Pu sample had a strong positive linear relation. Moreover, the peak of 241Am was identified by gamma spectrum measurement. Based on these results, we report actual findings of alpha emitters in the FDNPP reactor buildings originating from nuclear fuels. The surface contamination level of alpha emitters exceeded 4 Bq/cm2.

Another aim of measuring alpha emitters is obtaining information about the two-dimensional distribution of alpha emitters. In the previous study, alpha particles from Pu characteristically show up as spots in the two-dimensional distributions 20 . Information about the size and distribution of alpha emitters is very useful for radiation protection. For example, the particle size of alpha emitters is important for estimating internal exposure to a worker 12 . Moreover, information about the two-dimensional distribution of alpha emitters is valuable for distinguishing Pu from Rn progeny because the distributions of alpha particles between them are different 20,23,24 . An imaging plate (IP) has been widely used to obtain the two-dimensional distribution of alpha emitters 25,26 , but this plate is sensitive to beta particles and gamma rays as well. There are large quantities of beta and gamma emitters at the FDNPP site. The IP cannot distinguish alpha particles from beta particles and gamma rays and identify radionuclides of alpha emitters. Therefore, energy information is necessary to measure only alpha emitters in FDNPP.
We have developed an alpha particle imaging detector that can measure the energy spectrum and the two-dimensional distribution of alpha particles by combining a cerium-doped Gd 3 (Ga,Al) 5 O 12 (Ce: GAGG) scintillator with a silicon photomultiplier (SiPM). This detector has good energy resolution (~13% full width at half maximum (FWHM)) and high spatial resolution (0.6 mm FWHM) for 5.5 MeV alpha particles 27 . Furthermore, its beta and gamma sensitivity can be reduced by setting the thickness of the Ce: GAGG scintillator to 0.05 mm.
In this study, we measured alpha emitters in the reactor building of FDNPP for the first time by using the developed alpha particle imaging detector. The features of radionuclides of alpha emitters in the reactor building of FDNPP were clarified by these measurements.

Results
Preparation of smear papers. Surface contamination on the floor inside the FDNPP reactor building was wiped off by using smear papers for the measurement. Figure 1 shows photographs of the smear papers. The diameter of the smear papers was 5 cm. The smear papers were covered with a thin polyethylene film to prevent cross-contamination to the detector. The presence of surface dirt on smear papers nos 1, 3, and a little on no. 2, and red paint on smear paper no. 4 was confirmed visually. These smear papers were measured using a ZnS(Ag) survey meter and a Geiger-Muller (GM) counter, and measured results are listed in Table 1. The alpha count rate measured using the ZnS(Ag) survey meter was up to 5,000 cpm. The beta (gamma) count rate measured using the GM counter was higher than the upper measurement limit of the GM counter (>100,000) in the cases of smear papers no. 1 and 4. The beta (gamma) count rate was considerably higher than that of alpha particles (18-1200 times higher). Moreover, it was confirmed that the ZnS(Ag) survey meter was not influenced by beta and gamma rays by comparing the count rates obtained with and without paper.
The total number of smear papers was 4. Smear paper no. 3 yielded the highest count rate of 5,000 cpm for alpha particles, whereas smear paper nos1 and 4 yielded the highest count rates of >100,000 cpm for beta particles. Therefore, there was no correlation between the alpha and beta count rates. Fundamental performance of the alpha particle imaging detector. Figure 2 shows the energy spectrum of 5.5-MeV alpha particles from the 241 Am source. The energy resolution was calculated by Gaussian fitting. The energy resolution was 13.3% FWHM. The spatial resolution of the developed detector was evaluated by measuring FWHM of an intensity profile of the Pu point source shown in Fig. 3. The spatial resolution was found to be 0.59 mm FWHM.
The count rate of the 241 Am source measured using the alpha particle imaging detector for 2 min was 132700 cpm (2212 cps). The surface emission rate of the 241 Am source was 2217 s −1 . The efficiency was calculated using the following formula 28 : The measured count rate(cps) The surface emission rate of the 241Am source( ) 100 (1) 1 The efficiency was 99.8%. This efficiency was used to calculate surface contamination level in units of Bq/cm 2 .
Energy spectrum of the FDNPP smear papers measured using the alpha particle imaging detector. Figure 4 shows the energy spectra of the smear papers. In all the measured energy spectra, the peak at 2 MeV was created by beta particles from 137 Cs and 90 Sr( 90 Y). However, the shapes of energy spectra of each of the smear papers in the 4-6 MeV region, which corresponds to energies of alpha particles from Pu isotopes (5.  Table 1. List of smear paper numbers, place of wiping, and count rates measured using ZnS(Ag) survey meter, GM counter, and cadmium zinc telluride (CZT) spectrometer.  spectra were confirmed up to 4 MeV. In the cases of smear papers no. 2 and 4, the energy spectra were confirmed up to 6.5 MeV. By inserting a paper between the detector and the smear paper, only beta (gamma) spectrum can be measured because alpha particle can be shielded. The alpha spectrum was obtained by subtracting normalized beta spectrum. Figure 5 shows the alpha spectrum of the smear paper. In the cases of smear papers no. 1 and 3, the alpha energy spectrum was confirmed from 3 MeV to 4 MeV. By contrast, in the cases of smear papers no. 2 and 4, the energy spectrum was confirmed from 5 MeV to 6 MeV, which corresponds to an alpha particle energy of 5.5 MeV from 238 Pu. Dirt particles, for example, soil, can be seen in the photographs of smear papers no.1 and 3. Alpha particles can be absorbed by dirt. and as a result, their energy decreases in cases of smear papers no. 1 and 3. By contrast, smear papers no. 2 and 4 did not contain as many dirt particles as did smear papers no. 1 and 3, and the energies of the alpha particles on those papers were not affected by self-absorption. Therefore, their alpha spectra could be confirmed in the energy range of 5 MeV to 6 MeV. Figure 6 shows comparisons of the energy spectra of the smear papers and the Pu sample measured at the nuclear fuel facility 20 . The energy spectrum of smear papers no. 2 or 4 and that of the Pu sample confirmed up to 6.5 MeV. The shape of the spectra of smear papers no. 2 or 4 and the Pu samples were different because both smear paper and the Pu sample were affected by self-absorption and their peaks were shifted to lower energy. In the results of all smear papers, the alpha spectrum was not confirmed at energies exceeding 6 MeV, which means that this spectrum was not created by naturally occurring radionuclides such as Radon progeny (mainly 7.7 MeV alpha particle from 214 Po). To show the relationship between the relative counts of the Pu samples and the smear papers quantitatively, the correlation coefficient of the relative counts of the Pu sample and the smear papers in the energy range of 5-6.5 MeV was evaluated. To calculate the correlation coefficient, first, all spectra were fit using polynomial functions to standardize a width of an energy bin in the energy spectrum. Then, the correlation coefficient between the Pu sample and the smear papers was calculated. Figure 7 shows the correlation coefficient of the relative counts of the Pu sample and the smear papers in the energy range of 5-6.5 MeV. The correlation coefficient r was more than 0.93 in all cases, and the relative counts of the Pu sample and the smear papers share a strong, positive linear relationship. Therefore, this alpha spectrum must have been created by artificial radioactivity originating from a nuclear fuel such as 238 Pu (5.5 MeV alpha particle), 239 Pu (5.15 MeV alpha particle), and 241 Am (5.5 MeV alpha particle).
Two-dimensional distribution of the FDNPP smear papers measured using the alpha particle imaging detector. Figure 8 shows the two-dimensional distribution of the smear papers in the energy range of 0-8 MeV (alpha + beta particles). The two-dimensional distributions of each of the smear papers were different. In the case of smear paper no.1, a pattern with a few streaks was seen. In the case of smear paper no. 2, oblique lines and a spot were seen. In the case of smear paper no. 3, many spots and a patchy pattern were seen. In the case of smear paper no. 4, many spots were distributed in the image. Figure 9 shows the two-dimensional distributions of the smear papers in the energy range of 0-3 MeV (beta particles). The two-dimensional distributions in the 0-8 MeV region and the 0-3 MeV region were similar. Therefore, almost all spots on the two-dimensional distributions were formed by beta particles. It is possible that these particles have been discovered recently in several studies 29,30 . Figure 10 shows the two-dimensional distribution of the smear papers in the energy range of 5-6.5 MeV (alpha particles). Alpha particle emitters were widely distributed on the smear paper, as opposed to forming spots. In addition, Fig. 11 shows the two-dimensional distribution of the Pu sample obtained from a fabrication process of mixed oxide (MOX) fuel in the nuclear fuel facility. These two-dimensional distribution images were created in the same energy range of 5-6.5 MeV. The spots were confirmed in both images of the Pu samples. Even though the energy ranges of the FDNPP smear paper and the Pu sample are identical (5-6.5 MeV), their two-dimensional  Gamma-ray spectrum of the FDNPP smear papers using a CZT-based gamma spectrometer. Figure 13 shows the gamma-ray spectra of the smear papers measured using a CZT-based gamma spectrometer.
In the spectra of smear paper no. 2 and 4, a peak at 60 keV was confirmed, and it was formed by gamma rays from 241 Am. In the spectra of smear papers no. 1 and 3, the peak at 60 keV was not confirmed because the count rates of 134 Cs and 137 Cs were too high and the peak was buried in the Compton continuum.
Surface contamination level of smear paper. The measured alpha count was converted to surface contamination (SC) in Bq/cm 2 and it was compared with that obtained using the ZnS(Ag) survey meter. The SC can be calculated using the following formula 31 : where S represents wiped area [cm 2 ] (=100 cm 2 ), F represents removal factor (=0.1, as recommended in ISO-7503-1 32 ), ε s denotes source efficiency (=0.25), and ε i denotes instrument efficiency. The instrument efficiency of the alpha particle imaging detector was 99.8% and that of the ZnS(Ag) survey meter was 32.5%/2π ± 25%, which is more than 30%/2π 33 . Because the effective area of the alpha particle imaging detector was smaller than the area of the smear paper (5 cm diameter), area ratio (=area of smear paper/ effective area of alpha particle imaging detector) was applied as a calibration factor. Figure 14 shows the SC level of the smear papers. Because the ZnS(Ag) survey meter has a wide range of instrument efficiency, the SC levels measured by the ZnS(Ag) survey meter were 12.8-24.4 Bq/cm 2 (smear paper no. 1), 0.6-1.1 Bq/cm 2 (no. 2), 58.0-111.1 Bq/cm 2 (no. 3), and 47.5-91.1 Bq/cm 2 (no. 4). By contrast, the SC levels measured by the alpha particle imaging detector were 11.5 Bq/cm 2 (smear paper no. 1), 1.7 ± 0.06 Bq/cm 2 (no. 2), 59.3 ± 0.21 Bq/cm 2 (no. 3), and 65.9 ± 0.03 Bq/cm 2 (no.4). The standard deviation of repeat measurements was minimal. The SC measured using the ZnS(Ag) survey meter corresponded to the alpha particle imaging detector within the variation of instrument efficiency of the ZnS(Ag) survey meter. Evaluation of Air contamination level. We evaluated air contamination from the measured SC level using following formula.

Air contamination Bq cm Surface contamination Bq cm
The resuspension factor K (cm ) where, the value of K for plutonium is 2 × 10 −8 cm −1 31 . The maximum SC level measured by the alpha particle imaging detector was 65.9 ± 0.03 Bq/cm 2 (smear no.4). The air contamination can be calculated to be 1.3E-6 Bq/cm 3 .

Discussion
We measured the existing alpha emitter contamination in smear papers at the FDNPP site by using the alpha particle imaging detector, and revealed that the alpha emitters in the reactor buildings at the FDNPP site originated from the nuclear fuels. This result has been impossible to obtain using ZnS(Ag) detectors. The correlation coefficient of the alpha spectra of the smear papers and the Pu sample had a strong positive linear relation. The Pu sample was obtained from the MOX fuel facility. Although 239 Pu is the dominant Pu isotope in the MOX fuel ( 238 Pu: 3.5%, 239 Pu: 47.4%) 34 , 238 Pu has the highest specific activity ( 238 Pu: 600 GBq/g, 239 Pu is 2 GBq/g) among the plutonium isotopes. Considering the specific activity and abundance ratio of Pu isotopes, 5.5 MeV alpha particles from 238 Pu majorly contribute to the alpha spectra. The form of Pu on the smear paper will be PuO 2 if it originated from the fuel 35 . Moreover, the 241 Am peak at 60 keV was found by measuring the gamma spectrum. 241 Am is a decay product of 241 Pu and is often measured as the evidence of Pu contamination. Furthermore, in an evaluation report of the amount of radioactive nuclides in the reactor core by using the ORIGEN2 code, 238 Pu and 241 Am that emit 5.5 MeV alpha particles showed high activities among the alpha emitters in nuclear fuel (e.g., in the reactor building unit 2 after 5 years, 238 Pu:4.87E+6GBq/core, 239 Pu:8.95E+5GBq/core, 240 Pu:1.04E+6GBq/ core, and 241 Am:2.43+6GBq/core) 36 . Based on these results, the major alpha emitters on the smear papers are 238 Pu and 241 Am. The beta activity of the FDNPP smear papers was approximately 10-1000 times higher than their alpha activity. Therefore, the influence of beta particles must be considered in alpha spectrum measurement. In this study, we used a very thin (0.05-mm-thick) scintillator to reduce the influence of beta particles. However, in measured spectra of the smear papers, beta spectrum and alpha spectrum overlapped partly. By subtraction with/without  paper, the beta spectrum could be removed. In the future, a technique for discrimination of beta particles is necessary for accurate alpha spectrometry. The two-dimensional distribution of alpha particles on the FDNPP smear papers was different from that of the Pu sample; In the Pu sample, spots were observed, but spots and wide distribution of minute particles were  simultaneously observed in the FDNPP smear papers. The Pu sample originated from particles used to fabricate MOX fuel, and its diameter was approximately a few micrometers 12,37-39 . However, the alpha emitters in FDNPP would have different diameters and compositions compared to the Pu particle in the MOX fuel fabrication facility. Further studies are required to clarify the diameter and composition of the Pu in the FDNPP. Sato et al. subjected a rubble sample from reactor unit 2 to radiochemical analysis 40 . The radioactive concentration was (6.1 ± 0.3) × 10 1 Bq/g of 238 Pu, (2.5 ± 0.2) × 10 1 Bq/g of 239+240 Pu, (2.4 ± 0.2) × 10 1 Bq/g of 241 Am, and (5.1 ± 0.3) × 10 1 Bq/g of 244 Cm. Our measurement smear paper no. 4 showed that the radioactive concentration of alpha particles was (6.6 ± 0.003) × 10 1 Bq/cm 2 . Both measurement results are of the same order of magnitude.
Given that the SC levels of three of the four smear papers easily exceeded 4 Bq/cm 2 which is considered an acceptable level in Japan, the level of contamination can be thought to be extremely high. This contamination is loose as opposed to being fixed because it can be removed using smear papers. It can lead to internal exposure, and therefore, extreme caution is required to prevent inhalation of Pu contamination. A worker entering into FDNPP must put on a full-face mask. Considering the resuspension factor (2 × 10 −8 cm −1 ) and the protection  factor of the full-face mask (generally, 50), the air concentration inside the mask is estimated to be below the air concentration limit of 238 Pu (7 × 10 −7 Bq/cm 3 ) 41 .
We detected many beta-emitting particles with extremely high activity. The combination of multiple technologies such as Transmission Electron Microscope is useful to identify them 29,30 .
A chemical process is very difficult to apply on-site at the FDNPP. In Sato's chemical analysis, the samples had to be transported from the FDNPP to the Nuclear Science Research Institute of the Japan Atomic Energy Agency 40 . Even though our measurements were conducted on-site at the FDNPP without destroying the samples, the results showed the radioactive concentration of alpha particles, which was of the same order of magnitude as measured using a radiochemical analysis. Therefore, our measurement method proved to be valuable for detecting alpha emitter contamination at the FDNPP site.

Methods
Alpha particle imaging detector. An alpha particle imaging detector must have good energy resolution for identifying radionuclides of alpha emitters. Moreover, because there are large quantities of beta and gamma emitters at the FDNPP site, the influences of beta and gamma rays must be distinguished from those of alpha particles. Figure 15 shows a schematic drawing of the alpha particle imaging detector developed in this study. A very thin (0.05-mm-thick) GAGG scintillator was used employed to limit beta sensitivity 42 . The length and the width of the scintillator size 26 mm × 26 mm. The 0.05-mm-thick GAGG scintillator was coupled to a 1-mm-thick glass plate and a 1-mm-thick acrylic light guide. The bottom of the acrylic light guide was optically coupled to silicon photomultiplier (SiPM) arrays having 8 × 8 channels (Through Silicon Via (TSV) MPPC array S12642-0404PA-50, Hamamatsu Photonics K.K., Japan) by using an optical compound (Shin-Etsu silicone, Shin-Etsu Chemical Co., Ltd., Japan). The detector was covered with aluminized mylar to shield it from external light. The GAGG scintillator produces scintillation light when an alpha particle deposits energy. The SiPMs convert this scintillation light into electric signals. The analog signals produced by the 8 × 8 SiPMs were fed to preamplifiers and a weight-summing circuit, which creates X+, X−, Y+, Y− signals for position calculation. These signals are converted to digital signals by using analog-to-digital converters and subsequently fed to a field programmable gate array. Finally, the data are transferred to a laptop via a lan cable, and the two-dimensional position and energy spectrum of the alpha particles are displayed on the laptop. Figure 16 shows a photograph of the alpha particle imaging detector and the data acquisition system. The data acquisition system is the same as that in the previous alpha imaging system 20 . High voltage (HV) was adjusted to −67 V, and the temperature of the SiPM circuit  was monitored to compensate for temperature-dependent changes in gain. During measurement, the detector and sample were enclosed in a light-tight box to shield them from external light. Fundamental performance of alpha particle imaging detector. Before the measurement at the FDNPP, we confirmed the fundamental performance, that is, energy resolution, spatial resolution, and efficiency, of the alpha particle imaging detector. Energy resolution and efficiency were evaluated using an 241 Am alpha source of diameter 5 mm that emitted 5.5 MeV alpha particles. A thin polyethylene film was placed between the source and the alpha imaging detector to create the same conditions as those during the measurement of samples at the FDNPP site. Spatial resolution of the developed detector was evaluated using a Pu point source. We got the predicted standard deviation by taking the square root of the number of counts of each bin in the spectrum 28 .
Measurement of smear papers by using alpha particle imaging detector. Figure 17 shows a schematic drawing of the measurement of the smear papers by using the developed alpha particle imaging detector. The alpha particle imaging detector and the smear papers were placed in a container in the field of the FDNPP. Each of the smear papers was placed on the detector and close contact between the paper and the detector was achieved by placing a weight on top. The smear paper was placed in a light-tight box, and the door of the container was closed to prevent penetration of external light. The measurement was repeated twice during a 5 min period. Also, the measurement was started at around noon to maintain the room temperature. The dose-rate of gamma rays in the measurement environment was approximately 10 μSv/h. Moreover, we measured two Pu samples obtained from a MOX fuel fabrication facility for comparison with the FDNPP smear papers. The measurement was conducted under the same conditions as those used in FDNPP (measurement time, HV). Figure 17. Photo of measurement of smear papers: Top view. The alpha particle imaging detector was placed under the smear paper.