Crystal Growth and Energy Transfer Study in Ce³⁺ and Pr³⁺ Co-Doped Lu₂Si₂O₇

Abstract

Ce-doped Lu₂Si₂O₇ has a high density, high luminescence efficiency even at high temperatures, and a high effective atomic number, making it a promising candidate for use as a radiation detector in medical devices and resource exploration equipment. In this study, we grow and characterize Pr³⁺ and Ce³⁺-doped Lu₂Si₂O₇ single crystals by systematically varying the Ce³⁺ to Pr³⁺ ratio to further improve scintillation properties. The optical characterization results show a bidirectional energy transfer: from the Pr³⁺ 5d levels to the Ce³⁺ 5d levels and from the Ce³⁺ 5d levels to the Pr³⁺ 4f levels. Consistently with this result, the PL decay time of emission from the Pr³⁺ 5d–4f transition tends to become faster as the Ce³⁺/Pr³⁺ ratio increases, due to the energy transfer from the Pr³⁺ 5d levels to the Ce³⁺ 5d levels. Additionally, (Ce_0.0022 Pr_0.0016 Lu_0.9962)₂Si₂O₇ exhibits a high light yield comparable to Ce-doped Lu₂Si₂O₇ and a slightly faster decay time than Ce-doped Lu₂Si₂O₇.

1. Introduction

Scintillators have numerous applications, including medical imaging, homeland security, radiation dosimetry, geological exploration, high-energy physics, nuclear power plants, and space exploration. Among these applications, the improvement of the characteristics of scintillators is particularly crucial in medical imaging, as the radiation exposure can have significant effects on the human body. The typical scintillation properties are density, effective atomic number, light yield (LY), decay time, and energy resolution [1]. In the pursuit of an ideal scintillator that meets these criteria, different kinds of host materials have been developed, including Lu₂SiO₅, Gd₂SiO₅, and Bi₄Ge₃O₁₂ [1,2,3]. In particular, Lu₂Si₂O₇ (LPS) is one of the promising inorganic scintillator host materials due to its high density (6.23 g/cm³) and high effective atomic number (Z_eff = 64) [4]. As a luminescent center, the Ce³⁺ 5d–4f transition, being a parity- and spin-allowed transition, is anticipated to exhibit high luminescence efficiency and a fast decay time. Consequently, Ce³⁺-doped Lu₂Si₂O₇ (Ce:LPS) scintillators exhibit a relatively high LY up to 26,300 photons/MeV and a fast decay time of 38 ns. In addition, through their study, Pr³⁺-doped scintillators have been reported to exhibit a faster decay time than Ce³⁺ [1]. In Pr:LPS, the scintillation and optical properties have been investigated, demonstrating a faster decay time (~26 ns) compared to Ce:LPS [5,6,7]. Furthermore, recent studies have found that Pr:LPS shows the highest scintillation LY (9700 photons/MeV) among Pr:Y₂Si₂O₇ and Pr:Gd₂Si₂O₇ [7,8,9]. Other LPS scintillation materials doped with various luminescent centers, such as Tm:LPS and Nd:LPS, have also been developed [10,11]. However, these scintillation LYs remain relatively low at 2340 photons/MeV and 800 photons/MeV, respectively.

For further improvement of scintillation properties, e.g., afterglow, decay time, and LY, the co-doping method has been widely attempted. As an example of afterglow improvement, Bi³⁺ co-doped Tl:CsI [12] has been shown to successfully reduce the afterglow to 0.1% without compromising the light yield. Additionally, Yb²⁺ co-doped Tl:CsI [13] has been found to achieve a high LY of 90,000 photons/MeV, a good energy resolution of 7.9% (511 keV, FWHW), and a significantly reduced afterglow of 0.035%. On the other hand, Mg²⁺ co-doped Ce:Gd₃Al₅O₁₂ [14] is an example where the decay time is improved, with an increase in the fast component and an acceleration of the decay time from 90 ns to 40 ns. In addition, Na⁺ co-doped Ce:LiCaAlF₆ [15] has been reported as an improvement of LY by increasing the Ce³⁺ concentration and reducing defects in the crystal. Recent studies have shown a rising interest in improving scintillation properties by co-doping with Ce³⁺ and Pr³⁺. In fact, Pr³⁺ co-doped Ce:Lu_0.8Sc_0.2BO₃ [16] has been reported to exhibit a high scintillation efficiency, while Pr³⁺ co-doped Ce:Li₆Y(BO₃)₃ [17] shows promise as a fast scintillator for thermal neutron detection. In the case of Ce:LPS, Feng et al. examined a Pr³⁺ co-doped Ce:LPS scintillator with Pr³⁺ while keeping the concentrations of Ce³⁺ and Pr³⁺ fixed at 0.05 mol%, and reported that the Pr³⁺ co-doping enhances the luminescence intensity [18]. Therefore, in this study, we concentrate on optimizing the Ce³⁺/Pr³⁺ ratio to improve the scintillation properties. We grow and characterize Pr³⁺ co-doped Ce:LPS single crystals, systematically varying the Ce³⁺/Pr³⁺ ratios.

2. Materials and Methods

Powders of Lu₂O₃, CeO₂, and SiO₂ with purities of at least 99.99% and a Pr₂O₃ powder with a purity of 99.9% were used for crystal growth. After weighing these powders with a composition of (Ce_x Pr_y Lu_1−x−y)₂Si₂O₇ ((x, y) = (0.005, 0.000), (0.002, 0.005), (0.005, 0.005), (0.010, 0.005), (0.000, 0.005)), the powders were wet mixed with an agate mortar and ethanol. Subsequently, they were placed into the iridium crucible with a ϕ3 mm die and one capillary. The micro-pulling-down (μ-PD) method [19], employing a μ-PD furnace equipped with a radiofrequency induction heating system and a vacuum-tight chamber, was used for crystal growth. The LPS was utilized as the seed crystal and was pulled down at a rate of 0.03 mm/min in an argon and oxygen mixed gas flow. Following the crystal growth, the furnace output was slowly decreased to room temperature over the course of 5 h.

In the grown crystals, the crystal structures and actual compositions were determined by the powder X-ray diffraction (XRD) technique (D8 DISCOVER, Bruker, Yokohama, Japan, Cu Kα radiation, λ = 1.5418 Å) and the electron probe microanalyzer (EPMA, S–3400N, Hitachi, Tokyo, Japan), respectively. After that, we measured the absorption spectra with a spectrophotometer (V–550, JASCO, Tokyo, Japan). The photoluminescence (PL) excitation, emission spectra, and PL decay times were obtained by a spectrometer (FLS 1000, Edinburgh Instruments, Edinburgh, UK) with a Xe lamp as an excitation source and a nanosecond flash lamp, respectively. To evaluate the temperature dependence of the PL decay time, a vacuum liquid nitrogen cryostat (VPF–100–H, Lake Shore Cryotronics, Westerville, OH, USA) was employed, enabling temperature adjustments between 78 K and 790 K, in combination with the spectrometer (FLS1000, Edinburgh Instruments, Edinburgh, UK). Thereafter, we evaluated the light yield and the scintillation decay times at room temperature using a gamma-ray source (¹³⁷Cs, 662 keV). The light photons emitted by the scintillator were detected and converted by a photomultiplier tube (PMT, R7600U–200, Hamamatsu, Japan) into an electrical pulse. Following that, they were amplified by a preamplifier (113, ORTEC, Zoetermeer, The Netherlands), shaped by a shaper (572A, ORTEC), using a shaping time of 0.5 μs, and digitized with a multi-channel analyzer (MCA, 8000A, Amptec, Austin, TX, USA).

3. Results and Discussions

We successfully grew transparent crystals having a 3 mm diameter and lengths ranging from 15 to 50 mm (Figure 1). From the powder XRD results (see Figure 2), the diffraction patterns were found to correspond well with the reference pattern (Lu₂Si₂O₇, PDF 34–0509). Furthermore, the crystalline structure of all samples was considered to be monoclinic with a space group C2/m.

Table 1. Sample compositions obtained through EPMA analysis [atm%].

	Lu			Si			Ce			Pr
	C0	C	C/C0	C0	C	C/C0	C0	C	C/C0	C0	C	C/C0
Ce0.5	49.75	49.46	0.99	50.00	50.43	1.01	0.25	0.11	0.46	0.00	0.00	――
Ce0.2 Pr0.5	49.65	48.77	0.98	50.00	51.04	1.02	0.10	0.11	1.12	0.25	0.08	0.34
Ce0.5 Pr0.5	49.50	49.60	1.00	50.00	50.09	1.00	0.25	0.16	0.64	0.25	0.15	0.60
Ce1.0 Pr0.5	49.25	48.16	0.98	50.00	51.63	1.03	0.50	0.18	0.36	0.25	0.03	0.11
Pr0.5	49.75	48.15	0.97	50.00	51.73	1.03	0.00	0.00	――	0.25	0.12	0.48

(C₀: nominal concentration, C: actual concentration).

shows the composition of all samples evaluated by EPMA. The actual composition of the host material closely matched the nominal composition. On the other hand, the actual composition of Ce³⁺ and Pr³⁺ was lower than that of the nominal compositions. This is due to the difference in the ionic radius: the ionic radius of Ce³⁺ (1.034 Å) and Pr³⁺ (1.013 Å) at the six coordinates is larger than that of Lu³⁺ (0.848 Å), making it difficult for Ce³⁺ and Pr³⁺ to occupy the Lu³⁺ sites [20,21,22].

Table 2. The Ce³⁺/Pr³⁺ ratio for both the nominal and actual compositions of each sample and names based on these ratios.

	Ce3+/Pr3+		Name
	Nominal	Actual
Ce0.2 Pr0.5	0.40	1.32	Ce/Pr1.32
Ce0.5 Pr0.5	1.00	1.07	Ce/Pr1.07
Ce1.0 Pr0.5	1.00	6.80	Ce/Pr6.80

shows the Ce³⁺/Pr³⁺ ratio for both the nominal and actual compositions of each sample. The Ce_0.2 Pr_0.5 and Ce_0.5 Pr_0.5 samples show similar ratios. To facilitate the discussion of the Ce³⁺/Pr³⁺ ratio, the samples are described using the Ce³⁺ to Pr³⁺ ratio based on the actual composition.

Figure 3a shows the absorption spectra of (Ce_x Pr_y Lu_1−x−y)₂Si₂O₇, where absorption peaks associated with Ce³⁺ 4f–5d transitions were observed in the 265–370 nm range for all Ce³⁺ and Pr³⁺ co-doped samples. Moreover, in the samples doped with Pr³⁺, absorption peaks consistent with the Pr³⁺ 4f–5d transitions were detected in the 200–265 nm range, along with a minor peak attributed to the Pr³⁺ 4f–4f transition around 485 nm. Based on the absorption spectra, Figure 3b shows PL emission spectra for excitation at the Ce³⁺ 4f–5d₁ transition (349 nm). The emission spectra revealed not only the broad emission of the Ce³⁺ 5d–4f transitions, but also the sharp emissions originating from the Pr³⁺ 4f–4f transitions. From these observations, it can be concluded that the energy transfer from the Ce³⁺ 5d levels to the Pr³⁺ 4f levels occurs. The PL emission spectra, excited at the Pr³⁺ 4f–5d₁ transition (247 nm), is shown in Figure 3c. The emission peaks corresponding to the Pr³⁺ 5d–4f transitions, Ce³⁺ 5d–4f transitions, and Pr³⁺ 4f–4f transitions were observed in the wavelength range of 250–350 nm, 350–500 nm, and 545–670 nm, respectively. This suggests that the energy transfer from Pr³⁺ 5d levels to Ce³⁺ 5d levels is also taking place. The excitation spectra for Ce³⁺ 5d–4f emissions (383 nm) is shown in Figure 3d. The peaks attributed to both the Ce³⁺ 4f–5d transition and the Pr³⁺ 4f–5d transition were identified in the excitation spectra. The emission peaks’ wavelength derived from the Pr³⁺ 5d–4f transitions, shown in Figure 3c, overlaps with the excitation peak derived from the Ce³⁺ 4f–5d₂ transition. This suggests that both non-radiative and radiative energy transfers from the Pr³⁺ 5d levels to the Ce³⁺ 5d levels are expected to occur. Figure 4a shows the PL decay curves for the luminescence from the Ce³⁺ 5d–4f transition (λ_ex = 349 nm, λ_em = 383 nm). Each decay curve was fitted with a single exponential equation. The PL decay times are presented in Table 3, indicating that the co-doped samples exhibited slightly accelerated decay times compared to Ce:LPS. Since Ce:LPS has been reported to have no concentration quenching at Ce³⁺ concentrations below 1.3 atm%, the accelerated PL decay time is probably due to the energy transfer from Ce³⁺ 5d levels to Pr³⁺ 4f levels [23]. Figure 4b displays the PL decay curves for the luminescence from the Pr³⁺ 5d–4f transition (λ_ex = 247 nm, λ_em = 303 nm). The PL decay times of the co-doped samples were calculated using double exponential fitting, revealing that the PL decay times of the co-doped samples were faster compared to those of Pr:LPS, as shown in Table 3. In Pr:LPS, the PL decay time remained consistent across Pr³⁺ concentrations ranging from 0.1 to 2.0%, suggesting that the faster PL decay time is likely due to an energy transfer from the Pr³⁺ 5d levels to the Ce³⁺ 5d levels [7]. The PL decay time accelerated with increasing Ce³⁺/Pr³⁺ ratios, possibly due to the enhanced probability of energy transfer as the Ce³⁺ concentration relative to Pr³⁺ increases.

The PL decay curves of the Ce³⁺ 5d–4f transition and Pr³⁺ 5d–4f transition for the Ce/Pr_1.32 sample were measured in the temperature range of 78–790 K. Figure 5 (black square) depicts the temperature dependence of the PL decay time of the luminescence from the Ce³⁺ 5d–4f transition, and two major trends are observed. In the temperature range below 425 K, the PL decay time increased to 39 ns as the temperature increased. Then, around 450 K, the PL decay time rapidly decreased. These trends are consistent with those reported for Ce:LPS [4,23]. At low temperatures, radiative transitions are dominant, and the increase in the PL decay time can be attributed to the increasing influence of self-absorption with a rising temperature. On the other hand, at high temperatures, the decrease in the PL decay time is attributed to the thermal ionization from the Ce³⁺ 5d level to the conduction band [24]. Figure 5 (red circle) illustrates the temperature dependence of the PL decay time for the luminescence from the Pr³⁺ 5d–4f transition. The onset of the PL decay times shortening was observed around 450 K, a phenomenon which is also attributed to temperature quenching due to thermal ionization [25].

Figure 6 depicts the pulse–height spectra under excitation with ¹³⁷Cs (662 keV) as a gamma-ray source. The light yields were estimated by comparing the peak channel positions of the samples and those of the reference (Ce:Gd₂SiO₅, light yield: 10,000 photons/MeV), as shown in Table 3. Ce:LPS possessed a light yield of ~36,000 photons/MeV, which is significantly higher than that of Ce:LPS (~26,300 photons/MeV) reported in previous studies [4,23,26]. This is most probably influenced by the presence of several impurities and defect states (traps) [27]. As a study on electron and hole traps has also been reported [28], more detailed studies are needed to consider the effects of defects and other factors contributing to the high light yield observed in this study. Focusing on the light yield of Pr³⁺ co-doped Ce:LPS, the Ce/Pr_1.32 sample had a high light yield, comparable to that of the conventional Ce:LPS (~26,300 photons/MeV). We also evaluated the scintillation decay time excited by ¹³⁷Cs (662 keV) as a gamma-ray source, see Figure 7. The decay curve fittings were performed using a single exponential equation. The co-doping samples exhibited slightly faster decay times compared to Ce:LPS (Table 3). These results are in agreement with the PL decay results for the Ce³⁺ 5d–4f transitions. It is considered that the high light yield and fast decay time of the Ce/Pr_1.32 sample was associated with the energy transfer from non-radiative and radiative energy transfers from the Pr³⁺ 5d levels to the Ce³⁺ 5d levels and with the energy transfer from Ce³⁺ 5d levels to Pr³⁺ 4f levels, respectively.

Table 3. PL decay time and scintillation parameters of (Ce_x Pr_y Lu_1−x−y)₂Si₂O₇ samples.

	PL Decay Time [ns]			Light Yield [photons/MeV]	Decay Time [ns]
	λex = 349 nm λem = 383 nm	λex = 247 nm λem = 303 nm
	τ	τ1	τ2
Ce	37.0	—	—	~36,000 ~26,300 [4,23,26]	39.8 ± 0.4
Ce/Pr1.07	32.0	2.4	13.2	~25,000	34.1 ± 0.2
Ce/Pr1.32	35.2	1.2	15.4	~28,000	36.9 ± 0.3
Ce/Pr6.80	35.0	4.6	17.4	~6600	37.7 ± 0.2
Pr	——	18.2	—	——	——

Table 3. PL decay time and scintillation parameters of (Ce_x Pr_y Lu_1−x−y)₂Si₂O₇ samples.

	PL Decay Time [ns]			Light Yield [photons/MeV]	Decay Time [ns]
	λex = 349 nm λem = 383 nm	λex = 247 nm λem = 303 nm
	τ	τ1	τ2
Ce	37.0	—	—	~36,000 ~26,300 [4,23,26]	39.8 ± 0.4
Ce/Pr1.07	32.0	2.4	13.2	~25,000	34.1 ± 0.2
Ce/Pr1.32	35.2	1.2	15.4	~28,000	36.9 ± 0.3
Ce/Pr6.80	35.0	4.6	17.4	~6600	37.7 ± 0.2
Pr	——	18.2	—	——	——

4. Conclusions

Pr³⁺ co-doped Ce:LPS single crystals were grown using the micro-pulling-down method, with the Ce³⁺ to Pr³⁺ ratio systematically varied. The PL emission spectra demonstrate emissions not only from the Pr³⁺ 5d–4f and 4f–4f transitions, but also from the Ce³⁺ 5d–4f transitions under excitation at 247 nm (Pr³⁺ 4f–5d₁ transitions). In addition, under excitation at 349 nm (Ce³⁺ 4f–5d₁ transitions), emissions from both the Ce³⁺ 5d–4f transitions and the Pr³⁺ 4f–4f transitions were observed. Therefore, it is revealed that bidirectional energy transfers are occurring from the Pr³⁺ 5d levels to the Ce³⁺ 5d levels and from the Ce³⁺ 5d levels to the Pr³⁺ 4f levels. Moreover, the PL decay time of the Pr³⁺ 5d–4f transition became faster with an increasing Ce/Pr ratio due to the energy transfer from the Pr³⁺ 5d levels to the Ce³⁺ 5d levels. The scintillation characterization results revealed that the (Ce_0.0022 Pr_0.0016 Lu_0.9962)₂Si₂O₇ had a high light yield comparable to that of Ce:LPS, along with a slightly faster decay time compared to Ce:LPS.