Development of an alpha/beta/gamma detector for radiation monitoring

Seiichi Yamamoto, and Jun Hatazawa

Citation: Rev. Sci. Instrum. 82, 113503 (2011); doi: 10.1063/1.3658821View online: http://dx.doi.org/10.1063/1.3658821View Table of Contents: http://aip.scitation.org/toc/rsi/82/11Published by the American Institute of Physics

REVIEW OF SCIENTIFIC INSTRUMENTS 82, 113503 (2011)

Development of an alpha/beta/gamma detector for radiation monitoringSeiichi Yamamoto1,a) and Jun Hatazawa2

1Kobe City College of Technology, 8-3, Gakuen-Higashi-machi, Nishi-ku, Kobe, 651-2194, Japan2Osaka University of Graduate School of Medicine, 2-2, Yamadaoka, Suita-shi, Osaka, 565-0871, Japan

(Received 11 August 2011; accepted 16 October 2011; published online 14 November 2011)

For radiation monitoring at the site of nuclear power plant accidents such as Fukushima Daiichi, ra-diation detectors not only for gamma photons but also for alpha and beta particles are needed becausesome nuclear fission products emit beta particles and gamma photons and some nuclear fuels containplutonium that emits alpha particles. We developed a radiation detector that can simultaneously mon-itor alpha and beta particles and gamma photons for radiation monitoring. The detector consists ofthree-layered scintillators optically coupled to each other and coupled to a photomultiplier tube. Thefirst layer, which is made of a thin plastic scintillator (decay time: 2.4 ns), detects alpha particles. Thesecond layer, which is made of a thin Gd2SiO5 (GSO) scintillator with 1.5 mol.% Ce (decay time:35 ns), detects beta particles. The third layer made of a thin GSO scintillator with 0.4 mol.%Ce (decay time: 70 ns) detects gamma photons. By using pulse shape discrimination, the count ratesof these layers can be separated. With individual irradiation of alpha and beta particles and gammaphotons, the count rate of the first layer represented the alpha particles, the second layer representedthe beta particles, and the third layer represented the gamma photons. Even with simultaneous irra-diation of the alpha and beta particles and the gamma photons, these three types of radiation can beindividually monitored using correction for the gamma detection efficiency of the second and thirdlayers. Our developed alpha, beta, and gamma detector is simple and will be useful for radiationmonitoring, especially at nuclear power plant accident sites or other applications where the simulta-neous measurements of alpha and beta particles and gamma photons are required. © 2011 AmericanInstitute of Physics. [doi:10.1063/1.3658821]

I. INTRODUCTION

For radiation monitoring at nuclear power plant acci-dents such as Fukushima Daiichi, radiation detectors notonly for gamma photons but also for alpha and beta parti-cles are needed because some nuclear fission products emitbeta and gamma photons, and the mixed oxide fuel usedin one of the power plants contained plutonium that emitsalpha particles. In nuclear pollution environments, the al-pha and beta particles must also be monitored in additionto the gamma photons. Monitoring gamma photons is rel-atively easy and is usually performed with scintillation de-tectors. However, alpha and beta particles are difficult tomonitor because of their short ranges, and most of theseparticles are absorbed in the entrance window of radiationdetectors.

Radiation detectors exist for individually measuring al-pha and beta particles and gamma photons. Thin ZnS(Ag)scintillation detectors are the most common method for alphaparticle detection, such as TCS-232 B, Hitachi-Aloka Med-ical Co., Japan. Plastic scintillation based detectors are alsoused for beta particle detection, such as TCS-316 H, Hitachi-Aloka Medical Co., Japan. A NaI(Tl) scintillation detectoris most commonly used for gamma photon detection, suchas TCS-171 B, Hitachi-Aloka Medical Co., Japan. However,using three types of scintillation detectors is not economical

a)Author to whom correspondence should be addressed. Electronic mail:s-yama@kobe-kosen.ac.jp.

and may actually increase radiation doses in highly contami-nated areas because it takes a long time to measure with threedetectors.

To solve this problem, the phoswich technique was intro-duced for alpha and beta particle detectors made of ZnS(Ag)and plastic scintillators.1 A similar phoswich detector is com-mercially available (TCS-362, Hitachi-Aloka Medical Co.,Japan) that employed ZnS(Ag) and plastic scintillators andcombined two photomultiplier tubes (PMTs) to simultane-ously measure alpha and beta particles. In such a detector,there is a problem: the plastic scintillator for beta particle de-tection also detects gamma photons.

The phoswich technique was expanded to a three-layerconfiguration to measure three different kinds of radiation.2–4

One phoswich detector employed a ZnS(Ag) scintillator foralpha particles, a plastic scintillator for beta particles, andNaI(Tl) or Bi4Ge3O12 (BGO) for gamma photons.2 Anotheremployed ZnS(Ag) for alpha particles, CaF2 for beta parti-cles, and NaI(Tl) for gamma detectors.3 However, since thelight output of ZnS(Ag) is high and the energy of most al-pha particles is high (5–8 MeV), pulse height differences werefound between scintillators that complicate signal processing.The detection of gamma photons in scintillators for beta par-ticles must also be solved.

A thin plastic scintillator, which is another candidate foran alpha particle detector, is used for radon detectors.5, 6 It istransparent and can measure the energy spectrum for alphaparticles when it is combined with a PMT. Since its light out-put is smaller than other scintillators such as ZnS(Ag), it isa good candidate for a high energy alpha particle detector in

0034-6748/2011/82(11)/113503/6/$30.00 © 2011 American Institute of Physics82, 113503-1

113503-2 S. Yamamoto and J. Hatazawa Rev. Sci. Instrum. 82, 113503 (2011)

FIG. 1. (Color online) Conceptual drawing of alpha/beta/gamma phoswichdetector.

the phoswich configuration with other scintillators with highlight output.

Ce doped Gd2SiO5 (GSO), a scintillator mainly used forpositron emission tomography (PET) systems, has a uniquecharacteristic; the decay time can be controlled with its Ceconcentration7 and used for depth-of-interaction (DOI) detec-tors for PET.8, 9 Two GSO scintillators with different Ce con-centrations were used for a beta detector for a blood samplingsystem.10 The detector uses a phoswich detector made of0.5-mm thick GSO scintillators with different Ce concentra-tions (thus different decay times). The first layer detects betaparticles plus gamma photons and the second layer only de-tects gamma photons. By subtracting the counts of the secondlayer from the first, the beta counts can be estimated.

We found that a combination of a thin plastic scintillatorwith thin GSO scintillators with different Ce concentrationsis suitable as a phoswich detector for the simultaneous mea-surement of alpha and beta particles and gamma photons be-cause the pulse heights of these three scintillators are similarand their decay times are adequately different. With a simi-lar method as a gamma photon compensation method for aGSO blood sampling system,10 it may be possible to sepa-rately count the beta particles and the gamma photons de-tected in the scintillator for beta particles. Consequently, wedeveloped a phoswich detector that can separately monitor al-pha and beta particles and gamma photons by a thin plasticscintillator and two types of thin GSO scintillators.

II. MATERIALS AND METHODS

A. Principle of operation

Figure 1 shows a schematic drawing of a detector thatconsists of three layers of scintillators optically coupled toeach other and coupled to a PMT.

The first layer was made of an 8-mm diameter, 50-μmplastic scintillator (decay time: 2.4 ns). Since the range oftypical alpha particles (energy: ∼6 MeV) in plastic is around50 μm, all of the energy of the alpha particles is absorbedin the first layer.5, 6 The second layer was made of an 8-mmdiameter, 0.5-mm thick GSO scintillator with 1.5 mol.% Ce

FIG. 2. (Color online) Developed alpha/beta/gamma phoswich detector,electronics, and personal computer.

(decay time: 35 ns). The maximum range of high energybeta particles (∼2 MeV) in water is around 5 mm. Thus, a0.5-mm thick GSO can absorb most of the energy of the betaparticles because its density is approximately six times higherthan water.10 The third layer was made of an 8-mm diameter,0.5-mm GSO scintillator but with 0.4 mol.% Ce, so the de-cay time is longer (70 ns). Since the alpha and beta particlescannot reach the third layer because of their ranges, the thirdlayer only detects gamma photons.

The pulse shape of the output of the PMT has three types,depending on the number of layers (Fig. 1, right). For exam-ple, the detection of alpha particles in the first layer producesa very fast pulse because the decay time of the plastic scintil-lator is ∼2.4 ns. Similarly, detection in the second layer pro-duces pulses with medium decay time, and the third layer pro-duces slow pulses. By using pulse shape discrimination, thesethree pulses can be distinguished and separately monitored.

B. Developed alpha/beta/gamma detector

In Fig. 2, the entire developed three-layer phoswich al-pha/beta/gamma detector system is shown. We used a Hama-matsu 10-mm diameter photomultiplier tube (H3164-10). Theplastic scintillator was NE-102A (Ohyo Koken, Japan), andthe GSO scintillators with 1.5 mol.% Ce and 0.4 mol.% Cewere made by Hitachi Chemical, Japan. These three scintilla-tors were optically coupled to each other using silicone rubber(KE-420, Shin-etsu silicone, Japan) and optically coupled tothe PMT using the same silicone rubber. For the input lightshield, three layers of aluminum Mylar were used. The out-side diameter of the detector was 15 mm and its length was100 mm.

The signal from the PMT is processed digitally using a100-MHz analog-to-digital (A-D) converter. Signals higher

113503-3 S. Yamamoto and J. Hatazawa Rev. Sci. Instrum. 82, 113503 (2011)

FIG. 3. Pulse shape detected by developed detector for alpha particles fromAm-241 (energy: 6.0 MeV) (a), for beta particles from Sr-Y-90 (maximumenergy: 0.54 and 2.8 MeV) (b), and for gamma photons from Cs-137 (energy:0.66 MeV) (c).

than a threshold are digitally integrated for two different in-tegration times: partial (90 ns) and full (160 ns).11 Since fullintegration can collect most scintillation signals, it representsthe energy signal and is used for energy discrimination. Theratio of partial to full integration represents the pulse shapeof the scintillators and is used for the pulse shape discrimi-

nations. The energy and pulse shape discriminated signals arefed to a notebook personal computer (PC) and counted fora preset acquisition time. The energy window and the pulseshape discrimination levels can be digitally set by PC.

III. RESULTS

A. Pulse shape of detector

Figure 3(a) shows the pulse shape of the developed detec-tor for alpha particles from Am-241 (6.0 MeV). The measureddecay time is around 10 ns; the time is longer than the decaytime of the plastic scintillator (2.4 ns) because of the limita-tion of the band width of the amplifier. Figure 3(b) shows thepulse shape for the beta particles from Sr-Y-90 (maximumenergy: 0.54 and 2.8 MeV). The decay time is around 40 ns.Figure 3(c) shows the pulse shape for gamma photons fromCs-137 (0.66 MeV). The decay times have two components:∼40 ns and ∼80 ns. The former is the scintillation pulse fromthe second layer and the latter is from the third layer.

B. Pulse shape and energy spectra of detector

Figure 4(a) shows the pulse shape (left) and the energyspectrum (right) for the Am-241 alpha particles measuredby the detector. The pulse shape spectrum is the distributionshowing the scintillator’s decay time. The horizontal axis inthe spectrum is roughly proportional to the inverse of the de-cay time. The horizontal axes in Fig. 4 are relative values.The pulse shape spectrum showed a peak on the right, andthe energy spectrum showed a single peak that corresponds tothe alpha particles (6 MeV) detected by the plastic scintilla-tor. The energy resolution for the 6 MeV alpha particles was∼33% full-width at half-maximum (FWHM).

Figure 4(b) shows the pulse shape (left) and the energyspectrum (right) for the Sr-Y-90 beta particles measured bythe detector. The pulse shape spectrum showed a peak in themiddle position, and the energy spectrum showed broad dis-tribution, which is a typical beta energy spectrum.

Figure 4(c) shows the pulse shape (left) and the energyspectrum (right) for the Cs-137 gamma photons measuredby the detector. The pulse shape spectrum showed doublepeaks: from the second and third layers. The energy spectrumshowed a single peak that corresponds to the photo-peak ofthe gamma photons (0.66 MeV) detected by the GSO scintil-lators. The energy resolution for the 0.66 MeV gamma pho-tons was ∼19% FWHM.

C. Background count rate measurements

We measured the background count rate of our developeddetector. Without any radioisotopes around the detector, thecount rates for the three layers were measured for 1 h. The av-erage count rate for each layer was evaluated. Table I showsthe results of the average background count rate of the de-tector. The background count rates were low, and they wereprobably from natural radioisotopes such as K-40.

113503-4 S. Yamamoto and J. Hatazawa Rev. Sci. Instrum. 82, 113503 (2011)

FIG. 4. (Color online) Pulse shape spectra (left column) and energy spectra (right column): for Am-241 alpha particles (a), Sr-Y-90 beta particles (b), andCs-137 gamma photons (c). Vertical lines on left are threshold levels for pulse shape discrimination and those on right are energy windows.

D. Count rate curves individually irradiating alphaand beta particles and gamma photons

Figure 5(a) shows the raw count rates for each layer whileindividually irradiating the alpha and beta particles and thegamma photons. Am-241, Sr-Y-90, and Cs-137 were used forthe alpha, beta, and gamma sources, respectively. In the curvefor the gamma photons, the count rate of the second layer isalso observed in addition to the third layer because the sec-ond layer is made of GSO and has similar detection probabil-ity as the third layer. Because some Compton scattered lower

TABLE I. Average background count rate for three layers of developeddetector.

First layer (cps) Second layer (cps) Third layer (cps)

0.008 0.16 0.12

energy photons were absorbed in the second layer, the countrate of the third layer for the gamma photons is smaller thanthe second layer. The detection of gamma photons in the sec-ond layer complicates distinguishing the count rate of the betaparticles if the type of radiation source is not provided.

To solve this problem, we employed a subtractionmethod. By assuming that the detection probabilities of thegamma photons of the second to the third layers are constant,corrected beta count rate Cbeta was derived by subtraction:

Cbeta = C2 − cfC3, (1)

where C2 and C3 are the count rates of the second and thirdlayers. “cf” is the ratio of the probabilities of the detectionefficiencies for the gamma photons of the second to the thirdlayers. The value was determined experimentally to be 1.7.

In Fig. 5(b), the corrected count rates for these three lay-ers are shown. With correction, the count rates for the alpha

113503-5 S. Yamamoto and J. Hatazawa Rev. Sci. Instrum. 82, 113503 (2011)

FIG. 5. (Color online) Count rates for each layer while individually mea-suring alpha and beta particles and gamma photons: raw data (a) and aftercorrection (b).

and beta particles and the gamma photons were separatelydetected.

The detection probabilities of this detector after correc-tion are shown in Table II. When the alpha particles were ir-radiated, 7.1% were misidentified as beta particles. When irra-diating the beta particles, 0.8% were misidentified as gammaphotons. When the gamma photons were irradiated, they weremisidentified as beta particles −8.1% after correction.

TABLE II. Detection probabilities of detector after correction for alpha andbeta particles and gamma photons.

Alpha counts (%) Beta counts (%) Gamma counts (%)

Alpha irradiated 100 7.1 0Beta irradiated 0 100 0.8Gamma irradiated 0 −8.1 100

FIG. 6. (Color online) Count rates for each layer while simultaneously mea-suring alpha plus gamma, beta plus gamma and beta: raw data (a) and aftercorrection (b).

E. Count rate curves simultaneously irradiating alphaand beta particles and gamma photons

Figure 6(a) shows the raw count rate curves for each layerwhile simultaneously measuring the irradiated alpha particlesand the gamma photons and the beta particles and the gammaphotons. The gamma photons and the beta particles were alsoirradiated individually during the measurements.

The alpha particles and the gamma photons were sep-arately measured by the first and third layers, respectively.However, when the beta particles and the gamma photonswere simultaneously irradiated, we could not separate themin the second layer without correction.

The corrected count rate curve based on Eq. (1) is shownin Fig. 6(b). The count rates for the alpha and beta particles

113503-6 S. Yamamoto and J. Hatazawa Rev. Sci. Instrum. 82, 113503 (2011)

TABLE III. Average count rate for alpha and beta particles and gammaphotons irradiated simultaneously after correction.

Alpha count Beta count Gamma countrate (cps) rate (cps) rate (cps)

Gamma irradiated 0 − 2.2 12.6Alpha and gamma irradiated 27.1 0.4 11.7Beta and gamma irradiated 0.2 15.1 12 .0Beta irradiated 0 15.5 0

and the gamma photons were detected only when they wereirradiated.

Table III shows the average count rate values for the alphaand beta particles and the gamma photons after correction.The corresponding count rates increased only when the alphaor beta particles or the gamma photons were irradiated. Thecount rates did not significantly change with or without othertypes of radiation that were simultaneously irradiated.

IV. DISCUSSION

Using the three-layer phoswich detector, simultaneousmeasurements of alpha and beta particles and gamma pho-tons became possible. Our developed detector will be usefulat sites where the radioisotope information is not well known.In such situations, we must measure the alpha and beta par-ticles as well as the gamma photons. The most suitable sit-uation for this detector is nuclear power plant accident sites,such as the Fukushima Daiichi area, where the radioisotopeinformation is sometimes not obvious enough.

Our developed detector can also provide energy spec-tra from the full integration data. We observed single peaks

for alpha particles and gamma photons (Figs. 4(a) right and4(c) right). From the energy spectra, we can estimate theapproximate energies for the alpha particles and the gammaphotons that provide information for determining the type ofradioisotopes.

The subtraction method for determining beta counts us-ing Eq. (1) worked well in the simultaneously measured datawith beta particle and gamma photon irradiation (Fig. 6(b)).One problem in this method is that the subtraction method in-creases the statistical deviation of the measured data. Such anincrease was observed in the simultaneously irradiated peri-ods with beta particles and gamma photons in Fig. 6(b).

V. CONCLUSION

Simultaneous measurements of alpha and beta particlesand gamma photons became possible with a three layereddetector. This radiation detector will be useful for radiationmonitoring, especially at nuclear power plant accident sites.

1S. Usuda, H. Abe, and A. Mihara, J. Nucl. Sci. Technol. 29, 927 (1992).2S. Usuda, H. Abe, and A. Mihara, Nucl. Instrum. Methods Phys. Res. A340, 540 (1994).

3T. L. White and W. H. Miller, Nucl. Instrum. Methods Phys. Res. A 422,144 (1999).

4K. Yasuda, S. Usuda, and H. Gunji, IEEE Trans. Nucl. Sci. 47, 1337 (2000).5S. Yamamoto and T. Iida, Nucl. Instrum. Methods Phys. Res. A 418, 387(1998).

6S. Yamamoto, K. Yamasoto, and T. Iida, IEEE Trans. Nucl. Sci. 46(9), 1929(1999).

7H. Ishibashi, K. Shimizu, K. Susa, and S. Kubota. IEEE Trans. Nucl. Sci.36(1), 170 (1989).

8S. Yamamoto and H. Ishibashi. IEEE Trans. Nucl. Sci. 45(3), 1078 (1998).9S. Yamamoto, Nucl. Instrum. Methods Phys. Res. A 598, 480 (2009).

10S. Yamamoto, M. Imaizumi, E. Shimosegawa, Y. Kanai, Y. Sakamoto,K. Minato, K. Shimizu, M. Senda, and J. Hatazawa, Phys. Med. Biol.55(13), 3813 (2010).

11S. Yamamoto, Nucl. Instrum. Methods Phys. Res. A 587, 319 (2008).