lable at ScienceDirect Optical Materials 75 (2018) 297e303 Contents lists avai Optical Materials journal homepage: www.elsevier .com/locate/optmat Luminescent Eu3þ doped Al6Ge2O13 crystalline compounds obtained by the sol gel process for photonics Lauro J.Q. Maia a, *, Fausto M. Faria Filho a, Rog�eria R. Gonçalves b, Sidney J.L. Ribeiro c a Grupo Física de Materiais, Instituto de Física, Campus II, CP 131, 74001-970, Universidade Federal de Goi�as-UFG, Goiânia, GO, Brazil b Departamento de Química, Faculdade de Filosofia, Ciências e Letras de Ribeir~ao Preto, Av. Bandeirantes 3900, 14040-901, Universidade de S~ao Paulo-USP, Ribeir~ao Preto, SP, Brazil c Institute of Chemistry, CP 355, 14801-970, S~ao Paulo State University-UNESP, Araraquara, SP, Brazil a r t i c l e i n f o Article history: Received 21 July 2017 Received in revised form 27 September 2017 Accepted 23 October 2017 Available online 6 November 2017 Keywords: Al6Ge2O13 crystalline phase Eu3þ doped Structural properties Optical properties * Corresponding author. E-mail address: lauro@ufg.br (L.J.Q. Maia). https://doi.org/10.1016/j.optmat.2017.10.038 0925-3467/© 2017 Elsevier B.V. All rights reserved. a b s t r a c t We synthesized pure and Eu3þ doped Al6Ge2O13 samples by an easy and low-cost sol-gel route using the GeO2, Al(NO3)3$9H2O and Eu(NO3)3$6H2O as precursors, tetramethylammonium hydroxide and ethanol as solvents. The Al6Ge2O13 crystalline phase possesses orthorhombic structure and is a potential host for rare earth ions, especially due to high aluminum concentration. Homogeneous and transparent sols and gels were obtained. The samples containing 1 mol% of Eu3þ were heat-treated at 1000 �C to eliminate organic compounds, providing high optical quality and structural purity. All materials were characterized by thermogravimetric and differential thermal analysis, X-ray diffraction, Fourier transform infrared spectroscopy, high resolution transmission electron microscopy, selected area electron diffraction, diffuse reflectance spectra in the ultravioletevisibleenear infrared regions and photoluminescence measure- ments. High purity of Eu3þ doped Al6Ge2O13 orthorhombic phase and well crystallized grain dimensions of around 100 nm was obtained with high red photoluminescence emission. The decay lifetime of 5D0 level from Eu3þ (the emission at 612 nm) was determined, being between 0.97 and 2.12 ms, and an average quantum efficiency of 54% was determined (considering the average experimental lifetime of 1.77 ms). Moreover, it was calculated and analyzed some parameters of Judd-Ofelt theory applied to Eu3þ emissions from Al6Ge2O13 host. The results show that Eu3þ doped Al6Ge2O13 crystalline compounds have large potential to be used in displays and LED devices. © 2017 Elsevier B.V. All rights reserved. 1. Introduction Scientific community has great interest in rare earth ions doped inorganic materials due to their interesting optical properties. Rare earth doped inorganic materials have been used in solid-state la- sers, active planar waveguides, optical fiber amplifiers, lighting emitting diodes (LED's), memories, solar cells, and displays [1e3]. Rare-earth (RE) trivalent ions can emit light from the near-infrared (NIR) to the ultraviolet (UV) due to intra-4f or internal 4fe5d transitions [1]. Europium ions (Eu3þ) have interesting optical features such as well-defined spectral lines and red/orange emissions, which have been exploited for technological applications in displays, fluores- cent lamps, fluoroimmunoassays. The most standard commercial red phosphors are based on Eu3þ doped oxides (Y2O3:Eu3þ, YPO4:Eu3þ, YVO4:Eu3þ, Y2(WO4)3:Eu3þ, Y2O2S:Eu3þ). In order to make red phosphor with high red emission intensity, a high europium concentration is required. However, due to the onset of concentration quenching in commercial matrices, the relatively low efficiency can be achieved with high concentrations of Eu3þ. On the other hand, the Eu3þ ion is often used as structural probe when doped in crystalline or amorphous structure due to its particular optical transitions [4], and a detailed understanding of the local structure and bonding of dopant ions is important for optical device engineering [5]. Concerning improvement of optical properties, firstly a reduc- tion or elimination of non-radiative processes is required, such as OH groups which are responsible for luminescence quenching. Multiphonon relaxation represents one significant non-radiative process and, consequently, the choice and control of vibrational structure of the host is crucial to enhance the radiative process. Other important parameter is related to rare earth concentration mailto:lauro@ufg.br http://crossmark.crossref.org/dialog/?doi=10.1016/j.optmat.2017.10.038&domain=pdf www.sciencedirect.com/science/journal/09253467 www.elsevier.com/locate/optmat https://doi.org/10.1016/j.optmat.2017.10.038 https://doi.org/10.1016/j.optmat.2017.10.038 https://doi.org/10.1016/j.optmat.2017.10.038 Fig. 1. TG and DSC curves of the Eu3þ doped Al6Ge2O13 compound previously calcined at 400 �C. L.J.Q. Maia et al. / Optical Materials 75 (2018) 297e303298 and their distribution in the host, where ion-ion interactions give rise to non-radiative processes as energy migration, cross relaxa- tion, which reduce considerably the luminescence [6,7]. Besides, optical features improvement can be also achieved changing the microenvironment of the rare earth ions, defined by symmetry sites and refractive index, which influences the transitions probabilities. Therefore, there is a great interest to develop physically and chemically stable host matrices with high rare-earth ions solubility, especially for europium to have pure red emissions to be used in LED and display devices. One potential rare earth host candidate is the Al6Ge2O13 crys- talline phase, which is characterized by an orthorhombic structure containing high concentration of aluminum ions, which can allow noteworthy rare earth content incorporation to be used in solid- state lasers, optical amplifiers, LED's, and displays. Nevertheless, the literature reports that obtaining Al6Ge2O13 crystalline phase is difficult because high temperatures are required, as high as 1325 �C, and long heat-treatment periods (15 h), especially by solid state reaction [8e10] using Al(NO3)3$9H2O or Al2O3 as precursors for aluminum and GeO2 for germanium. Important to stress that the traditional sol-gel route, which is a soft route, has been used for producing homogeneously different samples using alkoxydes, mainly TEOG [Ge(OC2H5)4] as a germanium precursor [11e13]. However, germanium alkoxydes are very expensive and highly sensitive to atmospheric humidity, requiring its manipulation in a dry glove box. To overcome this inconvenience, we have developed in recent works [1,14] a meth- odology, which doesn't require the use of controlled atmosphere and humidity. Recently, Er3þ doped SiO2-Al2O3-GeO2 compounds were prepared and Al6Ge2O13 nanocrystallites were obtained around 1000 �C [1]. In the present work, we synthesized pure and Eu3þ doped Al6Ge2O13 samples by simple and lower cost sol-gel route than traditional procedures. The structural, microstructural and spec- troscopic properties of the pure and Eu3þ doped Al6Ge2O13 com- pounds were studied to be used in displays and LED's as red emitters. 2. Experimental procedure 2.1. Synthesis Transparent sols were prepared by the sol-gel route to form the polymeric precursor. First, citric acid (CA, C6H8O7, 99.5% purity, Synth) was diluted in ethanol (EtOH, C2H5OH, 99.5% purity, Chemycalis) for 30 min under vigorous stirring (solution 1). Al(NO3)3$9H2O (aluminum nitrate nonahydrate, 98% purity, Sigma Aldrich) and Eu(NO3)3$5H2O (europium nitrate pentahydrate, 99.9% purity, Sigma Aldrich) were diluted in ethanol separately (solution 2). In parallel, 200mg of GeO2 (germanium oxide, 99.998% purity, Sigma Aldrich) was reacted with 3 ml of TMAH solution (tetramethylammonium hydroxide, C4H13NO, 25% volume inwater, Sigma Aldrich) and 200 ml of deionized H2O, and stirred vigorously for 20 min to obtain the solution 3. Solution 3 was added to the solution 1, followed by solution 2 addition. The final solution was vigorously stirred for 20 min and remained resting at room tem- perature for 24 h, yielding stable and transparent sols. The molar ratio of the citric acid and metal was 3:1, and each 1 g of CA was dissolved in 2 ml of EtOH, and each 1 g of aluminum and europium nitrates was dissolved in 10 ml of EtOH. To eliminate the solvent and trigger gelation (hydrolysis and condensation reactions), the sols were oven-dried at 150 �C for 24 h. The dry gels were then calcined at 400 �C for 24 h. Each sample was ground into fine powder in an agate mortar and heat- treated at 1000 �C. The heat treatments were performed in a muffle furnace under air atmosphere for 1 h, using alumina crucibles. 2.2. Characterization The thermal decomposition and crystallization processes of the powders previously calcined at 400 �C/24 h were studied by the thermogravimetric (TGA) and differential thermal analyses (DTA) (Setaram LABSYS EVO), which were performed simultaneously under a continuous flow of O2. Samples (~20 mg) were heated in a platinum crucible at 10 �C/min from 25 �C to 1150 �C, and a plat- inum crucible was used as reference during the measurements. After annealing at 1000 �C, the samples were characterized structurally by X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR). The XRD measurements were taken with a Shimadzu XRD-6000 X-ray diffractometer with Bragg- Brentano theta-2 theta geometry, at a continuous scan speed of 0.04�/s from 15 to 65�. Ka radiation of 1.54059 Å from a Cu tube operating at 40 kV was used. The FTIR measurements were ob- tained with a Perkin Elmer Spectrum 400 FT-IR spectrometer. The experiments were performed from 4000 to 400 cm�1 with a res- olution of 2 cm�1, and 56 spectra were recorded. High-resolution transmission electron microscope (HRTEM) was used for obtaining images from a JEOL JEM 2010 operating at 200 keV, and electron diffraction (SAED) patterns from selected area. The diffuse reflectance spectra were obtained using a UVeViseNIR PerkinElmer Lambda 1200WB spectrophotometer and a Praying Mantis® accessory. The BaSO4 powder from Sigma- Aldrich was used as standard reflectance material. The photoluminescence (PL) emission spectra were performed at room temperature using a Horiba-Jobin Yvon Fluorolog FL3-221 spectrofluorimeter, equipped with double monochromator and a Hamamatsu photomultiplier tube. For excitation, a continuous 450 W Xe arc lamp was employed. PL emission spectra were cor- rected for the spectral response of the monochromators and the detector using a typical correction spectrum provided by the manufacturer. The PL decay curves at 612 nmwere obtained under excitation at 394 nm using a pulsed Xe lamp (3 ms bandwidth) in a Horiba-Jobin Yvon Fluorolog FL3-222 spectrofluorimeter. 3. Results and discussion Fig. 1 shows the TG and DTA curves of Eu3þ doped Al6Ge2O13 compound previously calcined at 400 �C. Although the technique is L.J.Q. Maia et al. / Optical Materials 75 (2018) 297e303 299 semi-quantitative, mass variation can be measured accurately. The TG curves revealed mass loss events, which were related to endo- thermic or exothermic reactions as indicated by the DTA curves. These endothermic or exothermic reactions are attributed to three main reactions: firstly, elimination of adsorbed water (peak around 147 �C) with a mass loss of 12%; secondly, carbon oxidation and its elimination in the form of CO and CO2 (combustion reaction, broad peak centered at 346 �C) with a mass loss of 11%; and finally, a crystallization process starting at 810 �C, with the peak at 830 �C, which is assigned to the Al6Ge2O13 crystalline phase formation. In the crystallization process, the mass of the sample reduces around 1% was related to an elimination of carbon and/or hydroxyl groups strongly linked to the metals, followed by a gradual increasing of the mass probably due to oxygen absorption from atmosphere (residual oxidation reaction) and some broad exothermic peaks related to changes in the AlO4 e AlO6 groups ratio in the crystalline phase (to be confirmed in future works for this phase). Someworks on mullite (3Al2O3:2SiO2) studies mention that the relation be- tween tetrahedral and octahedral groups is dependent of annealing temperature [15]. Fig. 2 shows XRD patterns of pure and Eu3þ doped samples heat- treated at 1000 �C. For comparison, the diffraction position from the JCPDS card number 71-1061 was included in Fig. 2. Only the Al6Ge2O13 phase in an orthorhombic structure and the Pbam (55) space group (mullite-type structure) crystallized at 1000 �C. Based on the XRD data displayed in Fig. 2 (peak position and Miller index of diffracted planes), and using the Rede 93 software program developed by Paiva Santos et al. [16], the following cell parameters were calculated: a ¼ 7.61(2) Å, b ¼ 7.97(8) Å, and c ¼ 2.93(4) Å for pure sample and a ¼ 7.66(2) Å, b ¼ 7.80(5) Å, and c ¼ 2.91(3) Å for Eu3þ doped sample. These values are similar to those listed on JCPDS card number 71-1061 (a ¼ 7.65(2) Å, b ¼ 7.779(2) Å, c ¼ 2.925(2) Å). The cell volume was 178(5) Å3 for pure, was 174(3) Å3 for Eu3þ doped, both similar to JCPDS card number 71-1061 with 174.06(6) Å3. Comparisonwith pure Al6Ge2O13 and JCPDS reference suggests that the insertion of rare earth ions don't change signifi- cantly the cell volume and parameters. Gao et al. [17] produced Al6Ge2O13 ceramic powder by the co-precipitation method, using Al(NO3)3 and Cl3GeCH2CH2COOH as precursors. In comparisonwith reported synthesis, pure and Eu3þ doped Al6Ge2O13 phase at 1000 �C were easily obtained using a friendly chemical route. The Al6Ge2O13 phase in an orthorhombic structure have the same Pbam (55) space group of mullite structure, as a consequence it has been assumed that both compounds has comparable Fig. 2. X-ray diffraction patterns of pure and Eu3þ doped Al6Ge2O13 heat treated at 1000 �C, and the diffraction pattern of the JCPDS card number 71-1061 for comparison. structures and consequently Ge4þ are distributed in the sites occupied by Si4þ. The backbone of the mullite structure is edge- sharing AlO6 octahedra (designated as M sites) forming chains running parallel to the crystallographic c-axis [18]. A part of the Si4þ is replaced by Al3þ. The compensation of the substitution-induced charge deficiency is achieved by removal of a number of oxygen atoms bridging the tetrahedral T2O5 groups [designated as O3 or alternatively O(C)]. This produces oxygen vacancies (designated as O3 or O vacancies), according to the coupled substitution 2Si4þ þ O2� / 2Al3þ þ vacancy (,), with x of the general formula related to the number of oxygen vacancies per unit cell in the composition range 0 < x < 0.67. The formation of O vacancies is accompanied by the displacement of the T positions close to the bridging O atoms to the new T* positions. The T*O4 tetrahedra form trimers or triclusters of three tetrahedral having a common bridging O atom. The studies mentioned that most of the T* sites are occupied by Al. The combination of octahedral chains with tetrahedral di and triclusters and O vacancies produces disturbed structural channels running parallel to the crystallographic c-axis. The distribution of the O vacancies and of the tetrahedral Al and Si atoms in a first approach appears random [18]. Aryal et al. [19] shown that mullite is an example of 2 x 2 x 2 supercell. All fully occupied sites were retained and those with partial occupations had the appropriate number of atoms removed. Oxygen vacancies were created by removing some O atoms that were close to such that the supercell was stoichiometric and charge neutral. The resulting 126-atom supercell for mullite consists of 36 Al, 12 Si, and 78 O atoms [20], being ~16 units of AlO6 octahedra and ~20 units of AlO4 tetrahedra. However, the tetrahedral to octahedral aluminum site ratio is dependent on the heat-treatment temperature [15] and certainly on the synthesis process. The solubility of Eu3þ in Al6Ge2O13 orthorhombic structure is not known, but can be strongly limited by the aluminum sites quantity. Note that Ge4þ, Al3þ and Eu3þ have ionic radius of 0.39, 0.535, and 0.947 Å respectively [20]. As a consequence, it can be assumed that Eu3þ should prefer replace Al3þ of AlO6 octahedra (designated as M sites), because they have same valence and smaller difference between its ionic radius, also the octahedral sites can accommodate easily lanthanide elements than tetrahedral sites. Note that in octahedral sites of mullite the average inter- atomic distance between Al and O atoms is 1.913(2) Å, and in tetrahedral sites is 1.700(2) Å [15], and consequently some octa- hedral sites can well accommodate Eu3þ ions. Fig. 3. Vibrational absorption spectra in the infrared region of pure and Eu3þ doped Al6Ge2O13 powders heat treated at 1000 �C. Fig. 4. (a) and (c) HRTEM images and (b) and (d) SAED of pure and Eu3þ doped Al6Ge2O13 heat treated at 1000 �C, respectively. L.J.Q. Maia et al. / Optical Materials 75 (2018) 297e303300 Fig. 3 shows the FTIR spectra of pure and Eu3þ doped samples heat-treated at 1000 �C. Absorption band attributed to stretching vibrational mode of hydroxyl groups (OH) at around 1639 and 3450 cm�1 were observed in both samples. Low intensity suggests significant elimination of OH content without relevant change in the presence of Eu3þ ions. The bands in the region of 400e1200 cm�1, centered at 475, 525, 580, 712, 805, 880, 945, 1035, and 1080 cm�1 correspond to vibrational modes of AlO4, AlO6 and GeO4 groups of the Al6Ge2O13 crystals, as illustrated in Fig. 3. Their assignments are listed in Table 1, and are in accordancewith those assigned to the vibrational modes of the Al6Ge2O13 crystalline phase reported byMeinhold and Mackenzie [10]. Fig. 4(a) and (c) shows HRTEM images of pure and Eu3þ doped Al6Ge2O13 samples, respectively. Well defined atomic planes of 0.34(3) nmwas observed in pure sample and attributed to the (210) plane of the orthorhombic Al6Ge2O13 phase, and 0.53(3) nm assigned to the (110) plane in Eu3þ doped sample, in agreement with crystallization of the Al6Ge2O13 phase. The SAED patterns in Fig. 4(b) and (d) reveal that the particles are monocrystals. Band gap values were calculated from the diffuse reflectance measurements displayed in Fig. 5(a) and (b), using the procedure previously described by Pessoni and co-workers [21]. In the Kubelka-Munk model [22], the remission function FðR∞Þ is defined as: FðR∞Þ ¼ ð1� R∞Þ2 2R∞ (1) where R∞ is the measured reflectance in % normalized by a stan- dard, which in our case was BaSO4. The band gap (Eg) and the ab- sorption coefficient (a) are correlated by ahn ¼ C1 ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi hn� Eg p , where hn is the photon energy and C1 is a proportionality constant. The remission function can be written by: ½FðR∞Þhn�2 ¼ C2 � hn� Eg � (2) Therefore, using C2 ¼ 1, Eg is obtained from the linear fit of ½FðR∞Þhn�2 versus hn. In Fig. 5(a) and (b) a high reflectivity can be seen from 350 nm up to 800 nm for both compounds, attesting the optical quality of the pure and Eu3þ doped Al6Ge2O13 crystalline powders. The decrease of the reflectance intensity starts ~350 nm and finishes ~250 nm wavelengths. Two low absorption peaks around 395 nm and 460 were attributed to Eu3þ due to 7F0 / 5L6 and 7F0 / 5D2 transitions, as indicated in Fig. 5(b). The insets in Fig. 5(a) and (b) shows the ½FðR∞Þhn�2 vs. hn curves with linear fitting at higher energies. The optical band gap energy (Eg) is the intercept of the fitting with the horizontal line ½FðR∞Þhn�2 ¼ 0. The determined Eg were 4.24(5) eV for pure sample and 4.30(5) eV for Eu3þ doped sample. Both samples can be considered non-conducting. Table 1 Frequency of vibrational modes from spectra in Fig. 2 for both samples in comparison to Pure Al6Ge2O13 (This work) Eu3þ doped Al6Ge2O13 (This w Frequency (cm�1) 475 475 525 525 580 580 712 712 805 805 880 880 945 945 1035 1035 1080 1080 Fig. 6 shows the excitation and emission PL spectra of pure and Eu3þ doped Al6Ge2O13 crystalline powders. Pure Al6Ge2O13 shows a those from Refs. [9,10]. ork) From Ref. [9] From Ref. [10] Assignments 469 O-Ge-O bend and Al-O-Al bend 541 510 Al-O stretch 593 O-Al-O bend 709 660 T-O-T bend, in-plane T is Ge or Al 779 792 Al-O stretch, in-plane 831 Al-O strect, out-of-plane 889 Ge-O stretch, out-of-plane 1035 Ge-O stretch, in-plane 1068 1077 Ge-O stretch, in-plane Fig. 5. Room temperature diffuse reflectance of (a) pure and (b) Eu3þ doped Al6Ge2O13 crystalline powders. Inset is the bandgap (Eg) determined by diffuse reflectance curves. Fig. 6. (a) Excitation photoluminescence spectra of pure and Eu3þ doped Al6Ge2O13 crystalline powders monitoring the emissions at 455 nm and 612 nm, respectively. (b) Emission photoluminescence spectra from pure sample under excitations at 360 nm and 394 nm, and (b) from Eu3þ doped sample under excitation at 394 nm. L.J.Q. Maia et al. / Optical Materials 75 (2018) 297e303 301 broad and low intensity emission band centered at 425 nm (violet) when excited at 360 nm, and under excitation at 394 nm the emission band was centered at 455 nm (blue). The origin of this violet-blue emission can be assigned to intrinsic defects in the host such as oxygen vacancies, and interstitial or metal ion (Ge, Al) va- cancies, which can coexist. Similar broad band occurs also for Eu3þ doped sample; strongly suggesting that is due to intrinsic defects in the Al6Ge2O13 host. Intrinsic defects are present in the similar structure, the mullite, as mentioned by Schneider et al. [18], probably Al6Ge2O13 host have the same defect type. Emission from Eu3þ doped sample at 612 nm attributed to 5D0 / 7F2 transition was observed under excitation at wavelengths between 300 and 600 nm, due to 4f-4f transitions assigned in Fig. 6(a). Under exci- tation at 394 nm, corresponding to Eu3þ 7F0 / 5L6 transition, typical Eu3þemission from the 5D0 to 7F0, 7F1, 7F2, 7F3, and 7F4 levels was measured between 570 nm and 710 nm. The most intense peaks are located at 612 nm (red emission-R) and at 590 nm, (or- ange emission-O) corresponding to the 5D0 / 7F2 and 5D0 / 7F1 transitions respectively. The intensity relation between red and orange emissions (R/O relation) was 2.7, indicating that the euro- pium ions are distributed in relatively high symmetry sites of the host, and the presence of the emission band at 579 nm assigned to the 5D0 / 7F0 attest for the presence in chemical environment without inversion center. In addition, even it was not observed splitting in the 5D0/ 7F0 transition, an inhomogeneous broadening suggests a distribution of Eu3þ ions in more than one symmetry site into the host. Furthermore, the emission spectra profile considering all 5D0 / 7FJ (J ¼ 0, 1, 2, 3, 4) transitions is broadened, which can be attributed to the effect of the distribution of different microenvi- ronment around the Eu3þ ion, i.e., distinct symmetry sites occupied by the Eu3þ ions in the host, producing inhomogeneous broadening due to the superposition of Stark emission levels. As Al6Ge2O13 host possess a structure close to that of mullite phase which is consti- tuted by AlO6 octahedra and AlO4 tetrahedra sites, can lead us to propose that mostly Eu3þ substitutes Al3þ in octahedral sites in Al6Ge2O13 orthorhombic structure. The 5D0 emission decay curve of the Eu3þ ions under excitation at 394 nm, was presented in Fig. 7, and a non-single exponential decay was observed. To estimate the 5D0 excited state lifetime, an averaged lifetime (tave) value was calculated by integrating the decay curve using the following equation, and the value obtained was 1.77 ms. tave ¼ Ztf t0 IðtÞ Iðt0Þ dt (3) Likewise, the emission decaywas fitted by an exponential curve, which the best result was achieved by considering a second order Fig. 7. PL decay curve from the 5D0 level of Eu3þ and its bi-exponential fitting. L.J.Q. Maia et al. / Optical Materials 75 (2018) 297e303302 exponential curve, with lifetime of t1 ¼ 2.12(3) ms and t1 ¼ 0.97(4) ms, and pre-exponential factors A1 ¼ 0.68(3) and A2 ¼ 0.32(3), respectively. This result suggests that at least two main species are contributing to the transition. The first one, corresponding to the longer lifetime possibly is due Eu3þ ions replacing Al3þ sites in the Al6Ge2O13 crystalline structure, which is coherent with presence of Eu3þ in relatively high symmetry site (long as 2.12 ms), probably distorted octahedral site. On the other hand, the fast one can be attributed due to Eu3þ on particle surface and/or influenced by OH groups. The hydroxyl groups (-OH) were previously detected by FTIR results. The long component has higher pre-exponential factor than that calculated for the fast one, demonstrating that the higher and isolated Eu3þ amounts are distributed into the Al6Ge2O13 crystalline structure. As previously presented and discussed by M.H.V. Werts et al. [23] and L.D. Carlos et al. [24] the spontaneous emission probability (A), the radiative lifetime (tRad), quantum efficiency for 5D0 / 7F2 emission, andU2 andU4 parameters can be calculated, as presented below. The pure magnetic-dipole character of the 5D0 / 7F1 transition enable the determination of intensity parameters from the emis- sion spectrum, because this transition does not depend on the local ligand field experimented by Eu3þ ions, that may be used as reference for the entire spectrum [23,24]. The A01 spontaneous decay rate of 5D0 / 7F1 transition is A01 ¼ A0 01n 3, with A0 01 ¼14.65 s�1 in vacuum and n is the refractive index of the host. Then, the intensity of the 5D0 / 7F0-6 transitions in terms of area of emission curves (S0J) is: S0J ¼ hcyA0JNð5D0Þ; (4) where N(5D0) is the 5D0 level population that emits. The total radiative decay rate can be write as: Table 2 Calculated A0J (J¼ 1, 2, 4) and AT radiative decay rates, radiative (tRad) experimental (tExp) l Al6Ge2O13 phase. Sample A01 (s�1) A02 (s�1) A04 (s�1) A 1 mol% Eu3þ doped Al6Ge2O13 (This work) 76.65 212.64 12.99 3 AT ¼ X6 J¼0 A0J ¼ A01hcy01 S01 X6 J¼0 S0J hcy0J (5) The branching ratio for 5D0 / 7F5,6 transitions must be neglected due to its low relative intensity, and the radiative contribution can be calculated using only 5D0 / 7F0-4 transitions [24]. The emission quantum efficiency (q) is defined by the experi- mental and radiative lifetimes ratio: q ¼ tExp tRad (6) The site symmetry and luminescence behavior of Eu3þ ions in Al6Ge2O13 host was carried out by calculations of Judd-Ofelt pa- rameters Ul (l ¼ 2, 4). From Judd-Ofelt theory, the intensity pa- rameters Ul are given by: Ul ¼ 3h 64p4e2y3 9 n � n2 þ 2 �2 1���5D0jUðlÞj7FJj2 A0J (7) The values for reduced matrix elements are 0.0032 of l ¼ J ¼ 2 and 0.0023 of l ¼ J ¼ 4. Table 2 list the determined values for A01, A02, A04, AT, tRad, q(%), U2 and U4. The emission quantum efficiency was calculated using the three tExp values (tav, t1, and t2) for comparison. For the calculation, we consider the n ¼ 1.72 from Ref. [25] for Al6Ge2O13 crystalline phase. The dependence of the radiative lifetime for different hosts originates from the radiation field polarization of the medium and photon density change in an optically dense medium [26]. The oscillator strength of the electric dipole transition for Eu3þ and refractive index are considered high in Al6Ge2O13, then the 5D0 radiative lifetime of Eu3þ in Al6Ge2O13 reduces, i.e., the radiative transition rates are higher, as we can see in Table 2. In this way, the 5D0 quantum efficiency (q in %) is relatively high, with high values of 64% regarding mostly the Eu3þ ions in the crystalline structure. Therefore, when the tave is used, the esti- mated q value is about 54%, and clearly the presence of non- radiative process was considered here. Then we take into account the two experimental lifetimes from decay curve fitted by bi- exponential function, the q values are 64% and 29% for t1 and t2, respectively, which are due to the Eu3þ probably replacing Al3þ of AlO6 octahedra, and Eu3þ ions on particle surface and/or influenced by OH groups. The polarization and asymmetry behavior of the rare-earth li- gands are determined by U2 parameter, whereas the other parameter U4 depend on long range effects [27]. The high U2 value for Eu3þ in Al6Ge2O13 host indicate a high asymmetry nature that is corroborated by the R/O ratio (R ¼ Red emission (5D0 / 7F2 tran- sition), and O ¼ orange emission (5D0 / 7F1 transition)) being of 2.7. The low U4 value imply that 5D0 / 7F2 transition efficiency increases, or be red colour. Certainly, the 5D0 / 7F2 transition is the main emission of Eu3þ indicating almost pure red emission in the material. ifetimes, quantum efficiency (q) andU2 andU4 Judd-Ofelt parameters for Eu3þ doped T (s�1) tRad (ms) tExp (ms) q (%) U2 (10�20 cm2) U4 (10�20 cm2) 02.29 3.31 1.77 (tave) 54 4.54 0.40 2.12 (t1) 64 0.97 (t2) 29 L.J.Q. Maia et al. / Optical Materials 75 (2018) 297e303 303 4. Conclusions In summary, the results indicated that the samples consisted of well crystallized Al6Ge2O13 phase in an orthorhombic structure. A simple sol-gel route was used successfully to obtain pure and Eu3þ doped Al6Ge2O13 particles at relative low temperatures, using germanium oxide, aluminum and europium nitrates as metal sources, citric acid as complexing agent, and tetramethylammo- nium hydroxide, water and ethanol as solvents. The new chemical methodology developed dispenses the use of germanium alcox- ydes, which is commonly reported in the literature to prepare compounds based on germanium from chemical procedure. The structural properties of pure and Eu3þ doped Al6Ge2O13 phase is composed by AlO6, AlO4 and GeO4 groups. The materials possess optical bandgap of 4.24 eV and 4.30 eV for pure and Eu3þ doped, respectively. The Eu 3þ doped materials have high PL emission around 612 nm under excitation at 394 nm. The 5D0 level possess two lifetime values: one ~2.12 ms associated to Eu3þ replacing Al3þ from AlO6 octahedra of Al6Ge2O13 crystalline structure and another ~0.97 ms due to Eu3þ on particle surface and/or influenced by hy- droxyl groups. The quantum efficiency is 64% and 29% for these two Eu3þ different sites (two lifetime values), respectively. An average quantum efficiency is 54% considering the average experimental lifetime of 1.77 ms. The Eu3þ in Al6Ge2O13 host have high U2 value that reveal its high asymmetry nature. Finally, good structural and optical emissions of Eu3þ doped Al6Ge2O13 samples were obtained with potential application in displays as red emitters. Acknowledgments We acknowledge financial support from the Brazilian Agencies: Conselho Nacional de Desenvolvimento Científico e Tecnol�ogico (CNPq), Instituto Nacional de Ciência e Tecnologia de Fotônica (INCT INFO), Coordenaç~ao de Aperfeiçoamento de Pessoal de Nível Supe- rior (CAPES), Fundaç~ao de Amparo �a Pesquisa do Estado de S~ao Paulo (FAPESP), Fundaç~ao de Apoio �a Pesquisa da Universidade Federal de Goi�as (FUNAPE), and Fundaç~ao de Amparo �a Pesquisa do Estado de Goi�as (FAPEG). Bolsista da CAPES/Est�agio Sênior/Processo nº 88881.121134/2016-01. References [1] F.M. Faria Filho, R.R. Gonçalves, S.J.L. Ribeiro, L.J.Q. Maia, Mater. Sci. Eng. B 194 (2015) 21. [2] A. Martucci, G. Brusatin, M. Guglielmi, C. Strohhofer, J. Fick, S. Pelli, J. Sol-Gel Sci. Technol. 13 (1998) 535. [3] G. Nunzi Conti, S. Berneschi, M. Brenci, S. Pelli, S. Sebastiani, G.C. Righini, C. Tosello, A. Chiasera, M. Ferrari, Appl. Phys. Lett. 89 (2006) 121102. [4] R. Reisfeld, E. Berman, M. Zelmer, A. Patra, J. Alloys Compd. D. 300e301 (2000) 147. [5] C. Zhu, A. Monteil, M. El-Jouad, N. Gaumer, S. Chaussedent, Opt. Lett. 34 (2009) 3749. [6] Y. Yan, A.J. Faber, H. Waal, J. Non-Cryst. Solids 181 (1995) 283. [7] W.A. Pisarski, J. Pisarski, W. Ryba-Romanowski, Chem. Phys. Lett. 380 (2003) 604. [8] D. Michel, Ph. Colomban, S. Abolhassani, F. Voyron, Kahn-Harari, J. Eur. Ceram. Soc. 16 (1996) 161. [9] D. Voll, P. Angerer, A. Beran, H. Schneider, Vib. Spectr. 30 (2002) 237. [10] R.H. Meinhold, K.J.D. Mackenzie, J. Mater. Chem. 10 (2000) 701. [11] S. Grandi, P. Mustarelli, S. Agnello, M. Cannas, A. Cannizzo, J. Sol-Gel Sci. Technol. 26 (2003) 915. [12] J.H. Jang, J. Koo, B.S. Bae, J. Non-Cryst. Solids 259 (1999) 144. [13] C. Jing, J. Hou, Y. Zhang, J. Phys. D. Appl. Phys. 39 (2006) 1174. [14] L.J.Q. Maia, R.R. Gonçalves, A.S.L. Gomes, S.J.L. Ribeiro, J. Braz. Chem. Soc. 26 (2) (2015) 2545. [15] R.M. Soares, A.C.S. Sabioni, I.S. Resck, V.A.S. Falcomer, J.A. Dias, N.M. Silva, S.M.C. Menezes, G.M. Costa, Mater. Res. 10 (2007) 75. [16] C.O. Paiva Santos, D. Garcia, Y.P. Mascarenhas, J.A. Eiras, Cerâmica 35 (1989) 153. [17] L. Gao, H. Wang, W.M. Wang, Z.Y. Fu, Chin. J. Inorg. Chem. 23 (2007) 1169. [18] H. Schneider, R.X. Fischer, J. Schreuer, J. Am. Ceram. Soc. 98 (2015) 2948. [19] S. Aryal, P. Rulis, W.-Y. Ching, J. Am. Ceram. Soc. 95 (2012) 2075. [20] R.D. Shannon, Acta Cryst. A 32 (1976) 751. [21] H.V.S. Pessoni, L.J.Q. Maia, A. Franco Jr., Mater, Sci. Semicond. Proc. 30 (2015) 135. [22] B. Hapke, Theory of Reflectance and Emittance Spectroscopy, Cambridge University Press, 2012. [23] M.H.V. Werts, R.T.F. Jukes, J.W. Verhoeven, Phys. Chem. Chem. Phys. 4 (2002) 1542. [24] L.D. Carlos, R.A. S Ferreira, V. Zea Bermudez, S.J.L. Ribeiro, Adv. Mater 21 (2009) 509. [25] M.J. Weber (Ed.), Handbook of Optical Materials, CRC Press, 2003. [26] G. Pan, H. Song, Q. Dai, R. Qin, X. Bai, B. Dong, L. Fan, F. Wang, J. Appl. Phys. 104 (2008) 084910. [27] S. Som, A.K. Kunti, V. Kumar, V. Kumar, S. Dutta, M. Chowdhury, S.K. Sharma, J.J. Terblans, H.C. Swart, J. Appl. Phys. 115 (2014) 193101. http://refhub.elsevier.com/S0925-3467(17)30671-7/sref1 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref1 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref2 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref2 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref3 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref3 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref4 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref4 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref4 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref5 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref5 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref6 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref7 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref7 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref8 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref8 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref9 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref10 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref11 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref11 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref12 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref13 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref14 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref14 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref15 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref15 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref16 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref16 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref16 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref17 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref18 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref19 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref20 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref21 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref21 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref22 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref22 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref23 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref23 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref24 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref24 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref25 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref26 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref26 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref27 http://refhub.elsevier.com/S0925-3467(17)30671-7/sref27 Luminescent Eu3+ doped Al6Ge2O13 crystalline compounds obtained by the sol gel process for photonics 1. Introduction 2. Experimental procedure 2.1. Synthesis 2.2. Characterization 3. Results and discussion 4. Conclusions Acknowledgments References