blue upconversion in Nd3+:TeO2 glass, effect of modifiers and heat treatment on the fluorescence bands

Spectrochimica Acta Part A 74 (2009) 776–780

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

UV/blue upconversion in Nd3+ :TeO2 glass, effect of modifiers and heat treatment on the fluorescence bands R.K. Verma, K. Kumar, S.B. Rai ∗ Laser Spectroscopy Laboratory, Department of Physics, Banaras Hindu University, Varanasi 221005, India

a r t i c l e

i n f o

Article history: Received 14 November 2008 Received in revised form 23 July 2009 Accepted 7 August 2009 PACS: 42.70.Hj 42.70.Ce Keywords: Optical properties Upconversion ESA Luminescence Fluorescence intensity ratio (FIR) Temperature sensor

a b s t r a c t Upconversion (UC) emissions in UV/blue region have been observed in Nd3+ doped tellurite glass on 532 nm excitation. The UC bands have been observed at 360, 387, 417 and 452 nm due to the 4 D3/2 → 4 I9/2 , 4 D3/2 → 4 I11/2 , 4 D3/2 → 4 I13/2 and 4 D3/2 → 4 I15/2 transitions, respectively and they show two photon character. The effect of BaCO3 , BaF2 and BaCl2 glass modifiers on the UC efficiency has been studied and Judd-Ofelt intensity parameters have been calculated and compared. The BaCl2 modified glass showed maximum UC intensity among the three modifiers and this enhancement in UC intensity has been related to the reduction in average phonon frequency of the glass sample. Heat treatments of the BaF2 and BaCl2 modified samples also show enhancement in UC intensity while the BaCO3 modified sample has no such effect. Lifetime of the 4 D3/2 level has been measured to understand the mechanism responsible for UC emission. Temperature dependent fluorescence studies have been done on the 4 F3/2 , 4 F5/2 and 2 S3/2 emitting levels and results show that Nd3+ doped tellurite glass can be used as a temperature sensor. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Rare earth ions doped in glass hosts have long history in developing solid state lasers and other optical devices [1–5], since these ions have favorable energy levels and transitions extends from NIR to visible region of the optical spectrum. The earlier devices based on rare earth ions were utilized the downconversion emissions property but from last two decades the efforts are concentrated on the development of lasers and other devices based on the upconversion (UC) emission processes. Applications of UC lasers include display devices, optical and thermal sensors, on-line amplifiers etc. [6–8]. The UC efficiency in a particular rare earth ion depends both on its own energy level scheme and on the host in which it is doped. The Nd3+ ion has densely packed energy levels and hence it is very difficult to achieve UC emissions due to large nonradiative relaxation rates. However, when the Nd3+ ions are doped in low phonon hosts it is possible to observe UC emissions. Little attempts have been made to observe the UC emission from Nd3+ ions using visible pump radiations [9–12] because of above said difficulty. Neodymium doped oxide glasses like silicates, borates; phosphates 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: [email protected] (S.B. Rai). 1386-1425/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2009.08.017

phonon frequencies. Bullock et al. [13] have shown the possibility of observing the UV upconversion emission from Nd3+ ions using 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 nm excitations. The frequency upconversion mediated by phonons has been analyzed by Auzel [15] and they have shown that it is possible to get the luminescence, even when the energy difference between the exciting radiation frequency and the level separation is larger than the maximum phonon frequency of the host material. Kumar et al. [16] have observed phonon assisted upconversion in tellurite glass on 800 nm excitation. Several rare-earth ions also possess thermally coupled levels which can be used as fluorescence temperature sensors [17–18]. Nd3+ ion has 4 F3/2 and 4 F5/2 levels that lie close to each other and can be used to measure the temperature. 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 material for optoelectronics. In this work, we have chosen tellurite glass as a host and BaCO3 , BaF2 and BaCl2 as glass modifiers. Effect of these modifiers on the UC intensity has been studied. We have also studied the effect of heat treatment on the emission intensity.

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

The downconversion emission studies show that two thermally coupled levels 4 F3/2 and 4 F5/2 can be used as temperature sensor. 2. Experiment The glass samples for present investigations have the following composition(80 − x)TeO2 + 20BaCO3 + xNd2 O3 (80 − x)TeO2 + 20BaF2 + xNd2 O3 (80 − x)TeO2 + 20BaCl2 + xNd2 O3 where x varies from 0.2, 0.4, 0.6, 0.75 and 1.0 mol%. The raw materials used were 99.9% pure. The melt-quench technique was used for the preparation of samples [19]. To produce nano-crystallites in the glass matrix, samples were heated at 330 ◦ C for different time intervals. The absorption spectra of the samples were measured using Cary Varian 500 UV–vis-NIR spectrometer in the 400–1100 nm region. FTIR spectra of the samples were recorded in 400–4000 cm−1 region using PerkinElmer Spectrum RX1. Second harmonic of Nd:YAG laser (Verdi V5, CW, Coherent) was used for fluorescence excitation. For lifetime measurements the second harmonic of a pulsed Nd:YAG laser (spitlight600, 7 ns pulse width, Innolas, Germany) was used. 3. Results and discussion 3.1. Absorption spectra and Judd-Ofelt analysis Samples with different concentrations of Nd3+ ion were prepared and quenching concentration of Nd3+ was searched by observing the change in UC intensity. Optimum intensity was observed for 0.75 mol% Nd3+ ion concentration. The ion concentration was fixed at this value for absorption and UC studies. The absorption spectra of samples modified with the three different modifiers were recorded in the 400–1100 nm region and used in Judd-Ofelt analysis [20,21]. The absorption spectra were found similar for the three modifiers with a little change in intensities of the absorption bands. Assignment of the absorption bands and their energies are given in Table 1. These transitions are within the levels arising from the 4fn configuration of rare earth ion and can be analyzed by Judd-Ofelt theory. The intensity of these transitions can be expressed in term of oscillator strength (fexp ), as:



fexp = 4.32 × 109

ε () d

(1)

where ␧(␯) is the molar absorptivity at frequency v (cm−1 ). According to Judd-Ofelt theory, the oscillator strength of electric dipole transition can be given as fcal =

82 mc (n2 + 2) 9n 3h(2J + 1)



2

˝ (j ||U  ||

j )

2

(2)

=2,4,6

Table 1 Observed absorption bands and their energies in barium modified tellurite glass. S. no.

Transition

Energy (cm−1 )

1 2 3 4 5 7 8 9 10 11

4

F3/2 ← I9/2 F5/2 + 2 H9/2 ← 4 I9/2 2 S3/2 + 4 F7/2 ← 4 I9/2 4 F9/2 ← 4 I9/2 2 H11/2 ← 4 I9/2 4 G5/2 + 2 G7/2 ← 4 I9/2 4 G7/2 ← 4 I9/2 4 G9/2 ← 4 I9/2 2 K18/2 + 2 G9/2 + 4 G11/2 ← 4 I9/2 2 P1/2 ← 4 I9/2

11273 12308 13843 14715 15757 16945 18843 20879 21219 23064

4

4

777

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

˝2 (×10−20 cm2 )

˝4 (×10−20 cm2 )

˝6 (×10−20 cm2 )

BaCO3 BaF2 BaCl2

5.6459 6.3792 4.4634

7.5614 8.8564 5.4028

6.3773 5.6630 4.1293

In the above equation m represents the electron mass, c is the speed of light, h Planck constant, n is refractive index and  is the transition frequency in wave number. As mentioned earlier the intensity of a transition depends on the ion as well as on the host matrix. The square term in the bracket represents the matrix element which is considered to be independent of host matrix. The oscillator strength of different transitions was calculated from the absorption data, and the three intensity parameters ˝␭ ( = 2, 4, 6) were obtained. These values are given in Table 2. From the table it is observed that ˝6 decreases monotonically when BaCO3 is replaced by BaF2 and BaCl2 successively and this trend indicates a decrease in covalence of Nd–X bond. 3.2. UC emission The UC spectra of Nd3+ doped tellurite glass have been recorded using 532 nm excitation. The optimum fluorescence intensity has been observed for sample doped with 0.75 mol% of Nd3+ ion. The observed emission bands with peak assignments are given below: 360 nm: 4 D3/2 → 4 I9/2 387 nm: 4 D3/2 → 4 I11/2 417 nm: 4 D3/2 → 4 I13/2 452 nm: 4 D3/2 → 4 I15/2 The upconversion intensity is observed maximum for 4D 4 3/2 → I13/2 transition (Fig. 1(a)). The figure also compares the UC spectra of samples with different modifiers. The overall fluorescence intensity is found maximum for sample modified with BaCl2 . The effect heat treatment on BaF2 modified glass has been shown in Fig. 2(b) which will be discussed latter. In order to get the knowledge about UC process, the dependence of UC emission intensity with incident pump power P was studied. The UC intensity (Iup ) vary with the mth power of the pump power P according to the relation Iup = Pm , where m is the number of pump photons absorbed per UV–vis photon emitted. The slope of ln–ln plots shows a clear quadratic dependence of fluorescence intensity upto a definite excitation power (Fig. 2(a)). On increasing the pump power above this limit a rapid decrease in intensity is observed for all the UC bands without showing the saturation. It is thought that this behavior results due to migration of population from 4 F3/2 level to the very close higher excited levels (viz. 4 F5/2 , 2H 2 4 9/2 , S3/2 , F7/2 ) through phonon absorption. The pump intensity generates temperature at incident area and at around a particular temperature the ions in 4 F3/2 level starts absorbing the energy from lattice and reaches to the higher close lying levels. The result decreases ion population in 4 F3/2 level which is intermediate level for the UC channel. Lifetime of the upper level of UC transition (4 D3/2 ) has been measured with pulsed 532 nm excitation. The decay curves obtained do not show detectable risetime. Since risetime differentiates between ESA and ET and absence of any risetime confirms the ESA process through phonon relaxation for all UC bands. The fluorescence emission lifetime is given as 1/ = WR + WMP + WCR + WETU

(3)

Where  is the lifetime of the level and WR , WMP , WCR , and WETU are the radiative transition probability of energy transfer, multiphonon relaxation, cross relaxation and energy transfer upconvesion, 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 emission intensity (b).

Fig. 2. ln–ln plot between excitation power and upconversion intensity of 4 D3/2 → 4 I13/2 band (a) and decay curve of 4 D3/2 level on excitation with 532 nm radiation (b).

of WETU and WCR terms are negligibly small and can be neglected. The absence of WETU and WCR rates decay curve shows exponential behavior. The measured lifetime of 4 D3/2 level with BaCl2 , BaF2 and BaCO3 modifiers were 340, 330 and 315 ␮s, respectively. The decay pattern of 4 D3/2 level with 532 nm excitation has been shown in Fig. 2(b). The curve is single exponential with good fit parameters. This figure clearly shows no risetime that support ESA process. The mechanism of upconversion emission has been shown schematically in Fig. 3. The most probable ESA through phonon relaxation pathway for the observed excitation is: 4I 4 4 9/2 + h → G7/2 → F3/2 + phonons emission 4 4 4F 3/2 + h → D7/2 → D3/2 + phonons emission 4D 4 I + photon emission → 3/2 ij i.e. on absorption with 532 nm photon, the Nd3+ ions are excited from ground level to the 4 G7/2 level. The excited ions then relax to 4 4F 3/2 level through multiphonon emission. The F3/2 is a metastable level and ions in this level reabsorb the incident photon and promoted to 4 D7/2 level. The ions from this level decay nonradiatively to 4 D3/2 level from which the UC emission bands are observed. 3.3. Effect of heat treatment The enhancement in optical properties of rare earth ions on heat treatment of samples is well known. We have studied the effect of heat treatment on the emission properties of Nd3+ ions with different modifiers on excitation with 532 nm. The samples were heat treated at around glass transition temperature for different time intervals 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

Fig. 4. Fluorescence spectra on Nd3+ doped BaF2 modified glass at different temperatures (a) and plot of fluorescence intensity ratio for emission bands centre at 814 and 874 nm as a function of temperature on a monolog scale (b).

spectra of heat treated samples with as-prepared sample for BaF2 modified glass. From the figure it is clear that the fluorescence intensity increases with the increase of heating time upto 6 h. On increasing the heating time further, a reduction in the emission intensity is observed due to decrease in transparency of the glass and increased excitation light scattering. A similar thing is observed with the BaCl2 also. However glass with BaCO3 modifier shows no effect of heating on the fluorescence intensity. A lifetime measurements also show an increase in lifetime of 4 D3/2 level after heat treatments in case of BaF2 as well as BaCl2 modifiers. The intensity enhancement in case of BaF2 and BaCl2 is expected to be due to the possible nano volume crystallization inside the glass but in case of BaCO3 crystallization does not occur. To detect the possible crystallite formation in the glass network, XRD patterns of these halide samples were monitored but no crystallization peak was observed in any of the two. There may be two possibilities for this; either the crystallite concentration is very low or there is no definite crystal phase formation rather than phase separation. It seems that second fact is true because it is very difficult to form crystals in the tellurite glass using alkaline earth modifiers [22]. This gives a strong probably of phase separation. On heating the samples, the fluoride/chloride ions got surround the rare earth ions and hence create a low phonon environment around the rare earth ions which may be the reason for enhancement of fluorescence intensity. This phase separation is not observed in the case of BaCO3 modifier [22]. An another reason for increase in UC intensity in BaF2 and BaCl2 samples may be the reduction in OH content in the prepared samples. The heat treatments reduce the OH content in the samples which has been confirmed from the FTIR studies. FTIR spectra clearly show a decrease in intensity of OH vibration band around 3600 cm−1 with the increase in heating time.

of the sample increases, the intensity of 874 nm band decreases while the intensity of 814 nm band increases but the rate of increase of 814 nm band intensity is slow than the rate of decrease of 874 nm band intensity. It shows a thermalization of the levels at higher temperatures as mentioned earlier. Since energy separation between the 2 S3/2 (4 F7/2 ), 4 F5/2 (2 H9/2 ) and 4 F3/2 levels is small so the two levels follow Boltzmann distribution law and can be used to measure the temperature of the sample. The 2 S3/2 (4 F7/2 ) level also lies ∼830 cm−1 above the 4 F5/2 level and only one phonon is needed to fill this gap. As the temperature of the sample increases ions in the 4F 3/2 level absorb thermal energy from the sample and increase the populations in the 2 H9/2 , 4 F5/2 and 4 S3/2 , 4 F7/2 levels. The intensity ratio of the two bands with temperature is shown in Fig. 4(b). The fluorescence intensity ratios (FIR) for 4 F3/2 and 4 F5/2 levels have been calculated at different temperatures. The population of such levels follows the Boltzmann distribution and FIR can be expressed as [18].

3.4. Temperature effect on fluorescence emission The effect of temperature on the Stokes emission bands has also been studied. At room temperature two bands are observed at 814 and 874 nm due to the 4 F5/2 → 4 I9/2 and 4 F3/2 → 4 I9/2 transitions, respectively. As temperature of the sample increases a new band appears at 755 nm whose intensity increases with the increase in temperature. This new peak has been assigned to the 2S 4 4 3/2 ( F7/2 ) → I9/2 transition. The spectra taken at different temperatures are shown in Fig. 4(a). At room temperature the 874 nm band is intense compared to the 814 nm band. When temperature

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

780

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

The fluorescence intensity emitted (4 F3/2 → 4 I9/2 ) at temperature T is given as I1,0 = g1,0 1,0 ω1,0 exp(−E1 /KT ) And from level 2 to 0

(4 F5/2

→ 4 I9/2 )

from

level

1to

0 (4)

of small sized crystallites. Also the temperature sensing capability of two thermally coupled level of Nd3+ ion doped in tellurite glass has been investigated on the basis of fluorescence intensity ratio technique.

is

I2,0 = g2,0 2,0 ω2, 0 exp(−E2 /KT )

(5)

The fluorescence intensity ratio (FIR) in such a case is given as FIR = I2,0 /I1,0 = N2 /N1 = B exp[−E/KT ]

(6)

where Ni , Iij are the number of ions, the fluorescence intensity from the upper (i = 2) and lower (i = 1) thermalizing levels to a terminal level (i = 0) g is the degeneracy of the levels involved, B is a constant and E is energy difference between two levels. The other terms have their usual meanings. In the present case the value of B is found to 5.95. Using the intensity ratios of the two bands from Eq. (5) the temperature of the source has been calculated. The temperature of glass was also measured using thermocouple and a good agreement has been observed. The fluorescence intensity ratio for emission bands centered at 814 and 874 nm as a function of temperature is shown in Fig. 4(b). The slope of the graph is very large which indicates its high sensitivity. Thus fluorescence detection scheme allows a real time measurement of temperature. One possible use of this sensor is for biological applications. Fig. 5 shows the schematic mechanism responsible for temperature dependent fluorescence emission. 4. Conclusions Nd3+ doped tellurite glass was synthesized using melt quench method and the effects of different barium compounds as modifiers have been analyzed. The BaCl2 modified glass sample has been found best among the modifiers studied. The Heat treated samples with BaF2 modifier however yields maximum fluorescence intensity. The increment in emission intensity was explained as the cumulative effect of reduction of OH impurity and formation

Acknowledgements Authors are grateful to DST, New Delhi, India for financial assistance. One of us (R. K. Verma) would like to thank Banaras Hindu University, Varanasi for providing scholarship. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

A.A. Kaminskii, Laser Crystals, Springer, Berlin, 1990. S.A. Wade, S.F. Collins, G.W. Baxter, J. Appl. Phys. 94 (2003) 4743. P. Xie, T.R. Gosnell, Opt. Lett. 20 (1995) 1014. R. Rolli, K. Gatterer, M. Wachtler, M. Bettinelli, A. Speghini, D. Ajo, Spectrochim. Acta A 57 (2001) 2009. A. Renuka Devi, C.K. Jayasankar, Mat. Chem. Phys. 42 (1995) 106. E. Downing, L. Hesselink, J. Ralson, R. Macfarlane, Science 273 (1996) 1185. M. Yamada, H. Ono, T. Kanamori, S. Sudo, Y. Ohishi, Electron. Lett. 33 (1997) 710. M. Naftaly, A. Jha, J. Appl. Phys. 87 (2000) 2098. L. Edilson, E.L. Falcao-Filho, C.B. de Araujo, Y. Messaddeq, J. Appl. Phys. 92 (2002) 3065. L.S. de Menezes, C.B. de Araujo, Y. Messaddeq, M.A. Aegerter, J. Non-Cryst. Solids 213–214 (1997) 256. L.S. de Menezes, C.B. de Araujo, G.S. Maciel, Y. Messaddeq, M.A. Aegerter, Appl. Phys. Lett. 70 (1997) 683. K. Kumar, S.B. Rai, D.K. Rai, Solid State Commun. 139 (2006) 363. S.R. Bullock, B.R. Reddy, C.L. Pope, S.K. Nash-Stevenson, J. Non-Cryst. Solids 212 (1997) 85. K. Kumar, S.B. Rai, Solid State Commun. 142 (2007) 58. F. Auzel, Phys. Rev. B 13 (1976) 2809. K. Kumar, D.K. Rai, S.B. Rai, Eur. Phys. J. Appl. Phys. 41 (2008) 143. V.K. Rai, S.B. Rai, Appl. Phys. B 87 (2007) 323. A.K. Singh, S.B. Rai, Appl. Phys. B 86 (2007) 66. K. Kumar, S.B. Rai, D.K. Rai, Spectrochim. Acta A 66 (2007) 1052. B.R. Judd, Phys. Rev. 127 (1962) 750. G.S. Ofelt, J. Chem. Phys. 37 (1962) 511. N.J. Kreidl, Glass: Scicence and Technology, vol. 1, Academic Press, 1983.