Concentration effect on the spectroscopic behavior of

Journal of Luminescence 165 (2015) 77–84

Contents lists available at ScienceDirect

Journal of Luminescence journal homepage: www.elsevier.com/locate/jlumin

Concentration effect on the spectroscopic behavior of Tb3 þ ions in zinc phosphate glasses C.R. Kesavulu a,n, Anielle Christine Almeida Silva b, M.R. Dousti a, Noelio Oliveira Dantas b, A.S.S. de Camargo a, Tomaz Catunda a a b

Instituto de Fisica de São Carlos, Universidade de São Paulo, Avenida Trabalhador Sãocarlense 400, São Carlos, SP, Brazil Laboratorio de Novos Materials Isolantes e Semicondutores (LNMIS), Instituto de Física, Universidade Federal de Uberândia, MG, Brazil

art ic l e i nf o

a b s t r a c t

Article history: Received 4 December 2014 Received in revised form 25 March 2015 Accepted 15 April 2015 Available online 23 April 2015

Zinc phosphate glasses (PZABPTb) in the compositional system: P2O5–ZnO–Al2O3–BaO–PbO doped with variable Tb3 þ concentrations (1–5 wt% Tb2O3) were prepared and characterized through absorption, excitation, emission and intensity decay rate measurements. The Judd–Ofelt model has been adopted to evaluate the radiative properties of the 5D4-7F6–3 emission transitions. The effect of Tb3 þ ion concentration on the emissions from the 5D3,4 excited levels is discussed in detail. Analysis of the intensity decay curves corresponding to blue and green emissions from levels 5D3 and 5D4, respectively, allowed determination of effective lifetimes, which confirmed the Tb3 þ ion concentration quenching of the blue emission in these glasses. The decay curves for the 5D3 level are found to be non-exponential in nature for all the studied concentrations due to ion–ion energy transfer through cross-relaxation. In an attempt to identify the origin of the energy transfer mechanism, the decay curves were well fitted to the Inokuti– Hirayama model for S ¼6, which indicates that the energy transfer process is of dipole–dipole type. The optical band gap energy (Eopt) has been evaluated taking into account the ultraviolet edge of absorption spectra. & 2015 Elsevier B.V. All rights reserved.

Keywords: Zinc phosphate glasses Tb3 þ ion Judd–Ofelt theory Luminescence Decay time

1. Introduction Trivalent rare-earth ions (RE3 þ ) present intense visible emissions as a consequence of 4f–4f transitions which are little sensitive to the ion's surroundings due to the shielding effect of outer 5s and 5p shell electrons [1–4]. Thus, these ions have been largely studied in fluorescent materials with diverse applications such as in fluorescent lamps, phosphors and up-conversion based devices [5]. Moreover, tunable fluorescence in semiconductors [6] and photochemical hole-burning [7] have been studied by doping RE3 þ ions in glasses. Materials activated by trivalent terbium (Tb3 þ ) are known as very good emitters of blue and green light under UV excitation of emitting levels 5D3 and 5D4 [8]. In oxide glasses, with high phonon energy, the emission lines corresponding to 5D4-7FJ (J¼6,5,4,3) transitions are favored around 490 (blue), 545 (green), 580 (yellow) and 620 nm (red), respectively. Because the intensity of the emission assigned to the 5D4-7F5 transition (545 nm) mostly dominates over all other emissions, the Tb3 þ luminescence usually appears green to the naked eye.

n

Corresponding author. Tel.: þ 55 16 33738093. E-mail address: [email protected] (C.R. Kesavulu).

http://dx.doi.org/10.1016/j.jlumin.2015.04.012 0022-2313/& 2015 Elsevier B.V. All rights reserved.

Among several possibilities of RE3 þ -doped materials, glasses are of special interest for active display and image processing panels due to their transparence, easy-shaping and cost-effective properties [9]. Particularly, phosphate glasses have attracted much attention given their unique physical properties such as high thermal expansion coefficients, low melting and softening temperatures, high electrical conductivities, extensive ultraviolet (UV) transmission, etc. [9,10]. Zinc phosphate glasses have been used in optical waveguides and solid state lasers, as well as solders and welds between glassy and metallic parts in electronic circuits and television tubes, because their thermal expansion coefficient is similar to that of many metals [11,12]. The introduction of the network intermediate [Al3 þ ] and the network modifier [Zn2 þ ] induces structural changes in the phosphate glass structure formation of non-bridging oxygen (NBOs). In addition, the presence of network modifiers such as Ba2 þ , Pb2 þ and Zn2 þ lead to the softening of the glass network. Such structural arrangement facilitates the incorporation of RE3 þ without formation of a large number of clusters, thereby resulting in a better statistical distribution of the dopant ions [13]. This fact, in addition to good chemical and mechanical stabilities, suggest that the zinc phosphate (PZABP) glass system could be a potentially good host for RE3 þ ions.

78

C.R. Kesavulu et al. / Journal of Luminescence 165 (2015) 77–84

In particular, the PZABP glasses exhibit high transparency from UV to the near infrared, with wide band gap energies and relatively low phonon energies (∼1100 cm 1) when compared to other phosphates. It is expected that 4fn-4fn 15d1 transitions of RE3 þ are observed with significant intensity and no interference from the host. Recently, Neto et al. [14] have shown that ZnTe nanocrystals (quantum dots and bulk-like) can be incorporated in PZABP glass matrices for optical device applications. In this paper, zinc phosphate glasses with the nominal composition of 60P2O5–15ZnO–5Al2O3–10BaO–10PbO (mol%) doped with Tb2O3 (1.0, 2.0, 3.0, 4.0 and 5.0 wt%, respectively) were studied to investigate the dopant ion concentration effect on spectroscopic properties. The Judd–Ofelt (J–O) theory [15,16] was applied to evaluate the radiative properties based on the phenomenological intensity parameters (Ωλ ¼ 2,4,6), which yield spontaneous transition probabilities (AR), luminescence branching ratios (βR) and the radiative lifetime (τrad). The luminescence decay curves of 5D3,4 excited levels and the energy transfer among the excited Tb3 þ ions are also analyzed and discussed in details.

2. Experimental studies The zinc phosphate glasses with the chemical compositions (mol%) 60P2O5–15ZnO–5Al2O3–10BaO–10PbO–xTb2O3, where x ¼ 1.0, 2.0, 3.0, 4.0 and 5.0 (wt%), were prepared by the conventional melt quenching technique. In the first step, 10 g batches of homogeneously mixed host composition were melted in alumina crucibles, at 1300 °C, for 30 min and under rich carbon atmosphere. The melts were quenched between metallic plates at 30 °C. The resulting pieces were then pulverized and sieved to yield powders with grains smaller than 53 mm. In the second step, batches of 5 g of the resulting micro powders, containing variable concentrations of Tb2O3, were melted in alumina crucibles at 1250 °C, for 30 min, in nitrogen gas atmosphere. The resulting melts were quenched between metallic plates at 0 °C. The obtained glasses were annealed at 380 °C for 12 h to relieve thermal stresses and strains and then slowly cooled to room temperature. The undoped host glass was also prepared as a reference. Finally, the glass samples were cut and polished for optical measurements. The refractive index (1.545) was determined using an Abbe refractometer at sodium wavelength (589.3 nm) with 1-bromon-

apthalene (C10H7Br) as a contact liquid. Density measurements were carried out by Archimedes' principle with xylene as immersion liquid. Absorption spectra were recorded on a UV–visible–NIR spectrophotometer (PerkinElmer, Lambda 1050). Emission, excitation and decay curves of different Tb3 þ -doped glasses have been recorded using a Horiba Fluorolog spectrofluorimeter with Xe arc lamp as an excitation source. All these measurements were carried out at room temperature.

3. Results and discussion 3.1. Absorption spectra – J–O intensity parameters The room temperature optical absorption spectra of PZABP:5Tb glass in UV–visible and near infrared spectral regions are shown in Fig. 1(a) and (b), respectively. As shown in Fig. 1(a), the absorption peaks centered at 352, 358, 368, 378 and 484 nm are assigned to the transitions from ground state 7F6 to the higher excited states 5 L9, 5G5, 5L10, (5D3,5G6) and 5D4, respectively, of 4f8 configuration of Tb3 þ ions. In the NIR region (Fig. 1(b)), the detected absorption bands at wavelengths 1894 and 2223 nm arise due to transitions from the ground state 7F6 to the excited states, 7F0,1,2 and 7F3, respectively. The observed transitions are due to the electric dipole interaction obeying the selection rules ΔS ¼0, Lr 6 and ΔJ r6 [17]. The spin forbidden (7F6-5D4) absorption band in the visible region is weak, while the rest of the absorption bands display moderate intensity. From the absorption spectrum, it is noticed that the bands corresponding to the 7F6-7F0, 7F6-7F1 and 7F6-7F2 transitions, overlap. The absorption band intensities are determined in terms of experimental oscillator strength (fexp) following Refs. [3,18]. Application of the J–O theory [15,16] requires room temperature absorption spectrum, because this theoretical model assumes that all sublevels of the ground state are equally populated. In order to obtain the J–O intensity parameters (Ωλ ¼ 2,4,6) the values of fexp are equaled to the expression for the theoretical oscillator strengths (fcal) (which contain Ωλ ¼ 2,4,6) and the parameters are obtained by fitting the system of equations with the least-square fit method. fexp and fcal are presented in Table 1. The small root-mean-square deviation value (δrms) of 70.23 10 6 obtained between the fexp and fcal values indicate the good fit. In the present work the values obtained for the phenomenological

Fig. 1. Optical absorption spectrum of PZABP:5Tb glass in (a) UV–vis and (b) NIR regions.


J–O parameters are Ω2 ¼ 7.75, Ω4 ¼1.13 and Ω6 ¼ 3.27 ( 10 20 cm2) are presented in Table 2 along with those reported for other Tb3 þ -doped systems [18–25]. In the present PZABP:5Tb glass, the J–O intensity parameters follow the trend as Ω2 4 Ω6 4 Ω4, which is similar to the trend observed in oxyfluorosilicate [19], BaF2–InF3–ThF4–ZnF2–NaF [20], borosilicate (SBNACZ) [21] and LBTAF [18] glasses. Many authors, including Jorgensen and Reisfeld [26], have proposed that the magnitude of Ω2 is not only an indicative of the extent of covalent bonding but also explains the local structure of the RE3 þ ion site. Compared to other Tb3 þ :systems [18–25], the present glass show higher value of Ω2 (7.75 10 20 cm2), indicating higher Tb–O covalence and higher asymmetry at the Tb3 þ ion site. Such larger site asymmetry suggests higher mixing of opposite parity electronic configurations, which are responsible for the high spectral intensities [27]. On the other hand, in the present study, the asymmetry around rare earth ions is lower than in BaF2–InF3–ThF4–ZnF2–NaF [20], LBTAF [18] and BaO–Al2O3–P2O5 [24] glasses. The Ω6 is related to the rigidity of the host material. From the value of Ω6 (3.27 10 20 cm2), in the present glass, it can be suggested that it possess less rigidity compared to oxyfluorosilicate [19], BaF2–InF3–ThF4–ZnF2–NaF [20] and fluorophosphate [25] glasses, but it possess higher rigidity than the other reported Tb3 þ -doped systems listed in Table 2. 3.2. Optical band gap energy (Eopt)

Table 1 Absorption band positions (λ, nm), experimental (fexp) and calculated (fcal) oscillator strengths ( 10 6) for 5.0 wt% Tb3 þ -doped PZABP:5Tb glass. Transition

λ (nm)

fexp

fcal

F6-5L9 F6-5G5 7 F6-5L10 7 F6-5D3,5G6 7 F6-5D4 7 F6-7F0,1,2 7 F6-7F3

352 358 368 378 484 1894 2223

0.21 0.03 0.09 0.21 0.11 2.02 0.84

0.51 0.18 0.61 0.17 0.03 2.01 0.84

7

concentration are shown graphically in Fig. 2(b). Eopt values are found to increase slightly with the increasing of Tb3 þ ion content in the present glasses (Fig. 2(b)) and it is due to the reduction of non-bridging oxygen's (NBO's). The decrease in NBO's lowers the valence band maximum which in turn leads to have an increase in the optical band gap values [30]. 3.3. Excitation spectrum Fig. 3 shows the excitation spectrum of PZABP:5Tb glass measured by monitoring the green emission of Tb3 þ ion at λeme ¼ 540 nm. As it can be seen, it is comprised of two intense peaks of f–f transitions, (i) a broad band centered at 376 nm attributed to superimposition of transitions 7F6-(5D3,5G6), 5L10, 5 G5, 5L9, (5L8,5L7), 5D1, 5H7,5H6 and (ii) a band at 484 nm assigned to 7F6-5D4 transition. In addition, another excitation band at 284 nm can be assigned to the spin forbidden 9D levels of 4f8-4f75d1 transition of Tb3 þ ion. Among these, the (5D3,5G6) transition is found to be most intense which will be excited to record the emission spectra of Tb3 þ -doped PZABP glasses. All these excitation transitions are analogous to those observed in optical absorption spectrum, except an excitation band at 284 nm (Fig. 1(a)). Similar excitation spectra were obtained in several Tb3 þ doped glasses. For instance, dos Santos et al. [31] observed a 4f75d1 transition at 267 nm. 3.4. Luminescence spectra – radiative properties

The optical energy band gap (Eopt) values were determined by plotting (αhν)1/2 as a function of photon energy hν for un-doped and different concentration of Tb2O3 doped samples, respectively. One can find the Eopt for indirect transitions by extrapolating the linear region of the curve to the hν axis [28] as shown in Fig. 2(a). The optical band gap energies of the present work lie in the range of 4.33–4.49 eV for undoped and 5 wt% Tb3 þ -doped PZABP glasses, respectively. This is small, but has the same order as those reported in literature for Tb3 þ -doped calcium aluminosilicate glasses [29]. The Eopt values as a function of varying Tb3 þ ion

7

79

RMS deviation (δrms) ¼ 7 0.23 10 6

Table 2 Comparison of Judd–Ofelt parameters ( 10 20 cm2) and their trend in Tb3 þ ions in different hosts. Host

Ω2

Ω4

Ω6

Trend

PZABP:5Tb (present work) Oxyfluorosilicate [19] BaF2–InF3–ThF4–ZnF2–NaF [20] Borosilicate (SBNACZ) [21] LBTAF [18] Phosphate (P–5Tb) [22] Sodium fluoro-borate [23] BaO–Al2O3–P2O5 [24] Fluorophosphate [25]

7.75 4.57 12.32 6.26 13.23 0.66 0.45 24 2.75

1.13 0.42 1.90 2.39 1.23 3.76 5.53 32 3.21

3.27 3.49 4.32 2.49 2.96 2.17 1.39 2.6 3.36

Ω2 4Ω6 4Ω4 Ω2 4Ω6 4Ω4 Ω2 4Ω6 4Ω4 Ω2 4Ω6 4Ω4 Ω2 4Ω6 4Ω4 Ω4 4Ω6 4Ω2 Ω4 4Ω6 4Ω2 Ω4 4Ω2 4Ω6 Ω6 4Ω4 4Ω2

Fig. 4 shows the normalized emission spectra with respect to the peak intensity of the 5D4-7F5 (540 nm) emission transition under excitation at 376 nm for different Tb3 þ -doped PZABP glasses. The luminescence spectra demonstrated the emission transitions arising from both 5D3 and 5D4 energy levels to 7FJ ground state multiplets. The emission peaks at 412, 434, 455 and 470 nm are attributed to 5D3-7F5,4,3,2 transitions and those located at 486, 540, 582 and 618 nm are ascribed to 5D4-7F6,5,4,3 transitions. It is observed that the emission intensities of 5D3 state decreases with increasing Tb3 þ ion concentration for all the glasses (Fig. 4). The intensity variation of blue (5D3-7F4 at 434 nm) and green (5D4-7F5 at 540 nm) emissions, as a function of Tb3 þ ion concentration has been graphically shown in Fig. 5. The relative intensity of blue emission IB (5D3-7F4) with respect to green emission IG (5D4-7F5) is shown in Fig. 6 as a function of Tb3 þ ions concentration. It is noticed that blue to green relative emission intensity ratio (IB/IG) decreases with increasing Tb3 þ ion concentration and this parameter indicates the degree of quenching. This trend implies that the intensity of emission transitions from 5 D3 is comparable to those from 5D4 level at lower Tb3 þ concentration that is accompanied by a decrease in 5D3 emission intensity with an increase of Tb3 þ ion concentration. This should be due to the concentration quenching of the 5D3 emission, which is more prominent for higher Tb3 þ ion concentration because of the reduction in the inter-ionic distances. The latter mechanism has been clearly described in the partial energy level diagram with cross-relaxation process (Fig. 7). By applying the J–O parameters, the radiative properties such as radiative rates AR, experimental and calculated branching ratios βR and stimulated emission cross-sections sem for the 5D4-7F3, 5 D4-7F4, 5D4-7F5, and 5D4-7F6, transitions of Tb3 þ ions in the present title glass were determined and are listed in Table 3. The details of calculations are given in earlier reported works [3,18]. Moreover, the effective bandwidth Δλeff, emission cross-section sem and their products (gain bandwidth parameter, sem Δλeff) are also presented. The branching ratio is a critical parameter, which determines the lasing capability of any specific transition by characterizing the possibility of attaining stimulated emission. The

80


Fig. 2. (a) hν versus (αhν)1/2 plot for un-doped and 5 wt% Tb3 þ ions in PZABP glasses (b) Variation of optical band gap energy values (Eopt) with Tb3 þ ion concentrations for PZABP glasses.

Fig. 3. Excitation spectrum for PZABP:5Tb glass monitored for the emission at 540 nm.

Fig. 4. The normalized emission spectra with respect to the peak intensity of the 5 D4-7F5 emission transition under excitation at 376 nm.

Fig. 5. Variation of green (5D4-7F5) and blue (5D3-7F4) emission intensity as a function of Tb3 þ ion concentration in PZABP glasses.

Fig. 6. Variation of fluorescence intensity ratio as a function of Tb3 þ concentration in PZABP glasses.


81

Fig. 7. Partial energy level diagram showing the luminescence channels and resonant energy transfer (RET) in Tb3 þ -doped PZABP glasses.

Table 3 Peak wavelengths (λp, nm), radiative transition probabilities (AR, s 1), branching ratios (βR), effective bandwidths (Δλeff, nm), stimulated emission cross-sections (sem 10 22 cm2) and gain band width parameter ((sem Δλeff) 10 28 cm3) for 5 D4-7F3–6 transitions of 5.0 wt% of Tb3 þ doped PZABP glass. 5

7

D4-

F3 F4 F5 7 F6 7 7

λp

619 582 540 486

AR

23.74 12.54 207.61 31.04

βR Exp.

Cal.

0.08 0.11 0.65 0.17

0.09 0.05 0.76 0.11

Δλeff

rem

(rem Δλeff)

9.81 11.39 8.62 10.76

1.97 0.70 11.39 0.89

1.94 0.80 9.81 0.96

ΣAR ¼275 s 1; τrad ¼ 3.6 ms

branching ratios are also used to predict the relative intensities of all emission transitions originating from a given excited state. It is known that a transition having branching ratio of Z0.50 can emit laser radiation more efficiently [27]. The calculated branching ratio of 0.76 for the 5D4-7F5 transition of 5.0 wt% Tb3 þ -doped PZABP glass suggests its suitability for green color display devices. The higher value of sem is also favorable for low threshold and high gain to obtain continuous wave laser action. The τrad for the 5D4 level of Tb3 þ ions in PZABP:5Tb glass is found to be 3.64 ms, which is comparable to that of the Tb3 þ doped in AlO3 crystal [32]. Recently, Silversmith et al. [33] have suggested that Tb3 þ has proven to be a useful probe of fluorescence quenching mechanisms. Due to the large energy (1500 cm 1) gap below the emitting level, 5D4-7FJ fluorescence is easily observed in all materials and is not affected by the fluorescence quenching mechanisms in the RE3 þ -doped glasses. Emission from 5D3, 5700 cm 1 above 5D4 is significantly quenched by nearby trapped hydroxyl ions or water molecules, by cross relaxation with other Tb3 þ ions, and by multiphonon relaxation. Thus, the 5D4 emission intensity serves as an internal calibration and the relative intensity of IB/IG (5D3/5D4) can be used as a measure of the degree of fluorescence quenching. The 5 D3-7FJ emission should be considered as representative of many potentially emitting transitions in RE systems that are susceptible to quenching in glasses. In the present zinc phosphate glass host, the highest lattice phonon energy is found to be around 1100 cm 1, which implies that at least five phonons are required to bridge the energy gap (5794 cm 1) of 5D3-5D4 levels and more than 15 phonons to depopulate 5D4 level to its lower ground state

energy levels (energy gap of 5D4-7FJ 415,000 cm 1). Hence, the multi-phonon relaxation can be neglected in the latter case. The hydroxyl ion (OH ) impurity in glasses acts as high-energy phonons (∼3200 cm 1), which may cause rapid quenching of 5D3 energy levels of the Tb3 þ ions. Besides the possibility of multiphonon assisted non-radiative relaxation of 5D3 to 5D4 level due to its low energy gap at around 5794 cm 1, a possibly dominating faster cross-relaxation (CR) phenomenon could be involved (see Fig. 7). Such CRs result from a resonant energy transfer (RET) through 5D3-5D4⇒7F6-7F0 process as the energy difference between 5D3 and 5D4 matches well with the energy gap between 7 F6 and 7F0 energy levels [34]. This process becomes more prominent at higher Tb3 þ ion concentrations. Choi et al. [34] and Sohn et al. [35] have suggested a second type of energy transfer quenching mechanism for the 5D3 and 5D4 excited levels based on the excited state absorption by a cooperative energy transfer from the latter to the upper laying levels (UL). According to them, this process has been proposed as in three possible ways: (i) 5D4-7F6⇒5D4-UL; (ii) 5D4-7F6⇒5D3-UL and (iii) 5D3-7F6⇒5D3-UL. In order to induce the latter mechanism, at least two excited active ions in close vicinity of each other are needed, so that by de-exciting one, other can be promoted to the upper levels such as 7D, 9D of 4f75d1 configuration, or to a charge transfer band (CTB) and sometimes even the host absorption band. The cross-relaxation by RET occurs among the excited active ions and its nearest neighboring unexcited active ions. Since it is more probable for active ions to be surrounded with unexcited ions rather than the excited ones, the cross-relaxation mechanism could contribute more to 5D3 quenching than the cooperative energy transfer to upper levels. In the case of 5D4 excited level, nonradiative relaxation is negligible and no quenching is observed for 5 D4 level with increasing Tb3 þ ion concentration. In general, the emission color of two glasses is compared by means of color coordinates and is good certification for photoluminescence applications. In 1931, the Commission International de I’Eclairage (CIE) established a universal quantitative model of color spaces. The chromaticity coordinates of Tb3 þ doped PZABP glasses are calculated and listed in Table 4 (using CIE calculated software) from their corresponding emission spectra excited by 376 nm as shown in Fig. 8, the obtained CIE color coordinates of all the concentration of Tb3 þ doped glasses lie in the green region. The changes in the color coordinates may be due to the variation of the asymmetric ratios of various Tb3 þ ion concentrations of

82


Table 4 The variation of 5D3-7F4 experimental lifetime (τexp, ms), concentration (N0 1020 ions/cm3), thickness (l, cm), optical band gap energy (Eopt, eV), energy transfer parameter (Q), critical transfer distance (R0, Å), dipole–dipole interaction parameter (CDA 10 39 cm6/s) and CIE coordinates (x, y) with respect to concentration (wt%) of Tb3 þ ions in PZABP glasses. System

Cons. l

Eopt

τexp

N0

Q

R0

CDA (x, y)

PZABP:0Tb PZABP:1Tb PZABP:2Tb PZABP:3Tb PZABP:4Tb PZABP:5Tb

0 1.0 2.0 3.0 4.0 5.0

4.33 4.43 4.44 4.44 4.46 4.49

– 0.93 0.77 0.70 0.50 0.42

– 0.557 1.114 1.659 2.136 2.773

– 0.44 0.95 1.50 1.59 2.64

– 10.2 10.5 10.7 10.0 10.8

– 1.2 1.4 1.5 1.0 1.6

0.224 0.220 0.228 0.225 0.228 0.227

– (0.284, 0.483) (0.298, 0.540) (0.307, 0.571) (0.310, 0.583) (0.313, 0.592)

Fig. 9. Decay curves for the 5D4 level of different Tb3 þ ion concentration in PZABP glasses.

Fig. 8. CIE chromaticity diagram for Tb3 þ doped PZABP glasses under UV (376 nm) excitation.

PZABP glasses. The present results suggest that the Tb3 þ -doped PZABP glasses could be a potential candidate as green color emitting display/lamp applications. 3.5. Decay curve analysis Luminescence decay analysis is very useful for understanding the energy transfer mechanism and luminescence quenching behavior of Tb3 þ ions. Fig. 9 shows the decay curves for the 5D4 level for different Tb3 þ ion concentrations obtained by monitoring the green emission around 540 nm attributed to 5D4-7F5 transition. It can be seen that decay curves exhibit single exponential nature for all the concentrations in the studied glasses. The experimental lifetimes are determined to be 2.94, 2.88, 2.66, 2.69 and 2.62 ms for 1.0, 2.0, 3.0, 4.0 and 5.0 wt% of Tb3 þ ions, respectively. It is noticed that the lifetime of the 5D4 level decreases very slightly as the Tb3 þ concentration increases. This indicates that energy migration does not affect the 5D4 lifetime as observed in most Tb3 þ -doped glasses [22,23,29]. The multiphonon emission rate is also expected to be negligible due to the large energy gap ( 15,000 cm 1) below the 5D4 level. The long fluorescence lifetime of Tb3 þ ions in these glasses can reduce the pump threshold to get the visible laser output. The luminescence quantum efficiency (η) is defined as the ratio of the number of photons emitted to the number of photons absorbed and is given by η ¼ τexp/τrad. The η of the 5D4 fluorescence level is found to be 81% for PZABP:1Tb glass, which is comparable to that of the Tb3 þ doped phosphate glasses [22]. Furthermore, the laser threshold or optical gain, i.e., the product of sem τrad, is found to

Fig. 10. Decay curves for the 5D3 level of different Tb3 þ ion concentration in PZABP glasses. The IH fit (S¼ 6) using in Eq. (2). The inset shows the variation of energy transfer parameter (Q) with Tb3 þ ion concentrations (ions/cm3).

be 41.46 10 25 cm2 s for the 5D4-7F5 transition. Such relatively high threshold nominates the PZABP:5Tb glass as a promising material to attain intense green laser emission. Fig. 10 shows the decay curves for the 5D3 level for different concentrations of Tb3 þ ions obtained by monitoring the blue emission around 434 nm attributed to 5D3-7F4 transition. As evidenced in Fig. 10, the decay curves exhibit non-exponential nature for all the terbium concentrations. Similar behavior has been reported for other Tb3 þ -doped glasses such as in Tb3 þ -doped sodium germinate [36] and Tb3 þ -doped calcium aluminosilicate glasses [29]. The experimental lifetimes (τexp) of the 5D3 level have been determined using the non-exponential decay curves by finding the average lifetime using the formula [37].

τexp =

∫ tI (t ) dt ∫ I (t ) dt

(1) 5

3þ

The τexp values for the D3 level of Tb ions are 0.93, 0.77, 0.70, 0.50 and 0.42 ms for 1.0, 2.0, 3.0, 4.0, and 5.0 wt% of Tb3 þ , respectively, which are lower than the lifetime for the 5D3 level in


calcium aluminosilicate [20] and low silica calcium aluminosilicate [31] glasses. As we discussed in earlier section, the contribution of multiphonon decay should be negligible for 5D3 state in Tb3 þ doped glasses. Therefore, the CR effect should be negligible in the low concentration limit (C-0). Extrapolating 5D3 level decay lifetime data for C-0, we obtained τ0∼0.97 ms from a plot of τexp 1 versus C. It is important to remark that we did not obtain a reliable radiative lifetime value for 5D3 state from the Judd–Ofelt method, probably due to the strong overlapping of the UV absorption lines. This is common problem in Tb3 þ glasses, where very few reliable Judd–Ofelt calculations are found in the literature [21]. Since phonon energy is nearly equal to zero, we assumed τrad∼τ0 ¼0.97 ms for 5D3 level in Tb3 þ doped PZABP glasses. The decay curves for 5D3-7F4 transition of Tb3 þ -doped PZABP glasses exhibit non-exponential nature for all the concentrations and obtained experimental lifetimes for 5D3-7F4 transition, are given in Table 4. The lifetime for the 5D3 level is found to decrease with increasing Tb3 þ concentration. The quenching of lifetime and its non-exponential nature could be due to the increased energy transfer processes through cross-relaxations between the Tb3 þ ions. In order to find out the nature of the energy transfer processes, the decay curves are fitted to the Inokuti–Hirayama (IH) model [38]. According to the IH-model, the fluorescence decay intensity, I, is given by

⎧ ⎛ t ⎞3/ S ⎫ ⎪ ⎪ t I (t ) = I0 exp ⎨ − −Q ⎜ ⎟ ⎬ ⎪ ⎪ τ τ ⎝ ⎠ 0 0 ⎩ ⎭

(2)

where t is the time after excitation, τ0 is the intrinsic decay time of the donors in the absence of acceptors (τ0 ¼ τrad in the respective case). Q is the energy transfer parameter defined as

Q=

4π ⎛ 3⎞ Γ ⎜1 − ⎟ N0 R 03 ⎝ 3 S⎠

(3)

Q depends on a number S and the gamma function Γ(x), which is equal to 1.77 for dipole–dipole (S¼ 6), 1.43 for dipole–quadrupole (S¼ 8), and 1.30 for quadrupole–quadrupole (S ¼10) interactions, respectively. N0 is the acceptors concentration, which is almost equal to total concentration of RE3 þ ions, and R0 is the critical transfer distance defined as the donor–acceptor separation for which the rate of energy transfer between a donor and acceptor is equal to the rate of intrinsic decay, τ0−1. The 5D3 luminescence decays in this paper are well described by Eq. (2). The donor–acceptor interaction parameter CDA is related to R0 as

CDA =

R 0(S ) τ0

(4)

The decay curves of PZABP glasses are well-fitted to the IH model for S¼ 6, indicative of dipole–dipole interaction between Tb3 þ ions. The analysis of non-exponential decay curves yields the values of Q that increases approximately linearly with Tb3 þ ion concentrations (shown in the inset of Fig. 10). Consequently, using Eqs. (3) and (4) nearly constant values of R0 and CDA were obtained, as shown in Table 4. As expected, the R0 and CDA values are found to be nearly constant for the five samples with a mean value of 10.770.03 Å and 1.5470.3 10 39 cm6/s, respectively. These results are comparable to that in Tb3 þ -doped fluorophosphate glasses [25] and other rare earth doped [39-41] and commercial [42] glasses. The decrease in values of τexp and increase in Q parameters by increasing the Tb2O3 content clearly indicate the boosted ET process between Tb3 þ ions through CRs. The increase in the Q value with increasing concentration is due to the increase of effective density of donors and acceptors. Therefore, the crossrelaxation process responsible for resonant energy transfer between Tb3 þ ions in PZABP glasses is mainly governed by dipole–dipole

83

interactions, which agrees well with the previous works of Duhamel-Henry et al. [25], Sontakke et al. [29] and Pisarska et al. [43].

4. Conclusions In summary, the concentration dependent luminescence of Tb3 þ -doped zinc phosphate glasses, prepared by the meltquenching technique was studied through absorption, excitation, luminescence and decay rate measurements. Two prominently green (5D4-7F5 at 540 nm) and blue (5D3-7F4 at 434 nm) emission bands are observed upon excitation at 376 nm. It is observed that the 5D3 emission intensity of Tb3 þ ions decreases with increase in Tb3 þ ion concentration, while the intensity of green emission from 5 D4 level has shown a linear increase. The decay curves of 5D4 level of Tb3 þ ion exhibit single exponential nature for all the concentrations and the lifetimes vary slightly from 2.94 to 2.62 ms. The 5 D3 fluorescence decay curves have shown non-exponential nature for all the concentrations and the lifetimes vary from 0.93 to 0.42 ms. The non-exponential decay rates are well-fitted to the Inokuti–Hirayama model for S¼ 6, indicating that the nature of the cross-relaxation energy transfer between Tb3 þ ions is of dipole– dipole type. The energy transfer parameter value increases with increasing Tb3 þ ion concentration due to the enhancement of energy transfer process between Tb3 þ ions through cross-relaxation. The long fluorescence lifetime of Tb3 þ ions in these glasses can reduce the pump threshold to get the visible laser output. The radiative properties of 5.0 wt% of Tb3 þ doped zinc phosphate glasses strongly suggest their utility to development of green color display devices and solid state visible lasers.

Acknowledgments This research was supported by the Fundação de Amparo a Pesquisa do Estado de São Paulo (FAPESP), Brazil Grant no. 2012/ 19338-9 and also the authors would like to thank the CEPID Project no. 2013/07793-6 (CERTEV-Center for Research, Technology and Education on Vitreous Materials).

References [1] B.C. Wybourne, Spectroscopic Properties of Rare Earths, Interscience, New York, 1965. [2] G.H. Dieke, Spectra and Energy Levels of Rare Earth Ions in Crystals, Interscience, New York, 1968. [3] C.R. Kesavulu, C.K. Jayasankar, Mater. Chem. Phys. 130 (2011) 1078. [4] C.R. Kesavulu, C.K. Jayasankar, J. Lumin. 132 (2012) 2802. [5] A. Taguchi, M. Taniguchi, K. Takahei, Appl. Phys. Lett. 60 (1992) 965. [6] K. Hirao, S. Todoroki, D.H. Cho, N. Soga, Opt. Lett. 18 (1993) 1586. [7] N. Nogami, Y. Abe, Appl. Phys. Lett. 65 (1994) 1227. [8] E. Nakazawa, S. Shionoya, J. Chem. Phys. 47 (1967) 3211. [9] J.A. Wilder, J. Non-Cryst. Solids 38–39 (1980) 879. [10] N.H. Ray, C.J. Lewis, J.N.C. Laycock, W.D. Robinson, Glass Technol. 14 (2) (1973) 50. [11] B. Tischendorf, J.U. Otaigbe, J.W. Wiench, M. Pruski, B.C. Sales, J. Non-Cryst. Solids 282 (2–3) (2001) 147. [12] B. Zhang, Q. Chen, L. Song, H.P. Li, F.Z. Hou, J.C. Zhang, J. Non-Cryst. Solids 354 (18) (2008) 1948. [13] N.O. Dantas, E.S.F. Neto, R.S. Silva, D.R. Jesus, F. Pelegrini, Appl. Phys. Lett. 93 (2008) 1 193115. [14] M.C. Neto, G.H. Silva, A.P. Carmo, A.S. Pinheiro, N.O. Dantas, M.J.V. Bell, V. Anjos, Chem. Phys. Lett. 588 (2013) 188. [15] B.R. Judd, Phys. Rev. 127 (1962) 750. [16] G.S. Ofelt, J. Chem. Phys. 37 (1962) 511. [17] S. Tanabe, T. Ohyagi, N. Soga, T. Hanada, Phys. Rev. B 46 (1992) 3305. [18] B.C. Jamalaiah, J. Suresh Kumar, A. Mohan Babu, T. Sasikala, L. Rama Moorthy, Physica B 404 (2009) 2020. [19] L. Huang, G. Qin, Y. Arai., R. Jose, T. Suzuki, Y. Ohishi, T. Yamashita, Y. Akimoto, J. Appl. Phys. 102 (2007) 1 093506. [20] G. Amaranath, S. Buddhudu, F.J. Bryant, J. Non-Cryst. Solids 122 (1990) 66. [21] T. Yamashita, Y. Ohishi, J. Appl. Phys. 102 (2007) 1 123107.

84


[22] L. Zhang, M. Peng, G. Dong, J. Qiu, Opt. Mater. 34 (2012) 1202. [23] D. Umamaheswari, B.C. Jamalaiah, T. Sasikala, T. Chengaiah, I.I.-Gon Kim, L. Rama Moorthy, J. Lumin. 132 (2012) 1166. [24] S.V.G.V.A. Prasad, M. Srinivasa Reddy, V. Ravi Kumar, N. Veeraiah, J. Lumin. 127 (2007) 637. [25] N. Duhamel-Henry, J.L. Adam, B. Jacquier, C. Linares, Opt. Mater. 5 (1996) 197. [26] C.K. Jorgensen, R. Reisfeld, J. Less-Common Met. 93 (1983) 107. [27] H. Lin, D. Yang, G. Liu, T. Ma, B. Zhai, Q. An, J. Yu, X. Wang, X. Liu, E.Y.B. Pun, J. Lumin. 113 (2005) 121. [28] E.A. Davis, N.F. Mott, Philos. Mag. 22 (1970) 903. [29] A.D. Sontakke, K. Biswas, K. Annapurna, J. Lumin. 129 (2009) 1347. [30] K. Maheshvaran, K. Marimuthu, J. Lumin. 132 (2012) 2259. [31] J.F.M. dos Santos, I.A.A. Terra, N.G.C. Astrath, F.B. Guimaraes, M.L. Baesso, L.O.A. Nunes, T. Catunda, J. Appl. Phys. 117 (2015) 053102. [32] D.K. Sardar, K.L. Nash, R.M. Yow, J.B. Gruber, U.V. Valiev, E.P. Kokanyan, J. Appl. Phys. 100 (2006) 1 083108. [33] A.J. Silversmith, N.T.T. Nguyen, D.L. Campbell, D.M. Boye, C.P. Ortriz, K.R. Hoffman, J. Lumin. 129 (2009) 1501.

[34] Yoon Yang Choi, Kee-Sun Sohn, Hee Dong Park, Se. Yong Choi, J. Mater. Res. 16 (2001) 881. [35] Kee-Sun Sohn, Namsoo Shin, Electrochem. Solid State Lett. 5 (2002) H21. [36] E. Alvarez, Ma.E. Zayas, D. Rodriguez-Carvajal, F. Felix-Dominguez, R.P. DuarteZamorano, C. Falcony, U. Caldino, Opt. Mater. 37 (2014) 451. [37] F. Lahoz, I.R. Martin, J. Mendez-Ramos, P. Nunez, J. Chem. Phys. 120 (2004) 6180. [38] M. Inokuti, F. Hirayama, J. Chem. Phys. 43 (1965) 1978. [39] C. Jacinto, S.L. Oliveira, L.A.O. Nunes, J.D. Myers, M.J. Myers, T. Catunda, Phys. Rev. B 73 (2006) 125107. [40] S.M. Lima, J.A. Sampaio, T. Catunda, A.S.S. de Camargo, L.A.O. Nunes, M.L. Baesso, D.W. Hewak, J. Non-Cryst. Solids 284 (2001) 274. [41] J.A. Caird, A.J. Ramponi, P.R. Staver, J. Opt. Soc. Am. B 8 (1991) 1391. [42] P.R. Ehrmann, J.H. Campbell, J. Am. Ceram. Soc. 85 (2002) 1061. [43] J. Pisarska, M. Soltys, L. Zur, W.A. Pisarski, C.K. Jayasankar, Appl. Phys. B 116 (2014) 837.