Journal of Sol-Gel Science and Technology 26, 943–946, 2003 c 2003 Kluwer Academic Publishers. Manufactured in The Netherlands.
Er3+ /Yb3+ Co-Activated Silica-Alumina Monolithic Xerogels A. CHIASERA, M. MONTAGNA, R. ROLLI AND S. RONCHIN∗ Dipartimento di Fisica and INFM, Universit`a di Trento, via Sommarive 14, I-38050 Povo-Trento, Italy
[email protected]
S. PELLI AND G.C. RIGHINI Optoelectronics and Photonics Department, IROE-CNR, via Panciatichi 64, I-50127 Firenze, Italy R.R. GONC¸ALVES Dipartimento di Ingegneria dei Materiali and INFM, Universit`a di Trento, via Mesiano 77, I-38100 Trento, Italy Y. MESSADDEQ AND S.J.L. RIBEIRO Instituto de Qu´ımica-UNESP, P.O. Box 355, 14801-970 Araraquara, SP, Brazil C. ARMELLINI, M. FERRARI AND L. ZAMPEDRI CNR-CeFSA, Centro Fisica Stati Aggregati, via Sommarive 14, I-38050 Povo-Trento, Italy
Abstract. Monolithic silica xerogels doped with different concentrations of Er3+ , Yb3+ and Al3+ were prepared by sol-gel route. Densification was achieved by thermal treatment in air at 950◦ C for 120 h with a heating rate of 0.1◦ C/min. We studied the luminescence properties of the 4 I13/2 → 4 I15/2 emission band of Er3+ as a function of the Al/Er/Yb concentration and we paid particular attention to the alumina effects. Raman spectroscopy and Vis-NIR absorption were used to monitor the degree of densification of the glasses and the residual OH content. Keywords: silica-alumina glasses, sol-gel, erbium-ytterbium, spectroscopic properties 1.
Introduction
Erbium-doped silica-based glasses prepared by sol-gel route are attractive materials for integrated optics devices such us optical amplifiers operating in C telecommunications band [1–4]. The main problems associated with sol-gel-based devices are the non-radiative relaxation channels due to both rare-earth concentration quenching and vibrations of the OH groups [1, 2]. In previous works we have studied the spectroscopic properties of Er2 O3 -SiO2 xerogels as a function of erbium concentration and the influence of the OH groups vibrations on the fluorescence decay at 1.5 µm [5–7]. ∗ To
whom all correspondence should be addressed.
It is well known that the aluminium co-doping can alleviate the problem of Er clustering [1, 2]. Furthermore, ytterbium co-doping can significantly enhance system absorption at 980 nm, thus making more efficient the pumping mechanism [8]. In this work we present a spectroscopic study of Er3+ /Yb3+ co-activated silica-alumina monolithic xerogels.
2.
Experimental
Silica gels were prepared using as starting solution a mixture of tetramethylortosilicate (TMOS), methanol, deionized water and nitric acid in the molar ratios 0.06:0.35:0.55:0.04, respectively [5, 9].
944
Chiasera et al.
Table 1. Erbium, ytterbium and aluminium composition of the silica xerogels. Sample labelling
Er/Si (ppm)
Yb/Si (ppm)
Al/Si (ppm)
EYA1
2000
4000
6000
EYA2
5000
10000
15000
EYA3
10000
20000
30000
EY1
10000
20000
0
Erbium, ytterbium, aluminium were introduced as Er(NO3 )3 ·5H2 O, Yb(NO3 )3 ·5H2 O and Al(NO3 )3 · 9H2 O, respectively. Samples with different Er/Si concentration were prepared; the Yb and Al were also added in different concentrations: Table 1 reports the Er/Yb/Al content for the samples investigated in this work. Densification was achieved by thermal treatment in air at 950◦ C for 120 hours with a heating rate of 0.1◦ C/min. The final samples were monolithic square pieces of about 10 × 10 × 3 mm and of good optical quality. Photoluminescence spectroscopy, in the region of the 4 I13/2 → 4 I15/2 transition of Er3+ ions, was performed by exciting at 514.5 nm (Ar laser) and 980 nm (diode laser). The luminescence was dispersed by a 320 mm single-grating monochromator with a resolution of 2 nm. The light was detected using an InGaAs photodiode and a lock-in technique. Decay curves were obtained recording the signal by a digital oscilloscope. All measurements were performed at room temperature. More details about the experimental set-up have been reported in previous papers [5, 7]. 3.
Results and Discussion
Figure 1 shows the absorption spectrum in the visibleNIR range obtained for the samples EYA3 and EY1. The material presents a wide transparency region extending from 2000 up to 280 nm. The spectrum is characteristic of Er3+ -doped oxide glasses [8, 10]. The absorption bands are identified with the transitions from the 4 I15/2 ground state to the excited states of the Er3+ ion. The sharp band at 10250 cm−1 is assigned to the transition from the 2 F7/2 ground state to the 2 F5/2 level of the Yb3+ ion; it is overlapped to erbium 4 I15/2 → 4 I11/2 absorption transition, whose cross section is more than one order of magnitude smaller than that of the ytterbium transition [8]. The spectra obtained for the other samples are similar, except for the band intensities, which depend on the rare earths concentration. The two reported spectra show
Figure 1. Room temperature absorption spectrum in the Vis-NIR spectral region of the 10000 Er/Si ppm, 20000 Yb/Si ppm SiO2 xerogels with 30000 (EYA3) and 0 (EY1) Al/Si ppm annealed at 950◦ C for 120 h. The final states of the 2S+1 LJ ← 4 I15/2 transitions are labelled.
practically the same band intensities, but for the sample EYA3 doped with 30000 ppm of aluminium there is a general increase in background absorption towards higher energies. In the NIR spectral region the two bands at 4525 and 7230 cm−1 are due to OH vibrations. The band located at 7230 cm−1 is the overtone of the SiO H stretching vibration of hydrogen bonded silanol groups [11]. Figure 2 shows the VV polarised Raman spectra. Two spectral regions are shown: (I) anti-Stokes and Stokes from low frequency region up to 1500 cm−1 and (II) the region between 3200 cm−1 and 3800 cm−1 . Figure 2(I) shows the so called Boson peak that is
Figure 2. Room temperature Raman spectra of all investigated silica xerogels doped with different Er/Yb/Al content annealed at 950◦ C for 120 h. (I) excitation at 458 nm; (II) excitation at 514.5 nm.
Er3+ /Yb3+ Co-Activated Silica-Alumina Monolithic Xerogels
characteristic of the glassy state and is related to the densification degree of the samples [5–7]. Furthermore, the bands at 430, 800, 1100 and 1190 cm−1 are assigned to the silica network and the bands at 490 and 610 cm−1 are due, respectively, to the D1 and D2 defects. The band at 960 cm−1 is due to the Si OH vibration. The Raman spectra in Fig. 2(II) show one band centred at about 3670 cm−1 which is assigned to the SiO H stretching vibration of hydrogen bonded silanol groups. These observations indicate that full densification is achieved for all the investigated samples. In the Raman spectrum of the EYA3 xerogel, the most doped sample, a weak shoulder appears around 335 cm−1 ; a similar feature has already been observed by Arai et al. in a Nd2 O3 -Al2 O3 -SiO2 sol-gel glass [12]. In any case, the aluminium addition does not increase the vibrational cut-off energy of the silica matrix. Figure 3(I) shows the room temperature photoluminescence spectra corresponding to the 4 I13/2 → 4 I15/2 transition of the Er3+ ions obtained upon CW excitation at 514.5 nm for the samples EYA1, EYA2, and EYA3. The spectra exhibit a main emission peak at 1.53 µm with a shoulder at about 1.55 µm. The spectral width of the emission band, due to inhomogeneous and homogeneous broadening plus additional Stark splitting of the excited and ground states, ranges from 48 to 55 nm: the largest one is observed for the sample EYA3. Some changes are also observable in the shape of the emission band: the most Al-doped glass (EYA3) exhibits the flattest band.
Figure 3. Room temperature photoluminescence spectra (I) of the 4I 4 4 13/2 → I15/2 transition and decay curves (II) from the I13/2 level of Er3+ ion after excitation at 514.5 nm for the samples EYA1 (a), EYA2 (b), EYA3 (c). Spectrum (dashed line) and decay profile (open squares) of EYA1 sample obtained after excitation at 980 nm are also reported.
945
Figure 3(II) shows the room temperature decay curves of the 4 I13/2 level after excitation at 514.5 nm for the same three samples. The decay curve of the sample EYA1 is a single-exponential with a lifetime of 8.7 ms. Furthermore, this measured value of lifetime is relatively high in comparison to those reported for Er-doped aluminosilicate materials [1, 3, 13, 14]. The decay curves for EYA2 and EYA3 samples are not single exponential and the values τ1/e (defined by the time for which the luminescence is 1/e of the intensity at t = 0), are about 7.5 and 5.6 ms, respectively. This decrease of τ1/e with increasing erbium concentration is due to energy transfer between erbium ions [1]. Figure 3 shows, as an example, also the fluorescence spectrum and decay curve at 1.53 µm after excitation at 980 nm for the EYA1 sample: no significant variations were observed in the shape of emission band and in the lifetime for the two excitation energies. Further experiments are in progress to estimate the effect of Yb3+ doping on the fluorescence intensity at about 1.53 µm and to clarify the Yb-Er energy transfer processes. Figure 4(a) and (b) compare the two samples EY1 and EYA3 containing 0 and 30000 Al/Si ppm, respectively. We can observe a slight increase in the flatness and in the broadening of the band for the Al-doped glass (bandwidth of 55 nm for the Al-doped sample versus 51 nm for the one without Al). These values are similar to those measured in Al co-doped silica glasses [14, 15] and larger than those in silicate glasses [6, 15]. Moreover, the peak position shifts from 1532.5 nm, for EY1
(a)
(b)
Figure 4. Room temperature photoluminescence spectra of the 4I 4 4 13/2 → I15/2 transition (a) and decay curves from the I13/2 level 3+ (b) of Er ion after excitation at 514.5 nm for the 10000 Er/Si ppm, 20000 Yb/Si ppm SiO2 xerogels with 30000 (EYA3) and 0 (EY1) Al/Si ppm. The vertical line is included as a guide for the eye.
946
Chiasera et al.
sample, to 1531 nm, for EYA3. Similar shifts have been reported in literature for melting and sol-gel glasses codoped with alumina [2]. In Fig. 4(b) the decay curves are shown: a strong difference is observed. Both profiles are not single exponential, but the sample EY1 without Al exhibits a faster initial decay: τ1/e is about 0.1 ms for the sample EY1 and 5.6 ms for the sample EYA3. This result confirms the role of aluminium in the reduction of the rare-earth clustering. A possible mechanism for the solubility of the rare-earth ions in aluminium codoped silica glasses has been proposed by Arai et al. [12] and by Zhou et al. [16] on the basis of a structural model of the SiO2 -Al2 O3 glass. When the aluminium ions are added, they could be incorporated in two local bonding configurations in the silica network: a tetrahedral bonding configuration, such as AlO4/2 groups, as a network former, and an octahedral coordination of oxygens atom, such as AlO6/2 groups, as a network modifier [12, 16]. These groups could act as solvation shells in the glass network for the rare-earth ions [1, 12, 16]. In the case of the AlO4/2 groups, due to charge compensation, the Er3+ ions are preferentially accommodated near to the aluminium sites. In the network modifier case, the aluminium ions break the silica structure producing non bridging Al-O groups which can coordinate the Er3+ ions. 4.
Conclusion
NIR absorption and Raman spectra show that complete densification of the Er2 O3 -Yb2 O3 -Al2 O3 -SiO2 xerogels is achieved for all investigated samples. The emission of Er3+ at 1.5 µm is observed for all samples. For the xerogel activated with 2000 Er/Si ppm a lifetime of 8.7 ms is measured for the metastable 4 I13/2 level of Er3+ . The 30000 ppm Al-doped sample exhibits a broader emission band and a much slower decay at 1.5 µ m than the analogous sample without alumina. These results indicate that the alumina co-doping provides a more uniform distribution of Er3+ ions in the
host glass. Further experiments are in progress to identify the advantage of a specific Er/Yb/Al glass composition for the spectroscopic properties at 1.53 µm. References 1. X. Orignac, D. Barbier, X.M. Du, R.M. Almeida, O. McCarthy, and E. Yeatman, Opt. Mater. 12, 1 (1999). 2. B.T. Stone and K.L. Bray, J. Non-Cryst. Solids 197, 136 (1996). 3. E.M. Yeatman, M.M. Ahmad, O. McCarthy, A. Vannucci, P. Gastaldo, D. Barbier, D. Mongardien, and C. Moronvalle, Opt. Commun. 164, 19 (1999). 4. G.C. Righini, S. Pelli, M. Brenci, M. Ferrari, C. Duverger, M. Montagna, and R. Dall’Igna, J. Non-Cryst. Solids 284, 223 (2001). 5. M. Ferrari, C. Armellini, S. Ronchin, R. Rolli, C. Duverger, A. Monteil, N. Balu, and P. Innocenzi, J. Sol-Gel Sci. Tech. 19, 569 (2000). 6. C. Duverger, M. Montagna, R. Rolli, S. Ronchin, L. Zampedri, M. Fossi, S. Pelli, G.C. Righini, A. Monteil, C. Armellini, and M. Ferrari, J. Non-Cryst. Solids 280, 261 (2001). 7. L. Zampedri, C. Tosello, F. Rossi, S. Ronchin, R. Rolli, M. Montagna, A. Chiasera, G.C. Righini, S. Pelli, A. Monteil, S. Chaussedent, C. Bernard, C. Duverger, M. Ferrari, and C. Armellini, SPIE 4282, 200 (2001). 8. M.P. Hehlen, N.J. Cockroft, T.R. Gosnell, and A.J. Bruce, Phys. Rev. B 56, 9302 (1997). 9. G. Pucker, S. Parolin, E. Moser, M. Montagna, M. Ferrari, and L. Del Longo, Spectrochim. Acta A 54, 2133 (1998). 10. J.A. Sampaio, T. Catunda, A.A. Coelho, S. Gama, A.C. Bento, L.C.M. Miranda, and M.L. Baesso, J. Non-Cryst. Solids 273, 239 (2000). 11. C.J. Brinker and G.W. Scherer, Sol-Gel Science: The Physics and Chemistry of Sol-Gel Processing (Academic Press, San Diego, 1990), p. 627. 12. K. Arai, H. Namikawa, K. Kumata, T. Honda, Y. Ishii, and T. Handa, J. App. Phys. 59, 3430 (1986). 13. G.C. Righini, M. Bettinelli, M. Brenci, C. Duverger, M. Ferrari, M. Fossi, M. Montagna, S. Pelli, and A. Speghini, SPIE 3749, 755 (1999). 14. R.M. Almeida, X.M. Du, D. Barbier, and X. Orignac, J. Sol-Gel Sci. Tech. 14, 209 (1999). 15. M. Dejneka and B. Samson, MRS Bull. 24, 39 (1999). 16. Y. Zhou, Y.L. Lam, S.S. Wang, H.L. Liu, C.H. Kam, and Y.C. Chan, Appl. Phys. Lett. 71, 507 (1997).
