Ut

RL

a

ARRA

P44

KOUELFT

1

oitoptvd[

oditlaidp

1d

Spectrochimica Acta Part A 74 (2009) 776–780

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular andBiomolecular Spectroscopy

journa l homepage: www.e lsev ier .com/ locate /saa

V/blue upconversion in Nd3+:TeO2 glass, effect of modifiers and heatreatment on the fluorescence bands

.K. Verma, K. Kumar, S.B. Rai ∗

aser Spectroscopy Laboratory, Department of Physics, Banaras Hindu University, Varanasi 221005, India

r t i c l e i n f o

rticle history:eceived 14 November 2008eceived in revised form 23 July 2009ccepted 7 August 2009

ACS:2.70.Hj2.70.Ce

a b s t r a c t

Upconversion (UC) emissions in UV/blue region have been observed in Nd3+ doped tellurite glass on532 nm excitation. The UC bands have been observed at 360, 387, 417 and 452 nm due to the 4D3/2→ 4I9/2,4D3/2→ 4I11/2, 4D3/2→ 4I13/2 and 4D3/2→ 4I15/2 transitions, respectively and they show two photon char-acter. The effect of BaCO3, BaF2 and BaCl2 glass modifiers on the UC efficiency has been studied andJudd-Ofelt intensity parameters have been calculated and compared. The BaCl2 modified glass showedmaximum UC intensity among the three modifiers and this enhancement in UC intensity has been relatedto the reduction in average phonon frequency of the glass sample. Heat treatments of the BaF2 and BaCl2

eywords:ptical propertiespconversionSAuminescence

modified samples also show enhancement in UC intensity while the BaCO3 modified sample has nosuch effect. Lifetime of the 4D3/2 level has been measured to understand the mechanism responsible forUC emission. Temperature dependent fluorescence studies have been done on the 4F3/2, 4F5/2 and 2S3/2

emitting levels and results show that Nd3+ doped tellurite glass can be used as a temperature sensor.© 2009 Elsevier B.V. All rights reserved.

luorescence intensity ratio (FIR)emperature sensor

. Introduction

Rare earth ions doped in glass hosts have long history in devel-ping solid state lasers and other optical devices [1–5], since theseons have favorable energy levels and transitions extends from NIRo visible region of the optical spectrum. The earlier devices basedn rare earth ions were utilized the downconversion emissionsroperty but from last two decades the efforts are concentrated onhe development of lasers and other devices based on the upcon-ersion (UC) emission processes. Applications of UC lasers includeisplay devices, optical and thermal sensors, on-line amplifiers etc.6–8].

The UC efficiency in a particular rare earth ion depends bothn its own energy level scheme and on the host in which it isoped. The Nd3+ ion has densely packed energy levels and hence

t is very difficult to achieve UC emissions due to large nonradia-ive relaxation rates. However, when the Nd3+ ions are doped inow phonon hosts it is possible to observe UC emissions. Little

ttempts have been made to observe the UC emission from Nd3+

ons using visible pump radiations [9–12] because of above saidifficulty. Neodymium doped oxide glasses like silicates, borates;hosphates etc. do not show UC emission because of their high

∗ Corresponding author. Tel.: +91 542 230 7308; fax: +91 542 236 9889.E-mail address: sbrai49@yahoo.co.in (S.B. Rai).

386-1425/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.saa.2009.08.017

phonon frequencies. Bullock et al. [13] have shown the possibil-ity of observing the UV upconversion emission from Nd3+ ionsusing heavy modifiers in the oxide hosts. de Menezes et al. [11]have observed UV/blue UC in Nd3+ doped fluorozirconate glass.Kumar and Rai [14] have reported the UC emissions in Nd3+

doped lithium oxide modified tellurite glass with 532 and 800 nmexcitations.

The frequency upconversion mediated by phonons has beenanalyzed by Auzel [15] and they have shown that it is possi-ble to get the luminescence, even when the energy differencebetween the exciting radiation frequency and the level separationis larger than the maximum phonon frequency of the host material.Kumar et al. [16] have observed phonon assisted upconversion intellurite glass on 800 nm excitation. Several rare-earth ions alsopossess thermally coupled levels which can be used as fluores-cence temperature sensors [17–18]. Nd3+ ion has 4F3/2 and 4F5/2levels that lie close to each other and can be used to measure thetemperature.

Since, tellurite glass has much lower phonon frequency(∼800 cm−1), it is expected to observe the UC emission in Nd3+

ions when doped in this host. Also, tellurite glass possesses high

chemical durability and thermal stability. It is an excellent materialfor optoelectronics. In this work, we have chosen tellurite glassas a host and BaCO3, BaF2 and BaCl2 as glass modifiers. Effect ofthese modifiers on the UC intensity has been studied. We havealso studied the effect of heat treatment on the emission intensity.

ica Acta Part A 74 (2009) 776–780 777

Tc

2

c

0

nnfwtiofhI

3

3

pootamJiaeeac

f

w

d

f

TO

Table 2Judd-Ofelt intensity parameters for Nd3+ in TeO2 glass with different modifiers.

Modifier ˝2 (×10−20 cm2) ˝4 (×10−20 cm2) ˝6 (×10−20 cm2)

R.K. Verma et al. / Spectrochim

he downconversion emission studies show that two thermallyoupled levels 4F3/2 and 4F5/2 can be used as temperature sensor.

. Experiment

The glass samples for present investigations have the followingomposition-

(80− x)TeO2 + 20BaCO3 + xNd2O3(80− x)TeO2 + 20BaF2 + xNd2O3(80− x)TeO2 + 20BaCl2 + xNd2O3where x varies from 0.2, 0.4, 0.6,

.75 and 1.0 mol%.The raw materials used were 99.9% pure. The melt-quench tech-

ique was used for the preparation of samples [19]. To produceano-crystallites in the glass matrix, samples were heated at 330 ◦C

or different time intervals. The absorption spectra of the samplesere measured using Cary Varian 500 UV–vis-NIR spectrometer in

he 400–1100 nm region. FTIR spectra of the samples were recordedn 400–4000 cm−1 region using PerkinElmer Spectrum RX1. Sec-nd harmonic of Nd:YAG laser (Verdi V5, CW, Coherent) was usedor fluorescence excitation. For lifetime measurements the secondarmonic of a pulsed Nd:YAG laser (spitlight600, 7 ns pulse width,

nnolas, Germany) was used.

. Results and discussion

.1. Absorption spectra and Judd-Ofelt analysis

Samples with different concentrations of Nd3+ ion were pre-ared and quenching concentration of Nd3+ was searched bybserving the change in UC intensity. Optimum intensity wasbserved for 0.75 mol% Nd3+ ion concentration. The ion concen-ration was fixed at this value for absorption and UC studies. Thebsorption spectra of samples modified with the three differentodifiers were recorded in the 400–1100 nm region and used in

udd-Ofelt analysis [20,21]. The absorption spectra were found sim-lar for the three modifiers with a little change in intensities of thebsorption bands. Assignment of the absorption bands and theirnergies are given in Table 1. These transitions are within the lev-ls arising from the 4fn configuration of rare earth ion and can benalyzed by Judd-Ofelt theory. The intensity of these transitionsan be expressed in term of oscillator strength (fexp), as:

exp = 4.32× 109

∫ε (�)d� (1)

here �(�) is the molar absorptivity at frequency v (cm−1).According to Judd-Ofelt theory, the oscillator strength of electric

ipole transition can be given as

cal =8�2mc�

3h(2J + 1)(n2 + 2)

2

9n

∑�=2,4,6

˝�(�j||U�|| j′ )2

(2)

able 1bserved absorption bands and their energies in barium modified tellurite glass.

S. no. Transition Energy (cm−1)

1 4F3/2← 4I9/2 112732 4F5/2 + 2H9/2← 4I9/2 123083 2S3/2 + 4F7/2← 4I9/2 138434 4F9/2← 4I9/2 147155 2H11/2← 4I9/2 157577 4G5/2 + 2G7/2← 4I9/2 169458 4G7/2← 4I9/2 188439 4G9/2← 4I9/2 2087910 2K18/2 + 2G9/2 + 4G11/2← 4I9/2 2121911 2P1/2← 4I9/2 23064

BaCO3 5.6459 7.5614 6.3773BaF2 6.3792 8.8564 5.6630BaCl2 4.4634 5.4028 4.1293

In the above equation m represents the electron mass, c is thespeed of light, h Planck constant, n is refractive index and � is thetransition frequency in wave number. As mentioned earlier theintensity of a transition depends on the ion as well as on the hostmatrix. The square term in the bracket represents the matrix ele-ment which is considered to be independent of host matrix. Theoscillator strength of different transitions was calculated from theabsorption data, and the three intensity parameters˝�(�= 2, 4, 6)were obtained. These values are given in Table 2. From the table it isobserved that˝6 decreases monotonically when BaCO3 is replacedby BaF2 and BaCl2 successively and this trend indicates a decreasein covalence of Nd–X bond.

3.2. UC emission

The UC spectra of Nd3+ doped tellurite glass have been recordedusing 532 nm excitation. The optimum fluorescence intensity hasbeen observed for sample doped with 0.75 mol% of Nd3+ ion. Theobserved emission bands with peak assignments are given below:

360 nm: 4D3/2→ 4I9/2387 nm: 4D3/2→ 4I11/2417 nm: 4D3/2→ 4I13/2452 nm: 4D3/2→ 4I15/2The upconversion intensity is observed maximum for

4D3/2→ 4I13/2 transition (Fig. 1(a)). The figure also comparesthe UC spectra of samples with different modifiers. The overallfluorescence intensity is found maximum for sample modifiedwith BaCl2. The effect heat treatment on BaF2 modified glass hasbeen shown in Fig. 2(b) which will be discussed latter. In orderto get the knowledge about UC process, the dependence of UCemission intensity with incident pump power P was studied. TheUC intensity (Iup) vary with the mth power of the pump powerP according to the relation Iup = Pm, where m is the number ofpump photons absorbed per UV–vis photon emitted. The slope ofln–ln plots shows a clear quadratic dependence of fluorescenceintensity upto a definite excitation power (Fig. 2(a)). On increasingthe pump power above this limit a rapid decrease in intensity isobserved for all the UC bands without showing the saturation. It isthought that this behavior results due to migration of populationfrom 4F3/2 level to the very close higher excited levels (viz. 4F5/2,2H9/2, 2S3/2, 4F7/2) through phonon absorption. The pump intensitygenerates temperature at incident area and at around a particulartemperature the ions in 4F3/2 level starts absorbing the energyfrom lattice and reaches to the higher close lying levels. The resultdecreases ion population in 4F3/2 level which is intermediate levelfor the UC channel.

Lifetime of the upper level of UC transition (4D3/2) has been mea-sured with pulsed 532 nm excitation. The decay curves obtained donot show detectable risetime. Since risetime differentiates betweenESA and ET and absence of any risetime confirms the ESA pro-cess through phonon relaxation for all UC bands. The fluorescenceemission lifetime is given as

1/� =WR +WMP +WCR +WETU (3)

Where � is the lifetime of the level and WR, WMP, WCR, andWETU are the radiative transition probability of energy trans-fer, multiphonon relaxation, cross relaxation and energy transferupconvesion, respectively. At low concentrations the contributions

778 R.K. Verma et al. / Spectrochimica Acta Part A 74 (2009) 776–780

Fig. 1. Comparison in upconversion emission spectra of samples modified with three modifiers (a) and effect of heat treatment of samples on upconversion emissionintensity (b).

F 4I13/2 band (a) and decay curve of 4D3/2 level on excitation with 532 nm radiation (b).

oTbBpFT

sr

f4

lpro

3

thfti

ig. 2. ln–ln plot between excitation power and upconversion intensity of 4D3/2→

f WETU and WCR terms are negligibly small and can be neglected.he absence of WETU and WCR rates decay curve shows exponentialehavior. The measured lifetime of 4D3/2 level with BaCl2, BaF2 andaCO3 modifiers were 340, 330 and 315 �s, respectively. The decayattern of 4D3/2 level with 532 nm excitation has been shown inig. 2(b). The curve is single exponential with good fit parameters.his figure clearly shows no risetime that support ESA process.

The mechanism of upconversion emission has been shownchematically in Fig. 3. The most probable ESA through phononelaxation pathway for the observed excitation is:

4I9/2 + h�→ 4G7/2→ 4F3/2 + phonons emission4F3/2 + h�→ 4D7/2→ 4D3/2 + phonons emission4D3/2→ 4Iij + photon emissioni.e. on absorption with 532 nm photon, the Nd3+ ions are excited

rom ground level to the 4G7/2 level. The excited ions then relax toF3/2 level through multiphonon emission. The 4F3/2 is a metastableevel and ions in this level reabsorb the incident photon andromoted to 4D7/2 level. The ions from this level decay non-adiatively to 4D3/2 level from which the UC emission bands arebserved.

.3. Effect of heat treatment

The enhancement in optical properties of rare earth ions on heat

reatment of samples is well known. We have studied the effect ofeat treatment on the emission properties of Nd3+ ions with dif-

erent modifiers on excitation with 532 nm. The samples were heatreated at around glass transition temperature for different timentervals to form glass ceramics. Fig. 1(b) compares the emission Fig. 3. Energy level diagram of Nd 3+ and frequency upconversion pathways.

R.K. Verma et al. / Spectrochimica Acta Part A 74 (2009) 776–780 779

F rature8

smiiiawemtstcstwfitf[ietfltiOttsh

3

aasni2

pb

fluorescence intensity ratios (FIR) for 4F3/2 and 4F5/2 levels havebeen calculated at different temperatures. The population of suchlevels follows the Boltzmann distribution and FIR can be expressedas [18].

ig. 4. Fluorescence spectra on Nd3+ doped BaF2 modified glass at different tempe74 nm as a function of temperature on a monolog scale (b).

pectra of heat treated samples with as-prepared sample for BaF2odified glass. From the figure it is clear that the fluorescence

ntensity increases with the increase of heating time upto 6 h. Onncreasing the heating time further, a reduction in the emissionntensity is observed due to decrease in transparency of the glassnd increased excitation light scattering. A similar thing is observedith the BaCl2 also. However glass with BaCO3 modifier shows no

ffect of heating on the fluorescence intensity. A lifetime measure-ents also show an increase in lifetime of 4D3/2 level after heat

reatments in case of BaF2 as well as BaCl2 modifiers. The inten-ity enhancement in case of BaF2 and BaCl2 is expected to be dueo the possible nano volume crystallization inside the glass but inase of BaCO3 crystallization does not occur. To detect the pos-ible crystallite formation in the glass network, XRD patterns ofhese halide samples were monitored but no crystallization peakas observed in any of the two. There may be two possibilities

or this; either the crystallite concentration is very low or theres no definite crystal phase formation rather than phase separa-ion. It seems that second fact is true because it is very difficult toorm crystals in the tellurite glass using alkaline earth modifiers22]. This gives a strong probably of phase separation. On heat-ng the samples, the fluoride/chloride ions got surround the rarearth ions and hence create a low phonon environment aroundhe rare earth ions which may be the reason for enhancement ofuorescence intensity. This phase separation is not observed inhe case of BaCO3 modifier [22]. An another reason for increasen UC intensity in BaF2 and BaCl2 samples may be the reduction inH content in the prepared samples. The heat treatments reduce

he OH content in the samples which has been confirmed fromhe FTIR studies. FTIR spectra clearly show a decrease in inten-ity of OH vibration band around 3600 cm−1 with the increase ineating time.

.4. Temperature effect on fluorescence emission

The effect of temperature on the Stokes emission bands haslso been studied. At room temperature two bands are observedt 814 and 874 nm due to the 4F5/2→ 4I9/2 and 4F3/2→ 4I9/2 tran-itions, respectively. As temperature of the sample increases a

ew band appears at 755 nm whose intensity increases with the

ncrease in temperature. This new peak has been assigned to theS3/2 (4F7/2)→ 4I9/2 transition. The spectra taken at different tem-eratures are shown in Fig. 4(a). At room temperature the 874 nmand is intense compared to the 814 nm band. When temperature

s (a) and plot of fluorescence intensity ratio for emission bands centre at 814 and

of the sample increases, the intensity of 874 nm band decreaseswhile the intensity of 814 nm band increases but the rate of increaseof 814 nm band intensity is slow than the rate of decrease of 874 nmband intensity. It shows a thermalization of the levels at higher tem-peratures as mentioned earlier. Since energy separation betweenthe 2S3/2 (4F7/2), 4F5/2 (2H9/2) and 4F3/2 levels is small so the twolevels follow Boltzmann distribution law and can be used to mea-sure the temperature of the sample. The 2S3/2 (4F7/2) level also lies∼830 cm−1 above the 4F5/2 level and only one phonon is needed tofill this gap. As the temperature of the sample increases ions in the4F3/2 level absorb thermal energy from the sample and increase thepopulations in the 2H9/2, 4F5/2 and 4S3/2, 4F7/2 levels. The intensityratio of the two bands with temperature is shown in Fig. 4(b). The

Fig. 5. A partial energy level diagram of Nd3+ and the mechanism of fluorescenceemission with temperature.

7 ica Ac

(

I

I

F

wtlahttghbiiatme

4

mfifpit

[

[

[[

[[[[17] V.K. Rai, S.B. Rai, Appl. Phys. B 87 (2007) 323.

80 R.K. Verma et al. / Spectrochim

The fluorescence intensity emitted from level 1to 04F3/2→ 4I9/2) at temperature T is given as

1,0 = g1,01,0ω1,0 exp(−E1/KT) (4)

And from level 2 to 0 (4F5/2→ 4I9/2) is

2,0 = g2,02,0ω2,0 exp(−E2/KT) (5)

The fluorescence intensity ratio (FIR) in such a case is given as

IR = I2,0/I1,0 = N2/N1 = B exp[−�E/KT] (6)

here Ni, Iij are the number of ions, the fluorescence intensity fromhe upper (i = 2) and lower (i = 1) thermalizing levels to a terminalevel (i = 0) g is the degeneracy of the levels involved, B is a constantnd �E is energy difference between two levels. The other termsave their usual meanings. In the present case the value of B is foundo 5.95. Using the intensity ratios of the two bands from Eq. (5) theemperature of the source has been calculated. The temperature oflass was also measured using thermocouple and a good agreementas been observed. The fluorescence intensity ratio for emissionands centered at 814 and 874 nm as a function of temperature

s shown in Fig. 4(b). The slope of the graph is very large whichndicates its high sensitivity. Thus fluorescence detection schemellows a real time measurement of temperature. One possible use ofhis sensor is for biological applications. Fig. 5 shows the schematic

echanism responsible for temperature dependent fluorescencemission.

. Conclusions

Nd3+ doped tellurite glass was synthesized using melt quenchethod and the effects of different barium compounds as modi-

ers have been analyzed. The BaCl2 modified glass sample has beenound best among the modifiers studied. The Heat treated sam-les with BaF2 modifier however yields maximum fluorescence

ntensity. The increment in emission intensity was explained ashe cumulative effect of reduction of OH impurity and formation

[[[[[

ta Part A 74 (2009) 776–780

of small sized crystallites. Also the temperature sensing capabilityof two thermally coupled level of Nd3+ ion doped in tellurite glasshas been investigated on the basis of fluorescence intensity ratiotechnique.

Acknowledgements

Authors are grateful to DST, New Delhi, India for financial assis-tance. One of us (R. K. Verma) would like to thank Banaras HinduUniversity, Varanasi for providing scholarship.

References

[1] A.A. Kaminskii, Laser Crystals, Springer, Berlin, 1990.[2] S.A. Wade, S.F. Collins, G.W. Baxter, J. Appl. Phys. 94 (2003) 4743.[3] P. Xie, T.R. Gosnell, Opt. Lett. 20 (1995) 1014.[4] R. Rolli, K. Gatterer, M. Wachtler, M. Bettinelli, A. Speghini, D. Ajo, Spectrochim.

Acta A 57 (2001) 2009.[5] A. Renuka Devi, C.K. Jayasankar, Mat. Chem. Phys. 42 (1995) 106.[6] E. Downing, L. Hesselink, J. Ralson, R. Macfarlane, Science 273 (1996)

1185.[7] M. Yamada, H. Ono, T. Kanamori, S. Sudo, Y. Ohishi, Electron. Lett. 33 (1997)

710.[8] M. Naftaly, A. Jha, J. Appl. Phys. 87 (2000) 2098.[9] L. Edilson, E.L. Falcao-Filho, C.B. de Araujo, Y. Messaddeq, J. Appl. Phys. 92 (2002)

3065.10] L.S. de Menezes, C.B. de Araujo, Y. Messaddeq, M.A. Aegerter, J. Non-Cryst. Solids

213–214 (1997) 256.11] L.S. de Menezes, C.B. de Araujo, G.S. Maciel, Y. Messaddeq, M.A. Aegerter, Appl.

Phys. Lett. 70 (1997) 683.12] K. Kumar, S.B. Rai, D.K. Rai, Solid State Commun. 139 (2006) 363.13] S.R. Bullock, B.R. Reddy, C.L. Pope, S.K. Nash-Stevenson, J. Non-Cryst. Solids 212

(1997) 85.14] K. Kumar, S.B. Rai, Solid State Commun. 142 (2007) 58.15] F. Auzel, Phys. Rev. B 13 (1976) 2809.16] K. Kumar, D.K. Rai, S.B. Rai, Eur. Phys. J. Appl. Phys. 41 (2008) 143.

18] A.K. Singh, S.B. Rai, Appl. Phys. B 86 (2007) 66.19] K. Kumar, S.B. Rai, D.K. Rai, Spectrochim. Acta A 66 (2007) 1052.20] B.R. Judd, Phys. Rev. 127 (1962) 750.21] G.S. Ofelt, J. Chem. Phys. 37 (1962) 511.22] N.J. Kreidl, Glass: Scicence and Technology, vol. 1, Academic Press, 1983.