Transparent nano-glass-ceramics for efficient infrared emission

Journal of Alloys and Compounds 451 (2008) 542–544

Transparent nano-glass-ceramics for efficient infrared emission V.K. Tikhomirov, C. G¨orller-Walrand, K. Driesen ∗ Chemistry Department, Katholieke Universiteit Leuven, B-3000 Leuven, Belgium Available online 19 April 2007

Abstract We summarise the experimental data and provide a general theoretical basis for efficient infrared emission in the rare-earth doped transparent oxyfluoride nano-glass-ceramics. In these glass-ceramics, more than 90% fraction of the rare-earth dopant, such as Ho3+ , Dy3+ , Eu3+ , Tm3+ , Er3+ , dissolve in the cubic ␤-PbF2 nano-crystals with a certain diameter of the order of 10 nm, whilst these nano-crystals are embedded in a robust aluminosilicate glass network. A remarkably low maximal phonon energy coupled to the rare-earth dopant in the ␤-PbF2 (at about 250 cm−1 ) permits the efficient infrared emission of the dopants from the levels, which are non-radiatively quenched in other glassy hosts. Further advantage of this nano-glass-ceramic material is in its robustness typical of the aluminosilicate glass, which offers fabrication of durable waveguide devices. The site of the rare-earth dopant is proposed to be nearly cubic with 8 fluorine ligands around the dopant resulting in prolonged lifetimes of the lasing levels and admixture of the vibronic coupling to the strength of some electric dipole transitions. © 2007 Elsevier B.V. All rights reserved. PACS: 81.05.Pj (glass-based composites; vitroceramics); 61.46.−w (nanoscale materials structure); 78.67.Bf (optical properties of nano-crystals and nano-particles) Keywords: Glass-ceramics; Infrared emission; Rare-earth; Lanthanide; Luminescence; Holmium; Dysprosium

1. Introduction Currently there is a demand for lasers and optical amplifiers operating in the near-infrared from 1.2 to 1.7 ␮m at the extended telecommunications window and in the mid-infrared from 1.8 to 5.0 ␮m for, e.g. chemical sensors, surgery, air navigation. One of the best candidates for lasing/optical amplification is the rareearth ion doped in the respective host because many rare-earth ions can emit in the near- and mid-infrared. However, the gaps between the involved and intermediate energy levels of the rareearth ions are rather small resulting usually in a substantial nonradiative quenching of the infrared emission. To overcome this problem, it is necessary to use a host for the rare-earth dopant, which has a low vibration frequency. Here we report on oxyfluoride nano-glass-ceramics (GC) 32(SiO2 )9(AlO1.5 )31.5(CdF2 )18.5(PbF2 )5.5(ZnF2 ):3.5(ReF3 ) host where heavy doping with rare-earth ions (Re) can be achieved, and where vibration energy affecting the dopant (less than 250 cm−1 ) is smaller than in other competitive glassy hosts. This host is robust taking the advantage of an oxide silica-based glass with respect to chemical and physical

∗

Corresponding author. Tel.: +32 16 327383; fax: +32 16 327992. E-mail address: [email protected] (K. Driesen).

0925-8388/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2007.04.112

durability and the possibility of waveguides shaping, such as buried channel waveguides. We report on the intense infrared emission of several rare-earth dopants at different wavelengths and provide a general theoretical basis for efficient infrared emission in this rare-earth doped transparent GC. The site of the rare-earth dopant is proposed to be highly symmetric, i.e. nearly octahedral cubic, resulting in prolonged lifetimes of the lasing levels and admixture of the vibronic coupling to the strength of some electric dipole transitions, in particular of Ho3+ . 2. Experimental The procedures for preparation of the precursor glass 32(SiO2 )9(AlO1.5 ) 31.5(CdF2 )18.5(PbF2 )5.5(ZnF2 ):3.5(ReF3 ) and daughter GC have been described elsewhere [1–4]. It was found that the rare-earth doped nanocrystalline phase can be produced avoiding a heat-treatment step [1–3] by irradiating the precursor glass with an infrared laser; this in particular allows a micro-fabrication of buried GC waveguides within the bulk of precursor glass [4]. The rare-earth dopant nucleates the growth of the nano-crystalline phase consisting of spherical-shaped fcc cubic ␤-PbF2 nano-crystals heavy doped with the respective dopants [1–5]. The size and shape of the nano-crystals have been evaluated by means of the transmission electron microscopy (TEM), X-ray diffraction (XRD) and Raman scattering techniques, e.g. [1]. The bulk rare-earth doped GCs have been found to be transparent in the spectral range from about 350 nm to 5 ␮m owing to very small size of the nano-crystals precluding attenuation of light due to

V.K. Tikhomirov et al. / Journal of Alloys and Compounds 451 (2008) 542–544

543

Fig. 1. Normalized emission spectra of the Ho3+ -doped GC at room (thick solid line) and 77 K (thin solid line) temperatures excited at 640 nm.

Fig. 3. Room (thick line) and 77 K (thin line) excitation spectra 5 F4 , 5 S2 ← 5 I8 of the Ho3+ -doped GC.

the Rayleigh scattering and close match of the refractive index of the nanocrystals and the environment glass network at about 1.75. The typical size of the bulk samples of the precursor glass and daughter GC could be as high as, e.g. 50 mm × 10 mm × 5 mm. The optical set-ups have been described elsewhere [1–5].

are substantially longer than reported to date for these dopants, e.g. [7–9], pointing out an advantage for applications, due to the reasons discussed further. The site of the rare-earth dopant in this GC is evaluated to be nearly cubic for more than 90% fraction of the rare-earth dopant dissolved in the cubic ␤-PbF2 nano-crystals [2,3]. The high symmetry of the dopant site allows a contribution of vibronic component to the strength of the forced electric dipole transitions [10], as it is seen in Fig. 3 for the 5 F4 , 5 S2 ← 5 I8 excitation band of the Ho3+ -doped GC. The side components that almost disappear when cooling down are proposed to be due to the coupling of the Stark components of this weak electronic transition to the phonons characteristic of the ␤-PbF2 , as discussed elsewhere [2].

3. Results The data presented in Fig. 1 (Ho3+ -dopant) and Fig. 2 (Dy3+ dopant) point out an advantage of this GC host for doping by rare-earth ions for obtaining infrared luminescence, which is either completely or substantially quenched in other materials. In addition, the doping level in this GC at 3.5 mol% is remarkably high suggesting application in short length laser/optical devices, such as planar or buried waveguides. Two emission bands of the Ho3+ at 1.2 ␮m (Fig. 1, 5 I6 → 5 I8 transition) and 1.47 ␮m (Fig. 1, 5 F5 → 5 I6 transition), and of Dy3+ at 1.35 ␮m (Fig. 2, 6 H9/2 , 6 F11/2 → 6 H15/2 transition) fall in the modern extended telecommunication window ranging from 1.2 to 1.7 ␮m [6]. The lifetime of the Ho3+ band at 1.47 ␮m equals to 24 ␮s at the room temperature and 60 ␮s at 77 K. The lifetime of the Ho3+ band at 1.2 ␮m equals to 1.60 ms both at the room temperature and at 77 K. The lifetime for the Dy3+ band at 1.35 ␮m equals to 150 ␮s at room temperature. These lifetimes

4. Discussion Fig. 4 shows energy level diagram of several rare-earth ions, such as Dy3+ , Ho3+ , Er3+ , Tm3+ , which potentially may emit at the multiple infrared wavelengths. It is seen that the gap between the spin-orbit split levels lies in the range of 1000–2000 cm−1 . Therefore, most of rare-earth doped hosts will not emit light at these infrared wavelengths from 1 to 5 ␮m since the luminescence will be non-radiatively quenched by emission of phonons in high energy hosts with a typical energy of 500–1500 cm−1 [11]. n (T ) for non-radiative/phonon assisted The probability Wnr decay of the lasing level at with simultaneous emission of n phonons at temperature T is involved in Eq. (1): 1 1 n = Wnr (T ) + τexp τrad

(1)

where τ exp is the experimental lifetime and τ rad is the radiative lifetime of the lasing level. It is seen from Eq. (1) that large n (T ) will preclude a radiative emission. On the other hand, Wnr n (T ) can also be expressed by Eq. (2) [12]: Wnr Fig. 2. Room temperature emission spectrum of the Dy3+ -doped GC excited at 1.06 ␮m. A structured deep between 1.34 and 1.40 ␮m is due to absorption of the OH radicals in the photodetector.

n (T ) Wnr

=

n Wnr (0) ×

   −¯hω −n 1 − exp kT

(2)

544

V.K. Tikhomirov et al. / Journal of Alloys and Compounds 451 (2008) 542–544

longed experimental lifetime τ exp according to Eq. (1), which is a prerequisite for obtaining a required inverse population for the lasing/optical amplification effect. 5. Conclusion An efficient infrared luminescence has been reported in a robust nano-glass-ceramics host doped with a few rare-earth dopants, such as Dy3+ , Ho3+ , Er3+ [1], Tm3+ [5], at a doping level of 3.5 mol%. It is shown that the low maximal phonon energy of 250 cm−1 and the high dopant site symmetry in these nano-glass-ceramics do ensure these efficient infrared emissions. References

Fig. 4. Energy level diagram of the Dy3+ , Ho3+ , Er3+ , Tm3+ , where arrows indicate some involved emission transitions. The energy gap between 6 H9/2 , 6F 6 3+ −1 11/2 and H11/2 of the Dy is indicated as a scale and it equals to 1800 cm . −1 The maximal phonon energy of the GC host equals to 250 cm .

where n is the number of phonons h ¯ ω, which bridges the gap between the lasing level and the nearest lower lying level. Thus, n (T ) increases drastically with h Wnr ¯ ω (or decreases drastically with n) and therefore, for efficient radiative decay/emission of the lasing level, according to Eqs. (1) and (2), h ¯ ω should be as small as possible, i.e. substantially smaller than 500 cm−1 . Actually not many materials respond to this requirement, especially bearing in mind that the host should be robust, durable and offer possibility of waveguide shaping [2,3]. The GC proposed in this paper offers a maximal phonon energy coupled to the rare-earth dopant as small as 250 cm−1 [13–15], suits the requirements of robustness, durability and waveguide shaping [4], and therefore may be a best candidate for devices using the infrared luminescence. The high dopant site symmetry in this GC, argued to be nearly cubic octahedral with eight fluorine ligands around the dopant [1–4], ensures a prolonged radiative lifetime τ rad , i.e. a pro-

[1] V.K. Tikhomirov, D. Furniss, I.M. Reaney, M. Beggiora, M. Ferrari, M. Montagna, R. Rolli, Appl. Phys. Lett. 81 (2002) 1937. [2] V.K. Tikhomirov, K. Driesen, C. G¨orller-Walrand, Phys. Status Sol. (a), available online from 14 December 2006. [3] K. Driesen, V.K. Tikhomirov, C. G¨orller-Walrand, V.D. Rodr´ıguez, A.B. Seddon, Appl. Phys. Lett. 88 (2006) (art.073111). [4] V.K. Tikhomirov, J. Koch, D. Wand, B. Chichkov, Phys. Status Sol. (a) 202 (2005) R73. [5] M. Mattarelli, V.K. Tikhomirov, M. Montagna, E. Moser, S. Chaussedent, S. Pelli, G.C. Righini, M. Ferrari, J. Non-Cryst. Solids 345–346 (2004) 354. [6] G.C. Thomas, B.I. Shraiman, P.F. Glodis, M.J. Stephen, Nature (London) 404 (2000) 262. [7] D.P. Machewirth, K. Wei, V. Krasteva, R. Datta, E. Snitzer, G.H. Sigel, J. Non-Cryst. Solids 213–214 (1997) 295. [8] Y.B. Shin, J. Heo, J. Non-Cryst. Solids 253 (1999) 23. [9] L. Wetenkamp, G.F. West, H. T¨obben, J. Non-Cryst. Solids 140 (1992) 35. [10] C. G¨orller-Walrand, K. Binnemans, in: K.A. Gschneidner Jr., L. Eyring (Eds.), Handbook on the Physics and Chemistry of Rare Earths, vol. 25, Elsevier, Amsterdam, The Netherlands, 1998, p. 101. [11] R. Reisfield, Ann. Chim. Fran. 7 (1982) 147. [12] B. Henderson, J.F. Imbusch, Optical Spectroscopy of Inorganic Solids, Oxford University Press, Oxford, 1989 (chapter 5). [13] V.K. Tikhomirov, M. Ferrari, M. Montagna, L.F. Santos, R.M. Almeida, J. Non-Cryst. Solids 337 (2004) 191. [14] M. Takahashi, M. Izuki, R. Kanno, Y. Kawamoto, J. Appl. Phys. 83 (1998) 3920. [15] V.K. Tikhomirov, M. Ferrari, M. Montagna, L.F. Santos, R.M. Almeida, Europhys. Lett. 64 (2003) 529.