Evaluation of pulse shape discrimination performance of scintillation materials and PSD methods by using statistical models

doi:10.1016/S0168-9002(98)00698-6

Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment

Volume 416, Issue 1, 11 October 1998, Pages 32-44

https://doi.org/10.1016/S0168-9002(98)00698-6 Get rights and content

Abstract

Two statistical models that include transit time spread and electronic noise effects are studied for evaluation of scintillation materials and for comparison of pulse shape discrimination (PSD) methods. The intrinsic PSD low-energy limitations for three scintillators are estimated by using these models, which can be used for PSD parameter optimization. Under the optimal conditions, the low-energy PSD limits of early NE213, trans-stilbene, and Borexino scintillators, are 100, 65, and 60 keV neutron energy, respectively. Calculated results show that under optimal conditions, the zero-crossing method is better for early NE213 and trans-stilbene scintillators, and the charge comparison is better for Borexino. The calculated best constant fraction discrimination (CFD) settings in the zero-crossing method for early NE213, trans-stilbene, and Borexino scintillators are 0.8, 0.8 and 0.9, respectively. The calculated best fast window settings in the charge comparison method are 0–40, 0–22, and 0–10 ns, respectively.

Introduction

Optimization of electronic components and pulse processing features for neutron-photon pulse shape discrimination (PSD) at low energies (<500 keV neutron energy) by using “trial-and-error” methods is a time consuming and difficult task. The experimental evaluation of scintillation properties of materials and the comparison of PSD methods are affected by many factors, such as the light coupling and equipment adjustment, and conflicting conclusions can be obtained. For example, Sabbah, et al. [1]mentioned two methods (zero-crossing and charge comparison), and they concluded that charge comparison is better than the zero-crossing method. Ranucci [2]evaluated PSD properties of scintillators with statistical analysis, and his results show that the charge comparison method offers superior PSD performance relative to the zero crossing approach. However, Wolski and Moszynski [3]experimentally show that the zero-crossing technique is superior to the charge comparison at low energies. In order to eliminate human and equipment errors, statistical models for calculation of intrinsic PSD performance characteristics of scintillators and of PSD methods are necessary, especially at low energies where the number of photoelectrons is small and statistical fluctuations are a dominant factor. Other advantages for using models to evaluate scintillation materials, to compare methods, and to optimize PSD parameters include time-savings and cost.

Present statistical models for the zero crossing and charge comparison methods are based on Ranucci's work [2], but they are extended by including transit time spread and electronic noise. The models are tested with characterization data for early NE213, trans-stilbene and Borexino scintillators, and results show that the transit time spread and electronic noise are important for determining PSD performance. The models are used for estimation of PSD low-energy limitation of three scintillators, for comparison of PSD methods, and for PSD parameter optimization.

Section snippets

Description of the models

Statistical models that include scintillation decay time, transit time spread, and electronic noise of the equipment are developed and programmed for zero-crossing and for charge comparison methods. Assumptions for the models are as follows:

1.
The light output of scintillators is proportional to energy deposition, and ambient conditions such as temperature are negligible. Photoelectrons produced from an energy deposition event are assumed to be Gaussian distributed in time when more than 30

Results and discussions

Computer codes are written to implement the previous statistical models under different conditions, such as scintillation materials, CFD values, time windows and transit time spread of PMTs. The slow component of the distribution is selected for computation because its photon and neutron peak locations are in the same order as in the rise-time distribution of the zero-crossing method, where the photon peak is on the left side of the neutron peak. This arrangement makes comparison more

Conclusions

Two statistical models, which simulate photoelectron pulse processing for the zero crossing and charge comparison PSD methods, are developed for evaluation of the PSD properties of scintillators and PSD methods. Transit time spread and electronic noise contributions are included in the models so that they represent real detection systems and can be used to predict the PSD performance of a given scintillation detector and associated electronics. Calculations shows that these two factors

Acknowledgements

The authors would like to acknowledge discussions and comments of Dr. N. Hill, Dr. P. Groer, Dr. L.W. Townsend, Dr. J. T. Mihalczo, Dr. D. J. Downing, and Dr. H. L. Dodds, and to thank Mr. G. L. Graves and Mr. R. J. Bailey for technical assistance.

References (5)

B. Sabbah et al.
Nucl. Instr. and Meth.
(1968)
G. Ranucci
Nucl. Instr. and Meth. A
(1995)

There are more references available in the full text version of this article.

Cited by (12)

Principal Component Analysis for pulse-shape discrimination of scintillation radiation detectors
2016, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
In this paper, we report on the application of Principal Component analysis (PCA) for pulse-shape discrimination (PSD) of scintillation radiation detectors. The details of the method are described and the performance of the method is experimentally examined by discriminating between neutrons and gamma-rays with a liquid scintillation detector in a mixed radiation field. The performance of the method is also compared against that of the conventional charge-comparison method, demonstrating the superior performance of the method particularly at low light output range. PCA analysis has the important advantage of automatic extraction of the pulse-shape characteristics which makes the PSD method directly applicable to various scintillation detectors without the need for the adjustment of a PSD parameter.
A comparison of digital zero-crossing and charge-comparison methods for neutron/γ-ray discrimination with liquid scintillation detectors
2015, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
Citation Excerpt :
The common methods of exploiting this property in scintillation detectors are the analog Charge-Comparison (CC) method which is based on the integrations of pulses over two different time intervals [2,3] and the analog Zero-Crossing (ZC) method [4,5] in which a zero-crossing time bears the pulse-shape information. A number of intercomparisons show that there are applications in which either may be preferable [6,7,8]. In recent years, there has been a strong interest in using digital techniques for n/γ discrimination applications and various digital algorithms, including the digital versions of the CC and the ZC methods have been used for this purpose [9,10].
In this paper, we have compared the performances of the digital zero-crossing and charge-comparison methods for n/γ discrimination with liquid scintillation detectors at low light outputs. The measurements were performed with a 2″×2″ cylindrical liquid scintillation detector of type BC501A whose outputs were sampled by means of a fast waveform digitizer with 10-bit resolution, 4 GS/s sampling rate and one volt input range. Different light output ranges were measured by operating the photomultiplier tube at different voltages and a new recursive algorithm was developed to implement the digital zero-crossing method. The results of our study demonstrate the superior performance of the digital zero-crossing method at low light outputs when a large dynamic range is measured. However, when the input range of the digitizer is used to measure a narrow range of light outputs, the charge-comparison method slightly outperforms the zero-crossing method. The results are discussed in regard to the effects of the quantization noise and the noise filtration performance of the zero-crossing filter.
Optimization of integration limit in the charge comparison method based on signal shape function
2014, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
A novel method is proposed to analyze neutron and gamma-ray signal shapes in liquid scintillation detectors. Specifically, the signal shape functions for a BC501 detector were characterized and a statistical model was used to analyze the discrimination of neutrons and gamma rays. The model varied the starting points of tail integration in the charge comparison method (CCM), and an optimized starting point was determined. Experimental measurements were performed to verify the model, and the results indicated good agreement. For a BC501 scintillator with 8.07 ns and 74.63 ns decay time constants we found optimal time to start the tail integration at 24 ns past the decay maximum.
Characterization of a tunable quasi-monoenergetic neutron beam from deuteron breakup
2007, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
Citation Excerpt :
The Stilbene detector is sensitive to both photons and neutrons. However, neutrons deposit energy in the detector through proton recoil rather than direct interaction with electrons [9]. Therefore, neutrons can be discriminated from photons by their longer pulse shape.
A neutron irradiation facility is being developed at the 88-inch cyclotron at Lawrence Berkeley National Laboratory for the purposes of measuring neutron reaction cross sections on radioactive targets and for radiation effects testing. Applications are of benefit to stockpile stewardship, nuclear astrophysics, next generation advanced fuel reactors and cosmic radiation biology and electronics in space. The facility will supply a tunable, quasi-monoenergetic neutron beam in the range of 10–30 MeV or a white neutron source, produced by deuteron breakup reactions on thin and thick targets, respectively. Because the deuteron breakup reaction has not been well studied at intermediate incident deuteron energies, above the target Coulomb barrier and below 56 MeV, a detailed characterization was necessary of the neutron spectra produced by thin targets.
Neutron time-of-flight (TOF) methods have been used to measure the neutron spectra produced on thin targets of low-Z (titanium) and high-Z (tantalum) materials at incident deuteron energies of 20 MeV and 29 MeV at 0°. Breakup neutrons at both energies from low-Z targets appear to peak at roughly half of the available kinetic energy, while neutrons from high-Z interactions peak somewhat lower in energy, owing to the increased proton energy due to breakup within the Coulomb field. Furthermore, neutron spectra appear narrower for high-Z targets. These centroids are consistent with recent preliminary proton energy measurements using silicon telescope detectors conducted at LBNL, though there is a notable discrepancy with spectral widths. Prospects for producing a tunable, quasi-monoenergetic neutron facility of 10⁶–10⁸ n/cm²/s at LBNL are promising.
Identification of shielded neutron sources with the liquid scintillator BC-501A using a digital pulse shape discrimination method
2007, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
Citation Excerpt :
This difference in the intensity of slow neutron and γ-ray components serves as a basis for PSD methods used to identify and characterize pulses. Several analog methods for PSD have been developed and studied in the past; examples are charge-comparison and rise-time methods [6,11,12]. Two main trends in the improvement of the PSD properties of liquid scintillators for applications in high count-rate fields have been pursued.
Fast and unambiguous identification of shielded neutron sources is of paramount importance in nuclear nonproliferation and international safeguards applications. Usually, the identification of neutron spectra relies on unfolding procedures that are unstable and time consuming. In this paper, we present a new application of an existing technique, which can be used for neutron source identification by direct analysis of the pulse height spectrum, i.e., without resorting to energy spectrum unfolding. This technique is based on the acquisition of neutron and γ-ray pulses by using a liquid scintillation detector and a fast waveform digitizer. An optimized digital pulse shape discrimination method based on standard charge integration is used to discriminate neutrons from γ-rays. The neutron pulses are then analyzed to generate a pulse height distribution. The accuracy of the neutron source identification technique was tested on three neutron sources: Cf-252, Am–Be, and Am–Li. Several source–shielding configurations were tested to assess the influence of potential shielding of the source and the sensitivity of the technique. For this investigation, lead and polyethylene shielding blocks were used. The measured pulse height distributions were compared with the distributions simulated with the MCNP-PoliMi code, and very good agreement was obtained. The results show that for both shielded and unshielded configurations the identification of Cf-252, Am–Be, and Am–Li shielded sources is easily achievable by direct analysis of the measured pulse height distributions, without the need of subsequent neutron energy spectrum unfolding.
Recent developments in neutron detection
2000, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
Developments in neutron detection technology during the past three years are reviewed with special emphasis on those technologies with known or possible application to safety, security, or industrial development. The technologies discussed include track detectors, image plates, activation detectors, bubble detectors, proton recoil instruments, time-of-flight instruments, moderating detectors, proportional counters, semiconductor detectors, and scintillation detectors. The review concentrates on advances in detection technology rather than on the use of existing methods for applications. For the purposes of this review, neutron spectrometry is considered to be a subset of the general area of neutron detection.

View all citing articles on Scopus

View full text