Development and Performance Characteristics of Personal Gamma ...
sensors Technical Note
Development and Performance Characteristics of Personal Gamma Spectrometer for Radiation Monitoring Applications Hye Min Park and Koan Sik Joo * Department of Physics, University of Myongji, Yongin 449-728, Korea; [email protected] * Correspondence: [email protected]; Tel.: +82-10-9288-6436 Academic Editor: Vittorio M. N. Passaro Received: 9 April 2016; Accepted: 17 June 2016; Published: 21 June 2016
Abstract: In this study, a personal gamma (γ) spectrometer was developed for use in applications in various fields, such as homeland security and environmental radiation monitoring systems. The prototype consisted of a 3 ˆ 3 ˆ 20 mm3 Ce-doped Gd–Al–Ga–garnet (Ce:GAGG) crystal that was coupled to a Si photomultiplier (SiPM) to measure γ radiation. The γ spectrometer could be accessed remotely via a mobile device. At room temperature, the implemented Ce:GAGG-SiPM spectrometer achieved energy resolutions of 13.5%, 6.9%, 5.8%, and 2.3% for 133 Ba at 0.356 MeV, 22 Na at 0.511 MeV, 137 Cs at 0.662 MeV, and 60 Co at 1.33 MeV, respectively. It consumed only about 2.7 W of power, had a mass of just 340 g (including the battery), and measured only 5.0 ˆ 7.0 cm2 . Keywords: Ce-doped Gd–Al–Ga–garnet (Ce:GAGG); Si photomultiplier (SiPM); spectrometer; energy resolution; homeland security (HLS); environmental radiation monitoring system (ERMS)
1. Introduction Gamma (γ)-ray spectrometry analysis methods can be classified based on whether they involve direct or indirect detection, according to the manner in which the γ-rays are converted into electrical signals. The direct detection methods employ semiconductor devices that generate electrical signals immediately, with no intermediate stages, through γ-ray absorption. The indirect detection methods involve scintillation detection systems; in such a system, a photoelectric device detects electrical signals generated by photons with wavelengths in the visible range that are produced through interactions between the γ-rays and the scintillator. CdZnTe, CdTe, and HgI2 are primarily used in room temperature semiconductor devices employing direct detection methods [1]. These materials are preferred in γ-ray spectrometry analysis because they yield energy resolutions superior to those achievable with scintillator crystals. However, the maintenance of these materials is more expensive than that of scintillator crystals, and their detection efficiencies and physical rigidities are very low [2]. Currently, compound semiconductor materials are also more expensive than scintillator crystals. NaI(Tl), CsI(Tl), and LYSO(Ce) are the primary scintillator crystals used in indirect detection. These materials yield energy resolutions lower than those of semiconductors, but they are more sensitive. Because of the specific characteristics and limitations of each of these types of materials and the corresponding analysis methods, further development of detectors that combine scintillators and semiconductors is necessary and ongoing [3]. Thus, scintillator detectors remain promising candidates for use in homeland security (HLS) and environmental radiation monitoring systems (ERMSs). In this study, a personal γ spectrometer was developed, and its feasibility was evaluated by analyzing its energy resolution using standard γ-ray sources. The device described herein, called the “Personal-Spect”, essentially consists of a scintillation detector, signal processing electronics, a voltage supply, signal analysis components, and display units. Sensors 2016, 16, 919; doi:10.3390/s16060919
www.mdpi.com/journal/sensors
Sensors 2016, 16, 919
2 of 6
“Personal-Spect”, essentially consists of a scintillation detector, signal processing electronics, a voltage supply, signal analysis components, and display units. Sensors 2016, 16, 919
2 of 6
2. Materials and Methods An S12572-100C (photo-sensitive area: 3 × 3 mm2, Hamamatsu Photonics Co., Hamamatsu, 2. Materials and Methods Japan) Hamamatsu single Si photomultiplier (SiPM)2was used in the Personal-Spect to minimize the An S12572-100C (photo-sensitive area: 3 ˆ 3 mm , Hamamatsu Photonics Co., Hamamatsu, Japan) overall system dimensions (Table 1). Hamamatsu single Si photomultiplier (SiPM) was used in the Personal-Spect to minimize the overall system dimensions (Table 1). Table 1. Specifications of Si photomultiplier (SiPM) used in this study. Table 1. Specifications of Si photomultiplier (SiPM) used in this study. Parameter Value
Photosensitive area Parameter Number of pixels Photosensitive area range Spectral response Number of pixels Peak photon detection efficiency (at 450 nm) Spectral response range Bias voltage Peak photon detection efficiency (at 450 nm) Breakdown Bias voltage voltage BreakdownGain voltage Gain temperature Operating Operating temperature
3 × 3 mm2 Value 900 2 3 ˆ 3 mm 320–900 nm 900 35% 320–900 nm Vbr + 1.4 V 35% 65 V +± 1.4 10 V V Vbr 6 652.8 V× ˘10 10 V ˆ 106°C −202.8 °C–40
´20 ˝ C–40 ˝ C
The Personal-Spect prototype contained a 3 × 3 × 20 mm3 Ce-doped Gd–Al–Ga–garnet (Ce:GAGG) crystal, whichprototype was fabricated via the growth method using a seed The Personal-Spect contained a Czochralski 3 ˆ 3 ˆ 20crystal mm3 Ce-doped Gd–Al–Ga–garnet crystal (Furukawa Denshi Co., Tokyo, Japan) and contained approximately 1% Ce. Ce:GAGG a (Ce:GAGG) crystal, which was fabricated via the Czochralski crystal growth method using ahas seed peak wavelength 520–530 nm, whichJapan) fits theand spectral response range of the SiPM. Also, sincehas thisa crystal (FurukawaofDenshi Co., Tokyo, contained approximately 1% Ce. Ce:GAGG −3, its stopping power is high. Furthermore, its decay time material has a density of 6.63 peak wavelength of 520–530 nm,g·cm which fits the spectral response range of the SiPM. Also, since this ´3 conventional (90 ns) ishas shorter thanofthose ofcm the scintillators (CsI:Tl and NaI:Tl) [4]. In order to material a density 6.63 g¨ , its stopping power is high. Furthermore, its decay time (90 ns) maximize the light the Ce:GAGG and to matchand it to the SiPM photosensitive area, the is shorter than thoseoutput of the from conventional scintillators (CsI:Tl NaI:Tl) [4]. In order to maximize the crystal geometry optimized and using Carlo n-particle extended code [5].the Figure 1 shows the light output fromwas the Ce:GAGG to Monte match it to the SiPM photosensitive area, crystal geometry detector head of the Personal-Spect. The SiPM was coupled a Ce:GAGG scintillator covered with was optimized using Monte Carlo n-particle extended code [5].toFigure 1 shows the detector head of the five layers of a white diffusive in order to optimize the scintillation light collection. Personal-Spect. The SiPM was (Teflon) coupled reflector to a Ce:GAGG scintillator covered with five layers of a white The optical coupling was achieved byoptimize using optical grease (n = light 1.465)collection. as the coupling medium. The diffusive (Teflon) reflector in order to the scintillation The optical coupling coupled detector was covered and sealed inside an Al tube to preventThe background noise from was achieved by using optical grease (n = 1.465) as the coupling medium. coupled detector was external covered light. and sealed inside an Al tube to prevent background noise from external light.
Figure × 3 mm22 Figure 1. Structure Structure of of Ce-doped Ce-doped Gd–Al–Ga–garnet Gd–Al–Ga–garnet (Ce:GAGG) (Ce:GAGG) detector detector head (left); 3 ˆ 3 3 Hamamatsu crystal Hamamatsu S12572-100C S12572-100C SiPM and 3 ׈33׈2020mm mmCe:GAGG Ce:GAGG crystal(right). (right).
Figure 2 shows the Personal-Spect structure and the overall spectroscopy system designed in Figure 2 shows the Personal-Spect structure and the overall spectroscopy system designed in this this study. The fabricated Ce:GAGG-SiPM detector and electronics were installed in an Al shielding study. The fabricated Ce:GAGG-SiPM detector and electronics were installed in an Al shielding box to box to prevent background noise from external light. prevent background noise from external light.
Sensors 2016, 16, 919 Sensors 2016, 16, 919
3 of 6 3 of 6
Figure 2. (A) Personal-Spect structure and (B) overall spectroscopy system. Figure 2. (A) Personal-Spect structure and (B) overall spectroscopy system.
Previously, charge-sensitive preamplifiers that convert scintillation detector charge signals into voltage signals, as shaping amplifiers that amplify signals and simultaneously signals,as well as well as shaping amplifiers thatconverted amplify voltage converted voltage signals and shape them intoshape Gaussian forms, were designed use designed in signal processing in γ-ray spectrometry [2]. simultaneously them into Gaussian forms,for were for use in signal processing in γ-ray However, such[2]. signal processing cause signal attenuation due to impedance mismatch, spectrometry However, suchprocedures signal processing procedures cause signal attenuation due to decrease themismatch, signal-to-noise ratio through the amplification and shaping and and causeshaping output impedance decrease the signal-to-noise ratio through the processes, amplification signal loss and due cause to detected overlap with the dead signal time. overlap with the dead time. processes, outputsignal signal loss due to detected Therefore, in this study, the signal processing stage was excluded due to the high electron amplification gains of SiPMs, and only the driver circuit of the scintillation detector was designed left-hand side side of of Figure Figure 2 shows a 700 mV and built to analyze the output signal of the circuit. The left-hand 22 22 pulse signal with a 2 µs μs pulse pulse width width that that was was generated generated by by aa Na source and measured using the driver circuit. voltage supply supply consisted consistedofofa amain main power supply a voltage-boosting The voltage power supply and and a voltage-boosting unit. unit. The The voltage-boosting unit designed was designed to increase an voltage input voltage 73V, V which ˘ 1.5 V, voltage-boosting unit was to increase an input of 7.2 Vofup7.2 toV73up V ±to1.5 is which is the input voltage of the SiPM. A Li-poly battery V, 850 mAh) was usedasasthe themain main power the input voltage of the SiPM. A Li-poly battery (7.4 (7.4 V, 850 mAh) was used supply with converter modulemodule (UltraVolt, Inc., Ronkonkoma, NY, USA). This witha DC–DC a DC–DC converter (UltraVolt, Inc., Ronkonkoma, NY,Personal-Spect USA). This 2 2 had dimensions of dimensions 5.0 ˆ 7.0 cm , and it cm could run for more 3 h than with3an average power Personal-Spect had of 5.0 × 7.0 , and it could run than for more h with an average consumption of 2.7 W. power consumption of 2.7 W. using a tablet computer, which constituted constituted the the display display unit, unit, The signal analysis was performed using and a miniature 4906-channel K102 Multichannel Analyser Analyser (Kromek (Kromek Ltd., Ltd., Sedgefield, Sedgefield, UK). UK). In the experiments described herein, a standard disc-type radiation source was used to evaluate 133Ba: 0.356 MeV, 22Na: 0.511 MeV, 22 0.511 MeV and the Personal-Spect. Personal-Spect. Several standard standard γ-ray γ-ray sources sourceswere wereused: used:133 137 60 1.27 MeV, Cs: 0.662 0.662 MeV, MeV, and 60Co: Co:1.17 1.17MeV MeVand and1.33 1.33MeV MeV(Spectrum (SpectrumTechniques, Techniques, Oak Oak Ridge, Ridge, TN, TN, MeV, 137Cs: USA). Each source had an activity of 1 μCi and was placed 10 mm from the surface of the developed device, and the live time was set to 300 s for each of the experiments.
Sensors 2016, 16, 919
4 of 6
USA). Each source had an activity of 1 µCi and was placed 10 mm from the surface of the developed device, and16, the Sensors 2016, 919live time was set to 300 s for each of the experiments. 4 of 6 Sensors 2016, 16, 919
4 of 6
3. Resultsand andDiscussion Discussion 3. Results 3. Results and Discussion Figure Figure 33 shows shows the the energy energy spectra spectra obtained obtained from from the the Personal-Spect. Personal-Spect. The The 0.356 0.356 MeV MeV peak peak of of the the 3exhibits shows the energy spectra obtained fromFor the the Personal-Spect. The 0.356 MeV of the 133 22 Napeak 133Ba 22 BaFigure source an energy resolution of 13.5%. 0.511 MeV peak in the spectrum source exhibits an energy resolution of 13.5%. For the 0.511 MeV peak in the Na spectrum that 133Ba source exhibits an energy resolution of 13.5%. For the 0.511 MeV peak in the 22Na spectrum that that generated through positron annihilation during β decay,ananenergy energyresolution resolution of of 6.9% 6.9% was was was generated through positron annihilation during β decay, was was generated through positron annihilation duringyielded β decay, an energy resolution of 6.9% was 137 137 measured. The 0.662 MeV peak in the Cs spectrum an energy resolution of 5.8%, while measured. The 0.662 MeV peak in the 137 Cs spectrum yielded an energy resolution of 5.8%, while for for measured. The 0.662 MeV peak in was the obtained Cs spectrum an1.33 energy ofat5.8%, whileand for 60 60Co, Co, an an energy energy resolution of 2.3% 2.3% at an an yielded energy of of MeV,resolution and peaks 1.17 MeV resolution of was obtained at energy 1.33 MeV, and peaks at 1.17 MeV and 60Co, an energy resolution of 2.3% was obtained at an energy of 1.33 MeV, and peaks at 1.17 MeV and 1.33 1.33 MeV MeV are are clearly clearly observable observable[6]. [6]. 1.33 MeV are clearly observable [6].
133Ba (0.356 MeV), 22Na 137Cs 22 (0.511 137(0.662 Figure 3. 3. Measured energy spectra for MeV), MeV), and Figure Measuredenergy energyspectra spectrafor for133133 (0.356 MeV), (0.511 MeV), Cs (0.662 MeV), 22Na Na 137Cs (0.662 Figure 3. Measured BaBa (0.356 MeV), (0.511 MeV), MeV), and 60Co60 (1.33 MeV). and Co (1.33 MeV). 60Co (1.33 MeV).
Figure 4 depicts the Personal-Spect energy calibration line, which was used to evaluate the Figure 4 depicts the Personal-Spect energy calibration line, which was used to evaluate the linearity of the γ-ray energy. The points on this graph were obtained and the line was estimated linearity of ofthe theγ-ray γ-rayenergy. energy. The points graph obtained andline thewas lineestimated was estimated linearity The points on on thisthis graph werewere obtained and the based 133Ba, 22Na, 137Cs, and 60Co photopeak energies and channels. based on the measured 133 Ba, 22 137 133Na, based the measured Ba, 22 Na, Cs, 60 and Co photopeak energies and channels. on the on measured Cs,137 and Co 60photopeak energies and channels.
133Ba, 22Na, 137Cs, and 60Co photopeak energies and channels. Figure 4. Linear fit of measured 133 22Na, 137 137Cs, Figure4.4.Linear Linearfitfitofofmeasured measured133 Ba, Ba, 22 Figure Cs,and and6060Co Cophotopeak photopeakenergies energiesand andchannels. channels.
The fit shown in Figure 4 confirms the linearity of the relationship between the γ-ray energy The fit shown in Figure 4 confirms the linearity of the relationship between the γ-ray energy and peak channel number for the energy range from 0.356 MeV to 1.33 MeV. The R22 value of the and peak channel number for the energy range from 0.356 MeV to 1.33 MeV. The R value of the linear relationship was found to be 0.9924. linear relationship was found to be 0.9924.
Sensors 2016, 16, 919
5 of 6
The fit shown in Figure 4 confirms the linearity of the relationship between the γ-ray energy and peak channel number for the energy range from 0.356 MeV to 1.33 MeV. The R2 value of the linear Sensors 2016, 16, 919 5 of 6 relationship Sensors 2016, 16,was 919 found to be 0.9924. 5 of 6 Figure 5 depicts the Personal-Spect detection efficiency curve, which was used to evaluate Figure 5 depicts the Personal-Spect detection efficiency curve, which was used to evaluate the the Figurepeak 5 depicts the Personal-Spect detection efficiency curve, which used to evaluate the full-energy efficiency (FEPE) forfor γ-rays of various energies. The FEPE is was theisefficiency with which full-energy peak efficiency (FEPE) γ-rays of various energies. The FEPE the efficiency with full-energy peak efficiency (FEPE) for γ-rays of various energies. The FEPE is the efficiency with γ-rays of a given produce full-energy peak pulses, rather than pulses otherofsizes. which γ-rays of aenergy given energy produce full-energy peak pulses, rather than of pulses other sizes. which γ-rays of a given energy produce full-energy peak pulses, rather than pulses of other sizes.
Figure 5. 5. Full-energy peak efficiency efficiency in in the the 0.356–1.3 0.356–1.3 MeV MeV energy energy range. range. Figure Full-energy peak Figure 5. Full-energy peak efficiency in the 0.356–1.3 MeV energy range.
The efficiency curve in Figure 5 shows a relatively linear decrease within the high-energy range The efficiency curve in 55 shows aa relatively linear within the range efficiency curve in Figure Figure shows relatively linear decrease decrease theofhigh-energy high-energy fromThe 0.356 MeV to 1.33 MeV. Clearly, to be useful, the detector must bewithin capable absorbing arange large from MeV to MeV. Clearly, to the detector must capable of absorbing a large from 0.356 0.356 MeV to 1.33 1.33 MeV.This Clearly, tobe beuseful, useful, detector must be capableof ofsuitable absorbing large fraction of the γ-ray energy. capability can be the realized by using abedetector sizea or by fraction of the γ-ray energy. This capability can be realized by using a detector of suitable size or by fraction of the γ-ray energy. This capability can be realized by using a detector of suitable size or by choosing a scintillator material with a sufficiently high Z [7]. choosing aa scintillator material with aa sufficiently high ZZ [7]. choosing scintillator material with sufficiently high [7]. 133 22 137 Figure 6 depicts the γ-ray energy spectrum for a mixed source containing Ba, Cs, 133 22Na, 22 Na, 137and Figure depicts the the γ-ray γ-rayenergy energyspectrum spectrumfor fora amixed mixedsource source containing Cs, 133Ba,Ba, 137Cs, and Figure 66results depicts containing Na, 60Co [8]. The presented in this figure confirm that γ-ray spectrometry analysis is possible for a 60 Co [8]. The results presented in this figure confirm that γ-ray spectrometry analysis is possible and 60Co [8]. The results in this figure confirm γ-ray spectrometry mixed source based presented on the γ-ray energy peaks in thethat range from 0.3 MeV to analysis 1.3 MeV.is possible for a for a mixed source onγ-ray the γ-ray energy the range 0.3 MeV 1.3 MeV. mixed source basedbased on the energy peakspeaks in theinrange fromfrom 0.3 MeV to 1.3toMeV.
Figure 6. Energy spectra emitted from 133Ba (0.303 MeV and 356 MeV), 22Na (0.511 MeV), 137Cs (0.662 133Ba (0.303 MeV and 356 MeV), 22Na22(0.511 MeV), 137Cs 137 Figureand 6. Energy spectra emitted 60Co (1.33 Figure 6. Energy spectra emittedfrom from 133 Ba (0.303 MeV and 356 MeV), Na (0.511 MeV), (0.662 Cs MeV). MeV), 60Co (1.33 60 MeV). MeV), and (0.662 MeV), and Co (1.33 MeV).
4. Conclusions 4. Conclusions 4. Conclusions A personal γ spectrometer was fabricated and evaluated to assess its usability in radiation A personal personal γ spectrometer spectrometer was is fabricated andrequiring evaluated to 2.7 assess its usability usability in radiation A γ was fabricated and evaluated to assess its monitoring applications. This device low-power, only W from a single in 7.4radiation V input, monitoring applications. This device is low-power, requiring only 2.7 W from a single 7.4 V input, 2 monitoring applications. Thisthe device is low-power, requiringofonly a single 7.4 V has a mass of 340 g (including battery), and has dimensions 5.0 ×2.7 7.0W cmfrom . It can be accessed 2. It can be accessed has a mass of 340 g (including the battery), and has dimensions of 5.0 × 7.0 cm 2 remotely a mobile device. This device exhibitedand higher resolutions, confirmed by the input, hasvia a mass of 340 g (including the battery), has energy dimensions of 5.0 ˆas7.0 cm . It can be remotely via a mobile device. This device exhibited higher energy resolutions, as confirmed by the spectra obtained fora several standard sources. The higher featuresenergy of the Personal-Spect were accessed remotely via mobile device. Thisγdevice exhibited resolutions, as confirmed spectra obtained several standard γ sources. The features of the were characterized, and for the best full-width at half-maximum energy resolution wasPersonal-Spect found to be 5.8% at characterized, and the best full-width at half-maximum energy resolution was found to be 5.8% 662 keV. Thus, the Personal-Spect is expected to be applicable in fields such as HLS and ERMS asat a 662 keV. Thus, the Personal-Spect is expected to be applicable in fields such as HLS and ERMS as a personal γ spectrometer and to replace expensive semiconductor detectors. personal γ spectrometer and to replace expensive semiconductor detectors. Acknowledgments: This work was supported by the 2015 Research Fund of Myongji University. Acknowledgments: This work was supported by the 2015 Research Fund of Myongji University.
Sensors 2016, 16, 919
6 of 6
by the spectra obtained for several standard γ sources. The features of the Personal-Spect were characterized, and the best full-width at half-maximum energy resolution was found to be 5.8% at 662 keV. Thus, the Personal-Spect is expected to be applicable in fields such as HLS and ERMS as a personal γ spectrometer and to replace expensive semiconductor detectors. Acknowledgments: This work was supported by the 2015 Research Fund of Myongji University. Author Contributions: Hye Min Park made substantial contributions in design of the gamma spectrometer, all experiments, data analysis, results interpretation and manuscript preparation. Koan Sik Joo made significant contributions in experimental design, data analysis, results interpretation, and manuscript preparation. Conflicts of Interest: The authors declare no conflict of interest.
References 1.
2.
3. 4.
5. 6. 7. 8.
Sordo, S.D.; Abbene, L.; Caroli, E.; Mancini, A.M.; Zappettini, A.; Ubertini, P. Progress in the development of CdTe and CdZnTe semiconductor radiation detectors for astrophysical and medical applications. Sensors 2009, 9, 3491–3526. [CrossRef] [PubMed] Kim, H.S.; Ha, J.H.; Park, S.H.; Cho, S.Y.; Kim, Y.K. Fabrication and performance characteristics of a CsI(Tl)/PIN diode radiation sensor for industrial applications. Nucl. Instrum. Appl. Radiat. Isot. 2009, 67, 1463–1465. [CrossRef] [PubMed] Sanaei, B.; Baei, M.T.; Sayyed-Alangi, S.Z. Characterization of a new silicon photomultiplier in comparison with a conventional photomultiplier tube. J. Mod. Phys. 2015, 6, 425–433. [CrossRef] Yeom, J.Y.; Yamamoto, S.; Derenzo, S.E.; Spanoudaki, V.C.; Kamada, K.; Endo, T.; Levin, C.S. First performance results of Ce:GAGG scintillation crystals with silicon photomultipliers. IEEE Trans. Nucl. Sci. 2013, 60, 988–992. [CrossRef] Park, H.M.; Joo, K.S. Performance characteristics of a silicon photomultiplier based compact radiation detector for Homeland Security applications. Nucl. Instrum. Meth. Phys. Res. A 2015, 781, 1–5. [CrossRef] David, S.; Georgiou, M.; Fysikopoulos, E.; Loudos, G. Evaluation of a SiPM array coupled to a Gd3Al2Ga3O12:Ce (GAGG:Ce) discrete scintillator. Phys. Med. 2015, 31, 763–766. [CrossRef] [PubMed] Canberra. Gamma and X-ray Detection. Available online: http://www.canberra.com/literature/ fundamental-principles/pdf/Gamma-Xray-Detection.pdf (accessed on 20 June 2016). Osovizky, A.; Ginzburg, D.; Manor, A.; Seif, R.; Ghelman, M.; Cohen-Zada, I.; Ellenbogen, M.; Bronfenmakher, V.; Pushkarsky, V.; Gonen, E.; et al. SENTIRAD—An innovative personal radiation detector based on a scintillation detector and a silicon photomultiplier. Nucl. Instrum. Meth. Phys. Res. A 2011, 652, 41–44. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
Development and Performance Characteristics of Personal Gamma Spectrometer for Radiation Monitoring Applications Hye Min Park and Koan Sik Joo * Department of Physics, University of Myongji, Yongin 449-728, Korea; [email protected] * Correspondence: [email protected]; Tel.: +82-10-9288-6436 Academic Editor: Vittorio M. N. Passaro Received: 9 April 2016; Accepted: 17 June 2016; Published: 21 June 2016
Abstract: In this study, a personal gamma (γ) spectrometer was developed for use in applications in various fields, such as homeland security and environmental radiation monitoring systems. The prototype consisted of a 3 ˆ 3 ˆ 20 mm3 Ce-doped Gd–Al–Ga–garnet (Ce:GAGG) crystal that was coupled to a Si photomultiplier (SiPM) to measure γ radiation. The γ spectrometer could be accessed remotely via a mobile device. At room temperature, the implemented Ce:GAGG-SiPM spectrometer achieved energy resolutions of 13.5%, 6.9%, 5.8%, and 2.3% for 133 Ba at 0.356 MeV, 22 Na at 0.511 MeV, 137 Cs at 0.662 MeV, and 60 Co at 1.33 MeV, respectively. It consumed only about 2.7 W of power, had a mass of just 340 g (including the battery), and measured only 5.0 ˆ 7.0 cm2 . Keywords: Ce-doped Gd–Al–Ga–garnet (Ce:GAGG); Si photomultiplier (SiPM); spectrometer; energy resolution; homeland security (HLS); environmental radiation monitoring system (ERMS)
1. Introduction Gamma (γ)-ray spectrometry analysis methods can be classified based on whether they involve direct or indirect detection, according to the manner in which the γ-rays are converted into electrical signals. The direct detection methods employ semiconductor devices that generate electrical signals immediately, with no intermediate stages, through γ-ray absorption. The indirect detection methods involve scintillation detection systems; in such a system, a photoelectric device detects electrical signals generated by photons with wavelengths in the visible range that are produced through interactions between the γ-rays and the scintillator. CdZnTe, CdTe, and HgI2 are primarily used in room temperature semiconductor devices employing direct detection methods [1]. These materials are preferred in γ-ray spectrometry analysis because they yield energy resolutions superior to those achievable with scintillator crystals. However, the maintenance of these materials is more expensive than that of scintillator crystals, and their detection efficiencies and physical rigidities are very low [2]. Currently, compound semiconductor materials are also more expensive than scintillator crystals. NaI(Tl), CsI(Tl), and LYSO(Ce) are the primary scintillator crystals used in indirect detection. These materials yield energy resolutions lower than those of semiconductors, but they are more sensitive. Because of the specific characteristics and limitations of each of these types of materials and the corresponding analysis methods, further development of detectors that combine scintillators and semiconductors is necessary and ongoing [3]. Thus, scintillator detectors remain promising candidates for use in homeland security (HLS) and environmental radiation monitoring systems (ERMSs). In this study, a personal γ spectrometer was developed, and its feasibility was evaluated by analyzing its energy resolution using standard γ-ray sources. The device described herein, called the “Personal-Spect”, essentially consists of a scintillation detector, signal processing electronics, a voltage supply, signal analysis components, and display units. Sensors 2016, 16, 919; doi:10.3390/s16060919
www.mdpi.com/journal/sensors
Sensors 2016, 16, 919
2 of 6
“Personal-Spect”, essentially consists of a scintillation detector, signal processing electronics, a voltage supply, signal analysis components, and display units. Sensors 2016, 16, 919
2 of 6
2. Materials and Methods An S12572-100C (photo-sensitive area: 3 × 3 mm2, Hamamatsu Photonics Co., Hamamatsu, 2. Materials and Methods Japan) Hamamatsu single Si photomultiplier (SiPM)2was used in the Personal-Spect to minimize the An S12572-100C (photo-sensitive area: 3 ˆ 3 mm , Hamamatsu Photonics Co., Hamamatsu, Japan) overall system dimensions (Table 1). Hamamatsu single Si photomultiplier (SiPM) was used in the Personal-Spect to minimize the overall system dimensions (Table 1). Table 1. Specifications of Si photomultiplier (SiPM) used in this study. Table 1. Specifications of Si photomultiplier (SiPM) used in this study. Parameter Value
Photosensitive area Parameter Number of pixels Photosensitive area range Spectral response Number of pixels Peak photon detection efficiency (at 450 nm) Spectral response range Bias voltage Peak photon detection efficiency (at 450 nm) Breakdown Bias voltage voltage BreakdownGain voltage Gain temperature Operating Operating temperature
3 × 3 mm2 Value 900 2 3 ˆ 3 mm 320–900 nm 900 35% 320–900 nm Vbr + 1.4 V 35% 65 V +± 1.4 10 V V Vbr 6 652.8 V× ˘10 10 V ˆ 106°C −202.8 °C–40
´20 ˝ C–40 ˝ C
The Personal-Spect prototype contained a 3 × 3 × 20 mm3 Ce-doped Gd–Al–Ga–garnet (Ce:GAGG) crystal, whichprototype was fabricated via the growth method using a seed The Personal-Spect contained a Czochralski 3 ˆ 3 ˆ 20crystal mm3 Ce-doped Gd–Al–Ga–garnet crystal (Furukawa Denshi Co., Tokyo, Japan) and contained approximately 1% Ce. Ce:GAGG a (Ce:GAGG) crystal, which was fabricated via the Czochralski crystal growth method using ahas seed peak wavelength 520–530 nm, whichJapan) fits theand spectral response range of the SiPM. Also, sincehas thisa crystal (FurukawaofDenshi Co., Tokyo, contained approximately 1% Ce. Ce:GAGG −3, its stopping power is high. Furthermore, its decay time material has a density of 6.63 peak wavelength of 520–530 nm,g·cm which fits the spectral response range of the SiPM. Also, since this ´3 conventional (90 ns) ishas shorter thanofthose ofcm the scintillators (CsI:Tl and NaI:Tl) [4]. In order to material a density 6.63 g¨ , its stopping power is high. Furthermore, its decay time (90 ns) maximize the light the Ce:GAGG and to matchand it to the SiPM photosensitive area, the is shorter than thoseoutput of the from conventional scintillators (CsI:Tl NaI:Tl) [4]. In order to maximize the crystal geometry optimized and using Carlo n-particle extended code [5].the Figure 1 shows the light output fromwas the Ce:GAGG to Monte match it to the SiPM photosensitive area, crystal geometry detector head of the Personal-Spect. The SiPM was coupled a Ce:GAGG scintillator covered with was optimized using Monte Carlo n-particle extended code [5].toFigure 1 shows the detector head of the five layers of a white diffusive in order to optimize the scintillation light collection. Personal-Spect. The SiPM was (Teflon) coupled reflector to a Ce:GAGG scintillator covered with five layers of a white The optical coupling was achieved byoptimize using optical grease (n = light 1.465)collection. as the coupling medium. The diffusive (Teflon) reflector in order to the scintillation The optical coupling coupled detector was covered and sealed inside an Al tube to preventThe background noise from was achieved by using optical grease (n = 1.465) as the coupling medium. coupled detector was external covered light. and sealed inside an Al tube to prevent background noise from external light.
Figure × 3 mm22 Figure 1. Structure Structure of of Ce-doped Ce-doped Gd–Al–Ga–garnet Gd–Al–Ga–garnet (Ce:GAGG) (Ce:GAGG) detector detector head (left); 3 ˆ 3 3 Hamamatsu crystal Hamamatsu S12572-100C S12572-100C SiPM and 3 ׈33׈2020mm mmCe:GAGG Ce:GAGG crystal(right). (right).
Figure 2 shows the Personal-Spect structure and the overall spectroscopy system designed in Figure 2 shows the Personal-Spect structure and the overall spectroscopy system designed in this this study. The fabricated Ce:GAGG-SiPM detector and electronics were installed in an Al shielding study. The fabricated Ce:GAGG-SiPM detector and electronics were installed in an Al shielding box to box to prevent background noise from external light. prevent background noise from external light.
Sensors 2016, 16, 919 Sensors 2016, 16, 919
3 of 6 3 of 6
Figure 2. (A) Personal-Spect structure and (B) overall spectroscopy system. Figure 2. (A) Personal-Spect structure and (B) overall spectroscopy system.
Previously, charge-sensitive preamplifiers that convert scintillation detector charge signals into voltage signals, as shaping amplifiers that amplify signals and simultaneously signals,as well as well as shaping amplifiers thatconverted amplify voltage converted voltage signals and shape them intoshape Gaussian forms, were designed use designed in signal processing in γ-ray spectrometry [2]. simultaneously them into Gaussian forms,for were for use in signal processing in γ-ray However, such[2]. signal processing cause signal attenuation due to impedance mismatch, spectrometry However, suchprocedures signal processing procedures cause signal attenuation due to decrease themismatch, signal-to-noise ratio through the amplification and shaping and and causeshaping output impedance decrease the signal-to-noise ratio through the processes, amplification signal loss and due cause to detected overlap with the dead signal time. overlap with the dead time. processes, outputsignal signal loss due to detected Therefore, in this study, the signal processing stage was excluded due to the high electron amplification gains of SiPMs, and only the driver circuit of the scintillation detector was designed left-hand side side of of Figure Figure 2 shows a 700 mV and built to analyze the output signal of the circuit. The left-hand 22 22 pulse signal with a 2 µs μs pulse pulse width width that that was was generated generated by by aa Na source and measured using the driver circuit. voltage supply supply consisted consistedofofa amain main power supply a voltage-boosting The voltage power supply and and a voltage-boosting unit. unit. The The voltage-boosting unit designed was designed to increase an voltage input voltage 73V, V which ˘ 1.5 V, voltage-boosting unit was to increase an input of 7.2 Vofup7.2 toV73up V ±to1.5 is which is the input voltage of the SiPM. A Li-poly battery V, 850 mAh) was usedasasthe themain main power the input voltage of the SiPM. A Li-poly battery (7.4 (7.4 V, 850 mAh) was used supply with converter modulemodule (UltraVolt, Inc., Ronkonkoma, NY, USA). This witha DC–DC a DC–DC converter (UltraVolt, Inc., Ronkonkoma, NY,Personal-Spect USA). This 2 2 had dimensions of dimensions 5.0 ˆ 7.0 cm , and it cm could run for more 3 h than with3an average power Personal-Spect had of 5.0 × 7.0 , and it could run than for more h with an average consumption of 2.7 W. power consumption of 2.7 W. using a tablet computer, which constituted constituted the the display display unit, unit, The signal analysis was performed using and a miniature 4906-channel K102 Multichannel Analyser Analyser (Kromek (Kromek Ltd., Ltd., Sedgefield, Sedgefield, UK). UK). In the experiments described herein, a standard disc-type radiation source was used to evaluate 133Ba: 0.356 MeV, 22Na: 0.511 MeV, 22 0.511 MeV and the Personal-Spect. Personal-Spect. Several standard standard γ-ray γ-ray sources sourceswere wereused: used:133 137 60 1.27 MeV, Cs: 0.662 0.662 MeV, MeV, and 60Co: Co:1.17 1.17MeV MeVand and1.33 1.33MeV MeV(Spectrum (SpectrumTechniques, Techniques, Oak Oak Ridge, Ridge, TN, TN, MeV, 137Cs: USA). Each source had an activity of 1 μCi and was placed 10 mm from the surface of the developed device, and the live time was set to 300 s for each of the experiments.
Sensors 2016, 16, 919
4 of 6
USA). Each source had an activity of 1 µCi and was placed 10 mm from the surface of the developed device, and16, the Sensors 2016, 919live time was set to 300 s for each of the experiments. 4 of 6 Sensors 2016, 16, 919
4 of 6
3. Resultsand andDiscussion Discussion 3. Results 3. Results and Discussion Figure Figure 33 shows shows the the energy energy spectra spectra obtained obtained from from the the Personal-Spect. Personal-Spect. The The 0.356 0.356 MeV MeV peak peak of of the the 3exhibits shows the energy spectra obtained fromFor the the Personal-Spect. The 0.356 MeV of the 133 22 Napeak 133Ba 22 BaFigure source an energy resolution of 13.5%. 0.511 MeV peak in the spectrum source exhibits an energy resolution of 13.5%. For the 0.511 MeV peak in the Na spectrum that 133Ba source exhibits an energy resolution of 13.5%. For the 0.511 MeV peak in the 22Na spectrum that that generated through positron annihilation during β decay,ananenergy energyresolution resolution of of 6.9% 6.9% was was was generated through positron annihilation during β decay, was was generated through positron annihilation duringyielded β decay, an energy resolution of 6.9% was 137 137 measured. The 0.662 MeV peak in the Cs spectrum an energy resolution of 5.8%, while measured. The 0.662 MeV peak in the 137 Cs spectrum yielded an energy resolution of 5.8%, while for for measured. The 0.662 MeV peak in was the obtained Cs spectrum an1.33 energy ofat5.8%, whileand for 60 60Co, Co, an an energy energy resolution of 2.3% 2.3% at an an yielded energy of of MeV,resolution and peaks 1.17 MeV resolution of was obtained at energy 1.33 MeV, and peaks at 1.17 MeV and 60Co, an energy resolution of 2.3% was obtained at an energy of 1.33 MeV, and peaks at 1.17 MeV and 1.33 1.33 MeV MeV are are clearly clearly observable observable[6]. [6]. 1.33 MeV are clearly observable [6].
133Ba (0.356 MeV), 22Na 137Cs 22 (0.511 137(0.662 Figure 3. 3. Measured energy spectra for MeV), MeV), and Figure Measuredenergy energyspectra spectrafor for133133 (0.356 MeV), (0.511 MeV), Cs (0.662 MeV), 22Na Na 137Cs (0.662 Figure 3. Measured BaBa (0.356 MeV), (0.511 MeV), MeV), and 60Co60 (1.33 MeV). and Co (1.33 MeV). 60Co (1.33 MeV).
Figure 4 depicts the Personal-Spect energy calibration line, which was used to evaluate the Figure 4 depicts the Personal-Spect energy calibration line, which was used to evaluate the linearity of the γ-ray energy. The points on this graph were obtained and the line was estimated linearity of ofthe theγ-ray γ-rayenergy. energy. The points graph obtained andline thewas lineestimated was estimated linearity The points on on thisthis graph werewere obtained and the based 133Ba, 22Na, 137Cs, and 60Co photopeak energies and channels. based on the measured 133 Ba, 22 137 133Na, based the measured Ba, 22 Na, Cs, 60 and Co photopeak energies and channels. on the on measured Cs,137 and Co 60photopeak energies and channels.
133Ba, 22Na, 137Cs, and 60Co photopeak energies and channels. Figure 4. Linear fit of measured 133 22Na, 137 137Cs, Figure4.4.Linear Linearfitfitofofmeasured measured133 Ba, Ba, 22 Figure Cs,and and6060Co Cophotopeak photopeakenergies energiesand andchannels. channels.
The fit shown in Figure 4 confirms the linearity of the relationship between the γ-ray energy The fit shown in Figure 4 confirms the linearity of the relationship between the γ-ray energy and peak channel number for the energy range from 0.356 MeV to 1.33 MeV. The R22 value of the and peak channel number for the energy range from 0.356 MeV to 1.33 MeV. The R value of the linear relationship was found to be 0.9924. linear relationship was found to be 0.9924.
Sensors 2016, 16, 919
5 of 6
The fit shown in Figure 4 confirms the linearity of the relationship between the γ-ray energy and peak channel number for the energy range from 0.356 MeV to 1.33 MeV. The R2 value of the linear Sensors 2016, 16, 919 5 of 6 relationship Sensors 2016, 16,was 919 found to be 0.9924. 5 of 6 Figure 5 depicts the Personal-Spect detection efficiency curve, which was used to evaluate Figure 5 depicts the Personal-Spect detection efficiency curve, which was used to evaluate the the Figurepeak 5 depicts the Personal-Spect detection efficiency curve, which used to evaluate the full-energy efficiency (FEPE) forfor γ-rays of various energies. The FEPE is was theisefficiency with which full-energy peak efficiency (FEPE) γ-rays of various energies. The FEPE the efficiency with full-energy peak efficiency (FEPE) for γ-rays of various energies. The FEPE is the efficiency with γ-rays of a given produce full-energy peak pulses, rather than pulses otherofsizes. which γ-rays of aenergy given energy produce full-energy peak pulses, rather than of pulses other sizes. which γ-rays of a given energy produce full-energy peak pulses, rather than pulses of other sizes.
Figure 5. 5. Full-energy peak efficiency efficiency in in the the 0.356–1.3 0.356–1.3 MeV MeV energy energy range. range. Figure Full-energy peak Figure 5. Full-energy peak efficiency in the 0.356–1.3 MeV energy range.
The efficiency curve in Figure 5 shows a relatively linear decrease within the high-energy range The efficiency curve in 55 shows aa relatively linear within the range efficiency curve in Figure Figure shows relatively linear decrease decrease theofhigh-energy high-energy fromThe 0.356 MeV to 1.33 MeV. Clearly, to be useful, the detector must bewithin capable absorbing arange large from MeV to MeV. Clearly, to the detector must capable of absorbing a large from 0.356 0.356 MeV to 1.33 1.33 MeV.This Clearly, tobe beuseful, useful, detector must be capableof ofsuitable absorbing large fraction of the γ-ray energy. capability can be the realized by using abedetector sizea or by fraction of the γ-ray energy. This capability can be realized by using a detector of suitable size or by fraction of the γ-ray energy. This capability can be realized by using a detector of suitable size or by choosing a scintillator material with a sufficiently high Z [7]. choosing aa scintillator material with aa sufficiently high ZZ [7]. choosing scintillator material with sufficiently high [7]. 133 22 137 Figure 6 depicts the γ-ray energy spectrum for a mixed source containing Ba, Cs, 133 22Na, 22 Na, 137and Figure depicts the the γ-ray γ-rayenergy energyspectrum spectrumfor fora amixed mixedsource source containing Cs, 133Ba,Ba, 137Cs, and Figure 66results depicts containing Na, 60Co [8]. The presented in this figure confirm that γ-ray spectrometry analysis is possible for a 60 Co [8]. The results presented in this figure confirm that γ-ray spectrometry analysis is possible and 60Co [8]. The results in this figure confirm γ-ray spectrometry mixed source based presented on the γ-ray energy peaks in thethat range from 0.3 MeV to analysis 1.3 MeV.is possible for a for a mixed source onγ-ray the γ-ray energy the range 0.3 MeV 1.3 MeV. mixed source basedbased on the energy peakspeaks in theinrange fromfrom 0.3 MeV to 1.3toMeV.
Figure 6. Energy spectra emitted from 133Ba (0.303 MeV and 356 MeV), 22Na (0.511 MeV), 137Cs (0.662 133Ba (0.303 MeV and 356 MeV), 22Na22(0.511 MeV), 137Cs 137 Figureand 6. Energy spectra emitted 60Co (1.33 Figure 6. Energy spectra emittedfrom from 133 Ba (0.303 MeV and 356 MeV), Na (0.511 MeV), (0.662 Cs MeV). MeV), 60Co (1.33 60 MeV). MeV), and (0.662 MeV), and Co (1.33 MeV).
4. Conclusions 4. Conclusions 4. Conclusions A personal γ spectrometer was fabricated and evaluated to assess its usability in radiation A personal personal γ spectrometer spectrometer was is fabricated andrequiring evaluated to 2.7 assess its usability usability in radiation A γ was fabricated and evaluated to assess its monitoring applications. This device low-power, only W from a single in 7.4radiation V input, monitoring applications. This device is low-power, requiring only 2.7 W from a single 7.4 V input, 2 monitoring applications. Thisthe device is low-power, requiringofonly a single 7.4 V has a mass of 340 g (including battery), and has dimensions 5.0 ×2.7 7.0W cmfrom . It can be accessed 2. It can be accessed has a mass of 340 g (including the battery), and has dimensions of 5.0 × 7.0 cm 2 remotely a mobile device. This device exhibitedand higher resolutions, confirmed by the input, hasvia a mass of 340 g (including the battery), has energy dimensions of 5.0 ˆas7.0 cm . It can be remotely via a mobile device. This device exhibited higher energy resolutions, as confirmed by the spectra obtained fora several standard sources. The higher featuresenergy of the Personal-Spect were accessed remotely via mobile device. Thisγdevice exhibited resolutions, as confirmed spectra obtained several standard γ sources. The features of the were characterized, and for the best full-width at half-maximum energy resolution wasPersonal-Spect found to be 5.8% at characterized, and the best full-width at half-maximum energy resolution was found to be 5.8% 662 keV. Thus, the Personal-Spect is expected to be applicable in fields such as HLS and ERMS asat a 662 keV. Thus, the Personal-Spect is expected to be applicable in fields such as HLS and ERMS as a personal γ spectrometer and to replace expensive semiconductor detectors. personal γ spectrometer and to replace expensive semiconductor detectors. Acknowledgments: This work was supported by the 2015 Research Fund of Myongji University. Acknowledgments: This work was supported by the 2015 Research Fund of Myongji University.
Sensors 2016, 16, 919
6 of 6
by the spectra obtained for several standard γ sources. The features of the Personal-Spect were characterized, and the best full-width at half-maximum energy resolution was found to be 5.8% at 662 keV. Thus, the Personal-Spect is expected to be applicable in fields such as HLS and ERMS as a personal γ spectrometer and to replace expensive semiconductor detectors. Acknowledgments: This work was supported by the 2015 Research Fund of Myongji University. Author Contributions: Hye Min Park made substantial contributions in design of the gamma spectrometer, all experiments, data analysis, results interpretation and manuscript preparation. Koan Sik Joo made significant contributions in experimental design, data analysis, results interpretation, and manuscript preparation. Conflicts of Interest: The authors declare no conflict of interest.
References 1.
2.
3. 4.
5. 6. 7. 8.
Sordo, S.D.; Abbene, L.; Caroli, E.; Mancini, A.M.; Zappettini, A.; Ubertini, P. Progress in the development of CdTe and CdZnTe semiconductor radiation detectors for astrophysical and medical applications. Sensors 2009, 9, 3491–3526. [CrossRef] [PubMed] Kim, H.S.; Ha, J.H.; Park, S.H.; Cho, S.Y.; Kim, Y.K. Fabrication and performance characteristics of a CsI(Tl)/PIN diode radiation sensor for industrial applications. Nucl. Instrum. Appl. Radiat. Isot. 2009, 67, 1463–1465. [CrossRef] [PubMed] Sanaei, B.; Baei, M.T.; Sayyed-Alangi, S.Z. Characterization of a new silicon photomultiplier in comparison with a conventional photomultiplier tube. J. Mod. Phys. 2015, 6, 425–433. [CrossRef] Yeom, J.Y.; Yamamoto, S.; Derenzo, S.E.; Spanoudaki, V.C.; Kamada, K.; Endo, T.; Levin, C.S. First performance results of Ce:GAGG scintillation crystals with silicon photomultipliers. IEEE Trans. Nucl. Sci. 2013, 60, 988–992. [CrossRef] Park, H.M.; Joo, K.S. Performance characteristics of a silicon photomultiplier based compact radiation detector for Homeland Security applications. Nucl. Instrum. Meth. Phys. Res. A 2015, 781, 1–5. [CrossRef] David, S.; Georgiou, M.; Fysikopoulos, E.; Loudos, G. Evaluation of a SiPM array coupled to a Gd3Al2Ga3O12:Ce (GAGG:Ce) discrete scintillator. Phys. Med. 2015, 31, 763–766. [CrossRef] [PubMed] Canberra. Gamma and X-ray Detection. Available online: http://www.canberra.com/literature/ fundamental-principles/pdf/Gamma-Xray-Detection.pdf (accessed on 20 June 2016). Osovizky, A.; Ginzburg, D.; Manor, A.; Seif, R.; Ghelman, M.; Cohen-Zada, I.; Ellenbogen, M.; Bronfenmakher, V.; Pushkarsky, V.; Gonen, E.; et al. SENTIRAD—An innovative personal radiation detector based on a scintillation detector and a silicon photomultiplier. Nucl. Instrum. Meth. Phys. Res. A 2011, 652, 41–44. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
