2 luminescence quantum yield

Optical Materials 28 (2006) 1325–1328 www.elsevier.com/locate/optmat

Erbium-activated modified silica glasses with high 4I13/2 luminescence quantum yield S.N.B. Bhaktha a,h, B. Boulard b, S. Chaussedent c, A. Chiappini a, A. Chiasera d, E. Duval e, C. Duverger b, S. Etienne f, M. Ferrari d,*, Y. Jestin d, M. Mattarelli a, M. Montagna a, A. Monteil c, E. Moser a, H. Portales g, K.C. Vishunubhatla a,h a

Dipartimento di Fisica, Universita` di Trento CSMFO Group, Trento, Italy Laboratoire des Oxydes et Fluorures, UMR CNRS 6010, Universite´ du Maine, Le Mans, France c Laboratoire POMA-UMR CNRS 6136, Universite´ d’Angers, France CNR–IFN, Istituto di Fotonica e Nanotecnologie, CSMFO Group, via Sommarive 14, 38050 Povo-Trento, Italy e LPCML-UMR CNRS 5620, Universite´ Lyon 1, France f Laboratoire de Physique des Mate´riaux-UMR CNRS 7556, Ecole des Mines, Nancy, France g Universite´ Pierre et Marie Curie, LM2N, 4, place Jussieu, F-75252 Paris Cedex 05, France h School of Physics, University of Hyderabad, Gachibowli, Hyderabad 500 046, India b

d

Received 1 August 2005; received in revised form 2 February 2006; accepted 3 February 2006 Available online 17 April 2006

Abstract In this paper we report on optical and spectroscopic properties of an innovative erbium-doped modified silicate Baccarat glass of molar composition: 77.29SiO2:11.86K2O:10.37PbO:0.48Sb2O3. Two different sets of samples were produced, containing 0.2 and 0.5 mol% Er3+ ions, respectively. Optical and spectroscopic assessment of these glasses was performed by absorption, photoluminescence, and refractive index measurements. The quantum efficiency was estimated comparing measured and radiative lifetime of the 4 I13/2 metastable state of the Er3+, the latter obtained by Judd–Ofelt analysis.  2006 Elsevier B.V. All rights reserved. PACS: 42.70.Ce; 78.20.Ci; 78.55.Qr Keywords: Silicate; Glasses; Erbium; Optical properties; Photoluminescence; Radiative lifetime; Quantum efficiency

1. Introduction Telecommunication market is driving the research and development activity in erbium-doped glasses for optical amplifiers. Among the key parameters acting on the performance of active optical devices, the host glass composition plays a crucial role on the amplification efficiency [1–4]. Different materials have been employed as host for erbium

*

Corresponding author. Tel.: +39 461881684; fax: +39 461881696. E-mail address: [email protected] (M. Ferrari).

0925-3467/$ - see front matter  2006 Elsevier B.V. All rights reserved. doi:10.1016/j.optmat.2006.02.018

ions in view to obtain the signal amplification required in optical communication devices, however silica glass still remains one of the more suitable hosts [5]. In addition to a high transparency around 1.5 lm, the long lifetime of the Er3+ 4I13/2 metastable level permits the highpopulation inversion that is necessarily involved in the optical amplifiers claiming high-gain and low signal-tonoise ratio. In this paper, we report on optical and spectroscopic properties of an innovative erbium-doped modified silicate Baccarat glass making it valuable candidate for further applications in optical technologies.

1326

S.N.B. Bhaktha et al. / Optical Materials 28 (2006) 1325–1328 4

2. Experimental The glasses have been produced at Cristallerie Baccarat by a conventional melt-quenching technique with the following molar composition: 77.29SiO2:11.86K2O:10.37PbO: 0.48Sb2O3. Two different sets of samples were produced, containing 0.2 and 0.5 mol% Er3+ ions, respectively labeled B02 and B05. The Er3+ concentrations have been determined in each case taking into account both the nominal composition and the measured density of the glass. The density of the samples has been measured by a gas pycnometer. Refractive index value at several wavelengths has been measured with an accuracy of ±0.001 and a resolution of ±0.0005, using a standard prism coupling method. Refractive index, Er3+ concentration, and density of the Baccarat glasses are reported in Table 1. Optical absorption experiments were performed at room temperature from the ultraviolet to the near infrared spectral range with a double beam spectrophotometer (UV– Vis–NIR Cary 5000 Varian). The Fourier transform infrared (FTIR) spectroscopy measurements were carried out by using a JASCO FTIR-660 plus spectrometer with a resolution set to 4 cm1. The 514.5 nm line of an Ar+-ion laser and the 980 nm line of a Ti:Sapphire laser were used as excitation sources for near infrared photoluminescence (PL) spectroscopy measurements. A Triax 320 mm focal length single grating monochromator was used to disperse the luminescence light onto an InGaAs photodiode and the signal acquisition was performed using a standard lock-in technique. To measure the PL decay of the excited 4 I13/2 level, the cw excitation laser was pulsed with a mechanical chopper and data were acquired with a digital oscilloscope. 3. Results and discussion The UV–Vis–NIR absorption spectrum obtained for the B02 glass is plotted in Fig. 1(A). The spectrum is characteristic of Er3+-doped oxide glasses [6]. The absorption bands are identified with the transitions from the 4I15/2 ground state to the excited states of the Er3+ ions. In particular the band located at around 1540 nm is related to the

Table 1 Refractive index measured at several wavelengths, density and Er3+ concentration of the ‘‘Baccarat’’ silica glasses Sample labelling 3

Density (g/cm ) Er3+-concentration (cm3) Refractive index (±0.0005)@ 457 nm 488 nm 514.5 nm 543.5 nm 632.8 nm 1319 nm 1542 nm

B02

B05

3.049 ± 0.001 8.9 · 1019

3.083 ± 0.006 2.2 · 1020

1.5762 1.5723 1.5688 1.5651 1.5602 1.5448 1.5427

1.5767 1.5732 1.5696 1.5657 1.5610 1.5456 1.5427

I15/2 ! 4I13/2 transition of the trivalent erbium ion. Several Stark structures of the erbium absorption band at 1.5 lm are clearly resolved [6]. The absorption spectrum of the B02 glass sample reported in Fig. 1(B) has been obtained from FTIR measurements in order to extend the previous plot to the infrared region up to the multi-phonon absorption edge which is, for this glass, at around 4500 nm. These samples present a wide transparency region extending from 350 to 2700 nm. The absorption spectra obtained for the B05 sample are similar, except for the band intensities, which depend on the rare earths concentration. The spectrum of Fig. 1(B) shows the multi-phonon tail of the absorption edge, which partially overlaps the wide band (3000–3700 cm1) assigned to the presence of hydroxyl groups in the glass matrix, centred at about 3570 cm1 [7,8]. The amplitude of this fundamental OH stretching band allows the estimation of the OH concentration by the Beer–Lambert law C = (a Æ 0.434)(1/e) where C is the concentration of the bonded species whose vibrations induce the IR light absorption, a is the absorption coefficient and e is the extinction coefficient. A rough estimation of water content is obtained assuming that the value of the extinction coefficient is comparable to what it is found for aluminosilicate glasses (e3520 nm = 40 lglass/molOH cmglass) [7] and andesitic glasses (e3570 nm = 70 ± 0.7 lglass/molOH cmglass) [8,9]. Taking a value of the extinction coefficient e3570 nm  70 lglass/molOH cmglass and a  1 cm1 the resulting COH concentration is low as 6 · 106 mol cm3, which corresponds to an OH content of about 3.6 · 1018 cm3. The PL spectra for the two samples in the region of the 4 I13/2 ! 4I15/2 transition of Er3+ ions, obtained upon excitation at 514.5 nm with an excitation power of 180 mW, are shown in Fig. 2. The PL spectra exhibit a main emission peak at 1537 nm. The spectral width of the two emission bands measured at 3 dB from the maximum of the intensity is 18 ± 1 nm, and the Stark structures at 1490, 1542, 1567 and 1617 nm appear well defined. This result indicates that Er3+ ions occupy sites characterized by similar local environment so that inhomogeneous broadening is not so important in respect to the amount of the Stark splitting. Moreover, is recognized that the homogeneous broadening is dominant in silicate glasses [1]. No change in the spectroscopic features of the 1.5 lm emission was observed exciting at 980 nm. Is important to note that no upconversion signal has been detected from Baccarat Er3+-doped glass upon 980 nm excitation. Characteristic green upconverted luminescence was not observed, even exciting the more concentrated sample B05 at 980 nm, with a pump power of around 400 mW and a beam waist of about 50 lm. An important issue in the optimisation of material and component design is the reduction of the energy transfer upconversion (ETU) mechanism stemming from the clustering of the Er3+ ions. The non-appearance of upconversion in ‘‘Baccarat’’ glasses indicate that most of Er3+ ions are homogeneously distributed and that interaction clusters as well as chemical clusters are practically absent [10].

S.N.B. Bhaktha et al. / Optical Materials 28 (2006) 1325–1328

6

(A)

2

2.0 1.5

G11/2 2

H11/2

1.0 4

I13/2

4

S3/2

0.5

4 4 F9/2 I9/2 I11/2

4

0.0 300

600

900

1200

1500

1800

Absorption coefficient (cm-1)

Absorption coefficient (cm-1)

2.5

1327

(B)

4

2

0 2000

3000

4000

5000

6000

7000

Wavenumber (cm-1)

Wavelength (nm)

Intensity (arbitrary units)

Intensity (arbitrary units)

Fig. 1. (A) Room temperature absorption spectrum in the UV–Vis–NIR spectral region of the B02 sample. Some of the final states of the 4I15/2 ! 2S+1LJ transitions are labelled; (B) FTIR absorption spectrum of the same sample in the IR region.

B05

B02

B05

B02 0

1400

1450

1500

1550

1600

1650

1700

Wavelength (nm)

20

40

60

Time (ms)

Fig. 2. Normalized room temperature photoluminescence spectra of the 4 I13/2 ! 4I15/2 transition of Er3+ ion for the B02 and B05 samples, obtained by exciting at 514.5 nm.

Fig. 3. Room temperature luminescence decay curve, from the 4I13/2 state of Er3+ ion in B02 (j) and B05 (d) samples, obtained after pumping at 514.5 nm with an excitation power of 180 mW. The solid lines represent single exponential decay fit to the experimental data.

Fig. 3 reports the luminescence decay curves from the I13/2 state of Er3+ ion in B02 and B05 samples, obtained upon excitation at 514.5 nm with an excitation power of 180 mW. The same decay profile was measured upon 980 nm excitation. Both the decay curves exhibit a singleexponential behavior. For the B05 sample doped with 0.5 mol% of erbium a lifetime of 11.5 ± 0.1 ms was measured and the sample B02 doped with 0.2 mol% of erbium exhibits a lifetime of 14.2 ± 0.1 ms. Such a measured lifetime is very close to the highest values which have already been reported for erbium in silicate glass hosts: 14.5 ms in Silicate L-22 [1] and 13.5 ms in Soda-lime silicate glasses [6] are typical values found in the literature. The measured lifetime (smes) must be compared with the radiative lifetime, srad, to obtain the radiative quantum efficiency QE defined by the ratio of the measured to the radiative lifetime: QE = smes/srad. The value of srad can be calculated via different theoretical approaches and numerical analysis. The Judd–Ofelt [11,12] theory yields an estimation of the oscillator strength characterizing the intensity of a transition

between two 2S+1LJ multiplets. The Judd–Ofelt parameters Xq are obtained by the chi-square method [13,14] using the absorption spectrum of Fig. 1(A). The radiative lifetime of the Er3+ 4I13/2 metastable state is then obtained from the oscillator strength of the 4I13/2 ! 4I15/2 transition. Both electric and magnetic dipole contributions of the spontaneous emission probability are considered in the calculation of the radiative lifetime. Table 2 reports the obtained Judd–Ofelt parameters together with the root mean square (r.m.s.) deviations of the oscillator strengths [13]. The radiative lifetimes obtained from the Judd–Ofelt theory and the estimated quantum efficiency are also reported. The estimated QE of the Er3+ 4I13/2 metastable level is about 62% for the B05 glass and of about 80% for the B02 glass. Such high quantum efficiency has already been observed in Er3+-doped tellurite glasses [14], but remains among the highest quantum efficiencies estimated in pure or modified-silica host glasses [6]. In a recent paper [15] 4I13/2 radiative lifetime, for the same samples presented in the present work, were calcu-

4

1328

S.N.B. Bhaktha et al. / Optical Materials 28 (2006) 1325–1328

Table 2 Intensity parameters Xq (in units of 1020 cm2) in the 77.29SiO2:11.86K2O:10.37PbO:0.48Sb2O3 ‘‘Baccarat’’ glass activated by 0.2 mol% Er3+ (sample B02) and 0.5 mol% Er3+ (sample B05) Sample

X2

X4

X6

r.m.s.

smes (ms)

srad (ms)

QE (%)

B02 B05

3.36 3.19

0.75 0.56

0.17 0.16

1.38 · 107 6.5 · 108

14.2 ± 0.1 11.5 ± 0.1

17.8 ± 0.1 18.4 ± 0.1

79.8 62.5

Calculated (srad) and measured (smes) lifetimes with corresponding radiative quantum efficiency QE, of the Er3+ 4I13/2 metastable level are also reported.

lated applying a simple method based on the Einstein’s relation for the emission probability of a two-level system [16] and the approximate McCumber procedure [4,17,18]. The various methods considered here and in the paper of Ref. [15] for calculating srad give values in good agreement among them in despite of the different procedures employed from one method to the other. The internal gain coefficient g at wavelength k can be estimated by means of the formula g(k) = rem(k) Æ N2  rabs(k)N1 where N1 and N2 are the densities of ions in the ground state and the excited state, respectively. (N1 + N2 = N, N being the density of erbium ions) [4]. The emission and absorption cross sections rem and rabs were determined according to McCumber procedure [4,17,18]. In the case of total inversion (N2 = N), at 1537 nm we obtain an internal gain coefficient of about 2.0 dB. 4. Conclusions The glasses, with molar composition: 77.29SiO2:11.86K2O:10.37PbO:0.48Sb2O3, have been produced at Cristallerie Baccarat by a conventional melt-quenching technique. Two different sets of samples were produced, containing 0.2 and 0.5 mol% of Er3+ ions. The refractive index was measure at several wavelengths. Absorption measurements indicate the high transparency of the ‘‘Baccarat’’ glass. The water content of these glasses is very low and we estimate a COH concentration low as 3.6 · 1018 cm3. Luminescence at 1.5 lm with a spectral width of 18 nm observed on both the glasses. No changes in the spectral width as a function of the excitation wavelength were observed and no NIR-to-visible upconversion signal has been detected. The non-appearance of upconversion in ‘‘Baccarat’’ glasses indicates that most of Er3+ ions are homogeneously distributed and that interaction clusters as well as chemical clusters are practically absent. The 4I13/2 metastable state of the Er3+ ions decay curves present a single exponential profiles, with a lifetime value of 11.5 ms for the sample doped with 0.5 mol% of erbium and a lifetime of 14.2 ms for the sample doped with 0.2 mol% of erbium. For the 0.2 mol% Er3+-activated glass, quantum efficiency reaches very high values, up to 79.8%. Such high quantum efficiency has already been observed in Er3+-doped tellurite glasses, but remains

among the highest quantum efficiencies estimated in pure or modified-silica host glasses. The optical and spectroscopic properties of this modified-silica glass make it well adapted as host medium for the light propagation and a good candidate for many applications in telecommunication systems. Acknowledgements Authors acknowledge the financial support of MIURFIRB RBNE012N3X, PAT 2004-2006 FAPVU, MIURPRIN2004 (2004–2005), CNR-CNRS (2004–2005), and ITPAR (2003–2006), projects. We gratefully acknowledge Cristallerie Baccarat for supplying samples. References [1] [2] [3] [4] [5]

[6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

[16] [17] [18]

W.J. Miniscalco, J. Lightwave Technol. 9 (1991) 234. A.J. Kenyon, Prog. Quant. Electron. 26 (2002) 225. A. Polman, J. Appl. Phys. 82 (1997) 1. E. Desurvire, Erbium-doped Fiber Amplifiers, Principles and Applications, John Wiley, New York, 1994. G.C. Righini, S. Pelli, M. Ferrari, C. Armellini, L. Zampedri, C. Tosello, S. Ronchin, R. Rolli, E. Moser, M. Montagna, A. Chiasera, S.J.L. Ribeiro, Opt. Quant. Electron. 34 (2002) 1151. M.P. Hehlen, N.J. Cockroft, T.R. Gosnell, Phys. Rev. B 56 (1997) 9302. S.N. Houde-Walter, P.M. Peters, J.F. Stebbins, Q. Zeng, J. NonCryst. Solids 286 (2001) 118. C.W. Mandeville, J.D. Webster, M.J. Rutherford, B.E. Taylor, A. Timbal, K. Faure, Am. Mineral. 87 (2002) 813. P.L. King, T.W. Vennemann, J.R. Holloway, R.L. Hervig, J.B. Lowenstern, J.F. Forneries, Am. Mineral. 87 (2002) 1077. F. Auzel, P. Goldner, Opt. Mater. 16 (2001) 93. B.R. Judd, Phys. Rev. 127 (1962) 750. G.S. Ofelt, J. Chem. Phys. 37 (1962) 511. K.A. Gschneidner, L. Eyring, Handbook on the Physics and Chemistry of Rare Earth, vol. 25, Elsevier, 1998. R. Rolli, M. Montagna, S. Chaussedent, A. Monteil, V.K. Tikhomirov, M. Ferrari, Opt. Mater. 21 (2003) 743. V. Benoit, S.N.B. Bhaktha, B. Boulard, S. Chaussedent, A. Chiappini, A. Chiasera, E. Duval, S. Etienne, M. Ferrari, B. GaillardAllemand, Y. Jestin, M. Mattarelli, M. Montagna, A. Monteil, E. Moser, G. Nunzi Conti, S. Pelli, H. Portales, D.N. Rao, G.C. Righini, K.C. Vishunubhatla, Proc. SPIE 5723 (2005) 79. B.J. Chen, G.C. Righini, M. Bettinelli, A. Speghini, J. Non-Cryst. Solids 322 (2003) 319. D.E. McCumber, Phys. Rev. 134 (1964) A299. W.J. Miniscalco, R.S. Quimby, Opt. Lett. 16 (1991) 258.