Scintillator light yield measurements with waveform digitizers
aa r X i v : . [ phy s i c s . i n s - d e t ] S e p Scintillator light yield measurements with waveform digitizers
T.A. Laplace a, ∗ , B.L. Goldblum a,b , J.A. Brown a , J.J. Manfredi a a Department of Nuclear Engineering, University of California, Berkeley, California 94720 USA b Nuclear Science Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720 USA
Abstract
The proton light yield of organic scintillators has been measured extensively in recent yearsusing fast waveform digitizers and large discrepancies exist in the values reported by dif-ferent authors. In this letter, we address principles of digital signal processing that mustbe considered when conducting scintillator light yield measurements. Digitized waveformpulse height values are only proportional to the amount of scintillation light if the temporalshape of the scintillation pulse is independent of the amount of energy deposited. This isnot the case for scintillation pulses resulting from fast neutron interactions in organic scin-tillators. Authors measuring proton light yield should therefore report pulse integral valuesand ensure that the integration length is long enough to capture most of the scintillationlight.
Keywords: organic scintillator, proton light yield, neutron detection, digital signalprocessingThe absolute light yield of a scintillator can be defined as the number of photons gen-erated per unit energy deposited by a particle slowing down in the scintillating medium[1]. The term light yield is often referred to in the literature as a ratio of the number ofscintillation photons to the energy deposited [2], but has also been used interchangeablywith the terms light output and scintillation response to represent a quantity proportionalto the number of scintillation photons [3, 4, 5].For fast neutron detection using organic scintillators, the primary mechanism for lightgeneration is n-p elastic scattering, and it is the recoiling proton that deposits energy result-ing in molecular excitation and ionization. For the purposes of this discussion, the protonlight yield is defined as the number of scintillation photons generated by a recoiling protonin an organic scintillator. By extension, the proton light yield relation is defined as therelationship between the amount of scintillation light generated by a recoiling proton as afunction of its energy deposited in the medium. This relationship varies non-linearly withrecoil energy due to quenching phenomena [6]. The potential use of organic scintillators asneutron detectors in various arms control and nuclear security applications has promptedefforts to precisely characterize the proton light yield of a variety of organic scintillators in ∗ Corresponding author
Email address: [email protected] (T.A. Laplace) he energy range relevant for fission neutrons. This renewed interest occurred coincidentlywith the broad availability of fast waveform digitizers.Representative traditional analog chains for measurement of the proton light yield oforganic scintillators coupled to photomultiplier tubes (PMTs) are detailed in Refs. [7, 8, 9].Two signals are output from the PMT: the signal from the anode is fed to an analog timingchain and the signal from one of the dynodes is fed to a preamplifier followed by a shapingamplifier. The shaping time plays a crucial role in estimating the collected charge. Whenthe shaping time is much less than the charge collection time, the voltage signal output ofthe shaping amplifier reproduces the current input. This case is recommended for timinginformation or when high event rates are prioritized over accurate energy information [10].If the shaping time is much greater than the charge collection time, the shaping amplifierwill act as a current integrator. The maximum voltage output (i.e., amplitude of the signal)will be proportional to the charge generated within the PMT as long as the load circuitcapacitance remains constant [10]. For scintillator signals, the maximum voltage of theoutput of the shaping amplifier is referred to as the “pulse height.” With a sufficiently longshaping time, this corresponds to the integral of the PMT charge, which is proportional tothe number of scintillation photons (assuming a linear PMT response [11]). For example,in their pioneering measurement of the proton light yield of the NE-213 organic liquidscintillator, Verbinski et al. used a 2.4 µ s shaping time [7]. This is long in comparison totypical temporal responses of organic scintillator pulses on the order of 1 −
100 ns [12, 13].Since the early 2000s, the availability of fast waveform digitizers has enabled a differentapproach to signal acquisition in nuclear instrumentation. Instead of feeding the PMToutput to analog modules to accomplish various operations, the signal is directly fed to thedigitizer, where full waveforms are digitized and then stored for offline signal processing. Thisallows for fine-tuning of various signal processing operations such as integration length, pulse-shape discrimination parameters, baseline estimation, pile-up rejection, etc. Measurementsof the proton light yield of organic scintillators using waveform digitizers have yielded largediscrepancies between different groups. Brown et al. [14] demonstrated that differences inproton light yield measurements of EJ-309 resulted in part from the chosen integration lengthof the digitized waveforms. The proton light yield has been reported in several publicationsby estimating the maximum of the scintillation pulse (i.e., the digitized waveform pulseheight) as opposed to the pulse height from the output of the shaping amplifier in an analogsignal chain [15, 16, 17, 18]. Such analyses reporting proton light yield obtained usingdigitized waveform pulse heights carry several flaws, detailed below.First, the temporal shapes of pulses resulting from electron and proton recoils in organicscintillators are different. For some materials, the pulse shape is sufficiently different suchthat a simple method like comparing the amount of light from the tail of the pulse to thetotal scintillation light allows for separation of events generated by different types of recoilingparticles, i.e., pulse shape discrimination (PSD) via a charge integration approach [19, 20].The proton light yield is often reported in units of MeV electron-equivalent (MeVee), whichcorresponds to the equivalent amount of scintillation light produced by recoiling electronsof a given energy. For a given number of scintillation photons, the average pulse originatingfrom a neutron interaction has less prompt light (due to increased ionization quenching) and2 P u l s e I n t e g r a l t o P u l s e H e i gh t R a ti o Organic Glass - Shin et al. (2019)EJ-309 - Lawrence et al. (2014)EJ-299 - Lawrence et al. (2014)EJ-204 - Laplace et al. (2018)
Figure 1: Ratio of the proton light yield obtained using the pulse integral relative to the digitized waveformpulse height as function of proton recoil energy. Note the zero-suppressed axes. more delayed light (due to increased triplet-triplet annihilation) than a pulse resulting froma γ -ray interaction. The use of the digitized waveform pulse height to obtain the proton lightyield in MeVee therefore results in an underestimation of the proton light yield relation.Second, the proton pulse shape is energy dependent. As a result, digitized waveformpulse heights are not proportional to the number of scintillation photons as function ofenergy. Figure 1 shows the ratio of the proton light yield obtained using the integral of thedigitized waveform to the height of that waveform for the same datasets for four differentorganic scintillators. Shin et al. examined an organic glass scintillator using an integrationlength of 150 ns to ensure >
95% of the scintillation light was captured, and thereforethe pulse integral value is expected to be proportional to the amount of scintillation lightgenerated in the detector [18]. Lawrence et al. used a 180 ns integration window for theliquid EJ-309 and plastic EJ-299-33 organic scintillators [16]. Although measurements ofthe temporal response of the scintillation light generated by EJ-309 are not available in theliterature, we estimated this integration length as capturing ∼
90% of the scintillation lightby examining average digitized waveforms in the MeV proton energy range. In the caseof non-PSD scintillators, data adapted from previous measurements of the EJ-204 protonlight yield are provided, which employed an integration length of 200 ns to capture at least95% of the scintillation light within the acquisition window [3]. The ratio shown in Fig. 1varies as a function of proton recoil energy for all materials examined, demonstrating thatthe two quantities are not proportional. For EJ-204, the ratio remains energy dependent,but to a lesser degree than that of the PSD materials. This variation suggests a change in3 L ong t o S ho r t I n t e g r a l R a ti o Figure 2: Ratio of the proton light yield of EJ-309 from Brown et al. [14] using a long (300 ns) to short(30 ns) integral as a function of proton recoil energy. Note the zero-suppressed axes. the temporal response of these organic scintillators as function of proton energy.Figure 2 shows a comparison of the EJ-309 proton light yield obtained using a long(300 ns) and short (30 ns) integration length from Brown et al. [14]. Similarly, the ratiois energy dependent, suggesting that a short integration length is not proportional to thetotal scintillation light. While the prompt fluorescent light is captured by a 30 ns integrationlength, only a fraction of the delayed light is integrated. The ratio of prompt-to-delayed lightis dependent upon the nature and energy of the incident particle, the degree of ionizationquenching in the medium, and the resulting density of excited and ionized molecules.Pulse height data may be used in an attempt to recover information in high countrate experiments where scintillator pulses overlap. Algorithms have been developed anddemonstrated for slow inorganic scintillators to deconvolve digitized waveforms with pulsepileup [21], though in this case knowledge of the temporal response of the scintillation lightwas required. The complexity is increased with organic scintillators, given a pulse shapethat is both particle and energy dependent.In summary, the shape of scintillation pulses resulting from neutron interactions in or-ganic scintillators is dependent upon the energy deposited. Therefore, the proton light yieldcannot be measured using the maximum pulse height of a digitized waveform. This valuewill be proportional to the peak photon fluence on the photocathode and not to the numberof scintillation photons created. Authors measuring proton light yield should report pulseintegral values and ensure that the integration length is long enough to capture most of thescintillation light. 4 cknowledgements
This work was performed under the auspices of the U.S. Department of Energy byLawrence Berkeley National Laboratory under Contract DE-AC02-05CH11231 and is basedupon work supported by the U.S. Department of Energy, National Nuclear Security Adminis-tration, Office of Defense Nuclear Nonproliferation Research and Development (DNN R&D)via the Nuclear Science and Security Consortium under Award Number DE-NA0003180.
ReferencesReferences [1] I. Holl, E. Lorenz, G. Mageras, A measurement of the light yield of common inorganic scintillators,IEEE Transactions on Nuclear Science 35 (1) (1988) 105–109. doi:10.1109/23.12684 .[2] E. Sysoeva, V. Tarasov, O. Zelenskaya, Comparison of the methods for determination ofscintillation light yield, Nuclear Instruments and Methods in Physics Research Section A:Accelerators, Spectrometers, Detectors and Associated Equipment 486 (1) (2002) 67 – 73. doi:10.1016/S0168-9002(02)00676-9 .[3] T. Laplace, B. Goldblum, J. Brown, D. Bleuel, C. Brand, G. Gabella, T. Jordan, C. Moore, N. Munshi,Z. Sweger, A. Sweet, E. Brubaker, Low energy light yield of fast plastic scintillators, Nuclear Instrumentsand Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and AssociatedEquipment doi:10.1016/j.nima.2018.10.122 .[4] J. B. Birks, The Theory and Practice of Scintillation Counting, Pergamon Press, 1964, Ch. 3, p. 55.[5] F. D. Brooks, Development of organic scintillators, Nuclear Instruments and Methods 162 (1) (1979)477 – 505. doi:10.1016/0029-554X(79)90729-8 .[6] J. B. Birks, The Theory and Practice of Scintillation Counting, Pergamon Press, 1964, Ch. 6, p. 187.[7] V. V. Verbinski, W. R. Burrus, T. A. Love, W. Zobel, N. W. Hill, R. Textor, Calibration of anorganic scintillator for neutron spectrometry, Nuclear Instruments and Methods 65 (1) (1968) 8 – 25. doi:10.1016/0029-554X(68)90003-7 .[8] D. Smith, R. Polk, T. Miller, Measurement of the response of several organic scintillators toelectrons, protons and deuterons, Nuclear Instruments and Methods 64 (2) (1968) 157 – 166. doi:10.1016/0029-554X(68)90189-4 .[9] J. Iwanowska, L. ´Swiderski, T. Krakowski, T. Koz lowski, M. Moszy´nski, G. Pausch, Calibration ofEJ309 liquid scintillator for neutron spectrometry, in: 2012 IEEE Nuclear Science Symposium andMedical Imaging Conference Record (NSS/MIC), 2012, pp. 149–152.[10] G. F. Knoll, Radiation Detection and Measurement, 4th Edition, John Wiley & Sons, 2012, Ch. 4, p.110.[11] M. Friend, G. Franklin, B. Quinn, An LED pulser for measuring photomultiplier linearity, NuclearInstruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors andAssociated Equipment 676. doi:10.1016/j.nima.2012.02.015 .[12] L. M. Bollinger, G. E. Thomas, Measurement of the time dependence of scintillation inten-sity by a delayed-coincidence method, Review of Scientific Instruments 32 (9) (1961) 1044–1050. doi:10.1063/1.1717610 .[13] J. A. Brown, B. L. Goldblum, L. A. Bernstein, D. L. Bleuel, N. M. Brickner, J. A. Caggiano, B. H.Daub, G. S. Kaufman, R. Hatarik, T. W. Phillips, S. A. Wender, K. van Bibber, J. Vujic, N. P. Zaitseva,Relative light yield and temporal response of a stilbene-doped bibenzyl organic scintillator for neutrondetection, Journal of Applied Physics 115 (19) (2014) 193504. doi:10.1063/1.4878238 .[14] J. A. Brown, B. L. Goldblum, T. A. Laplace, K. P. Harrig, L. A. Bernstein, D. L. Bleuel, W. Younes,D. Reyna, E. Brubaker, P. Marleau, Proton light yield in organic scintillators using a double time-of-flight technique, Journal of Applied Physics 124 (4) (2018) 045101. doi:10.1063/1.5039632 .
15] A. Enqvist, C. C. Lawrence, B. M. Wieger, S. A. Pozzi, T. N. Massey, Neutron light output responseand resolution functions in EJ-309 liquid scintillation detectors, Nuclear Instruments and Methods inPhysics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 715(2013) 79 – 86. doi:http://dx.doi.org/10.1016/j.nima.2013.03.032 .[16] C. C. Lawrence, M. Febbraro, T. N. Massey, M. Flaska, F. Becchetti, S. A. Pozzi, Neutron responsecharacterization for an EJ299-33 plastic scintillation detector, Nuclear Instruments and Methods inPhysics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 759(2014) 16 – 22. doi:10.1016/j.nima.2014.04.062 .[17] H. Wang, D. Carter, T. N. Massey, A. Enqvist, Neutron light output function and resolution investiga-tion of the deuterated organic liquid scintillator EJ-315, Radiation Measurements 89 (2016) 99 – 106. doi:https://doi.org/10.1016/j.radmeas.2016.03.009 .[18] T. H. Shin, P. L. Feng, J. S. Carlson, S. D. Clarke, S. A. Pozzi, Measured neutron light-output responsefor trans-stilbene and small-molecule organic glass scintillators, Nuclear Instruments and Methods inPhysics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 939(2019) 36 – 45. doi:10.1016/j.nima.2019.05.036 .[19] R. B. Owen, The decay times of organic scintillators and their application to the discrimination betweenparticles of differing specific ionization, IRE Transactions on Nuclear Science 5 (3) (1958) 198–201. doi:10.1109/TNS2.1958.4315657 .[20] F. Brooks, Organic scintillators, in: O. Frisch (Ed.), Progress in Nuclear Physics, Vol. 5, PergamonPress, 1956, pp. 252–313.[21] W. Guo, R. P. Gardner, C. W. Mayo, A study of the real-time deconvolution of digitized waveformswith pulse pile up for digital radiation spectroscopy, Nuclear Instruments and Methods in PhysicsResearch Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 544 (3) (2005)668–678. doi:10.1016/j.nima.2004.12.036 ..