New Materials for Fast Neutron Detection

Authors: L. Adamowski ¹, K. Brylew ¹, J.S. Carlson ², P.L. Feng ³, M. Grodzicka-Kobylka ¹, M. Moszyński^† ¹, L. Nguyen ³, T. Szczesniak ¹, L. Swiderski ¹, J.J. Valiente-Dobón ^4,5

¹) National Centre for Nuclear Research (NCBJ), Otwock, Poland; ²⁾ Blueshift Optics LLC, Oakland, USA; ³⁾ Sandia National Laboratories, Livermore, USA; ⁴) Laboratori Nazionali di Legnaro, INFN, Legnaro, Italy; ⁵) Instituto de Física Corpuscular, CSIC-Universidad de Valencia, 46980, Valencia, Spain

Growing demand for large size and robust fast neutron detectors with neutron/gamma-ray (n/g) discrimination capability has been a driving force for development of organic solid-state scintillators. In recent years, a couple of new materials appeared like EJ-276 plastic (from Eljen Technologies), trans-stilbene (from InRad Optics), polyurethane based M-600 (from Target SystemElektronik) or Organic Glass Scintillator (OGS – from Blue Shift Optics). A performance study of these materials compared to a benchmark liquid scintillator EJ-309 was published in [1]. Amongst presented materials, trans-stilbene shows superior neutron/gamma discrimination, however, high light yield (LY) and faster response time of OGS with n/g discrimination capability almost as good as in EJ-309 liquid scintillator provides a reasonable solid-state alternative at reduced production cost. Possibility of casting large size, custom shape samples is another advantage for using OGS in fast neutron detection.

Optimization of gate selection for n/γ discrimination

As a continuation of this study, the influence of gate selection in charge comparison method on the n/g discrimination capability was investigated. Charge comparison method is one of the simplest approaches to make use of difference in pulse shapes induced by various types of radiation in order to identify the type of detected particles or quanta. Fig.1 shows typical gamma-ray and neutron induced pulses recorded with a cylindrical, 1 inch diameter and height OGS sample. Examples of gates used for pulse shape discrimination (PSD) are marked with relevant rectangles.

Fig. 1: An example of gamma and neutron induced pulses in OGS recorded by a digital
analyzer. The corresponding CCM parameters are: long gate, short gate and pre-gate.

The signals are integrated in short and long gates. Denoting Q as the integral of a pulse in relevant gate, PSD parameter is defined as:

$\text{PSD} = \frac{Q_\text{long} - Q_\text{short}}{Q_\text{long}}$

Next, PSD is plotted as a function of $Q_\text{long}$ , which is proportional to energy deposited in the scintillator. The resulting energy dependent PSD histogram in a form of a 2D heatmap is presented in Fig.2. The pre-gate determines the start point of long- and short gate. The baseline is calculated as the average of pulse heights within the baseline gate.

Fig. 2: An example of 2D heatmap of the PSD parameter versus total charge (energy
deposition) for OGS measured with a PuBe neutron source. Two branches of neutron and
gamma-ray induced events are clearly separated.

To quantify n/g discrimination capability, figure of merit (FoM) is usually proposed. One of the most popular estimates is built by selecting part of entire energy rangy of the 2D heatmap and projecting data onto the PSD axis. As a result, neutron and gamma-ray induced events, if separated, form two peaks in the histogram, see Fig. 3.

Fig. 3: The projection of the 2D heatmap onto the PSD axis at (500+/-50) keVee energy.

Then, FoM is defined as follows:

$\text{FoM} = \frac{\text{CTR}_\text{n} - \text{CTR}_\text{g}}{\text{FWHM}_\text{n} - \text{FWHM}_\text{\gamma}}$

The analysis presented in [2] was aimed at investigation of optimal set of parameters (readout triggering method and gate ranges) for optimizing the n/γ discrimination performance in OGS and trans-stilbene. Example of results obtained for OGS is presented in Fig.4.

Fig. 4: 3D plot of FOM values for OGS. Gates with maximum FOM value for a given
energy range are marked in red.

Influence of sample size on n/γ discrimination

OGS has relatively high scintillation LY, however it is also known to show some significant self-absorption. As light yield affects the n/g discrimination capability, we decided to investigate it for samples with different aspect ratio in terms of diameter versus height. Fig. 5 presents a set of cylindrical OGS samples with 1 inch diameter and height increased from 1 inch up to 5 inches. A 1 inch diameter and height trans-stilbene was also used for comparison.

Fig. 5: Photo of trans-stilbene (left) and five organic glass scintillators (right) used in this research.

As shown in Fig.6, the photoelectron yield, a quantity proportional to LY, drops down by a factor of 0.5% / mm in the case of 1 inch diameter OGS samples. The LY reduction rate is about 2 times higher than in the case of EJ276 plastic scintillator.

Fig. 6: The decrease of photoelectron yield of OGS (green circles) compared to EJ-276 (blue crosses) with the scintillator height. The yields were normalized to 100% at 25.4 mm (1 inch).

Despite the fact of light loss with increased height of the cylindrical samples, the FoM values maintain high values, showing a small decrease for events in energy range between 100 keV_ee and 1000 keV_ee, see Fig.7. The details of analysis are presented in [3].

Fig. 7: OGS FoM in wide energy range 100-1000 keVee decreases with scintillator height. Several measurements were made with every sample. An apparent single point on the graph for 55 mm sample is a result of three independent measurements, all with the same FOM value 1.6. The uncertainty of FoM values was estimated at 3% from standard deviation of all measurements.

Fast neutron detection performance of BSO-406 – further development of OGS

In response to issues with durability, OGS has been modified by adding polystyrene. This modified material is currently available from Blue Shift Optics under commercial name BSO-406. Investigation of fast neutron detection performance has been recently published in [4], where studies carried out with 2 inch diameter and height cylindrical samples are presented. Despite lower light yield and slightly slower response, BSO-406 offers high quality n/γ discrimination, at 100 keV_ee almost as good as pure OGS or even liquid EJ‑309, see Fig.8.

Fig. 8: Comparison of the Figure of Merit (FOM) for 2″×2″ scintillator samples: BSO-406, OGS, EJ-276, EJ-309, and M‑600.

References

M. Grodzicka-Kobyłka et al., “2 inch molecular organic glass scintillator for neutron–gamma discrimination”, Nucl. Instr. Meth. A Vol. 1047 (2023) 167702, DOI: https://doi.org/10.1016/j.nima.2022.167702.
Open access version: https://doi.org/10.48550/arXiv.2301.06766
Lukasz Adamowski et al., “Advantages of off-line analysis of digitally recorded pulses in case of neutron-gamma discrimination in scintillators”, submitted to Nucl. Sci. Tech.
Open access version: https://doi.org/10.48550/arXiv.2504.11963
Lukasz Adamowski et al., “”, submitted to ________
M. Grodzicka-Kobyłka et al., “Comparison of an OGS/Polystyrene scintillator (BSO-406) with pure OGS (BSO-100), EJ-276, EJ-309, and M600 scintillators”, submitted to Nucl. Instr. Meth. A.
Open access version: https://doi.org/10.48550/arXiv.2504.10072