Synthesis and Radiation Response Properties of Tm-activated Na 2 O–ZnO–TeO 2 –B 2 O 3 Glasses

Na


Introduction
(3) They have been used in many fields, for example, medical imaging, (4) security inspection, (5) environmental measurement, (6) and resource exploration. (7)The following properties are generally important for scintillators: high light yield (LY), fast decay, low afterglow, and mechanical and chemical stabilities.There are no scintillators satisfying all the above properties.However, the required properties vary in applications; therefore, suitable scintillators are chosen on the basis of their physical and chemical properties.(20)(21)(22) Glasses are an attractive form owing to their low cost, ease of formability, high mechanical strength, and high freedom in composition selection.In particular, Ce-doped lithium silicate glasses have been intensively studied for thermal neutron detection, (19,(23)(24)(25) and GS20, which shows the LY of 6000 photons/ neutron (26) with no hygroscopicity and low density, is a commercial glass scintillator.For thermal neutron detection, a 3 He gas counter is mainly applied by using 3 He(n, p) 3 H reactions; (27) however, alternative detectors are extensively studied because of the limitation of 3 He resources. (28)orate glasses to be used for thermal neutron detection in addition to Li-containing glasses have also received attention because 10 B has a larger thermal neutron capture cross section (3840 barn) than 6 Li (940 barn). (29)Some compositions with borate have shown response to neutrons or α-rays; (30)(31)(32)(33) however, there are no commercial ones.In this study, Na 2 O-ZnO-TeO 2 -B 2 O 3 (NZTB) glasses were developed.The prepared glasses have almost the same effective atomic number (29) as GS20 ( 26), (34) and the relatively low value is an advantage for distinguishing signals of neutrons from those of γ-rays.TeO 2 improves thermal strength, chemical stability, and light output when combined with B 2 O 3 . (35,36)Borate glasses generally show high hygroscopicity; therefore, Na 2 O was added as a modifier. (37)Moreover, alkali-metal-oxide-containing ZnO-TeO 2 -B 2 O 3 glasses were discovered to show good transparency in visible ranges. (38,39)This optical characteristic is suitable for the host material for scintillators with luminescence centers.(45) The main peaks appear in the 300-500 nm range, (46)(47)(48)(49) which matches the high wavelength sensitivity regions of conventional photodetectors such as photomultiplier tubes and Si photodiodes.Here, we fabricated NZTB glasses doped with different Tm concentrations, and investigated their optical and scintillation properties.

Materials and Methods
25Na 2 O-20ZnO-5TeO 2 -50B 2 O 3 glasses doped with Tm (0, 0.1, 0.5, 1, and 2 mol%) were synthesized by the melt quenching method.First, Tm 2 O 3 (4N), Na 2 CO 3 (4N), ZnO (4N), TeO 2 (4N), and B 2 O 3 (5N) powders were homogeneously mixed with an agate mortar.Then, they were transferred to an alumina crucible and melted at 900 ℃ for 1 h.After that, the melt was flowed onto a preheated stainless-steel plate to quench.The obtained samples were annealed at 300 ℃ for 1 h to remove thermal and mechanical strains.The annealing temperature was determined by measuring the glass transition temperature (T g ) of the undoped sample with a TG-DTA system (Hitachi High-Tech Corporation, STA7200).

Results and Discussion
Figure 1 shows the photograph and densities of the prepared samples.The surfaces were polished for the following PL and scintillation measurements.All the samples appeared transparent and colorless under room light.The T g of the undoped sample was 500 ℃.The densities were changed in the range of 2.4-2.7 g/cm 3 with respect to the Tm concentration.As the dopant concentration increased, the densities increased owing to the molecular mass of Tm 2 O 3 being larger than that of the compounds composing the host.The values were comparable to that of GS20 (2.5 g/cm 3 ). (26)Figure 2 shows the XRD patterns of the samples.Some parts of the glasses not used in the PL and scintillation measurements were crushed into powders, and XRD measurements were conducted.All the samples showed only halo peaks; hence, the samples had no periodical structures and formed glass phases.
Figure 3 shows the diffuse transmission spectra of NZTB glasses.The transmittance was 80-95% in the 350-850 nm range.Absorption peaks due to the electronic transitions of Tm 3+ (52,53) were clearly observed at 350, 470, 680, and 790 nm in the 1 and 2% Tm-doped samples.Absorption edges were confirmed at 260-280 nm.They were slightly shifted to lowenergy regions as the dopant concentration increased.As the Tm concentration increased, the peaks were shifted to low-energy regions.Figure 4 shows the PL excitation and emission spectra of the undoped and 0.5% Tm-doped NZTB glasses.Although the undoped sample did not show any emissions, the Tm-doped sample showed several emissions at 450, 480, 650, and 750 nm under excitation at 360 nm.These emissions were respectively considered to be derived from the 1 D 2 -3 F 4 , 1 G 4 -3 H 6 , 1 G 4 -3 F 4 , and 1 D 2 -3 F 3 transitions of Tm 3+ . (54,55)The QYs of the 0.1, 0.5, 1, and 2% Tm-doped samples were respectively 4.3, 8.6, 7.2, and 3.5% when monitored at 400-800 nm upon excitation at 360 nm. Figure 5 shows the PL decay curves of the Tm-doped samples.The  obtained curves matched with an approximation by a single exponential decay function.The decay times of the 0.1, 0.5, 1, and 2% Tm-doped samples were obtained to be 14.8, 13.7, 12.6, and 10.0 μs, respectively.From the above decay times and QYs, radiative (k f ) and nonradiative (k nr ) transition rates were estimated and are shown in Table 1.In the estimation, the following equations were applied: k f = QY/τ and k nr = (1 − QY)/τ.Here, the PL decay time was denoted as τ.k f decreased as the Tm concentration increased from 0.5 to 1%.The tendency can be described as concentration quenching that occurred at 1% Tm doping.
Figure 6 shows the X-ray-induced scintillation spectra of the NZTB glasses.Sharp emission peaks were observed at 350, 360, 450, and 480 nm.Similar scintillation peaks were observed in other Tm-doped materials; (56,57) hence, they were considered to be attributed to the 4f-4f transitions of Tm 3+ .Figure 7 shows the X-ray-induced scintillation decay curves of the Tmdoped samples.All the curves were approximated by a single exponential function when the instrumental response function (IRF) was deconvoluted.)(59) Therefore, they originated from the 4f-4f transitions of Tm 3+ .When compared with the PL decay times, they became much longer because of the conversion and transportation processes of scintillation in addition to the direct excitation and emission processes. (60)igure 8 shows the pulse height spectrum of 241 Am α-rays (5.5 MeV) measured using the 1% Tm-doped NZTB glass.The spectrum of 137 Cs γ-rays (0.662 MeV) measured using a Ce-doped Gd 2 SiO 5 sample (GSO, 8000 photons/MeV) was also displayed as a reference for calculating the LY of a prepared sample.A full absorption peak appeared only in the 1% Tm-doped sample, whereas this peak was unclear in the spectra.The rest of the prepared samples did not show full absorption peaks owing to their low LYs.From Robbins' model, (61) LY is considered directly proportional to QY, but inversely proportion to bandgap energy (E g ).The QY of the 0.5% Tmdoped sample was higher than that of the 1% Tm-doped sample; however, the 0.5% sample did not show a full absorption peak.This would be due to the difference in E g : a high concentration of Tm doping would broaden the impurity bands and tails would reduce E g . (62)This tendency was confirmed in the diffuse transmission spectra shown in Fig. 3. Therefore, the LY of the 1% Tm-doped sample would become higher than that of the 0.5% Tm-doped sample and the full absorption peak would appear.By comparing the channel of a full absorption peak (230 ch) with a photoabsorption peak channel of the reference (23750 ch), which was taking the difference in gain into account, the LY of the 1% Tm-doped glass was estimated to be 51 photons/5.5MeV-α.

Conclusions
NZTB glasses doped with Tm (0.1, 0.5, 1, and 2%) were fabricated by a conventional melt quenching method, and their physical, PL, and scintillation properties were studied.The prepared samples had halo peaks and no sharp diffraction patterns.Their densities were 2.4-2.7 g/cm 3 , which were comparable to that of GS20.Under the irradiation of both ultraviolet light and X-rays, some emission peaks appeared in visible regions, and they originated from the 4f-4f transitions of Tm 3+ .Although the prepared samples could not detect thermal neutrons, they showed signals under 241 Am α-ray irradiation (5.5 MeV).The LY of the 1% Tm-doped sample was estimated to be 51 photons/5.5MeV-α.Therefore, the result revealed that the glass with a composition of NZTB had a potential for thermal neutron detection because neutrons were detected through the observation of the α-rays generated by neutron capture reactions.There is still huge room for research on NZTB glass scintillators because the molar ratio of host compositions, the optimum dopants, and their concentrations have not been investigated.By considering them, the QY, LY, and capability of neutron detection would be improved.

Fig. 4 .
Fig. 4. (Color online) PL excitation and emission spectra of undoped and 0.5% Tm-doped NZTB glasses.Color scales indicate the intensities, where white and black indicate high and low intensities, respectively.

Table 1
Summary of PL characteristics of Tm-doped NZTB glasses.