Distortion induced structural characteristics of Ba2R2/3TeO6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb and Lu) double perovskites and their multifunctional optical properties for lighting and ratiometric temperature sensing

Group theoretical predictions obtained from Raman and IR spectra emphasize the structures of B-site ordered double perovskites Ba2R2/3TeO6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) with ubiquitous BO6 or BO6 octahedral tilting are monoclinic (P21/n), which was further confirmed via the Rietveld refinement of XRD patterns. The wide band gaps of these compounds were consistent with their ability to absorb UV radiation and hence their potential to act as photoluminescent host materials. Photoluminescence spectra of Eu-activated Ba2Y2/3TeO6, irrespective of substitution at Aor B-sites, yielded a dominant magnetic dipole transition at 592 nm and a less intense electric dipole transition at 610 nm when excited at 276 nm suggested the preference of Eu ions to occupy Y sites. The thermal activation energy of the Ba2Y2/3TeO6:Eu 3+ phosphor was 0.15 eV and its relative sensitivity in the temperature range 300–500 K was estimated to have a maximum of 0.18% K 1 at 300 K for FIR measurements, which is better than its decay lifetime counterpart. The red shift and the peak broadening of the highenergy Raman modes at elevated temperatures of Ba2Y2/3TeO6:Eu 3+ are due to enhanced B-site octahedral distortions and are correlated with the thermal quenching of Eu emission. Temperaturedependent lifetimes in the range 80–500 K imply intensification of non-radiative relaxation within the limit. Furthermore, the phosphorescence lifetime, color parameters, and JuddOfelt parameters are explained in detail. The multifunctional optical properties of this phosphor can thus be utilized for technological applications, such as solid-state lighting and ratiometric optical thermometry.


Introduction
Among the different structural families of ceramic compounds, perovskites (ABX 3 ) form the most often encountered materials since they are well known for their technological applications owing to their physical and chemical properties. 1 The ideal cubic perovskite structure has a twelve-coordinated A-site cation, a six-coordinated B-site cation and an anion. The perovskite family can be extended by substituting cations at either A-or B-sites to form complex perovskites or double perovskites with the general formulae A 2 BB 0 X 6 , AA 0 BB 0 X 6 and AA 0 B 2 X 6 , where the anion site is usually occupied by oxygen. [1][2][3][4] Cationic substitution may cause expansion/contraction or tilting of the BX 6 octahedra in order to compensate for the mismatch in ionic radii of the cations and hence, the symmetry of these compounds reduces to tetragonal, monoclinic, trigonal, etc. Amongst the quaternary perovskites, the B-site ordered A 2 BB 0 O 6 arouse particular attention among researchers and have been extensively studied for their intriguing structuredependent properties. Double perovskites with cationic substitution in the B-sites usually adopt 1 : 1 rock salt ordering, where B and B 0 cations alternate in all three dimensions. The possibility of different cations at the B-site allows novel combinations of elements, including transition metals, lanthanides, actinides and some main group elements. The cation substitution may cause structural alterations, which result in symmetry changes from aristotype cubic (Fm% 3m) to hettotypes such as tetragonal (I4/m), rhombohedral (R% 3) or monoclinic (P2 1 /n or I2/m). Moreover, the B-site ordered double perovskites exhibit interesting magnetic, optical, and dielectric as well as superconducting properties. [5][6][7][8][9][10][11] Extensive reviews have been carried out on B-site ordered double perovskites, exploring their structural and physical properties. [1][2][3] Several types of inter-lanthanide perovskites showing remarkable physical properties, for example magnetic and optical properties, which can be attributed to the deeplying 4f electrons of the rare earth ions, were discovered in the 1960s. 12,13 These perovskites also have potential applications in the field of solid oxide fuel cells when doped with bivalent ions. In addition, perovskites containing lanthanides have been probed for their properties and applications. 13 Later, in 2017, Kong and Cava 14 reported the crystal structure and magnetic properties of the double perovskites Ba 2 R 2/3 TeO 6 (R = Y, La, Pr, . . ., Lu). The crystal structure was reported to be cubic. However, our group 15 re-examined the crystal symmetry of five of these double perovskites Ba 2 Ln 2/3 TeO 6 (Ln = La, Pr, Nd, Sm and Eu) using Raman spectroscopy and Rietveld refinement of the XRD patterns and found that the most likely space group is P2 1 /n with monoclinic symmetry. Moreover, their optical properties were also investigated, and a red-emitting phosphor Ba 2 La 2/3 TeO 6 :Eu 3+ was also developed in anticipation of using it in white light emitting diodes.
Phosphor-converted white light emitting diodes (WLEDs) are considered to be the next generation of lighting sources. [16][17][18][19][20][21] However, the deficient emission in the red spectral region results in a low color rendering index (CRI) for these WLEDs. In order to improve the performance of WLEDs, highly efficient and stable red phosphors are usually incorporated. [22][23][24][25][26][27][28][29][30][31][32][33][34] Phosphor-converted red emitting LEDs have been a breakthrough in plant cultivation research, since red and far red light serve important functions in photosynthesis as well as in photomorphogenesis. 35 Optical thermometry based on rare earth ion doped oxides is a promising method to gather large-scale imaging and contactless measurement of temperature compared to other conventional thermometers. 36 Recently, optical thermometry has become a major research interest owing to its fast response time, contactless measurement, high precision and accuracy, high signal yield, etc. Optical temperature sensing is based on measuring the change in emission intensity against temperature when interacting with physical systems. There are several parameters of luminescence, such as intensity, spectrum bandwidth, spectral shift and lifetime, that can be used to calculate the temperature change, among which measuring the fluorescence intensity ratio (FIR) is the most favourable technique. 36 Trivalent rare earth ion doped systems are used for optical thermometric applications due to the abundant 4f energy levels of these ions. Among the different rare earth ions, Eu 3+ ions, owing to their thermally coupled energy levels (TCLs) 5 D 1 and 5 D 0 , are also used for optical thermometry applications. 37 Therefore, it is essential to synthesize and characterize novel red phosphors that can be used in various technological applications.
Rare earth doped red phosphors require a suitable host matrix that has good absorbance in the near UV region and excellent thermal and chemical stability. Double perovskites are among such host materials for luminescent centers or socalled activator ions. Red phosphors based on rare earth substituted complex perovskites are widely reported for lighting applications. [38][39][40][41][42][43][44][45][46][47] In addition, several double perovskite phosphors, Gd 2 ZnTiO 6 :Pr 3+ , Y 2 MgTiO 6 :Mn 4+ , LiLaMgWO 6 :Er 3+ , etc., have been explored for their temperature-sensing potential. [45][46][47] In the present work, we successfully synthesized Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) double perovskites and their crystal structure is examined by XRD, Raman and FTIR spectroscopy with B-site substitution of rare earth ions. The previously reported Ba 2 Ln 2/3 TeO 6 (Ln = La, Pr, Nd, Sm and Eu) show no symmetry variation when the B-site is substituted with La 3+ to Eu 3+ . However, the total number of modes in the Raman spectra decrease from 21 to 16 when the B-site changes from La 3+ to Eu 3+ . In addition to this, the tilting or distortion of the BO 6 and B 0 O 6 octahedra also reduces and the B-O-B 0 (B = La, Pr, Nd, Sm, Eu; B 0 = Te) bond angle approaches 1801 with B-site variation. 15 This gives an intuition of a possible change in symmetry for Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) from monoclinic to cubic when the B-site undergoes substitution from Gd 3+ to Lu 3+ . The photoluminescence properties of both A-and B-site substituted Ba 2 Y 2/3 TeO 6 :Eu 3+ , temperature-dependent luminescence, optical thermometric parameters, photoluminescence decay curves and Judd-Ofelt theory are discussed in detail. This is carried out on the assumption that, unlike Ba 2 La 2/3 TeO 6 :Eu 3+ which was previously reported, Ba 2 Y 2/3 TeO 6 :Eu 3+ shows enhanced luminescence performance as yttrium-based host matrices like Y 2 O 2 S:Eu 2+ have been established as potential red phosphors that can be excited using ultraviolet (UV) radiation due to their wide band gap characteristics (4.6-4.8 eV) and hence utilized in the field of light emitting diode applications. Characterization techniques X-ray diffractograms were collected using a Bruker D8 Advance diffractometer (40 kV, 40 mA) equipped with a Cu-Ka source (l = 1.5406 Å), a nickel filter and a Lynx eye position sensitive detector. Topas 4.2 software was used to perform the Rietveld refinement of the XRD patterns. A Chebychev polynomial was used to fit the background. Cell parameters, atomic coordinates, occupancy and isotropic temperature factors were refined. An Ametek EDAX -Octane Plus system was used to carry out EDS analysis. Room temperature micro Raman scattering spectra were recorded using a LABRAM HR-800 Raman spectrometer with a 633 nm He-Ne laser (5 mW) with a resolution of 1 cm À1 . The FTIR spectra of the samples in the far-IR and mid-IR range were recorded by the ATR method in vacuum using a JASCO FTIR 6800 spectrometer. In the mid-IR region (400-1000 cm À1 ), a Ge-KBr beam splitter and triglycerine sulfate (TGS) detector and in the far-IR region (50-630 cm À1 ), a Myla beam splitter and a polyethylene triglycerine sulfate (PE-TGS) detector were used. The diffuse reflectance spectra were obtained in the range 190-900 nm using a JASCO V-750 UV-Vis spectrophotometer. A Fluorolog spectrofluorometer with a 450 W xenon lamp was used to record the photoluminescence excitation (PLE) and emission (PL) spectra (slit width 1 nm, integration time 0.5 s). The temperature-dependent photoluminescence and lifetime measurements were collected using a cryostat designed by Oxford Instruments attached with a FLS 1000 spectrophotometer from Edinburgh Instruments in a temperature range of 80-500 K. Decay curves were recorded with a flash lamp frequency of 25 Hz with a time range of 40 ms and maximum peak counts of 10 000. Temperature-dependent Raman spectra were recorded using a Horiba Jobin Yvon LABRAM HR-evolution Raman spectrometer with a 785 nm high-power single-frequency diode laser (90 mW). Quantum yield was measured with the aid of a Quanta-j integrating sphere set-up associated with a Fluorolog spectrofluorometer using a 450 W xenon lamp.

Results and discussion
XRD patterns of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) X-ray diffraction technique was used to analyse the crystallinity and phase purity of all the calcined powders. Fig. 1a shows the XRD patterns of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) indexed with pseudocubic lattice planes. XRD patterns of Ba 2 Gd 2/3 TeO 6 and Ba 2 Lu 2/3 TeO 6 are shown separately in Fig. S1 and S2 (ESI †). For double perovskites, if the charge difference is Z2 and the size difference is greater than 0.1 Å, then the most favourable ordering among the B-site cations is rocksalt ordering. For Ba 2 R 2/3 TeO 6 , the charge difference between R 3+ and Te 6+ is +3 and the size difference is greater than 0.1 Å, denoting a possible rocksalt ordering of B-site cations. The characteristic R point 111 reflection around 2y E 181 also indicates the 1 : 1 rocksalt ordering of these compounds. The splitting of reflections at 440 planes in Ba 2 Lu 2/3 TeO 6 and Ba 2 Gd 2/3 TeO 6 shown in Fig. 1b and c points to the existence of a possible octahedral distortion and hence a lower symmetry instead of the aristotype cubic (Fm% 3m) structure. Another parameter that defines the extent of distortion is the tolerance factor, which is equal to one for an ideal cubic structure and is less than/greater than unity for lower symmetries. A tolerance factor that deviates from unity also suggests the structural variation of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) from cubic (Fm% 3m) structure. The tolerance factor is determined by considering the effective size of vacancies the same as that of the B-site cations (R 3+ ) and varies from 0.9904 to 1.0085 when the B-site is substituted with Gd 3+ to Lu 3+ . This implies a cubic symmetry for these double perovskites since the tolerance factor lies within the limits defined for a cubic structure. In addition to this, our earlier report suggests some evidence for an increase in symmetry when going from La 3+ to Eu 3+ in the B-site of Ba 2 Ln 2/3 TeO 6 (Ln = La, Pr, Nd, Sm, Eu). The tolerance factors are given in Table S3 (ESI †). Interestingly, the tolerance factor exceeds as well as falls short of unity, when the B-site is occupied by R 3+ ions (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu). However, these striking values of the tolerance factor may be due to the assumption of taking the effective size of vacancies (concentration E33%) and lanthanide ions to be the same. There is a possibility that the effective size of vacancies gets larger along with lanthanide contraction and hence compensates for the reduction in ionic size, which is the reason for obtaining a tolerance factor greater than one. Nonetheless, the tolerance factor is only a rough estimate since it varies in accordance with the value of ionic radii taken for determining it and in this case the tolerance factor suggests the possibility of a structural transition to cubic when Ba 2 R 2/3 TeO 6 is substituted from Gd 3+ to Lu 3+ . In summary, we are expecting a symmetry change from XRD analysis which could be more easily identifiable from vibrational spectroscopic techniques such Raman and IR spectroscopy.
The EDS indicate that Ba 2 R 2/3 TeO 6 (R = Gd, Dy, Y, Er, Lu) have an elemental composition of Ba, Te, O, and the rare earths Gd, Dy, Y, Er and Lu. The obvious adventitious carbon peak is due to surface contamination. Elemental mapping shows that all the elements are distributed uniformly in the compounds. Table S8 (ESI †) gives the atomic percentage of elements obtained from EDS.
Raman and infrared spectra of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) The room-temperature Raman spectra shown in Fig. 3a exhibit similar spectral profiles imply same symmetry for Ba 2 R 2/3 TeO 6 . For an ideal double perovskite with aristotype cubic Fm% 3m symmetry, four Raman active modes are possible. However, more than four modes are observed in the Raman spectra of Ba 2 R 2/3 TeO 6 similar to Ba 2 Ln 2/3 TeO 6 (Ln = La, Pr, Nd, Sm and Eu) evidently point to a lower symmetry for these double perovskites. The exact number of modes is determined by deconvoluting the spectra using the Lorentzian function and the bands obtained are assigned in Table 1. The deconvoluted Raman spectra of Ba 2 Lu 2/3 TeO 6 and Ba 2 Gd 2/3 TeO 6 are shown in Fig. 3b and c, respectively, and those of Ba 2 R 2/3 TeO 6 (R = Y, Tb, Dy, Ho, Er, Tm, Yb) are shown in Fig. S9-S15 (ESI †). The total number of Raman modes for all these double perovskites varies from 12 to 15 in the range 100-1000 cm À1 and are lower than previously reported Ba 2 Ln 2/3 TeO 6 (Ln = La, Pr, Nd, Sm and Eu). Eventhough, we expected a gradual decrease in the number of modes with lanthanide contraction of rare earth ions, as shown in Table 1, they do not follow a trend. For example, in Ba 2 Gd 2/3 TeO 6 , the number of Raman modes is 15 and decreases to 12 for Ba 2 Ho 2/3 TeO 6 . Again, for Ba 2 Y 2/3 TeO 6 , the number of modes increases to 15. This might be due to the presence of sparse amount of impurities that could not be identified from the XRD patterns. Though XRD patterns signify the phase purity of all the compounds, Raman spectroscopy is able to find very small traces of impurities which do not even affect the crystallography of these compounds, since this technique is more sensitive than XRD. The addition of impurities lighter or heavier than the host atoms gives rise to new vibrational modes of higher or lower frequencies, respectively, in the Raman and IR spectra. The n 1 modes due to the symmetric oxygen stretching vibrations are located in the region 750-800 cm À1 and the n 2 modes due to the asymmetric oxygen stretching vibrations of the octahedra are located in the region 550-740 cm À1 , whereas the vibrational bands due to the oxygen bending motion of the octahedra are observed in the region 350-510 cm À1 . The lowenergy external modes in the range 100-350 cm À1 are due to translational and liberational modes, which are sensitive to lowering of symmetry due to octahedral tilting. Octahedral distortions are evident from the presence of low-energy modes. For modes above 650 cm À1 , a significant shift to the longer wavenumber region (Fig. S16, ESI †) is noticed with a decrease in the ionic radii of the B-site cation from Gd 3+ to Lu 3+ . As the unit cell gets contracted with the substitution of less massive ions, the ionic bonds become stronger and thus more energy is needed for the vibrations of the molecules.
In addition, this space group is ascribed to double perovskites with Jahn-Teller ions like Mn 3+ , Ni 3+ or Cu 3+ as in the case of Sr 2 MnWO 6 48 and Sr 2 MnMoO 6 . 49 The space groups P4/mnc has 14 Raman active modes and C2/c has 27 Raman active modes. However, these space groups can be ruled out since the Rietveld refinement parameters are not reliable compared to the space group P2 1 /n. Thus, in accordance with the previously reported evidences and confirmation of the symmetry of Ba 2 Ln 2/3 TeO 6 (Ln = La, Pr, Nd, Sm and Eu), the most likely symmetry of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) is found to be monoclinic (P2 1 /n). The monoclinic (P2 1 /n) symmetry is a combination of in-phase and anti-phase tilting of the BO 6 and B'O 6 octahedra around the 001 axes of the ideal cubic structure. The irreducible representation of P2 1 /n space group is given by G = T(3A g + 3B g ) + L(3A g + 3B g ) + n 1 (A g + B g ) + n 2 (2A g + 2B g ) Thus, there are 24 Raman active modes possible for a double perovskite that crystallizes in monoclinic (P2 1 /n) symmetry. The absence of other bands predicted by group theory might be due to the polycrystallinity of the compounds, the weak bands which could not be identified among the background noise and some modes may fall outside the range of the spectrum. The IR transmittance spectra obtained using the ATR method are shown in Fig. 4. The spectra look almost identical and their visual inspection shows six well-defined modes. A careful inspection of the IR spectra indicates the presence of additional bands and the exact number of modes can only be identified by fitting the IR reflectivity spectra. However, group theory predicts 33 infrared active modes for a monoclinic structure belonging to the P2 1 /n space group. Strong vibration coupling exists between the BO 6 and B 0 O 6 octahedral moieties. This simple argument proves that the isolated octahedral groupings will dominate the vibrations in infrared as well as in Raman spectroscopy. Among the 6 well-defined modes, two modes around 340 cm À1 and 605 cm À1 are dominant, which can be assigned to the deformational modes and the antisymmetric stretching vibrations of the BO 6 octahedral moieties, respectively. All vibrational modes located in the ranges 150-175 cm À1 , 180-220 cm À1 , 325-350 cm À1 , 450-500 cm À1 , 592-610 cm À1 and 660-680 cm À1 essentially come from the vibrations of RO 6 and TeO 6 octahedra. Similar to Raman spectra, the modes above 590 cm À1 show a spectral shift towards the high wavenumber region with lanthanide contraction. The translational symmetry elements, such as glide planes and screw axes, give rise to systematic absences/extinctions of some (hkl) reflections in the XRD patterns. The systematic absences of (hkl) reflections, such as (h0l) with h + l = 2n + 1 and (0k0) with k = 2n + 1, in the XRD patterns implies the monoclinic symmetry (P2 1 /n) of the crystals. The presence of the (330) reflection for atomic displacement and (204) and (404) reflections for in-phase tilting also suggest monoclinic symmetry. The refined XRD patterns of Ba 2 Gd 2/3 TeO 6 and Ba 2 Lu 2/3 TeO 6 are shown in Fig. 5a and b. The Rietveld refinement parameters of Ba 2 Gd 2/3 TeO 6 and Ba 2 Lu 2/3 TeO 6 are given in Table 2. From the refined results, b is close to 901, suggesting the possibility of a structural transition to orthorhombic along with the lanthanide contraction. Among the 12 different space groups predicted by Howard, only pnnn belongs to orthorhombic. Very few reports have been published about ordered double perovskites with octahedral tilting belonging to the space group pnnn. This space group does not permit the splitting of (hhh) reflections which we could not observe in XRD patterns of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) and hence rule out orthorhombic structure. However, in general for compounds crystallizing in P2 1 /n symmetry, the value of b is larger than 901 if the size of at least one cation in the B-site is larger such as the rare earth ions in this case. These large cations can afford the distortions caused by the neighboring octahedral environment better than smaller cations with higher charges. Thus, it is plausible to conclude that the structure of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) double perovskites with a larger difference in ionic radii could be a distinct monoclinic unit cell rather than orthorhombic. To sum up, the monoclinic crystal structure derived from Raman and IR spectra is confirmed from the Rietveld refinement of the XRD patterns. The refined XRD patterns and parameters of Ba 2 R 2/3 TeO 6 (R = Y, Tb, Dy, Ho, Er, Tm, Yb) are given in  Table S24 (ESI †). Lattice parameters as well as unit cell volume are decreased along with lanthanide contraction of the rare earth ions.
In this system, Ba 2+ and O 2À occupy the Wyckoff position 4e, R 3+ and Te 6+ occupy the Wyckoff positions 2d and 2c respectively. Occupancy is fixed at unity except for R 3+ which has only 2/3rd occupancy of the atomic site. The satisfactory R factors, R wp , R P , R exp and goodness of fit confirm the P2 1 /n space group of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu). Selected bond distances of Ba 2 Gd 2/3 TeO 6 and Ba 2 Lu 2/3 TeO 6 are shown in Table 3 and bond distances of Ba 2 R 2/3 TeO 6 (R = Y, Tb, Dy, Ho, Er, Tm, Yb) are shown in Table S25 ( Diffuse reflectance spectra of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) The optical band gaps of all the synthesized compounds are determined using diffuse reflectance spectroscopy (DRS) in order to discover their potential to act as host materials for luminescence centers. The band gaps are determined using the where F(R N ) is the Kubelka-Munk function, hn is the energy in eV and E g is the optical band gap. Similar to Ba 2 Ln 2/3 TeO 6 (Ln = La, Pr, Nd, Sm and Eu), all the compounds exhibit a direct band transition. Fig. 6 shows the diffuse reflectance spectra (Kubelka-Munk method) of Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu). The obtained band gaps are shown in the inset to Fig. 6 Fig. 7 and the refinement parameters are given in Table S29 (ESI †). In order to confirm the occupancy of activator ions in the Y 3+ site and check the possibility of the crystallographic site swapping to the Ba 2+ site, Eu 3+ ions are distributed at different occupancies in both A-and B-sites in accordance with their stoichiometry, which results in poor refinement parameters. However, it is reported that such swapping cannot be detected directly from XRD at lower concentrations; only the Rietveld refinement of XRD patterns of higher concentrations will give a feasible result. 52 On the other hand, the relative intensity of electric and magnetic dipole transitions obtained from photoluminescence spectroscopy is a probe to detect any crystallographic off-centering of the Eu 3+ ions. The EDS of Ba 2 Y 2/3 TeO 6 :10%Eu 3+ have an elemental composition of Ba, Te, O, and the rare earths Y and Eu. The elemental mapping shown in Fig. S30 (ESI †) exhibits a uniform distribution of elements in the phosphor. Table S8 (ESI †) gives the atomic percentage of elements obtained from EDS. DRS of Ba 2 Y 2/3 TeO 6 :x%Eu 3+ (x = 5, 10, 15 mol%) and their Kubelka-Munk plots for finding the optical band gap are shown in Fig. S31(a) and (b) (ESI †). The reflectance spectra present absorption peaks at 394 nm and 464 nm corresponding to the characteristic f-f transitions of the Eu 3+ ions. As the Eu 3+ concentration increases, the band gap reduces from 4.9 eV of the host Ba 2 Y 2/3 TeO 6 to 4.3 eV of Ba 2 Y 2/3 TeO 6 :15%Eu 3+ possibly due to the rise in density of the 4f band gap states within the forbidden band with Eu 3+ content.
The band gap energy of 4.9 eV of Ba 2 Y 2/3 TeO 6 which corresponds to the wavelength of absorption in the UV region suggests the ability of this material to act as a host material for luminescent centres. Fig. 8a and b show the photoluminescence excitation (PLE) spectra of Ba 2 Y 2/3 TeO 6 :Eu 3+ and the  emission spectra (PL) of Ba 2 Y 2/3 TeO 6 :x% Eu 3+ (x = 2.5, 5, 7.5 10, 12.5, 15, 17.5 and 20). The photoluminescence excitation spectrum monitored at 592 nm presents a broad band ranging from 230 to 340 nm attributed to the charge transfer states of EuO 6 groups and the narrow peaks at 362, 382, 394 and 464 nm can be assigned to 7 F 0 -5 D 4 , 7 F 0 -5 G 2,3 , 7 F 0 -5 L 6 and 7 F 0 -5 D 2 transitions, respectively, which correspond to the characteristic f-f electronic transitions in the Eu 3+ ions. The luminescence spectra present emissions at 592 ( 5 D 0 -7 F 1 ), 610 ( 5 D 0 -7 F 2 ), 635 ( 5 D 0 -7 F 3 ) and 710 ( 5 D 0 -7 F 4 ) nm in which the magnetic dipole transition ( 5 D 0 -7 F 1 ) is the most intense. The emission spectrum with high spectral resolution in the 570-590 nm spectral region is also shown in Fig. S32 (ESI †). The peak at 582 nm corresponds to the 5 D 0 -7 F 0 transition of the Eu 3+ ions. The energy level diagram of Ba 2 Y 2/3 TeO 6 :Eu 3+ is shown in Fig. S33 (ESI †). When excited at 276 nm, ligand-to-metal charge transition occurs from the O 2À 2p orbitals from the valence band to the empty 4f orbitals of Eu 3+ . The absorbed energy is transferred to Eu 3+ ions and the electrons from the highest energy levels relaxes non-radiatively to the 5 D 0 state. From there, electrons further transit to 7 F J ground levels giving visible emission. It can be seen that the photoluminescence emission intensity increases with Eu 3+ ions concentration but decreases when Eu 3+ concentration reaches 12.5 mol%. As the doping concentration increases, the electric multipolar interactions between Eu 3+ causes energy migration among them. This non-radiative energy transfer among the activator ions increases with doping concentration and correspondingly the emission intensity decreases. The critical distance (R c ) of energy transfer among the activator ions can be determined using the equation: 53 where V is the volume of the unit cell, x c is the concentration of the activator ions beyond which quenching happens and N is the number of Y 3+ ions in the unit cell. Substituting the values for V, N and x c as 295 Å 3 , 6 and 0.1 respectively, the critical distance is estimated to be 9.79 Å. The value of R c greater than 4 Å implies electric multipolar interaction among the activator ions. Based on Dexter's theory, if the interaction among the activator ions is multipolar, then it can be quantitatively explained using the equation 54 where I is the luminescence intensity, x is the activator ion concentration after quenching, k and b are constants for the different types of interactions for the same system and y represents the type of interaction between the activator ions. The relative intensity of the electric (610 nm) and magnetic (592 nm) dipole transition is a strong probe to determine the site symmetry of activator ions. Thus, hypersensitivity or the strong influence of the ligand field on the activator ions and hence on the emission intensity gives an idea about the site symmetry in which the ions remain. The dominant magnetic dipole transition in the PL spectra indicates that Eu 3+ ions occupy a site with an inversion center or B-site. However, the emission at 610 nm denotes the occupancy of some of the Eu 3+ ions in a non-center of symmetric environment or A-site of the lattice. Further, to investigate the photoluminescence properties of A-site substituted Ba 2 Y 2/3 TeO 6 , two different concentrations, a concentration lower than optimal concentration (5 mol%) and optimal concentration (10 mol%) for B-site substituted Ba 2 Y 2/3 TeO 6 :Eu 3+ , are substituted in the A-site of the lattice. The combined XRD patterns of A-site substituted Ba 2 Y 2/3 TeO 6 are shown in Fig. S35 (ESI †). The XRD patterns show an indistinct additional impurity phase around 2y E 301. Interestingly the same luminescence spectrum is observed with more intense magnetic transitions as well as electric dipole transitions. The fluorescence intensity ratio between the electric ( 5 D 0 -7 F 2 ) and magnetic dipole ( 5 D 0 -7 F 1 ) transitions defined as the asymmetry ratio, is a measure of the distortion of Eu 3+ ions from the symmetric environment of the host matrix. We expected a slightly larger asymmetry ratios for A-site substituted Ba 2 Y 2/3 TeO 6 :Eu 3+ , since in that case the probability of occurrence of electric dipole transitions is greater. Here the asymmetry ratio for Ba 1.85 Y 2/3 Eu 0.1 TeO 6 , Ba 1.925 Y 2/3 Eu 0.05 TeO 6 , Ba 2 Y 0.567 Eu 0.1 TeO 6 and Ba 2 Y 0.517 Eu 0.05 TeO 6 are obtained as 0.46, which implies Eu 3+ ions prefer to occupy B-site rather than A-site even if the substitution is done on the A-site. This is in contradiction to the photoluminescence behavior of other double perovskites, which show a dominant electric dipole transition when the Eu 3+ ions are substituted in an asymmetric site. Fig. 9 shows a comparison of PL spectra of A-and B-site substituted Ba 2 Y 2/3 TeO 6 :Eu 3+ . Raman spectra of A-and B-site substituted Ba 2 Y 2/3 TeO 6 :Eu 3+ are shown in Fig. S36 (ESI †). The Raman spectrum does not exhibit any considerable changes in the low-energy lattice modes when Eu 3+ is substituted in the A-site, which further confirms the above result.

Temperature-dependent photoluminescence
Temperature-dependent photoluminescence spectra of Ba 2 Y 2/3 TeO 6 :10%Eu 3+ in the range 80-500 K are shown in Fig. 10a. Below 300 K, the emission intensity almost maintains a steady value with no remarkable variation. However, above 300 K, the intensity of photoluminescence shows an appreciable change with rise in temperature. It is essential to probe the thermal stability of phosphors above room temperature, because temperature can influence the color stability and emission intensity, which is a major concern during fabrication of LEDs since they usually work above room temperature. At 480 K, the maximum intensity reaches 41% of the emission intensity at room temperature. Thermal quenching is attained at 400 K and the effect can be better understood by determining the thermal activation energy. The thermal activation energy, DE is determined from the Arrhenius equation where I 0 is the original intensity, I T is the intensity at a temperature T, C is a constant, k is the Boltzmann constant (8.617 Â 10 À5 eV K À1 ) and DE is the activation energy. A plot of 1/kT vs. ln(I 0 /I T À 1) is given in Fig. 10b and the activation energy, DE is obtained as 0.15 eV. The exact mechanism of thermal quenching is given in Fig. S37 (ESI †). Thermal activation energy is defined as the energy required by the electrons to cross the 5 D 0 excited state to the Eu-O CTS and a larger value of DE indicates good thermal stability of the phosphor. A lower activation energy indicates faster degradation of emission intensity of the phosphor. Generally, as temperature increases, phonon-assisted nonradiative transitions become significant and will suppress the luminescence intensity. However, in case of tellurate glasses, the phonon energy is very low, in the range of 800-900 cm À1 , which suggests that Eu 3+ in tellurate host materials are less affected by phonon-assisted non-radiative transitions. 55 Hence, we expect Ba 2 Y 2/3 TeO 6 activated with Eu 3+ ions to exhibit exceptional photoluminescence behaviour and to be a potential candidate for solid-state lighting device applications. Even though its spectral characteristics are good enough for LED applications, unlike other tellurate phosphors, its thermal activation energy is slightly low. Thermal activation energies of various double perovskite tellurates are shown in Table 4.
Temperature-dependent Raman spectra of Ba 2 Y 2/3 TeO 6 :10%Eu 3+ In order to analyze the dependence of thermal quenching of Eu 3+ ions on the structural changes in the host matrix, the temperature-dependent Raman spectra of Ba 2 Y 2/3 TeO 6 :10% Eu 3+ are recorded in a temperature range of 93 K to 873 K. The spectral profile of the Raman spectra of phosphor show no evidence for phase transitions above or below room temperature. Fig. S38 (ESI †) shows the temperature-dependent Raman spectra of Ba 2 Y 2/3 TeO 6 :10% Eu 3+ . The Raman bands around 730-820 cm À1 above room temperature corresponding to the symmetric stretching vibrations of the B-site octahedral cations are shown in Fig. S39 (ESI †). The band near 770 cm À1 at room temperature undergoes a red shift and broadens at elevated temperature, which implies that more vibrations occur even at low energies. The distortion of a B-site octahedra as well as tilting get enhanced at higher temperatures. Hence, distortions also affect the B-site Eu 3+ ions, which further results in faster degradation of the Eu 3+ emission due to the thermally stimulated non-radiative energy transfer from Eu 3+ to the distortion defects. 59 In brief, the peak broadening and red shift above room temperature occur due to the temperature effects. The peak position as well as FWHM values are plotted in Fig. 11.

Temperature-dependent PL decay curves
The temperature-dependent PL lifetime decay curves in the range 80-500 K are obtained by exciting the phosphor Ba 2 Y 2/3 TeO 6 :10%Eu 3+ using 276 nm with l em = 592 nm. Fig. 12a and b show the low-temperature and high-temperature behavior of the decay curves, respectively. The decay curves are well fitted using a double exponential function given by I(t) = A 1 exp(Àt/t 1 ) + A 2 exp(Àt/t 2 ) + I 0 (6) where I 0 is the emission intensity at time t = 0 ms and I(t) is the intensity at t = t ms. A 1 and A 2 are both constants, and t 1 and t 2 are the decay components. The average lifetimes at various temperatures can be determined using the fitting parameters obtained from eqn (6) and substituted in the following equation The lifetime falls off from 4.826 AE 0.259 ms at 80 K to 2.906 AE 0.139 ms at 500 K due to the intensification of non-radiative relaxation rate with a rise in temperature, which in turn decreases the decay rate. Usually, the 5 D 0 measured lifetimes of Eu 3+ are dominated by non-radiative relaxation mechanisms, such as multiphonon emission, direct or indirect phonon assisted energy transfer to the impurity traps or crossover to a low-lying ligand to the metal CT state. 60 Even though nonradiative relaxations are predominant at low temperatures, their effects become more evident at high temperatures in the form of a rapid decrease in lifetime. Fig. S Table 4. Even though the quantum yield of Ba 2 Y 2/3 TeO 6 :Eu 3+ is comparatively lower than that of other tellurates, it is much better than those of commercially available Y 3+ -based red phosphors used in WLED applications, as shown in Table 4. Chromaticity coordinates obtained as (0.615, 0.384) imply that the phosphor exhibits an orange-red color, as shown in the inset of Fig. 12b    where n = (x À x e )/(y À y e ). (x, y) are the chromaticity coordinates and (x e , y e ) are (0.3320, 0.1858), the coordinates corresponding to the epicenter of convergence of the isotemperature lines of the CIE 1931 chromaticity diagram. CCT is calculated as 1657 K, which indicates warmer yellow-red light.
Optical thermometric properties of Ba 2 Y 2/3 TeO 6 :10%Eu 3+ Optical thermometers using various phosphors explore certain parameters of luminescence like the fluorescence intensity ratio (FIR), lifetime, spectral bandwidth, etc. for calculating temperature sensitivity. The intensity ratio, 5 D 0 -7 F 1 / 5 D 0 -7 F 2 , and the emission lifetimes of Eu 3+ ions in the temperature range 300-500 K are used to probe the sensitivity of the Ba 2 Y 2/3 TeO 6 :10%Eu 3+ phosphor. Hou et al. 37 proposed a new temperature-sensing formula for evaluating the FIR of Eu 3+doped Ba 2 TiGe 2 O 8 , which is equivalent to the general fluorescence intensity ratio formula FIR = B Â exp(ÀDE/k B T). The new FIR is the ratio of 5 D 0 -7 F 1 and 5 D 0 -7 F 2 transitions. A plot of temperature vs. FIR is shown in Fig. S41 (ESI †). The relative sensitivity is defined as The relative sensitivity of the Ba 2 Y 2/3 TeO 6 :10%Eu 3+ phosphor is determined using the TeSen calculator, developed by Kaczmarek et al. 62 for determining thermometric parameters in ratiometric optical thermometry. A plot of temperature vs. relative sensitivity obtained from FIR is given in Fig. 13a. The maximum relative sensor sensitivity is 0.18% K À1 at 300 K and it reduces to 0.06% K À1 at 500 K. The dependence of lifetime on temperature can be fitted using the Struck and Fonger model given by 63 where t(T) is the lifetime at temperature T, A is a constant, B is the frequency factor, k is the Boltzmann constant and DE is the thermal activation energy. Temperature vs. lifetime data was plotted and is shown in Fig. S42 (ESI †). The relative sensitivity, S r is therefore given by A plot of temperature vs. relative sensitivity obtained from lifetime is given in Fig. 13b. The maximum sensitivity is 0.136% K À1 at 500 K and it reduces to 0.100% K À1 at 300 K. From a comparison of both results, the temperature sensor sensitivity for FIR is better than the lifetime measurement. A comparison of relative sensitivities of different phosphors  is tabulated in Table 5 and the relative sensitivity of Ba 2 Y 2/3 TeO 6 :10% Eu 3+ indicates the ability of this phosphor to be used for temperature-sensing applications in the range 300-500 K. Thus, Ba 2 Y 2/3 TeO 6 :Eu 3+ phosphor can be used in non-contact optical thermometers.

Judd-Ofelt analysis
Further analysis of the site symmetry modifications of Eu 3+ ions are investigated by determining the Judd-Ofelt intensity parameters O l (l = 2, 4). [64][65][66][67][68][69] The O 2 parameter is an indicator of site asymmetry, whereas the O 4 parameter is sensitive to macroscopic properties, such as rigidity of the host materials, viscosity and dielectric constant. In this work, the intensity parameters were calculated from the emission spectrum, taking the 5 D 0 -7 F 1 transition as the reference. The Judd-Ofelt intensity parameters and radiative parameters are presented in Table 6. The low value of the J-O parameter O 2 suggests that most of the Eu 3+ ions occupy a site with inversion symmetry, which is in good agreement with the PL spectrum. In addition, the 5 D 0 -7 F 1 transition contributes almost 71% of the total radiative transition. The phosphor has a maximum internal quantum efficiency of 42.86%, which suggests the potential application of these phosphors in various technological applications, including WLEDs and optical thermometers. A detailed explanation of the Judd-Ofelt analysis is given in the ESI. †

Conclusions
In summary, Ba 2 R 2/3 TeO 6 (R = Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) were synthesized via a solid-state ceramic route, and their crystal structures and optical properties were investigated for the first time. Even though we expected a structural transition from monoclinic to cubic when the B-site ions varied from Gd 3+ to Lu 3+ , Raman and IR spectra presented more than four modes which contradicts our assumption. The Raman spectra were deconvoluted using the Lorentzian function and the space group was identified as P2 1 /n, which was further confirmed by the Rietveld refinement of the XRD patterns. The optical band gaps suggested that the compounds can act as host materials for luminescence centers, and their photoluminescence properties were broadly investigated via substituting Eu 3+ in the B-site of Ba 2 Y 2/3 TeO 6 . The luminescence spectra implied the presence of Eu 3+ ions at both A-and B-sites but they were mostly present at the B-sites. Eu 3+ ions substituted at the A-site of Ba 2 Y 2/3 TeO 6 exhibited similar PL spectra with an asymmetry ratio almost equal to that of B-site-substituted Ba 2 Y 2/3 TeO 6 . The thermal activation energy and the color parameters indicated the potential application of the phosphor in the field of solidstate lighting. Temperature-dependent Raman spectra of the phosphor with an optimal concentration suggested the dependence of structural distortions at elevated temperatures on thermal quenching. The lifetime decreases from 4.826 AE 0.259 ms at 80 K to 2.906 AE 0.139 ms at 500 K due to the intensification of non-radiative relaxation rate with a rise in temperature. Additionally, the phosphor shows a relative temperature sensitivity of 0.18% K À1 at 300 K for FIR measurements, which is better than for lifetime measurements, and hence it could also find applications in ratiometric optical temperature sensing. The PL spectrum and J-O parameters together suggest the occupation of symmetric B-sites by the majority of Eu 3+ ions in the host lattice. In brief, Ba 2 Y 2/3 TeO 6 :Eu 3+ is the first double perovskite phosphor with defects that shows distinct photoluminescence behavior as well as temperature-sensing potential. The substitution of Eu 3+ ions as well other rare-earth ions into Ba 2 R 2/3 TeO 6 (R = Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) double perovskites might provide better photoluminescence properties.

Conflicts of interest
There are no conflicts to declare.  Table 6 Judd-Ofelt intensity parameters and radiative parameters obtained from the emission spectrum and lifetime data Judd-Ofelt intensity parameters O 2 (Â10 À20 cm 2 ) 0.26 AE 10 À4 O 4 (Â10 À20 cm 2 ) 0.04 AE 10 À4