Transparent Microcomposite Films Based on a Ce-Doped Li6Gd(BO3)3 Scintillator for Radiation Detection.

Scintillators are widely used for high-energy radiation detection. Hybrid inorganic-organic composite scintillators with high light yields, high light decay rates, excellent stability, and low costs are in great demand. Here, we report a novel scintillator composed of Ce-doped Li6Gd(BO3)3 (LGBO) microphosphors (MPs) and polymethyl methacrylate for X-ray and thermal neutron detection. The Ce-doped LGBO MPs, fabricated using a facile high-temperature solid-state reaction method, exhibit intense blue light at 416 nm under X-ray and UV excitation and have a high photoluminescence quantum yield of ∼63%. More importantly, the composite scintillator based on these MPs has excellent transparency and luminescence intensity. The luminescence integral intensity of composite scintillators is superior to that of commercial CsI:Na under X-ray excitation, and the light yield under thermal neutron irradiation is 21,000 photons/thermal neutron. The scintillation decay time is found to be below 600 ns. The neutron-gamma signal discrimination and neutron detection efficiency of the composite scintillators are acceptable for practical application. There is an excellent separation between neutron and background events. It represents significant improvements in scintillator performances, especially for reliable thermal neutron scintillators that are likely to improve the data qualities of scientific instruments, including charge-coupled device-based imagers and Anger logic-based position-sensitive detectors in neutron user facilities.

Introduction

Scintillator-based detectors are considerably important in fields including neutron and X-ray scattering science, high-energy physics, medical imaging, and industrial testing and exploration.[1−3] As a core part of the detectors, scintillators convert the energy of high-energy rays or particles into visible photons. A higher light yield, shorter decay time, better stability, and more reasonable production cost have always been pursued in developing these materials.[4] Compared with conventional inorganic scintillation single crystals or transparent ceramics, inorganic particle–polymer composite scintillators not only have easier preparation processes, low cost, and flexibility but also have reasonable performances in detection efficiency, response speed, and so forth.[5,6] Meanwhile, compared to conventional organic scintillators, the addition of inorganic particles increases the effective atomic number (Zeff) of the scintillator, thus improving the ability to stop X-rays and photoelectron production. Inorganic particle–polymer composite scintillators are typically fabricated by combining luminescent quantum dot nanoparticles such as perovskites or CdSe/ZnS composited with organic polymers such as polymethyl methacrylate (PMMA), polystyrene (PS), polydimethylsiloxane (PDMS), and so forth.[7−9] For X-ray detection, excellent chemical stability and intense photon emission are required. Quantum dot materials such as perovskites still have problems including poor stability and low loading fraction (<5 wt %) of inorganic nanoparticles in its composite scintillators, which result in a poor attenuation of X-rays.[5,10] In the field of thermal neutron detection, the light yield, detection efficiency, and decay time of a scintillator are important parameters. In order to capture thermal neutrons, it is necessary to incorporate or contain 6Li, 10B, and 157Gd in the scintillator material. 157Gd has the highest thermal neutron absorption cross-section but suffers from the problem of difficult neutron−gamma discrimination.[11] A common approach is to add 6LiF to the scintillator, as in the typical thermal neutron scintillator ZnS:Ag/6LiF, but this leads to the opaqueness of the materials, further reducing the light yield and neutron detection efficiency.[12]

As a common route for scintillator developments, Ce-doped oxide inorganic microparticles prepared by a high-temperature solid-state reaction may have excellent stability and good scintillation properties. Previously, we have provided a solution to the problem of neutron−gamma discrimination for Ce-doped Gd-containing oxide scintillators in thermal neutron detection.[13,14] However, there are few investigations on the possible applications of Ce-doped Gd-containing oxide inorganic particle–polymer composite scintillators for either thermal neutron or X-ray detection. Among these oxides, Ce-doped Li6Gd(BO3)3 (LGBO) has a high light yield and short decay time and potentially provides a high detection efficiency and high count rate capability for neutrons or X-rays.[15] The light yields of single crystals under X-ray and thermal neutron irradiation are 14,000 photons/MeV and 40,000 photons/neutron, respectively.[12,16]

In this work, we have investigated a series of samples with different sintering temperatures and doping concentrations in oxides by high-temperature solid-state reactions. Micro-sized phosphors with a high photoluminescence quantum yield (PLQY) and short decay time have been prepared through optimizing sintering conditions and doping concentrations. Luminescence and thermal quenching mechanism of oxide powders are investigated in detail. Furthermore, the obtained powders (np = 1.65) were mixed with PMMA (nm = 1.49) in dichloromethane to obtain films with different phosphor loading fractions. A comprehensive study of the scintillation properties under X-ray, γ-ray, and neutron irradiation was performed. The results indicate that the transparent microcomposite scintillator under X-ray excitation emits intense blue light with wavelengths in the sensitive range of photomultiplier tubes (PMTs); the best of them has a light yield of 21,000 photons/thermal neutron under thermal neutron irradiation. The pulse shape discrimination of neutron/γ is implemented by digital signal processing methods.[13,14] The present work facilitates the development of excellent microcomposite scintillators for the urgent applications in the fields such as neutron/X-ray scattering science, nondestructive industrial tests, medical imaging, and high-energy physics research.

Experimental Section

Preparation of Ce-Doped LGBO Microphosphors

The raw materials were Gd2O3, Li2CO3, H3BO3, and CeF3, all of which have a purity of 99.99%. The raw powdered materials were mixed and ground for 2 h, then were put into an alumina crucible for presintering in a muffle furnace, and were sintered at 400 °C for 8 h. Subsequently, a series of samples were resintered in a muffle furnace at 600–800 °C and sintered again in air for 8 h to optimize the sintering temperature. Another series of samples with different doping concentrations were resintered in a tube furnace at 750 °C and sintered again for 8 h with a reducing gas (8 vol % H2/92 vol % Ar2).

Preparation of Ce-Doped LGBO/PMMA Composite Scintillators

The 1 mol % Ce-doped LGBO microphosphor (MP) sintered at 750 °C under a reducing atmosphere was used to prepare composite scintillators. A mixture of 0–0.2 g of the MP and 0.2 g of PMMA was ground for 10 min and transferred to a beaker, and 2 mL of dichloromethane was added. The resulting mixture was stirred in a sonicator for 3–5 min and then poured into circular holes (diameter = 25 mm, depth = 2 mm) of the grinder, and the transparent composite scintillator films with thicknesses of 0.30–0.35 mm were obtained after overnight solidification.

Characterization

The crystal structure of the prepared Ce-doped LGBO powders was studied using an X-ray diffractometer (Rigaku, Miniflex 600). The microstructures of the obtained powder and composite scintillators were studied using a scanning electron microscope (Tescan Brno, Vega3). Photoluminescence (PL) and radioluminescence (RL) spectra were measured using a home-built X-ray emission spectra test system shown in Figure S1. Temperature-dependent emission spectra and PLQY of MPs and composite scintillators were measured using a fluorescence spectrophotometer (Zolix, OmniFluo 900) with an X-ray tube, a heating apparatus, and an integrating sphere. The PLQY was calculated by the following formulawhere Lsam, Eref, and Esam represent the integral intensity of sample emission, reference (BaSO4) scattering, and sample scattering, respectively. The PL decay time was detected using a microscopic ultrafast spectroscopic measurement system (Light Conversion HARPIA). The transmission spectra of the composite scintillators were measured using a UV–vis–NIR absorption spectrometer (Hitachi UH4150).

Thermal Neutron and Gamma Ray Scintillation Measurements

The pulse-height spectra were measured using a PMT (ET Enterprises, 9124B), a preamplifier (ORTEC, 9305), a shaping amplifier (ORTEC, 572A, 3 μs shaping time), and a multichannel analyzer (MCA, ORTEC, Eesy-MCA-8k). The scintillation decay curves were obtained using an oscilloscope (Agilent, MSO9254A). The measurement details are shown in Figure S2. γ-ray sources were 60Co (average energy = 1.2 MeV) and 137Cs (average energy = 0.662 MeV). A neutron source of 252Cf (9 μg, 1.8 × 108 Bq) was moderated by polyethylene (thickness = 500 mm) to obtain thermal neutrons with an energy-obeying Maxwell distribution (average energy = 25 meV, emission rate ≈ 2000 n/s). The relative light yields under thermal neutron irradiation were calculated using the following equationwhere LY and CN represent the light yield and full energy peak channel number, respectively. Considering that the photons emitted by GS20 and the Ce-doped LGBO/PMMA composite scintillator both originated from the Ce luminescence center and the emission band wavelengths were basically identical (Figure S3), the sensitivity correction factor of PMT was neglected. The neutron−gamma discrimination is performed by our previously proposed pulse-shape algorithms, which has been applied to Gd3Al2Ga3O12:Ce (GAGG) scintillators. A detailed explanation can be found in refs (13) and (14).

Results and Discussion

Crystal Structure and Microscopic Morphology

The cell structure of LGBO is shown in Figure a. LGBO is monoclinic with the space group P21/c. There are three cation positions in the cell, that is, B3+, Li+, and Gd3+; Ce3+ enters the lattice by replacing Gd3+ which has a similar radius. The X-ray diffraction (XRD) pattern at different sintering temperatures is shown in Figure b. When the sintering temperature is 600 °C, the monoclinic phase of LGBO has been formed. As the sintering temperature increases, the crystallinity and grain size increase, resulting in a gradual increase in the intensity of the diffraction peaks. The intensity of the diffraction peak reaches its maximum at a sintering temperature of 750 °C. As the temperature continues to increase, the intensity of the diffraction peaks decreases, which may be caused by a decrease in crystallinity as the temperature approaches the melting point of LGBO (832 °C, Figure S4). As shown in Figure c, all diffraction peaks are consistent with the standard card, indicating that Ce3+ has successfully entered the lattice of the host without forming other crystalline phases. The strongest diffraction peaks are enlarged as shown in Figure d; with increasing doping concentration, the diffraction peaks gradually shift toward a smaller angle. According to Bragg’s law (2d sin θ = nλ), a larger lattice parameter leads to a smaller angle of the diffraction peak, which indicates that the small-radius Gd3+ is replaced by the large-radius Ce3+.[17]

Figure 1

Phases and crystal structures of Ce-doped LGBO. (a) Schematic diagram of the cell structure. (b) XRD patterns of powders with different sintering temperatures (600–800 °C). (c) XRD patterns of powders with different doping concentrations. (d) Enlarged diffraction peaks of (11–2). (e) XRD pattern refinement of 1 mol % CeF3-doped LGBO.

To further investigate the crystalline phase structure of the sample, the XRD pattern of the 1 mol % Ce-doped sample sintered at 750 °C was refined using Fullprof software.[18] As shown in Figure e, the diffraction pattern of the sample was well fitted. The refinement yielded cell parameters of a = 7.2237 Å, b = 16.5041 Å, c = 6.6901 Å, and β = 105.362°, which are similar to those in the previous reports.[19]

Figure a shows photographs of xCe-doped LGBO/yPMMA (x = 0, 0.01, 0.02, 0.03, 0.10, 0.20 g, y = 0.20 g) composite scintillator films under visible light and X-ray irradiation. The luminescence intensity gradually increases with increasing loading, but the transparency decreases. To observe the microscopic morphology of the Ce-doped LGBO MPs and the composite scintillator, scanning electron microscopy (SEM) photographs were taken. As shown in Figure b,c, the average size of the micropowder obtained by the high-temperature solid-phase reaction was ∼45 μm (the particle size distribution is shown in Figure S5). An uneven particle size was found, probably due to the uneven distributions of temperature and gas during the sintering process. The microscopic morphology of the Ce-doped LGBO/PMMA composite scintillator is shown in Figure d,e. The LGBO luminescent micropowder particles are uniformly distributed in PMMA by mixing with an organic solvent.

Figure 2

Morphology of the Ce-doped LGBO powder with composite scintillator films. (a) Photographs of scintillator films with different loading fractions (0–0.20 g) under X-ray irradiation. (b,c) SEM photographs of micrometer phosphors. (d,e) SEM photographs of the 10 wt % Ce-doped LGBO/PMMA composite scintillator.

Luminescence Properties

Emission spectra under X-ray excitation and steady-state and transient fluorescence spectra were used to analyze the luminescence properties of Ce-doped LGBO phosphors. The emission spectra of 1 mol % Ce-doped LGBO phosphors with different sintering temperatures under X-ray excitation are shown in Figure a. The four emission peaks are located at 313, 421, 544, and 626 nm, respectively. The 313 nm emission peak is attributed to the transition of the Gd3+ ion 6P → 8S7/2. Similar emission peaks of Gd3+ ions are also observed for Gd-containing host materials.[20] The broad emission peak at 350–550 nm originates from the radiative transitions of the excited state of Ce3+, which can be produced by the absorption of holes by Ce3+ or electrons by Ce4+. The corresponding processes are described as follows.[20]

Figure 3

Luminescence properties of Ce-doped LGBO phosphors. (a) Emission spectra of phosphors with different sintering temperatures for X-ray excitation. (b) Emission spectra of phosphors with different CeF3 doping concentrations for X-ray excitation. (c) Excitation and emission spectra; (d) 3D spectrum; and (e) quantum efficiency measurement spectra of 1 mol % Ce-doped LGBO phosphors. (f) PL decay time for different CeF3 doping concentrations.

Meanwhile, emission peaks at 544, 579, and 626 nm were also observed, which may be attributed to impurity rare-earth ions, the exact cause of which is still under study. With the increase in sintering temperature, the intensity of the Ce3+/Ce4+ ion emission peak gradually increases and then decreases, and the luminescence intensity reaches a maximum when the sintering temperature is 750 °C. This regularity is consistent with the variation in the XRD peak intensity at different sintering temperatures (Figure b), indicating that the luminescence intensity is influenced by the crystallinity and grain size.[21] With the fixed amount of the doping concentration, the more the crystallization, the brighter the emission will be. The emission spectra of Ce-doped LGBO with different doping concentrations are shown in Figure b, and the same four emission peaks consistent with Figure a are observed. With the increase in doping concentration, the intensity of the emission peak reaches its maximum at a doping concentration of 1 mol %. According to the Forster–Dexter theory, an excessive doping concentration will lead to the intensification of energy transfer between ions, which in turn results in concentration quenching.[22,23]

Figure c shows the excitation and emission spectra of 1 mol % Ce-doped LGBO. Three peaks exist in the excitation spectrum, located at 274, 310, and 347 nm, which are consistent with the excitation spectra of Ce-doped LGBO single crystals reported in the literature.[24] The excitation peak at 274 nm corresponds to the 8S7/2 → 6I transition of the Gd3+ ion and those at 310 and 347 nm correspond to the 4f → 5d2 and 4f → 5d1 transitions of the Ce3+ ion, respectively. The 350–550 nm broad emission peaks were obtained under UV excitation at 347 nm in general accordance with the X-ray excitation. The three-dimensional (3D) fluorescence spectrum corresponding to excitation and emission is shown in Figure d.

To investigate the luminescence performance of Ce-doped LGBO, the PLQY test was performed, as shown in Figure e. The calculated PLQY was 62.66%. This indicates that the prepared micrometer-sized Ce-doped LGBO has a high PL quantum efficiency. The light yield is calculated according to the following equation[25]where Y is the scintillator light yield, E is the energy absorbed by the scintillator, β is a parameter indicating the average energy required to produce a thermalized electron–hole pair, Egap is the band gap energy value of the host material, S is the energy transfer efficiency of the thermalized electron–hole pair to the excited state of the luminescence center, and Q is the PLQY of the luminescence center. Based on eq , a high PLQY may result in a high scintillator light yield. The lifetimes of Ce3+ in different host materials are variable due to the influence of the host material crystal field and electron–phonon interactions.

Different applications require different decay times for luminescent materials. For lighting, pc-LEDs are usually required for a long decay time (microsecond to millisecond). While for scintillators, short decay times (1–100 ns) are typically required in order to facilitate the collection of photon signals when the incident flux of high-energy particles is high.[26]Figure f shows the doping concentration-dependent PL decay time under UV-pulsed excitation at 347 nm. The decay curves were fitted using the single exponential formula[27]where I(t) is the time-dependent PL intensity, A is the fitting parameter, and τ is the PL decay time. The fitted luminescence lifetimes were obtained in the range of 25.29–29.49 ns (detailed fitting is shown in Figure S6), and the results indicate that the PL of Ce3+ in LGBO has a fast decay.

PL Thermal Quenching, PL, and Scintillation Mechanisms

The thermal stability of the emission centers will determine the scintillation efficiency and light yield under high-temperature conditions.[28] The increase in temperature intensifies electron–phonon coupling and nonradiation transition processes, leading to the thermal quenching of luminescence. In order to investigate the thermal quenching phenomenon of Ce-doped LGBO in detail, the emission spectra at different temperatures were measured. As shown in Figure a, the luminescence intensity gradually decreases as the temperature increases from 300 to 500 K. To further analyze the variation in the luminescence intensity with temperature, the curve of the integrated intensity with temperature was obtained by integrating the emission peaks at different temperatures. According to the current research on the thermal quenching of Ce3+ in the literature, the thermal quenching mechanism of Ce3+ is mainly in the form of crossover relaxation and thermal ionization,[29,30] both of which can be fitted using the following equation[31]where I0 is the integrated intensity corresponding to 0 K, I(T) is the integrated intensity at temperature T, A is a constant related to the matrix, kB is the Boltzmann constant, and ΔE is the thermal quenching activation energy (or thermal ionization energy). Here, we use ΔE to measure the thermal stability of the phosphor. The thermal quenching activation energy of 1 mol % Ce-doped LGBO is 380 meV, as obtained by the fitting, which is very similar to the results of Ogorodnikov et al.[32] The large ΔE indicates its excellent stability. The measurement results of thermoluminescence (TL) spectroscopy showed that no significant emission peak was observed (Figure S7), and considering the large Stokes shift (ΔS = 4780 cm–1), it is believed that the thermal quenching phenomenon is more possibly caused by crossover relaxation.[33]

Figure 4

Mechanistic diagram of variable-temperature PL with scintillation and PL processes: (a) emission spectra of 347 nm excitation at different temperatures. (b) Integrated intensity of the emission spectra at different temperatures and its fitting. (c) Schematic diagram of the PL and scintillation processes.

From the previous results and analysis, combined with the reports in the literature,[33] the model of the Ce3+ luminescence process in the LGBO host material is established, as shown in Figure c. The left and right sides of Figure c correspond to the scintillation and fluorescence processes, respectively. For the scintillation process, there are four main steps:[2] absorption of ionizing radiation by the scintillator to generate primary excitation (the photoelectric effect and Compton scattering); production and thermalization of secondary electron–hole pairs (the Auger process and electron inelastic scattering); energy transfer to the luminescence center; and visible light emission. As for the fluorescence process, the host material is usually not involved, and the luminescent centers absorb and emit visible photons directly.

X-ray Scintillation Performance and the Transmittance of Composite Scintillators

The emission spectra of the Ce-doped LGBO/PMMA composite scintillator with different loading fractions under X-ray excitation are shown in Figure a. The luminescence intensity of the composite scintillator increases with the loading fraction. As shown in the inset of Figure a, when the phosphor loading fraction exceeds 15 wt %, the RL integrated intensity of the composite scintillator will be superior to that of the commercial CsI:Na. Figure b shows the corresponding transmittance curves. Pure PMMA has a transmittance of about 80% in the 300–800 nm band. The transmittance of the 10 wt % Ce-doped LGBO/PMMA composite scintillator is close to 40%. Multiple repetitions of samples prepared with different loading fractions showed that the uniformity is poor when the loading amount is 5 wt %, and the transparency is always lower than that of the samples with 10 wt % loading.

Figure 5

X-ray scintillation properties and transmission spectra of composite scintillators. (a) X-ray emission spectra of scintillators with different loading fractions. (b) Transmission spectra. (c,d) Integrated emission intensity of the Ce-doped LGBO phosphor, composite scintillator, and commercial CsI:Na scintillator at different X-ray tube currents and tube voltage.

To further investigate the X-ray scintillation performance of the Ce-doped LGBO phosphor and composite scintillator, the luminescence intensity versus X-ray tube current and voltage curves were obtained (Figure c,d). The 10 wt % Ce-doped LGBO/PMMA composite scintillator was detected to have lower luminescence intensity than the commercial CsI:Na powder due to the comparatively low phosphor loading. A comparison between the phosphors shows that the luminescence intensity of the 1 mol % Ce-doped LGBO phosphor under X-ray excitation is higher than that of the commercial CsI:Na powder. The integrated luminescence intensity has an essentially linear increase with increasing X-ray tube current and voltage. This indicates that the Ce-doped LGBO/PMMA composite scintillator has potential applications in the radiation detection of X-rays.

Light Yield

In order to investigate the light yield of composite scintillators in detail, the pulse height spectra of samples with different loading fractions irradiated by different radioactive sources were measured. The pulse distribution of the composite scintillator with different loading fractions under the irradiation of a 252Cf thermal neutron source is shown in Figure a. The light yield of the Ce-doped LGBO/PMMA composite scintillator is mainly determined by the transparency and the powder loading fraction. As the content of the Ce-doped LGBO phosphors increases, the cut-off channel number of the pulse height spectra gradually increases, which indicates more pulsed photons of a high amplitude and higher light yield. However, when the loading fraction is 100%, the cut-off channel number of the samples decreases, which is due to a drastic decrease in transparency, and most of the photons are lost by light scattering. The cut-off channel number of the 50 wt % sample can reach 5000, and the channel number of the full energy peak is about 4000.

Figure 6

Pulse height spectra of composite scintillators irradiated with different radioactive sources. (a) Pulse height spectra of composite scintillators with different Ce-doped LGBO MP loading fractions under thermal neutron irradiation. (b) Pulse height spectra of the 1 mol % Ce-doped LGBO (50 wt %)/PMMA composite scintillator compared with that of commercial scintillation glass GS20. (c) Light yield of our composite scintillator compared with that of GS20 and a typical commercial scintillation screen EJ-426 HD. (d,e) Pulse height spectra of composite scintillators with different loading fractions under two types of γ-ray irradiation, 137Cs and 60Co, respectively.

The other samples failed to observe a significant full energy peak, which may be due to the loss of high-amplitude pulsed photons due to light scattering. To obtain the relative optical yield of the composite scintillator, the pulse height spectrum of the 50 wt % loading fraction sample was compared with that of the commercial scintillator 6Li-glass GS20 (Figure b). Based on the GS20 light yield (6000 photons/thermal neutron),[12] a relative light yield of ∼21,000 photons/thermal neutron was obtained for the sample with a 50 wt % loading fraction. In addition, the pulse height spectra of samples with different loading fractions were tested under irradiation with 60Co and 137Cs γ-ray sources (Figure d,e). The γ-ray light yield varies with the loading fraction in the same pattern as thermal neutrons. Compared to thermal neutrons, the cut-off channel numbers of pulse height spectra of composite scintillators with different loading fractions under γ-ray irradiation are less than 2000.

Scintillation Decay Time and n–γ Discrimination

In applications involving high count rates of incident particles or time-of-flight measurements, the timing characteristics of the scintillator become essential. Scintillators with a shorter decay time will allow detectors to have a higher count rare capability or higher time resolution. To investigate the scintillation decay time components of the LGBO/PMMA composite scintillator, the decay times under irradiation from different radioactive sources were collected (Figure a). All decay variations can be well fitted by the triple exponential function[34]where I(t) is the time-dependent scintillation light intensity; A (i = 1, 2, 3) is the fitting parameter; and τ (i = 1, 2, 3) is the decay time of the fast, intermediate, and slow components, respectively. The decay time components obtained by the fitting are shown in Table . The decay time of all three types components can be corresponded to the data reported previously in the literature studies.[34−36] The fast decay components of 3.60–3.76 ns and the intermediate decay components of 56.22–95.10 ns are attributed to the intrinsic scintillation decay of Ce3+ ions, while the slow decay components of 257.80–594.44 ns are attributed to the energy transfer of Gd3+ → Ce3+.[36]

Figure 7

(a) Scintillation decay under irradiation from different radioactive sources and the corresponding fitted curves. (b) Neutron−gamma discrimination for the 50 wt % LGBO phosphor-loaded composite scintillator (S1–S2 scatter plot).

Table 1

Scintillation Decay Time Components under the Excitation of Different Radioactive Sources for the 50 wt % LGBO Phosphor-Loaded Composite Scintillator

radiation sources	τ1 (ns) Ce3+	τ2 (ns) Ce3+	τ3 (ns) Gd3+ → Ce3+
60Co	3.69 (18%)	95.10 (32%)	594.44 (50%)
137Cs	3.76 (6%)	56.22 (29%)	257.80 (65%)
252Cf	3.60 (7%)	88.19 (25%)	515.93 (68%)

For efficient neutron detection, an excellent neutron−gamma discrimination (NGD) ratio (or neutron−gamma efficiency ratio) is required for a practical application such as neutron scattering, security, or nondestructive testing in industry. From the fitting results in Table , the decay time components of 252Cf neutrons and 60Co γ-rays are different. The neutron−gamma discrimination using pulse shape difference is performed via a conventional two-feature S1–S2 plot,[13] as shown in Figure b. Here, S1 and S2 are the integration values of individual pulses in the time windows of [−50, 0] ns and [0, 50] ns, respectively. A two-dimensional scatter plot of S1–S2 can separate most of the neutron from the γ signal using the blue solid line as the class boundary. The boundary is set to obtain both a neutron efficiency and an NGD ratio as high as possible. Events above and below this line are assigned to neutrons and γ-rays, respectively. However, the conventional S2–S1 method can only provide a neutron detection efficiency of about 1.4% and an NGD ratio of about 102, which are too low to have any practical applications.

To improve both the neutron detection efficiency and NGD ratio, we applied a graph-embedded non-negative matrix factorization (GNMF) method for particle discrimination.[14] In this method, an event graph (or network) is established to describe the event similarity using reference events selected from the data measured under 252Cf and 60Co sources and background. A batch of unlabeled events is regarded as graph signals, whose low-dimensional representation is searched under the assumption of smoothness in the high-dimensional-event graph. More description can be found in ref (14). Using the GNMF data model, we found that P2 versus V2 can be used to separate neutron events from γ and background events, where V2 and P2 are the second components of the signal matrix and projected matrix = and is the Laplacian operator of the graph. As shown in Figure , a boundary composed of three lines can separate neutron and γ. There is an almost clean separation between neutron and background events. A thermal neutron detection efficiency of 55% (relative to a GS20 detector) and an NGD ratio of about 1.3 × 105 are obtained, which are much better than those obtained by the traditional S2–S1 method and are good enough for practical applications. Further improvement in the NGD and neutron detection efficiency may reside in the following: (i) the improvement of the average fitting residual value (4.4 × 10–2) of R2, which is slightly larger than that for a bright GAGG scintillator.[14] (ii) The optimization of the film thickness, component fractions, and LGBO particle size.

Figure 8

Neutron−gamma discrimination using the GNMF method for the 50 wt % LGBO phosphor-loaded composite scintillator.

Conclusions

In conclusion, we prepared Ce-doped LGBO MPs with a high stability and high PL quantum efficiency by a facile high-temperature solid-state reaction. These MPs exhibit high luminescence intensity under X-ray and UV excitation and a decay time below 600 ns. We have discussed the PL and scintillation processes in detail and have investigated the mechanism of the luminescence thermal quench phenomenon of Ce3+ in the LGBO system. Furthermore, the transparent scintillator films have been fabricated by mixing LGBO MPs with PMMA. A comparison with commercial scintillators CsI:Na and GS20 shows that our scintillator films exhibit high light yields under irradiation with X-rays and thermal neutrons. Due to the large thermal neutron absorption cross-section of 157Gd and the easy solid-state sintering route to obtain MPs, fabricated Ce-doped LGBO/PMMA scintillators would effectively reduce the production cost and time. Our results suggest that Ce-doped LGBO/PMMA is a low-cost, high-performance scintillator for X-ray and thermal neutron detection.