Preparation and Scintillating Properties of Sol-Gel Eu3+, Tb3+ Co-Doped Lu2O3 Nanopowders

Nanocrystalline Eu3+, Tb3+ co-doped Lu2O3 powders with a maximum size of 25.5 nm were prepared by the sol-gel process, using lutetium, europium and terbium nitrates as precursors, and ethanol as a solvent. Differential thermal analysis (DTA) and infrared spectroscopy (IR) were used to study the chemical changes during the xerogel annealing. After the sol evaporation at 100 °C, the formed gel was annealed from 300 to 900 °C for 30 min under a rich O2 atmosphere, and the yielded product was analyzed by X-ray diffraction (XRD) to characterize the microstructural behavior and confirm the crystalline structure. The results showed that Lu2O3 nanopowders start to crystallize at 400 °C and that the crystallite size increases along with the annealing temperature. A transmission electron microscopy (TEM) study of samples annealed at 700 and 900 °C was carried out in order to analyze the microstructure, as well as the size, of crystallites. Finally, in regard to scintillating properties, Eu3+ dopant (5 mol%), Tb3+ codoped Lu2O3 exhibited a typical red emission at 611 nm (D°→7F2), furthermore, the effect of Tb3+ molar content (0.01, 0.015 and 0.02% mol) on the Eu3+ radioluminiscence was analyzed and it was found that the higher emission intensity corresponds to the lower Tb3+ content.


Introduction
Since Lu 2 O 3 :Eu 3+ first attracted attention as a potential X-ray phosphor [1], many efforts have been conducted in the last few decades to process it, due to a growing need for new materials to be employed in high-resolution X-ray imaging systems, including high-definition X-ray radiographers, positron emission tomography (PET) scanners as well as many others industrial measuring systems [2][3][4][5]. What makes lutetia attractive for such applications is its high efficiency in absorbing any kind of ionizing radiation [6]. Indeed, the combination of its high density (9.42 gcm −3 ) and its high atomic number of Lu (Z = 71) [7], along with its good absorption in the diagnostic medical energy range (15-150 keV) makes it a very interesting materials for the mentioned applications [8]. Furthermore, it is a convenient host lattice for activators forming scintillating materials [9], especially for Eu 3+ and Tb 3+ , due to the fact that its band gap is large enough to accommodate the energy levels of the ions [10]. On the other hand, lutetia systems doped with Eu 3+ are expected to replace the typical CsI:Tl scintillators in digital imaging, since, along with their stated properties, they present a reddish emission wavelength (~611 nm) that matches the spectral sensitivity of CCD detectors [11,12], and, moreover, CsI:Tl suffer from radiation damage at high doses, probably due to its low density (4.51 gcm −3 ) and is slightly hygroscopic [13,14]. Another common material, the bismuth orthogermanate (BGO), presents a low density (7.13 gcm −3 ) and emits less than 10 photons/keV [15], compared to 30 photons /keV for Lu 2 O 3 :Eu 3+ [16]. Finally, the well known Gd 2 O 2 S:Tb 3+ (GOS), tends to be replaced in scintillating devices due to its chemical instability and sensitivity to moisture [17,18]. Furthermore, it has been demonstrated that Lu 2 O 3 :Eu 3+ presents a light yield similar to that of GOS [19] and that a scintillating screen made of this material can produce images of better quality than those obtained by the standard GOS [20]. On the other hand, it has been noted that different methods of synthesis lead to particles with diverse morphologies and specific surface areas, and, therefore, different luminescent properties [21,22]. Because it is essential for scintillating applications to yield ultrafine, mono-sized, low-aggregated and spherical powders [23], many soft chemical routes have been used to produce lutetia nanopowders, such as co-precipitation [24][25][26], solvothermal reaction [27][28][29][30][31], molten salts synthesis [32], and the combustion method [33][34][35][36][37]. Oxalic acid and urea have been used extensively in some of these methods, but these present some drawbacks for rare earth compounds, particularly due to the fact that rare earth oxalates produce particles that grow rapidly, along with severe agglomeration, whereas urea limits the yield from the homogeneous precipitation process [38]. Therefore, the sol-gel process, due to its unique advantages, like high chemical homogeneity, the possibility of achieving several compositions by simples changes in the process, and the ability to vary the nature and concentration of doping ions [39], it is ideal to produce rare earth doped Lu 2 O 3 nanopowders [40][41][42][43]. Finally, in order to improve the light yield of these materials, the Tb 3+ ion has been proposed for co-doping Lu 2 O 3 :Eu 3+ [44]. In Y 2 O 3 :Eu 3+ ,Tb 3+ , it has been found that there is an increment of the Eu 3+ luminescence, due to an energy transfer from Tb +3+ to Eu 3+ ions [45,46]; furthermore in previous work [47], it has been determined that, in Gd 2 O 3 :Eu 3+ nanopowders, the radioluminescence emission is enhanced by the incorporation of Tb 3+ ions.
The aim of the current work is to synthesize for first time Eu 3+ ,Tb 3+ co-doped Lu 2 O 3 scintillating nanopowders by a simple sol-gel process, starting with nitrates as precursors and without the use of urea or oxalic acid, and instead, diethyleneglycol (DEG) C 4 H 10 O 3 as polymerization agent. It was studied the chemical changes during the sol gel process of the nanopowders by means of a Thermogravimetric (TGA), differential thermal analyses (DTA) and infrared spectroscopy (IR). The structural evolution during the annealing process was analyzed by X-ray diffraction (XRD), the morphology by means of TEM observations, and their scintillating properties under X-ray (30 keV) radiation as a function of the heat treatment temperature and the Tb 3+ content. Figure 1 shows the TGA-DTA curves of the as-prepared Lu 2 O 3 xerogel 5 mol % Eu 3+ , 0.01 mol % Tb 3+ . As observed, the TGA curve shows three stages of weight loss. The first stage (I), from

Thermal Analysis
100-175 °C, shows a weight loss of 9% and two endothermic events, at 120 and 170 °C. The first event can be attributed to the release of water of hydration and OH, and the second one to the decomposition of the organic matter, this event suggest the formation of lutetium hydrated species or even carbonate species to form precipitates with Lu 3+ ions [48]. The second stage (II), from 175-360 °C, corresponding to a weight loss of 25%, involves two types of endothermic events, occurring at 210, 310 and 340°C. The first can be ascribed to the evaporation of DEG (b. p. 240 °C), whereas the last two correspond to the elimination of carbonyl groups. The last stage (III) exhibits a weight loss of 14% and includes two major events: a strong exothermic peak at 360 °C, which can be associated with the crystallization process of the ceramic sample into the cubic phase, and a strong endothermic peak at 390 °C, which was related to the pyrolisis of the remnant's carbon groups.

Infrared Analysis
The Fourier transformed infrared spectra of Lu 2 O 3 : Eu 3+ 5 mol %, Tb 3+ 0.01 mol% are depicted in Figure 2. This study was carried out in the range 4000-400 cm −1 on the dried precursor sol, thermally treated at different temperatures, in order to determine the evolution of the decomposed products of the xerogel powders calcined up to the crystallization process. For the xerogel at 100 °C, bands observed at 3400 cm −1 (ν), 1650 cm −1 (δ) and 750 cm −1 (δ) can be ascribed to O-H stretching (ν) and deformation (δ) vibrations, due the presence of water and alcohol groups. Since the heating at 400 °C was not enough to remove these species, we have to conclude that they were structurally built into the host and not merely adsorbed on the surface. After a 600 °C heat treatment, these O-H vibrations were less intense than those observed at lower temperatures, exerting an influence on the powder's microstructure, as demonstrated by XRD analysis. While the heating of the powders at much higher temperatures (necessarily to stimulate the growth of the crystallites and their partial sintering), the vibrations related with the O-H impurities [49] are almost absent, i.e., these vibrations are nearly missing at 900 °C. The absorption peak around 1380 cm −1 indicates the N-O stretching vibration [50] of  [51], and one is observed at 820-880 cm −1 , characteristic of the C-C bond [52]. The intensity of all these bands decreased with the annealing temperature, as has been observed for lutetia ceramics fabricated by nitrate sources [53]. However, after 800 °C sintering, the N-O vibrations were still present, suggesting that some NO 3 − residues were adsorbed in the sample, which could be removed after 900 °C thermal treatment. Finally, the bands at 580 and 489 cm −1 , observed from 400 °C, and attributed to the Lu-O stretching vibrations of cubic Lu 2 O 3 (a Lu 2 O 3 host lattice vibration) [54][55][56], indicate that the crystallization was just beginning at 400 °C, which was confirmed by XRD and TEM observations. Figure 3 shows the evolution of the X-ray diffraction patterns of the Lu 2 O 3 : Eu 3+ 5 mol %, Tb 3+ 0.01 mol% nanopowders annealed in air at temperatures ranging from 300 to 900 °C. At 300 °C, the xerogel exhibits an almost amorphous behavior; however, at 400 °C, the system possesses an aspect of an amorphous phase characterized by broad diffraction peaks, which also indicates that the crystallites are very small, lower than 5 nm. This result is in good agreement with the DTA observations, since the crystallization process began between 300 and 400 °C. As the annealing temperature reaches 500 °C, it becomes evident that the crystallization process has ended in a cubic Lu 2 O 3 structure (JCPDS 431021) with a spatial group Iā3 (lattice parameter 10.391 Å). With an increments the annealing temperature, the diffraction peaks become narrower, which reflects an increase in the size of the Lu 2 O 3 crystallites. Table 1 shows the calculated crystal sizes according to Scherer's formula D = 0.9λ/β cosθ [57], taking into account the broadening line of the diffracted peak, due to the effect of crystal size, where D is the crystal size of the powder,  (0.15406 nm) is the wavelength of the diffracted X-ray, β is the full-width radiation at half-maximum (FWHM) of the peak, and  is the Bragg angle of the diffracted X-ray. The crystallite size ranges from 5.3 nm at 500 °C to 25.5 nm at 900 °C; these observations were confirmed by TEM observations for the 700 and 900 °C co-doped lutetia nanopowders.   Figure 4a shows a TEM bright-field micrograph of selected Lu 2 O 3 : Eu 3+ 5 mol %, Tb 3+ 0.01 mol % scintillating nanopowders, annealed at 700 °C. As observed, the morphology of the powders is mainly angular; however, some of the faces are rounded and the particles are highly agglomerated. Figure 4b shows the particles' indexed diffraction pattern, which exhibits typical nanometric ring-type behavior and confirms the cubic structure. Figure 4c shows a dark field micrograph of the area at the (2 2 2) direction. The average size determined from these observations was ≈ 17 nm, in good accord with XRD results, and presents a normal centered distribution in the range of 5-40 nm (Figure 4d).

Structural Properties
Bright field, diffraction pattern and dark field micrographs, as well as particle size distribution of the sample annealed at 900 °C are shown in Figure 5a-b. As noted, similar observations can be made for this sample, whose average size was ≈ 27 nm. The kinetics growing of the sample can be related to the sintering and agglomeration processes activated by the thermal process.  Figure 6 presents the scintillating properties of Lu 2 O 3 : Eu 3+ 5 mol%, Tb 3+ 0.01 mol% nanopowders annealed at 700 and 900 °C, under X-ray excitation of 30 kV and 40 mA. The scintillating spectra of the sample display a group of emission lines situated in the 575-to 725-nm spectral region, corresponding the Eu 3+ transitions from the excited 5 D 0 level to 7 F j (J = 0,1,2,3,4) levels, with no evidence of any Tb 3+ emission (545 nm). The maximum emission line at 611 nm, which matches well the spectral sensitivity range of CCD cameras, corresponds to the 5 D 0 → 7 F 2 electric dipole transition, and is a product of the Eu 3+ ion located at C 2 sites in the Lu 2 O 3 host cubic lattice [58], and its intensity is evidence of an efficient channel of energy transfer from the lutetia matrix to Eu 3+ emission centers according to a recombination mechanism [59]. The less intense emission lines at 532, 580, 630 and 665 nm, correspond to the 5 D 0 → 7 F 0 , 5 D 0 → 7 F 1 , 5 D 0 → 7 F 3 and 5 D 0 → 7 F 4 transitions, respectively, and are in good agreement with the energy levels described elsewhere [60]. The difference in the light yield output between the two samples, of about ~110%, can be explained in terms of two effects. First, it is well known that the crystallinity achieved at higher temperatures is related to a more efficient activation of the Eu 3+ ion in the nanopowder sample. Furthermore, it has been demonstrated that until the complete elimination of residual OH contamination, the non-radiative multiphonon relaxation process can be stimulated with high-energy phonons introduced by this impurity [33]. In the current work, as can be established by IR experiments, at 900 °C the residual OH − impurity has been almost completely eliminated, which is a temperature considerably lower to the 1300 °C range that has been previously reported [61].  Figure 7 shows the light yield variation of the Eu 3+ 5 D 0 → 7 F 2 (611 nm) emission as a function of the Tb 3+ concentration for nanopowders annealed at 700 and 900 °C. In both cases, the lutetia co-doped at 0.01 mol % Tb 3+ presents the highest light yield, and for reasons explained earlier, the emission of the sample heat-treated at 900 °C is higher than that of the 700 °C one. The enhancement of the Eu 3+ emission in the nanopowders is due to a non-radiative energy transfer from Tb 3+ to Eu 3+ , as Tb can absorb more X-ray radiation. At higher Tb 3+ concentrations, a self-quenching mechanism [45] exists, a product of the fact that the Tb-Tb energy transfer mechanism is more efficient than the Tb-Eu one, and, therefore, presents a higher possibility that a Tb ion is closer than another sort, resulting in a drop in Eu 3+ emission. Similar results have been observed for Gd 2 O 3 co-doped Eu, Tb nanopowders [47].

Experimental Section
Lu 2 O 3 :Eu 3+ , Tb 3+ nanopowders were prepared by a simple sol-gel process method. Figure 8 shows the flow scheme for this process. The starting materials used were lutetium nitrate, Lu(NO 3 ) 3 •6H 2 O (Alfa Aesar, 99.96%), europium nitrate Eu(NO 3 ) 3 , terbium nitrate Tb(NO 3 ) 3 (99.5% and 99.6% respectively, Alfa Aesar), acetic acid CH 3 COOH (Fermont 98%) used as a catalyst, and diethyleneglycol (DEG) C 4 H 10 O 3 (Alfa Aesar, 99%) used as a polymerization agent. Initially, the lutetium nitrate was dissolved in an ethanol-deionized water solution (95-5 vol%) using a hot stirrer at 40 °C for 3 hours to obtain 25 mL of a 0.95 M Lu sol. Thereafter, the solution pH was adjusted by incorporating acetic acid (0.17 M), and the DEG was added (0.42 M). Finally, the co-doping elements, previously dissolved in ethanol, were incorporated the initial lutetium sol. The sol obtained was mixed for 2 hours to produce a transparent solution, stable for more than 3 months. Subsequently, the sol was dried at 100 °C for 24 hours, and the yielded gel was calcined at different temperatures for 1 hour to obtain the Lu 2 O 3 :Eu 3+ , Tb 3+ scintillating nanopowders. Thermogravimetric (TGA) and differential thermal analyses (DTA) of dried gel were conducted using a SDT Q600 TA instrument; the studies were performed in air with a heating rate of 10 °C min −1 . The IR spectra of the samples were recorded in the range of 4000-400 cm −1 using Fourier transform infrared spectroscopy (FTIR 2000, Perkin Elmer) and the KBr pelleting technique. The phase composition of the powders was identified by X-ray diffraction at room temperature on a powder diffractometer (Bruker D8Advance) using Cu K radiation (1.5418 Å). The morphology and particle size of the powders were observed with a transmission electronic microscope (JEOL 2200) operating at 200 keV. The scintillating properties were recorded employing an X-ray generator (PW-1830 Phillips) operating at 30 kV and 40 mA, using W K radiation (0.2086 Å) and a photon flux of 10 12 ph s −1 .

Conclusions
Lu 2 O 3 :Eu 3+ , Tb 3+ nanopowders have been prepared by a simple sol-gel method, using lutetium, europium and terbium nitrate as precursors, and DEG as a polymerization agent. This nanopowder crystallizes into a cubic system from 400 °C, and the process is completed at 500 °C, resulting in particles composed of crystallites ranging in size from 5.3 to 25.5 nm. The increments in crystallite sizes depend on the heat treatment temperatures. The powder exhibits interesting scintillating properties, and at 611 nm presents a reddish emission corresponding with the Eu 3+ 5 D 0 → 7 F 2 transition, which makes it a promising material for X-ray detection systems, since this emission matches the maximum efficiency of the CCD cameras. The light yield of the nanopowders was analyzed at two different annealing temperatures: 700 and 900 °C. It was established that at the sample heat treated at the higher temperature, presents an enhanced light output, presumably due to better crystallinity and to the complete removal of OH -. Finally, it was determined that with the co-doping of Tb 3+ at a 0.01 mol% level, the light yield was enhanced compared to trials with higher Tb contents, presumably due to the effect of a self-quenching process.