Tailoring chromatic dispersion in chalcogenide–tellurite microstructured optical fiber

Optical Fiber Technology 20 (2014) 409–413

Contents lists available at ScienceDirect

Optical Fiber Technology www.elsevier.com/locate/yofte

Tailoring chromatic dispersion in chalcogenide–tellurite microstructured optical fiber Tomas Kohoutek a, Zhongchao Duan a, Hiroyasu Kawashima a, Tonglei Cheng a, Takenobu Suzuki a, Morio Matsumoto b, Takashi Misumi b, Yasutake Ohishi a,⇑ a b

Research Center for Advanced Photon Technology, Toyota Technological Institute, 2-12-1 Hisakata, Tempaku, Nagoya 468-8511, Japan Furukawa Denshi Co., Ltd., 2-3-2 Marunouchi, Chiyoda-Ku, Tokyo 100-8370, Japan

a r t i c l e

i n f o

Article history: Received 18 December 2013 Revised 3 April 2014 Available online 13 June 2014 Keywords: Glasses and optical materials Optical properties of glasses Optical fibers

a b s t r a c t We report fabrication of a highly nonlinear hybrid microstructured optical fiber composed of chalcogenide glass core and tellurite glass cladding. The flattened chromatic dispersion can be achieved in such an optical fiber with near zero dispersion wavelength at telecommunication wavelengths k = 1.35–1.7 lm, which cannot be achieved in chalcogenide glass optical fibers due to their high refractive index, i.e. n > 2.1. We demonstrate a hybrid 4-air hole chalcogenide–tellurite optical fiber (Dn = 0.25) with flattened chromatic dispersion around k = 1.55 lm. In optimized 12-air hole optical fiber composed of the same glasses, the chromatic dispersion values were achieved between 20 and 32 ps/nm/km in a broad wavelength range of 1.5–3.8 lm providing the fiber with extremely high nonlinear coefficient 86,000 km1W1. Hybrid chalcogenide/tellurite fibers pumped with the near infrared lasers give good promise for broadband optical amplification, wavelength conversion, and supercontinuum generation in the near- to mid-infrared region. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Since the first demonstration of microstructured optical fiber (MOF) by Russell et al. [1], MOFs have attracted much attention due to their high nonlinearity, low loss and strong light confinement, and tunable chromatic dispersion (group velocity). These parameters are necessary for efficient nonlinear optical processes such as self-phase modulation, cross-phase modulation, four wave mixing, soliton self-frequency shift or super-continuum generation (SC). Except for high nonlinearity, MOFs brought enhanced options for tuning of the chromatic dispersions of optical fibers, whose zero or near zero values [2,3] are required for effective generation of nonlinear processes. Chromatic dispersion tuning was first developed on silica glass fibers [4,5]. However, silica glass fibers have provided low nonlinear coefficients, i.e. less than 70 km1W1 [6], so far. Therefore, recent efforts have been also devoted to the study of promising nonlinear non-silica glasses such as heavy oxide doped silica glasses [7], tellurite [8,9] or fluoride [10] glasses with transparency window extended up to 6 or 8 lm, and their microstructured fibers, respectively. Furthermore, chalcogenide ⇑ Corresponding author. E-mail address: [email protected] (Y. Ohishi). http://dx.doi.org/10.1016/j.yofte.2014.05.004 1068-5200/Ó 2014 Elsevier Inc. All rights reserved.

glasses (ChG) have broadened infrared transparency up to 12, 18, and 23 lm for sulfides, selenides, and tellurides, respectively. Also the third order nonlinearity coefficients (v3) of ChG are approximately 100–300 times larger than that of silica glass [11], which together with high refractive index (n > 2.1) elevates nonlinearity of chalcogenide MOFs to highest reported values. ChG optical fiber fabrication brings some difficulties such as costly purification, drawing under inert atmosphere, lower mechanical and thermal stability of the fibers, higher sensitivity to humidity and certain health issues. Despite these difficulties, ChG are worth to study since they make possible the extension of the application wavelengths of the optical fibers to the middle infrared region, where they can be used for environmental sensing, medical diagnosis, and cure applications [11–14]. In fabrication of chalcogenide MOFs, two main techniques have been used successfully. The capillary-stacking ‘‘stack and draw’’ technique is commonly used for fabrication of silica MOFs. We have used this method for fabrication of our hybrid MOF. The main advantage of the technique is its ability to fabricate structured preform with a complex geometry. On the other hands complex geometries require multiple preparation steps, whose increase exposure of the glass to the atmosphere, and also require very precise pressure control during the processing in order to preserve the structure of the preform for the microstructure of the optical

410

T. Kohoutek et al. / Optical Fiber Technology 20 (2014) 409–413

fiber. The second commonly used technique is the extrusion technique, which has simpler processing with the advantage of reduced number of fabrication steps. However, the extrusion technique is still quite challenging for fabrication of optical fibers with complex microstructures. In this paper, we demonstrate the ability of a hybrid chalcogenide/tellurite MOF to achieve flattened and near zero chromatic dispersion (important for broad SC generation) at the wavelengths of currently available pico- and femto-second pulse lasers operating at telecommunication wavelengths, typically at k = 1.55 lm. We show the influence of various cladding materials on the chromatic dispersion of a hybrid chalcogenide/tellurite MOF with 4-air hole structure. We also demonstrate flattened chromatic dispersions in a complex chalcogenide/tellurite MOF with near zero chromatic dispersion values, i.e. from 20 to 32 ps/nm/km, over a wide range of infrared wavelengths k = 1.5–3.8 lm. Such a MOF has extremely high nonlinear coefficient c  86,000 km1W1.

Fig. 1. Chromatic dispersion calculated for a hybrid 4-air holes GGSS–TZLB MOF. The comparison shows the dispersion curves calculated for changing core diameter d = 0.8–1.3 lm.

2. Calculations 2.1. Chromatic dispersion calculations Chromatic dispersion of hybrid chalcogenide–tellurite MOF was calculated based on its geometrical microstructure taking into account the refractive index dispersion of both glasses given by the Sellmeier Eq. (1):

n2 ¼ 1 þ

Xi¼3 h i¼1

Ai k2 =ðk2  L2i Þ

i

ð1Þ

where the fitting coefficients, Ai and Li2, for GGSS and TZLB glasses are summarized in Table 1 and k is the wavelength of the light. By using a Finite Element Method (FEM), we calculated chromatic dispersion of the fiber (D) as (2):

D¼

  1 d db k2 2pc dk dk

ð2Þ

where c is the speed of the light in vacuum and b is the propagation constant of the fiber.The nonlinear coefficient was then calculated according to Eq. (3) [2],

c¼

2p k

R R1

n ðx; yÞjFðx; yÞj4 dxdy 1 2 R R 2 1 jFðx; yÞj2 dxdy 1

ð3Þ

where F(x,y) is the profile of the electric field, and n2(x,y) is the nonlinear refractive index, which we took from Refs. [15,16], i.e. n2 = 1.8  1017, 5.9  1019, and 2.3  1023 m2/W for GGSS glass, TZLB glass, and air hole, respectively. 2.2. Chromatic dispersion in one air hole ring chalcogenide–tellurite MOF – compositional aspects We calculated the chromatic dispersion of our hybrid chalcogenide–tellurite MOF, where the refractive index contrast between the core and the cladding glass was Dn = 0.25 at 1.55 lm. The refractive index dispersions and fiber structures resulted in fiber

Table 1 Sellmeier coefficients determined for GGSS and TZLB glasses based on the refractive index measurements by using a prism coupler method. Materials

i=1 i=2 i=3

GGSS

TZLB

Ai

Li2

Ai

Li2

2.51696 1.41612 0.0601

0.0218087 0.1034606 153.82497

1.67189 1.34862 0.6218

0.0004665 0.0574608 46.725427

chromatic dispersion values depicted in Fig. 1. Dispersion curves shown in Fig. 1 were calculated from the microstructure of the fabricated fiber shown in Fig. 2 but the diameter of chalcogenide glass core varies from 0.8 to 1.3 lm. The chromatic dispersion of our prepared fiber corresponds to the curve with d = 0.9 lm. The chromatic dispersion values at 1.55 lm increased from 360 to 190 ps/nm/km with increasing core diameter. The values are still far from near zero dispersion values suitable for generating efficient nonlinear effects. On the other hand, the chromatic dispersion of the hybrid fiber was much more flattened than that achievable by any chalcogenide MOF at telecommunication wavelengths [17]. This is given by the fact that the effective refractive index of the hybrid MOF is reduced due to low refractive index tellurite glass cladding. Indeed, there is a way how to achieve flattened and near zero chromatic dispersion values at wavelengths around k = 1.55 lm even for structured fibers with chalcogenide glass cores. The key issue is the tuning of the refractive index contrast between chalcogenide glass core and tellurite glass cladding. Fig. 3 shows the chromatic dispersion curves calculated for MOFs with the similar microstructure such as that shown in Fig. 2, but with changing the cladding glass and thus changing the refractive index contrast between the core and the cladding materials. Fig. 3a shows chromatic dispersion curve for a hybrid chalcogenide–tellurite MOF with the index contrast Dn = 0.45. The core remained the same chalcogenide GGSS glass but the cladding glass whose material dispersion we considered was TeO2-ZnO-Li2ONa2O-P2O5 (TZLNP). For this material choice, we should note that it is only a theoretical material choice, and the chromatic dispersion values of D = 20 to +20 ps/nm/km were obtained for a MOF with the core diameter of d = 1 lm in the wavelength range of k = 1.1–1.6 lm. Similar chromatic dispersion values were obtained for a MOF with the core diameter of d = 1.3 lm in the wavelength range of k = 1.6–2.5 lm, which gives promise for the efficient pumping of the fiber at k > 2 lm, i.e. suitable for mid-IR SC generation in chalcogenide core MOFs [18,19]. Similarly, by replacing tellurite TZLNP glass for lower refractive index phosphate P2O5-ZnO-Na2O-K2O (PZNK) glass cladding, again this is a theoretical material choice, we obtained chromatic dispersion curves shifted to normal dispersion values (see Fig. 3b) caused by high refractive index contrast between the core and the cladding glass corresponding to Dn = 0.7. Combination of high refractive index GGSS core and low refractive index phosphate glass cladding did not provide flattened dispersion curves. Certain advantage can be seen in the fact that normal dispersion values were achieved in relatively broad interval of wavelengths, i.e.

T. Kohoutek et al. / Optical Fiber Technology 20 (2014) 409–413

411

Fig. 2. Optical micrographs recorded from cleaved preform (a), hybrid chalcogenide–tellurite MOF (b) and the core part of a hybrid MOF (c), respectively.

Fig. 3. Chromatic dispersion calculated for a hybrid 4-air holes GGSS–TZLNP and GGSS–PZNK MOFs. The comparison shows the dispersion curves calculated for changing core diameter d = 0.8–1.3 lm.

k = 1–2.5 lm. The TZLNP and PZNK glasses [20] were considered only in terms of theoretical dispersion calculations. These glasses can be adopted as cladding glasses as their softening temperatures are comparable with the chalcogenide glass one. However, the thermal expansion coefficient difference is larger, which makes them less suitable for practical fabrication of composite fibers than in case of GGSS and TZLB choice. Sellmeier coefficients for the calculations were taken from Ref. [20]. 3. MOF fabrication The hybrid chalcogenide–tellurite MOF was fabricated by a capillary stacking method. The composition of the tellurite glass was 78 TeO2-5ZnO-12Li2O-5Bi2O3 mol% (TZLB). The raw materials were of analytic grade. The composition of chalcogenide glass Ge15Ga3

Sb12S70 (GGSS) was prepared in a form of a rod with outer diameter of 12 mm by a direct synthesis from elements of 5N purity (Furukawa Denshi Co. Ltd.) in an evacuated silica ampoule. The refractive index dispersions were obtained by fitting the refractive index values measured by using Prism coupler method with Sellmeier equation, see spectral dependences of the refractive index of these two glasses in Ref. [17]. Tellurite glass rod with the shape of tetragon with outside diameter of 12 mm and air hole in the center of 3 mm was prepared by the rotational casting method. The TZLB glass melt was cast in a shaped metallic mold and consequently annealed at the glass transition temperature. Three other tellurite glass tubes were prepared by the rotational casting method with inner diameter of 4 mm and outer diameter of 12 mm. The chalcogenide rod was firstly elongated to match the inner diameter (i.e. 4 mm) of the tellurite tube, then inserted in it and elongated together in order to obtain chalcogenide–tellurite rod with outer diameter of 3 mm. This rod was inserted in the center hole of a tetragonally-shaped tellurite glass rod and elongated together into a cane with the outer diameter of 4 mm. The cane was inserted into a tellurite glass ‘‘jacket’’ tube and elongated into a preform with outer diameter near 4 mm. Then, the preform was inserted into another tellurite glass jacket tube and drawn into the fiber. During the fiber drawing, positive pressure of nitrogen gas was applied into the air holes of the cane in order to preserve the structure of the preform for the microstructured optical fiber. The entire fabrication process is depicted in Fig. 4. Fabrication of a hybrid MOF demands a good choice of the glass materials with similar thermal properties, i.e. softening temperature (Ts) and thermal expansion coefficient, in order to avoid cracking at their interface or disruptions of the core along the fiber length. Thermo-mechanical analysis of selected glasses showed matching their Ts at T = 307 °C [17]. Glass transition temperatures (Tg) of these glasses were found to be differed about DT  19 °C by to the same measurements. The fabricated chalcogenide–tellurite MOF had a 4-air hole microstructure with the hole diameters of approximately 7 lm. The center part of the MOF included square 5  5 lm area of TZLB glass with GGSS core of the diameter of 0.9 lm. The outer fiber diameter was about 120 lm. The preform and fiber cross-sections are depicted in Fig. 2.

4. Measurement of the supercontinuum generation in an optical fiber The length of hybrid chalcogenide–tellurite MOF we used for the measurement of the SC spectra was 5 cm, i.e. one half of our previous experiment [17]. Both fiber ends were cleaved by using a diamond stylus. The fiber loss was estimated by using a cut-back method several times. The attenuation loss of the fiber was about 4–6 dB/cm. Higher loss could influence the width of the spectral

412

T. Kohoutek et al. / Optical Fiber Technology 20 (2014) 409–413

Fig. 4. Schematic shows fabrication of a hybrid chalcogenide–tellurite MOF.

broadening we obtained. To observe SC generation, we launched the 1557 nm pump pulse with the pulse width of 400 fs and the repetition rate of 16.75 MHz from a custom-made fiber laser and amplifier system into the fiber input end by using an aspheric lens with 0.47 NA and a precision positioning system. The coupling efficiency was estimated about 10%. The output end of the fiber was connected by using a standard multimode silica fiber to an

optical spectrum analyzer (OSA, Yokogawa AQ6375) with a measurement range of 1200–2400 nm. Fig. 5 shows the SC spectra recorded from our prepared 4-air hole GGSS–TZLB MOF pumped with a custom made 400 fs pulse laser operating at k = 1.55 lm, i.e. corresponding to D = 280 ps/ nm/km. The power dependence spectra showed broadening of the pump wavelength with increasing laser power. At lower pump powers, the self-phase modulation dominates the broadening effect, while at higher pump powers the spectral broadening was mainly due to Raman soliton dynamics. 5. Chromatic dispersion in two air hole ring chalcogenide– tellurite MOF – geometrical aspects

Fig. 5. Supercontinuum generation from a hybrid GGSS–TZLB MOF recorded after pumping the fiber with a custom made 400 fs fiber laser at k = 1557 nm. Thermal damage of used 5 cm long fiber was found at pulse powers higher than 600 mW.

For GGSS–TZLB glass choice with the refractive index contrast Dn = 0.25 a chance for obtaining a hybrid MOF with flattened and near zero dispersion values still exists. The solution is in a use of optimized MOF with more complex microstructure, which brings enhanced options for further dispersion tuning. In the text below, we demonstrate possibilities of an optimized fiber microstructure for a hybrid GGSS–TZLB MOF. We considered two 6-air hole rings in tellurite glass cladding and chalcogenide glass core as depicted in Fig. 6. To calculate the chromatic dispersion of a complex hybrid chalcogenide–tellurite MOF, we defined the diameters as H1 (inner) and H2 (outer) and the pitches as K1 (inner) and K2 (outer). We calculated the chromatic dispersion in this MOF for various values of the inner air hole pitch, K1. The other structural parameters were fixed to the core diameter of d = 1.0 lm, H1 = 0.36 lm, H2 = 2.2 lm, and K2 = 2.04 lm. Fig. 7 shows the calculated chromatic dispersion curves. For K1 = 0.71 lm, the flattened dispersion was observed with four

Fig. 6. Schematic shows the microstructure of the two 6-air-hole rings chalcogenide–tellurite MOF with defined diameters (H1 and H2, and pitches of the air holes in the fiber geometry.

T. Kohoutek et al. / Optical Fiber Technology 20 (2014) 409–413

413

Our results revealed influence of the refractive index contrast between the chalcogenide core and the tellurite cladding glasses, i.e. materials choice, for the design of a simple suspended core MOF in order to achieve flattened and near zero chromatic dispersion at the telecommunication wavelengths. We also showed enhanced options for the chromatic dispersion tailoring in a hybrid MOF with more complex microstructure providing flattened dispersion over wide spectral region, i.e. 1.5–3.8 lm with extremely high nonlinear coefficient 86,000 km1W1. We believe that hybrid MOFs give promise for enhanced generation of nonlinear processes applicable for broadband optical amplification, wavelength conversion, and supercontinuum generation at wavelengths from near- to mid-infrared region. Acknowledgments Fig. 7. Chromatic dispersion curves calculated for a hybrid GGSS–TZLB MOF depicted in Fig. 6. The parameter K1 parameter changed K1 = 0.68, 0.71 and 0.72. Other parameters were kept constant at d = 1.0 lm, H1 = 0.36 lm, H2 = 2.2 lm, and K2 = 2.04 lm.

This work was supported by Japanese Ministry of Culture, Sports and Education (MEXT) via the Support Program for Forming Strategic Research Infrastructure (2011–2015). References

Table 2 Summarizes nonlinear coefficients determined at k = 1.55 lm for various microstructured optical fibers, see Refs. [25–29]. MOFs

Nonlinearity coefficient c (km1W1) at k = 1.55 lm

Bismuth(1) Silica SF57 Tellurite Bismuth(2) Bismuth(3) Our MOF

64 70 640 675 735 1360 86,000

ZDWs at k = 1.52, 2.11, 2.86, and 3.72 lm and with value between D = 20 and 32 ps/nm/km from k = 1.5 to 3.8 lm. In the inset of Fig. 7, we can see that the electric mode field intensities in the MOF calculated for three different wavelengths, i.e. (a) k = 1.5 lm, (b) 2.5 lm, (c) 3.5 lm. It is apparent from the result that at longer wavelengths, the more light intensity was dispersed from the chalcogenide core to the tellurite cladding. This behavior seems to be effective for tailoring the chromatic dispersion and it helps to satisfy the phase-matching condition in the fiber. Optimized microstructure with flattened chromatic dispersion [21–23] can significantly improve the stability of SC generation in a hybrid chalcogenide–tellurite MOF. Now we are trying to fabricate 2 air hole ring hybrid chalcogenide–tellurite MOF but their processing is technologically challenging and will require more efforts and time. Another remarkable feature of complex hybrid GGSS–TZLB MOF is its extremely high nonlinearity. The nonlinear coefficient c at k = 1.55 lm was calculated to be as high as 86,000 km1W1, which is about 80,000 times larger than that of standard SMF28 fiber [24], and one or two orders of magnitude larger than those c of other highly nonlinear MOFs reported so far except for tapered fibers, see Table 2. 6. Conclusions We demonstrated options for the chromatic dispersion tuning in a hybrid chalcogenide–tellurite microstructured optical fiber.

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]

[17] [18] [19]

[20] [21] [22] [23] [24]

[25] [26] [27] [28] [29]

P. Russell, Science 299 (2003) 358. G.P. Agrawal, Nonlinear Fiber Optics, 4th ed., Academic Press, Boston, 2007. J.M. Dudley, J.R. Taylor, Nat. Photon 3 (2009) 85. J.C. Knight, Nature 424 (2003) 847. F. Zhang, J. Wu, S. Fu, K. Xu, Y. Li, X. Hong, P. Shum, J. Lin, Opt. Express 18 (2010) 15870. F. Poli, F. Adami, M. Foroni, L. Rosa, A. Cucinotta, S. Selleri, Appl. Phys. B 81 (2005) 251. J.F. Van Derlofske, Proc. SPIE 4445 (2001) 119. P. Domachuk, N.A. Wolchover, M. Cronin-Golomb, A. Wang, A.K. George, C.M.B. Cordeiro, J.C. Knight, F.G. Omenetto, Opt. Express 16 (2008) 7161. M. Liao, X. Yan, G. Qin, C. Chaudhari, T. Suzuki, Y. Ohishi, Opt. Express 17 (2009) 15481. G. Qin, X. Yan, C. Kito, M. Liao, C. Chaudhari, T. Suzuki, Y. Ohishi, Appl. Phys. Lett. 95 (2009) 161103. A. Zakery, S.R. Elliott, J. Non-Cryst, Solids 330 (2000) 1. F. Charpentier, B. Bureau, J. Troles, C. Boussard-Plédel, K. Michel-Le Pierrès, F. Smektala, J.L. Adam, Opt. Mater. 31 (2009) 496. J.S. Sanghera, L.B. Shaw, I.D. Aggarwal, IEEE J. Sel. Top. Quantum Electron. 15 (2009) 114. D. Lezal, J. Pedlikova, J. Zavadil, Chalcogenide Lett. 1 (2004) 11. Z. Duan, M. Liao, X. Yan, C. Kito, T. Suzuki, Y. Ohishi, Appl. Phys. Express 4 (2011) 072502. W. Gao, M. Liao, X. Yan, C. Kito, T. Kohoutek, T. Suzuki, M. El-Amraoui, J.-C. Jules, G. Gadret, F. Desevedavy, F. Smektala, Y. Ohishi, Appl. Phys. Express 4 (2011) 102601. T. Kohoutek, Z. Duan, X. Yan, T. Suzuki, M. Matsumoto, Takashi Misumi, Y. Ohishi, Proc. SPIE 8257 (2012) 82570D. H. Kawashima, T. Kohoutek, X. Yan, T. Suzuki, Y. Ohishi, Phys. Stat. Sol. C 9 (12) (2012) 2621. I. Savelli, O. Mouawad, J. Fatome, B. Kibler, F. Désévédavy, G. Gadret, J.-C. Jules, P.-Y. Bony, H. Kawashima, W. Gao, T. Kohoutek, T. Suzuki, Y. Ohishi, F. Smektala, Opt. Express 20 (24) (2012) 27083. H.T. Tong, Ch. Kito, T. Suzuki, Y. Ohishi, Opt. Mater. 34 (2012) 1795. G. Genty, S. Coen, J.M. Dudley, J. Opt. Soc. Am. B 24 (2007) 1771. K.M. Hilligsøe, T. Andersen, H. Paulsen, C. Nielsen, K. Mølmer, S. Keiding, R. Kristiansen, K. Hansen, J. Larsen, Opt. Express 12 (2004) 1045. V. Finazzi, T.M. Monro, D.J. Richardson, J. Opt. Soc. Am. B 20 (2003) 1427. J.C. Hernandez Garciaa, O. Pottieza, R. Grajales-Coutin´o, B. Ibarra-Escamilla, E.A. Kuzin, J.M. Estudillo-Ayala, J. Gutierrez-Gutierrez, Laser Phys. 21 (8) (2011) 1. M. Liao, C. Chaudhari, G. Qin, X. Yan, C. Kito, T. Suzuki, Y. Ohishi, M. Matsumoto, T. Misumi, Opt. Express 17 (2009) 21608. K. Kikuchi, K. Taira, N. Sugimoto, Electron. Lett. 38 (2002) 166. J.H. Lee, W. Belardi, K. Furusawa, P. Petropoulos, Z. Yusoff, T.M. Monro, D.J. Richardson, IEEE Photon. Technol. Lett. 15 (2003) 440. P. Petropoulos, H. Ebendorff-Heidepriem, V. Finazzi, R. Moore, K. Frampton, D.J. Richardson, T. Monro, Opt. Express 11 (2003) 3568. H. Ebendorff-Heidepriem, P. Petropoulos, S. Asimakis, V. Finazzi, R. Moore, K. Frampton, F. Koizumi, D. Richardson, T. Monro, Opt. Express 12 (2004) 5082.