Influence of the synthesis pathway upon the thermoluminescent response of gadolinium aluminate matrix phosphors

The phosphors endowed with a glow curve have a peak much more intense than the others, sensitization to low doses, and a linear relationship between the thermoluminescent response and the dose, which are in demand in the dosimetry market. The gadolinium aluminate has been successfully used as a host lattice of phosphors with luminescent properties. Principally, the luminescent response of the said matrix has been improved by doping with rare earth ions. The thermoluminescent signal depends on the synthesis route because the defect density of a crystal is closely connected to the production pathway that the material went through. Therefore, a crucial step in sensitizing a phosphor is selecting a synthesis route that enhances the efficiency of its thermoluminescent mechanisms. The research herein focused on synthesizing phosphors composed of gadolinium aluminate with no dopant and doped with two molar percent of dysprosium, using two different routes. Among the two synthesis pathways, it was determined to be the most suitable for enhancing the thermoluminescent response of the aforementioned phosphors. The methods employed were reverse coprecipitation and the modified citrate precursor. The phosphors obtained by the reverse coprecipitation method showed the most sensitive thermoluminescent signal. Mainly, the phosphor activated with the dysprosium ion produced the most intense signal, suggesting an improvement in the thermoluminescent mechanisms due to the dopant insertion.


Introduction
In recent years, thermoluminescent research has focused on improving the thermoluminescent (TL) response of phosphors that have already demonstrated TL sensitivities.A phosphor must meet several criteria to be employed in measurements of ionizing radiation doses using TL signal assessment.The main criterion is that the shape of the glow curve should be featured either by one peak or, in case of more than one peak, there must be a peak that is significantly more intense, surpassing the others.Besides, the main peak should be narrow, and it should be located at a far temperature from room temperature to avoid fading under storage conditions materials [1].The optimization process of the TL phosphor starts with the suitable pick of the host and activator material.Once the components of the phosphor have been chosen, a crucial step that rules on its density of lattice defect is the synthesis pathway.Currently, ceramics production in the industrial environment relies on the solid-state route.However, wet chemical routes allow for the control of numerous variables of synthesis and the reduction of the huge energy waste associated with the solid-state pathways.For this reason, wet chemical synthesis has been used in many studies to produce TL phosphors [2][3][4].
Phosphors based on GdAlO3 (GAO) host have mainly been produced for photo-luminescent and neutron rod control applications due to numerous useful properties such as thermal and chemical stabilities, low cut-of phonon energy, high fusion point, and low thermal expansion [5,6].Nonetheless, the TL properties of the phosphors relied on the GAO host have only been reported by a few previous studies, including two published by our research team, which focused on the characterization of powders synthesized by the reverse coprecipitation method.The results showed that the phosphors were sensitive to beta doses ranging between 1.1 Gy and 44 Gy.Besides, doping the host with low amounts of rare earth ions yielded an enhanced signal intensity [4,7].In addition to the mentioned works, two more were made by Tamrakar R et al. [8] and Tiwari S et al. [9], respectively, that studied the stimulated luminescent response of the GAO matrix synthesized by solid-state reaction.They arrived at the same conclusion regarding the favorable influence of rare earth ion activation on matrix sensitization.
In the research described herein, two types of wet chemical routes were chosen to synthesize phosphors based on orthorhombic gadolinium aluminate (GAO).One of them, the reverse coprecipitation method (RCM), was picked because of its simplicity and low cost [10].The other method chosen was the citrate precursor modified method (CPM), a variation of the Pechini method, excluding the chelation process.It is distinguished by the excellent dispersion of the starter ions achieved through the citrate process, which decreases the probability of segregation [11].
Keeping in view the importance of the synthesis pathway upon the potential TL applications of phosphors, the research was focused on establishing a comparison among the glow curves recorded for powders with Gd1-xDyxAlO3 stoichiometries, where the value x=0 represents the undoped material.The most recommended method to synthesize the said phosphor was selected based on both the TL sensitivity (translated into the intensity of the TL response) and the shape of the glow curve.The TL response is strongly connected to the lattice defect density, but other factors influence the electron transitions that result in emission, such as retrapping or non-radiative recombination.Besides, the density of lattice defects is not totally proportional to TL sensitivity; each host and activator combination is associated with a suitable amount of lattice defects that maximize the TL emission.For these reasons, it is really difficult to draw a conclusion about why the material was TL sensitive or not due to the inevitable lack of information.Hence, the work herein was limited to deciding which synthesis method yielded the greatest improvement in TL response.Any other conclusion about the causes must be regarded as a conjecture and should be thoroughly studied in further investigations.

Materials and methods
The methods and materials used to pursue the investigation will be described in the following sections.

Synthesis methods
In order to establish the influence of the synthesis method upon the TL response of based GAO host phosphors, four samples were produced.Each sample differed from the others in the composition or the synthesis pathway.Since the TL response of a no-activated host could be strongly modified by doping, in the first step, the undoped GAO phosphors were produced.Subsequently, the host was activated by Dy 3+ ion with a Gd0.99Dy0.02AlO3stoichiometry.Regarding the starter materials, Gd2O3 (99.9%+) from Alfa Aesar, and Al(NO3)3•H2O (98.7%+) from J.T.Baker were used as sources of Gd 3+ and Al 3+ ions, respectively.On the other hand, the source of the dopant ions was the Dy2O3, which was produced by calcining Dy(NO3)3•xH2O at 1000 °C.The Dy 3+ ions were obtained from the oxide form instead of the nitrate form to avoid complications arising from the unknown hydration state of the nitrate.To yield the starter solutions, the nitrate was diluted in distilled water, while the oxides were dissolved in HNO3 from Meyer after proper dilution in water.As it was mentioned at the section above, the phosphors were synthesized by reverse coprecipitation and citrate precursor methods.Both methods started with preparing stoichiometrically correct mixtures of starter ions, followed by a divergent step.The subsequent subsections going to describe the further steps of each method.For the RCM, the starter ion mixture was added dropwise into a vessel containing a sufficient amount of NH4OH to produce the precipitation of the starter ionoxyhydroxides.Consequently, the precipitate was filtered, washed in distilled water at 70 °C, and finally filtered again.
2.1.2.Citrate precursor modified method.For the CPM, the starter ion mixture was heated up to 80 °C while it was stirred continuously.Achieved the said temperature, a citric acid solution was added.The stoichiometry relationship between citric acid and host lattice starter ions was 2:1.To achieve the optimal development of the starter ion citrates, the mixture was stirred for 30 minutes the said temperature.After the citrate formation period completed, the mixture was converted into powder using an automatic spray drying (YAMATO-ADL31).The input parameters submitted to the spray dryer device are shown in Table 1.

Calcination
Both products from the two synthesis routes were dried at 100 °C in an oven for 24 hours.After being removed from the oven, the RCM precipitate transformed into small opaque rocks from its original white mud form, while the CPM products retained their brown powder form.The RCM products were manually ground to convert them into powder.Both from RCM and CPM powders were submitted to a calcination cycle, which started at room temperature and reached 1500 °C with a heating rate of 5°C/min.Then, the temperature remained at the maximum for 5 hours, and finally, decreased to room temperature trough heat exchange with the environment.

X ray diffraction
The crystal phases of the synthetized powders were identified using a X ray diffraction (XRD) device (Bruker D8 Advance) that employs a Cu Kα radiation with a wavelength of 1.541 Å.To record high quality diffraction patterns, a 2θ step of 0.02° and a time step of de 0.6 s were applied.

Scanning electron microscopy
The morphologies of the synthetized powders were analyzed by scanning electron microscopy (SEM) micrographs carried out by a JEOL JSM 7600F electron microscope using upper secondary electrons in-lens.The images were quality improved by covering the samples with copper layers.Besides, all powders were fixed on carbon bands to prevent the contamination of the device.The agglomerated powders were composed of grains whose sizes were determined by averaging 100 grains per sample, using the Lince software [12].

Thermoluminescent measurements
The TL response depends on the amount of material exposed to radiation.Then, samples of 10 mg were prepared to ensure comparable results.An automatic reader (lexsygsmart) from Freiberg Instruments, which holds a 90 Sr/ 90 Y beta source capable of delivering a dose rate of 0.11 Gy/s, was employed to irradiate and read out the glow curves.Recording the TL response, the samples were heated from room temperature to 400 °C using a heating rate of 10 °C/s.

Deconvolution of the glow curves
The glow curves recorded for some of the synthesized powders were deconvoluted using the TLanal software [13].The said software uses the Figure of merit (FOM) to assess the goodness of fit.The criterion applied to stop the iteration process of a deconvolution was to reduce the FOM by less than 2.5 % [14].To avoid wrong hypothesis about the mechanisms governing the TL emissions, a model of general kinetic order was applied.The electron traps were characterized by the quantitative values of the following parameters: trap depth, frequency factor and kinetic order.

Results and discussions
The results of the research are presented and discussed in the following sections, according to the technique used.

X ray diffraction
The Figure 1 shows the XRD patterns recorded for each sample and the peak indexation.The powders synthetized by RCM were composed of two phases.In addition to the GAO phase (PDF 00-046-0395), which was identified as principal, there was a secondary phase with stoichiometry Gd2Al4O9, whose crystal system is monoclinic (PDF 00-046-0396).The minority phase detected (GAM) has been identified in previous syntheses of the Al2O3-Gd2O3 system [5].The appearance of the remaining phase different that GAO phase, may be associated with a segregation of precursor ions during the synthesis process.However, the powders synthesized by CPM were exclusively composed by the GAO phase.
The phase purity of the CPM products confirmed the high degree of starter ion dispersion that can be achieved by the citrate formation.The greater dispersion of starter ions implies a lower likelihood of segregation, thereby preventing the formation of phases with undesired stoichiometries.

Scanning electron microscopy
The Figure 2 shows the SEM micrographs recorded for the synthetized powders.The surfaces of all particles were constituted by well-defined grain edges that formed at the high temperatures reached by the samples.The images also show some cracks and microporosities along the grain edges, mostly in the doped samples.Defects in the crystal lattice could be localized both in grain edges and in cracks, thereby enhancing the TL emission of the phosphor.It is also observed that with the CPM synthesis route the particles grew more, which indicates that there was a greater diffusion during the grain growth process due to the calcination temperature.

Thermoluminescent measurements
The TL responses recorded for the synthesized powders irradiated with a beta dose of 39.6 Gy are shown in the Figure 3. Regarding the glow curve shape of the doped samples, doping enhanced the intensity of the signals for both synthesis methods.The glow curve recorded for the sample synthesized by RCM consisted of three sharp peaks located at 150°C, 219 °C and 355 °C, with the first having the highest intensity and surpassing the others.By contrast, the peaks were wider and shifted to higher temperatures for the signal ascribed the sample synthesized by CPM.For the CPM synthesized sample, the TL peaks appeared overlapped, suggesting a quasi-continuous distribution of electron traps, unlike the RCM synthesized sample whose glow curve consisted of well-separated peaks associated to electron traps having well localized energy levels [15].Additionally, the electron traps in the sample synthetized by CPM required higher energies to be emptied, which is why the peaks emerged at higher temperatures.
It is worth emphasizing the enhancement in the intensity of the signal yielded by doping.Commonly, the insertion of a suitable number of foreign ions into the host lattice is related to an improvement in efficiency of the TL mechanisms [16,17].Thus, for the RCM could, there may have been an increase in the likelihood of radiative recombination, resulting in a sensitivity increase of more than five times due to doping.For the samples synthesized by CPM was harder to propose a conjecture, since the samples passed from an almost null signal to an increment in intensity of four orders.It is possible that the undoped sample yielded peaks located far from the threshold of temperature applied for recording the TL responses, and the insertion of Dy 3+ ions may have caused the peak shifting toward low temperatures.The previous hypothesis could be confirmed by recording the glow curve at much higher temperatures.

Deconvolutions of glow curves
Table 2, Table 3, and Table 4 show the results of deconvoluting the glows curves recorded for the synthetized powders, excluding the undoped sample synthesized by CPM due to its poor response.First of all, it is worth highlighting that all deconvolutions accomplished the stopping criterion, as they were ascribed to FOM values that were less than 2.5 % for all cases.The glow curve recorded for the undoped sample synthesized by CPM consisted of five peaks, whereas the doped sample synthetized by the same method was ascribed to a glow curve with only three peaks.The changes in the number of peaks observed upon doping could be related to an approach of the energy levels for the electron traps.Likewise, the doped sample synthetized by CPM was featured by a glow curve with three peaks, indicating the substitution of the two percentage of Gd 3+ ions by Dy 3+ ions into the GAO lattice could be associated to the formation of three electron traps with well-defined energy levels.Additionally, the Peak 2 of the sample synthetized by CPM exhibited striking features, since its depth and frequency factor values were the significantly lower than those of the rest of peaks, whereas its kinetic order indicated a small probability of retrapping during its appearance.The trap parameters of the peak 2 suggested that it was associated to a small number of active electron traps.However, the Figure 4 demonstrates that Peak 2 was the most intense.Therefore, in the formation of the other peaks shown in the Figure 4 must have also been involved the non-radiative recombination [1,18].

Conclusions
Powders composed of either the gadolinium aluminate lattice as host or this matrix activated with dysprosium ions were successfully synthesized using reverse coprecipitation and citrate precursor routes.A remnant phase, in addition to the orthorhombic gadolinium aluminate phase, was detected in the samples synthesized using the reverse coprecipitation method.In contrast, the samples yielded using the alternative method consisted of pure orthorhombic gadolinium aluminate phase.Some of the lattice defects, upon which the thermoluminescent sensitivity of the powders relied, could have been located in the grain boundaries, porosities, and cracks observed using scanning electron microscopy.The activation of the host lattice using dysprosium ions as a dopant sensitized the thermoluminescence response fivefold for the reverse coprecipitation method and by more than four orders of magnitude for the citrate precursor method.This sensitization may be attributed to an approach in the energy levels of the electron traps.Regarding the shape of the glow curves and the sensitivity of the samples, the reverse coprecipitation method was the most recommended method for producing phosphors based on an orthorhombic gadolinium aluminate matrix.Non-radiative recombination processes may have influenced the thermoluminescent mechanisms of the doped sample synthesized using the citrate precursor method, resulting in a loss of sensitivity.

Figure 3 .
Figure 3. TL responses recorded for the synthesized powders irradiated with a beta dose of 39.6 Gy.

Figure 4 .
Figure 4. Deconvolution of the glow curve recorded for the doped sample synthesized by CPM irradiated with 39.6 Gy of beta dose.

Table 1 .
Input parameters of the spray dryer device.

Table 2 .
Trap parameters obtained for the synthesized powders.

Table 3 .
Trap parameters obtained for the synthesized powders.

Table 4 .
Trap parameters obtained for the synthesized powders.