Observation of intermediate bands in Eu 3 + doped YPO 4 host : Li + ion effect and blue to pink light emitter

Related Articles Beam profile indicator for swift heavy ions using phosphor afterglow AIP Advances 2, 032116 (2012) Analysis of the radiative lifetime of Pr3+ d-f emission J. Appl. Phys. 112, 013536 (2012) Mn-activated K2ZrF6 and Na2ZrF6 phosphors: Sharp red and oscillatory blue-green emissions J. Appl. Phys. 112, 013506 (2012) Photoluminescence investigation of the indirect band gap and shallow impurities in icosahedral B12As2 J. Appl. Phys. 112, 013508 (2012) Strong blue light emission from a-SiNx:O films via localized surface plasmon enhancement Appl. Phys. Lett. 101, 013106 (2012)


I. INTRODUCTION
2][3][4][5][6][7][8] The great interest in doping of trivalent lanthanide ions in host arises from their unique intra-configurational f-f transitions, which occur as sharp and intense emission lines.They are extensively used in plasma display panels (PDPs), field emission displays (FEDs), cathode ray tubes (CRTs), fluorescent lamps and laser devices, etc. [9][10][11] However the production of phosphor with uniform particle size distribution, small particle size, shape control and easy dispersion in polar solvents (like ethanol, methanol) are challenging area.Moreover, decrease in particle size results more surface dangling bonds which can absorb OH -and CO 3 2-from the surrounding environment such as aqueous medium or atmosphere during wet chemical synthesis route.Such small particles can be dispersible in polar medium to some extent.If capping agent is added to such particles, the extent of dispersion in medium with long duration increases.On other hand, variation of luminescence intensity is also dependent on type of capping agents.i.e., Capping agent with long chain hydrocarbon having functional group -O-H, -COOH (e. g. ethylene di-ammine tetra-acetate (EDTA), citric acid) can reduce luminescence intensity; whereas capping agent with short chain hydrocarbon (e. g. ethylene glycol (EG)) improves luminescence as compared to those prepared in water medium. 124][15] Also, some other methods such as hydrothermal and Pechini's methods are used for the preparation of LaF 3 and Y 2 O 3 nanoparticles. 16,17 ong the oxide phosphors, YPO 4 has a large indirect band gap (∼8.6 eV), high dielectric constant (∼7-10), optically isotropic with a refractive index (∼1.72),9][20][21][22] Such small vibration energy is a good choice of the host materials and will allow for effective radiative transitions between electronic energy levels of the rare earth ions in YPO 4 host.YPO 4 has tetragonal structure with space group I4 1/amd (D 19 4h , zircon type, Z = 4) and the Y 3+ ion occupies D 2d site symmetry.The Y 3+ ion is coordinated to the 8 oxygen atoms to form dodecahedron and the PO 4 tetrahedrons are isolated from each other in such a manner -YO 8 -PO 4 -YO 8 -PO 4 -.YO 8 group has two different types of Y-O bonds; whereas PO 4 group has one type of P-O bond.The bond length of four Y-O bonds is ∼2.313Å and remaining four bonds have ∼2.374Å. 22 It was reported that Y-O bond length changes whereas P-O bond length almost remains the same on annealing YPO 4 nanoparticles at higher temperatures. 23n past few years, Ningthoujam and his co-workers 24 has examined the luminescence properties of Ln 3+ ion doped YPO 4 nanoparticles.It was found that the presence of water molecules up to 800 • C in Ce 3+ co-doped YPO 4 :Eu was observed and thus the luminescence intensity was quenched significantly.There are many research articles available to show the enhancement of luminescence in Eu 3+ doped YPO 4 nanophosphor by co-doping Li + and Bi 3+ appreciably, in which the intensity of magnetic dipole transition is predominant over electric dipole transition. 2,25,26 B][29][30][31][32] Existence of intermediate bands observed in YPO 4 system has not been discussed much in literature.This understanding will be useful in transfer process from such intermediate bands to Eu 3+ .Generation of light emitting diodes (LED) in green and red regions has been reported.However, material with efficient blue emitter is challenging to LED applications.Could the blue emitter be produced from YPO 4 :Eu 3+ ?
In this study, we have prepared 5 at.%Eu 3+ doped YPO 4 (YPO 4 :5Eu) and Li + co-doped YPO 4 :5Eu at relatively low temperature of ∼100-120 • C for 1 h using polyol route where EG molecules act as capping agent as well as reaction medium.Their detail crystal structure and luminescence are studied.Variation in luminescence intensities of magnetic and electric dipole transitions with different concentrations of Li + and heat treatment is observed.Interestingly, intermediate bands between band gap of YPO 4 are observed.Variation in color can be observed in this study by modification of surface and particle size/annealing temperature.

A. Materials and synthesis
The nanoparticles of 5 at.%Eu 3+ doped YPO 4 (YPO 4 :5Eu) and Li + (Li + = 3, 5, 7 and 10 at.%) co-doped YPO 4 :5Eu are prepared at low temperature (∼100-120 • C) for 1 h in ethylene glycol (EG).The starting materials for Y 3+ , PO 4 3-, Eu 3+ , Li + are yttrium oxide (Y 2 O 3 , 99.99%, Sigma Aldrich), ammonium dihydrogen phosphate (NH 4 H 2 PO 4 , 99.999%, Sigma Aldrich), europium oxide (Eu 2 O 3 , 99.9%, Sigma Aldrich) and lithium hydroxide (LiOH, 99.99%, Sigma Aldrich), respectively.In typical synthesis procedure of 5 at.%Li + and 5 at.%Eu 3+ doped YPO 4 nanoparticles, 1 g of Y 2 O 3 , 0.086 g of Eu 2 O 3 and 0.012 g of LiOH were dissolved together in concentrated nitric acid (HNO 3 ) in a 250 ml two necked round bottom flask and were heated at 80 • C with addition of double distilled water at least five times in order to remove the excess acid.To this 1.13 g of NH 4 H 2 PO 4 dissolved in 10 ml of double distilled water and 100 ml of ethylene glycol (EG) were added.The solution was stirred for 2 h and followed by 1 h sonication for uniform mixing.The reaction mixture was heated at ∼100-120 • C for 1 h under refluxing condition until the white precipitation is completed.EG molecules act as a solvent as well as capping agent during the reaction for YPO 4 nanoparticles.When the nucleation starts, surrounding EG molecules cap smaller particles and thus, particle growth is slow.The agglomeration among the particles is hindered.Dielectric medium of reaction is also help in controlling particle size.The precipitate so obtained was washed two times by centrifugation in ethanol to remove the excess of EG, and then dried at room temperature for four days.Finally, the as-prepared sample is divided into two parts: one part of the sample was annealed at 500 • C in ambient atmosphere at the rate of 2 • C/min for 4 h.During this, organic capping agent (EG) is removed as CO 2 and H 2 O gaseous due to burning.In addition, only uncapped particles remain and growth of particles occurs.

B. Material characterization
The crystal structure of the material was identified by PW 1071 Philips powder X-ray diffractometer (XRD) with Ni filtered Cu-k α (1.5405 Å) radiation at 30 kV and 20 mA.All patterns were recorded over the angular range 10 ≤ 2θ /deg ≤ 70 with a step size of 2θ = 0.02 • .The Scherrer relation was used to calculate the average crystallite size from XRD spectrum.The relation is expressed as follows: where λ is the wavelength of the X-ray and β hkl the full width at half maximum (FWHM) of peak of the XRD pattern. 33The contribution of instrument to FWHM is removed by using standard Si. 13 Transmission electron microscopy (TEM) image of samples was recorded using a JEOL at an acceleration voltage of 200 kV.For TEM measurement, the samples were grinded and mixed together with EG and dispersed particles could be achieved under ultrasonic vibration for 1 h.A drop of the dispersed particles was put over the carbon coated copper grid and evaporated to dryness in the ambient atmosphere.The high resolution TEM (HRTEM) images were recorded at 300 keV using FEI Titan Microscopy.Infrared (IR) spectrum was measured with a FTIR spectrometer (Bomem MB 102) with a resolution of 1 cm -1 .The sample was mixed with KBr (Sigma Aldrich, 99.99%) in 1:5 ratio and pellet was prepared.Such pellet was used to record the spectra.The photoluminescence (PL) spectra of these powder phosphors were recorded using Hitachi F-4500 spectrometer with a 150 W Xe lamp as a source at the spectral resolution of 3 nm.All the measurements were carried out at room temperature.PL decay was recorded with Edinburgh instrument F920 equipped with Nd-YAG laser pumped optical parameters oscillator (OPO) having a pulse width of 10 ns and repetition frequency of 10 Hz as the excitation source.Chemical bonding energies of ions in sample were measured using X-ray photoelectron spectroscopy (XPS) SPECS, Germany (Mg K α X-ray source, hv = 1253.6eV).UV-Visible absorption spectra were recorded using Simadzu UV 310PC spectrophotometer.

A. XRD study
The XRD patterns of as-prepared and 500 • C annealed 0-10 at.%Li + co-doped YPO 4 :5Eu samples are matched with JCPDS card no: 11-0254.The patterns show that samples have single phase and exhibit tetragonal structure of YPO 4 having space group I4 1/amd .Figure 1 shows the typical XRD pattern of 10 at.%Li + co-doped YPO 4 :5Eu annealed at 500 • C. The average crystallite sizes of 10 at.%Li + co-doped YPO 4 :5Eu as-prepared and 500 • C annealed samples are found to be ∼18 and 42 nm, respectively.The samples show high crystalline nature and intensity increases with Li + ion concentration.Also annealing at 500 • C confirms the increase in crystalline nature.

B. FTIR study
Figure 2 shows the FTIR spectra of 5 at.%Li + co-doped YPO 4 :5Eu nanoparticles of as-prepared and 500 • C annealed samples.Both show similar pattern in peaks except at ∼883 and 1384 cm -1 in the case of as-prepared sample.The peaks at ∼538 and 630 cm -1 correspond to the bending vibrations of PO 43-(termed as v 4 region).The strong bands centered at 1043 and 1084 cm -1 correspond to stretching vibrations of PO 43-(termed as v 3 region). 34,35 he vibration intensity of v 3 region is found to be 18% stronger than that of v 4 region.The integrated intensities of both v 3 and v 4 are found to increase with annealing temperature.We do not find any extra phosphorus containing groups such as P 2 O 7 4-which is generally observed at ∼1265 cm -1 . 36The intensities of peaks corresponding to bending and stretching vibrations of O-H/H 2 O centered at ∼1644 and 3400 cm -1 decrease with annealing temperature.In the case of as-prepared sample, the weak peaks due to stretching vibrations of H-C-H group of EG are found at 2880 and 2948 cm -1 and the peaks centered at ∼883 and 1384 cm -1 correspond to NO 3 -group which may be originated from the presence of HNO 3 which is added in reaction. 37The OH -ions on the surface of particle act as source of quencher to luminescence intensity and the luminescence intensity can be significantly enhanced after removing OH -ions by heat treatment (discussed later).The peak at ∼2349 cm -1 corresponds to asymmetric stretching vibration of CO 2 , which arises from absorption over particle or IR lamp passing through air medium.

C. TEM study
Transmission electron microscope (TEM) and High resolution TEM (HRTEM) are used to characterize the morphology and structure of as-prepared and 500 • C annealed 5 at.%Li + co-doped YPO 4 :5Eu samples.Figures 3(a

D. XPS study
To study the Li + ion effect on chemical binding energies of ions in YPO 4 :5Eu compound, XPS measurements of as-prepared 0 and 10 at.%Li + co-doped YPO 4 :5Eu samples are studied (Fig. 4).The XPS spectrum in Fig. 4(a) shows the peak corresponding to Li (1s) having core binding energy (BE) ∼44.5 eV for 10 at.%Li + co-doped YPO 4 :5Eu sample.Figure 4(b) shows the peaks at ∼156.17 for 0 and 10 at.%Li + co-doped YPO 4 :5Eu samples, respectively.Notably, the peak for P(2p) falls on 131-132 eV.It is difficult to distinguish Eu 2+ (4d 3/2 ) and P(2p).When the spectra in range of 125-144 eV is expanded (Fig. 4(c)), we observe a small peak at ∼141.1 eV which corresponds to Eu 3+ (4d 3/2 ) and no peak corresponding to Eu 2+ (4d 5/2 ) is observed at ∼127.1 eV. 24This confirms the high probability of Eu 3+ present in samples.Figure 4(d) shows the XPS spectra of Y(2p 1/2 ) and its corresponding peaks at ∼299.4,300.1 eV for 0 and 10 at.%Li + YPO 4 :5Eu samples are observed.The O(1s) peaks for 0 and 10 at.%Li + YPO 4 :5Eu samples are found to be at 529.1 and 529.8 eV, are observed (Fig. 4(e)).From this study, BE of individual ions increases on Li + co-doping.This suggests the improvement of crystallinity on Li + co-doping (i.e., defect decreases).Shifting of BE signifies that possibility of change in positive charge and/or chemical environment around Y 3+ /Eu 3+ .XPS study confirms that the Li + co-doping changes the chemical environment around Y 3+ /Eu 3+ .It is to be noted that there is asymmetric nature of O(1s) peak at the higher energy site (i.e., hump at 530.5 eV), which is signature of oxygen vacancy in lattice. 38Deconvolution of peak (O(1s)) of as-prepared 0 and 10 at.%Li + co-doped YPO 4 :5Eu samples using Lorentzian distribution function is shown in Figs.S1(a) and S1(b). 39Here, symbol 'S' refers to Supplemental Material.Two peaks (∼ 529 and 530 eV) could be fitted well.

E. Luminescence study 1. Excitation study
Figure 5(a) shows excitation spectra of as-prepared samples of Li + (0, 3, 5, 7 and 10 at.%) co-doped YPO 4 :5Eu by monitoring the emission wavelength at 594 nm.The excitation spectrum consists of strong absorption band between 225-280 nm with center at ∼255 nm and full width at half maxima (FWHM) ∼25 nm, which can be assigned to the charge transfer from O 2-to Eu 3+ (Eu-O CT).There are sharp absorption peaks at 366, 386 and 399 nm, which correspond to 7 F 0,1 → 5 D 4 , 7 F 0,1 → 5 G 1 , 5 L 7 and 7 F 0 → 5 L 6 of Eu 3+ , respectively. 3,24 he absorption intensity of 7 F 0 → 5 L 6 transition of Eu 3+ ion at 399 nm (FWHM ∼6 nm) is 2.2 times stronger than Eu-O CT absorption indicating a weak energy transfer from Eu-O CT band to Eu 3+ . 40,41 Peaks are similar to as-prepared samples.However, Eu-O CT band is found at ∼241 nm and its intensity is less than that of as-prepared sample.The blue shift of Eu-O CT band with respect to as-prepared sample indicates increase of ionicity with annealing.In pure Eu 2 O 3 , 42 there is no absorption peak at ∼340-350 nm and also intensity of Eu-O CT band is more than that of Eu 3+ (399 nm).In this study, absorption intensity in ∼340-350 nm is very high and also intensity of Eu-O FIG. 6. Luminescence spectra of YPO 4 :5Eu: as-prepared and 500 • C annealed samples at different excitation wavelengths, 375 nm filter is used for 240-350 nm excitations and no filter for 399 nm excitation.
CT band is much less than that of Eu 3+ (399 nm).It means that there is host/intermediate band absorption.However, pure YPO 4 has band gap of 8.6 eV (145 nm), which is more than the energy of Eu-O CT band.Large absorption band in ∼300-500 nm in excitation spectra (Fig. 5) indicates that there are intermediate bands/localized levels between the band gap of YPO 4 .Similar observations were reported in Eu 3+ /Li co-doped LaPO 4 and ZnO. 29,43 n overall, Li + doping improves absorption intensity.

Emission study
We have recorded the emission spectra of Li + (0, 3, 5, 7 and 10 at.%) co-doped YPO 4 :5Eu at different excitation wavelengths: 240-300 nm, 340, 350 and 399 nm.Samples with Li + = 0, 5 and 10 at.% are shown in Figs.6-8 and remaining samples are shown in Figures S2 and S3. 39We have used 375 nm filter for recording emission spectra in all excitations except 399 nm (no filter is used).In the emission spectra of YPO 4 :5Eu (Fig. 6), the broad emission intensity in 400-550 nm with maximum at 430 nm in case of as-prepared sample and at 460 nm in case of 500 • C annealed sample when excited at λ ≥ 270 nm are observed.This broad peak emission is related to host or intermediate bands.The red-shift in emission peak is related to increase of crystallite size on annealing.The intensity of broad emission increases with increasing excitation wavelength from 240 to 399 nm and even is much more than that of Eu 3+ emission at 593 and 615 nm (magnetic 5 D 0 → 7 F 1 and electric 5 D 0 → 7 F 2 dipole transitions, respectively).It is expected that luminescence intensity of Eu 3+ should be more on excitation at 240-250 nm (Eu-O CT band) as compared to that at 399 nm because of energy transfer from Eu-O CT band to Eu 3+ , but this is opposite to this observation.It means that there are intermediate absorption bands between band gap of YPO 4 .This may be related to interaction with absorbed gases (CO 3 2-, H 2 O) or capping agent (EG) in case of as-prepared sample or carbon remained in case of 500 • C annealed sample or defects present in lattice.The emission intensities after excitation at 340, 350 and 399 nm in case of as-prepared sample are very high and found to saturate, whereas 500 • C annealed sample shows the saturation at 399 nm excitation.In the case of 500 • C annealed sample, the host/intermediate bands emission intensity decreases and the broadening of emission peak in ∼400-570 nm occurs as compared to as-prepared sample.This may be related to the extent of decrease of defect or absorbed gases or capping agent (EG) on annealing and thus it enhances energy transfer rate from host/intermediate band to Eu 3+ .
When Li + is co-doped in to YPO 4 :5Eu, the intensity related to host/intermediate bands increases up to Li + = 3 at.%and then decreases with further increase of Li + in case of as-prepared samples (Figs. 7 and 8 and Figs.S2 and S3). 39In case of 500 • C, this intensity remains almost unchanged except 10 at.% of Li + where intensity is found to be high.In all Li + co-doping systems, the 500 • C annealed samples show lower luminescence intensity related to host/intermediate as compared to as-prepared sample.This trend is similar to that without Li + co-doping.Overall, the peak maximum corresponding to the host/intermediate shifts to the higher wavelength by ∼10-20 nm with Li + co-doping up to 10 at.%.This may be related to the increase in crystallite/particle size on annealing.
In addition to strong host/intermediate band emission, the electric and magnetic dipole transitions of Eu 3+ at 593 and 615 nm are observed.Their intensity variations with excitation wavelengths as well as Li + concentrations are found.These are clearly shown in Figs.S4-S13. 39Emission FIG. 8. Luminescence spectra of 10 at.%Li + co-doped YPO 4 :5Eu: as-prepared and 500 • C annealed samples at different excitation wavelengths, 375 nm filter is used for 240-350 nm excitations and no filter for 399 nm excitation.
intensities at different excitation wavelengths are compared by fitting the area under electric and magnetic dipole transitions centered at 593 and 615 nm, respectively using Gaussian distribution function: where I is the observed intensity, I B the background intensity, w i the width at half maximum intensity of the curve and A i area under the curve.λ is wavelength and λ i is the mean wavelength value corresponding to the transition.During calculations of peak areas under electric and magnetic dipole transitions the wavelength range 580-630 nm is used.The asymmetric environment of europium ion (Eu 3+ ) in host lattice can be calculated by intensity ratio of the electric ( 5 D 0 → 7 F 2 ) to magnetic ( 5 D 0 → 7 F 1 ) dipole transitions.This is known as asymmetric ratio represented by A 21 , where subscripts '2' and '1' refer transitions of 5 D 0 to 7 F j , j = 2 and 1, respectively.The A 21 is defined as   A 21 values for as-prepared samples at different excitation wavelengths are calculated.Table I gives the A 21 values of all as-prepared and 500 • C annealed samples at different excitation wavelengths.
In case of as-prepared samples, luminescence intensity of Eu 3+ increases with Li + up to 3 at.%and then decreases for excitations at 240-270, 340 and 399 nm.Figures 9(a)-9(c) show the variation of intensity of electrical dipole transition (A 2 ), its FWHM and asymmetric ratio (A 21 ) with Li + at 399 nm excitation.A 2 increases with Li + up to 3 at.%and decreases with further increase of Li + .Its FWHM decreases from ∼8.1 to 7.8 nm as Li + concentration increases up to 10 at.%.A 21 value is found to be less than 1.0, indicating higher intensity for magnetic dipole transition over that for electric dipole transition.It is to be noted that A 21 is found to be ≥1.0 at 280-300 and 350 nm excitations (Table I).When we see site symmetry of Eu 3+ in YPO 4 , it should be D 2d which has asymmetric environment with two different Eu-O bond lengths. 22,23,44 I such view, the electric dipole transition should be more than the magnetic dipole transition (i.e., A 21 ≥ 1.0).Variation in intensities of these two dipole transitions at different excitation wavelengths was not discussed in the previous studies to the best of authors' knowledge.This may be related to the interaction of incoming excitation light with Eu 3+ environment (nearest environment (O) and second nearest environment (PO 4 ).The interaction parameter will depend on incoming wavelength and intensity.This revelation after analysis is one of present findings.
In case of 500 • C annealed samples, variation of luminescence intensity with Li + is similar to as-prepared samples, but their luminescence intensity and FWHM are higher than as-prepared samples for 399 nm excitation (Figs.9(a) and 9(b)).Increase in luminescence intensity is related to decrease of non-radiative rate from surface dangling bonds and capping ligand (EG) on annealing.A 21 is found to close to 1.0 at all excitation wavelengths and such behavior is different from asprepared samples.Highest A 21 is found to be ∼1.34 for 7 at.%Li + co-doped YPO 4 :5Eu under 260 and 280 nm for 500 • C annealed sample.Its value varies between ∼0.9-4.0 in some glass systems and ∼7-10 in CaMoO 4 hosts. 3,45,46 Sn et al. 27 reported the effect of excitation wavelengths at ∼147 nm (host excitation) and 254 nm (Eu-O CT excitation)) on emission intensity and A 21 of Eu 3+ doped YPO 4 .At 254 nm excitation, A 21 is 1:1, whereas this becomes 1:0.8 at ∼147 nm excitation.Ningthoujam and his coworkers 47 have examined the luminescence properties of Eu 3+ ion doped YPO 4 nanoparticles capped by ethylenediamine tetraacetic acid (EDTA).It is found that the excitation wavelength significantly influences the emission intensity and A 21 due to the presence of capped layer on the surface of nanoparticles.Recently, Li and his coworkers 29 studied white light emission from oleic acid capped LaPO 4 :Eu 3+ nanorods.Due to the formation of intermediate state/mid-gap states as a consequence of the interaction of chemical bonding of oleic acid to LaPO 4 :Eu 3+ , the color coordinates from Red to White under different excitation wavelengths (270-395 nm) were obtained.Guo and his coworkers 31 found the higher luminescence intensity at electric dipole transition than that at magnetic dipole transition in glass-ceramic having Eu 3+ doped YPO 4 nano crystals.In addition to Eu 3+ they found the presence of Eu 2+ in glass matrix which results broad emission peak centered at ∼436 nm due to the 5d-4f transition of Eu 2+ .Conversion from Eu 3+ to Eu 2+ is related to the reduction process occurred from fluoride used in the preparation of glass-ceramic.Ray and his coworkers 32 prepared YPO 4 :Eu nanorods and nanoparticles, which have dominant in electric dipole transition than magnetic dipole transition.From above literatures, variation of luminescence intensities of electric and magnetic dipole transitions depends on sample characteristics.Now, we have a question for origin of broad luminescence observed in ∼400-500 nm whether it is from the host or due to Eu 2+ present in the sample.In our study, the probability of reduction of Eu 3+ to Eu 2+ is very low since there is no fluoride precursor during preparation.Also, our XPS study (Fig. 4) confirms the no peak corresponding to Eu 2+ in the sample spectrum.This suggests that the observed luminescence in ∼400-500 nm is related to the presence of intermediate bands within host, which arises from oxygen defect in lattice (from XPS study) or capping ligand interaction with surface of particles or adsorbed gases over surface of particles (from IR study).
Figures 10(a) and 10(b) show the photographs of 5 at.%Li + co-doped YPO 4 :5Eu as-prepared and 500 • C annealed nanoparticles under the Nd-YAG laser excitation at 266 nm (power ∼ 0.3 W at focusing spot).As-prepared sample shows strong blue color whereas 500 • C annealed sample shows pink colour.When we see emission spectrum for 260-270 nm excitation (Fig. 7, in which μW power is used), it is expected that blue-green colour should be instead of pink colour.In our opinion, the high power of laser makes electric and magnetic dipole transition intensities more prominent over host emission for 500 • C sample (pink colour).
The presence of EG molecules capped on surface of as-prepared Li + co-doped YPO 4 :5Eu particles are useful to easy dispersion of nanoparticles in polar solvents like ethanol and polyvinyl alcohol (PVA).For a typical dispersion, 10 mg of as-prepared 5 at.%Li + co-doped YPO 4 :5Eu is dispersed in 5 ml of ethanol followed by ultrasonication.Figure 10(c) shows photograph of redispersed as-prepared 5 at.%Li + co-doped YPO 4 :5Eu, before and after exposure of 266 nm laser excitation.Further, to make a thin film of 5 at.%Li + co-doped YPO 4 :5Eu, 10 mg of as-prepared sample is mixed with 2.5 ml of distilled water.To that, 1 g of PVA and 2.5 ml of ethanol are added.The solution is ultrasonicated for 30 min to make uniform dispersion.This solution was placed over poly petri dish.It is kept for 4 days at room temperature for drying.In this way polymer films of 5 at.%Li + co-doped YPO 4 :5Eu having thickness ∼0.2-0.3 mm and ∼10 cm diameter are prepared.The film shows very bright blue under 266 nm laser excitation (Fig. 10(d)).Both dispersed particles in the ethanol and the film show blue colour with the CIE (Commission internationale de l'Eclairage) chromaticity co-ordinates (0.17, 0.17).Li et al. 29 reported strong blue emission of oleic acid capped Eu 3+ doped LaPO 4 having CIE values around (0.22, 0.13) under 380 nm excitation.Uniform brightness of the thin film confirms the homogeneous distribution of particles in PVA matrix.It is found that peak positions of electric and magnetic dipole transitions are unaffected after dispersion of nanoparticles in PVA polymer matrix (which is not shown here).The emission intensity of as-prepared nanoparticles re-dispersed in ethanol and PVA matrix is slightly less than the powder sample.This is due to the presence of less no of Eu 3+ ions per unit volume of dispersion.This film will be useful in the development of optoelectronic devises.
Figure 11 shows the schematic diagram of energy levels in Li + co-doped YPO 4 :5Eu.Band gap of YPO 4 (8.6 eV = 145 nm) is more than Eu-O charge transfer band (4.8-5.1 eV = 240-260 nm) and their band edges are crossing each other. 18Because of defect on surface or capping ligands on particles, there is possibility to have intermediate bands, which is also supported by UV-Visible absorption measurement (Fig. 12).Absorption band extends up to 500 nm.Such intermediate band absorption helps in enhancement of luminescence intensity by energy transfer (ET) to Eu 3+ .Similar type of results are shown in case of LaPO 4 nanoparticles. 29After excitation at ∼280-399 nm, the electrons from the valence band go to excited state which falls on intermediate bands/mid gap states and holes are created at the valence band.After the removing the excitation source, the electrons comes to the valence band (ground state).During this, the electron-hole recombination takes place resulting to emission at blue to pink colour regions.In addition, emission due to Eu 3+ ( 5 D 0 → 7

F. Luminescence Decay Study
The luminescence decay curves of the level 5 D 0 (593 nm) of Eu 3+ for as-prepared and 500 • C annealed samples of Li + (0, 3 and 10 at.%) co-doped YPO 4 :5Eu have been shown in Figs.13(a)-13(c).Excitation wavelength is fixed at 464 nm from Nd-YAG laser.The decay curves are not well fitted by mono-exponential equation (I = I 0 exp(t/τ ) ).The typical mono-exponential curve fitting to data of 3 at.%Li + doped YPO 4 :5Eu annealed at 500 • C with χ 2 = 3.195 is shown in Figure S14 and parameters obtained after fitting is given in Table S1. 39The fitting behavior can be clearly understood by plotting ln(I) vs. t, which is shown in Fig. S14 (inset). 39The fitted straight line does not match with decay data points.This suggests that environment of Eu 3+ ions in lattice are not same in different positions.
All decay data are fitted by using bi-exponential decay equation, which is expressed as where I 1 and I 2 are the intensities at different time intervals and τ 1 and τ 2 their corresponding lifetimes.The bi-exponential fitting to luminescence decay data of 500 • C annealed 3 at.%Li + co-doped YPO 4 :5Eu nanoparticles is shown in Fig. 13(d).The fitting parameters are given in figure itself.The average lifetime can be calculated using the equation, The parameters obtained after bi-exponential equation fitted to data are given in Table II.In case of as-prepared samples, the lifetime of Eu 3+ increases from 1.05 to 1.26 ms with increasing Li + concentration from 0 to 3 at.%and decreases with further increase of Li + (i.e. 10 at.%Li + , where lifetime is 1.09 ms).This behavior is similar to variation in luminescence intensity with Li + (Fig. 9).In case of 500 • C annealed samples, lifetime value of 1.52 ms is observed for Li + = 0 and 3 at.%and 1.85 ms for Li + = 10 at.%.The lifetime value increases on annealing.The bi-exponential decay suggests the availability of Eu 3+ ions on surface (τ 1 ) and core (τ 2 ) of particles.On annealing, the percentage of τ 1 decreases and that of τ 2 increases.This indicates a decrease of non-radiative rate from surface of particles on annealing.

IV. CONCLUSIONS
Li + (Li + = 0, 3, 5, 7 and 10 at.%) co-doped YPO 4 :5Eu nanoparticles are prepared using polyol method at reaction temperature 100-120 • C. As-prepared samples are annealed to 500 • C to remove the presence of dangling bonds on the surface of nanoparticles.In the case of as-prepared /n, χ 2 is goodness of fitting, w k is weighting factor for data points (w k = 1/ √ F k ), X k is the calculated lifetime and F k is the measured lifetime data.
samples, the formation of intermediate bands/mid gap states is found.The intensities of both the intermediate bands (400-500 nm) and Eu 3+ (500-700 nm) emission of the samples increase with Li + up to 3 at.%and then decrease with increasing Li + in the case of as-prepared samples.However, significant change has not been observed in intermediate band emission intensities for 500 • C annealed samples with/without co-doping of Li + and its emission intensity is slightly less than as-prepared sample except 10 at.%Li + co-doping.However the intensity of Eu 3+ emission is significantly enhanced on annealing at 500 • C. The intensities of intermediate band and Eu 3+ emission are significantly influenced by the excitation wavelengths (240-399 nm).Blue emission is not due to the Eu 2+ confirmed by XPS study and this is assigned to intermediate bands within host band gap where electron-hole recombination takes place after excitation.We observe core binding energy with a small peak at ∼141.1 eV in XPS spectra which corresponds to Eu 3+ (4d 3/2 ) and no peak corresponding to Eu 2+ (4d 5/2 ) is observed at ∼127.1 eV.This confirms the high probability of Eu 3+ present in samples.The as-prepared nanoparticles are re-dispersed in ethanol and PVA thin film is prepared form the re-dispersed particles.The film shows uniform bright blue.The CIE chromaticity coordinates of as-prepared samples are close to (0.17, 0.17).They can be potential candidate for life science activity and LEDs applications.The emission intensity ratio of electric to magnetic dipole transitions varies with the excitation wavelength.In case of as-prepared samples, the lifetime of Eu 3+ ( 5 D 0 ) increases from 1.05 to 1.26 ms with increasing Li + concentration up to 3 at.%and decreases with Li + (i.e. 10 at.%Li + ) where lifetime is ∼1.09 ms and such behavior is similar to variation in luminescence intensity.In case of 500 • C annealed samples, lifetime values are found to be ∼1.52 and 1.85 ms for 0 and 10 at.%Li + co-doping, respectively.This work demonstrates the preparation of high quality luminescence material by introducing Li + ion, which helps in enhancement of luminescence intensity of Eu 3+ .
) and 3(b) show TEM and HRTEM images of as-prepared sample.The well-resolved lattice fringes of as-prepared sample having inter-planar distance (2.79 Å) are observed and are assigned to (211) plane of tetragonal system after Fast-Fourier-Transform (FFT).

Figure 3 (
c) shows lattice fringes of its 500 • C annealed sample.Two types of lattice fringes having inter-planar spacing of 3.43 and 2.82 Å corresponding to the (200) and (211) planes of tetragonal phase of YPO 4 are found.The selected area electron diffraction pattern of as-prepared sample is shown in Figure3(d).The high crystallinity is confirmed by the rings as shown in Figure3(d).The assignment of the rings is shown in figure itself.Spherical particle sizes of 40 and 70 nm are observed from as-prepared and 500 • C annealed samples, respectively.
Figure5(a) shows excitation spectra of as-prepared samples of Li + (0, 3, 5, 7 and 10 at.%) co-doped YPO 4 :5Eu by monitoring the emission wavelength at 594 nm.The excitation spectrum consists of strong absorption band between 225-280 nm with center at ∼255 nm and full width at half maxima (FWHM) ∼25 nm, which can be assigned to the charge transfer from O 2-to Eu 3+ (Eu-O CT).There are sharp absorption peaks at 366, 386 and 399 nm, which correspond to 7 F 0,1 → 5 D 4 , 7 F 0,1 → 5 G 1 , 5 L 7 and 7 F 0 → 5 L 6 of Eu 3+ , respectively.3,24The absorption intensity of 7 F 0 → 5 L 6 transition of Eu 3+ ion at 399 nm (FWHM ∼6 nm) is 2.2 times stronger than Eu-O CT absorption indicating a weak energy transfer from Eu-O CT band to Eu 3+ .40,41 Figure 5(b) shows the excitation spectra of 500 • C annealed samples by monitoring the emission wavelength at ∼594 nm.Peaks are similar to as-prepared samples.However, Eu-O CT band is found at ∼241 nm and its intensity is less than that of as-prepared sample.The blue shift of Eu-O CT band with respect to as-prepared sample indicates increase of ionicity with annealing.In pure Eu 2 O 3 , 42 there is no absorption peak at ∼340-350 nm and also intensity of Eu-O CT band is more than that of Eu 3+ (399 nm).In this study, absorption intensity in ∼340-350 nm is very high and also intensity of Eu-O
FIG. 11.Schematic diagram for energy transfer process among the Eu-O, intermediate bands and Eu 3+ state in YPO 4 :5Eu.

FIG. 13 .
FIG. 13.Lifetime decay spectra of as-prepared and 500 • C annealed Li + doped YPO 4 :5Eu nanoparticles of (a) Li + = 0, (b) Li + = 3 and (c) Li + = 10 at.% under 464 nm laser excitation.Emission is monitored at 590 nm.(d) Bi-exponential fitting to luminescence decay data of 3 at.%Li + doped 500 • C annealed sample and fitting parameters are shown in the figure itself.The y-axis in (a)-(c) are represented on log scale.

TABLE II .
Parameters obtained after bi-exponential decay fit to data of as-prepared (ASP) and 500 • C annealed samples (500 • C).Excitation and emission wavelengths are fixed at 464 and 594 nm, respectively.Nd-YAG laser source is used.