Investigation of concentration quenching and 1.3 μm emission in Nd3+-doped bismuth glasses

Available online at www.sciencedirect.com

Spectrochimica Acta Part A 70 (2008) 537–541

Investigation of concentration quenching and 1.3 ␮m emission in Nd3+-doped bismuth glasses Qiuhua Nie a , Xujie Li a,∗ , Shixun Dai a , Tiefeng Xu a , Zhenjuan Jin a , Xianghua Zhang a,b a

Faculty of Information Science and Engineering, The State Key Laboratory Base of Novel Functional Materials and Preparation Science, Ningbo University, Zhejiang 315211, PR China b Universit´ e de Rennes I, Rennes 35042, France Received 16 April 2007; received in revised form 13 July 2007; accepted 27 July 2007

Abstract Nd2 O3 -doped 70Bi2 O3 –20B2 O3 –10SiO2 –xNd2 O3 (x = 0.1, 0.3, 0.5, 0.7, 1.0, 1.5 mol%) bismuth glasses were prepared by the conventional melt-quenching method, and the Nd3+ :4 F3/2 → 4 I13/2 fluorescence properties had been studied for different Nd3+ concentrations. The Judd–Ofelt analysis for Nd3+ ions in bismuth boron silicate glasses was also performed on the base of absorption spectrum. The transition probabilities, excited state lifetimes, the fluorescence branching ratios, quantum efficiency and the stimulated emission cross-sections of 4 F3/2 → 4 I13/2 transition were calculated and discussed. Based on the electric dipole–dipole interaction theory, the interaction parameters: CDD , for the energy migration rate 4 F3/2 , 4 I9/2 → 4 F3/2 , 4 I9/2 and CDA , for cross-relaxation rate 4 F3/2 , 4 I9/2 → 4 I15/2 , 4 I15/2 , and/or 4 F3/2 , 4 I9/2 → 4 I13/2 , 4 I15/2 in bismuth boron silicate glasses were about 18.4 × 10−40 cm6 /s and 3.4 × 10−40 cm6 /s, respectively. © 2007 Elsevier B.V. All rights reserved. Keywords: Bismuth glasses; Fluorescence properties; Quantum efficiency; Concentration quenching; Energy migration

1. Introduction The need for an optical amplifier in the second telecommunication window at 1.3 ␮m has stimulated a great deal of research activity in both rare earth-doped devices and in alternative approaches such as the semiconductor amplifier and Raman fiber amplifier [1–3]. Nd3+ ions present radiative emission at 1350 nm originating from the electronic transitions between 4F 4 3+ are 3/2 → I13/2 . And the possible host glasses for Nd interesting. HEAVY-METAOLX IDE (HMO) glasses might be arbitrarily defined as those glasses containing over 50 cation percent (cat %) of bismuth and/or lead which participate in the glass structure as network formers. Among the HMO glasses, bismuth boron silicate glasses are of growing interest. They are characterized by high density, high refractive index, high thermal expansion, low transformation temperature and excellent infrared transmission [4]. So HMO glasses-doped with trivalent rare earth ions are considered like promising hosts for optical amplifiers [5]. Tanabe et al. [6] reported that the Bi2 O3 -based

∗

Corresponding author. Tel.: +86 574 8760 0358; fax: +86 574 8760 0946. E-mail address: [email protected] (X. Li).

1386-1425/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2007.07.046

borosilicate glasses show broad emission spectra, and the glasses with Bi2 O3 show properties superior to the tellurite. The 1.3 ␮m amplifier based on bismuth glasses always needs a higher Nd3+ concentration to obtain sufficient gains. At the high Nd3+ concentration, however, the luminescence will be quenched by energy transfer processes due to interactions between Nd3+ ions. Until now much attention has been given to the research on concentration quenching of Er3+ [7,8], but little is known about the concentration quenching in Nd3+ ions-doped bismuth glasses. The present work is devoted to the investigation property of 1.3 ␮m emission and the analysis of the concentration quenching in bismuth glasses using a model based on energy transfer. 2. Experiment Glasses with the composition of 70Bi2 O3 –20B2 O3 – 10SiO2 –xNd2 O3 (x = 0.1, 0.3, 0.5, 0.7, 1.0, 1.5 mol%) are prepared. Table 1 shows the glass compositions of the samples and corresponding sample no. The starting materials are reagent grade Bi2 O3 , B2 O3 and SiO2 , Nd2 O3 with more than 99.99% purity. About 25 g batches of starting materials are fully mixed and then melted in the Pt crucibles at 1100–1200 ◦ C in an elec-

538

Q. Nie et al. / Spectrochimica Acta Part A 70 (2008) 537–541

Table 1 Glass sample no., glass composition, density, refractive index, Nd3+ concentration Sample no.

BBS1 BBS2 BBS3 BBS4 BBS5 BBS6

Composition (mol%) Bi2 O3

B2 O3

SiO2

Nd2 O3

70 70 70 70 70 70

25 25 25 25 25 25

10 10 10 10 10 10

0.1 0.3 0.5 0.7 1.0 1.5

tronic furnace. After completely melting, the glass liquids are poured into a stainless mold and then annealed to room temperature. The obtained glasses are cut and polished carefully to 10 mm × 10 mm × 1.2 mm in order to meet the requirements for optical measurements. The densities are measured according to the Archimedes principle. Refractive indices are measured at 632.8 nm on SAIRON-SPA4000 prism coupler. The absorption spectra are recorded between 500 and 2000 nm with a Perkin-Elmer Lambda 950 UV–vis–NIR spectrophotometer. The fluorescence spectra are obtained with a TRIAX550 spectrofluorimeter upon excitation of 800 nm LD with a maximum power of 2 W. The fluorescence lifetime of Nd3+ :4 F3/2 level is measured with light pulses of 800 nm LD. The decay traces are recorded on a digital oscilloscope and fitted by single exponential functions to obtain the decay rates. All the measurements are performed at room temperature.

Density (g/cm3 )

Refractive index

Nd3+ concentration (1019 cm−3 )

7.428 7.392 7.387 7.342 7.271 7.335

2.1972 2.1980 2.1978 2.1942 2.1797 2.1937

2.6 7.8 12.9 18.0 25.4 38.5

Fig. 2. 1.3 ␮m emission spectra of Nd3+ for different Nd3+ concentrations. 4S

3. Results and discussion 3.1. Absorption spectra and Judd–Ofelt analysis The absorption spectrum of Nd3+ singly-doped bismuth glass (BBS5) is shown in Fig. 1, For the Nd3+ singly-doped sample, the absorption spectrum consists of eight absorption bands located at 1650, 881, 807, 750, 685, 629, 587 and 528 nm, corresponding to the ground state 4 I9/2 to the excited states 4 I15/2 , 4 F3/2 , 4 F5/2 ,

4F

4F

4 13/2 + G7/2 , respec3+ tively. Fig. 2 shows 1.3 ␮m emission spectra of Nd for different 3/2 +

7/2 ,

9/2 ,

2H

11/2 ,

4G

5/2 +

2G

7/2 ,

2K

Nd3+ concentrations. The intensity drastically increases and has a maximum around 1 mol% Nd2 O3 (BBS5) and decreases with an increase of the Nd2 O3 content. It is important to mention at this point that the trend of Nd3+ :4 F3/2 → 4 I15/2 and 4F 4 3+ 3/2 → I13/2 emission for different Nd concentrations are the same as 1.3 ␮m emission spectra. For the Nd3+ content more than 1 mol%, the luminescence will be quenched by energy transfer processes due to interactions between Nd3+ ions, in other words concentration quenching occurs in Nd3+ ions-doped bismuth boron silicate glasses. According to the Judd–Ofelt theory [9,10], Judd–Ofelt parameters of different Nd3+ concentrations in bismuth glasses and predicted spontaneous-radiative rates electric-dipole line strengths and fluorescence branching ratio of Nd3+ ions are determined. The corresponding values are listed in Tables 2 and 3. From Table 2, it can be seen that Ω2 > Ω4 > Ω6 . Table 2 Judd–Ofelt parameters of different Nd3+ concentrations in bismuth glasses

Fig. 1. Absorption spectra of Nd3+ singly-doped bismuth glass.

Sample no.

Ω2 × 10−20 (cm2 )

Ω4 × 10−20 (cm2 )

Ω6 × 10−20 (cm2 )

BBS1 BBS2 BBS3 BBS4 BBS5 BBS6

4.50 4.25 4.20 4.45 4.42 3.84

2.54 3.05 2.96 2.91 3.06 3.08

4.03 3.53 3.60 3.28 3.26 3.24

Q. Nie et al. / Spectrochimica Acta Part A 70 (2008) 537–541

539

Table 3 Predicted spontaneous-radiative rates electric-dipole line strengths and fluorescence branching ratio of Nd3+ ions Sample no.

Initial manifold

Final manifold

4F 3/2

4I 15/2 1858a

4I 13/2 1346a

4I 11/2 1063a

4I 9/2 881a

BBS1

Aed (s−1 ) Sed (10−20 cm2 ) β

14.3892 0.1128 0.0055

226.6655 0.8544 0.1089

1362.8407 2.0009 0.5175

969.48 0.8099 0.3682

BBS2

Aed (s−1 ) Sed (10−20 cm2 ) β

11.9916 0.0988 0.0046

238.9150 0.7484 0.0912

1343.6230 2.0733 0.5131

1024.14 0.8992 0.3911

BBS3

Aed (s−1 ) Sed (10−20 cm2 ) β

12.8522 0.1008 0.0049

255.9439 0.7632 0.0981

1283.6260 1.8855 0.4922

1055.76 0.8824 0.4048

BBS4

Aed (s−1 ) Sed (10−20 cm2 ) β

11.6294 0.0918 0.0048

231.7068 0.6954 0.0950

1182.4992 1.7482 0.4847

1014.02 0.8530 0.4156

BBS5

Aed (s−1 ) Sed (10−20 cm2 ) β

11.2915 0.0913 0.0047

224.8075 0.6911 0.0926

1163.0779 1.7613 0.4790

1028.71 0.8864 0.4237

BBS6

Aed (s−1 ) Sed (10−20 cm2 ) β

12.5529 0.0907 0.0046

250.0455 0.6869 0.0921

1297.6443 1.7560 0.4778

1155.61 0.8898 0.4255

a

Wavelength (nm).

The Ω2 parameters is affected by the covalency, the Ω6 parameter is related to the rigidity of the glass hosts, and Ω4 parameters is determined by Ω2 parameter and Ω6 parameter [11,12]. The stimulated emission cross-section at the peak wavelength is estimated with the formula [13] σp =

λ4p 8πcn2 λeff

A[(4 F3/2 ; 4 I13/2 )]

(1)

where λp is the peak wavelength of the emission band and λeff is the effective bandwidth determined by the relation [13]  1 I(λ) dλ (2) λeff = Ip where Ip is the peak intensity of the emission band. According Eq. (1), the stimulated emission cross-section at the peak wavelength is proportion to the electric-dipole spontaneous emission probabilities A[(4 F3/2 ;4 I13/2 )] which is listed in Table 3. Obviously, the larger of A[(4 F3/2 ;4 I13/2 )], the larger of the stimulated emission cross-section. From Table 3, it can be seen that the values of A[(4 F3/2 ;4 I13/2 )] are suffi-

cient large in bismuth boron silicate glasses. Table 4 shows that an important reduction of the 4 F3/2 level lifetime is clearly observed when increasing Nd3+ concentration, which indicates the increases of the energy transfer rate involving both Nd3+ ions. The stimulated emission cross-section of 4 F3/2 → 4 I13/2 transition, calculated using Eq. (1) is found to reach the maximum 4.1406 × 10−21 cm2 in BBS3 sample. As is well known, a good candidate of laser material should have large effective bandwidth, long fluorescence lifetime, stimulated emission cross-section and high branching ratios. As a comparison, some important spectroscopic parameters of Nd3+ in glasses are listed in Table 5. It is evident that, for BBS3 sample, the effective bandwidth, fluorescence lifetime and stimulated emission crosssection are close to that of glasses which have been reported as good candidate of laser material. 3.2. Energy transfer among Nd3+ ions Fig. 3 shows simplified energy level diagram of Nd3+ ions. The main energy transfer processes between Nd3+ ions, named,

Table 4 level lifetime, Nd3+ :4 F3/2 → 4 I13/2 effective bandwidth and the stimulated emission cross-section at the peak wavelength

4F 3/2

Sample no.

λeff (nm) FWHM (nm) τ rad (ms) τ f (ms) η (%) σ (10−21 cm2 )

BBS1

BBS2

BBS3

BBS4

BBS5

BBS6

55.73 46 0.380 0.354 93.16 3.6562

55.89 51 0.382 0.336 87.96 3.7950

55.05 50 0.383 0.325 84.86 4.1406

55.02 51 0.410 0.333 81.22 3.7404

55.76 51 0.412 0.317 76.94 3.6287

55.47 52 0.368 0.303 82.34 3.9936

540

Q. Nie et al. / Spectrochimica Acta Part A 70 (2008) 537–541

Table 5 Comparison of J–O parameters Ωt (10−20 cm2 ), branching ratios (β), radiative lifetime (ms), and emission parameters effective bandwidth (nm), stimulated emission cross-section (10−21 cm2 ) in different glass Glass composition

Ω2

Ω4

Ω6

β

λeff

σ

τ rad

Reference

Borophosphate glass Tellurite glass Oxyfluoride glass Borate lead glass BBS3

1.22 3.80 5.28 4.72 4.20

4.25 4.94 3.86 2.12 2.96

1.28 4.54 1.37 3.93 3.60

0.060 0.109 0.090 0.123 0.0981

– 45.9 62.0 57.0 55.05

2.71 – 5.70 7.00 4.14

– 0.153 0.417 0.426 0.325

[14] [15] [16] [17] This work

Fig. 3. Simplified energy level diagram of Nd3+ ions presenting the crossrelaxation and excitation migration.

Fig. 4. The absorption cross-section and emission cross-section of Nd3+ :4 F3/2 → 4 I9/2 transition in BBS5 glasses.

the energy migration 4 F3/2 , 4 I9/2 → 4 F3/2 , 4 I9/2 (CDD ), and the cross-relaxation 4 F3/2 , 4 I9/2 → 4 I15/2 , 4 I15/2 , and/or 4 F3/2 , 4I 4 4 9/2 → I13/2 , I15/2 (CDA ). Normalized emission spectra are obtained from the relation:

the relation can be written as [16]    em FD (E)σXabs (E) 3 hc 4 6 RDX = dE 4π n E4

I(λ) g(λ) =  ∞ 0 I(λ) dλ

where X = D or A for donor–donor or donor–acceptor transfer, respectively, σXabs (E) is the absorption cross-section of species X at photon energy E, Eq. (5) conversion to an integral over wavelength yield  3c 6 RDX = σDem (λ)σXabs (λ) dλ (6) 8π4 n2 Ar

(3)

where I(λ) is the emission spectrum. We obtained the crosssection for stimulated emission σ em , from the normalized emission spectrum and the spontaneous emission rate, using the relations [18,19] σ em (λ) =

λ2 A

r g(ν) 2 8πn

=

λ4 A

r g(λ) 8πcn2

(4)

where g(ν) is the emission probability per unit frequency, g(λ) is the corresponding value per unit wavelength, and n is the refractive index of the material. F¨orster and Dexter derived expressions for calculating the critical range for nonradiative dipole–dipole energy transfer, using the emission spectrum of the donor and the absorption spectrum of the acceptor [20,21]. Dexter’s version of

(5)

It is important to remember that the critical range is actually independent of the radiative decay rate of the donor, since σDem (λ) is directly proportional to Ar , as indicated in Eq. (4). CDA = Ar R6DA ,

CDD = Ar R6DD

(7)

The parameters of CDX and RDX (X = A or D) are obtained by using the spectral overlap integrals between the donor emission (σDem ) and the acceptor (σXabs ) absorption cross-section according Eqs. (3)–(7). The absorption cross-section and emission crosssection of Nd3+ :4 F3/2 → 4 I9/2 transition in BBS5 glasses are

Table 6 Comparison of CDX and RDX (X = A or D) Host glasses

CDD (cm6 /s)

RDD (A)

CDA (cm6 /s)

RDA (A)

BBS (this work) Lithium lanthanum metaphosphate-like [22] Fluoroindogallat [23]

18.4 × 10−40 40 × 10−40 7.0 × 10−40

9.4 4.2 8.2

3.4 × 10−40 – 2.5 × 10−40

7.1 – 6.9

Q. Nie et al. / Spectrochimica Acta Part A 70 (2008) 537–541

plot in Fig. 4 and the absorption cross-section and emission cross-section of Nd3+ :4 F3/2 , 4 I9/2 → 4 I15/2 , 4 I15/2 , and/or 4 F3/2 , 4I 4 4 9/2 → I13/2 , I15/2 transitions are calculated based on absorption spectrum by using Eq. (4). The obtained values of these parameters are listed in Table 6. The parameters of CDX (X = A or D) are higher than the values of fluoroindogallat glass and lower than the values of lithium lanthanum metaphosphate-like glass. 4. Conclusions In this paper, the Nd3+ :4 F3/2 → 4 I13/2 fluorescence properties have been studied for different Nd3+ concentrations. It is found that the fluorescence intensity drastically increases and has a maximum around 1 mol% Nd2 O3 and decreases with an increase of the Nd2 O3 content. Quantum efficiency and the stimulated emission cross-sections of 4 F3/2 → 4 I13/2 transition are calculated. Based on the electric dipole–dipole interaction theory, the interaction parameter: CDD , for the energy migration rate 4 F3/2 , 4 I9/2 → 4 F3/2 , 4 I9/2 and CDA , for cross-relaxation rate 4F , 4I 4 4 4 4 4 4 3/2 9/2 → I15/2 , I15/2 , and/or F3/2 , I9/2 → I13/2 , I15/2 −40 in bismuth boron silicate glasses are about 18.4 × 10 cm6 /s −40 6 cm /s, respectively. The corresponding critical and 3.4 × 10 range are RDD = 9.4, and RDA = 7.1 A, respectively. Acknowledgements This work was supported by the Science and Technology Department of Zhejiang Province under Grant No. 2006C21082, the Natural Science Foundation of Zhejiang Province under Grant No. 601011.

541

References [1] M. Naftaly, A. Jha, J. Appl. Phys. 87 (2000) 5. [2] E.O. Serqueira, N.O. Dantas, A.F.G. Monte, M.J.V. Bell, J. Non-Cryst. Solids 352 (2006) 3623–3628. [3] L.R.P. Kassab, N.D.R. Junior, S.L. Oliveira, J. Non-Cryst. Solids 352 (2006) 324–3229. [4] W.H. Dumbaugh, J.C. Lapp, J. Am. Ceram. Soc. 75 (1992) 2315–2326. [5] H. Yamauchi, Y. Ohishi, Opt. Mater. 27 (2005) 679. [6] S. Tanabe, N. Sugimoto, S. Ito, T. Hanada, J. Lumin. 87–89 (2000) 670– 672. [7] S.X. Dai, T.F. Xu, Q.H. Nie, et al., Phys. Lett. A 359 (2006) 330–333. [8] S.X. Dai, C.L. Yu, G. Zhou, et al., J. Lumin. 117 (2006) 35–45. [9] R. Judd, Phys. Rev. 127 (1962) 750. [10] G.S. Ofelt, J. Chem. Phys. 37 (1962) 511. [11] H.E.Y. Lin, B. Pun, X.R. Liu, J. Non-Cryst. Solids 283 (2001) 27. [12] R. Rolli, K. Gatterer, M. Wachtler, Spectrochim. Acta Part A 57 (2001) 2009. [13] M.J. Weber, D.C. Ziegler, C.A. Angell, J. Appl. Phys. 53 (1982) 4344. [14] S.S.L. Surana, Y.k. Sharma, S.P. Tandon, Mater. Sci. Eng. B83 (2001) 204–209. [15] K.U. Kumar, V.A. Prathyusha, P. Babu, C.K. Jayasankar, A.S. Joshi, A. Speghini, M. Bettinelli, Spectrochim. Acta Part A 67 (2007) 702–708. [16] D. Chen, Y.S. Wang, Y.L. Yu, E. Ma, F. Liu, J. Phys. Chem. Solids 68 (2007) 193–200. [17] B. Karthikeyan, Reji. Philip, S. Mohan, Opt. Commun. 246 (2005) 153–162. [18] W. Koechner, Solid-State Laser Engineering, vol. 1, Springer-Verlag, New York, 1976, p. 17. [19] A.A. Kaminskii, Laser Crystals, vol. 14, Springer-Verlag, Berlin, 1981, p. 14. [20] T. F¨orster, Ann. Phys. (N.Y.) 2 (1948) 55. [21] D.L. Dexter, J. Chem. Phys. 21 (1953) 836. [22] C. Lurin, C. Parent, G. Leflem, et al., J. Less-Common Metals 126 (1986) 155–160. [23] D.F. de Sousa, F. Batalioto, M.J.V. Bell, et al., J. Appl. Phys. 90 (2001) 7.