Effect of Li4P2O7 and Li2Cu2P6O18 minor phases on the luminescentproperties of Cuþ doped Li3PO4

Rujuta A. Barve a,1, Rajesh R. Patil a,n, Mukund S. Kulkarni b, Bhuwan C. Bhatt c,Sanjeev V. Moharil d

a Institute of Science, Nagpur, Indiab Radiation Safety Systems Division, Bhabha Atomic Research Centre, Mumbai, Indiac D.S.T Fellow, C/o RPAD, BARC, Mumbai, Indiad R.T.M Nagpur University, Nagpur, India

a r t i c l e i n f o

Article history:Received 6 November 2013Received in revised form10 June 2014Accepted 27 June 2014Available online 8 July 2014

Keywords:Lithium compoundsOrthophosphatePhotoluminescenceThermoluminescenceOSL

a b s t r a c t

The Li4P2O7 and Li2Cu2P6O18 phase formation and its correlation with luminescent properties in Cuþ

doped Li3PO4 were studied. The study shows that in air annealed samples minor phase of Li4P2O7 alongwith Li2Cu2P6O18 is formed. The amount of Li4P2O7 is inversely proportional to the annealingtemperature whereas amount of Li2Cu2P6O18 increases with annealing temperature. These phases affectluminescence properties of the material significantly. The Cuþ luminescence decreases with the increasein Li2Cu2P6O18 whereas, the TL–OSL sensitivity is found to be a function of amount of Li4P2O7 phasepresent in the sample. Of the sample annealed at various temperatures maximum OSL sensitivity isobtained for the 600 1C annealed sample for which 20% Li4P2O7 phase is detected which is maximumamong all air heated samples. The minimum sensitivity is observed in case of sample melted in air inwhich 3% of this phase is detected. Vacuum heated sample also shows minimum sensitivity, in which noLi4P2O7 phase is detected.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Among various orthophosphates, Li3PO4 has received lot ofattention because of its chemical and physical stability and abilityto form solid solutions with various compounds. Li3PO4 is knownto exist in three different phases α, β and γ which form at normalpressure and are stable at progressively higher temperatures[1–3]. β-Li3PO4 has a wurtzite superstructure and is a lowtemperature phase. This low temperature phase undergoes aphase transition to γ-Li3PO4 (high temperature phase) at 500 1C[4]. But this transition is found to be irreversible under normal dryconditions [5]. At ambient temperature, Li3PO4 may exist in bothpolymorphic forms. At 1170 1C, γ-Li3PO4 transforms into α-Li3PO4

which is stable up to the melting point at 1220 1C [1]. Li3PO4

possesses high ionic conductivity in this phase which is suitablefor battery applications. The ionic conductivity of alkali orthopho-sphates is also found to be enhanced significantly when synthe-sized under humid or hydrogen containing atmosphere [6].Attempts were also made to dope nitrogen to enhance the ionic

conductivity over several orders of magnitude. This led to thedevelopment of a new material named as LIPON.

Luminescence of 3dn ions such as Cr4þ , Mn5þ and Fe is studiedin Li3PO4 as part of study in tetrahedral configuration for possibleapplications as NIR laser material. Li3PO4:Mn5þ is found to be apotential laser material [7]. The spectroscopic properties of thisphosphor have also been used for studying the pressure inducedphase transitions of Li3PO4 [8]. LiZn1�xPO4:Mnx is another potentialphosphor that can be used as a UV radiation-converting phosphor forLEDs [9]. The Zeff of Li3PO4 is 10.6 very near to that of tissue (Zeff¼7.4)and other routinely used dosimetry phosphor like Al2O3:C. Therefore,attempts were made to develop this phosphor for applications inradiation dosimetry. Cuþ emission in Li3PO4 is reported as early as2001 [10]. Thermoluminescence in Li3PO4 is studied by Naranje et al.[11] with the intention to study role of phosphorous in LiF:Mg,Cu,P.This study led to the development of new Li3PO4 based phosphor forTL dosimetry. Detailed characterization of this phosphor was laterdone by Bhatt et al. [12].

Several luminescence properties of solids such as luminescenceefficiency, emission spectra and excited state lifetimes are affectedby the phase of the material. Discontinuities have been observed tooccur in these luminescence parameters during phase transitions[13]. In our earlier work changes in luminescence spectra of Cuþ

with respect to phase are observed in Na2SO4 [14]. Recently, the

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jlumin

Journal of Luminescence

http://dx.doi.org/10.1016/j.jlumin.2014.06.0570022-2313/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author. Mobile: þ91 9890359291; fax: þ91 712 2565581.E-mail address: rvapatil@yahoo.com (R.R. Patil).1 Present address: Radiological Safety Division, IGCAR, Kalpakkam, Tamil Nadu,

India.

Journal of Luminescence 156 (2014) 25–29

discontinuities in the 110 1C TL peak intensity are reported in bio-glass with the hydroxyl apatite formation [15]. In this paper wereport the photoluminescence, thermoluminescence, and opticallystimulated luminescence in Cu-doped Li3PO4 and its correlation tothe Li4P2O7 and Li2Cu2P6O18 phases which are formed during theannealing of the material.

2. Experimental

Cu doped Li3PO4 was synthesized using co-precipitationmethod in which lithium salt is dissolved in double distilled waterto which CuCl2 solution in desired amount was added. Thephosphorous precursor in stoichiometric proportion was dissolvedin double distilled water separately. Both solutions were mixed toget the white precipitate of Li3PO4. This precipitate was filtered,washed several times by double distilled water and then driedunder drying lamp to get the dried powder of Li3PO4. Severalsamples with varying Cu concentration (0.01 mol% to 0.5 mol%)were prepared however sample with 0.05 mol% Cu impurityconcentration yielded maximum photoluminescence. Hence thisconcentration is referred to as optimized concentration andsample with this concentration was chosen for further study.All the Li3PO4 samples discussed hereafter are with 0.05 mol% ofCu concentration. The sample was heated in air in graphitecrucible for 30 min in the temperature range 600–800 1C. Part ofthe sample was melted at 1200 1C in graphite crucible andquenched. Melted sample is labeled as sample B whereas pre-cipitated sample is labeled as sample A. All the samples weresynthesized three times and were found to be reproducible within75%. For taking measurements sieved particles in the 90–210 μmwere used.

For studying the TL/OSL response, all the samples were exposedto a test dose of 100 mGy using 90Sr/90Y beta source with the doserate of 20 mGy per min. All the samples were weight normalizedand the measurements were carried out under identical experi-mental conditions. Photoluminescence studies were carried onHitachi-4000 Spectrofluorometer. For recording PL, 200 mg pow-der of the samples having particle size in the range of 90–210 μmwas used. Emission spectra were recorded with excitation bandpass 5 nm and emission band pass 1.5 nm, while excitation spectrais recorded with excitation band pass 1.5 nm and emission bandpass 5 nm. 150 W Xenon lamp with ozone self deozonatinghousing is used as a light source. Large stigmatic concave gratinghaving 900 lines per mm is used for the excitation and emissionside with eagle mounting (F:3). Blaze wavelengths are 300 nm onexcitation side and 400 nm on emission side. Photomultiplier Tube(PMT) R 372F with flat spectral response is used for measure-ments. The spectra are corrected for wavelength response of theinstrument in the range 220–600 nm.

Optical stimulated luminescence (OSL) and thermolumines-cence was studied on the fabricated set up [16]. The assemblyconsists of array of blue/green LEDs as a stimulation source withadjustable power from 11 mW/cm2 to 48 mW/cm2. Two opticalfilters viz. UG-1 (across PMT), to prevent stimulation signal fromreaching PMT (9111B, 25 mm diameter end window PMT) and GG-435 (across LEDs), to cut-off the stimulation wavelengths below435 nm, were used in the assembly. All the operations in theassembly are controlled by the suitable software.

During all the OSL measurements the LED power was kept at11 mW/cm2 and signal was recorded for 200 s with the acquisitiontime 0.1 s. All the thermoluminescence measurements were takenat the heating rate 4 1C/s. The TL was recorded in air atmospherewith HA-3 filter used on PMT side. The details regarding theuncertainty in the temperature measurement are discussed else-where [16]

3. Results and discussions

Fig. 1 shows the X-ray diffraction pattern of various Li3PO4

samples. XRD pattern of as-prepared sample A (Fig. 1(a)) showsthat the intense diffraction peaks can be assigned to Li3PO4 inorthorhombic phase with space group Pmnb(62) and is similar toobserved earlier (JCPDS No. 83-0339). No other peaks that can beassigned to different compounds of Lithium and Phosphorous areobserved indicating purity of phase. Fig. 1(b) shows the XRDpattern of sample heated at 600 1C. In this pattern along withmajor phase of Li3PO4, minor phase of Li4P2O7 (JCPDS No.:77-1045)along with Li2Cu2P6O18 (JCPDS No.:27-1232) is detected. Theseminor phases are present in all annealed samples and the relativepercentage is found to be the function of annealing temperature.This could be judged by comparing the relative ratio of mostintense diffraction line of Li3PO4 to that of 100% intense line ofthese phases. Table 1 shows tabulation of the relative percentageof these phases in the sample annealed at different temperature.

Fig. 1. X-ray diffraction pattern of various Li3PO4 samples. (a) X-ray diffractionpattern of as-prepared sample A. (b) X-ray diffraction pattern of sample A heated at600 1C. (c) X-ray diffraction pattern of sample A heated at 800 1C. (d) X-raydiffraction pattern of sample B. (e) X-ray diffraction pattern of vacuum heatedsample.

R.A. Barve et al. / Journal of Luminescence 156 (2014) 25–2926

The variation of relative percentage of these phases with respect toannealing temperature is shown in Fig. 7(a).

It can be seen that for sample annealed at 600 1C, thecontribution of minor phase of Li4P2O7 is maximum. Furtherheating the sample at 800 1C decreases the relative ratio of 100%intensity line of Li4P2O7 with respect to Li3PO4 (Fig. 1(c)) whereaslines corresponding to Li2Cu2P6O18 phase start increasing. Fig. 1(d) shows the XRD pattern of the melted sample (sample B). In thispattern Li4P2O7 is found to reduce drastically with increase inLi2Cu2P6O18 phase. Effect of change in annealing atmosphere alsohas profound effect on the formation of these minor phases. Invacuum heated sample pure phase of Li3PO4 is observed as nolines corresponding to these minor phases are observed (Fig. 1(e)).

Fig. 2 shows the PL spectra of Cu doped Li3PO4 (sample A).The PL intensity of Cu doped sample is found to be dependent onannealing temperature. As-prepared and 450 1C heated sample donot exhibit Cu luminescence. Heating sample progressively athigher temperature increases the PL intensity. Sample heated at600 1C shows emission around 365 nm (Fig. 2(e)) with the excita-tion at 254 nm (Fig. 2(a)). Similar emission with shoulder peakaround 385 nm (Fig. 2(f)) and excitation is observed for the sampleheated at 700 1C. The intensity of emission is almost doublecompared to the PL intensity for the sample heated at 600 1C.The sample heated at 800 1C shows intense PL emission around365 nm with shoulder peak around 385 nm (Fig. 2(g)). Theexcitation is found at 257 nm (Fig. 2(b)). The PL intensity in thesample B melted at 1200 1C (Fig. 2(h)) is found to reduced by 10times compared to sample heated at 800 1C (Fig. 2(g)) with furtherincrease in shoulder peak around 385 nm The excitation for thisemission is around 257 nm (Fig. 2(c)). The observed Cuþ emissioncould be attributed to 3d94s1-3d10 transitions of Cu ion and theincrease in the PL intensity could be correlated to the decrease ofLi4P2O7 in Li3PO4 phase with the increase in annealing tempera-ture (Fig. 7b). For the samples heated beyond 800 1C this phase isfound to be reduced significantly (3% for sample melted at1200 1C) with the increase in the relative contribution ofLi2Cu2P6O18 phase. In this phase the valency of Cu is 2þ and

hence substantial Cuþ ions get converted into Cu2þ ions reducingthe Cuþ photoluminescence at 365 nm in the sample annealed at900 1C and that melted at 1200 1C. The increase in the peak heightof shoulder peak observed around 385 nm could also be correlatedto increase of this phase. However vacuum heated sample A showsenhancement in the PL intensity (Fig. 2(i)) and excitation (Fig. 2(d)). The observed increase in the PL intensity is due to absence ofLi4P2O7 as well as Li2Cu2P6O18 phase which are observed in the airheated samples. The results presented above reveal that formationof Li4P2O7 and Li2Cu2P6O18 phases is responsible for quenching ofCu luminescence (as per the PL results presented here) in Li3PO4.On the other hand presence of these phases imparts TL/OSLsensitivity to the sample.

Fig. 3 shows the TL response of Cu doped sample A heated atdifferent temperature. All the TL glow curves are recorded at theheating rate of 4 1C/s. The sample heated at 600 1C found to bemost sensitive and shows intense TL peak at 120 1C havingshoulder at 175 1C (Fig. 3(a)). The TL peak is found to be shiftedto 123 1C when the sample is heated at 700 1C and 800 1C (Fig. 3(b) and (c)) with decrease in the TL sensitivity. Area-wise, thesensitivity is found to be 0.9 and 0.7 times the sensitivity of thesample heated at 600 1C. Further heating of sample at 900 1C shiftsTL peak at 126 1C with a small shoulder around 178 1C (Fig. 3(d)).The sensitivity was found to be 0.6 times the sensitivity of thesample heated at 600 1C. When compared with the TL sensitivityof commercial Al2O3:C (Landauer Inc), the TL sensitivity of 600 1Cheated sample is found to be 1.3 times to that of the commercialAl2O3:C (Landauer Inc.).

The intense low temperature peak could be attributed to theminor phase of Li4P2O7. Naranje et al. [11] observed such peak inCu doped Li4P2O7. The variation of TL sensitivity with respect toannealing temperature is shown in Fig. 7(c). It can be seen that theTL sensitivity is maximumwhen the relative percentage of Li4P2O7

is maximum (Fig. 7(a)).This indicates that the formation of Li4P2O7 phase produces

large number of defects responsible for TL. Since the contributionof this phase decreases with the increase in annealing tempera-ture, the number of defects responsible for TL also reducesresulting into decrease in the TL sensitivity. Fig. 4 shows the TLresponse of sample B, and vacuum heated sample A. Sample Bshows TL peak at 110 1C and a small broad peak around 340 1C(Fig. 4(a)). Significant changes were observed in the TL of vacuumheated sample A. The sample exhibits two TL peaks; one peak at124 1C and other broad peak around 370 1C (Fig. 4(b)). However,the sensitivity is reduced drastically and found to be 25% to thatfor sample B. This decrease in the peak height of low temperature

Table 1Ratio of relative intensity of 100% line of Li4P2O7 and Li2Cu2P6O18 with Li3PO4.

Li3PO4 sample Li4P2O7 (%) Li2Cu2P6O18 (%)

As-prepared 0 0Sample A heated at 450 1C in air 6.6 0Sample A heated at 600 1C in air 20 2.2Sample A heated at 800 1C in air 11 4Sample B 3 14.7Sample A heated at 600 1C under vacuum 0 0

Fig. 2. Photoluminescence spectra of Cu doped Li3PO4. (a) Excitation curve ofsample A heated at 600 1C. (b) Excitation curve of sample A heated at 800 1C.(c) Excitation curve of sample B. (d) Excitation curve of sample A heated at 600 1Cunder vacuum. (e) Emission curve of sample A heated at 600 1C. (f) Emission curveof sample A heated at 700 1C. (g) Emission curve of sample A heated at 800 1C.(h) Emission curve of sample B. (i) Emission curve of sample A heated at 600 1Cunder vacuum.

Fig. 3. Thermoluminescence response of sample A heated at different temperature.(a) Thermoluminescence response of sample A heated at 600 1C. (b) Thermolumine-scence response of sample A heated at 700 1C. (c) Thermoluminescence response ofsample A heated at 800 1C. (d) Thermoluminescence response of sample A heated at900 1C.

R.A. Barve et al. / Journal of Luminescence 156 (2014) 25–29 27

peak indicates decrease in the relative contribution of Li4P2O7

phase responsible for such TL. Observation of high temperaturepeak indicates the formation of pure phase of Li3PO4. The observa-tion of such peak in Cu doped Li3PO4 is well reported in theliterature [11].

Intense BSL (Blue stimulated luminescence) is observed in allannealed samples. The sensitivity as well as decay is found to bedependent on annealing temperature. Fig. 5 shows the OSL decayof various samples with the inset showing the BSL decay ofcommercial Alumina (Landauer). The BSL of Cu doped sample Aannealed at 600 1C (Fig. 5b curve (i)) decays fast as compared tocommercial Alumina and to those annealed at 700 1C, 800 1C and900 1C (Fig. 5b curve (ii, iii, iv)) and exhibits highest sensitivity. Thesensitivity of this sample is found to be 3.2 times more than thecommercial Al2O3:C (Landauer Inc.) when compared using totalarea integration method. The BSL signal of sample annealed at600 1C decays within 12 s (Fig. 5(b) curve (i)). However, the BSLsignal of samples annealed at 700 1C, 800 1C and 900 1C found todecay within 20 s (Fig. 5(b) curve (ii, iii, iv)). The BSL sensitivity ofsamples annealed at 700 1C, 800 1C and 900 1C is found to be 81%,71% and 60% respectively as compared to the sample heated at600 1C. The BSL signal of vacuum heated sample A decaysrelatively fast within 8 s (Fig. 6(b)). However the BSL decay ofsample B is slow and the complete signal was found to decaywithin 25 s (Fig. 6(a)). The BSL sensitivity of these samples is alsovery low as compared with the sensitivity of air heated sample A.The BSL sensitivity of vacuum heated sample and sample B isfound to be 1% and 3% respectively to that of sample heated at600 1C. The variation of OSL sensitivity with respect to annealingtemperature is shown in Fig. 7(d). It can be seen that the OSLsensitivity is maximum when the relative percentage of Li4P2O7 ismaximum (Fig. 7(a)) and this is similar to what is observed incase of TL.

The results presented above reveal that the observed TL–OSLsensitivity is due to the presence of minor phase of Li4P2O7. Thispyrophosphate phase which seems to be responsible to act astrapping center results in imparting TL–OSL sensitivity to thematerial. Such observations were also made earlier in LiF:Mg,Cu,PTLD phosphor in which the TL sensitivity as well as shape of glowcurve was found to be dependent on Li4P2O7 phase which is presentin small amount [17].

Thus the observed decrease in the OSL sensitivity of the sampleB, and vacuum heated sample is due to absence of Li4P2O7 phasewhich reduces the number of defects responsible for TL and OSL.

4. Conclusion

Cu doped Lithium orthophosphate shows good OSL and theobserved OSL sensitivity is found to be function of annealingtemperature as well as atmosphere. In all annealed samples, minorphases of Li4P2O7 and Li2Cu2P6O18 are detected and the amount isdependent on annealing temperature. The presence of these

Fig. 4. Thermoluminescence response of sample B and vacuum heated sample A.(a) Thermoluminescence response of sample B. (b) Thermoluminescence responseof sample A heated at 600 1C under vacuum.

Fig. 5. (a) CW-OSL (BSL) response of sample A heated at different temperature(Inset shows the BSL response of commercial Alumina (Landauer)). (i) BSL responseof sample A heated at 600 1C. (ii) BSL response of sample A heated at 700 1C.(iii) BSL response of sample A heated at 800 1C. (iv) BSL response of sample Aheated at 900 1C. (b) Normalized BSL response of sample A heated at 600 1C.(i) Normalized BSL response of sample A heated at 700 1C. (ii) Normalized BSLresponse of sample A heated at 800 1C. (iii) Normalized BSL response of sample Aheated at 900 1C.

Fig. 6. CW-OSL (BSL) response of sample B and vacuum heated sample (Insetshows the normalized BSL response of sample B and vacuum heated sample).(a) BSL response of sample B. (b) BSL response of sample A heated at 600 1C undervacuum.

R.A. Barve et al. / Journal of Luminescence 156 (2014) 25–2928

phases affects the luminescent properties of the material. Thepresence of these phases is found to be detrimental to Cuþ

luminescence. However better TL–OSL is observed in the samplesin which Li4P2O7 phase is present in significant amount. The TLand OSL sensitivity of the samples is found to be reduced with thedecrease in the contribution of Li4P2O7 phase as this reduces thenumber of defects responsible for TL and OSL. Amongst thestudied phosphors, intense OSL with fast decay compared tocommercial Al2O3:C is observed in the Cu doped sample A, heatedin air at 600 1C and its sensitivity is 3.2 times higher as comparedwith commercial Al2O3:C on the basis of area integration method;integrated over the time period of 200 s.

Acknowledgements

The authors are grateful to B.R.N.S. (Sanction no. 2008/37/20)for funding this work.

References

[1] E. Reculeau, A. Elfakir, M. Quarton, J. Solid State Chem. 79 (1989) 205.[2] Von J. Zemann, Acta Crystallogr 13 (1960) 863.[3] R.D. Shannon, Acta Crystallogr., Sect. A: Found. Crystallogr. 32 (1976) 751.[4] C. Keffer, A. Mighell, F. Mauer, H. Swanson, S. Block, Inorg. Chem. 6 (1967) 119.[5] A.R. West, F.P. Glasser, NBS Special Publ. Solid State Chem 364 (1972) 457.[6] Bin Zhu, Bengt - Erik Mellander, Mater. Res. Bull. 28 (1993) 321.[7] M.F. Hazenkamp, H.U. Gudel, S. Kuck, G. Huber, W. Rauw, D. Reinen, Chem.

Phys. Lett 265 (1997) 264.[8] T. Riedener, Y. Shen, R.J. Smith, K.L. Bray, Chem. Phys. Lett. 321 (2000) 445.[9] Ting-Shan Chan, Ru-Shi Liu, I. Baginskiy, Chem. Mater. 20 (2008) 1215.[10] R.R. Patil, S.V. Moharil, Phys. Status Solidi A 187 (2001) 557.[11] S.M. Naranje, S.V. Moharil, Phys. Status Solidi A 165 (1998) 489.[12] B.C. Bhatt, B.S. Dhabekar, S.S. Sanaye, S.S. Shinde, S.V. Moharil, T.K. Gundu Rao,

Radiat. Prot. Dosim. 100 (2002) 251.[13] P.D. Townsend, M. Maghrabi, B. Yang, Nucl. Instrum. Methods Phys. Res., Sect.

B 191 (2002) 767.[14] S.U. Gaikwad, R.R Mistry, R.A. Barve, R.R. Patil, S.V Moharil, Indian J. Pure Appl.

Phys. 51 (2013) 235.[15] G.S. Polymeris, O.M. Goudouri, E. Kontonasaki, K.M. Paraskevopoulos, N.

C Tsirliganis, G. Kitis, J. Phys. D: Appl. Phys 44 (2011) 395501.[16] M.S Kulkarni, D.R Mishra, D.N Sharma, Nucl. Instrum. Methods Phys. Res., Sect.

B 262 (2007) 348.[17] F. Sun, L. Jiao, J. Wu, Z. Yang, SYuan, G. Dai, Radiat. Prot. Dosim. 51 (1994) 183.

Fig. 7. Variations in relative percentage intensity of Li4P2O7 and Li2Cu2P6O18, PL intensity, TL and OSL sensitivity with respect to change in annealing temperature. (a) Plot ofrelative percentage of Li4P2O7 and Li2Cu2P6O18 versus annealing temperature. (b) Plot of PL emission intensity versus annealing temperature. (c) Plot of TL sensitivity versusannealing temperature. (d) Plot of OSL sensitivity versus annealing temperature.

R.A. Barve et al. / Journal of Luminescence 156 (2014) 25–29 29