Spectroscopic and structural characterization of barium tellurite glass ...

Spectroscopic and structural characterization of barium tellurite glass fibers for mid-infrared ultra-broad tunable fiber lasers W. C. Wang,1 W. J. Zhang,2 L. X. Li,1 Y. Liu,1 D. D. Chen,1 Q. Qian,1 and Q. Y. Zhang1,* 1

State Key Laboratory of Luminescent Materials and Devices, Guangdong Provincial Key Laboratory of Fiber Laser Materials and Applied Techniques, and Institute of Optical Communication Materials, South China University of Technology, Guangzhou 510641, China 2 Shandong Provincial Key Lab of Preparation and Measurement of Building Materials, School of Material Science and Engineering, University of Jinan, Jinan 250022, China * [email protected]

Abstract: To meet the growing demand for mid-infrared tunable fiber lasers, the spectroscopic and structural properties of Tm3+/Ho3+ co-doped barium tellurite glass fibers are systematically evaluated by absorption, Raman, and photoluminescence spectra measurements. The density, molar volume, refractive index, and glass transition temperature are assessed in detail to fully understand their basic physical and thermal properties. Benefitting from the multiple structural sites in a barium tellurite glass system, the maximum doping concentration of Tm2O3 reaches up to 6.0 wt.% without inducing any crystallization or phase separation. Such a high ion concentration is conducive to reducing the fiber length and obtaining an efficient laser output. Furthermore, an intense ~2.0 μm ultra-broad emission with a full width at half maximum (FWHM) of 382 nm is achieved in the Tm3+/Ho3+ co-doped sample upon excitation at 808 nm by properly adjusting Tm3+ concentration and fiber length. The larger emission crosssections and higher gain coefficients along with excellent thermal stability indicate that this barium tellurite glass could be an attractive gain medium for mid-infrared ultra-broad tunable fiber lasers. ©2016 Optical Society of America OCIS codes: (300.6340) Spectroscopy, infrared; (250.5230) Photoluminescence; (160.4670) Optical materials; (160.5690) Rare-earth-doped materials.

References and links 1. 2. 3. 4. 5. 6. 7. 8. 9.

S. D. Jackson, “Towards high-power mid-infrared emission from a fibre laser,” Nat. Photonics 6(7), 423–431 (2012). K. F. Li, G. N. Wang, J. J. Zhang, and L. L. Hu, “Broadband ~2 μm emission in Tm3+/Ho3+ co-doped TeO2– WO3–La2O3 glass,” Solid State Commun. 150(39), 1915–1918 (2010). W. J. Zhang, Q. Y. Zhang, Q. J. Chen, Q. Qian, Z. M. Yang, J. R. Qiu, P. Huang, and Y. S. Wang, “Enhanced 2.0 microm emission and gain coefficient of transparent glass ceramic containing BaF2: Ho3+,Tm3+ nanocrystals,” Opt. Express 17(23), 20952–20958 (2009). K. Li, G. Zhang, X. Wang, L. Hu, P. Kuan, D. Chen, and M. Wang, “Tm3+ and Tm3+-Ho3+ co-doped tungsten tellurite glass single mode fiber laser,” Opt. Express 20(9), 10115–10121 (2012). G. Xue, B. Zhang, K. Yin, W. Yang, and J. Hou, “Ultra-wideband all-fiber tunable Tm/Ho-co-doped laser at 2 μm,” Opt. Express 22(21), 25976–25983 (2014). J. H. Geng, Q. Wang, Y. W. Lee, and S. B. Jiang, “Development of eye-safe fiber lasers near 2 μm,” IEEE Sel. Top. Quantum Electron. 20(5), 0904011 (2014). A. Jha, B. Richards, G. Jose, T. Teddy-Fernandez, P. Joshi, X. Jiang, and J. Lousteau, “Rare-earth ion doped TeO2 and GeO2 glasses as laser materials,” Prog. Mater. Sci. 57(8), 1426–1491 (2012). J. Yuan, Q. Yang, D. D. Chen, Q. Qian, S. X. Shen, Q. Y. Zhang, and Z. H. Jiang, “Compositional effect of WO3, MoO3, and P2O5 on Raman spectroscopy of tellurite glass for broadband and high gain Raman amplifier,” J. Appl. Phys. 111(10), 103511 (2012). Z. X. Jia, L. Liu, C. F. Yao, G. S. Qin, Y. Ohishi, and W. P. Qin, “Supercontinuum generation and lasing in thulium doped tellurite microstructured fibers,” J. Appl. Phys. 115(6), 063106 (2014).

#262277 © 2016 OSA

Received 31 Mar 2016; revised 23 May 2016; accepted 23 May 2016; published 25 May 2016 1 Jun 2016 | Vol. 6, No. 6 | DOI:10.1364/OME.6.002095 | OPTICAL MATERIALS EXPRESS 2095

10. W. C. Wang, J. Yuan, L. X. Li, D. D. Chen, Q. Qian, and Q. Y. Zhang, “Broadband 2.7 μm amplified spontaneous emission of Er3+ doped tellurite fibers for mid-infrared laser applications,” Opt. Mater. Express 5(12), 2964–2977 (2015). 11. A. Ghosh and R. Debnath, “Judd-Ofelt analysis of Er3+ activated lead free fluoro-tellurite glass,” Opt. Mater. 31(4), 604–608 (2009). 12. R. Debnath, A. Ghosh, and S. Balaji, “Synthesis and luminescence properties of an (Er2Te4O11) nanocrystals dispersed highly efficient upconverting lead free tellurite glass,” Chem. Phys. Lett. 474(4), 331–335 (2009). 13. S. Balaji, A. D. Sontakle, R. J. Sen, and A. Kalyandurg, “Efficient ~2.0 μm emission from Ho3+ doped tellurite glass sensitized by Yb3+ ions: Judd-Ofelt analysis and energy transfer mechanism,” Opt. Mater. Express 1(2), 138–150 (2011). 14. S. Balaji, K. Biswas, A. D. Sontakle, G. Gupta, and K. Annapurna, “Enhanced 1.8 μm emission in Yb3+/Tm3+ codoped tellurite glass: effects of Yb3+ ↔ Tm3+ energy transfer and back transfer,” J. Quant. Spectrosc. RA. 147(2), 112–120 (2014). 15. H. A. S. Sidek, R. El-Mallawany, S. S. Badaron, H. M. Kamari, and K. A. Matori, “Optical properties of erbium zinc tellurite glass system,” Adv. Mater. Sci. Eng. 2015, 628954 (2015). 16. H. A. S. Sidek, R. El-Mallawany, K. Hariharan, and S. Rosmawati, “Effect of concurrent ZnO addition and AlF3 reduction on the elastic properties of tellurite based glass system,” Adv. Condens. Matter Phys. 2014, 174362 (2014). 17. S. X. Peng, F. J. Yang, L. B. Wu, Y. W. Qi, S. C. Zheng, D. D. Yin, X. S. Wang, and Y. X. Zhou, “Tm3+/Ho3+/Yb3+ codoped tellurite glass for multicolor emission—structure, thermal stability and spectroscopic properties,” J. Alloys Compd. 609, 14–20 (2014). 18. J. J. Zhang, J. B. Qiu, and Y. Kawamoto, “New oxyfluorotellurite glass: thermal analysis and structural analysis by means of Raman scattering,” Mater. Lett. 55(1), 77–82 (2002). 19. A. Jha, S. Shen, and M. Naftaly, “Structural origin of spectral broadening of 1.5-μm emission in Er3+-doped tellurite glasses,” Phys. Rev. B 62(10), 6215–6227 (2000). 20. S. Tanabe, K. Tamai, K. Hirao, and N. Soga, “Excited-state absorption mechanisms in red-laser-pumped uv and blue upconversions in Tm3+-doped fluoroaluminate glass,” Phys. Rev. B 47(5), 2507–2514 (1993). 21. E. Rukmini and C. K. Jayasankar, “Spectroscopic properties of Ho3+ ions in zinc borosulphate glasses and comparative energy level analyses of Ho3+ ions in various glasses,” Opt. Mater. 4(4), 529–546 (1995). 22. M. Li, G. X. Bai, Y. Y. Guo, L. L. Hu, and J. J. Zhang, “Investigation on Tm3+-doped silicate glass for 1.8 μm emission,” J. Lumin. 132(7), 1830–1835 (2012). 23. S. Rai and A. L. Fanai, “Spectroscopic studies of Ho3+ doped SiO2-TiO2 nanoparticle for photonic applications,” Sci. Vis. 14(3), 112–117 (2014). 24. M. Li, Y. Y. Guo, G. X. Bai, Y. Tian, L. L. Hu, and J. J. Zhang, “~2 μm luminescence and energy transfer characteristics in Tm3+/Ho3+ co-doped silicate glass,” J. Quant. Spectrosc. RA. 127, 70–77 (2013). 25. A. S. S. de Camargo, S. L. de Oliveira, D. F. de Sousa, L. A. O. Nunes, and D. W. Hewak, “Spectroscopic properties and energy transfer parameters of Tm3+ ions in gallium lanthanum sulfide glass,” J. Phys. Condens. Matter 14(41), 9495–9505 (2002). 26. K. Kadono, M. Shojiya, M. Takahashi, H. Higuchi, and Y. Kawamoto, “Radiative and non-radiative relaxation of rare-earth ions in Ga2S3-GeS2-La2S3 glasses,” J. Non-Cryst. Solids 259(1), 39–44 (1999). 27. W. C. Wang, J. Yuan, X. Y. Liu, D. D. Chen, Q. Y. Zhang, and Z. H. Jiang, “An efficient 1.8 μm emission in Tm3+ and Yb3+/Tm3+ doped fluoride modified germanate glasses for a diode-pump mid-infrared laser,” J. NonCryst. Solids 404, 19–25 (2014). 28. C. A. Evans, Z. Ikonić, B. Richards, P. Harrison, and A. Jha, “Numerical rate equation modeling of a 2.1 μm Tm3+/Ho3+ co-doped tellurite fiber laser,” J. Lightwave Technol. 27(19), 4280–4288 (2009). 29. J. Żmojda, D. Dorosz, and J. Dorosz, “2.1 µm emission of Tm3+/Ho3+-doped antimony-silicate glasses for active optical fibre,” Bull. Pol. Acad. Sci.- Technol. Soc. 59(4), 381–387 (2011). 30. R. B. Li, C. Tian, Y. Tian, T. Wei, B. P. Li, X. F. Jing, F. P. Ruan, and F. C. Wang, “Mid-infrared emission properties and energy transfer evaluation in Tm3+ doped fluorophosphate glasses,” J. Lumin. 162, 58–62 (2015). 31. H. F. Chen, F. Z. Chen, T. Wei, Q. H. Liu, R. X. Shen, and Y. Tian, “Ho3+ doped fluorophosphate glasses sensitized by Yb3+ for efficient 2 μm laser applications,” Opt. Commun. 321, 183–188 (2014). 32. R. S. Wang, X. W. Meng, F. X. Yin, Y. Feng, G. S. Qin, and W. P. Qin, “Heavily erbium-doped low-hydroxyl fluorotellurite glasses for 2.7 μm laser applications,” Opt. Mater. Express 3(8), 1127–1136 (2013).

1. Introduction With the rapid development of fiber technology and commercial semiconductor lasers in the past decades, mid-infrared (MIR) broad tunable fiber lasers have aroused intense interest for their potential applications in minimally invasive surgery, remote sensing, and eye-safe laser radar [1]. Ultra-wideband gain spectra in the 2.0 μm wavelength region have a significant impact in many different fields of science and technology. In this respect, Tm3+ and Ho3+ ions are the main rare earth (RE) activators due to the Tm3+: 3F4 → 3H6 and Ho3+: 5I7 → 5I8 transitions, which permits the generation of broadband emission covering 1.7–2.1 μm. Tm3+ is of intense interests for its special advantages: the strong absorption band around 808 nm that could be conveniently excited by commercial available laser diode (LD), broad emission band

#262277 © 2016 OSA


around 2.0 μm (with a FWHM of about 300 nm) suitable for tunable lasers, and high quantum efficiency (near 200%) benefited from the cross relaxation process. Compared with Tm3+, the emission of Ho3+ is characterized by a red-shift of 200 nm and a much larger stimulated emission cross section which makes it suitable for wideband tunable 2.0 μm lasers with high gain characteristic. However, the major limitation of Ho3+-doped fiber laser is the lack of readily available LD pumping source. To solve this problem, an attractive way has been proposed by utilizing Tm3+ as an ideal codopant with Ho3+ to obtain efficient broad tunable 2.0 μm laser output by using the commercial 808 nm LD [2–4]. A broad tuning range of 303 nm from 1727 to 2030 nm has been realized in a Tm3+/Ho3+ co-doped silica fiber lasers recently [5]. However, silica glass has a poor solubility of RE ions, which can easily lead to clustering, phase separation, and concentration quenching. Currently the highest doping level in silica glass is only limited to 2 wt.%, even though extensive efforts have been devoted to increasing the solubility of RE through doping chemical additives (e.g., Al2O3, B2O3, and P2O5) and improving fabrication techniques (e.g., nano-particle technique) [6]. The intrinsic limitation of silica glass network structure makes it difficult to miniaturize devices. Meanwhile, it owns a higher phonon energy (~1100 cm–1) and is only transparent for wavelength shorter than ~2.2 μm. To overcome this wavelength limitation, multicomponent oxide glasses, especially tellurite glasses have been renewed as an important class of materials for photoelectric devices (e.g. fiber lasers, amplifiers, and supercontinuum light sources) in recent years because of their lower phonon energy, higher refractive index (up to 2.25), wide transmission region (up to 5 μm), larger absorption and emission cross sections, good thermal stability and chemical durability, and relatively larger RE ions solubility [7–9]. Considerable attention has recently been paid to the newly emerged barium tellurite glass (TeO2–BaO– La2O3) as enormously potential host for all-solid-state fiber lasers and amplifiers. Compared with the other tellurite glass, the barium tellurite glass exhibits excellent thermal stability and higher glass transition temperature than the classical zinc tellurite glass (TeO2–ZnO–Na2O) [7,10]. Moreover, the phonon energy of barium tellurite glass is much lower than that of the tungsten tellurite glass (TeO2–WO3–La2O3) [8,10], which is extremely beneficial to enhancing the radiative decay rate of Tm3+ and Ho3+, especially corresponding ~2.0 μm MIR emission. For example, Ghosh and Debnath have reported that Er3+-doped barium tellurite glass and glass ceramic with high upconversion and near-infrared (NIR) luminescence efficiency shows great potential for infrared concentrator and NIR sensor [11,12]. Then Balaji et al. have proposed to obtain an efficient ~2.0 μm emission by direct excitation and/or Yb3+ sensitization in the Tm3+- and Ho3+-doped barium tellurite glass [13,14]. More recently, we also successfully achieved a broadband 2.7 μm amplified spontaneous emission in the Er3+doped barium tellurite glass fibers. The results indicate that this glass host is a promising candidate for efficient MIR laser [10]. These special optical properties encourage us to explore whether barium tellurite glass is suitable for new potential applications of fiber lasers. In particular, to further increase its possibility in the application of MIR ultra-broad tunable fiber lasers, a more detailed quantitative study of spectroscopic and structural properties of barium tellurite glass and glass fiber is desirable. Herein, we systematically investigate the thermal, structural, and spectroscopic properties of barium tellurite glasses doped with Tm3+ and Ho3+ by differential scanning calorimetry (DSC), Raman, and photoluminescence spectra measurements, etc. Upon excitation at 808 nm LD, an intense ultra-broad tunable emission at ~2.0 μm is obtained in the Tm3+/Ho3+ co-doped tellurite glass fiber by properly adjusting Tm3+ concentration and fiber length. A quantitative study has been further carried out to evaluate the radiative properties of both Tm3+ and Ho3+ in this barium tellurite glass according to the Jude-Ofelt (J-O) theory and McCumber theory. Subsequently, single cladding barium tellurite glass fibers are fabricated using a convenient suction technique after optimizing the RE doping concentration. The amorphous nature and ultra-broad emission spectra of the glass fibers are also evaluated with the interest in highgain broadband tunable MIR fiber lasers.

#262277 © 2016 OSA


2. Experimental The glasses composed of 80TeO2–10BaF2–5BaO–5La2O3 (mol%) were prepared by melting thoroughly batched raw materials (purity ≥ 99.99%) at 850°C for 1 h. The glasses were doped with x wt.% Tm2O3 and y wt.% Ho2O3 (x = 0, 0.5, 2.0, 4.0, 6.0; y = 0, 0.5). Here we introduced 10 mol% BaF2 to reduce the OH– contents and this is very conducive to reducing the multi-phonon decay rate of Tm3+ and Ho3+ and realizing efficient 2.0 µm luminescence from this barium tellurite glass and fibers. The concentration of RE ions were determined by reviewing some related literatures and our previous experiences [2–6,10]. The melts were then cast into a preheated stainless mold, followed by annealing at 350 °C for 2 h to relinquish the inner stresses. After that, the annealed samples were cut and polished into a shape of 10 × 20 × 1 mm3 for measurements. Glass density was measured by Archimedes’ principle using distilled water as an immersed liquid. The refractive index was obtained by a Metricon Model 2010 prism coupler. The measured values were calculated for many times to determine their error margin. Optical absorption spectra were measured on Perkin-Elmer Lambda 900 UV/VIS/NIR double beam spectrophotometer (Waltham, MA) with the resolution of 1 nm. The fluorescence spectra were taken on a computer controlled TRIAX320 spectrofluorimeter (Jobin-Yvon Corp., France) equipped with an 808 nm LD (Coherent Corp.) as excitation source. The molar composition of the core glass was set to be 80TeO2–10BaF2–5BaO–5La2O3 with an addition of the optimized doping concentration of 6.0Tm2O3/0.5Ho2O3 (wt.%). The composition of cladding glass was 82TeO2–10BaF2–5BaO–3La2O3. Then the glass fibers were fabricated using a convenient suction technique and the detailed preparation process was described in our previous work [10]. The vitreous and/or crystalline nature of the core and cladding glasses were identified by a Philips Model PW1830 X-ray diffractometer (XRD) with Cu-Kα radiation (λ = 1.5406 Å) at 40 kV tube voltage and 40 mA tube current. 3. Results and discussions 3.1 Density, molar volume and refractive index Figure 1 shows the variation of density (ρ) and molar volume (Vm) for a series of barium tellurite glasses with Tm2O3 contents when Ho2O3 content is fixed at 0.5 wt.%. With the increase of Tm2O3 concentration, the density gradually increases from 5.409 ± 0.011 to 5.577 ± 0.011 g·cm–3. This can be related to the heavier thulium and holmium atomic mass as compared to other elements in the glass system. The molecular weight of Tm2O3 (385.8) and Ho2O3 (377.8) are greater than that of TeO2 (159.6), BaO (153.3), BaF2 (175.3) and La2O3 (325.8), whereas the atomic weight of individual constituent atom descends following the order of Tm (168.9) > Ho (164.9) > La (138.9) > Ba (137.3) > Te (127.6) > F (19.0) > O (16.0) and the atomic radius of each of the constituent atoms (in ppm) is Ba (224) > La (187) > Ho (177) > Tm (175) > Te (170) > O (66) > F (64). Moreover, it is found that the density exhibits a close relationship with the content of RE ions, which can be well fitted by the following Eq.:

ρ = A exp ( − x / B ) + C

(1)

where A, B, and C are constants. The fitting degree R2 of the experimental data is greater than 0.99, manifesting that the fitting results are reliable. This also shows that exponential equation is more appropriate than quadratic equation adopted by Sidek et al. in this situation [15]. The change of density is in accord with the molecular weight of compounds, which increases with more atoms added into the glass network. The molar volume of glass samples can be calculated according to the following Eqs [15]: Vm =

#262277 © 2016 OSA

M

ρ

(2)


M =

80M TeO2 + 10M BaF2 + 5M BaO + 5M La2 O3 + x + 0.5 100 + x / M Tm2O3 + 0.5 / M Ho2O3

(3)

where M is the average molecular mass of glass. From the Eq. (2), it can be found that the density and molar volume show an opposite trend. The calculated glass molar volume is slightly reduced from 31.07 ± 0.06 to 30.14 ± 0.06 ( × 10−6 cm−3) in a way close to exponential with the addition of RE ions, as shown in Fig. 1. The decrease in molar volumes for all glass samples is attributed to a decrease in bond length or interatomic spacing [16]. On the other hand, the refractive indexes of the studied glasses with various RE contents range from 2.0411 ± 0.0001 to 2.0521 ± 0.0001. This is also linked closely with the larger molecular weight of the introduced RE ions. The measurement of refractive index is very useful for the calculation of J-O parameters and the design of glass fibers.

Fig. 1. The variation of density and molar volume for barium tellurite glass with Tm2O3 contents when Ho2O3 content is fixed at 0.5 wt.%.

3.2 Thermal stability Thermal analysis is employed to study the effect of RE contents on glass stability, which determines whether the working temperature range of fiber drawing is wide enough. Since the fiber drawing is a reheating process, any crystallization or phase separation will ultimately increase the optical loss and worsen the transmission characteristics of the fiber. Generally, four technological parameters including glass transition temperature (Tg), onset crystallization temperature (Tx), peak crystallization temperature (Tp) and their temperature difference (T = Tx – Tg) are frequently used to evaluate the glass thermal properties. The first three temperature parameters are determined from the tangent intersections of DSC curves. A larger T indicates the glass possesses an excellent thermal ability against the nucleation and crystallization. Larger Tg means a strong resistance to the thermal damage aroused by the transmitted high-power laser, namely, higher laser-induced damage threshold. Therefore, glass hosts with large enough Tg and T is desirable to achieve better heat resistance and wider working range from a practical point of view. Figure 2 displays the DSC curves of glass samples in the temperature range of 225–585 °C. It can be seen that the T of all glasses exceed 100 °C, indicating the present glasses are very stable and thus suitable for fiber drawing. Besides, Tg, Tx and Tp gradually increase with the addition of RE oxides, which is a similar phenomenon found by Peng et al. This can be ascribed to the increase of bond numbers with greater number of cations and average crosslinking density after introducing the RE ions [17]. It is noted that the glass crystallization temperature is approximately between

#262277 © 2016 OSA


518 and 540 °C in the present doping level. Hence it should pass this temperature range as quickly as possible at a carefully controlled heating rate during the fiber drawing process.

Fig. 2. DSC curves of barium tellurite glass with Tm2O3 contents when Ho2O3 content is fixed at 0.5 wt.%.

3.3 Raman spectrum

Fig. 3. Raman spectrum of the undoped barium tellurite glass with fitting data.

Raman spectrum is an effective way to study the structure of glass materials. Figure 3 shows the measured Raman spectra of the undoped barium tellurite glass with fitting data in the spectral range of 200–1000 cm–1. There mainly exist two broad continuous scattering peaks attributed to the disordered structures in the present glass. The spectra can be further decomposed into five symmetrical Gaussian peaks (denoted as A, B, C, D, and E), including three medium peaks around 200, 329, and 462 cm–1, and two strong peaks around 670 and 770 cm–1. These fitted peak positions are derived from the data reported for other similar tellurite glasses [18]. All of these peaks are ascribed to the vibrations of the coordination polyhedral tellurium. The peaks around 462 and 670 cm–1 can be assigned to the asymmetric stretching vibrations of Te–O bond in [TeO4] trigonal bipyramidal structural units. Those around 329 and 770 cm–1 may originate from the bending vibrations of Te–O bond and Te =

#262277 © 2016 OSA


O double bonds in [TeO3] and distorted [TeO3+δ] trigonal pyramidal. Jha et al. have already done a great deal of research work and proven the function of spectral broadening of tellurite glasses doped with Er3+ [7,19], hence the presence of multiple structural sites in the present glass may be in favor of yielding an inhomogeneously broadened spectrum and improving the solubility of RE ions. In addition, it can be found that the maximum phonon energy of barium tellurite glass only extends to 770 cm–1. The lower phonon energy will be very conducive to reducing the non-radiative relaxation probability of RE ions. 3.4 Absorption spectra and Judd-Ofelt analysis Figure 4 shows the absorption spectra of Tm3+, Ho3+, and Tm3+/Ho3+ co-doped barium tellurite glasses over the wavelength region of 300–2200 nm. For the Tm3+-doped sample, the absorption spectrum is characterized by five bands at 465, 662 + 687, 793, 1212, and 1697 nm, which can be assigned to the transitions from the 3H6 ground state to the 1G4, 3F2 + 3F3, 3 H4, 3H5, and 3F4 excited states, respectively. For the Ho3+-doped sample, seven absorption bands centered at 418, 450 + 454, 536 + 545, 644, 889, 1155, and 1952 nm are observed, corresponding to the transitions from the 5I8 ground state to the 5G5, 5G6 + 5F1, 5F2,3 + 3K8, 5F4 + 5S2, 5F5, 5I6, and 5I7 excited states, respectively. No energy levels higher than 1G4 of Tm3+ and 5G5 of Ho3+ are observed because of the intrinsic bandgap absorption in the host glass, which is slightly different from tungsten tellurite glass and silicate glass ceramic [2,3]. The solubility of RE ions can be estimated roughly by the variation of the absorption intensity with doping concentrations. The integral absorption intensity at 793 nm as a function of Tm3+ concentrations are presented in the inset of Fig. 4. The good linearity fitting reveals RE ions are homogeneously distributed in the present barium tellurite glasses, which confirms that such barium tellurite glass possesses excellent RE solubility. The highest doping concentrations are much larger than that in other Tm3+/Ho3+ co-doped tellurite fiber lasers [4]. Higher doping concentration not only facilitates the cross-relaxation (CR) process between Tm3+ ions and energy transfer process from Tm3+ to Ho3+, but also enables higher pump absorption and higher gain per unit length, which is particularly essential to cases that need high laser gain generated from a short piece of active fiber [6].

Fig. 4. Absorption spectra of Tm3+, Ho3+, and Tm3+/Ho3+ co-doped barium tellurite glasses. Inset: integral absorption intensities at 793 nm as a function of Tm3+ concentrations.

Based on the absorption spectra, the J-O intensity parameters Ωt (t = 2, 4, 6), spontaneous transition probabilities A, radiative lifetimes τrad, and fluorescence branching ratios β of the related transitions for both Tm3+ and Ho3+ in this barium tellurite glass are calculated to fully understand their radiative properties. The detailed procedure and matrix elements have been described elsewhere [20,21]. In this process, four absorption bands associated with 3F4, 3H4, 3 F2,3, and 1G4 states of Tm3+ and five absorption bands corresponding to 5I6, 5F5, 5F4 + 5S2, 5F1

#262277 © 2016 OSA


+ 5G6, and 5G5 of Ho3+ are used for the calculation. Table 1 presents the J-O parameters of Tm3+- and Ho3+-doped barium tellurite glasses in comparison to other hosts [22–27]. The root-mean-square deviations δrms in this case are 0.49 and 1.47 ( × 10−6), respectively, indicating the fitting results are reliable. The values of J-O intensity parameters were calculated repeatedly to determine the computation error. It is well-known that larger Ω2 parameter means relatively high covalence of chemical bond between RE and oxygen ions. The parameters Ω4 and Ω6 have a connection with the rigidity and viscosity of the host glass. The calculated Ω2 in the present glass is higher than that of silicate, fluorophosphate, and fluoride glasses but smaller than that of germanate glasses, which means that there exhibit the moderate asymmetry and covalent environment around Tm3+ and Ho3+ ions. The Ω4 and Ω6 of the present glasses are smaller than that of most other hosts, which indicate the weak mechanical properties of tellurite glasses. Moreover, the spontaneous emission probabilities, branching ratios, and radiative lifetimes of the optical transitions for both Tm3+ and Ho3+ in the barium tellurite glass are listed in Tables 2 and 3. It is noted that the predicted spontaneous radiative probabilities for the Tm3+: 3F4 → 3H6 and Ho3+: 5I7 → 5I8 transitions are 531.99 and 166.22 s–1, respectively, which are larger than that of other tellurite glasses [28]. Higher spontaneous radiative probability is beneficial to efficient laser operation. Table 1. Comparison of J-O intensity parameters for Tm3+ and Ho3+ in several glass systems. Tm3+ (10−20 cm2)

Host glass

Ho3+ (10−20 cm2)

Ω2

Ω4

Ω6

Ω2

Ω4

Ω6

Silica [22,23]

6.23

1.91

1.36

1.97

1.47

1.23

Silicate [22,24]

3.08

0.99

0.40

3.14

3.04

0.94

Fluorophosphate [22,24]

3.01

2.56

1.54

2.05

3.77

1.28

5.80 1.96 6.11 4.73 ± 0.05

1.60 1.36 1.41 0.84 ± 0.01

1.30 1.16 1.31 1.15 ± 0.01

0.10 1.86 7.83

4.97 1.90 6.37 2.49 ± 0.02

0.98 1.32 2.05 0.97 ± 0.01

Chalcohalide [25,26] Fluoride [22,24] Germanate [24,27] Tellurite

4.33 ± 0.04

Table 2. Predicted spontaneous emission probabilities, branching ratios and radiative lifetimes of Tm3+ in the present glass. Transition 3 F4 3 H5 3

H4

3

3

1

F3

F2

G4

#262277 © 2016 OSA

Final state 3 H6 3 H6 3 F4 3 H6 3 F4 3 H5 3 H6 3 F4 3 H5 3 H4 3 H6 3 F4 3 H5 3 H4 3 F3 3 H6 3 F4 3 H5 3 H4 3 F3 3 F2

Sed (10−20 cm2) 3.57 1.44 0.60 1.78 0.95 0.48 1.23 0.21 3.26 1.03 0.30 1.45 0.92 1.71 0.08 0.29 0.11 0.93 1.15 0.44 0.13

Aed (s–1) 531.99 469.15 4.29 2527.01 204.42 28.19 3513.84 122.67 769.00 7.64 1325.48 1442.98 386.83 36.21 0.02 1980.62 284.49 1492.08 547.88 96.02 22.00

Amd (s–1) 128.38 0.22 31.08 13.10 90.04 0.54

0.04 13.29 224.94 54.68 5.26

β 1.00 0.99 0.01 0.90 0.08 0.01 0.78 0.05 0.18 0.00 0.42 0.45 0.12 0.01 0.00 0.42 0.06 0.36 0.13 0.02 0.00

τrad (ms) 1.88 1.66 0.36

0.22

0.31

0.21


Table 3. Predicted spontaneous emission probabilities, branching ratios and radiative lifetimes of Ho3+ in the present glass. Transition 5 I7 5 I6 5

I5

5

F5

5

F4

5

F3

5

G5

Final state 5 I8 5 I8 5 I7 5 I8 5 I7 5 I6 5 I8 5 I7 5 I6 5 I5 5 I8 5 I7 5 I6 5 I5 5 F5 5 I8 5 I7 5 I6 5 I5 5 F5 5 F4 5 I8 5 I7 5 I6 5 I5 5 F5 5 F4 5 F3 5 F2 5 K8 5 G6

Sed (10−20 cm2) 1.92 0.80 1.37 0.12 0.93 1.17 1.61 1.33 0.84 0.26 1.29 0.52 0.81 0.79 1.10 0.34 0.84 0.43 0.56 0.45 0.59 1.31 2.73 1.09 0.24 1.72 1.44 1.10 0.38 0.02 1.14

Aed (s–1) 111.52 248.71 27.15 95.24 122.15 12.99 3453.29 862.41 166.58 12.72 5700.89 862.66 552.51 212.75 108.53 2653.78 2793.51 685.35 427.69 52.47 4.29 10694.06 10857.05 2349.54 293.51 624.47 129.37 25.56 5.06 0.24 3.34

Amd (s–1) 54.70 27.62 15.39

48.05

5.04

0.94

β 1.00 0.82 0.18 0.39 0.50 0.12 0.77 0.19 0.04 0.00 0.76 0.12 0.07 0.03 0.02 0.40 0.42 0.10 0.06 0.01 0.00 0.43 0.43 0.68 0.09 0.18 0.04 0.01 0.00 0.00 0.00

τrad (ms) 6.12 3.30 4.07

0.22

0.14

0.15

0.04

3.5 Emission spectra and optical parameters of the barium tellurite glass and fiber The emission spectra of Tm3+/Ho3+ co-doped barium tellurite glasses with varying Tm3+ doping concentrations are shown in Fig. 5. Upon excitation at 808 nm, three continuous emission bands at 1.46, 1.8, and 2.0 μm originating from Tm3+: 3H4 → 3F4, Tm3+: 3F4 → 3H6, and Ho3+: 5I7 → 5I8 transitions are observed clearly. The emission intensities of 1.8 and 2.0 μm are enhanced at first with the incorporation of Tm3+, which is due to the improved cross relaxation and energy transfer process. A flat ultra-wideband emission from about 1580 to 2300 nm with a maximum full width at half maximum (FWHM) of 382 nm is obtained when the content of Tm2O3 reaches up to 6.0 wt.%. This is larger than that of Tm3+/Ho3+ co-doped tungsten tellurite glass (370 nm) and silicate glass (356 nm) [2,29]. The involved energy transfer mechanisms for this Tm3+/Ho3+ co-doped system can be explained by the simplified energy level diagram, as shown in the inset of Fig. 5. The Tm3+: 3H4 state is excited under the excitation of 808 nm LD, followed by a radiative process with emitting 1.4 μm emission and a non-radiative cross relaxation (CR: Tm3+: 3H4 + Tm3+: 3H6 → Tm3+: 3F4 + Tm3+: 3F4) process between two adjacent Tm3+ ions. After that, Ho3+: 5I7 level is populated through energy transfer from Tm3+ to Ho3+ (ET: Tm3+: 3F4 + Ho3+: 5I8 → Tm3+: 3H6 + Ho3+: 5I7). Finally, the radiative transitions of Tm3+: 3F4 → 3H6 and Ho3+: 5I7 → 5I8 occur, yielding 1.8 and 2.0 μm emissions [2–4].

#262277 © 2016 OSA


Fig. 5. Emission spectra of Tm3+/Ho3+ co-doped barium tellurite glasses upon excitation of 808 nm LD with various Tm2O3 concentrations. Inset: simplified energy level diagram for Tm3+Ho3+ co-doped system.

Figure 6 shows the absorption and emission cross-sections (σa and σe) corresponding to the Tm3+: 3H6 ↔ 3F4 and Ho3+: 5I8 ↔ 5I7 transitions in barium tellurite glasses calculated according to McCumber theory. The detailed computing process can be found elsewhere [27]. The emission cross section of Ho3+ in this barium tellurite glass equals to 1.20 × 10−20 cm2, which is nearly double that of Tm3+ (0.63 × 10−20 cm2). It is also higher than that in the silicate, fluoride, germanate, and fluorophosphate glasses [24]. Larger broadband emission cross section in the present glass is conducive to achieving ultra-broad tunable laser output at 2.0 μm. In addition, significant overlaps between Tm3+ emission and Ho3+ absorption crosssections suggests the great possibility of energy transfer from Tm3+ to Ho3+. In contrast, Tm3+ absorption and Ho3+ emission cross-sections have a smaller overlap than the former, which indicate that the backward energy transfer process is relatively inefficient.

Fig. 6. Absorption and emission cross sections corresponding to the Tm3+: 3H6 ↔ 3F4 and Ho3+: 5 I8 ↔ 5I7 transitions in barium tellurite glasses.

Figure 7 shows the calculated gain coefficients of Tm3+: 3F4 → 3H6 and Ho3+: 5I7 → 5I8 transitions versus wavelengths in barium tellurite glasses by assuming a set of population inversion values (p) ranging from 0 to 1 in interval of 0.2. It is noted that the maximum gain

#262277 © 2016 OSA


coefficients at ~2.0 μm are as high as 6.66 and 1.05 cm–1, respectively. They are much larger than that of silicate and fluorophosphate glasses, as shown in Table 4 [22,24,30,31]. The excellent high-gain properties can be attributed to the larger RE doping level and refractive index of barium tellurite glasses. According to the inset of Fig. 4, there is every reason to believe these values of gain coefficients can be further improved because the glass is still unsaturated with such high RE doping. An amazing RE doping content as high as 9.5 mol% has been demonstrated in the barium tellurite glass with a similar composition by Wang et al., this encouraging result may further support our conjecture [32]. On the other hand, it is found that the gain coefficient becomes positive once the population inversion level reaches 40%, implying that a low laser threshold is required for the Ho3+: 5I7 → 5I8 transition in the present glass. For a population inversion level of 0.6, it is possible to obtain smooth tuning laser in rather wide spectral region of 1970–2130 nm for Ho3+. Table 4. Comparison of doping concentrations and gain coefficients for several Tm3+- and Ho3+-doped glasses. Tm2O3 Host glass

Concentration

Silicate [22,24] Fluorophosphate [30,31] Tellurite

0.75 mol 6.0 mol(TmF3) 6.0 wt.%

Ho2O3

Gain coefficient(cm– 1 ) 1.50

0.4 mol

Gain coefficient(cm– 1 ) 0.84

0.97

1.0 mol(HoF3)

0.80

6.66

0.5 wt.%

1.05

Concentration

Fig. 7. Calculated gain coefficients versus wavelengths of (a) Tm3+: 3F4 → 3H6 and (b) Ho3+: 5I7 → 5I8 transitions in barium tellurite glasses.

Based on the above experimental and theoretical analysis, it can be found that the flattest broad emission spectrum at 2.0 μm combined with high-gain and excellent thermal properties can be achieved in the 6.0Tm2O3/0.5Ho2O3 co-doped barium tellurite glass. Therefore, this composition is adopted to further fabricate the glass fiber. The fiber is prepared by the way described in the Experimental part. Continuous tellurite fibers are successfully prepared with the diameter of about 60 and 250 μm for core and cladding, respectively. Meanwhile, the numerical aperture (NA) is 0.214. The fiber loss measured by traditional cut back method is nearly equal to our previous result [10]. Their basic physical parameters are presented in Table 5. Then XRD analysis is carried out to test the vitreous and/or crystalline nature of the core and cladding glasses, as shown in Fig. 8. As can be seen, both core and cladding glasses #262277 © 2016 OSA


exhibit similar broad XRD patterns at 27°, which is the characteristic peak of the tellurite glass. The absence of obvious sharp crystallization peak confirms the amorphous nature of the prepared glass samples. The XRD results are in good accordance with the DSC measurement, indicating that this kind of barium tellurite glass exhibits excellent thermal stability. Table 5. Basic physical parameters of the core and cladding glasses. Glasses Core Cladding

Tg (°C) 389 ± 3 373 ± 3

Tx (°C) 508 ± 3 485 ± 3

T (°C) 119 ± 6 112 ± 6

Refractive index 2.0523 ± 0.0001 2.0411 ± 0.0001

NA 0.214 ± 0.002

Fig. 8. XRD patterns for the core and cladding glasses.

Figure 9 compares the emission spectra of 6.0Tm2O3/0.5Ho2O3 co-doped barium tellurite bulk glass and glass fibers with various lengths upon excitation of 808 nm. The laser spot and output power of pump light as well as the distance between pump source and glass sample remain unchanged to ensure a same test condition. It is noted that the ~2.0 and 1.47 μm emissions increase simultaneously with the increment of the fiber lengths, which indicates that the pump energy is more effectively absorbed by the fiber. At the same time, the enhancement of ~2.0 μm emission is more significant than that of 1.46 μm, indicating that the energy transfer process from Tm3+: 3F4 to Ho3+: 5I7 is more and more efficient. With rising the fiber length from 4 to 12 cm, the intensity of luminescence is gradually reduced, which is due to the enhanced negative effects of fiber loss. Moreover, the FWHM of the emission spectra decreases monotonously from 382 nm to 216 nm because of the comprehensive role of complex energy transfer processes and limited energy divergence of glass fiber. Upon closer inspection, it can be found that the emission peak near 2.0 μm undergoes a red-shift from 1971 to 2022 nm with the increase of fiber lengths. This phenomenon may be caused by the growing radiation trapping along the fiber. Based on the above analysis, it can be concluded that the broader tunable MIR fiber lasers will be possibly realized by using this excellent barium tellurite glass fiber. The ultra-broad tunable all-fiber lasers will be further designed and performed in our subsequent study.

#262277 © 2016 OSA


Fig. 9. Comparison of emission spectra for bulk glass and glass fibers with various lengths upon excitation of 808 nm LD.

4. Conclusion In summary, we systematically studied the spectroscopic and structural properties of Tm3+/Ho3+ co-doped barium tellurite glasses and fibers. Upon excitation at 808 nm, an intense ultra-broad (FWHM = 382 nm) tunable emission at ~2.0 μm is obtained by properly adjusting Tm3+ concentration and fiber length. Absorption, Raman, DSC, and XRD measurements consistently present the evidences of multiple structural sites in this barium tellurite glass system, which allow a maximum doping concentration of 6.0 wt.% Tm2O3 without sacrificing the glass formation ability and thermal stability. Moreover, the investigation on radiative properties of both Tm3+ and Ho3+ shows that the emission cross section of Ho3+ (1.20 × 10−20 cm2) is nearly double that of Tm3+ (0.63 × 10−20 cm2). The gain coefficients of Tm3+ and Ho3+ are also as high as 6.66 and 1.05 cm–1 in the present glass, respectively. Besides, it is found that the glass density, molar volume, refractive index, and glass transition temperature exhibit a close relationship with the content of RE ions. These results manifest that this barium tellurite glass fiber is a very promising candidate for potential applications in the field of MIR ultra-broad tunable fiber lasers. Acknowledgments This work is financially supported by National Natural Science Foundation of China (NSFC) (Grant Nos. 51125005, 51472088 and 51302086), Fundamental Research Funds for the Central Universities, SCUT, and Natural Science Foundation of Shandong Province (No. SZR1408).

#262277 © 2016 OSA
