- Email: [email protected]
Response of the Li-7-enriched Cs2LiYCl6:Ce (CLYC-7) scintillator to 6–60 MeV neutrons
Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
Response of the Li-7-enriched Cs2LiYCl6:Ce (CLYC-7) scintillator to 6–60 MeV neutrons$ Richard S. Woolf n, Anthony L. Hutcheson, Bernard F. Phlips, Eric A. Wulf Space Science Division, U. S. Naval Research Laboratory, 4555 Overlook Ave., SW Washington, DC 20375, United States
art ic l e i nf o
a b s t r a c t
Article history: Received 30 June 2015 Received in revised form 24 August 2015 Accepted 31 August 2015 Available online 24 September 2015
We discuss a test campaign designed to irradiate the 7Li-enriched Cs2LiYCl6:Ce3 þ (CLYC-7) scintillator with 6–60 MeV neutrons using the cyclotron located at the Crocker Nuclear Laboratory in Davis, CA. CLYC-7 is a newly developed scintillator that exhibits the ability to make good γ-ray measurements and has the ability to detect and discriminate fast neutrons via pulse shape discrimination (PSD) while functioning as a spectrometer. This allows a single detector to make good measurement of both stimuli simultaneously. The response of this scintillation detector has been investigated below 20 MeV [1] but has yet to be explored for energies greater than 20 MeV. Understanding the spectral and pulse shape response across a broad energy range is important for any radiation detection instrumentation capable of detecting multiple species. At the highest energies sampled, the CLYC-7 PSD demonstrated not only the standard electron/proton separation expected in a mixed γ/n field but the ability to discriminate locally produced deuterons, tritons and α particles. We show the results from the four different neutron beam energies sampled during the experiment. Lastly, we present the results obtained for relating the light output equivalence between electrons and protons/deuterons. Published by Elsevier B.V.
Keywords: CLYC-7 Fast neutrons Cyclotron Pulse shape discrimination Light output
1. Introduction Cs2LiYCl6:Ce3 þ (CLYC) is an inorganic crystal scintillation detector that is unique in its ability to make measurements of γray lines with good energy resolution (δE/E 4.5% at 662 keV) and the ability to detect and discriminate incident neutrons – from γ rays – based on differences in the scintillation light decay times (on the order of nanoseconds vs. microseconds). One of the components of CLYC is Li, which in its natural abundance contains 7.5% 6Li. This isotope has a large neutron capture cross-section ( 1 kilobarn) at thermal neutron energies and is typically employed for thermal neutron detection. An incident neutron captured by the 6Li yields a detectable α particle (and triton) in an exothermic reaction that imparts 4.78 MeV to the two particles. Previous works [2–6] have reported on the development and performance of CLYC for thermal neutron detection. The large thermal neutron cross-section of 6Li makes CLYC ideal for thermal neutron measurements; when CLYC is used for fast neutron detection, as shown by [7–10], the large thermal neutron peak present at a fixed energy in pulse shape space ☆
The lead author is Richard S. Woolf and will handle all correspondence. Corresponding author. Tel.: þ 1 202 404 2886; fax: þ 1 202 767 6473. E-mail addresses: [email protected] (R.S. Woolf), [email protected] (A.L. Hutcheson), [email protected] (B.F. Phlips), [email protected] (E.A. Wulf). n
http://dx.doi.org/10.1016/j.nima.2015.08.080 0168-9002/Published by Elsevier B.V.
reduces the effectiveness of this modality. To utilize CLYC as a fast neutron detector, depleting the material of 6Li and hence enriching with 7Li, would lead to a scintillator nearly devoid of isotopes with a large thermal neutron capture cross-section. The 7Li-enriched CLYC scintillator (hereafter CLYC-7) can thus be used as a γ-ray/fast neutron detector. CLYC-7 has all the same properties as CLYC but with the ability to serve as a spectrometer for both γ rays and fast neutrons without the thermal neutron component. Aside from standard nuclear physics or homeland defense applications for which fast neutron/γ-ray detectors are employed, the ideal platform for CLYC-7 would be on a spacecraft, either from Earth orbit at 1 astronomical unit (A.U.) or near the Sun in the inner heliosphere, for studying the fast neutron/γ-ray emission from the Sun. One issue to consider in near-Earth orbit for spacebased solar γ-ray/neutron detection is the high-flux, chargedparticle environment. To combat these effects, one would use anticoincidence shielding (plastic scintillator) around the detector array and the appropriate triggering electronics set-up to produce a veto signal to reject these events. This method has been previously implemented for space instrumentation and successfully aided detection systems measuring high-energy, neutral emission from the Sun [11]. As in the example instrument given in the previous reference, typically instrumentation for the detection and measurement of these stimuli is designed to make either a good measurement of one species – with the other species detected for free, at less than optimal performance – or require disparate
48
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
instrumentation for both. Making a good measurement of both with a single instrument has not been feasible in the MeV regime prior to the development of this scintillator. Additionally, the modestly sized CLYC-7 scintillator, coupled to a compact light readout device (e.g., silicon photomultiplier) and operating with low power, would meet the strict criteria for mass, power, and size imposed on spacecraft instrumentation. Recent work [1] has been performed to characterize the response of CLYC-7 with monoenergetic neutrons up to 20 MeV. The work presented herein discusses a campaign to determine the response of CLYC-7 to 6 MeV – 60 MeV neutrons.
2. Experimental methodology
(PMT) (R6231-100) and associated PMT base (E-11897). The PMT was coupled to the quartz window via optical grease and wrapped with black electrical tape (Fig. 2). Because the surface area of the PMT is larger than the quartz window, a layer of a diffuse white reflector was placed on the exposed area of the PMT prior to wrapping. The fast output signal from the PMT anode was read directly into one channel of a 16-channel VME-based flash analog-todigital converter (FADC) manufactured by Struck Innovative Systeme (model no. SIS3316) [14]. The module digitizes the input pulse (5 V range, 50 Ω impedance) and performs pulse height analysis and PSD. The digitizer was set to internally trigger based on a trapezoidal filter; the system was enabled to run in CFD mode such that analysis was always performed at the same part of the pulse, regardless of the amplitude. A second (programmable) trapezoidal filtering method, based on a moving average window, was used to achieve higher quality performance in energy resolution (δE/E) compared to simply integrating the charge over the entire pulse, i.e., we observed an improvement of 1–2% in the full width at half maximum (FWHM) between the two methods. The energy filter algorithm allows the user to set a peaking time (P) and gap time (G) that builds two sums over P, separated by G, from the raw ADC input values, resulting in a trapezoid [15]. This is a standard method used in digital signal processing, allowing the user to optimize the trapezoid shape and flat top to obtain the best performance for a given detector. An additional parameter, the decay correction (α), was implemented to ensure that the trapezoid was symmetric and returned to baseline. PSD is performed by allowing the user to tune programmable gates (accumulators) that can be set to integrate the input pulse over a given region for a given length of time. Offline PSD analysis was performed using the traditional charge integration method [16] given by the following expression:
The CLYC-7 scintillation detector was irradiated with highenergy neutron beams at the Crocker Nuclear Laboratory (CNL) located on the campus of the University of California – Davis in March of 2015. The CNL facility houses a 7600 -isochronous cyclotron, which is capable of accelerating charged particles (protons, deuterons, α particles, and helions) from several 10 s of MeV to greater than 100 MeV. Fast neutrons are produced via a spallation reaction between accelerated protons and a thick beryllium (Be) target: 9Be(p,n)9B (Q value ¼ 1.85 MeV). The Be target was encapsulated in a copper (Cu) housing with a front thickness of 0.51 mm, through which the protons traversed before striking the Be target, and a rear thickness of 6.35 mm. The proton beam energies used were: Ep ¼20.4 MeV, 30 MeV, 45 MeV, and 67.5 MeV (highest proton energy achievable). The specific energy loss (dE/ dx) due to the proton passing through the Be target and the Cu foil, as determined via SRIM simulations [12], and the Q value of the reaction lead to a continuum neutron spectrum with endpoint energies of En ¼ 5.9 MeV, 18.7 MeV, 36.1 MeV, and 60.5 MeV. The beam current used at each energy was 200 nA, 50 nA, 200 nA, and 80 nA–100 nA, respectively. A steel collimator (length ¼ 1.6 m) forms an on-axis beam of neutrons that exits from an aluminum window 2.85 m away from the target. The expected beam flux at the exit was on the order of 104n cm2 s 1 at 30 MeV and above; the flux at lower energies is more difficult to ascertain [13]. Data from 12.5 MeV and 14.6 MeV have yielded measured beam fluxes of 30 n cm2 s 1 and 90 n cm2 s 1, respectively. The 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillator cube (enriched to 99.93% 7Li) is housed on five sides in 3.175-mm-thick aluminum with a 23 mm 23 mm 3.1 mm quartz window on the sixth side (Fig. 1). The scintillator is wrapped with Teflon tape and supported on the top and bottom with 1-mm-thick optical pressure pads. The light-readout device is coupled to the scintillator volume at the quartz window. For this campaign, we used a 51 mm (diameter) Hamamatsu super bialkali photomultiplier tube
where QS (QI) is the short (intermediate) integration gate length. The length of the long integration gate, QL, was set to integrate the total charge. The bias voltage on the detector was provided by one channel of a 32-channel high-current ISEG high voltage supply [17] powered via a Wiener MPOD mixed crate [18]. At CNL, the CLYC-7 detector was placed 2 m from the beam exit to achieve high statistics in a reasonable amount of time given that the efficiency to fast neutrons is on the order of a few tenths of a percent at a few MeV and increases to 3% at 20 MeV [1,10]. Extrapolating to 60 MeV, one would expect a continued increase in efficiency. To reduce the γ-ray flux from the spallation reaction, lead bricks (total thickness of 10 cm) were placed in front of the beam exit.
Fig. 1. The 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector packaged in custom-designed aluminum housing.
Fig. 2. The CLYC-7 scintillator (white arrow) shown with 51 mm super bialkali Hamamatsu PMT set-up at the Crocker Nuclear Laboratory experiment room.
1
QS QI
ð1Þ
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
3. Results 3.1. Calibration The detector output read into the FADC was tuned to achieve optimal δE/E and calibrated using standard radioactive check sources in the laboratory prior to the test campaign. Using a 137Cs source and the PMT operating at a positive bias voltage of 1300 V, the δE/E at 662 keV was found to be 6.2% (FWHM) with P ¼3.4 μs, G ¼0.3 μs and α ¼1.5 μs (Fig. 3). The performance was unexpected given that this detector demonstrated a δE/E of 4.4% (FWHM) at 662 keV when using an analog shaping amplifier and multichannel analyzer (MCA) for data acquisition. We determined that the memory allotted by the SIS3316 firmware for shaping the pulse greatly affects scintillators that require a long shaping time. This sample of the CLYC-7 scintillator required an analog shaping time of 10 μs to achieve optimal performance. The length of the long gate (QL), which integrates over the entire pulse, was 10 μs, but the energy filter only allowed for a full range of 4 μs. (An updated version of the SIS3316 firmware allots more memory for the energy filter to accommodate a full range of greater than 10 μs. Work to improve the energy performance is currently under investigation). At CNL, to accommodate the dynamic range for large amplitude pulses produced by 60.5 MeV neutrons, the high voltage on the PMT was lowered to 1150 V, thus reducing the response at low energies. A recalibration at this voltage was not completed. However, by comparing the shape of long background runs completed at each voltage for equal run times, we find that they differ by a multiplicative gain factor. This scaling factor, A, is then applied to the two-point linear energy calibration obtained at 1300 V using 60 Co data (1173 keV and 1333 keV) to correct the calibration for data obtained at the lower voltage. The two-point calibration was linear; for higher energies, we assumed that the calibration maintained linearity. The calibration curve is given as follows:
49
optimization was done in the laboratory prior to the campaign with the PMT operating at a positive bias of 1300 V. PSD event selection criteria for subsequent analysis with the PMT operating at 1150 V were determined from the neutron beam data. A measure of the PSD performance is given by the Figure of Merit (M), determined by the ratio of the difference in centroids of the γ-ray and neutron distributions to sum of the FWHM of each distribution. Gate widths of 0.08 μs (QS) and 1.2 μs (QI) with QI starting 0.12 μs after the start of QS were used to achieve an M of 2.4 for electron-equivalent energies greater than 750 keVe.e. (for reference, this electron equivalent energy corresponds to a pulse height channel of 13 in Fig. 4). The CLYC-7 scintillation detector was exposed to the four neutron beams for a typical period between 300 s and 900 s. The run times were determined by the amount of beam time allotted for the experiment and to ensure that Gaussian statistics would be achieved for the total number of counts registered by the detector. The total rate measured by the detector at each energy was 70 Hz (5.9 MeV), 160 Hz (18.7 MeV), and 3 kHz (36.1 MeV and 60.5 MeV). 3.2. Neutron-induced reactions in CLYC-7 Figs. 5–8 show the 2-d scatter plots of the calculated pulse shape vs. pulse height acquired at each beam energy. The pulse shape was calculated by Eq. (1), and the pulse height represents the energy deposited by the ionizing particle; both scales have been rebinned in terms of 28 bins. For reference, an electron equivalent energy of 1.5 MeVe.e. corresponds to a pulse height channel of 25. Investigation of these standard scatter plots reveal the different reaction channels available from neutron-induced reactions in the CLYC-7 with increasing neutron beam energy. The following subsections discuss our analysis to determine which reactions were occurring and their attribution.
where Ee.e. is the electron-equivalent energy (in keV), A is 0.63, and x is the channel. To optimize the PSD, we exposed the detector to a 160 μCi 252Cf laboratory source, located behind 15-cm-thick lead to reduce the intense γ-ray contribution from the source (Fig. 4). The
3.2.1. Neutron beam energy: 5.9 MeV For the neutron beam with 5.9 MeV endpoint energy, the pulse shape vs. pulse height shown in Fig. 5 displays the well-separated bands corresponding to electrons (lower band) and protons (upper band). The proton band results from the reaction 35Cl(n,p)35S (Q¼ þ 615 keV), which is dominant in the 1 MeV–10 MeV energy range due to its large cross-section and the fact that 35Cl is the
Fig. 3. A 10 μCi 137Cs source irradiated the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector read out by a 51 mm super bialkali Hamamatsu PMT and one channel of the 16-channel flash ADC. The FWHM of the 662 keV photopeak is 6.2%. For comparison, the same detector setup read out by analog shaping electronics measured the 662 keV photopeak with a FWHM of 4.4%.
Fig. 4. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated with a 160 μCi 252Cf γ/n source.
Ee:e: ¼ Að1:12 10 2 x þ 1:25Þ
ð2Þ
50
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
most abundant isotope in the CLYC-7 scintillator, as previously reported by D'Olympia et al. [1]. An additional proton-producing reaction is from 35Cl(n,p)35S*, i.e., excited states of the 35S. Starting 4 MeV and above, the cross-section for the 1st and 2nd excited states of 35S from the (n,p) reaction on 35Cl is comparable to the cross-section of the ground state reaction ( 5 10 2 b). These excited state reactions ultimately contribute to the counts in the proton band. Fig. 5 also demonstrates that the thermal neutron contribution from the 6Li reaction is negligible. If the scintillator was not depleted of this isotope, as shown in Ref. [7], a strong thermal neutron capture peak, independent of the incident neutron energy, would appear at a fixed energy in the proton band in pulse shape space. 3.2.2. Neutron beam energy: 18.7 MeV Fig. 6 shows the pulse shape vs. pulse height for the neutron beam with endpoint energy of 18.7 MeV. At this energy, we still observe the two bands seen at the 5.9 MeV endpoint; however, note the emergence of some structure in the upper band. The bottom part of the upper band is attributed to the neutron-capture protons given that these events follow the same trend (in terms of the mean of the rebinned pulse shape value) as observed at lower energies but have increased pulse height due to the increased endpoint energy of the beam. This fact alone demonstrates the utility of the CLYC-7 detector as a spectrometer. The top part of the upper band begins to show splitting above channel 40. The bands become more defined and separated as the incident neutron energy is increased. 3.2.3. Neutron beam energy: 36.1 MeV and 60.5 MeV Continuing with the trend observed at 18.7 MeV, for the higher energy neutron beams we observed the following in the scatter plots (Figs. 7 and 8): the well-defined γ-ray-related band as seen in previous energies (rebinned pulse shape value ¼70); three welldefined, intense bands between rebinned pulse shape values of 110–140; a less-intense but still defined band at rebinned pulse shape value 125; and hints of two separate bands with rebinned pulse shape value centroids of 95 and 185. This was the first time observing the fine banding structure in what is typically thought of as the proton only band as no data at this neutron energy existed for CLYC-7 at the time of the measurement. To determine the attribution of these bands, we began investigating all possible neutron-induced reactions that could occur for the
Fig. 5. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 5.9 MeV neutron beam.
Fig. 6. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 18.7 MeV neutron beam.
Fig. 7. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 36.1 MeV neutron beam.
Fig. 8. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 60.5 MeV neutron beam. This plot is annotated to indicate the reaction channels and accidental events observed.
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
given isotopes that comprise the CLYC-7 scintillator. The 35Cl reaction is the main attribute for the (n,p) reaction that leads to fast neutron detection at lower energies. Other reactions to consider are those produced by light ionizing particles, such as the deuteron via (n,d) reaction, the triton via (n,t) reactions, and α particles via (n,α) reactions. Important considerations for each of these reactions are the cross-section for a given energy range, the reaction threshold energy, the Q value (exothermic or endothermic), and the isotopic abundance. Based on previous work studying the relationship of the light output (dL/dx) and dE/dx [19,20], we know that charged particles heavier than the proton (deuteron, triton, alphas, 3He, carbons, etc.) populate a greater proportion of excited states in the constituent scintillation material, thus producing an increase in the slow component of the scintillation light and hence leading to longer decay times back to baseline. Because our method of PSD is a measure of the total integrated charge, longer decay times lead to larger values for the pulse shape. Similarly, the dL/dx is related to dE/dx, which in turn is proportional to Z2. Comparing the dL/dx for protons to deuterons and tritons, because Z¼1 in all cases, any differences observed are attributed to the Q value of the reaction. Aside from the Q value, for α particles, the charge state of Z¼ 2 leads to a reduction in the light output due to quenching effects from the radiation-less loss of energy when populating more excited states. Based on previous results with other scintillators exposed to light ions, one would start with the ansatz that the four main bands in Figs. 7 and 8 between rebinned pulse shape values 110–140 correspond to protons, deuterons, tritons and α particles in order of increasing value. For proton-producing reactions, we investigated (n,p), (n,np), and (n,2n) reactions. We know from the lower energy beam data that the lowest part of the upper band corresponds to protons. At these energies, the reactions considered for the isotopes associated with CLYC-7, in addition to the 35Cl(n,p)35S reaction, were 37 Cl(n,p)37S, 133Cs(n,p) 133Xe, 89Y(n,p)89Sr, 140Ce(n,p)140La, and 142Ce (n,p)142La. Cross-section data were obtain using the NNDC ENDF database [21]; for the reactions listed here and elsewhere in the manuscript, we required that the cross-section data have endpoint energies of at least 60 MeV. The only currently available values of the cross-sections for the isotopes in question are based on the TENDL (TALYS-based Evaluated Nuclear Data Library) 2009 and 2014 numerical models [22]. TALYS is a software modeling package used for the simulation of nuclear reactions up to 200 MeV, and specifically treats protons, deuterons, tritons, and α particles as both the target and ejectiles [23]. For the energy region of 20 MeV– 60 MeV, the listed reactions have comparable cross-sections with values ranging from 3 10 2 b to 3 10 3 b (see Fig. 9a). The reactions involving 133Cs and 35Cl are exothermic with Q values of 355 keV and 615 keV, respectively, whereas the reactions involving 37 Cl, 89Y, 140Ce, and 142Ce are endothermic with Q values ranging from 718 keV to 4.1 MeV. Fig. 9a clearly shows that for incident neutrons with energies less than 20 MeV, the 35Cl(n,p) reaction is dominant, as discussed in Section 3.2.1. For energies greater than 20 MeV, the other constituents begin to play a more dominant role in producing protons from the neutron capture reaction. The (n,np) and (n,2n) reactions are typically endothermic with Q values between 5 MeV to 10 MeV. Regardless, these reactions lead to either additional neutrons or a neutron and a proton, which would show up as a lower-energy proton in pulse shape space or could lead to some additional reactions at lower energies. Therefore, one can conclude that the observed proton band is caused by a combination of the listed reactions. The endpoint of the pulse height spectrum serves as a calibration point for proton equivalent energy, once the highest Q value is accounted for. Beginning with the proton band, the second band encountered with increasing rebinned pulse shape value corresponds to deuterons from (n,d) reactions. The reactions of interest for (n,d) are 133 Cs(n,d)132Xe, 89Y(n,d)88Sr, 35Cl(n,d)34S, and 37Cl(n,d)36S with Q
51
values ranging from 3.85 MeV to 6.1 MeV. The reaction crosssection for these isotopes (Fig. 9b) begin to turn on 10 MeV and are comparable in the 20 MeV–60 MeV energy range with values varying between 5 10 1 b to 1 10 2 b. At 20 MeV the Cl isotopes are the main deuteron contributors; above 30 MeV the 89 Y(n,d) reaction yields the largest cross-section and is greater than the Cl isotopes by less than a factor of 2. Compared to the 133 Cs, the 89Y(n,d) reaction cross-section is greater by a factor of 3 at 30 MeV then reducing to a factor of 2 at 60 MeV. Aside from the (n,d) reaction, other deuteron-producing reactions are (n, nd) or (n, pd). As was the case with the proton-producing reactions, these events would either give rise to additional lowerenergy neutrons or register in the lower-energy region of the proton/deuteron pulse shape space. The third band encountered with increasing rebinned pulse shape values in Figs. 7 and 8 is attributed to neutron-induced tritons from (n,t) reactions. The main reaction of interest is 35Cl(n, t)33S with a cross-section of 1 10 2 b in the 20 MeV region and falling off by an order of magnitude at 60 MeV (Fig. 9c). Compared to the (n,p) and (n,d) reactions, there is roughly an order of magnitude reduction in cross-section at the higher energies, which contributes to the reduction in the observed counts in this band. The Q value for this endothermic reaction is 9.3 MeV, requiring that this amount of energy be removed by the reaction products, which leads to a shift in the pulse height spectrum for incident neutrons of the same energy. Another reaction to consider in the 20 MeV range is 7Li(n,t)5He. The cross-section is 1 10 2 b at 20 MeV but then falls off precipitously (by three orders of magnitude) at 60 MeV. This reaction would lead to an increase in the signal at lower energies where the bands are not well separated. The fourth band observed in Figs. 7 and 8 is attributed to neutroninduced α-particle reactions (n,α). The reactions of 35Cl(n,α)32P, 37Cl(n, α)34P,133Cs(n,α)130I, and 140Ce(n,α)137Ba have the largest crosssections at 20 MeV, ranging from 5 10 2 b to 1 10 2 b; at 60 MeV, the cross-section for these reactions has dropped to 5 10–4 b, with the 35Cl(n,α) reaction remaining the largest over the entire energy range of interest (Fig. 9d). One can see the effect caused by the reduction in cross-section for increasing incident neutron energy in Figs. 7 and 8 where the α particle band becomes less intense, in comparison to the (n,p) and (n,d) bands. We conclude that the dominant source of α particles is from the 35Cl(n,α) reaction, which is exothermic with a Q value of þ937 keV. Lastly, as discussed at the beginning of this section, two additional banding features appear in the 2-d scatter plot in Figs. 7 and 8: one clustered above the main upper band (rebinned pulse shape value¼185) and a band located between the electron and proton bands (rebinned pulse shape value¼ 95). These bands appear in the data at the highest beam energies and are nearly absent at lower energies. For the band clustered above the main upper band, at first glance, one may surmise this is the result of another ionizing particle, one with a smaller cross-section (due to the number of observed counts) and higher dE/dx (due to its location in pulse shape space). However, closer inspection reveals that these events result from accidentals. A 1-d histogram (Fig. 10) of the pulse shape parameter – obtained by collapsing the y-axis of the 2-d scatter plot and applying cuts between 6.5 MeVe.e.–8.5 MeVe.e. – show each band and their relative contribution. The γ-ray-induced events have mainly been filtered out by the energy cuts; the main neutron-induced reactions are shown accordingly. The expression used to calculate the PSD is shown in Eq. (1). Typical values observed for the electrons, protons, deuterons, tritons, and alphas range from 0.80 to 0.91. The values for this band peak are around unity. (For clarity on the 2-d plots, a rebinned pulse shape value of 256 is equal to 1.1). Values around unity indicate that the integrated pulse within the short gate is much less than the integrated pulse within the intermediate gate window (QS {QI). These events stem from accidental triggers in the digitizer
52
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
Fig. 9. a: Cross-section values vs. incident energy for relevant (n,p) reactions. b: Cross-section values vs. incident energy for relevant (n,d) reactions. c: Cross-section values vs. incident energy for relevant (n,t) reactions. d: Cross-section values vs. incident energy for relevant (n,α) reactions.
system; the rate of accidentals is dependent on the input event rate, the pulse amplitude and timing sequence. For the accidentals we observe in the 2-d plot, where QI registers a value but QS does not (or is very small in comparison), these events can be attributed to a large amplitude pulses followed by secondary pulses occurring at the end of the integration window, and causing these accidental events to occur in the subsequent triggered event. We calculate the accidental (pile up) rate expected for the gate settings used and the total rate, as given by equation 3 [19], where rch is the rate of accidental events, rs is the source rate, and tr is the resolving time. r ch ¼ r 2s t r
Fig. 10. The 1-d histogram of the pulse shape for the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 60.5 MeV neutron beam at the Crocker Nuclear Laboratory. This plot is annotated to indicate the reaction channels and accidental events observed.
ð3Þ
For a total integration length of 10 μs and with 50 ns of dead time between pulses [15], tr ¼10.05 μs. For a source rate (rs) at the highest energies of 3 kHz, we find rch ¼ 90 Hz. For a 300-s run, the number of accidental events produced should be on the order of 3 104. By selecting out the accidental events in the 2-d scatter plot, we find a total on the order of 5 103 events. We note that the discrepancy is because not all events with pile up will show up in the accidental band; this band corresponds to the extreme case where QS { QI that was previously described. For comparison, at the lower energies, the rs has been reduced by more than an order
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
of magnitude. Correspondingly, the accidental rate expected is 0.049 Hz and 0.26 Hz at 5.9 MeV and 18.7 MeV, respectively. For the 900-s run completed at 5.9 MeV, Eq. (3) yields 40 events in an accidental band. Selecting on events above the proton band yields one accidental event, which was not unexpected considering the low rate and amplitude of pulses at this energy. For a 600-s run completed at 18.7 MeV, Eq. (3) yields 160 events in the accidental band. Selecting on events above the proton band yields 15 accidental events. Again, this result is not unexpected given the ratio of observed-to-calculated accidentals at 36.1 and 60.5 MeV and the fact that the rates and amplitudes of the pulses are much lower than at higher energies. The band located between the electron and proton bands is attributed to the neutron-capture protons escaping from the scintillator volume without fully depositing their energy [24]. Similar results have been observed in organic scintillators [25]. Figs. 8 and 10 have been annotated to illustrate the particles attributed to each of these bands.
53
3.3. Equivalent energy relationship One of the main goals of this experiment was to determine the light output response of protons from high-energy neutrons and relate this quantity to the light output of known-energy electrons from γ-ray-induced events. With the emergence of the additional reaction channels, we then wanted to additionally obtain a relation for the heavier particles. A measure of the γ-ray response comes from the calibration discussed in the last section (Eq. (2)), relating the light output from electrons to known γ-ray energy (Ee.e.). The light output response for heavier ions was obtained by first making 2-d selections on the pulse shape vs. pulse height, then reconstructing the resulting pulse height spectra of those events. Fig. 11a–d shows the pulse height spectra for each particle band. To determine the point in the spectrum that corresponds to the maximum energy deposited in the scintillator by the ion, we used a method analogous to one used previously in measurements with organic scintillator where the half height of the abrupt edge near the maximum pulse height corresponds to the maximum energy [26,27]. Depending on the reaction, one must also assign the proper reaction Q value to account for the particle’s total energy.
Fig. 11. a: The proton-selected pulse height spectra resulting from the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillator irradiated by the four differing energy neutron beams (5.9 MeV, 18.7 MeV, 36.1 MeV, and 60.5 MeV). b: Pulse height spectra for deuteron-selected events at the two energies (36.1 MeV and 60.5 MeV) in which this band was most prominent. c: Pulse height spectra for triton-selected events at the two energies (36.1 MeV and 60.5 MeV) in which this band was most prominent. d: Pulse height spectra for α-particle-selected events at the two energies (36.1 MeV and 60.5 MeV) in which this band was most prominent.
54
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
Table 1 Energy equivalence relationship between protons and electrons. The first column corresponds to the neutron beam energy, whereas the second column represents the beam energy with the Q value of the reaction taken into account. The third column shows the value, in terms of ADC channel, representing the location in the pulse height spectrum that corresponds to the maximum energy deposit. The last column is the electron equivalent energy as calculated by Eq. (2). Neutron energy, En (MeV)
Proton energy, Ep (MeV)
Maximum energy deposit location (ADC channel)
Electron equivalent energy, Ee.e. (MeV)
18.7 36.1 60.5
19.3 36.7 61.1
616,994 1,429,620 2,190,760
4.4 10.1 15.5
Table 2 Energy equivalence relationship between deuterons and electrons. The first column corresponds to the neutron beam energy, whereas the second column represents the beam energy with the Q value of the reaction taken into account. The third column shows the value, in terms of ADC channel, representing the location in the pulse height spectrum that corresponds to the maximum energy deposit. The last column is the electron equivalent energy as calculated by Eq. (2). Neutron energy, En (MeV)
Deuteron energy, Ed (MeV)
Electron Equivalent Maximum energy deposit location (ADC Energy, Ee.e. (MeV) channel)
36.1 60.5
32.3 56.7
1,477,130 2,214,960
10.5 15.7
Figs. 11a shows the reconstructed pulse height spectra for protons at each neutron beam energy. The previously described method for finding the location (in terms of ADC channel) of the maximum energy deposit was used at 18.7 MeV, 36.1 MeV, and 60.5 MeV. For 5.9 MeV neutrons, no abrupt edge was observed. The corresponding ADC channel was input into Eq. (2) to obtain an electron equivalent energy (Ee.e.) for a given proton energy. The same method was applied to the deuteron-selected pulse height spectra shown in Figs. 11b. The main results are outlined in Tables 1 and 2, which show: the neutron beam energy, the particle energy corrected for the reaction Q value, the ADC channel corresponding to the location of maximum energy deposited by the particle, and the electron equivalent energy derived from Eq. (2). The Ee.e. is comparable for protons and deuterons except for the fact that the deuteron energy is lower than that of the proton due to the Q value. For the neutron-induced triton and α-particle reactions, the number of events collected in each band was not sufficient to adequately produce pulse height spectra that demonstrated an abrupt edge. As shown in Fig. 11c and d, the resultant spectra follow an exponential fall off with no distinct features. Without an abrupt edge to determine the maximum energy deposit by the ion, we were unable to obtain a relationship for the light yield produced by tritons or α particles and how that related to the electron light yield. At the least, the endpoint observed in the pulse height spectra places a lower limit on the relationship between triton/αparticle energy and the electron equivalent light yield.
4. Conclusions This work presents the results from a campaign conducted at the Crocker Nuclear Laboratory in Davis, CA, in which we irradiated a 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillator with 6 MeV– 60 MeV neutron beams. Prior work investigated monoenergetic neutron beams characterizing CLYC-7 at o20 MeV. Thus, it was our goal to determine the response of this novel scintillator for
420 MeV neutrons. For the highest beam energies tested (36.1 MeV and 60.5 MeV), in addition to the standard electron/proton bands expected in a mixed γ/n field, we observed splitting of the traditional proton band in pulse shape space attributed to high-energy reaction channels becoming available for the production of deuterons, tritons, and α particles. We also observed escaping protons from the scintillator volume and an increase in the accidental events. Based on these new-found results, we determine the heretofore undetermined light output equivalent above 20 MeV relating light output for protons and deuterons to equivalent energy electrons. Unfortunately, the statistics in the triton/α-particle bands did not allow for a determination of the light output relationship between these particles and the equivalent energy electrons. The results presented establish the previously unknown response to highenergy fast neutrons with CLYC-7 and show that this scintillator can be used as a spectrometer based on not only the proton reaction but also the other reactions channels that appear at higher energy. The heavier ion reactions lead to the ability to perform spectrometry over a larger dynamic energy range. Characterization and understanding the properties of the CLYC-7 scintillator are important for future implementation in the field of solar physics or other applications where high-energy neutron fields are encountered.
Acknowledgments This work was sponsored by the Chief of Naval Research (CNR). The authors would like to thank the following people: Spencer Hartman from the Crocker Nuclear Laboratory for his assistance during the neutron beam campaign; Joshua Tower from Radiation Monitoring Devices for loaning our group the super bialkali PMT; J. Eric Grove and Lee Mitchell from the Naval Research Laboratory for useful discussions on data analysis.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]
N. D’Olympia, et al., Nuclear Instruments and Methods A 716 (2014) 433. C.M. Combes, et al., Journal of Luminescence 82 (1999) 299. E.V.D. van Loef, et al., Journal of Physics: Condensed Matter 14 (2002) 8481. A. Bessiere, et al., Nuclear Instruments and Methods A 537 (2005) 242. J. Glodo, et al., IEEE Transaction on Nuclear Science NS55 (3) (2008) 1206. M. Bourne, et al., Nuclear Instruments and Methods A 736 (2014) 124. N. D’Olympia, et al., Nuclear Instruments and Methods A 694 (2012) 140. J. Glodo, et al., IEEE Transaction on Nuclear Science NS60 (2) (2013) 864. M. Smith, et al., IEEE Transaction on Nuclear Science NS60 (2) (2013) 855. M. Smith, et al., Nuclear Instruments and Methods A 784 (2015) 162. J. Ryan, et al., Advances in Space Research 13 (9) (1993) 255. SRIM – the Stopping and Range of Ions in Matter. 〈http://www.srim.org〉, 2013. S. Hartmann, in Private Communications, 2015. Struck Innovative Systeme, SIS3316-250-14 16 channel 250 MSPS 14-bit. 〈http://www.struck.de/sis33166.html〉, Online 2015. SIS3316, 16 Channel VME Digitizer, User Manual, version: SIS3316-M-1-1V110, Online 2015. F.D. Brooks, Nuclear Instruments and Methods 4 (1959) 151. ISEG Germany, EDS. 〈http://www.iseg-hv.com/en/products/product-details/ product/20/〉. Wiener Plein and Baus Elektronik, MPOD Mixed Crate. 〈http://www.wiener-d. com/sc/power-supplies/mpod–lvhv/mpod-mix-crate.html〉. G.F. Knoll, Radiation Detection and Measurement, 3rd ed., Wiley, New York, 2000. V. Avdeichikov, et al., Nuclear Instruments and Methods A 484 (2002) 251. ENDF/B-VII.1. 〈http://www.nndc.bnl.gov〉. A.J. Koning, et al., Nuclear Data Sheets 113 (2012) 2841. A.J. Koning, et al. 〈http://www.talys.eu/tendl-2014.html〉. M. Moszynski, et al., Nuclear Instruments and Methods A 343 (1994) 563. A. Comrie, et al., Nuclear Instruments and Methods A 772 (2015) 43. V. Verbinski, et al., Nuclear Instruments and Methods 65 (1968) 8. R.S. Woolf, et al., Nuclear Instruments and Methods A 784 (2015) 80.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
Response of the Li-7-enriched Cs2LiYCl6:Ce (CLYC-7) scintillator to 6–60 MeV neutrons$ Richard S. Woolf n, Anthony L. Hutcheson, Bernard F. Phlips, Eric A. Wulf Space Science Division, U. S. Naval Research Laboratory, 4555 Overlook Ave., SW Washington, DC 20375, United States
art ic l e i nf o
a b s t r a c t
Article history: Received 30 June 2015 Received in revised form 24 August 2015 Accepted 31 August 2015 Available online 24 September 2015
We discuss a test campaign designed to irradiate the 7Li-enriched Cs2LiYCl6:Ce3 þ (CLYC-7) scintillator with 6–60 MeV neutrons using the cyclotron located at the Crocker Nuclear Laboratory in Davis, CA. CLYC-7 is a newly developed scintillator that exhibits the ability to make good γ-ray measurements and has the ability to detect and discriminate fast neutrons via pulse shape discrimination (PSD) while functioning as a spectrometer. This allows a single detector to make good measurement of both stimuli simultaneously. The response of this scintillation detector has been investigated below 20 MeV [1] but has yet to be explored for energies greater than 20 MeV. Understanding the spectral and pulse shape response across a broad energy range is important for any radiation detection instrumentation capable of detecting multiple species. At the highest energies sampled, the CLYC-7 PSD demonstrated not only the standard electron/proton separation expected in a mixed γ/n field but the ability to discriminate locally produced deuterons, tritons and α particles. We show the results from the four different neutron beam energies sampled during the experiment. Lastly, we present the results obtained for relating the light output equivalence between electrons and protons/deuterons. Published by Elsevier B.V.
Keywords: CLYC-7 Fast neutrons Cyclotron Pulse shape discrimination Light output
1. Introduction Cs2LiYCl6:Ce3 þ (CLYC) is an inorganic crystal scintillation detector that is unique in its ability to make measurements of γray lines with good energy resolution (δE/E 4.5% at 662 keV) and the ability to detect and discriminate incident neutrons – from γ rays – based on differences in the scintillation light decay times (on the order of nanoseconds vs. microseconds). One of the components of CLYC is Li, which in its natural abundance contains 7.5% 6Li. This isotope has a large neutron capture cross-section ( 1 kilobarn) at thermal neutron energies and is typically employed for thermal neutron detection. An incident neutron captured by the 6Li yields a detectable α particle (and triton) in an exothermic reaction that imparts 4.78 MeV to the two particles. Previous works [2–6] have reported on the development and performance of CLYC for thermal neutron detection. The large thermal neutron cross-section of 6Li makes CLYC ideal for thermal neutron measurements; when CLYC is used for fast neutron detection, as shown by [7–10], the large thermal neutron peak present at a fixed energy in pulse shape space ☆
The lead author is Richard S. Woolf and will handle all correspondence. Corresponding author. Tel.: þ 1 202 404 2886; fax: þ 1 202 767 6473. E-mail addresses: [email protected] (R.S. Woolf), [email protected] (A.L. Hutcheson), [email protected] (B.F. Phlips), [email protected] (E.A. Wulf). n
http://dx.doi.org/10.1016/j.nima.2015.08.080 0168-9002/Published by Elsevier B.V.
reduces the effectiveness of this modality. To utilize CLYC as a fast neutron detector, depleting the material of 6Li and hence enriching with 7Li, would lead to a scintillator nearly devoid of isotopes with a large thermal neutron capture cross-section. The 7Li-enriched CLYC scintillator (hereafter CLYC-7) can thus be used as a γ-ray/fast neutron detector. CLYC-7 has all the same properties as CLYC but with the ability to serve as a spectrometer for both γ rays and fast neutrons without the thermal neutron component. Aside from standard nuclear physics or homeland defense applications for which fast neutron/γ-ray detectors are employed, the ideal platform for CLYC-7 would be on a spacecraft, either from Earth orbit at 1 astronomical unit (A.U.) or near the Sun in the inner heliosphere, for studying the fast neutron/γ-ray emission from the Sun. One issue to consider in near-Earth orbit for spacebased solar γ-ray/neutron detection is the high-flux, chargedparticle environment. To combat these effects, one would use anticoincidence shielding (plastic scintillator) around the detector array and the appropriate triggering electronics set-up to produce a veto signal to reject these events. This method has been previously implemented for space instrumentation and successfully aided detection systems measuring high-energy, neutral emission from the Sun [11]. As in the example instrument given in the previous reference, typically instrumentation for the detection and measurement of these stimuli is designed to make either a good measurement of one species – with the other species detected for free, at less than optimal performance – or require disparate
48
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
instrumentation for both. Making a good measurement of both with a single instrument has not been feasible in the MeV regime prior to the development of this scintillator. Additionally, the modestly sized CLYC-7 scintillator, coupled to a compact light readout device (e.g., silicon photomultiplier) and operating with low power, would meet the strict criteria for mass, power, and size imposed on spacecraft instrumentation. Recent work [1] has been performed to characterize the response of CLYC-7 with monoenergetic neutrons up to 20 MeV. The work presented herein discusses a campaign to determine the response of CLYC-7 to 6 MeV – 60 MeV neutrons.
2. Experimental methodology
(PMT) (R6231-100) and associated PMT base (E-11897). The PMT was coupled to the quartz window via optical grease and wrapped with black electrical tape (Fig. 2). Because the surface area of the PMT is larger than the quartz window, a layer of a diffuse white reflector was placed on the exposed area of the PMT prior to wrapping. The fast output signal from the PMT anode was read directly into one channel of a 16-channel VME-based flash analog-todigital converter (FADC) manufactured by Struck Innovative Systeme (model no. SIS3316) [14]. The module digitizes the input pulse (5 V range, 50 Ω impedance) and performs pulse height analysis and PSD. The digitizer was set to internally trigger based on a trapezoidal filter; the system was enabled to run in CFD mode such that analysis was always performed at the same part of the pulse, regardless of the amplitude. A second (programmable) trapezoidal filtering method, based on a moving average window, was used to achieve higher quality performance in energy resolution (δE/E) compared to simply integrating the charge over the entire pulse, i.e., we observed an improvement of 1–2% in the full width at half maximum (FWHM) between the two methods. The energy filter algorithm allows the user to set a peaking time (P) and gap time (G) that builds two sums over P, separated by G, from the raw ADC input values, resulting in a trapezoid [15]. This is a standard method used in digital signal processing, allowing the user to optimize the trapezoid shape and flat top to obtain the best performance for a given detector. An additional parameter, the decay correction (α), was implemented to ensure that the trapezoid was symmetric and returned to baseline. PSD is performed by allowing the user to tune programmable gates (accumulators) that can be set to integrate the input pulse over a given region for a given length of time. Offline PSD analysis was performed using the traditional charge integration method [16] given by the following expression:
The CLYC-7 scintillation detector was irradiated with highenergy neutron beams at the Crocker Nuclear Laboratory (CNL) located on the campus of the University of California – Davis in March of 2015. The CNL facility houses a 7600 -isochronous cyclotron, which is capable of accelerating charged particles (protons, deuterons, α particles, and helions) from several 10 s of MeV to greater than 100 MeV. Fast neutrons are produced via a spallation reaction between accelerated protons and a thick beryllium (Be) target: 9Be(p,n)9B (Q value ¼ 1.85 MeV). The Be target was encapsulated in a copper (Cu) housing with a front thickness of 0.51 mm, through which the protons traversed before striking the Be target, and a rear thickness of 6.35 mm. The proton beam energies used were: Ep ¼20.4 MeV, 30 MeV, 45 MeV, and 67.5 MeV (highest proton energy achievable). The specific energy loss (dE/ dx) due to the proton passing through the Be target and the Cu foil, as determined via SRIM simulations [12], and the Q value of the reaction lead to a continuum neutron spectrum with endpoint energies of En ¼ 5.9 MeV, 18.7 MeV, 36.1 MeV, and 60.5 MeV. The beam current used at each energy was 200 nA, 50 nA, 200 nA, and 80 nA–100 nA, respectively. A steel collimator (length ¼ 1.6 m) forms an on-axis beam of neutrons that exits from an aluminum window 2.85 m away from the target. The expected beam flux at the exit was on the order of 104n cm2 s 1 at 30 MeV and above; the flux at lower energies is more difficult to ascertain [13]. Data from 12.5 MeV and 14.6 MeV have yielded measured beam fluxes of 30 n cm2 s 1 and 90 n cm2 s 1, respectively. The 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillator cube (enriched to 99.93% 7Li) is housed on five sides in 3.175-mm-thick aluminum with a 23 mm 23 mm 3.1 mm quartz window on the sixth side (Fig. 1). The scintillator is wrapped with Teflon tape and supported on the top and bottom with 1-mm-thick optical pressure pads. The light-readout device is coupled to the scintillator volume at the quartz window. For this campaign, we used a 51 mm (diameter) Hamamatsu super bialkali photomultiplier tube
where QS (QI) is the short (intermediate) integration gate length. The length of the long integration gate, QL, was set to integrate the total charge. The bias voltage on the detector was provided by one channel of a 32-channel high-current ISEG high voltage supply [17] powered via a Wiener MPOD mixed crate [18]. At CNL, the CLYC-7 detector was placed 2 m from the beam exit to achieve high statistics in a reasonable amount of time given that the efficiency to fast neutrons is on the order of a few tenths of a percent at a few MeV and increases to 3% at 20 MeV [1,10]. Extrapolating to 60 MeV, one would expect a continued increase in efficiency. To reduce the γ-ray flux from the spallation reaction, lead bricks (total thickness of 10 cm) were placed in front of the beam exit.
Fig. 1. The 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector packaged in custom-designed aluminum housing.
Fig. 2. The CLYC-7 scintillator (white arrow) shown with 51 mm super bialkali Hamamatsu PMT set-up at the Crocker Nuclear Laboratory experiment room.
1
QS QI
ð1Þ
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
3. Results 3.1. Calibration The detector output read into the FADC was tuned to achieve optimal δE/E and calibrated using standard radioactive check sources in the laboratory prior to the test campaign. Using a 137Cs source and the PMT operating at a positive bias voltage of 1300 V, the δE/E at 662 keV was found to be 6.2% (FWHM) with P ¼3.4 μs, G ¼0.3 μs and α ¼1.5 μs (Fig. 3). The performance was unexpected given that this detector demonstrated a δE/E of 4.4% (FWHM) at 662 keV when using an analog shaping amplifier and multichannel analyzer (MCA) for data acquisition. We determined that the memory allotted by the SIS3316 firmware for shaping the pulse greatly affects scintillators that require a long shaping time. This sample of the CLYC-7 scintillator required an analog shaping time of 10 μs to achieve optimal performance. The length of the long gate (QL), which integrates over the entire pulse, was 10 μs, but the energy filter only allowed for a full range of 4 μs. (An updated version of the SIS3316 firmware allots more memory for the energy filter to accommodate a full range of greater than 10 μs. Work to improve the energy performance is currently under investigation). At CNL, to accommodate the dynamic range for large amplitude pulses produced by 60.5 MeV neutrons, the high voltage on the PMT was lowered to 1150 V, thus reducing the response at low energies. A recalibration at this voltage was not completed. However, by comparing the shape of long background runs completed at each voltage for equal run times, we find that they differ by a multiplicative gain factor. This scaling factor, A, is then applied to the two-point linear energy calibration obtained at 1300 V using 60 Co data (1173 keV and 1333 keV) to correct the calibration for data obtained at the lower voltage. The two-point calibration was linear; for higher energies, we assumed that the calibration maintained linearity. The calibration curve is given as follows:
49
optimization was done in the laboratory prior to the campaign with the PMT operating at a positive bias of 1300 V. PSD event selection criteria for subsequent analysis with the PMT operating at 1150 V were determined from the neutron beam data. A measure of the PSD performance is given by the Figure of Merit (M), determined by the ratio of the difference in centroids of the γ-ray and neutron distributions to sum of the FWHM of each distribution. Gate widths of 0.08 μs (QS) and 1.2 μs (QI) with QI starting 0.12 μs after the start of QS were used to achieve an M of 2.4 for electron-equivalent energies greater than 750 keVe.e. (for reference, this electron equivalent energy corresponds to a pulse height channel of 13 in Fig. 4). The CLYC-7 scintillation detector was exposed to the four neutron beams for a typical period between 300 s and 900 s. The run times were determined by the amount of beam time allotted for the experiment and to ensure that Gaussian statistics would be achieved for the total number of counts registered by the detector. The total rate measured by the detector at each energy was 70 Hz (5.9 MeV), 160 Hz (18.7 MeV), and 3 kHz (36.1 MeV and 60.5 MeV). 3.2. Neutron-induced reactions in CLYC-7 Figs. 5–8 show the 2-d scatter plots of the calculated pulse shape vs. pulse height acquired at each beam energy. The pulse shape was calculated by Eq. (1), and the pulse height represents the energy deposited by the ionizing particle; both scales have been rebinned in terms of 28 bins. For reference, an electron equivalent energy of 1.5 MeVe.e. corresponds to a pulse height channel of 25. Investigation of these standard scatter plots reveal the different reaction channels available from neutron-induced reactions in the CLYC-7 with increasing neutron beam energy. The following subsections discuss our analysis to determine which reactions were occurring and their attribution.
where Ee.e. is the electron-equivalent energy (in keV), A is 0.63, and x is the channel. To optimize the PSD, we exposed the detector to a 160 μCi 252Cf laboratory source, located behind 15-cm-thick lead to reduce the intense γ-ray contribution from the source (Fig. 4). The
3.2.1. Neutron beam energy: 5.9 MeV For the neutron beam with 5.9 MeV endpoint energy, the pulse shape vs. pulse height shown in Fig. 5 displays the well-separated bands corresponding to electrons (lower band) and protons (upper band). The proton band results from the reaction 35Cl(n,p)35S (Q¼ þ 615 keV), which is dominant in the 1 MeV–10 MeV energy range due to its large cross-section and the fact that 35Cl is the
Fig. 3. A 10 μCi 137Cs source irradiated the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector read out by a 51 mm super bialkali Hamamatsu PMT and one channel of the 16-channel flash ADC. The FWHM of the 662 keV photopeak is 6.2%. For comparison, the same detector setup read out by analog shaping electronics measured the 662 keV photopeak with a FWHM of 4.4%.
Fig. 4. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated with a 160 μCi 252Cf γ/n source.
Ee:e: ¼ Að1:12 10 2 x þ 1:25Þ
ð2Þ
50
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
most abundant isotope in the CLYC-7 scintillator, as previously reported by D'Olympia et al. [1]. An additional proton-producing reaction is from 35Cl(n,p)35S*, i.e., excited states of the 35S. Starting 4 MeV and above, the cross-section for the 1st and 2nd excited states of 35S from the (n,p) reaction on 35Cl is comparable to the cross-section of the ground state reaction ( 5 10 2 b). These excited state reactions ultimately contribute to the counts in the proton band. Fig. 5 also demonstrates that the thermal neutron contribution from the 6Li reaction is negligible. If the scintillator was not depleted of this isotope, as shown in Ref. [7], a strong thermal neutron capture peak, independent of the incident neutron energy, would appear at a fixed energy in the proton band in pulse shape space. 3.2.2. Neutron beam energy: 18.7 MeV Fig. 6 shows the pulse shape vs. pulse height for the neutron beam with endpoint energy of 18.7 MeV. At this energy, we still observe the two bands seen at the 5.9 MeV endpoint; however, note the emergence of some structure in the upper band. The bottom part of the upper band is attributed to the neutron-capture protons given that these events follow the same trend (in terms of the mean of the rebinned pulse shape value) as observed at lower energies but have increased pulse height due to the increased endpoint energy of the beam. This fact alone demonstrates the utility of the CLYC-7 detector as a spectrometer. The top part of the upper band begins to show splitting above channel 40. The bands become more defined and separated as the incident neutron energy is increased. 3.2.3. Neutron beam energy: 36.1 MeV and 60.5 MeV Continuing with the trend observed at 18.7 MeV, for the higher energy neutron beams we observed the following in the scatter plots (Figs. 7 and 8): the well-defined γ-ray-related band as seen in previous energies (rebinned pulse shape value ¼70); three welldefined, intense bands between rebinned pulse shape values of 110–140; a less-intense but still defined band at rebinned pulse shape value 125; and hints of two separate bands with rebinned pulse shape value centroids of 95 and 185. This was the first time observing the fine banding structure in what is typically thought of as the proton only band as no data at this neutron energy existed for CLYC-7 at the time of the measurement. To determine the attribution of these bands, we began investigating all possible neutron-induced reactions that could occur for the
Fig. 5. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 5.9 MeV neutron beam.
Fig. 6. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 18.7 MeV neutron beam.
Fig. 7. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 36.1 MeV neutron beam.
Fig. 8. The pulse shape vs. pulse height 2-d scatter plot of the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 60.5 MeV neutron beam. This plot is annotated to indicate the reaction channels and accidental events observed.
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
given isotopes that comprise the CLYC-7 scintillator. The 35Cl reaction is the main attribute for the (n,p) reaction that leads to fast neutron detection at lower energies. Other reactions to consider are those produced by light ionizing particles, such as the deuteron via (n,d) reaction, the triton via (n,t) reactions, and α particles via (n,α) reactions. Important considerations for each of these reactions are the cross-section for a given energy range, the reaction threshold energy, the Q value (exothermic or endothermic), and the isotopic abundance. Based on previous work studying the relationship of the light output (dL/dx) and dE/dx [19,20], we know that charged particles heavier than the proton (deuteron, triton, alphas, 3He, carbons, etc.) populate a greater proportion of excited states in the constituent scintillation material, thus producing an increase in the slow component of the scintillation light and hence leading to longer decay times back to baseline. Because our method of PSD is a measure of the total integrated charge, longer decay times lead to larger values for the pulse shape. Similarly, the dL/dx is related to dE/dx, which in turn is proportional to Z2. Comparing the dL/dx for protons to deuterons and tritons, because Z¼1 in all cases, any differences observed are attributed to the Q value of the reaction. Aside from the Q value, for α particles, the charge state of Z¼ 2 leads to a reduction in the light output due to quenching effects from the radiation-less loss of energy when populating more excited states. Based on previous results with other scintillators exposed to light ions, one would start with the ansatz that the four main bands in Figs. 7 and 8 between rebinned pulse shape values 110–140 correspond to protons, deuterons, tritons and α particles in order of increasing value. For proton-producing reactions, we investigated (n,p), (n,np), and (n,2n) reactions. We know from the lower energy beam data that the lowest part of the upper band corresponds to protons. At these energies, the reactions considered for the isotopes associated with CLYC-7, in addition to the 35Cl(n,p)35S reaction, were 37 Cl(n,p)37S, 133Cs(n,p) 133Xe, 89Y(n,p)89Sr, 140Ce(n,p)140La, and 142Ce (n,p)142La. Cross-section data were obtain using the NNDC ENDF database [21]; for the reactions listed here and elsewhere in the manuscript, we required that the cross-section data have endpoint energies of at least 60 MeV. The only currently available values of the cross-sections for the isotopes in question are based on the TENDL (TALYS-based Evaluated Nuclear Data Library) 2009 and 2014 numerical models [22]. TALYS is a software modeling package used for the simulation of nuclear reactions up to 200 MeV, and specifically treats protons, deuterons, tritons, and α particles as both the target and ejectiles [23]. For the energy region of 20 MeV– 60 MeV, the listed reactions have comparable cross-sections with values ranging from 3 10 2 b to 3 10 3 b (see Fig. 9a). The reactions involving 133Cs and 35Cl are exothermic with Q values of 355 keV and 615 keV, respectively, whereas the reactions involving 37 Cl, 89Y, 140Ce, and 142Ce are endothermic with Q values ranging from 718 keV to 4.1 MeV. Fig. 9a clearly shows that for incident neutrons with energies less than 20 MeV, the 35Cl(n,p) reaction is dominant, as discussed in Section 3.2.1. For energies greater than 20 MeV, the other constituents begin to play a more dominant role in producing protons from the neutron capture reaction. The (n,np) and (n,2n) reactions are typically endothermic with Q values between 5 MeV to 10 MeV. Regardless, these reactions lead to either additional neutrons or a neutron and a proton, which would show up as a lower-energy proton in pulse shape space or could lead to some additional reactions at lower energies. Therefore, one can conclude that the observed proton band is caused by a combination of the listed reactions. The endpoint of the pulse height spectrum serves as a calibration point for proton equivalent energy, once the highest Q value is accounted for. Beginning with the proton band, the second band encountered with increasing rebinned pulse shape value corresponds to deuterons from (n,d) reactions. The reactions of interest for (n,d) are 133 Cs(n,d)132Xe, 89Y(n,d)88Sr, 35Cl(n,d)34S, and 37Cl(n,d)36S with Q
51
values ranging from 3.85 MeV to 6.1 MeV. The reaction crosssection for these isotopes (Fig. 9b) begin to turn on 10 MeV and are comparable in the 20 MeV–60 MeV energy range with values varying between 5 10 1 b to 1 10 2 b. At 20 MeV the Cl isotopes are the main deuteron contributors; above 30 MeV the 89 Y(n,d) reaction yields the largest cross-section and is greater than the Cl isotopes by less than a factor of 2. Compared to the 133 Cs, the 89Y(n,d) reaction cross-section is greater by a factor of 3 at 30 MeV then reducing to a factor of 2 at 60 MeV. Aside from the (n,d) reaction, other deuteron-producing reactions are (n, nd) or (n, pd). As was the case with the proton-producing reactions, these events would either give rise to additional lowerenergy neutrons or register in the lower-energy region of the proton/deuteron pulse shape space. The third band encountered with increasing rebinned pulse shape values in Figs. 7 and 8 is attributed to neutron-induced tritons from (n,t) reactions. The main reaction of interest is 35Cl(n, t)33S with a cross-section of 1 10 2 b in the 20 MeV region and falling off by an order of magnitude at 60 MeV (Fig. 9c). Compared to the (n,p) and (n,d) reactions, there is roughly an order of magnitude reduction in cross-section at the higher energies, which contributes to the reduction in the observed counts in this band. The Q value for this endothermic reaction is 9.3 MeV, requiring that this amount of energy be removed by the reaction products, which leads to a shift in the pulse height spectrum for incident neutrons of the same energy. Another reaction to consider in the 20 MeV range is 7Li(n,t)5He. The cross-section is 1 10 2 b at 20 MeV but then falls off precipitously (by three orders of magnitude) at 60 MeV. This reaction would lead to an increase in the signal at lower energies where the bands are not well separated. The fourth band observed in Figs. 7 and 8 is attributed to neutroninduced α-particle reactions (n,α). The reactions of 35Cl(n,α)32P, 37Cl(n, α)34P,133Cs(n,α)130I, and 140Ce(n,α)137Ba have the largest crosssections at 20 MeV, ranging from 5 10 2 b to 1 10 2 b; at 60 MeV, the cross-section for these reactions has dropped to 5 10–4 b, with the 35Cl(n,α) reaction remaining the largest over the entire energy range of interest (Fig. 9d). One can see the effect caused by the reduction in cross-section for increasing incident neutron energy in Figs. 7 and 8 where the α particle band becomes less intense, in comparison to the (n,p) and (n,d) bands. We conclude that the dominant source of α particles is from the 35Cl(n,α) reaction, which is exothermic with a Q value of þ937 keV. Lastly, as discussed at the beginning of this section, two additional banding features appear in the 2-d scatter plot in Figs. 7 and 8: one clustered above the main upper band (rebinned pulse shape value¼185) and a band located between the electron and proton bands (rebinned pulse shape value¼ 95). These bands appear in the data at the highest beam energies and are nearly absent at lower energies. For the band clustered above the main upper band, at first glance, one may surmise this is the result of another ionizing particle, one with a smaller cross-section (due to the number of observed counts) and higher dE/dx (due to its location in pulse shape space). However, closer inspection reveals that these events result from accidentals. A 1-d histogram (Fig. 10) of the pulse shape parameter – obtained by collapsing the y-axis of the 2-d scatter plot and applying cuts between 6.5 MeVe.e.–8.5 MeVe.e. – show each band and their relative contribution. The γ-ray-induced events have mainly been filtered out by the energy cuts; the main neutron-induced reactions are shown accordingly. The expression used to calculate the PSD is shown in Eq. (1). Typical values observed for the electrons, protons, deuterons, tritons, and alphas range from 0.80 to 0.91. The values for this band peak are around unity. (For clarity on the 2-d plots, a rebinned pulse shape value of 256 is equal to 1.1). Values around unity indicate that the integrated pulse within the short gate is much less than the integrated pulse within the intermediate gate window (QS {QI). These events stem from accidental triggers in the digitizer
52
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
Fig. 9. a: Cross-section values vs. incident energy for relevant (n,p) reactions. b: Cross-section values vs. incident energy for relevant (n,d) reactions. c: Cross-section values vs. incident energy for relevant (n,t) reactions. d: Cross-section values vs. incident energy for relevant (n,α) reactions.
system; the rate of accidentals is dependent on the input event rate, the pulse amplitude and timing sequence. For the accidentals we observe in the 2-d plot, where QI registers a value but QS does not (or is very small in comparison), these events can be attributed to a large amplitude pulses followed by secondary pulses occurring at the end of the integration window, and causing these accidental events to occur in the subsequent triggered event. We calculate the accidental (pile up) rate expected for the gate settings used and the total rate, as given by equation 3 [19], where rch is the rate of accidental events, rs is the source rate, and tr is the resolving time. r ch ¼ r 2s t r
Fig. 10. The 1-d histogram of the pulse shape for the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillation detector irradiated by a 60.5 MeV neutron beam at the Crocker Nuclear Laboratory. This plot is annotated to indicate the reaction channels and accidental events observed.
ð3Þ
For a total integration length of 10 μs and with 50 ns of dead time between pulses [15], tr ¼10.05 μs. For a source rate (rs) at the highest energies of 3 kHz, we find rch ¼ 90 Hz. For a 300-s run, the number of accidental events produced should be on the order of 3 104. By selecting out the accidental events in the 2-d scatter plot, we find a total on the order of 5 103 events. We note that the discrepancy is because not all events with pile up will show up in the accidental band; this band corresponds to the extreme case where QS { QI that was previously described. For comparison, at the lower energies, the rs has been reduced by more than an order
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
of magnitude. Correspondingly, the accidental rate expected is 0.049 Hz and 0.26 Hz at 5.9 MeV and 18.7 MeV, respectively. For the 900-s run completed at 5.9 MeV, Eq. (3) yields 40 events in an accidental band. Selecting on events above the proton band yields one accidental event, which was not unexpected considering the low rate and amplitude of pulses at this energy. For a 600-s run completed at 18.7 MeV, Eq. (3) yields 160 events in the accidental band. Selecting on events above the proton band yields 15 accidental events. Again, this result is not unexpected given the ratio of observed-to-calculated accidentals at 36.1 and 60.5 MeV and the fact that the rates and amplitudes of the pulses are much lower than at higher energies. The band located between the electron and proton bands is attributed to the neutron-capture protons escaping from the scintillator volume without fully depositing their energy [24]. Similar results have been observed in organic scintillators [25]. Figs. 8 and 10 have been annotated to illustrate the particles attributed to each of these bands.
53
3.3. Equivalent energy relationship One of the main goals of this experiment was to determine the light output response of protons from high-energy neutrons and relate this quantity to the light output of known-energy electrons from γ-ray-induced events. With the emergence of the additional reaction channels, we then wanted to additionally obtain a relation for the heavier particles. A measure of the γ-ray response comes from the calibration discussed in the last section (Eq. (2)), relating the light output from electrons to known γ-ray energy (Ee.e.). The light output response for heavier ions was obtained by first making 2-d selections on the pulse shape vs. pulse height, then reconstructing the resulting pulse height spectra of those events. Fig. 11a–d shows the pulse height spectra for each particle band. To determine the point in the spectrum that corresponds to the maximum energy deposited in the scintillator by the ion, we used a method analogous to one used previously in measurements with organic scintillator where the half height of the abrupt edge near the maximum pulse height corresponds to the maximum energy [26,27]. Depending on the reaction, one must also assign the proper reaction Q value to account for the particle’s total energy.
Fig. 11. a: The proton-selected pulse height spectra resulting from the 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillator irradiated by the four differing energy neutron beams (5.9 MeV, 18.7 MeV, 36.1 MeV, and 60.5 MeV). b: Pulse height spectra for deuteron-selected events at the two energies (36.1 MeV and 60.5 MeV) in which this band was most prominent. c: Pulse height spectra for triton-selected events at the two energies (36.1 MeV and 60.5 MeV) in which this band was most prominent. d: Pulse height spectra for α-particle-selected events at the two energies (36.1 MeV and 60.5 MeV) in which this band was most prominent.
54
R.S. Woolf et al. / Nuclear Instruments and Methods in Physics Research A 803 (2015) 47–54
Table 1 Energy equivalence relationship between protons and electrons. The first column corresponds to the neutron beam energy, whereas the second column represents the beam energy with the Q value of the reaction taken into account. The third column shows the value, in terms of ADC channel, representing the location in the pulse height spectrum that corresponds to the maximum energy deposit. The last column is the electron equivalent energy as calculated by Eq. (2). Neutron energy, En (MeV)
Proton energy, Ep (MeV)
Maximum energy deposit location (ADC channel)
Electron equivalent energy, Ee.e. (MeV)
18.7 36.1 60.5
19.3 36.7 61.1
616,994 1,429,620 2,190,760
4.4 10.1 15.5
Table 2 Energy equivalence relationship between deuterons and electrons. The first column corresponds to the neutron beam energy, whereas the second column represents the beam energy with the Q value of the reaction taken into account. The third column shows the value, in terms of ADC channel, representing the location in the pulse height spectrum that corresponds to the maximum energy deposit. The last column is the electron equivalent energy as calculated by Eq. (2). Neutron energy, En (MeV)
Deuteron energy, Ed (MeV)
Electron Equivalent Maximum energy deposit location (ADC Energy, Ee.e. (MeV) channel)
36.1 60.5
32.3 56.7
1,477,130 2,214,960
10.5 15.7
Figs. 11a shows the reconstructed pulse height spectra for protons at each neutron beam energy. The previously described method for finding the location (in terms of ADC channel) of the maximum energy deposit was used at 18.7 MeV, 36.1 MeV, and 60.5 MeV. For 5.9 MeV neutrons, no abrupt edge was observed. The corresponding ADC channel was input into Eq. (2) to obtain an electron equivalent energy (Ee.e.) for a given proton energy. The same method was applied to the deuteron-selected pulse height spectra shown in Figs. 11b. The main results are outlined in Tables 1 and 2, which show: the neutron beam energy, the particle energy corrected for the reaction Q value, the ADC channel corresponding to the location of maximum energy deposited by the particle, and the electron equivalent energy derived from Eq. (2). The Ee.e. is comparable for protons and deuterons except for the fact that the deuteron energy is lower than that of the proton due to the Q value. For the neutron-induced triton and α-particle reactions, the number of events collected in each band was not sufficient to adequately produce pulse height spectra that demonstrated an abrupt edge. As shown in Fig. 11c and d, the resultant spectra follow an exponential fall off with no distinct features. Without an abrupt edge to determine the maximum energy deposit by the ion, we were unable to obtain a relationship for the light yield produced by tritons or α particles and how that related to the electron light yield. At the least, the endpoint observed in the pulse height spectra places a lower limit on the relationship between triton/αparticle energy and the electron equivalent light yield.
4. Conclusions This work presents the results from a campaign conducted at the Crocker Nuclear Laboratory in Davis, CA, in which we irradiated a 25.4 mm 25.4 mm 25.4 mm CLYC-7 scintillator with 6 MeV– 60 MeV neutron beams. Prior work investigated monoenergetic neutron beams characterizing CLYC-7 at o20 MeV. Thus, it was our goal to determine the response of this novel scintillator for
420 MeV neutrons. For the highest beam energies tested (36.1 MeV and 60.5 MeV), in addition to the standard electron/proton bands expected in a mixed γ/n field, we observed splitting of the traditional proton band in pulse shape space attributed to high-energy reaction channels becoming available for the production of deuterons, tritons, and α particles. We also observed escaping protons from the scintillator volume and an increase in the accidental events. Based on these new-found results, we determine the heretofore undetermined light output equivalent above 20 MeV relating light output for protons and deuterons to equivalent energy electrons. Unfortunately, the statistics in the triton/α-particle bands did not allow for a determination of the light output relationship between these particles and the equivalent energy electrons. The results presented establish the previously unknown response to highenergy fast neutrons with CLYC-7 and show that this scintillator can be used as a spectrometer based on not only the proton reaction but also the other reactions channels that appear at higher energy. The heavier ion reactions lead to the ability to perform spectrometry over a larger dynamic energy range. Characterization and understanding the properties of the CLYC-7 scintillator are important for future implementation in the field of solar physics or other applications where high-energy neutron fields are encountered.
Acknowledgments This work was sponsored by the Chief of Naval Research (CNR). The authors would like to thank the following people: Spencer Hartman from the Crocker Nuclear Laboratory for his assistance during the neutron beam campaign; Joshua Tower from Radiation Monitoring Devices for loaning our group the super bialkali PMT; J. Eric Grove and Lee Mitchell from the Naval Research Laboratory for useful discussions on data analysis.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]
N. D’Olympia, et al., Nuclear Instruments and Methods A 716 (2014) 433. C.M. Combes, et al., Journal of Luminescence 82 (1999) 299. E.V.D. van Loef, et al., Journal of Physics: Condensed Matter 14 (2002) 8481. A. Bessiere, et al., Nuclear Instruments and Methods A 537 (2005) 242. J. Glodo, et al., IEEE Transaction on Nuclear Science NS55 (3) (2008) 1206. M. Bourne, et al., Nuclear Instruments and Methods A 736 (2014) 124. N. D’Olympia, et al., Nuclear Instruments and Methods A 694 (2012) 140. J. Glodo, et al., IEEE Transaction on Nuclear Science NS60 (2) (2013) 864. M. Smith, et al., IEEE Transaction on Nuclear Science NS60 (2) (2013) 855. M. Smith, et al., Nuclear Instruments and Methods A 784 (2015) 162. J. Ryan, et al., Advances in Space Research 13 (9) (1993) 255. SRIM – the Stopping and Range of Ions in Matter. 〈http://www.srim.org〉, 2013. S. Hartmann, in Private Communications, 2015. Struck Innovative Systeme, SIS3316-250-14 16 channel 250 MSPS 14-bit. 〈http://www.struck.de/sis33166.html〉, Online 2015. SIS3316, 16 Channel VME Digitizer, User Manual, version: SIS3316-M-1-1V110, Online 2015. F.D. Brooks, Nuclear Instruments and Methods 4 (1959) 151. ISEG Germany, EDS. 〈http://www.iseg-hv.com/en/products/product-details/ product/20/〉. Wiener Plein and Baus Elektronik, MPOD Mixed Crate. 〈http://www.wiener-d. com/sc/power-supplies/mpod–lvhv/mpod-mix-crate.html〉. G.F. Knoll, Radiation Detection and Measurement, 3rd ed., Wiley, New York, 2000. V. Avdeichikov, et al., Nuclear Instruments and Methods A 484 (2002) 251. ENDF/B-VII.1. 〈http://www.nndc.bnl.gov〉. A.J. Koning, et al., Nuclear Data Sheets 113 (2012) 2841. A.J. Koning, et al. 〈http://www.talys.eu/tendl-2014.html〉. M. Moszynski, et al., Nuclear Instruments and Methods A 343 (1994) 563. A. Comrie, et al., Nuclear Instruments and Methods A 772 (2015) 43. V. Verbinski, et al., Nuclear Instruments and Methods 65 (1968) 8. R.S. Woolf, et al., Nuclear Instruments and Methods A 784 (2015) 80.