Boron-Based Neutron Scintillator Screens for Neutron Imaging

Journal of Imaging

Article Boron-Based Neutron Scintillator Screens for Neutron Imaging

William Chuirazzi 1,*, Aaron Craft 1, Burkhard Schillinger 2, Steven Cool 3 and Alessandro Tengattini 4,5

1 Advanced Post-Irradiation Examination Department, Idaho National Laboratory, Idaho Falls, ID 83401, USA; [email protected] 2 Heinz Maier-Leibnitz Zentrum (FRM II) and Faculty for Physics E21, Heinz Maier-Leibnitz Zentrum (FRM II), Technische Universität München, 85748 Garching, Germany; [email protected] 3 DMI/Reading Imaging, Reading, MA 01867, USA; [email protected] 4 Le Centre National de la Recherche Scientiﬁque (CNRS), Université Grenoble Alpes, Grenoble INP, 3SR, 38000 Grenoble, France; [email protected] 5 Institut Laue-Langevin (ILL), 71 Avenue des Martyrs, 38000 Grenoble, France * Correspondence: [email protected]

Received: 27 October 2020; Accepted: 16 November 2020; Published: 19 November 2020

Abstract: In digital neutron imaging, the neutron scintillator screen is a limiting factor of spatial resolution and neutron capture efficiency and must be improved to enhance the capabilities of digital neutron imaging systems. Commonly used neutron scintillators are based on 6LiF and gadolinium oxysulfide neutron converters. This work explores boron-based neutron scintillators because 10B has a neutron absorption cross-section four times greater than 6Li, less energetic daughter products than Gd and 6Li, and lower γ-ray sensitivity than Gd. These factors all suggest that, although borated neutron scintillators may not produce as much light as 6Li-based screens, they may offer improved neutron statistics and spatial resolution. This work conducts a parametric study to determine the effects of various boron neutron converters, scintillator and converter particle sizes, converter-to-scintillator mix ratio, substrate materials, and sensor construction on image quality. The best performing boron-based scintillator screens demonstrated an improvement in neutron detection efficiency when compared with a common 6LiF/ZnS scintillator, with a 125% increase in thermal neutron detection efficiency and 67% increase in epithermal neutron detection efficiency. The spatial resolution of high-resolution borated scintillators was measured, and the neutron tomography of a test object was successfully performed using some of the boron-based screens that exhibited the highest spatial resolution. For ome applications, boron-based scintillators can be utilized to increase the performance of a digital neutron imaging system by reducing acquisition times and improving neutron statistics.

Keywords: neutron imaging; neutron radiography; digital imaging; neutron scintillator; scintillator screen; boron scintillator; scintillator characterization; scintillator detection eﬃciency; epithermal neutron imaging; neutron sensor

1. Introduction

1.1. Background Neutron imaging is a nondestructive examination technique that measures neutron transmission to examine a sample’s internal structure. This technique has been used in a variety of studies, including those of nuclear fuels and materials [1–6], cultural heritage objects [7–9], fuel cells [10–14], turbine blades [15–17], and many others. Digital neutron imaging techniques are now an essential capability

J. Imaging 2020, 6, 124; doi:10.3390/jimaging6110124 www.mdpi.com/journal/jimaging J. Imaging 2020, 6, x FOR PEER REVIEW 2 of 22 J. Imaging 2020, 6, 124 2 of 22 essential capability for any state‐of‐the‐art neutron imaging facility [18]. The most common digital neutron imaging system consists of a digital camera optically coupled to a neutron‐sensitive for any state-of-the-art neutron imaging facility [18]. The most common digital neutron imaging system scintillator screen. consistsA critical of a digital component camera of optically digital neutron coupled imaging to a neutron-sensitive systems is the scintillatorneutron scintillator screen. , because the scintillatorA critical dictates component the neutron of digital capture neutron efficiency imaging and systems is a is limiting the neutron factor scintillator, in spatial because resolution. the scintillator dictates the neutron capture efficiency and is a limiting factor in spatial resolution. Common Common scintillator screens for thermal and cold neutron imaging utilize either a 6Li(n,α)3H reaction 6 α 3 scintillatordue to its large screens neutron for thermal absorption and cold cross neutron‐section imaging and relatively utilize eitherlarge areactionLi(n, )energyH reaction, or a dueGd(n,γ to its + large neutron absorption cross-section and relatively large reaction energy, or a Gd(n,γ + ICe ) reaction ICe−) reaction because of gadolinium’s high neutron absorption cross‐section [19–23]. Although− these becausescintillator of gadolinium’sscreens are the high state neutron of the art absorption in the digital cross-section neutron [imaging19–23]. Although community, these the scintillator continual screensimprovement are the state of neutron of theart scintillators in the digital are neutron necessary imaging to meet community, the ever the‐higher continual user improvement demands for of neutronincreased scintillators spatial resolution are necessary to toexamine meet the smaller ever-higher samples user demandsand increased for increased detection spatial efficiency resolution to todecrease examine measurement smaller samples time. and increased detection efficiency to decrease measurement time. Boron-based screens offer a promising alternative to existing 6Li- and Gd-based neutron scintillator Boron‐based screens offer a promising alternative to existing 6Li‐ and Gd‐based neutron screens for a number of reasons. First, the thermal neutron cross-section of 10B (3840 b) is approximately scintillator screens for a number of reasons. First, the thermal neutron cross‐section of 10B (3840 b) is four times greater than that of 6Li (980 b), as shown in Figure1, which should increase the neutron approximately four times greater than that of 6Li (980 b), as shown in Figure 1, which should increase detection efficiency for a comparable 6Li-based screen. Second, the daughter products created by the neutron detection efficiency for a comparable 6Li‐based screen. Second, the daughter products neutron absorption of 10B have less total energy (2.31 MeV or 2.79 MeV) than those of 6Li (4.78 MeV) created by neutron absorption of 10B have less total energy (2.31 MeV or 2.79 MeV) than those of 6Li 10 α 7 and(4.78 Gd MeV) (7.937 and MeV Gd or(7.937 8.536 MeV MeV), or and8.536 the MeV) lower, and energy the fromlower a energyB(n, )fromLi reaction a 10B(n,α) is distributed7Li reaction on is 6 α 3 heavierdistributed daughter on heavier products daughter compared products to a comparedLi(n, ) H to reaction. a 6Li(n,α) The3H lowerreaction energy. The lower and larger energy size and of thelarger daughter size of products the daughter reduces products the range reduces the daughter the range products the will daughter travel, products a fundamental will travel, limiting a factorfundamental of a screen’s limiting spatial factor resolution, of a screen’s potentially spatial enabling resolution, borated potentially scintillator enabling screens borated to exhibit scintillator higher 6 10 spatialscreens resolutionto exhibit compared higher spatial to the resolution screens created compa withred toLi the or Gd.screens Additionally, created withB has6Li a or lower Gd. γ Additionally,-ray sensitivity 10B than has gadoliniuma lower γ‐ray due sensitivity to its lower than atomic gadolinium number. due Scintillator to its lower screens atomic with number. a boron 6 converterScintillator can screens theoretically with a boron be made converter much can thinner theoretically than current be madeLi- much and Gd-basedthinner than screens, current while 6Li‐ 6 γ exhibitingand Gd‐based superior screens, neutron while detection exhibiting efficiency superior to Li neutron scintillator detection screens, efficiency improved to-ray 6Li insensitivity scintillator 10 overscreens, Gd improved scintillator γ screens,‐ray insensitivity and improved over Gd spatial scintillator resolution. screens, The and combination improved ofspatialB’s resolution. improved γ neutronThe combination detection eofffi ciency,10B’s improved spatial resolution, neutron detection and -ray efficiency, insensitivity spatial could resolution, potentially and provide γ‐ray a high-performanceinsensitivity could neutronpotentially scintillator. provide a high‐performance neutron scintillator.

1.E+06106 BB-1010 101.E+055 LiLi66

101.E+044

101.E+033

101.E+022

1.E+0110 Cross Section(b) 1.E+001

101.E-01-1

101.E-02-2

10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 1 10 102

1.E-11 1.E-10 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01

1.E+01 1.E+02 Energy (MeV) 1.E+00 Figure 1.1. Microscopic absorption cross-sectioncross‐section of 1010B and 6LiLi from the Evaluated Nuclear Data File (ENDF(ENDF//B-VIII)B‐VIII) cross-sectioncross‐section libraries.libraries.

Boron-basedBoron‐based neutron scintillators scintillators for for neutron neutron imaging imaging applications applications were were conceived conceived as asearly early as asthe the1950s 1950s [24,25] [24, 25and] andtheir their superior superior neutron neutron detection detection efficiency eﬃciency has been has documented been documented [26]. Glass [26]. Glassscintillators scintillators [27,28] [27, scintillators,28], scintillators with with a mixture a mixture of neutron of neutron absorbing absorbing materials materials [29,30] [29,30],, and and plastic scintillators [[3131–3434]] have all utilized boron for neutronneutron detectiondetection in didifferentﬀerent contexts.contexts. Despite their usefulness in other applications, applications, previous studies have generally concluded concluded thatthat boron-basedboron‐based scintillator screens’screens’ poor light output prevented them from becoming a viable scintillator for neutron imaging [[3535–3737]]. However, However, Nakamura et al.’s work [38] [38] suggests suggests smaller boron particles may increase

J. Imaging 2020, 6, 124 3 of 22 light output performance. More recent efforts by other colleagues in the neutron imaging field have recently explored using 10B/CsI:Tl scintillator screens for neutron imaging [39]. The working hypothesis of this study is that the low probability of 10B’s daughter products to escape from the boron-containing particle and arrive as a scintillating particle is the likely cause of the low light output observed in previous investigations of boron-based neutron scintillator screens. This work aims to use smaller boron converter particles to enhance light output. The goal of this investigation is to benefit the broader neutron imaging community by exploring boron-based scintillator screens that may deliver improved neutron detection efficiency, light output, and potentially spatial resolution.

1.2. Theory of Boron-Based Scintillator Screens The range of daughter products created when a neutron interacts with a neutron converter is one of the limiting factors of a scintillator screen’s spatial resolution. When a thermal neutron (0.025 eV) is absorbed by 10B, a 10B(n,α)7Li reaction occurs with a 94% branching ratio of producing a 7Li particle in an excited state (Equation (1)) and a 6% branching ratio of producing a 7Li in the ground state (Equation (2)), with an average reaction energy of 2.34 MeV. The daughter products of 10B have less energy than those created by thermal neutron interactions with 6Li (Equation (3)) and Gd (Equations (4) and (5)), causing them to deposit energy more locally to the interaction site and improve the theoretical spatial resolution limit. Additionally, 10B’s daughter products are more massive and deposit their energy over a shorter range than those of 6Li:

10 1 7 4 B + n Li∗(0.84 MeV) + α(1.47 MeV) (1) 5 0 → 3 2 10B + 1n 7Li(1.01 MeV) + 4α(1.78 MeV) (2) 5 0 → 3 2 6Li + 1n 3H(2.73 MeV) + 4α(2.05 MeV) (3) 3 0 → 1 2 157 1 158 158 Gd + n Gd∗ Gd + γ rays + ICe−; Q = 7.937 MeV (4) 64 0 → 64 → 64 155 1 156 156 Gd + n Gd∗ Gd + γ rays + ICe−; Q = 8.536 MeV (5) 64 0 → 64 → 64 Natural gadolinium, which consists of seven stable isotopes, has a total thermal neutron cross-section of 48,800 barns. Two isotopes that have high thermal neutron capture cross-sections and a relatively high natural abundance are 155Gd (60,900 b, 14.8% abundance) and 157Gd (254,000 b, 15.7% abundance) [40]. As shown in Equations (4) and (5), gadolinium neutron absorption reactions produce prompt γ-rays and internal conversion electrons (ICe ), which also produce secondary characteristic − X-rays and Auger electrons, including Coster–Kronig electrons. The γ-rays represent 99% of the Q-value energy [41], but the emitted electrons deposit their energy more locally than photons. Thus, the response of a gadolinium oxysulfide scintillator (e.g., GOS, Gd2O2S, informally known as Gadox) to thermal neutron capture is a combination of the effects of emitted photons and electrons. Different GOS scintillators (such as GOS:Tb, GOS:Pr, GOS:Eu) have different properties, such as peak light emission, photons/MeV, and decay time. These properties can be matched to the specific application requirements to maximize performance. Thermal neutron capture by 157Gd, for example, releases an average of 3.288 photons including prompt and secondary γ-rays and X-rays with energies ranging from 10s of eV up to several MeV and a mean energy of 2.394 MeV [42,43]. This same reaction also releases ICe − with energies ranging from 29 keV to 6.9 MeV, with the most intense discrete ICe emissions of 29 and − 71 keV [40,43,44]. The range of a 71 keV ICe in gadolinium is approximately 20 µm [45]. Goorley and − Nikjoo have compiled a more complete table of gadolinium spectral photon emissions [43]. Less energetic daughter products help improve spatial resolution because they have a shorter range in the surrounding material. The probability that the daughter products will escape the boron converter particles and reach the scintillator particles decreases with increasing converter particle size. When the converter particles themselves prevent daughter products from reaching the scintillator material and producing photons, the light output of the scintillator screen is reduced despite any apparent J. Imaging 2020, 6, 124 4 of 22

J. Imaging 2020, 6, x FOR PEER REVIEW 4 of 22 improvement in the detection eﬃciency from using 10B compared to 6Li. Thus, the converter particle the converter particle size could explain the poor light output reported by previous studies on size could explain the poor light output reported by previous studies on borated scintillator screens. borated scintillator screens. A Transmission of Ions in Matter (TRIM) simulation [46] calculated the ranges of the daughter A Transmission of Ions in Matter (TRIM) simulation [46] calculated the ranges of the daughter products from the neutron absorption reactions of interest in the boron and lithium-based scintillator products from the neutron absorption reactions of interest in the boron and lithium‐based scintillator screens.screens TRIM. TRIM does does not not account account for for isotopic isotopic information,information, so natural elemental elemental abundances abundances were were used used in thein thecalculation. calculation. The The range range of internal of internal conversion conversion electrons electrons in GOS in scintillator GOS scintillator screens was screens calculated was usingcalculated the Penetration using the and Pen Energyetration Loss and ofEnergy Positrons Loss andof Positrons Electrons and (PENELOPE) Electrons (PENELOPE) software (Version) software [47 ]. Table(Version)1 lists the [47] daughter. Table 1 particleslists the daughter in the converter particles material, in the converter the converter material material, the converter densities material used for thedensities simulation, used and for the the resulting simulation daughter, and productthe resulting ranges, daughter which are product also displayed ranges, which in Figure are 2 also. displayed in Figure 2. Table 1. Simulated converter daughter product range in converter medium. Table 1. Simulated converter daughter product range in converter medium. Converter Density (g/cm3) Daughter Product Range Converter Density (g/cm3) 7Li (1.01 MeV) Daughterα (1.78 MeV) Product7Li (0.84Range MeV) α (1.47 MeV) 7Li (1.01 MeV) α (1.78 MeV) 7Li (0.84 MeV) α (1.47 MeV) B2O3 2.46 2.45 µm 4.84 µm 2.19 µm 3.97 µm NaBB25OO38 2.2.0646 2.45 2.98 µmµm 4.84 5.88 µmµm 2.19 2.66 µmµm 3.97 4.82 µµmm BN 2.10 2.61 µm 5.21 µm 2.34 µm 4.24 µm NaB5O8 2.06 2.98 µm 5.88 µm 2.66 µm 4.82 µm BN 2.10 2.613H (2.73 µm MeV) α5.21(2.05 µm MeV) 2.34 µm 4.24 µm LiF 2.64 3H (2.73 62.7 MeV)µm α (2.05 6.04 MeV)µm

LiF 2.64 62.7ICe µm(29 keV) IC6.04e (71 µm keV) − − 157 e− e− Gd2O 2S:Tb 7.32 IC (29 1.31 keV)µm IC 5.46(71 keV)µm 157Gd2O2S:Tb 7.32 1.31 µm 5.46 µm

7Li (0.84 MeV )

8 7

O Li (1.01 MeV ) 5

α (1.47 MeV ) NaB

α (1.78 MeV )

7Li (0.84 MeV )

3 Li (1.01 MeV )

O 2

B α (1.47 MeV ) α (1.78 MeV )

7Li (0.84 MeV )

7Li (1.01 MeV ) BN α (1.47 MeV ) α (1.78 MeV )

3H (2.73 MeV )

α (2.05 MeV ) LiF

0 5 10 15 20 60 70 Path Length (µm) FigureFigure 2. Transmission2. Transmission of of Ions Ions in in Matter Matter (TRIM) (TRIM) simulatedsimulated daughterdaughter product product path path length length in in various various converterconverter materials. materials The. The10 10B(n,B(n,α)α)77LiLi daughterdaughter products deposit deposit their their energy energy more more local locallyly than than the the 6Li(n,6Li(n,α)α)3H3H daughter daughter products. products. This This demonstrates demonstrates that 10BB-based‐based scintillators scintillators offer oﬀer an an improved improved theoreticaltheoretical spatial spatial resolution resolution limit limit compared compared to to6 Li-based6Li‐based screens.

J. Imaging 2020, 6, 124 5 of 22

These simulations show 6Li daughter products traveling through LiF, as well as 10B daughter 10 products traveling through a layer of B2O3, NaB5O8, or BN. The densities of NaB5O8 and Na B5O8 were essential to analyzing the imaging results, but the density of anhydrous NaB5O8 was not available in the literature. The density of its hydrous form, Na2O 5(B2O3) 10(H2O) (i.e., NaB5H10O13), is known nat 3 10 3 · · 10 (with B, 1.707 g/cm [48]; with B, 1.67 g/cm ). The density of Na B5O8 was experimentally 3 3 measured to be approximately 2.02 g/cm , fixing the density of natural NaB5O8 at 2.06 g/cm . The triton from 6Li(n,α)3H has a range of 62.7 µm, which can easily escape a ~10 µm diameter converter particle to deposit most of its 2.78 MeV energy into the surrounding scintillator material. However, the 10B daughters deposit their energy almost an order of magnitude closer to the event origin than the 3H from 6Li. Thus, the majority of daughter products from a 10B(n,α)7Li reaction in a 10 µm diameter particle would not escape the particle to interact with any surrounding material, demonstrating the need for smaller boron particles to maximize the probability of conversion from absorbed neutrons in the converter to photon production from surrounding scintillator material. The neutrons transmitted through a sample carry attenuation information about the sample, which are converted to photons by the scintillator, and read by a digital camera and sampled into pixels to form an image. Thus, the neutron counting statistics dictate image quality as long as the camera detects enough light to represent the number of neutrons absorbed in the scintillator screen. Scintillators with a higher detection efficiency can offer shorter acquisition times and improved image quality (e.g., signal-to-noise ratio, SNR). Scintillator screens containing one or more effective converter materials detect neutrons by attenuating them through absorption, so screens that absorb more neutrons demonstrate increased neutron detection efficiency. Neutron attenuation is described by the Beer–Lambert law, shown in Equation (6), where φ0 is the initial neutron flux and φ(t) is the neutron flux after attenuation through a material of thickness, t, and total attenuation macroscopic cross-section, ΣTotal. The total attenuation cross-section (ΣTotal) is the sum of the scattering and absorption cross-sections (Σs and Σa, respectively). For the strongly absorbing materials considered here, Σa >> Σs, meaning the attenuation is driven by absorption and scattering effects are negligible, and Σ Σa. Thus, a screen’s detection efficiency is the neutron attenuation fraction, (t)/φ : Total ≈ 0

Σtotalt φ(t) = φ0e− (6)

Calculated neutron detection efficiencies of various scintillator materials are listed in Table2 and displayed in Figure3. The attenuation curves for both thermal and cold neutrons show that gadolinium-based scintillators have the highest neutron detection efficiency. However, they have a large daughter product energy and are also sensitive to γ-rays. 10B-based scintillators have a significantly higher attenuation than 6Li scintillators, so a thinner borated scintillator screen can maintain the same detection efficiency as a thicker 6Li screen, further improving resolution by minimizing the potential for photon diffusion within the screen. This suggests that 10B-based screens should exhibit higher detection efficiency than 6Li-based screens, lower γ-ray sensitivity than gadolinium-based screens, and improved spatial resolution compared to both gadolinium-based and 6Li-based screens.

Table 2. Total macroscopic cross-section, mean free path, and the 20% absorption thickness of diﬀerent borated scintillator screens compared to 6LiF and GOS screens.

Σ 20% Absorption total Mean Free Path (µm) 1 Thickness (µm) Scintillator Mixture (cm− ) Thermal Cold Thermal Cold Thermal Cold 6LiF:ZnS (1:2) 12.8 28.5 780 350 174 78 natBN:ZnS (1:1) 14.3 32.0 700 313 156 70 6LiF:ZnS (1:1) 17.8 39.7 562 252 125 56 10 Na B5O8:ZnS (1:1) 41.5 92.4 241 108 54 24 10 B2O3:ZnS (1:1) 52.9 118 189 85 42 19 10BN:ZnS (1:1) 69.6 156 144 64 32 14 Gd2O2S (Gadox) 799 1467 13 6.8 2.8 1.5 J. Imaging 2020, 6, 124 6 of 22 J. Imaging 2020, 6, x FOR PEER REVIEW 6 of 22

100% 90% 80% 70% 60%

50%

40% 66LiFLiF (1:2)(1:2) ZnSZnS 66LiFLiF (1:1)(1:1) ZnSZnS

% Attenuation % 30% 10 enrB2O3(1:1)ZnSB2O3 (1:1) ZnS 20% natnat.BN BN(1:1) (1:1) ZnS ZnS 10enr.BN BN (1:1) (1:1) ZnS ZnS 10 10% NaenNaB5O8(1:1)ZnSB5O8 (1:1) ZnS GdGadoxO S (Gadox) 0% 2 2 0 50 100 150 200 250 300 Thickness (μm) 100% 90% 80% 70% 60% 50%

40% 66LiFLiF (1:2)(1:2) ZnSZnS 66LiFLiF (1:1)(1:1) ZnSZnS

% Attenuation % 30% 10 enrB2O3(1:1)ZnSB2O3 (1:1) ZnS 20% natnat.BN BN(1:1) (1:1) ZnS ZnS 10enr.BN BN (1:1) (1:1) ZnS ZnS 10 10% NaenNaB5O8(1:1)ZnSB5O8 (1:1) ZnS GadoxGd O S (Gadox) 0% 2 2 0 50 100 150 200 250 300 Thickness (μm)

Figure 3. Calculated a attenuationttenuation fraction for the thermal ( top) and cold neutrons ( bottom) in different different scintillator materials. 2. Materials and Methods 2. Materials and Methods This study was divided into two phases, beginning with the initial scoping measurements of a This study was divided into two phases, beginning with the initial scoping measurements of a narrow set of screen variations that informed the creation of a broader variety of screens for more narrow set of screen variations that informed the creation of a broader variety of screens for more detailed measurements. The following sections describe the scintillator screens created as part of this detailed measurements. The following sections describe the scintillator screens created as part of this work along with the methods employed to characterize their performance. work along with the methods employed to characterize their performance. 2.1. Scintillator Screens 2.1. Scintillator Screens The purpose of this study was to conduct a combinatorial study of boron-based neutron The purpose of this study was to conduct a combinatorial study of boron‐based neutron scintillators to determine if (1) borated scintillator screens could offer improved neutron imaging scintillators to determine if (1) borated scintillator screens could offer improved neutron imaging capabilities compared to current widely-used screens and (2) which parameters produced the best capabilities compared to current widely‐used screens and (2) which parameters produced the best performing screens. performing screens. 2.1.1. Scintillator Screen Compositions 2.1.1. Scintillator Screen Compositions A total of 25 different boron-based scintillator screens were evaluated as part of the initial scoping studies,A total which of are 25 listed different in Table boron3.‐based The results scintillator of this screens experiment were allowed evaluated for as a down-selection part of the initial of scintillatorscoping studies, screen which parameters are listed that in informed Table 3. the The second results phase of this of borated experiment scintillator allowed screen for a design down‐ andselection evaluation. of scintillator All screens screen employed parameters a ZnS:Cu that scintillator informed and the had second an active phase area of borated of 50 mm scintillator50 mm. × Fourscreen screens design layered and evaluation. the scintillator All screens and converter employed materials a ZnS:Cu separately, scintillator while and the had rest an had active a mixture area of scintillator50 mm × 50 and mm. converter Four screens material layered particles. the Thesescintillator scoping and studies converter included materials a larger separately variety of, while converter the materials,rest had a andmixture only of the scintillator top few performing and converter converters material were particles pursued. These further scoping in subsequent studies included testing. a larger variety of converter materials, and only the top few performing converters were pursued further in subsequent testing.

J. Imaging 2020, 6, 124 7 of 22

Table 3. Parameters of screens used in initial scoping studies.

Substrate Converter Material Atomic Ratio (B:ZnS) Thickness (µm) Aluminum 10B-enriched boron powder layered 5050 Aluminum 10B-enriched boron powder layered 100,100 Aluminum 10B-enriched boron powder layered 200,100 Aluminum 10B-enriched boron powder layered 200,200 Aluminum 10B-enriched Boric Acid 2:1 200 Aluminum 10B-enriched Boric Acid 1:1 200 Aluminum 10B-enriched Boric Acid 1:2 200 Aluminum 10B-enr. Sodium Pentaborate 2:1 200 Aluminum 10B-enr. Sodium Pentaborate 1:1 200 Aluminum 10B-enr. Sodium Pentaborate 1:2 200 Aluminum Boron Oxide (B2O3) 2:1 200 Aluminum Boron Oxide (B2O3) 1:1 200 Aluminum Boron Oxide (B2O3) 1:2 200 1.9 mm FSM Boron Oxide (B2O3) 2:1 200 1.9 mm FSM Boron Oxide (B2O3) 1:1 200 1.9 mm FSM Boron Oxide (B2O3) 1:2 200 Aluminum Boron Nitride (BN) 2:1 200 Aluminum Boron Nitride (BN) 1:1 200 Aluminum Boron Nitride (BN) 1:2 200 Aluminum Wurtzite-Boron Nitride (w-BN) 2:1 200 Aluminum Wurtzite-Boron Nitride (w-BN) 1:1 200 Aluminum Wurtzite-Boron Nitride (w-BN) 1:2 200 Aluminum Anhydrous Sodium Tetraborate 2:1 200 Aluminum Anhydrous Sodium Tetraborate 1:1 200 Aluminum Anhydrous Sodium Tetraborate 1:2 200

The second phase of this work examined 50 different boron-based scintillator screens, which are listed in Table4. At least two of each screen variety were fabricated and tested to reduce any anomalous effects from fabrication. Some additional screens were fabricated with a single thickness for the acquisition of test images. This study utilized a total of 118 borated scintillator screens and six different 6LiF screens. Except for a few cases that used a 1.8 mm-thick first-surface mirror (FSM), all screens were mounted on 63 mm 63 mm square aluminum substrates that were 1 mm thick. × Additional substrate materials were also tested. The active area of scintillator and converter material measured 50 mm 50 mm and a ZnS:Cu scintillator was used for each screen. The ZnS:Cu × scintillator was chosen because cameras used in this study have a higher quantum efficiency in the wavelengths emitted from ZnS:Cu than for ZnS:Ag. Screen performance was measured for sensors with 10 nat 10 a varying converter material ( B2O3, BN, and Na B5O8), scintillator median particle size (11.5 and 4.7 µm), converter-to-scintillator (boron atom to ZnS molecule) mix ratio (2:1, 1:1, and 1:2), and sensor substrate (FSM, matte black aluminum, polished aluminum, matte finish aluminum, broad-spectrum white polymer-coated aluminum, and broad-spectrum white polymer sheet adhered to aluminum). nat 10 10 While the BN and Na B5O8 scintillator materials were obtained commercially, the B2O3 was 10 10 fabricated specifically for this project using a precursor of B-enriched boric acid (H3 BO3) and basic chemical and material processes. Three separate types of scintillator screens were tested during this study. First, wedged screens consisting of mixed converter and scintillator material were tested to determine the impact of scintillator layer thickness on imaging performance. Second, screens with separate scintillator and converter layers as wedges of each layer, referred to as double wedge scintillator screens, were fabricated with the two wedges deposited in perpendicular directions so that a single double-wedged screen represents the full range of thicknesses for both the converter and scintillator materials. Third, screens with a thin, single-thickness scintillator layer were used to test some boron-based screens for neutron radiography. J. Imaging 2020, 6, 124 8 of 22

Table 4. Scintillator screen parameter combinations of second testing phase.

Particle Phosphor Converter Converter to Converter to Substrate Converter Size Maximum Thickness Scintillator Scintillator Surface Notes Material (µm) Thickness (µm) At. Ratio (B:ZnS) Weight Ratio Finish 11.5 300 10B metal 100 - - matte black double Wedge 4.7 300 10B metal 100 - - matte black double Wedge 10 11.5 300 B4C 100 - - matte black double Wedge 10 4.7 300 B4C 100 - - matte black double Wedge 10 11.5 300 B2O3 - 2:1 1:2.5 matte black single wedge 10 11.5 300 B2O3 - 1:1 1:4 matte black single wedge 10 11.5 300 B2O3 - 1:2 1:7 matte black single wedge 10 4.7 300 B2O3 - 2:1 1:2.5 matte black single wedge 10 4.7 300 B2O3 - 1:1 1:4 matte black single wedge 10 4.7 300 B2O3 - 1:2 1:7 matte black single wedge 10 11.5 300 Na B5O8 - 2:1 1:2.3 matte black single wedge 10 11.5 300 Na B5O8 - 1:1 1:3.5 matte black single wedge 10 11.5 300 Na B5O8 - 1:2 1:6.1 matte black single wedge 10 4.7 300 Na B5O8 - 2:1 1:2.3 matte black single wedge 10 4.7 300 Na B5O8 - 1:1 1:3.5 matte black single wedge 10 4.7 300 Na B5O8 - 1:2 1:6.1 matte black single wedge 11.5 300 natBN - 2:1 1:3.1 matte black single wedge 11.5 300 natBN - 1:1 1:5.1 matte black single wedge 11.5 300 natBN - 1:2 1:9.2 matte black single wedge 4.7 300 natBN - 2:1 1:3.1 matte black single wedge 4.7 300 natBN - 1:1 1:5.1 matte black single wedge 4.7 300 natBN - 1:2 1:9.2 matte black single wedge 10 11.5 300 B2O3 - 2:1 1:2.5 polished single wedge 10 11.5 300 B2O3 - 1:1 1:4 polished single wedge 10 11.5 300 B2O3 - 1:2 1:7 polished single wedge 10 4.7 300 B2O3 - 2:1 1:2.5 polished single wedge 10 4.7 300 B2O3 - 1:1 1:4 polished single wedge 10 4.7 300 B2O3 - 1:2 1:7 polished single wedge 10 11.5 300 Na B5O8 - 2:1 1:2.3 polished single wedge 10 11.5 300 Na B5O8 - 1:1 1:3.5 polished single wedge 10 11.5 300 Na B5O8 - 1:2 1:6.1 polished single wedge 10 4.7 300 Na B5O8 - 2:1 1:2.3 polished single wedge 10 4.7 300 Na B5O8 - 1:1 1:3.5 polished single wedge 10 4.7 300 Na B5O8 - 1:2 1:6.1 polished single wedge 11.5 300 natBN - 2:1 1:3.1 polished single wedge 11.5 300 natBN - 1:1 1:5.1 polished single wedge 11.5 300 natBN - 1:2 1:9.21 polished single wedge 4.7 300 natBN - 2:1 1:3.1 polished single wedge 4.7 300 natBN - 1:1 1:5.1 polished single wedge 4.7 300 natBN - 1:2 1:9.2 polished single wedge white poly 11.5 300 10B O - 1:1 1:4 single wedge 2 3 coating 10 11.5 300 B2O3 - 1:1 1:4 FSM single wedge 10 11.5 300 B2O3 - 1:1 1:4 matte single wedge 11.5 300 6LiF - 1:2 1:9.2 matte black single wedge 6.7 300 6LiF - 1:2 1:9.2 matte black single wedge 4.7 300 6LiF - 1:2 1:9.2 matte black single wedge 11.5 300 6LiF - 1:2 1:9.2 matte single wedge 6.7 300 6LiF - 1:2 1:9.2 matte single wedge 4.7 300 6LiF - 1:2 1:9.2 matte single wedge 10 11.5 300 B2O3 - 1:1 1:4 white poly sheet single wedge 10 4.7 20 B2O3 - 2:1 1:2.5 matte black 10 4.7 20 B2O3 - 1:1 1:4 matte black 10 4.7 20 B2O3 - 1:2 1:7 matte black 10 4.7 20 Na B5O8 - 2:1 1:2.3 matte black 10 4.7 20 Na B5O8 - 1:1 1:3.5 matte black 10 4.7 20 Na B5O8 - 1:2 1:6.1 matte black 4.7 20 natBN - 2:1 1:5.7 matte black 4.7 20 natBN - 1:1 1:10.4 matte black 4.7 20 natBN - 1:2 1:19.9 matte black 4.7 20 pure 10B 10 - - matte black layered 10 4.7 20 B4C 10 - - matte black layered J. Imaging 2020, 6, x FOR PEER REVIEW 9 of 22

4.7 20 10B2O3 - 1:1 1:4 matte black 4.7 20 10B2O3 - 1:2 1:7 matte black 4.7 20 Na10B5O8 - 2:1 1:2.3 matte black 4.7 20 Na10B5O8 - 1:1 1:3.5 matte black 4.7 20 Na10B5O8 - 1:2 1:6.1 matte black 4.7 20 natBN - 2:1 1:5.7 matte black 4.7 20 natBN - 1:1 1:10.4 matte black 4.7 20 natBN - 1:2 1:19.9 matte black 4.7 20 pure 10B 10 - - matte black layered 4.7 20 10B4C 10 - - matte black layered

Three separate types of scintillator screens were tested during this study. First, wedged screens consisting of mixed converter and scintillator material were tested to determine the impact of scintillator layer thickness on imaging performance. Second, screens with separate scintillator and converter layers as wedges of each layer, referred to as double wedge scintillator screens, were fabricated with the two wedges deposited in perpendicular directions so that a single double‐wedged screen represents the full range of thicknesses for both the converter and scintillator materials. Third, screens with a thin, single‐thickness scintillator layer were used to test some boron‐based screens for neutron radiography.

2.1.2. Wedged Scintillator Screens Screen thickness is one of the most impactful parameters that determines the performance of a J.scintillator Imaging 2020 screen., 6, 124 To determine screen performance as a function of thickness, one could fabricate9 of 22 multiple screens with a range of discrete thicknesses, measure the performance of each, and then interpolate between the data points. However, this is expensive and time‐consuming. 2.1.2. Wedged Scintillator Screens This work simultaneously studies a continuous range of thicknesses using a wedged scintillator screen,Screen where thickness scintillator is one la ofyer the thickness most impactful varies continuouslyparameters that across determines the screen the, performance as described of in a scintillatorprevious work screen. [49].To The determine wedged screens screen performancewere fabricated as by a function first depositing of thickness, a nominally one could 300 μm fabricate‐thick multiplescintillator screens layer onto with the a range substrate. of discrete Then, a thicknesses, rigid blade was measure used theto scrape performance the scint ofillator each, layer and away then interpolateuntil the target between high theand data low points.thicknesses However, were thisreached is expensive with a monotonic and time-consuming. gradient in between them. A jigThis was workused to simultaneously keep the blade studies at the aproper continuous angle. rangeThe resulting of thicknesses screens using had a nominal wedged scintillatormaximum screen,thickness where of 300 scintillator μm that layerdecreased thickness to a variesthin layer continuously a few microns across thick. the screen, The as as‐fabricated described inscintillator previous workthickness [49]. deposited The wedged on screensthe substrate were fabricated was measured by ﬁrst with depositing two different a nominally commercial 300 µm-thick coating scintillator thickness layergauges onto that the use substrate. eddy current Then, probes a rigid. The blade measurements was used to from scrape both the gauges scintillator were layer taken away at the until thicker the targetportion high of andthe screenlow thicknesses and averaged were reached together with to produce a monotonic a maximum gradient inas between‐fabricated them. thickness. A jig was A usedschematic to keep of the the wedged blade at scintillator the proper design angle. Theemphasizing resulting the screens wedge had shape a nominal (exaggerated, maximum not thickness to scale) ofis shown 300 µm in that Figure decreased 4. In total, to a 84 thin wedged layer a screens few microns were part thick. of Thethis as-fabricatedstudy. scintillator thickness depositedFigure on 5 theshows substrate a photograph was measured of a representative with two diﬀ wedgederent commercial screen taken coating with thicknessan oblique gauges light thatapplied use to eddy enhance current the probes. appearance The measurements of the surface fromtexture both in gaugesthe picture. were Multiple taken at surface the thicker defects portion can ofbe theseen screen in these and prototype averaged screens, together but to these produce disco a maximumntinuities are as-fabricated accounted thickness.for in the post A schematic‐processing of theof experimental wedged scintillator data [49] design. The emphasizing thicker section the wedge of the wedgeshape (exaggerated, is at the top not of the to scale) image, is shown with the in Figurethickness4. In decreasing total, 84 wedged continuously screens toward were part the bottom of this study. of the screen.

Dimensions Not to Scale 300 µm Scintillator Layer Substrate 1 mm

Figure 4. Schematic of a wedged scintillator concept [[49]49]..

Figure5 shows a photograph of a representative wedged screen taken with an oblique light applied to enhance the appearance of the surface texture in the picture. Multiple surface defects can be seen in these prototype screens, but these discontinuities are accounted for in the post-processing of experimental data [49]. The thicker section of the wedge is at the top of the image, with the thickness decreasingJ. Imaging 2020 continuously, 6, x FOR PEER towardREVIEW the bottom of the screen. 10 of 22

Figure 5.5. Photograph of a wedged scintillator screen with an oblique light applied to enhance the 1010 1010 appearance of thethe surfacesurface texture.texture. TheThe screenscreen shownshown isis aa BB22O3 neutron converter in a 1:11:1 (( B:ZnSB:ZnS)) mix ratio with a ZnS:Cu scintillator withwith 11.511.5 µμmm medianmedian particleparticle size.size.

2.1.3. Double Double-Wedged‐Wedged Scintillator ScreensScreens Mixing thethe converterconverter and and scintillator scintillator particles particles is feasibleis feasible for for white white converter converter particles particles because because they dothey not do absorb not absorb the light the emitted light emitted from the from scintillator the scintillator like a dark-colored like a dark material‐colored would. material This would study tested. This the performance of dark boron-containing converters, including 10B C and 10B metal, but separated study tested the performance of dark boron‐containing converters, including4 10B4C and 10B metal, but themseparated into them a distinct into a layer distinct from layer the from scintillator the scintillator in the form in the of form a double-wedged of a double‐wedged scintillator scintillator screen. screen. The converter layer was deposited directly onto an aluminum substrate and the thickness ranged from 100 µm to a few microns. The scintillator layer was deposited on top of the converter layer and varied from 300 μm to a few microns thick. The thickest parts of the two wedges were placed perpendicularly to each other, so each position of the scintillator screen represented a different combination of converter and scintillator layer thickness. This approach allowed a continuous range of thickness combinations to be assessed using a single screen. Nine separate double‐wedged screens were included in this study. A double‐wedged screen is displayed in Figure 6. The scintillator thickness varies from top (thicker) to bottom (thinner), while the converter is thickest on the right and decreases toward the left of the screen.

Figure 6. Photograph of a double‐wedged scintillator screen consisting of a boron carbide converter

(10B4C) and a ZnS:Cu scintillator with 11.5 μm median particle size.

2.1.4. High‐Resolution Scintillator Screens Boron‐based screens with uniformly thick scintillator coatings deposited as thin as possible were also explored in this study. The scintillator layers for these screens all had target maximum thicknesses of ≤ 20 μm. Light dispersion within the scintillator layer is a source of image unsharpness, so thin screens should exhibit higher spatial resolution. This part of the study explored the feasibility of achieving higher spatial resolution by exploiting boron’s increased neutron detection efficiency to offset the lower light output of a thinner screen. Figure 7 displays a photograph of a high‐resolution screen under oblique lighting to emphasize screen texture. The missing portion of the coating was caused by the insufficient adhesion of the coating to the substrate during a delicate scraping operation

J. Imaging 2020, 6, x FOR PEER REVIEW 10 of 22

Figure 5. Photograph of a wedged scintillator screen with an oblique light applied to enhance the

appearance of the surface texture. The screen shown is a 10B2O3 neutron converter in a 1:1 (10B:ZnS) mix ratio with a ZnS:Cu scintillator with 11.5 μm median particle size.

2.1.3. Double‐Wedged Scintillator Screens Mixing the converter and scintillator particles is feasible for white converter particles because they do not absorb the light emitted from the scintillator like a dark‐colored material would. This 10 10 J.study Imaging tested2020, 6the, 124 performance of dark boron‐containing converters, including B4C and B metal,10 ofbut 22 separated them into a distinct layer from the scintillator in the form of a double‐wedged scintillator screen. The converter layer was deposited directly onto an aluminum substrate and the thickness Theranged converter from 100 layer µm was to depositeda few microns. directly The onto scintillator an aluminum layer substratewas deposited and the on thickness top of the ranged converter from 100layerµ mand to avaried few microns. from 300 The μm scintillator to a few layermicrons was thick. deposited The onthickest top of part thes converter of the two layer wedge ands varied were fromplaced 300 perpendicularlyµm to a few microns to each thick. other, The so thickesteach position parts of the scintillator two wedges screen were placedrepresented perpendicularly a different tocombination each other, of so converter each position and of scintillator the scintillator layer screenthickness. represented This approach a diﬀerent allowed combination a continuous of converter range andof thickness scintillator combinations layer thickness. to be Thisassessed approach using alloweda single ascreen continuous. Nine separate range of double thickness‐wedged combinations screens towere be assessed included using in this a single study. screen. A double Nine‐ separatewedged double-wedged screen is displayed screens in wereFigure included 6. The in scintillator this study. Athickness double-wedged varies from screen top is (thicker) displayed to in bottom Figure 6(thinner),. The scintillator while the thickness converter varies is thickest from top on (thicker) the right to bottomand decreases (thinner), toward while the the left converter of the screen. is thickest on the right and decreases toward the left of the screen.

Figure 6. Photograph of a double-wedgeddouble‐wedged scintillator screen consistingconsisting of a boron carbide converter 1010 (( BB44C)C) and and a a ZnS:Cu ZnS:Cu scintillator scintillator with with 11.5 11.5 μmµm median median particle particle size size..

2.1.4. High High-Resolution‐Resolution Scintillator Screens Boron-basedBoron‐based screens withwith uniformly thickthick scintillator coatings depositeddeposited asas thin as possible were also explored explored in in this this study. study. The Thescintillator scintillator layers layers for these for screens these all screens had target all had maximum target thicknesses maximum of 20 µm. Light dispersion within the scintillator layer is a source of image unsharpness, so thin thicknesses≤ of ≤ 20 μm. Light dispersion within the scintillator layer is a source of image unsharpness, screensso thin screens should should exhibit exhibit higher higher spatial spatial resolution. resolution This. This part part of the of studythe study explored explored the the feasibility feasibility of achievingof achieving higher higher spatial spatial resolution resolution by by exploiting exploiting boron’s boron’s increased increased neutron neutron detectiondetection eefficiencyfficiency to ooffsetffset thethe lowerlower lightlight outputoutput ofof aa thinnerthinner screen.screen. Figure7 7 displays displays aa photographphotograph ofof aa high-resolutionhigh‐resolution screen under oblique lighting to emphasize screen texture. The missing portion of the coating was J.screen Imaging under 2020, 6 , obliquex FOR PEER lighting REVIEW to emphasize screen texture. The missing portion of the coating11 ofwas 22 caused by the insufficient insufficient adhesion ofof thethe coating to the substrate during a delicate scraping operation to make the wedgedwedged thicknessthickness profile.profile. However,However, this this defect defect did did not not prohibit prohibit measurements measurements because because it itis is outside outside the the central central field field of of view view (FOV) (FOV) of of the the screen. screen.

1010 Figure 7. Photograph of a high-resolutionhigh‐resolution scintillator screenscreen consistingconsisting ofof aa B22O3 neutronneutron converter inin a 1:1 ( 10B:ZnS)B:ZnS) mix mix ratio ratio with with a a ZnS:Cu ZnS:Cu scintillator scintillator with with 4.7 4.7 μmµm median median particle particle size size..

2.2. Neutron Neutron Source Source and Instrumentation The measurements for this study were taken using the A ANTARESNTARES cold neutron beam located at 7 2 the FRM II reactor [50]. All radiographs were acquired with a neutron ﬂux of 5.7 10 7 n/cm2 /s at the the FRM II reactor [50]. All radiographs were acquired with a neutron flux of 5.7 ×× 10 n/cm /s at the sample position and a length‐to‐diameter ratio (L/D) of 540, where L is the beamline length measured from the aperture to the image plane and D is the effective diameter of the aperture. A 1 mm‐thick cadmium filter was placed in the beamline to filter thermal neutrons for the testing of the screens’ sensitivity to epithermal neutrons. ANTARES has two imaging systems, one with a larger FOV and another for high resolution applications. The large FOV system uses an Andor iKon‐L camera (Andor iKon-L, Oxford Instruments, Belfast, UK). The high‐resolution imaging system utilizes an Andor Neo camera (Andor Neo, Oxford Instruments, Belfast, UK). The peak light emission of the ZnS:Cu used in this study is ~530 nm, which matches well with the camera used to study light output, as the camera has > 90% quantum efficiency between ~480 and ~710 nm with a peak near 550 nm [51]. The large FOV system was used to simultaneously characterize multiple screens for detection efficiency and light output, and the high‐resolution system was used for resolution testing and neutron tomography.

2.3. Testing Methodology The goal of this study was to characterize the performance of a variety of boron‐based scintillator screens to understand how different screen parameters affected the final imaging performance. Previous reports of low light output were a concern, so initial scoping studies focused on identifying borated screens that produced sufficient light output for neutron imaging applications and down‐ selecting the boron converter chemical form to the few most promising candidates for a subsequent set of screens and measurements. The second phase of testing measured the detection efficiency and light yield of the candidate scintillator screens for a wider variety of screen parameters. Spatial resolution was quantified by taking a radiograph of a 1 cm‐wide, 75 µm‐thick gadolinium strip and calculating the modulation transfer function (MTF) from the edge profile. Light output and detection efficiency were measured using previously developed methodology [49]. The larger FOV imaging system was used for these measurements because four screens could be simultaneously measured in a single radiograph. First, radiographs were acquired with the borated screens optically coupled directly to the imaging system to measure the relative light output of each screen. For cold neutron radiographs, this consisted of a single six‐second exposure. Epithermal testing was performed by acquiring five 20 s exposures and summing them together. A diagram of this measurement and a sample radiograph from it are displayed in Figure 8. Once the radiograph measuring the light yield of the screens was captured, the borated screens were removed from the imaging system and a common 6LiF/ZnS:Cu screen was mounted to the imaging system. An open beam image was acquired for the post‐process correction of subsequent

J. Imaging 2020, 6, 124 11 of 22 sample position and a length-to-diameter ratio (L/D) of 540, where L is the beamline length measured from the aperture to the image plane and D is the effective diameter of the aperture. A 1 mm-thick cadmium filter was placed in the beamline to filter thermal neutrons for the testing of the screens’ sensitivity to epithermal neutrons. ANTARES has two imaging systems, one with a larger FOV and another for high resolution applications. The large FOV system uses an Andor iKon-L camera (Andor iKon-L, Oxford Instruments, Belfast, UK). The high-resolution imaging system utilizes an Andor Neo camera (Andor Neo, Oxford Instruments, Belfast, UK). The peak light emission of the ZnS:Cu used in this study is ~530 nm, which matches well with the camera used to study light output, as the camera has > 90% quantum efficiency between ~480 and ~710 nm with a peak near 550 nm [51]. The large FOV system was used to simultaneously characterize multiple screens for detection efficiency and light output, and the high-resolution system was used for resolution testing and neutron tomography.

2.3. Testing Methodology The goal of this study was to characterize the performance of a variety of boron-based scintillator screens to understand how different screen parameters affected the final imaging performance. Previous reports of low light output were a concern, so initial scoping studies focused on identifying borated screens that produced sufficient light output for neutron imaging applications and down-selecting the boron converter chemical form to the few most promising candidates for a subsequent set of screens and measurements. The second phase of testing measured the detection efficiency and light yield of the candidate scintillator screens for a wider variety of screen parameters. Spatial resolution was quantified by taking a radiograph of a 1 cm-wide, 75 µm-thick gadolinium strip J. Imaging 2020, 6, x FOR PEER REVIEW 12 of 22 and calculating the modulation transfer function (MTF) from the edge profile. radiographs.Light output The andborated detection screens effi wereciency then were placed measured on the using source previously‐side of developed the 6LiF/ZnS:Cu methodology screen [and49]. Thea radiograph larger FOV of the imaging borated system screens was was used acquired for these to measure measurements the detection because efficiency four screens of each could screen. be simultaneouslyFigure 9 shows a measured schematic in of a this single setup, radiograph. as well as First, a representativ radiographse radiograph were acquired of this with measurement. the borated screensThere was optically a gap of coupled only 8 mm directly between to the the imaging image systemplane and to measurethe borated the screens, relative so light the output magnification of each screen.effects were For coldnegligible. neutron This radiographs, procedure thiswas consistedperformed of for a singlethe double six-second‐ and single exposure.‐wedge Epithermald screens testingwith the was cold performed neutron beam. by acquiring Epithermal five 20 measurements s exposures and using summing the same them process together. with A a diagramcadmium of‐ thisfiltered measurement beam were and also a conducted sample radiograph with the single from it wedge are displayed screens. in Figure8.

Figure 8. SchematicSchematic diagram diagram showing showing the the light light yield yield measurement measurement of ofthe the borated borated screens screens (left (left). A). Arepresentative representative radiograph radiograph showing showing the the light light output output from from four four borated borated screens screens (right (right). ).

Once the radiograph measuring the light yield of the screens was captured, the borated screens were removed from the imaging system and a common 6LiF/ZnS:Cu screen was mounted to the imaging system. An open beam image was acquired for the post-process correction of subsequent radiographs. The borated screens were then placed on the source-side of the 6LiF/ZnS:Cu screen and a radiograph of the borated screens was acquired to measure the detection efficiency of each screen. Figure9 shows a schematic of this setup, as well as a representative radiograph of this measurement. There was a gap of only 8 mm between the image plane and the borated screens, so the magnification effects were negligible. This procedure was performed for the double- and single-wedged screens with the coldFigure neutron 9. Schematic beam. diagramEpithermal of the measurements detection efficiency using measurementthe same process (left) withand a cadmium-filteredrepresentative beamradiograph were also of conducted four borated with screens the single obtained wedge with screens.this setup (right).

2.4. Data Processing A median filter (3 × 3‐pixel radius) was applied to all images to reduce noise in the radiographs used to measure relative light output and detection efficiency. All radiographs were open‐beam and dark‐field corrected. Acquisition times were six seconds for the images taken in the unfiltered cold beam. For the epithermal (i.e., cadmium‐filtered) beam, a set of five 20‐second images were taken for each measurement and summed together. Spatial resolution images were integrated over ten seconds. A python code was written to correlate data from both the light‐yield and detection‐ efficiency measurements for each screen. Once correlated, relative light yield and detection efficiency as a function of scintillator layer thickness was calculated. A separate python script calculated the effective spatial resolution in terms of MTF. More details on the measurement procedure and data analysis methods can be found in Chuirazzi et al. [49].

3. Results

3.1. Results of Initial Scoping Studies The light outputs of the first set of borated neutron scintillator screens was measured, which are shown in Figure 10. Scintillator screens with layered converter and scintillator material produced minimal light output. The performance of sensors with a converter and scintillator mixture varied based on the converter material and mix ratios. More converter means less scintillator, and vice versa. Too much converter means there is not enough scintillator to produce photons, and too much scintillator means there is not enough converter to absorb the neutrons. There will be an optimal mixture ratio. Thus, the light output does not necessarily change monotonically with the mix ratio.

J. Imaging 2020, 6, x FOR PEER REVIEW 12 of 22 radiographs. The borated screens were then placed on the source‐side of the 6LiF/ZnS:Cu screen and a radiograph of the borated screens was acquired to measure the detection efficiency of each screen. Figure 9 shows a schematic of this setup, as well as a representative radiograph of this measurement. There was a gap of only 8 mm between the image plane and the borated screens, so the magnification effects were negligible. This procedure was performed for the double‐ and single‐wedged screens with the cold neutron beam. Epithermal measurements using the same process with a cadmium‐ filtered beam were also conducted with the single wedge screens.

J. ImagingFigure2020 8., 6 Schematic, 124 diagram showing the light yield measurement of the borated screens (left). A12 of 22 representative radiograph showing the light output from four borated screens (right).

FigureFigure 9. SchematicSchematic diagram diagram of of the the detection detection efficiency eﬃciency measurement measurement ( (leftleft)) and and a a representative radiographradiograph of of four four borated borated screens obtained with this setup ( right))..

2.4.2.4. Data Data Processing Processing A median filter (3 3-pixel radius) was applied to all images to reduce noise in the radiographs A median filter (3 × 3‐pixel radius) was applied to all images to reduce noise in the radiographs usedused to to measure measure relative relative light light output output and and detection detection efficiency efficiency.. All All radiographs radiographs were were open open-beam‐beam and and darkdark-field‐field corrected. corrected. Acquisition Acquisition times times were were six six seconds seconds for for the the images images taken taken in in the the unfiltered unfiltered cold cold beam.beam. For For the the epithermal epithermal (i.e. (i.e.,, cadmium cadmium-filtered)‐filtered) beam, beam, a set a set of five of five 20‐ 20-secondsecond images images were were taken taken for eachfor each measurement measurement and and summed summed together. together. Spatial Spatial resolution resolution images images were were integrated integrated over over ten ten seconds.seconds. A python python code code was was written written to correlate to correlate data datafrom bothfrom theboth light-yield the light and‐yield detection-e and detectionfficiency‐ efficiencymeasurements measurements for each screen.for each Oncescreen. correlated, Once correlated, relative relative light yield light yield and detection and detection efficiency efficiency as a asfunction a function of scintillator of scintillator layer lay thicknesser thickness was calculated. was calculated. A separate A separate python python script calculated script calculated the effective the effectivespatial resolution spatial resolution in terms in of terms MTF. Moreof MTF. details More on details the measurement on the measurement procedure procedure and data and analysis data analysismethods methods can be found can be in found Chuirazzi in Chuirazzi et al. [49 ].et al. [49]. 3. Results 3. Results 3.1. Results of Initial Scoping Studies 3.1. Results of Initial Scoping Studies The light outputs of the first set of borated neutron scintillator screens was measured, which are shownThe in light Figure output 10.s Scintillator of the first set screens of borated with layeredneutron converterscintillator and screens scintillator was measured, material which produced are shownminimal in lightFigure output. 10. Scintillator The performance screens with of sensors layered with converter a converter and andscintillator scintillator material mixture produced varied minimalbased on light the converteroutput. The material performance and mix of ratios.sensors Morewith a converter converter means and scintillator less scintillator, mixture and varied vice basedversa. on Too the much converter converter material means and there mix ratios is not. enoughMore converter scintillator means to produce less scintillator, photons, and and vice too versa. much Tooscintillator much converter means there means is not there enough is not converter enough to scintillator absorb the to neutrons. produce There photons, will and be an too optimal much scinmixturetillator ratio. means Thus, there the lightis not output enough does converter not necessarily to absorb change the neutrons. monotonically There withwill thebe an mix optimal ratio. mixtureJ. Imaging ratio. 20 20Thus,, 6, x FOR the PEER light REVIEW output does not necessarily change monotonically with the mix13 of ratio.22 1200

1000

800

600

400 Grey Values per Second per Values Grey

200

0 50-50 100-100 100-200 200-200 1:2 1:1 2:1 1:2 1:1 2:1 1:2 1:1 2:1 1:2 1:1 2:1 1:2 1:1 2:1 1:2 1:1 2:1 1:2 1:1 2:1 10B O -ZnS 10B O -ZnS layered 10B-ZnS H 10BO -ZnS Na10B O -ZnS 2 3 2 3 BN-ZnS wBN-ZnS Na B O -ZnS 3 3 5 8 on aluminum on FSM 2 4 7

FigureFigure 10. The10. The measured measured relative relative light light outputoutput of borated borated neu neutrontron scintillator scintillator screens screens from from the initial the initial scopingscoping studies. studies.

The highest performing screens in these initial scoping measurements included 10B2O3, natBN, and Na10B5O8. The light output of the 10B2O3 screens was notably higher for screens on FSM than for a matte black aluminum substrate, indicating that further study of substrate variations may be warranted. Results of this testing informed the fabrication of additional boron‐based scintillator screens. This second generation of screens focused on 10B2O3, natBN, and Na10B5O8 converter materials.

3.2. Additional Testing Converter material, converter grain size, converter‐to‐scintillator mix ratio, and substrate material were all varied to investigate their impact on the scintillator screens’ performance. The results of the data taken in the cold neutron beam are reported first, followed by epithermal neutron results. Duplicate screens were fabricated with the same combination of screen parameters, and the results of duplicate screens were averaged to reduce the effects of any outliers. Relative light yield is normalized to the single brightest screen to facilitate a relative comparison between all screens.

3.2.1 Cold Neutron Results Figure 11 shows relative light yield as a function of screen thickness for the same 10B2O3(1:1)ZnS:Cu scintillator composition deposited on various substrates. Reflective materials such as white poly and a first surface mirror (FSM) increase light output because they simply reflect photons originally traveling away from the camera back toward the camera. However, photons reflected off the substrate undergo further diffusion in the scintillator material with the increased path‐length, degrading the spatial resolution. The matte black backing absorbs photons so they do not reflect back and contribute to the image, which reduces light output but can improve spatial resolution. These measurements show that a reflective substrate can increase the relative light yield by > 20% compared to a matte black substrate.

J. Imaging 2020, 6, 124 13 of 22

10 nat The highest performing screens in these initial scoping measurements included B2O3, BN, 10 10 and Na B5O8. The light output of the B2O3 screens was notably higher for screens on FSM than for a matte black aluminum substrate, indicating that further study of substrate variations may be warranted. Results of this testing informed the fabrication of additional boron-based scintillator screens. 10 nat 10 This second generation of screens focused on B2O3, BN, and Na B5O8 converter materials.

3.2. Additional Testing Converter material, converter grain size, converter-to-scintillator mix ratio, and substrate material were all varied to investigate their impact on the scintillator screens’ performance. The results of the data taken in the cold neutron beam are reported ﬁrst, followed by epithermal neutron results. Duplicate screens were fabricated with the same combination of screen parameters, and the results of duplicate screens were averaged to reduce the eﬀects of any outliers. Relative light yield is normalized to the single brightest screen to facilitate a relative comparison between all screens.

3.2.1. Cold Neutron Results

10 Figure 11 shows relative light yield as a function of screen thickness for the same B2O3(1:1)ZnS:Cu scintillator composition deposited on various substrates. Reflective materials such as white poly and a first surface mirror (FSM) increase light output because they simply reflect photons originally traveling away from the camera back toward the camera. However, photons reflected off the substrate undergo further diffusion in the scintillator material with the increased path-length, degrading the spatial resolution. The matte black backing absorbs photons so they do not reflect back and contribute to the image, which reduces light output but can improve spatial resolution. These measurements show that aJ. reflectiveImaging 2020 substrate, 6, x FOR PEER can increaseREVIEW the relative light yield by > 20% compared to a matte black substrate.14 of 22

1.0

0.9

0.8

0.7

0.6

0.5 Ultra White Poly 0.4 White Poly Coating on Aluminum FSM

Relative Relative LightYield 0.3 MatteClear SatinAluminum Aluminum 0.2 White Poly Sheeton Aluminum Adhered to Aluminum 0.1 MatteBlack SatinBlack Aluminum Aluminum PolishedBrite Clear Aluminum Aluminum 0.0 0 50 100 150 200 250 300 Screen Thickness (µm) 10 Figure 11. Relative light yield as a function of thickness for a B O (1:1)ZnS:Cu10 scintillator composition Figure 11. Relative light yield as a function of thickness2 for3 a B2O3(1:1)ZnS:Cu scintillator depositedcomposition onto deposited various substrates.onto various substrates.

nat 10 10 Converter materials of BN, B2O3, and Na B5O8 were mixed with ZnS:Cu scintillator material Converter materials of natBN, 10B2O3, and Na10B5O8 were mixed with ZnS:Cu scintillator material in B:ZnS atomic ratios of 2:1, 1:1, and 1:2. Results of borated screens with different converter materials, in B:ZnS atomic ratios of 2:1, 1:1, and 1:2. Results of borated screens with different converter materials, scintillator particles, and scintillator particle size are displayed in Figures 12–14. All screens in these scintillator particles, and scintillator particle size are displayed in Figures 12–14. All screens in these results were mounted on a matte black substrate to remove effects of photon reflection by the substrate. results were mounted on a matte black substrate to remove effects of photon reflection by the substrate. Relative neutrons per graylevel, which is the ratio of the detection efficiency to the relative light yield, provides a measure of how many neutrons are represented in a grayscale value and a useful measure to compare screen performance. Detection efficiency scales with the amount of boron converter material, simply due to the increased macroscopic absorption cross‐section of the converter–scintillator mixture. However, the screens with higher detection efficiency exhibited fewer photons produced per detected neutron because an increase in the amount of boron decreases the amount of phosphor material, inhibiting photon production. 2.0 10 1.0 B2O3B2O3 (2:1)ZnS11.5um (2:1) 11.5 µm 1.0 2.0 1.8 10 B2O3B2O3 (1:1)ZnS11.5um (1:1) 11.5 µm 0.9 10 0.9 1.8 1.6 B2O3B2O3 (1:2)ZnS11.5um (1:2) 11.5 µm 10 0.8 B2O3B2O3 (2:1)ZnS4.7um (2:1) 4.7 µm 0.8 1.6 1.4 10 B2O3B2O3 (1:1)ZnS4.7um (1:1) 4.7 µm 10 1.2 0.7 B2O3B2O3 (1:2)ZnS4.7um (1:2) 4.7 µm 0.7 1.4 1.0 0.6 0.6 1.2 0.8 0.5 0.5 1.0 0.6 0.4 0.4 0.8 Neutronsper Grayscale 0.4

0.3 0.3 0.6 Relative Relative LightYield 0.2 Detection Efficiency 0.2 0.2 0.4 0.0 0 100 200 3000.1 0.1 Relative Neutronsper Grayscale 0.2 Screen Thickness (µm)0.0 0.0 0.0 0 100 200 300 0 100 200 300 0 100 200 300 Screen Thickness (µm) Screen Thickness (µm) Screen Thickness (µm)

Figure 12. Results for 10B2O3 screens with different mix ratios and scintillator particle sizes.

J. Imaging 2020, 6, x FOR PEER REVIEW 14 of 22

1.0

0.9

0.8

0.7

0.6

0.5 Ultra White Poly 0.4 White Poly Coating on Aluminum FSM

Figure 11. Relative light yield as a function of thickness for a 10B2O3(1:1)ZnS:Cu scintillator composition deposited onto various substrates.

Converter materials of natBN, 10B2O3, and Na10B5O8 were mixed with ZnS:Cu scintillator material in B:ZnS atomic ratios of 2:1, 1:1, and 1:2. Results of borated screens with different converter materials, scintillator particles, and scintillator particle size are displayed in Figures 12–14. All screens in these results were mounted on a matte black substrate to remove effects of photon reflection by the substrate. Relative neutrons per graylevel, which is the ratio of the detection efficiency to the relative light yield, provides a measure of how many neutrons are represented in a grayscale value and a useful measure to compare screen performance. Detection efficiency scales with the amount of boron converter material, simply due to the increased macroscopic absorption cross‐section of the converter–scintillator mixture. However, the screens with higher detection efficiency exhibited fewer photonsJ. Imaging 2020produced, 6, 124 per detected neutron because an increase in the amount of boron decreases14 ofthe 22 amount of phosphor material, inhibiting photon production. 2.0 10 1.0 B2O3B2O3 (2:1)ZnS11.5um (2:1) 11.5 µm 1.0 2.0 1.8 10 B2O3B2O3 (1:1)ZnS11.5um (1:1) 11.5 µm 0.9 10 0.9 1.8 1.6 B2O3B2O3 (1:2)ZnS11.5um (1:2) 11.5 µm 10 0.8 B2O3B2O3 (2:1)ZnS4.7um (2:1) 4.7 µm 0.8 1.6 1.4 10 B2O3B2O3 (1:1)ZnS4.7um (1:1) 4.7 µm 10 1.2 0.7 B2O3B2O3 (1:2)ZnS4.7um (1:2) 4.7 µm 0.7 1.4 1.0 0.6 0.6 1.2 0.8 0.5 0.5 1.0 0.6 0.4 0.4 0.8 Neutronsper Grayscale 0.4

0.3 0.3 0.6 Relative Relative LightYield 0.2 Detection Efficiency 0.2 0.2 0.4 0.0 0 100 200 3000.1 0.1 Relative Neutronsper Grayscale 0.2 Screen Thickness (µm)0.0 0.0 0.0 6 0 100 200 300 0 100 200 300 0 100 200 300 J. Imaging 2020,Screen 6, x FOR Thickness PEER (µm) REVIEW Screen Thickness (µm) Screen Thickness (µm) 15 of 22 6 5 J. Imaging 2020, 6, x FOR PEER REVIEW 15 of 22 10 Figure 12. Results for 10 B2O3 screens with diﬀerent mix ratios and scintillator particle sizes. 5 Figurenat 12. Results for B2O3 screens with different mix ratios and scintillator particle sizes. 4 1.0 BNBN 11.5um(2:1)ZnS (2:1) 11.5 µm 1.0 6.0 BNnatBN 11.5um(1:1)ZnS (1:1) 11.5 µm 0.91.0 BNnatBN 11.5um(2:1)ZnS (2:1) 11.5 µm 0.91.0 6.0 4 BNnatBN 11.5um(1:2)ZnS (1:2) 11.5 µm 3 BNnatBN 11.5um(1:1)ZnS (1:1) 11.5 µm 5.0 0.80.9 BNnatBN 4.7um(2:1)ZnS (2:1) 4.7 µm 0.80.9 BNnatBN 11.5um(1:2)ZnS (1:2) 11.5 µm 3 natBN(1:1)ZnS 4.7 µm 5.0 0.7 BNnat 4.7um (1:1) 0.7 2 0.8 BNBN 4.7um(2:1)ZnS (2:1) 4.7 µm 0.8 Graylevel natBNBN 4.7um(1:2)ZnS (1:2) 4.7 µm 4.0 0.6 natBNBN 4.7um(1:1)ZnS (1:1) 4.7 µm 0.6

Neutronsper Graylevel 0.7 0.7 2 Graylevel natBNBN 4.7um(1:2)ZnS (1:2) 4.7 µm 4.0

1 0.50.6 0.50.6 3.0 Neutronsper Graylevel 1 0.40.5 0.40.5 0 3.0 2.0

0 100 200 3000.30.4 0.30.4 Relative Relative LightYield 0 DetectionEfficiency 2.0

0 Screen100 Thickness200 (µm) 3000.20.3 0.20.3 Relative Relative LightYield DetectionEfficiency 1.0

Screen Thickness (µm) 0.10.2 0.10.2 Relative Neutronsper 1.0 0.00.1 0.00.1 Relative Neutronsper 0.0 0.0 0 100 200 300 0.0 0 100 200 300 0.0 0 100 200 300 Screen Thickness (µm) Screen Thickness (µm) Screen Thickness (µm) 0 100 200 300 0 100 200 300 0 100 200 300 Screen Thickness (µm) Screen Thickness (µm) Screen Thickness (µm) Figure 13. Results for natBN screens with different mix ratios and scintillator particle sizes. Figure 13. Results for natBN screens with diﬀerent mix ratios and scintillator particle sizes. 3.0 Figure 13. Results for natBN screens with different mix ratios and scintillator particle sizes.

3.0 10 2.5 1.0 NaB5O8Na B5O 11.5um8(2:1)ZnS (2:1) 11.5 µm 1.0 2.0 10 NaB5O8Na B5O 11.5um8(1:1)ZnS (1:1) 11.5 µm 10 0.91.0 NaB5O8Na10 B5O 11.5um8(2:1)ZnS (2:1) 11.5 µm 0.91.0 1.82.0 2.5 NaB5O8Na B5O 11.5um8(1:2)ZnS (1:2) 11.5 µm 2.0 10 NaB5O8Na10 B5O 11.5um8(1:1)ZnS (1:1) 11.5 µm NaB5O8Na B5O 4.7um8(2:1)ZnS (2:1) 4.7 µm 0.80.9 10 0.80.9 1.61.8 NaB5O8Na10 B5O 11.5um8(1:2)ZnS (1:2) 11.5 µm NaB5O8Na B5O 4.7um8(1:1)ZnS (1:1) 4.7 µm 2.0 10 0.7 NaB5O8Na10 B5O 4.7um8(2:1)ZnS (2:1) 4.7 µm 0.7 1.4 1.5 0.8 NaB5O8Na B5O 4.7um8(1:2)ZnS (1:2) 4.7 µm 0.8 1.6 10 NaB5O8Na B5O 4.7um8(1:1)ZnS (1:1) 4.7 µm 1.5 0.60.7 10 0.60.7 1.21.4 1.0 NaB5O8Na B5O 4.7um8(1:2)ZnS (1:2) 4.7 µm 0.50.6 0.50.6 1.01.2

Neutronsper Grayscale 1.0 0.5 0.40.5 0.40.5 0.81.0

Neutronsper Grayscale 0.5 0.30.4 0.30.4 0.60.8 Relative LightYield 0.0 DetectionEfficiency

0 100 200 0.20.3 300 0.20.3 0.40.6 Relative LightYield 0.0 DetectionEfficiency 0 Screen100 Thickness200 (µm)0.10.2 300 0.10.2 Relative Neutronsper Grayscale 0.20.4

Screen Thickness (µm)0.00.1 0.00.1 Relative Neutronsper Grayscale 0.00.2 0.0 0 100 200 300 0.0 0 100 200 300 0.0 0 100 200 300 Screen Thickness (µm) Screen Thickness (µm) Screen Thickness (µm) 0 100 200 300 0 100 200 300 0 100 200 300 Screen Thickness (µm) Screen Thickness (µm) Screen Thickness (µm) 10 Figure 14. Results for Na B5O8 screens with different mix ratios and scintillator particle sizes. Figure 14. Results for Na10B5O8 screens with different mix ratios and scintillator particle sizes. RelativeFigure neutrons 14. Results per for graylevel, Na10B5O8 screens which with is the different ratio of mix the ratios detection and scintillator efficiency particle to the sizes. relative light yield,Two provides sizes of a measurescintillator of particles, how many 11.5 neutrons and 4.7 are μm, represented were used inin athe grayscale fabrication value of andthe screens. a useful Two sizes of scintillator particles, 11.5 and 4.7 μm, were used in the fabrication of the screens. However,measure to there compare was screen not aperformance. consistent trend Detection in the effi ciency performance scales with of screensthe amount with of borondifferent converter sized However, there was not a consistent trend in the performance of screens with different sized scintillatormaterial, simply particles. due Detection to the increased efficiency macroscopic was not affected absorption by scintillator cross-section particle of the size converter–scintillator because the bulk scintillator particles. Detection efficiency was not affected by scintillator particle size because the bulk boronmixture. concentration However, theof the screens mixture with is higher independent detection of escintillatorfficiency exhibited particle size fewer. Optical photons examination produced boron concentration of the mixture is independent of scintillator particle size. Optical examination didper detectednot reveal neutron any differences because an in increase homogeneity in the amountbetween of screens boron decreaseswith different the amount scintillator of phosphor particle did not reveal any differences in homogeneity between screens with different scintillator particle sizesmaterial,. inhibiting photon production. sizes. ATwo comparison sizes of scintillator of the performance particles, 11.5 of theand different 4.7 µm, were neutron used converters in the fabrication with a 2:1of the mix screens. ratio, A comparison of the performance of the different neutron 10 converters2 3 with a 2:1 mix ratio, mountedHowever, to there a polished was not aluminum a consistent substrate, trend in theis shown performance in Figure of screens15. B O with produced different the sized highest scintillator light 10 mounted to a polished10 5 8 aluminum substrate, is shown in Figure 15nat. B2O3 produced the highest light outputparticles. while Detection Na B O effi providedciency was the notbest adetectionffected by efficiency. scintillator The particle BN contained size because unenriched the bulk boron, boron whichoutput explains while Na its10B relatively5O8 provided poor the performance best detection compared efficiency. to the The other natBN screens contained that unenrichedused enriched boron, 10B. However,which explains based its solely relatively on the poor limited performance amount of compared 10B present to thein naturally other screens occurring that used boron enriched (19.9% 10B,B. 10 10 80.1%However, 11B), based10BN may solely be on su periorthe limited to 10B amount2O3 and Naof 10BB 5presentO8 if it were in naturally available occurring for future boron testing. (19.9% B, 80.1% 11B), 10BN may be superior to 10B2O3 and Na10B5O8 if it were available for future testing.

J. Imaging 2020, 6, 124 15 of 22

concentration of the mixture is independent of scintillator particle size. Optical examination did not reveal any differences in homogeneity between screens with different scintillator particle sizes. A comparison of the performance of the different neutron converters with a 2:1 mix ratio, mounted 10 to a polished aluminum substrate, is shown in Figure 15. B2O3 produced the highest light output 10 nat while Na B5O8 provided the best detection efficiency. The BN contained unenriched boron, which explains its relatively poor performance compared to the other screens that used enriched 10B. However, based solely on the limited amount of 10B present in naturally occurring boron (19.9% 10B, 80.1% 11B), J. Imaging 2020, 6, x FOR PEER REVIEW 16 of 22 J.10 Imaging 2020, 6, x FOR PEER REVIEW10 10 16 of 22 BN may be superior to B2O3 and Na B5O8 if it were available for future testing. 3.5 3.5 10 1.0 B2O3B2O 34.7um(2:1)ZnS (2:1) 4.7 µm 1.0 3.0 3.0 10 1.0 10B2O3B2O 34.7um(2:1)ZnS (2:1) 4.7 µm 1.0 3.0 3.0 B2O3B2O 311.5um(2:1)ZnS (2:1) 11.5 µm 0.9 10 0.9 BNnatB2O3BBN2 O4.7um(2:1)ZnS 311.5um(2:1)ZnS (2:1) (2:1)4.7 11.5 µm µm 2.5 0.9 nat 0.9 2.5 BNnatBN 11.5um4.7um(2:1)ZnS (2:1) (2:1) 11.54.7 µm µm 2.5 0.8 nat 0.8 2.5 BNBN10 11.5um(2:1)ZnS (2:1) 11.5 µm 0.8 NaB5O8Na B5O 84.7um(2:1)ZnS (2:1) 4.7 µm 0.8 2.0 10 0.7 NaB5O8Na10B5O 84.7um(2:1)ZnS (2:1) 4.7 µm 0.7 2.0 NaB5O8Na B5O 811.5um(2:1)ZnS (2:1) 11.5 µm Graylevel 0.7 10 0.7 2.0 NaB5O8Na B O 11.5um(2:1)ZnS (2:1) 11.5 µm Graylevel 1.5 0.6 5 8 0.6 2.0 1.5 0.6 0.6 1.0 0.5 0.5 1.5 1.0 0.5 0.5 1.5

Neutrons Graylevel per 0.4 0.4

Neutrons Graylevel per 0.5 0.4 0.4 1.0

0.5 0.3 0.3 1.0 Relative Relative LightYield

0.3 Detection Efficiency 0.3 Relative Relative LightYield 0.0 0.2 Detection Efficiency 0.2 0.0 0.5 0 100 200 3000.2 0.2

Relative Relative Neutronsper 0.5 0 100 200 0.1300 0.1 Screen Thickness (µm) 0.1 0.1 Relative Neutronsper Screen Thickness (µm) 0.0 0.0 0.0 0.0 0.0 0.0 0 100 200 300 0 100 200 300 0 100 200 300 0 Screen100 Thickness200 (µm) 300 0 Screen100 Thickness200 (µm) 300 0 Screen100 Thickness200 (µm) 300 Screen Thickness (µm) Screen Thickness (µm) Screen Thickness (µm) FigureFigure 15. 15. ResultsResults for for neutron neutron scintillator scintillator screens screens with with different diﬀerent neutron neutron converters. Figure 15. Results for neutron scintillator screens with different neutron converters. 6 TheThe top top performing boron boron-based‐based screens screens were compared with a 6LiFLiF/ZnS:Cu/ZnS:Cu screen screen comprised The 6top performing boron‐based screens were compared with a 6LiF/ZnS:Cu screen comprised ofof a a 1:2 ( 6LI:ZnS)LI:ZnS) mix mix ratio ratio under under cold cold neutron neutron exposure exposure,, for for which which the the results results are are summarized in in of a 1:2 (6LI:ZnS) mix ratio under cold neutron exposure, for which the results are summarized in FigureFigure 1616.. All All scintillator scintillator s screenscreens had had a a m matteatte b blacklack backing. backing. Figure 16. All scintillator screens had a matte black backing. 4.5 4.5 2.0 1.0 4.0 66LiFLiF(1:2)ZnS 4.7 um (1:2)4.7 µm 4.0 6 2.0 1.0 4.0 106LiFLiF(1:2)ZnS 4.7 um (1:2)4.7 µm 4.0 B2O3B2O3 (2:1)ZnS4.7um (2:1) 4.7 µm 10 1.8 0.9 10B2O3B2O3 (2:1)ZnS4.7um (2:1) 4.7 µm 3.5 3.5 B2O3B2O3 (1:1)ZnS4.7um (1:1) 4.7 µm 1.8 0.9 10 3.5 natB2O3 4.7um (1:1) 1.6 0.8 3.5 BNBBN2 O4.7um3(2:1)ZnS(1:1)ZnS (2:1) 4.7 4.7 µm µm nat 1.6 0.8 3.0 3.0 natBNBN 4.7um(1:1)ZnS(2:1)ZnS (1:1)(2:1) 4.7 µm nat 1.4 0.7 3.0 3.0 BNBN10 4.7um(1:1)ZnS (1:1) 4.7 µm Graylevel NaNaB5O8B5O8 4.7um(2:1)ZnS (2:1) 4.7 µm 1.4 0.7 10 2.5 Graylevel NaNaB5O810 B5O8 4.7um(2:1)ZnS (2:1) 4.7 µm 2.5 2.5 NaNaB5O8B5O8 (1:1)ZnS4.7um (1:1) 4.7 µm 1.2 0.6 2.5 NaNaB5O810B O (1:1)ZnS4.7um (1:1) 4.7 µm 1.2 0.6 2.0 5 8 1.0 0.5 2.0 2.0 1.0 0.5 1.5 2.0 0.8 0.4 1.5 1.5 0.8 0.4 1.5

0.6 0.3 Neutrons Graylevel per 1.0

Neutrons Graylevel per Relative Relative LightYield

0.6 DetectionEFficiency 0.3 1.0 1.0 Relative Relative LightYield 0.4 DetectionEFficiency 0.2 0.5 1.0 0.4 0.2 0.5 0.5 0.2 0.1 0.0 Relative Neutronsper 0.5 0.2 0.1 0.0 Relative Neutronsper 0.0 0.0 0 1000.0 200 300 0.0 0.0 0 1000.0 200 300 0 100 200 300 0 100 200 300Screen Thickness0 (µm)100 200 300 0 Screen100 Thickness200 (µm) 300 0 Screen100 Thickness200 (µm) 300Screen Thickness0 (µm)Screen100 Thickness200 (µm) 300 Screen Thickness (µm) Screen Thickness (µm) Screen Thickness (µm) Figure 16. Top-performing boron-based screens under a cold neutron beam compared to a Figure 16. Top‐performing boron‐based screens under a cold neutron beam compared to a 6LiF:ZnS Figure6LiF:ZnS 16. screen. Top‐performing boron‐based screens under a cold neutron beam compared to a 6LiF:ZnS screen. screen. The 6LiF-based screens produce significantly higher relative light yield than the boron-based The 6LiF‐based screens produce significantly higher relative light yield than the boron‐based screens,The which6LiF‐based is to be screens expected produce based onsignificantly the higher reactionhigher relative energy andlight longer yield rangethan the of 6 Li’sboron daughter‐based screens, which is to be expected based on the higher reaction energy and longer range of 6Li’s screens,products. which However, is to be the expected boron-based based screens on the exhibit higher a reaction much higher energy detection and longer efficiency range than of 6Li’s the daughter products. However, the boron‐based screens exhibit a much higher detection efficiency daughter6Li-based products. screen, with However, some screens the boron demonstrating‐based screens 125% exhibit higher a neutron much higher detection detection efficiency efficiency for the than the 6Li‐based screen, with some screens demonstrating 125% higher neutron detection efficiency thansame the thickness. 6Li‐based Their screen, higher with detection some screens efficiency demonstrating enables boron 125% screens higher to neutron provide detection improved efficiency neutron for the same thickness. Their higher detection efficiency enables boron screens to provide improved forcounting the same statistics, thickness. potentially Their higher accelerating detection image efficiency acquisition enables times boron proportionally screens to toprovide the improvement improved neutron counting statistics, potentially accelerating image acquisition times proportionally to the neutronin detection counting efficiency. statistics, potentially accelerating image acquisition times proportionally to the improvement in detection efficiency. improvementThe metric in ofdetection relative efficiency. neutrons per graylevel indicates that all boron-based scintillator screens The metric of relative neutrons per graylevel indicates that all boron‐based scintillator screens correlateThe metric with improved of relative neutron neutrons statistics per graylevel compared indicates to 6LiF screens.that all boron These‐based measurements scintillator show screens that correlate with improved neutron statistics compared to 6LiF screens. These measurements show that correlate6LiF screens with are impr brighter,oved neutron but the statistics brightness compared is not caused to 6LiF byscreens. the detection These measurements of additional show neutrons. that 6LiF screens are brighter, but the brightness is not caused by the detection of additional neutrons. 6LiF screens are brighter, but the brightness is not caused by the detection of additional neutrons. Borated scintillator screens that are thinner than 6LiF could offer the same neutron detection efficiency Borated scintillator screens that are thinner than 6LiF could offer the same neutron detection efficiency but improved resolution. These results align with theoretical predictions and justify the further but improved resolution. These results align with theoretical predictions and justify the further exploration of boron‐based scintillators in neuron imaging applications. exploration of boron‐based scintillators in neuron imaging applications. 3.2.2. Epithermal Neutron Results 3.2.2. Epithermal Neutron Results The absorption cross‐section of 10B is also higher than that of 6Li in the epithermal energy range, The absorption cross‐section of 10B is also higher than that of 6Li in the epithermal energy range, suggesting that 10B may exhibit superior performance in epithermal neutron imaging applications. suggesting that 10B may exhibit superior performance in epithermal neutron imaging applications. Epithermal neutron imaging is useful when examining large, dense samples because epithermal Epithermal neutron imaging is useful when examining large, dense samples because epithermal neutrons have greater penetration than thermal neutrons, usually at the expense of contrast [52]. neutrons have greater penetration than thermal neutrons, usually at the expense of contrast [52].

J. Imaging 2020, 6, 124 16 of 22

Borated scintillator screens that are thinner than 6LiF could oﬀer the same neutron detection eﬃciency but improved resolution. These results align with theoretical predictions and justify the further exploration of boron-based scintillators in neuron imaging applications.

3.2.2. Epithermal Neutron Results The absorption cross-section of 10B is also higher than that of 6Li in the epithermal energy range, suggesting that 10B may exhibit superior performance in epithermal neutron imaging applications. Epithermal neutron imaging is useful when examining large, dense samples because epithermal J.neutrons Imaging 2020 have, 6, x greaterFOR PEER penetration REVIEW than thermal neutrons, usually at the expense of contrast17 of [52 22]. Epithermal neutron imaging is used to examine both archeological artifacts [53,54] as well as nuclear Epithermal neutron imaging is used to examine both archeological artifacts [53,54] as well as nuclear fuels [4,55], among other objects. The epithermal region, which overlaps with the resonance region, fuels [4,55], among other objects. The epithermal region, which overlaps with the resonance region, contains many absorption resonances for different materials. These resonances can be exploited using contains many absorption resonances for different materials. These resonances can be exploited using pulsed neutrons and time-of-flight techniques to determine the isotopic composition of a material [56]. pulsed neutrons and time‐of‐flight techniques to determine the isotopic composition of a material Neutron scintillator screens with an improved detection efficiency for epithermal neutrons could also [56]. Neutron scintillator screens with an improved detection efficiency for epithermal neutrons could be useful in resonance imaging techniques [57–59]. also be useful in resonance imaging techniques [57–59]. In this work, a 1 mm-thick sheet of cadmium was placed in the neutron beam to absorb thermal In this work, a 1 mm‐thick sheet of cadmium was placed in the neutron beam to absorb thermal neutrons, leaving primarily epithermal neutrons. Borated scintillator screen performance was measured neutrons, leaving primarily epithermal neutrons. Borated scintillator screen performance was under epithermal neutron exposure following the same procedure as with the unfiltered neutron beam. measured under epithermal neutron exposure following the same procedure as with the unfiltered Figure 17 shows the top performing boron-based scintillators compared to a common 6LiF scintillator. neutron beam. Figure 17 shows the top performing boron‐based scintillators compared to a common The 6LiF scintillator exhibited superior light output compared to the borated screens under epithermal 6LiF scintillator. The 6LiF scintillator exhibited superior light output compared to the borated screens neutron exposure, as was the case with cold neutrons. However, the boron-based screens exhibit a under epithermal neutron exposure, as was the case with cold neutrons. However, the boron‐based much higher detection efficiency than the 6Li-based screen, with some screens exhibiting 67% higher screens exhibit a much higher detection efficiency than the 6Li‐based screen, with some screens epithermal neutron detection efficiency for the same thickness. Epithermal neutron tomography can exhibiting 67% higher epithermal neutron detection efficiency for the same thickness. Epithermal take more than a day to acquire a full set of projections, but these results show that boron-based neutron tomography can take more than a day to acquire a full set of projections, but these results screens have the potential to reduce the time necessary for an epithermal neutron tomography by show that boron‐based screens have the potential to reduce the time necessary for an epithermal approximately 60%. Additionally, the relative neutrons per graylevel is significantly higher for the neutron tomography by approximately 60%. Additionally, the relative neutrons per graylevel is boron-based scintillators than for the 6LiF scintillator, which could indicate improved SNR for the significantly higher for the boron‐based scintillators than for the 6LiF scintillator, which could images produced using the boron-based scintillators. indicate improved SNR for the images produced using the boron‐based scintillators. 1.0 1.4 0.10 1.0 LiF6LiF(1:2)ZnS 4.7um (1:2) 4.7 µm 10 B2O3B2O 34.7um(2:1)ZnS (2:1) 4.7 µm 0.09 0.80.9 10 1.2 B2O3B2O 34.7um(1:1)ZnS (1:1) 4.7 µm BNnat 4.7um (2:1) 0.08 Graylevel 0.8 BN(2:1)ZnS 4.7 µm nat 1.0 0.6 BNBN 4.7um(1:1)ZnS (1:1) 4.7 µm 0.07 0.7 10 Graylevel NaNaB5O8B5O8 (2:1)ZnS4.7um (2:1) 4.7 µm 0.8 0.06 0.40.6 0.05 0.5 0.6 0.2 0.04 0.4

0.4 0.03 0.00.3

Relative Relative LightYield Detection Efficiency

0.02 Relative Neutronsper 0.20 100 200 300 0.2

0.01 Relative per Neutrons 0.1 Screen Thickness (μm) 0.0 0.00 0.0 0 100 200 300 0 100 200 300 0 100 200 300 Screen Thickness (μm) Screen Thickness (μm) Screen Thickness (μm) Figure 17. Top-performing boron-based screens under an epithermal neutron beam compared to a Figure 17. Top‐performing boron‐based screens under an epithermal neutron beam compared to a 6LiF:ZnS screen. 6LiF:ZnS screen. 3.2.3. Spatial Resolution 3.2.3. Spatial Resolution Spatial resolution of the high-resolution boron screens was measured using the high-resolution imagingSpatial system resolution at ANTARES. of the high The‐ eresolutionﬀective pixel boron pitch screens of this was system measured was 57 pixelsusing/ themm high (17.5‐resolutionµm/pixel), imagingcorresponding system to at a ANTARES. Nyquist frequency The effective of 28.5 pixel lp/mm. pitch Figure of this 18 system shows was the MTF 57 pixels/mm results for (17.5 the µm/pixel),high-resolution corresponding screens, which to a Nyquist consisted frequency of thinly depositedof 28.5 lp/mm. (10–20 Figureµm) scintillator18 shows the layers MTF with results 4.7 µform themedian high‐ scintillatorresolution particlescreens, size.which The consisted spatial resolutionof thinly deposited of all high (10 resolution–20 μm) scintillator screens is displayed layers with in 4.7 μm median scintillator particle size. The spatial resolution10 of all high resolution natscreens is Table5 and the MTF curves of the top two performers, a layered B4C/ZnS screen and a BN/ZnS displayed in Table 5 and the MTF curves of the top two performers, a layered 10B4C/ZnS screen and a natBN/ZnS screen, are displayed in Figure 18. Discrepancies in the spatial resolution for screens of the same material can be attributed to variations in the already the thin (≤20 μm) target scintillator layer thickness. At thin scintillator layers, a slight variation in the amount of deposited scintillator material or surface defects can drastically affect screen performance.

J. Imaging 2020, 6, x FOR PEER REVIEW 18 of 22

Table 5. Effective spatial resolution for uniform thickness screens.

Converter to Scintillator Resolution at 10% Converter Atomic Ratio Modulation Transfer Material (Boron atom to ZnS:Cu) Function (MTF) (lp/mm) 10B2O3 2:1 10.6 10B2O3 2:1 7.6 10B2O3 1:1 9.6 10B2O3 1:1 9.6 10B2O3 1:2 7.4 10B2O3 1:2 7.3 10NaB5O8 2:1 10.5 10NaB5O8 2:1 8.8 10NaB5O8 1:1 10.2 10NaB5O8 1:1 9.9 10NaB5O8 1:2 7.6 10NaB5O8 1:2 8.2 natBN 2:1 4.8 J. Imaging 2020, 6, 124natBN 2:1 9.7 17 of 22 natBN 1:1 8.3 nat screen, are displayedBN in Figure 18. Discrepancies1:1 in the spatial resolution6.7 for screens of the same nat material can be attributedBN to variations in the1:2 already the thin ( 20 µm) target6.6 scintillator layer thickness. nat ≤ At thin scintillatorBN layers, a slight variation1:2 in the amount of deposited scintillator15.2 material or surface 10 10 defects canLayered drasticallyB a4Cﬀect screen10 μm performance.B4C, 20 μm ZnS:Cu 16.0

1.0 10 LayeredLayered B4C-ZnS:CuB4C-ZnS:Cu 0.9 natBN(1:2)ZnS:CuBN(1:2)ZnS:Cu 0.8 10% MTF 0.7 0.6 0.5 0.4 0.3

0.2 Modulation Transfer Function 0.1 0.0 0 5 10 15 20 Spatial Resolution (lp/mm) Figure 18. MTF curves of high-resolution scintillators with the best spatial resolution. The boron Figure 18. MTF curves of high‐resolution scintillators with the best spatial resolution. The boron carbide screen is layered with ZnS, while the boron nitride screen is mixed with the ZnS scintillator. carbide screen is layered with ZnS, while the boron nitride screen is mixed with the ZnS scintillator. Table 5. Eﬀective spatial resolution for uniform thickness screens. 3.2.4. Neutron Computed Tomography Converter to Scintillator Atomic Ratio Resolution at 10% Modulation Converter Material Radiographs were acquired with(Boron some atom of to the ZnS:Cu) high‐resolutionTransfer scintillator Function screens. (MTF) These (lp/ mm)screens were chosen10 based on their promising performance, but also for the relatively uniform surface texture B2O3 2:1 10.6 10 in the centralB2 Ofield3 of view of the screens. Fi2:1gure 19 (left) shows a radiograph of 7.6a small mechanical 10 wristwatch Bacquired2O3 using a thin 10B2O3/ZnS:Cu1:1 screen. A set of 1229 radiographs 9.6 was acquired for 10 neutron computedB2O3 tomography, and the resulting1:1 3D rendering of the wristwatch 9.6 is displayed in the 10 B2O3 1:2 7.4 right side 10of Figure 19. The mechanical gears inside the watch are clearly visible, demonstrating that B2O3 1:2 7.3 these prototype10 screens are feasible to use in neutron radiography and tomography applications. NaB5O8 2:1 10.5 10 NaB5O8 2:1 8.8 10 NaB5O8 1:1 10.2 10 NaB5O8 1:1 9.9 10 NaB5O8 1:2 7.6 10 NaB5O8 1:2 8.2 natBN 2:1 4.8 natBN 2:1 9.7 natBN 1:1 8.3 natBN 1:1 6.7 natBN 1:2 6.6 natBN 1:2 15.2 10 10 Layered B4C 10 µm B4C, 20 µm ZnS:Cu 16.0

3.2.4. Neutron Computed Tomography Radiographs were acquired with some of the high-resolution scintillator screens. These screens were chosen based on their promising performance, but also for the relatively uniform surface texture in the central ﬁeld of view of the screens. Figure 19 (left) shows a radiograph of a small mechanical 10 wristwatch acquired using a thin B2O3/ZnS:Cu screen. A set of 1229 radiographs was acquired for neutron computed tomography, and the resulting 3D rendering of the wristwatch is displayed in the J. Imaging 2020, 6, 124 18 of 22 right side of Figure 19. The mechanical gears inside the watch are clearly visible, demonstrating that these prototype screens are feasible to use in neutron radiography and tomography applications. J. Imaging 2020, 6, x FOR PEER REVIEW 19 of 22

Figure 19. A neutron radiographradiograph ((leftleft)) andand aa cut-awaycut‐away of of a a 3D 3D tomographic tomographic reconstruction reconstruction (right (right) of) of a mechanicala mechanical wristwatch. wristwatch.

4. Discussion Boron-basedBoron‐based scintillator scintillator screens screens demonstrated demonstrated superior superior neutron neutron detection detection effi ciency efficiency for both for both cold 6 andcold epithermal and epithermal neutron neutron applications applications when compared when compared to widely to used widelyLiF /ZnSused neutron 6LiF/ZnS scintillator neutron screens.scintillator Borated screens. screens Borated demonstrated screens demonstrated up to 125% higher up coldto 125% neutron higher detection cold eneutronfficiency detection and 67% 6 higherefficiency epithermal and 67% neutronhigher epithermal detection e neutronfficiency detection than representative efficiency thanLiF representative screens. Although 6LiF boratedscreens. 6 screensAlthough produce borated less screens light output produce than less representative light outputLiF than screens, representative the borated 6LiF screens screens deliver, the improved borated 6 neutronscreens deliver counting improved statistics neutron compared counting to representative statistics comparedLiF screens. to representative The higher neutron 6LiF screens detection. The ehigherfficiency neutron of boron-based detection screensefficiency off ofers boron the potential‐based screens for higher offers quality the potential neutron imagesfor higher and quality lower acquisitionneutron imag times.es and Epithermal lower acquisition neutron tomographytimes. Epithermal acquisition neutron time tomography can potentially acquisition be reduced time from can dayspotentially to hours, be reduced depending from on days the application. to hours, depending on the application. Thin boron-basedboron‐based scintillator screens were measured to compare compare theirtheir spatial spatial resolution resolution performance.performance. WhileWhile some some of the of boron-based the boron screens‐based provided screens relatively provided high-resolution relatively high performance,‐resolution theperformance, resolution the performance resolution wasperformance limited by was the limited Nyquist by limit the Nyquist of the imaging limit of systemthe imaging employed system in thisemployed study. in Further this study. measurements Further measurements should be pursued should using be pursued a neutron using imaging a neutron system imaging with a system higher Nyquistwith a higher limit toNyquist determine limit the to determine spatial resolution the spatial potential resolution of boron-based potential of screens. boron‐based screens. The top performing screens successfully producedproduced a useful tomographictomographic reconstruction of a test object. Boron Boron-based‐based screens may be able to deliver imp improvedroved spatial resolution over common widely used screens based on thethe fundamentalfundamental properties of the neutronneutron absorptionabsorption daughter products of 10 6 10BB compared to those of 6LiLi.. Provided Provided an an imaging imaging setup setup has has sufficient sufficient light light sensitivity, boron-basedboron‐based neutron scintillatorsscintillators cancan provide improved neutronneutron detectiondetection statistics.statistics. The higher neutron detection eefficiencyfficiency of boron-basedboron‐based screens stands to benefitbenefit thethe world-wideworld‐wide neutron imaging community by increasingincreasing neutronneutron countingcounting statisticsstatistics andand reducingreducing measurementmeasurement times.times.

Author Contributions: Contributions:Conceptualization, Conceptualization W.C.,, W.C., A.C., A.C., S.C., S.C., and andB.S.; methodology,B.S.; methodology, W.C. and W.C. A.C.; and formal A.C. analysis,; formal W.C. and A.C.; investigation, W.C., A.C., B.S., A.T.; resources, S.C.; data curation, W.C.; writing—original draft preparation,analysis, W.C. W.C.; and writing—review A.C.; investigation, and editing,W.C., A.C., W.C., B.S., A.C., A.T B.S.,.; resources, S.C. and A.T.;S.C.; visualization,data curation, W.C.; W.C. supervision,.; writing— A.C.;original project draft administration, preparation, W.C. A.C.;; writing funding—review acquisition, and editing, A.C. All W.C., authors A.C., have B.S., readS.C. and and A.T agreed.; visualization, to the published W.C.; versionsupervision, of the A.C manuscript..; project administration, A.C.; funding acquisition, A.C. All authors have read and agreed to Funding:the publishedWork version funded of through the manuscript. the INL Laboratory Directed Research & Development (LDRD) Program under DOEFunding: Idaho Work Operations funded Othroughﬃce Contract the INL DE-AC07-05ID14517. Laboratory Directed LDRDResearch Project & Development ID# 20A44-200FP. (LDRD) Program under DOE Idaho Operations Office Contract DE‐AC07‐05ID14517. LDRD Project ID# 20A44‐200FP.

Acknowledgments: This work is based upon experiments performed at the ANTARES instrument operated by FRM II at Heinz Maier‐Leibnitz Zentrum (MLZ), Garching, Germany. The authors would like to thank the FRM II staff as well as Amanda Smolinski for their assistance during the measurement process. DMI/Reading Imaging fabricated and prepared the various materials (e.g., the anhydrous 10B converter compounds and scintillators) and the coating formulations, fabricated the sensors used in this study, and measured the Na10B5O8 density.

J. Imaging 2020, 6, 124 19 of 22

Acknowledgments: This work is based upon experiments performed at the ANTARES instrument operated by FRM II at Heinz Maier-Leibnitz Zentrum (MLZ), Garching, Germany. The authors would like to thank the FRM II staff as well as Amanda Smolinski for their assistance during the measurement process. DMI/Reading Imaging fabricated and prepared the various materials (e.g., the anhydrous 10B converter compounds and scintillators) 10 and the coating formulations, fabricated the sensors used in this study, and measured the Na B5O8 density. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Lehmann, E.H.; Vontobel, P.; Hermann, A. Non-destructive analysis of nuclear fuel by means of thermal and cold neutrons. Nucl. Instrum. Methods Phys. Res. A 2003, 515, 745–759. [CrossRef] 2. Sim, C.-M.; Kim, T.; Oh, H.S.; Kim, J.C. Measurement of Ballooning Gap Size of Irradiated Fuels Using Neutron Radiography Transfer Method and HV Image Filter. J. Korean Soc. Nondestruct. Test. 2013, 33, 212–218. [CrossRef] 3. Tremsin, A.S.; Vogel, S.C.; Mocko, M.; Bourke, M.A.M.; Yuan, V.; Nelson, R.O.; Brown, D.W.; Feller, W.B. Non-destructive studies of fuel pellets by neutron resonance absorption radiography and thermal neutron radiography. J. Nucl. Mater. 2013, 440, 633–646. [CrossRef] 4. Craft, A.E.; Wachs, D.M.; Okuniewski, M.A.; Chichester, D.L.; Williams, W.J.; Papaioannou, G.C.; Smolinski, A.T. Neutron Radiography of Irradiated Nuclear Fuel at Idaho National Laboratory. Phys. Procedia 2015, 69, 483–490. [CrossRef] 5. Tremsin, A.S.; Craft, A.E.; Papaioannou, G.C.; Smolinski, A.T.; Boulton, N.M.; Ruddell, M.A.; Littell, B.J.; Riley, K.D. On the possibility to investigate irradiated fuel pins non-destructively by digital neutron radiography with a neutron-sensitive microchannel plate detector with Timepix readout. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2019, 927, 109–118. [CrossRef] 6. Schulthess, J.; Woolstenhulme, N.; Craft, A.; Kane, J.; Boulton, N.; Chuirazzi, W.; Winston, A.; Smolinski, A.; Jensen, C.; Kamerman, D.; et al. Non-Destructive Post-irradiation Examination Results of the First Modern Fueled Experiments in TREAT. J. Nucl. Mater. 2020, 541, 152442. [CrossRef] 7. Lehmann, E.H.; Vontobel, P.; Deschler-Erb, E.; Soares, M. Non-invasive studies of objects from cultural heritage. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2005, 42, 68–75. [CrossRef] 8. Rant, J.; Miliˇc,Z.; Isteniˇc,J.; Knific, T.; Lengar, I.; Rant, A. Neutron radiography examination of objects belonging to the cultural heritage. Appl. Radiat. Isot. 2006, 64, 7–12. [CrossRef] 9. Mannes, D.; Schmid, F.; Frey, J.; Schmidt-Ott, K.; Lehmann, E. Combined Neutron and X-ray Imaging for Non-invasive Investigations of Cultural Heritage Objects. Phys. Procedia 2015, 69, 653–660. [CrossRef] 10. Iranzo, A.; Boillat, P.; Rosa, F. Validation of a three dimensional PEM fuel cell CFD model using local liquid water distributions measured with neutron imaging. Int. J. Hydrogen Energy 2014, 39, 7089–7099. [CrossRef] 11. Makowska, M.G.; Strobl, M.; Lauridsen, E.M.; Frandsen, H.L.; Tremsin, A.S.; Kardjilov, N.; Manke, I.; Kelleher, J.F.; Theil Kuhn, L. Effect of stress on NiO reduction in solid oxide fuel cells: A new application of energy-resolved neutron imaging. J. Appl. Crystallogr. 2015, 48, 401–408. [CrossRef] 12. Cochet, M.; Forner-Cuenca, A.; Manzi, V.; Siegwart, M.; Scheuble, D.; Boillat, P.Novel Concept for Evaporative Cooling of Fuel Cells: An Experimental Study Based on Neutron Imaging. Fuel Cells 2018, 18, 619–626. [CrossRef] 13. Cho, J.I.S.; Neville, T.P.; Trogadas, P.; Meyer, Q.; Wu, Y.; Ziesche, R.; Boillat, P.; Cochet, M.; Manzi-Orezzoli, V.; Shearing, P.; et al. Visualization of liquid water in a lung-inspired flow-field based polymer electrolyte membrane fuel cell via neutron radiography. Energy 2019, 170, 14–21. [CrossRef] 14. Kulkarni, N.; Meyer, Q.; Hack, J.; Jervis, R.; Iacoviello, F.; Ronaszegi, K.; Adcock, P.; Shearing, P.R.; Brett, D.J.L. Examining the effect of the secondary flow-field on polymer electrolyte fuel cells using X-ray computed radiography and computational modelling. Int. J. Hydrogen Energy 2019, 44, 1139–1150. [CrossRef] 15. Koerner, S.; Schillinger, B.; Vontobel, P.; Rauch, H. A neutron tomography facility at a low power research reactor. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2001, 471, 69–74. [CrossRef] J. Imaging 2020, 6, 124 20 of 22

16. Sim, C.M.; Oh, H.S.; Kim, T.J.; Lee, Y.S.; Kim, Y.K.; Kwak, S.S.; Hwang, Y.H. Detecting Internal Hot Corrosion of In-service Turbine Blades Using Neutron Tomography with Gd Tagging. J. Nondestruct. Eval. 2014, 33, 493–503. [CrossRef] 17. Peetermans, S.; Lehmann, E.H. Simultaneous neutron transmission and diffraction imaging investigations of single crystal nickel-based superalloy turbine blades. NDT E Int. 2016, 79, 109–113. [CrossRef] 18. Lehmann, E.H. Neutron imaging facilities in a global context. J. Imaging 2017, 3, 52. [CrossRef] 19. Stedman, R. Scintillator for thermal neutrons using Li6F and ZnS (Ag). Rev. Sci. Instrum. 1960, 31, 1156. [CrossRef] 20. Mantler-Niederstätter, F.; Bensch, F.; Grass, F. An extremely thin scintillation detector for thermal neutrons. Nucl. Instrum. Methods 1977, 142, 463–466. [CrossRef] 21. Knoll, G.F. Radiation Detection and Measurements, 3rd ed.; Wiley and Sons: Hoboken, NJ, USA, 2000; ISBN 0471073385. 22. Katagiri, M.; Sakasai, K.; Matsubayashi, M.; Nakamura, T.; Kondo, Y.; Chujo, Y.; Nanto, H.; Kojima, T. Scintillation materials for neutron imaging detectors. Nucl. Instrum. Methods Phys. Res. A 2004, 529, 274–279. [CrossRef] 23. Crha, J.; Vila-Comamala, J.; Lehmann, E.; David, C.; Trtik, P. Light Yield Enhancement of 157-Gadolinium Oxysulfide Scintillator Screens for the High-Resolution Neutron Imaging. MethodsX 2019, 6, 107–114. [CrossRef][PubMed] 24. Sun, K.H.; Malmberg, P.R.; Pecjak, F.A. High-Efficiency Slow-Neutron Scintillation Counters. Nucleonics 1956, 14, 46–49. 25. McGonnagle, W.J. Symposium on Physics and Nondestructive Testing; Argonne National Laboratory: Argonne, IL, USA, 1960; ANL-6346. 26. Kojima, T.; Katagiri, M.; Tsutsui, N.; Imai, K.; Matsubayashi, M.; Sakasai, K. Neutron scintillators with high detection efficiency. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2004, 529, 325–328. [CrossRef] 27. Katagiri, M.; Sakasai, K.; Matsubayashi, M.; Kojima, T. Neutron/γ-ray discrimination characteristics of novel neutron scintillators. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2004, 529, 317–320. [CrossRef] 28. Ishii, M.; Kuwano, Y.; Asai, T.; Asaba, S.; Kawamura, M.; Senguttuvan, N. Boron based oxide scintillation glass for neutron detection. Nucl. Instrum. Methods Phys. Res. A Accel. Spectrometers Detect. Assoc. Equip. 2005, 537, 282–285. [CrossRef] 29. Czirr, J.B.; Macgillivray, G.M.; Macgillivray, R.R.; Seddon, P.J. Performance and characteristics of a new scintillator. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 1999, 424, 15–19. [CrossRef] 30. Williams, A.M.; Beeley, P.A.; Spyrou, N.M. Response of a lithium gadolinium borate scintillator in monoenergetic neutron fields. Radiat. Prot. Dosim. 2004, 110, 497–502. [CrossRef] 31. Normand, S.; Mouanda, B.; Haan, S.; Louvel, M. Discrimination methods between neutron and gamma rays for boron loaded plastic scintillators. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2002, 484, 342–350. [CrossRef] 32. Britvich, G.I.; Vasil’chenko, V.G.; Gilitsky, Y.V.; Chubenko, A.P.; Kushnirenko, A.E.; Mamidzhanyan, E.A.; Pavluchenko, V.P.; Pikalov, V.A.; Romakhin, V.A.; Soldatov, A.P.; et al. A neutron detector on the basis of a boron-containing plastic scintillator. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2005, 550, 343–358. [CrossRef] 33. Mahl, A.; Yemam, H.A.; Stuntz, J.; Remedes, T.; Sellinger, A.; Greife, U. Bis(pinacolato)diboron as an additive for the detection of thermal neutrons in plastic scintillators. Nucl. Instrum. Methods Phys. Res. A 2016, 816, 96–100. [CrossRef] 34. Mahl, A.; Yemam, H.A.; Fernando, R.; Koubek, J.T.; Sellinger, A. 10B enriched plastic scintillators for application in thermal neutron detection. Nucl. Inst. Methods Phys. Res. A 2018, 880, 1–5. [CrossRef] 35. Verhaeghe, J.L.; Thielens, G.; Creten, W.L. Preparation and properties of a boron containing scintillator for the detection of slow neutrons. Appl. Sci. Res. Sect. B 1962, 10, 247–256. [CrossRef] 36. Brenizer, J.S.; Berger, H.; Stebbings, C.T.; Gillies, G.T. Performance characteristics of scintillators for use in an electronic neutron imaging system for neutron radiography. Rev. Sci. Instrum. 1997, 68, 3371–3379. [CrossRef] J. Imaging 2020, 6, 124 21 of 22

37. McMillan, J.E.; Cole, A.J.; Kirby, A.; Marsden, E. Thermal neutron scintillators using unenriched boron nitride and zinc sulfide. J. Phys. Conf. Ser. 2015, 620, 012011. [CrossRef] 38. Nakamura, T.; Katagiri, M.; Tsutsui, N.; Toh, K.; Rhodes, N.; Schooneveld, E.; Ooguri, H.; Noguchi, Y.; Sakasai, K.; Soyama, K. Development of a ZnS/10B2O3 scintillator with low- afterglow phosphor Development of a ZnS/10B2O3 scintillator with low-afterglow phosphor. Int. Work. Neutron Opt. Detect. 2014, 528, 1–8. 39. Miller, S.R.; Marshall, M.S.J.; Wart, M.; Crha, J.; Trtik, P.; Nagarkar, V.V. High Resolution Thermal Neutron Imaging with 10Boron/CsI:Tl Scintillator Screen. IEEE Trans. Nucl. Sci. 2020, 9499, 1–5. 40. Van Delinder, K.W.; Khan, R.; Gräfe, J.L. Neutron activation of gadolinium for ion therapy: A Monte Carlo study of charged particle beams. Sci. Rep. 2020, 10, 1–11. [CrossRef] 41. Miller, G.A.; Hertel, N.E.; Wehring, B.W.; Horton, J.L. Gadolinium neutron capture therapy. Nucl. Technol. 1993, 103, 320–331. [CrossRef] 42. Kandlakunta, P.; Cao, L.R.; Mulligan, P. Measurement of internal conversion electrons from Gd neutron capture. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2013, 705, 36–41. [CrossRef] 43. Goorley, T.; Nikjoo, H. Electron and photon spectra for three gadolinium-based cancer therapy approaches. Radiat. Res. 2000, 154, 556–563. [CrossRef] 44. Abdushukurov, D.A.; Abduvokhldov, M.A.; Bondarenko, D.V.; Mumlnov, K.K.; Toshov, T.A.; Chlstyakov, D.Y. Modeling the registration efficiency of thermal neutrons by gadolinium foils. J. Instrum. 2007, 2, 04001. [CrossRef] 45. Berger, M.J.; Coursey, J.S.; Zucker, M.A.; Chang, J. Stopping-Power and Range Tables For Electrons, Protons, and Helium Ions; NIST Physics Laboratory: Gaithersburg, MD, USA, 2017. 46. Ziegler, J.F.; Ziegler, M.D.; Biersack, J.P. SRIM-The stopping and range of ions in matter (2010). Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 2010, 268, 1818–1823. [CrossRef] 47. Salvat, F.; Fernandez-Varea, J.M.; Sempau, J. PENELOPE-2006: A code system for Monte Carlo simulation of electron and photon transport. Work. Proc. 2006, 4, 6222. 48. Detail Information of NaB5H10O1; Catalyst Hub. Available online: http://www.catalysthub.net/materials.php? id=1830324,2016 (accessed on 27 October 2020). 49. Chuirazzi, W.; Craft, A. Measuring Thickness-Dependent Relative Light Yield and Detection Efficiency of Scintillator Screens. J. Imaging 2020, 6, 56. [CrossRef] 50. Schulz, M.; Schillinger, B. ANTARES: Cold neutron radiography and tomography facility. J. Large-Scale Res. Facil. JLSRF 2015, 1, 8–11. [CrossRef] 51. Andor, iKon-L Series iKon-L Series. Available online: https://andor.oxinst.com/products/ikon-xl-and-ikon- large-ccd-series/ikon-l-936 (accessed on 12 February 2020). 52. Schillinger, B.; Craft, A.E. Epithermal Neutron Radiography and Tomography on Large and Strongly Scattering Samples. Mater. Res. Proc. 2020, 15, 142–148. 53. Schooneveld, E.M.; Tardocchi, M.; Gorini, G.; Kockelmann, W.; Nakamura, T.; Perelli Cippo, E.; Postma, H.; Rhodes, N.; Schillebeeckx, P. A new position-sensitive transmission detector for epithermal neutron imaging. J. Phys. D Appl. Phys. 2009, 42, 152003. [CrossRef] 54. Cippo, E.P.; Borella, A.; Gorini, G.; Kockelmann, W.; Moxon, M.; Postma, H.; Rhodes, N.J.; Schillebeeckx, P.; Schoonenveld, E.M.; Tardocchi, M.; et al. Imaging of cultural heritage objects using neutron resonances. J. Anal. At. Spectrom. 2011, 26, 992–999. [CrossRef] 55. Vogel, S.C. A Review of Neutron Scattering Applications to Nuclear Materials. ISRN Mater. Sci. 2013, 2013, 1–24. [CrossRef] 56. Vogel,S.C.; Bourke, M.A.M.; Craft, A.E.; Harp, J.M.; Kelsey,C.T.; Lin, J.; Long, A.M.; Losko, A.S.; Hosemann, P.; McClellan, K.J.; et al. Advanced Postirradiation Characterization of Nuclear Fuels Using Pulsed Neutrons. Jom 2020, 72, 187–196. [CrossRef] 57. Tremsin, A.S.; McPhate, J.B.; Vallerga, J.V.; Siegmund, O.H.W.; Kockelmann, W.; Schooneveld, E.M.; Rhodes, N.J.; Bruce Feller, W. High resolution neutron resonance absorption imaging at a pulsed neutron beamline. IEEE Trans. Nucl. Sci. 2012, 59, 3272–3277. [CrossRef] J. Imaging 2020, 6, 124 22 of 22

58. Tremsin, A.S.; Kockelmann, W.; Pooley, D.E.; Feller, W.B. Spatially resolved remote measurement of temperature by neutron resonance absorption. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2015, 803, 15–23. [CrossRef] 59. Dangendorf, V.; Laczko, G.; Kersten, C. Fast neutron resonance radiography in a pulsed neutron beam. arXiv 2003, arXiv:nucl-ex/0301001v1.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional aﬃliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).