3124
IEEE TRANSACTIONS ON NUCLEAR SCIENCE, VOL. 52, NO. 6, DECEMBER 2005
CdWO4 Crystal in Gamma-Ray Spectrometry M. Moszyn´ski, Fellow, IEEE, M. Balcerzyk, M. Kapusta, Member, IEEE, A. Syntfeld, Member, IEEE, D. Wolski, G. Pausch, J. Stein, and P. Schotanus
Abstract—The properties of CdWO4 (CWO) crystals in gamma spectrometry were studied. Several small samples of 10 10 3 mm size, typically used in CT X-ray detectors, were tested and then compared to the performance of a larger crystal of 20 mm in diameter and 20 mm in height. The light output, energy resolution, and nonproportionality of the CWO response versus gamma-ray energy, were measured and compared with those of a small BGO to discuss further the origin of the intrinsic resolution of pure undoped scintillating crystals. A high light output of 6500 200 phe/MeV and a good energy resolution of 6 6 0 2% for 662 keV gamma rays from a 137 Cs source were measured for the small samples coupled to an XP3212 photomultiplier. Common nonproportionality curves and consequently common intrinsic resolutions of small CWO and BGO suggest that they represent fundamental characteristics of the heavy oxide scintillating material themselves. Index Terms—CdWO4 , energy resolution, non-proportionality, scintillators.
I. INTRODUCTION
N
EW high-Z scintillator detectors are needed to significantly improve current border monitoring instrumentation utilizing plastic detectors, as well as, for the development of highly efficient handheld instrumentation. One of the possible candidates is an undoped CdWO (CWO) crystal [1]–[8]. A high density of 7.9 g/cm and a high Z value of tungsten (74) assure comparable detection efficiency to that of BGO. The emission maximum is situated at 475–480 nm as measured in [2] with 2 cm thick sample. The relative light output corresponds to 38% of that of NaI(Tl), as measured with a bialkali photocathode photomultiplier (PMT) [2]. A good energy resolution of 6.8% was reported for a 2.5 cm in diameter and 1.2 cm thick CWO crystal in [2]. A comparable energy resolution of 6.8% was reported in [3] for a larger, 3 cm 3 cm, CWO crystal. An application of CWO in the gamma spectrometry is limited because of a long main decay time constant of about 15 s of the light pulse [2], [4], which does not allow measurements at high counting rates. However, in the border monitoring equipment, one is looking for traces of gamma rays emitted by Manuscript received October 12, 2004; revised July 26, 2005. This work was supported in part by the International Atomic Energy Agency, Research Contract No. 12596/Nuclear Security Multi-donors Fund and by the Polish Committee for Scientific Research Grant No. 3 T10C 010 26 and SPUB no. 621/E78/SPB/IAEA/0-13/DWM21/2004-2006. M. Moszyn´ski is with the Soltan Institute for Nuclear Studies, PL 05–400 Swierk-Otwock, Poland (e-mail:
[email protected]). M. Balcerzyk, M. Kapusta, A. Syntfeld, and D. Wolski are with the Soltan Institute for Nuclear Studies, PL 05-400 S´wierk-Otwock, Poland. G. Pausch and J. Stein are with Target Systemelectronic GmbH, D-42651 Solingen, Germany (e-mail:
[email protected]). P. Schotanus is with SCIONIX Holland B.V., 3980 CC Bunnik, The Netherlands (e-mail:
[email protected]). Digital Object Identifier 10.1109/TNS.2005.855704
well-shielded radioactive sources. Thus, high rate capabilities of the detectors are less important. Still the main goal is the construction of detectors with high efficiency to rule out innocent alarms on one side and being able to detect a heavily shielded nuclear material on the other side. In this respect, the good temperature stability of the light output of CWO must be pointed out [1], [5]. In [3], the CWO crystal was proposed as the compact gamma ray spectrometer for planetary lander missions. In [4] the pulse shape discrimination capability of CWO was studied in order to detect charged particles in a gamma ray background. Finally, the CWO crystal was also proposed for a detection of thermal neutrons [6], [7]. Since the afterglow of CWO upon an X-ray irradiation is very low [5], [8], the crystal is widely applied for X-ray detection in CT scanners. The aim of this work was to study the properties of CWO in gamma spectrometry. Four small samples of 10 10 3 mm size, typical for use in CT X-ray detectors, were tested. Two of them were characterized by a low afterglow, according to the CT detector requirements. Two others showed about a 5 times higher afterglow. Finally, the performance of a larger crystal with 20 mm diameter and 20 mm height, applicable in the border monitoring, was studied. The light output, energy resolution and nonproportionality of the CWO response were measured versus gamma-ray energy for all studied crystals. The nonproportionality and energy resmm olution characteristics were compared to those of BGO [9] to discuss further the origin of the intrinsic resolution of pure undoped scintillating crystals. A study [9] suggested that both curves represent the fundamental characteristics of the BGO crystal material itself. A collection of new data in this respect was postulated at the Non-proportionality Workshop held in Portland in 2003 during IEEE NSS-MIC conference [10]. II. EXPERIMENTAL DETAILS All the CWO crystals were delivered by Scionix. Four crystals of 10 10 3 mm size had the exit face polished and the others grounded. Two crystals, marked by L, showed very low afterglow, according to the CT detector requirements. The other two, marked by H, exhibited about 5 times higher afterglow according to the manufacturer data. The large crystal of 20 mm in diameter and 20 mm in height was grounded at the side surface, while the exit and top faces were polished. Unfortunately, the large crystal showed a slight yellow color suggesting a selfabsorption of the light [1]. In all measurements the crystals coated with Teflon tape were coupled by silicone grease to a Photonis XP3212 photomultiplier having a high blue sensitivity of 12.1 A/lm blue. A signal
0018-9499/$20.00 © 2005 IEEE
´ SKI et al.: CdWO CRYSTAL IN GAMMA-RAY SPECTROMETRY MOSZYN
3125
TABLE I NUMBER OF PHOTOELECTRONS AND ENERGY RESOLUTION MEASURED WITH CWO CRYSTALS. FOOTNOTE TO TABLE I (a) ACCORDING TO [14], MEASURED FOR THE SAME CRYSTAL BY THE XP2020Q PMT’s WITH THE CALIBRATED QUANTUM EFFICIENCY AND HAMAMATSU PIN PHOTODIODES
Fig. 1. Energy spectra of 662 keV -rays from a Cs source, as recorded with the 10 10 3 mm CWO (bottom panel) and with the 25 mm 30 mm NaI(Tl) (upper panel). An escape peak of K X-rays of Tungstate is seen at the 662 keV peak in the CWO spectrum.
2 2
;
2
from the PMT anode was processed by a modified Canberra 2005E preamplifier and sent to a Tennelec TC 244 spectroscopy amplifier. Most of the measurements were carried out with a 12 s triangle shaping time constant in the amplifier to catch most of the light emitted by CWO. The PC-based multichannel analyzer (Tukan 8 k [11]) was used to record energy spectra. Gaussian functions were fitted to the full energy peaks by using procedures in the analyzer to determine their position and energy resolution. It included also the analysis of complex double peaks characteristic for K X-rays. Although all the small samples of CWO were studied in the course of this work, particularly in respect to their energy resolution and nonproportionality of the light yield, the presented data below are limited to those collected for the CWO 1L only. III. RESULTS A. CWO Crystals in Gamma Spectrometry Fig. 1 shows the energy spectra of 662 keV -rays from a Cs source measured with a small CWO 1L crystal and with a 25 mm in diameter and 30 mm high NaI(Tl).
% for the 662 Note the high energy resolution of keV peak obtained with the CWO, which is comparable to that of NaI(Tl). Note also the comparable photofractions of 26% and 23% in both spectra for the CWO and the NaI(Tl) crystals, respectively, while the volume of the NaI(Tl) crystal is 50 times larger. A shift down of 32 keV KX-ray peak in CWO spectrum in relation to that of NaI(Tl) is seen. It reflects an excess of light at low energies in NaI(Tl) observed in the nonproportionality curve [12], [13] and a reduced light yield in the case of CWO, see Section III.C. The photoelectron number produced by the CWO in the XP3212 PMT, determined by the relation to the single phophe/MeV. Both toelectron peak, was equal to quantities, energy resolution and photoelectron number, are superior over those measured in [9] for small BGO crystals, see Table I. A comparable detection efficiency of CWO and BGO is expected. Note a systematically, about 5% higher photoelectron number measured for the CWO samples with a low afterglow (marked L), in comparison to that of the crystals with a high afterglow (marked H). Although the observed difference is at the limit of the accuracy, this observation was confirmed by series of independent measurements. It suggests that a reduction of afterglow, and more general, slow components of light pulses, may increase the light output of scintillators. % is The energy resolution of the small samples of comparable to that measured typically with good NaI(Tl) crystals, see Fig. 1, and much better than that of BGO. It reflects promising properties of CWO crystals for gamma spectrometry. In the last column of Table I, the light output of CWO crystals is collected. It was determined by a comparative method to that of the BGO crystal. The light output of the reference BGO crystal was measured previously in [14] using XP2020Q PMT’s with calibrated quantum efficiency and Si PIN photodiodes. Note that the peak emission of 490 nm of BGO is comparable to that of CWO. The absolute light output of the CWO
3126
IEEE TRANSACTIONS ON NUCLEAR SCIENCE, VOL. 52, NO. 6, DECEMBER 2005
Fig. 2. Energy spectrum of 662 keV -rays from a Cs source, as recorded with a large CWO crystal at 12 s shaping time constant in a spectroscopy amplifier.
crystal, determined in this way, corresponds to about 70% of that of NaI(Tl) [15]. The study of the large CWO crystal showed its poorer performance. Fig. 2 presents the energy spectrum of 662 keV -rays Cs source, as measured with a large CWO crystal at from a 12 s shaping time constant in the amplifier. % is observed, as comA poorer energy resolution of pared to those of the small samples and to that reported in [3] for even larger crystal. It confirms an expected self-absorption of light because of a light yellow color of the tested crystal. The phe/MeV is lower number of photoelectrons with at about a factor of two than that measured with the small samples. A poorer performance of the larger CWO crystals in this respect was reported in [1]. There is no doubt that further efforts are needed to improve the quality of CWO crystals. In contrast, a high photofraction of 47% was estimated for the 662 keV peak, comparable to that of a 3” 3” NaI(Tl) crystal with 51% [16]. These results and the spectra presented in Fig. 1 demonstrate a very good detection efficiency of CWO crystal. This is further confirmed by the spectra of gamma rays from Bi source and that of the background of the laboratory, a see Fig. 3. High photofractions are observed for the high energy gamma peaks of 1063 keV and 1770 keV. In the background spectrum, well defined peaks of 1460 keV for K, 1764 keV due to Radium series and 2614 keV due to Thorium series, are present. All of them are typically observed in background spectra. All the above results were collected making measurements at 12 s triangle shaping time constant in the amplifier. In turn, Fig. 4 presents the number of photoelectrons versus peaking time in the amplifier as measured for the CWO 1L crystal. It reflects the integral response of the decay of the light pulse. However, the full light is not yet collected even at 12 s shaping corresponding to 26.4 s peaking time. The continuous line presents the fit of the integral of the exponential function corresponding to the light pulse decay with the decay time of s. It shows a reasonable correspondence to that measured in [2] and [4]. Note that the estimated
Fig. 3. Energy spectra of -rays from a Bi source (top trace) and that of the laboratory background (bottom trace), as recorded with a large CWO crystal at 12 s shaping time constant in a spectroscopy amplifier.
Fig. 4. Photoelectron number measured versus peaking time in amplifier for the CWO 1L sample.
decay time of the light pulse is affected by a shorter component of about 5 s [1], [4], which is not distinguishable in the present measurement. B. Analysis of Energy Resolution The energy resolution, , of the full energy peak measured with a scintillator coupled to a photomultiplier can be written as [12], [13] (1) is the intrinsic resolution of the crystal, is the stawhere tistical contribution and is the transfer resolution associated with the variation of light and photoelectron collection [12], [13]. This quantity is, in fact, negligible in modern PMT readouts with good quality optical contact [12].
´ SKI et al.: CdWO CRYSTAL IN GAMMA-RAY SPECTROMETRY MOSZYN
3127
Fig. 5. Energy resolution, statistical contribution and intrinsic resolution versus peaking time in amplifier for the CWO 1L sample. Fig. 6. Comparison of the nonproportionality curves measured for the small CWO and BGO crystals. The curve for BGO follows that of [9].
The intrinsic resolution of a crystal is mainly associated with the nonproportional response of the scintillator [12], [13], [17] and various effects such as inhomogeneities in the scintillator causing local variations of the light output and nonuniform reflectivity of the crystal covering. The statistical uncertainty of the signal from the PMT is described, as (2) where is the number of photoelectrons and is the variance of the electron multiplier gain, equal to 0.2 for the XP3212. Fig. 5 shows the energy resolution measured with the small CWO 1L sample plotted versus the peaking time constant in the amplifier. It shows that above 6 s shaping time constant (13.2 s peaking time) the energy resolution is weakly improved. To understand better the measured energy resolution, the statistical contribution was calculated according to (2) and then the intrinsic resolution was found using (1), see Fig. 5. It shows that for shaping time constants above 6 s the intrinsic resolution of the scintillator is the main contribution to the total energy resolution. This is an important observation, since it allows improving somewhat a poor counting rate capability of CWO by a shorter shaping. C. Non-Proportionality and Intrinsic Resolution of CWO — Comparison With BGO The large contribution of the intrinsic resolution shown in Fig. 5 suggested inspecting the nonproportionality of the light yield versus -ray energy. Fig. 6 presents the nonproportionality characteristics of the small CWO 1L in comparison to that of a BGO crystal, according to [9]. The nonproportionality is defined here as the ratio of the photoelectron yield measured for photopeaks at a specific -ray energy relative to the yield at 662 keV -peak [12], [13], [17]. Note a comparable shape of the measured curve for CWO to that reported in [13] and [18]. In the whole range of energies, the curves are well matched. Since BGO showed the same nonproportionality at room and temperatures, it was postulated in [9] that its nonproat portionality is a fundamental characteristic of BGO material.
Fig. 7. Energy resolution of CWO and BGO crystals versus gamma rays energy. Error bars are within the size of the points. Data for BGO are taken from [9].
This may imply a further conclusion that the curve presented in Fig. 6 also represents fundamental characteristics of the scintillating materials themselves, i.e., characteristic of heavy oxide crystals. Note that one can expect a high purity of CWO crystals, developed for CT scanners. Both BGO and CWO crystals belong to the scintillators with the lowest afterglow. A further doping or impurities may modify these curves. This was particularly observed for pure NaI and CsI crystals [19], [20], and recently for LSO with different defect concentration [21]. Fig. 7 presents the energy resolution versus the energy of gamma rays obtained with the CWO 1L crystal in comparison to that measured in [9] with BGO crystals of 4 mm thickness. Note a much better energy resolution of the CWO crystal compared to BGO in entire range of energy. There is no doubt that this is due to about 3.5 times higher photoelectron number of the CWO; see Table I. Fig. 8 presents the calculated intrinsic energy resolution of the CWO 1L according to (1), based on the measurements presented in Fig. 7. The points for BGO are shown as well, following the data of [9].
3128
IEEE TRANSACTIONS ON NUCLEAR SCIENCE, VOL. 52, NO. 6, DECEMBER 2005
Common nonproportionality characteristics and, consequently almost identical contributions of the intrinsic resolution observed with the CWO and BGO suggest that in both cases, they represent fundamental characteristics of the scintillating materials themselves, which are characteristic for heavy oxide crystals free of afterglow. REFERENCES
Fig. 8. Intrinsic resolution of CWO and BGO crystals versus gamma ray energies. The points for BGO follow those of [9].
Note again the common straight line in the double logarithmic plot for both the crystals. This observation agrees well with the common nonproportionality characteristics presented in Fig. 6. It further confirms that the intrinsic resolution is strongly correlated with the nonproportionality of the scintillator response [12], [13], [17]. A similar dependency of nonproportionality and intrinsic resolution on the gamma ray energy was observed previously for LSO, GSO and YSO [21]. In all the cases, the nonproportionality curves showed a monotonic decrease of the light yield below 300 keV energy. In the case of the best LSO crystals, the nonproportionality approaches those measured for CWO and BGO [21]. In turn, the NaI(Tl) intrinsic energy resolution and nonproportionality dependence on energy reported in [12] differ clearly fromtheCWOandBGOdata.Theintrinsicresolutionof theCWO and BGO is approximately inversely proportional to the square root of energy, while a step-like dependence of NaI(Tl) resolution [12] is evidently correlated with the shape of the light nonproportionality curve showing an excess of the light yield atlow energies. The measurements carried out with the samples exhibiting a higher afterglow did not show any deterioration of the intrinsic resolution or nonproportionality. IV. CONCLUSION The study of the small CWO crystals showed a high light ph/MeV and an energy resolution of output of % for the 662 keV -rays from a Cs source, both placing the CWO crystal within bright scintillators with a good energy resolution. The high photofraction estimated for the CWO crystals studied here confirmed the high efficiency for gamma rays detection and the high potential of CWO crystals for application in the border monitoring equipment. A further improvement of the quality of larger crystals is required to improve their energy resolution. A comparison of the light output of the CWO with different intensities of afterglow indicates the importance of a reduction of the afterglow to increase the light output of scintillators.
[1] C. L. Melcher, R. A. Manente, and J. S. Schweitzer, “Applicability of Barium fluoride and cadmium tungstate scintillators for well logging,” IEEE Trans. Nucl. Sci., vol. 36, no. 1, pp. 1188–1192, Feb. 1989. [2] D. R. Kinloch, W. Novak, P. Raby, and I. Toepke, “New developments in cadmium tungstate,” IEEE Trans. Nucl. Sci., vol. 41, no. 4, pp. 752–754, Aug. 1994. [3] Y. Eisen, L. G. Evans, R. Starr, and J. I. Trombka, “CdWO scintillator as a compact gamma ray spectrometer for planetary lander missions,” Nucl. Instrum. Meth., vol. A490, no. 3, pp. 505–517, Sep. 2002. [4] T. Fazzini, P. G. Bizzeti, P. R. Maurenzig, C. Stramaccioni, F. A. Danevich, V. V. Kobychev, V. I. Tretyak, and Yu. G. Zdesenko, “Pulse-shape discrimination with CdWO crystal scintillators,” Nucl. Instrum. Meth., vol. A410, no. 2, pp. 213–219, Jun. 1998. [5] P. Schotanus, G. Stam, E. Gerritse, B. Utts, and B. Briaux, “Crismatec Scintillation Detectors Catalog,”, Numerous, France, 1992. [6] V. Ryzhikov, L. Nagornaya, V. Volkov, V. Chernikov, and O. Zelenskaya, “Thermal neutron detectors based on complex oxide crystals,” Nucl. Instrum. Meth., vol. A486, no. 1–2, pp. 156–159, Jun. 2002. [7] L. Nagornaya, V. Ryzhikov, A. Apanasenko, I. Tupitsyna, V. Chernikov, and V. Vostretsov, “Application prospects of cadmium-containing crystals based on tungstates and double tungstates,” Nucl. Instrum. Meth., vol. A486, no. 1–2, pp. 268–273, Jun. 2002. [8] M. Kobayashi, M. Ishii, Y. Usuki, and H. Yahagi, “Cadmium Tungstate scintillators with excellent radiation hardness and low background,” Nucl. Instrum. Meth., vol. A349, no. 2–3, pp. 407–411, Oct. 1994. [9] M. Moszyn´ ski, M. Balcerzyk, W. Czarnacki, M. Kapusta, W. Klamra, A. Syntfeld, and M. Szawlowski, “Intrinsic energy resolution and light yield nonproportionality of BGO,” IEEE Trans. Nucl. Sci., vol. 51, no. 3, pp. 1074–9, Jun. 2004. [10] P. Dorenbos, M Moszyn´ ski, and J. Valentine, presented at the Conclusions of Non-Proportionality Workshop, Portland, OR, Oct. 2003. [11] Tukan 8 k [Online]. Available: http://www.ipj.gov.pl/tukan [12] M. Moszyn´ ski, J. Zalipska, M. Balcerzyk, M. Kapusta, W. Mengeshe, and J. D. Valentine, “Intrinsic resolution of NaI(Tl),” Nucl. Instrum. Meth. A, vol. A484, no. 1–3, pp. 259–269, May 2002. [13] P. Dorenbos, J. T. M. de Haas, and C. W. E. van Eijk, “Non-proportionality in the scintillation response and the energy resolution obtainable with scintillation crystals,” IEEE Trans. Nucl. Sci., vol. 42, no. 6, pp. 2190–2202, Dec. 1995. [14] M. Moszyn´ ski, M. Kapusta, M. Mayhugh, D. Wolski, and S. O. Flyckt, “Absolute light output of scintillators,” IEEE Trans. Nucl. Sci., vol. 44, no. 3, pp. 1052–1061, Jun. 1997. [15] I. Holl, E. Lorenz, and G. Mergas, “A measurement of the light yield of common inorganic scintillators,” IEEE Trans. Nucl. Sci., vol. 35, no. 1, pp. 105–1059, Feb. 1987. [16] G. F. Knoll, Radiation Detection and Measurements, 3rd ed. New York, NY: Wiley, 2000. [17] J. D. Valentine, B. D. Rooney, and J. Li, “The light yield nonproportionality component of scintillator energy resolution,” IEEE Trans. Nucl. Sci., vol. 45, no. 3, pp. 512–517, Jun. 1998. [18] P. Dorenbos, J.T.M. De Haas, and C. W. E. Van Eijk, “Non-proportional response in the light yield of inorganic scintillators,” in Proc. IEEE Nuclear Science Symp. Medical Imaging Conf. 1994 IEEE Conf. Record (Cat. no. 94CH35762), vol. 1, New York, 1995, pp. 304–6. [19] M. Moszyn´ ski, M. Balcerzyk, W. Czarnacki, M. Kapusta, W. Klamra, P. Schotanus, A. Syntfeld, and M. Szawlowski, “Study of pure NaI at room and liquid nitrogen temperatures,” IEEE Trans. Nucl. Sci., vol. 50, no. 4, pp. 767–773, Aug. 2003. [20] M. Moszyn´ ski, M. Balcerzyk, W. Czarnacki, M. Kapusta, W. Klamra, P. Schotanus, A. Syntfeld, and M. Szawlowski, “Energy resoluton and non-proportionality of the light yield of pure CsI at liquid nitrogen temperature,” Nucl. Instrum. Meth., vol. A537, no. 1–2, pp. 357–362, Jan. 2005. [21] M. Balcerzyk, M. Moszyn´ ski, M. Kapusta, D. Wolski, J. Pawelke, and C. L. Melcher, “YSO, LSO, GSO and LGSO. A study of energy resolution and nonproportionality,” IEEE Trans. Nucl. Sci., vol. 47, no. 4, pp. 1319–1323, Aug. 2000.
